From f4b42421d80cac4291dd82bd97baa1bc06b351fc Mon Sep 17 00:00:00 2001 From: Mark Tinguely Date: Tue, 4 Dec 2012 17:18:02 -0600 Subject: xfs: use b_maps[] for discontiguous buffers Commits starting at 77c1a08 introduced a multiple segment support to xfs_buf. xfs_trans_buf_item_match() could not find a multi-segment buffer in the transaction because it was looking at the single segment block number rather than the multi-segment b_maps[0].bm.bn. This results on a recursive buffer lock that can never be satisfied. This patch: 1) Changed the remaining b_map accesses to be b_maps[0] accesses. 2) Renames the single segment b_map structure to __b_map to avoid future confusion. Signed-off-by: Mark Tinguely Reviewed-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Ben Myers --- fs/xfs/xfs_buf.c | 12 ++++++------ fs/xfs/xfs_buf.h | 6 +++--- 2 files changed, 9 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 26673a0b20e7..56d1614760cf 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -175,7 +175,7 @@ xfs_buf_get_maps( bp->b_map_count = map_count; if (map_count == 1) { - bp->b_maps = &bp->b_map; + bp->b_maps = &bp->__b_map; return 0; } @@ -193,7 +193,7 @@ static void xfs_buf_free_maps( struct xfs_buf *bp) { - if (bp->b_maps != &bp->b_map) { + if (bp->b_maps != &bp->__b_map) { kmem_free(bp->b_maps); bp->b_maps = NULL; } @@ -377,8 +377,8 @@ xfs_buf_allocate_memory( } use_alloc_page: - start = BBTOB(bp->b_map.bm_bn) >> PAGE_SHIFT; - end = (BBTOB(bp->b_map.bm_bn + bp->b_length) + PAGE_SIZE - 1) + start = BBTOB(bp->b_maps[0].bm_bn) >> PAGE_SHIFT; + end = (BBTOB(bp->b_maps[0].bm_bn + bp->b_length) + PAGE_SIZE - 1) >> PAGE_SHIFT; page_count = end - start; error = _xfs_buf_get_pages(bp, page_count, flags); @@ -640,7 +640,7 @@ _xfs_buf_read( xfs_buf_flags_t flags) { ASSERT(!(flags & XBF_WRITE)); - ASSERT(bp->b_map.bm_bn != XFS_BUF_DADDR_NULL); + ASSERT(bp->b_maps[0].bm_bn != XFS_BUF_DADDR_NULL); bp->b_flags &= ~(XBF_WRITE | XBF_ASYNC | XBF_READ_AHEAD); bp->b_flags |= flags & (XBF_READ | XBF_ASYNC | XBF_READ_AHEAD); @@ -1709,7 +1709,7 @@ xfs_buf_cmp( struct xfs_buf *bp = container_of(b, struct xfs_buf, b_list); xfs_daddr_t diff; - diff = ap->b_map.bm_bn - bp->b_map.bm_bn; + diff = ap->b_maps[0].bm_bn - bp->b_maps[0].bm_bn; if (diff < 0) return -1; if (diff > 0) diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h index 23f5642480bb..433a12ed7b17 100644 --- a/fs/xfs/xfs_buf.h +++ b/fs/xfs/xfs_buf.h @@ -151,7 +151,7 @@ typedef struct xfs_buf { struct page **b_pages; /* array of page pointers */ struct page *b_page_array[XB_PAGES]; /* inline pages */ struct xfs_buf_map *b_maps; /* compound buffer map */ - struct xfs_buf_map b_map; /* inline compound buffer map */ + struct xfs_buf_map __b_map; /* inline compound buffer map */ int b_map_count; int b_io_length; /* IO size in BBs */ atomic_t b_pin_count; /* pin count */ @@ -330,8 +330,8 @@ void xfs_buf_stale(struct xfs_buf *bp); * In future, uncached buffers will pass the block number directly to the io * request function and hence these macros will go away at that point. */ -#define XFS_BUF_ADDR(bp) ((bp)->b_map.bm_bn) -#define XFS_BUF_SET_ADDR(bp, bno) ((bp)->b_map.bm_bn = (xfs_daddr_t)(bno)) +#define XFS_BUF_ADDR(bp) ((bp)->b_maps[0].bm_bn) +#define XFS_BUF_SET_ADDR(bp, bno) ((bp)->b_maps[0].bm_bn = (xfs_daddr_t)(bno)) static inline void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref) { -- cgit From b94381737e9c4d014a4003e8ece9ba88670a2dd4 Mon Sep 17 00:00:00 2001 From: Mark Tinguely Date: Tue, 4 Dec 2012 17:18:03 -0600 Subject: xfs: rename bli_format to avoid confusion with bli_formats Rename the bli_format structure to __bli_format to avoid accidently confusing them with the bli_formats pointer. Signed-off-by: Mark Tinguely Reviewed-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_buf_item.c | 22 +++++++++++----------- fs/xfs/xfs_buf_item.h | 2 +- fs/xfs/xfs_trans_buf.c | 24 ++++++++++++------------ 3 files changed, 24 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index becf4a97efc6..1975b3d9007a 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -71,7 +71,7 @@ xfs_buf_item_log_debug( chunk_num = byte >> XFS_BLF_SHIFT; word_num = chunk_num >> BIT_TO_WORD_SHIFT; bit_num = chunk_num & (NBWORD - 1); - wordp = &(bip->bli_format.blf_data_map[word_num]); + wordp = &(bip->__bli_format.blf_data_map[word_num]); bit_set = *wordp & (1 << bit_num); ASSERT(bit_set); byte++; @@ -237,7 +237,7 @@ xfs_buf_item_size( * cancel flag in it. */ trace_xfs_buf_item_size_stale(bip); - ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL); + ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL); return bip->bli_format_count; } @@ -278,7 +278,7 @@ xfs_buf_item_format_segment( uint buffer_offset; /* copy the flags across from the base format item */ - blfp->blf_flags = bip->bli_format.blf_flags; + blfp->blf_flags = bip->__bli_format.blf_flags; /* * Base size is the actual size of the ondisk structure - it reflects @@ -371,7 +371,7 @@ xfs_buf_item_format_segment( nbits++; } } - bip->bli_format.blf_size = nvecs; + bip->__bli_format.blf_size = nvecs; return vecp; } @@ -405,7 +405,7 @@ xfs_buf_item_format( if (bip->bli_flags & XFS_BLI_INODE_BUF) { if (!((bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF) && xfs_log_item_in_current_chkpt(lip))) - bip->bli_format.blf_flags |= XFS_BLF_INODE_BUF; + bip->__bli_format.blf_flags |= XFS_BLF_INODE_BUF; bip->bli_flags &= ~XFS_BLI_INODE_BUF; } @@ -485,7 +485,7 @@ xfs_buf_item_unpin( ASSERT(bip->bli_flags & XFS_BLI_STALE); ASSERT(xfs_buf_islocked(bp)); ASSERT(XFS_BUF_ISSTALE(bp)); - ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL); + ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL); trace_xfs_buf_item_unpin_stale(bip); @@ -631,7 +631,7 @@ xfs_buf_item_unlock( */ if (bip->bli_flags & XFS_BLI_STALE) { trace_xfs_buf_item_unlock_stale(bip); - ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL); + ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL); if (!aborted) { atomic_dec(&bip->bli_refcount); return; @@ -644,8 +644,8 @@ xfs_buf_item_unlock( * If the buf item isn't tracking any data, free it, otherwise drop the * reference we hold to it. */ - if (xfs_bitmap_empty(bip->bli_format.blf_data_map, - bip->bli_format.blf_map_size)) + if (xfs_bitmap_empty(bip->__bli_format.blf_data_map, + bip->__bli_format.blf_map_size)) xfs_buf_item_relse(bp); else atomic_dec(&bip->bli_refcount); @@ -716,7 +716,7 @@ xfs_buf_item_get_format( bip->bli_format_count = count; if (count == 1) { - bip->bli_formats = &bip->bli_format; + bip->bli_formats = &bip->__bli_format; return 0; } @@ -731,7 +731,7 @@ STATIC void xfs_buf_item_free_format( struct xfs_buf_log_item *bip) { - if (bip->bli_formats != &bip->bli_format) { + if (bip->bli_formats != &bip->__bli_format) { kmem_free(bip->bli_formats); bip->bli_formats = NULL; } diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h index 6850f49f4af3..16def435944a 100644 --- a/fs/xfs/xfs_buf_item.h +++ b/fs/xfs/xfs_buf_item.h @@ -104,7 +104,7 @@ typedef struct xfs_buf_log_item { #endif int bli_format_count; /* count of headers */ struct xfs_buf_log_format *bli_formats; /* array of in-log header ptrs */ - struct xfs_buf_log_format bli_format; /* embedded in-log header */ + struct xfs_buf_log_format __bli_format; /* embedded in-log header */ } xfs_buf_log_item_t; void xfs_buf_item_init(struct xfs_buf *, struct xfs_mount *); diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c index 4fc17d479d42..f7510bf68284 100644 --- a/fs/xfs/xfs_trans_buf.c +++ b/fs/xfs/xfs_trans_buf.c @@ -93,7 +93,7 @@ _xfs_trans_bjoin( xfs_buf_item_init(bp, tp->t_mountp); bip = bp->b_fspriv; ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); - ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL)); + ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL)); ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED)); if (reset_recur) bip->bli_recur = 0; @@ -432,7 +432,7 @@ xfs_trans_brelse(xfs_trans_t *tp, bip = bp->b_fspriv; ASSERT(bip->bli_item.li_type == XFS_LI_BUF); ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); - ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL)); + ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL)); ASSERT(atomic_read(&bip->bli_refcount) > 0); trace_xfs_trans_brelse(bip); @@ -519,7 +519,7 @@ xfs_trans_bhold(xfs_trans_t *tp, ASSERT(bp->b_transp == tp); ASSERT(bip != NULL); ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); - ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL)); + ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL)); ASSERT(atomic_read(&bip->bli_refcount) > 0); bip->bli_flags |= XFS_BLI_HOLD; @@ -539,7 +539,7 @@ xfs_trans_bhold_release(xfs_trans_t *tp, ASSERT(bp->b_transp == tp); ASSERT(bip != NULL); ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); - ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_CANCEL)); + ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_CANCEL)); ASSERT(atomic_read(&bip->bli_refcount) > 0); ASSERT(bip->bli_flags & XFS_BLI_HOLD); @@ -598,7 +598,7 @@ xfs_trans_log_buf(xfs_trans_t *tp, bip->bli_flags &= ~XFS_BLI_STALE; ASSERT(XFS_BUF_ISSTALE(bp)); XFS_BUF_UNSTALE(bp); - bip->bli_format.blf_flags &= ~XFS_BLF_CANCEL; + bip->__bli_format.blf_flags &= ~XFS_BLF_CANCEL; } tp->t_flags |= XFS_TRANS_DIRTY; @@ -657,8 +657,8 @@ xfs_trans_binval( */ ASSERT(XFS_BUF_ISSTALE(bp)); ASSERT(!(bip->bli_flags & (XFS_BLI_LOGGED | XFS_BLI_DIRTY))); - ASSERT(!(bip->bli_format.blf_flags & XFS_BLF_INODE_BUF)); - ASSERT(bip->bli_format.blf_flags & XFS_BLF_CANCEL); + ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_INODE_BUF)); + ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL); ASSERT(bip->bli_item.li_desc->lid_flags & XFS_LID_DIRTY); ASSERT(tp->t_flags & XFS_TRANS_DIRTY); return; @@ -668,10 +668,10 @@ xfs_trans_binval( bip->bli_flags |= XFS_BLI_STALE; bip->bli_flags &= ~(XFS_BLI_INODE_BUF | XFS_BLI_LOGGED | XFS_BLI_DIRTY); - bip->bli_format.blf_flags &= ~XFS_BLF_INODE_BUF; - bip->bli_format.blf_flags |= XFS_BLF_CANCEL; - memset((char *)(bip->bli_format.blf_data_map), 0, - (bip->bli_format.blf_map_size * sizeof(uint))); + bip->__bli_format.blf_flags &= ~XFS_BLF_INODE_BUF; + bip->__bli_format.blf_flags |= XFS_BLF_CANCEL; + memset((char *)(bip->__bli_format.blf_data_map), 0, + (bip->__bli_format.blf_map_size * sizeof(uint))); bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY; tp->t_flags |= XFS_TRANS_DIRTY; } @@ -775,5 +775,5 @@ xfs_trans_dquot_buf( type == XFS_BLF_GDQUOT_BUF); ASSERT(atomic_read(&bip->bli_refcount) > 0); - bip->bli_format.blf_flags |= type; + bip->__bli_format.blf_flags |= type; } -- cgit From 820a554f2f83d21f7e9e608377c8c86af70917bd Mon Sep 17 00:00:00 2001 From: Mark Tinguely Date: Tue, 4 Dec 2012 17:18:04 -0600 Subject: xfs: fix segment in xfs_buf_item_format_segment Not every segment in a multi-segment buffer is dirty in a transaction and they will not be outputted. The assert in xfs_buf_item_format_segment() that checks for the at least one chunk of data in the segment to be used is not necessary true for multi-segmented buffers. Signed-off-by: Mark Tinguely Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_buf_item.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 1975b3d9007a..c48e60bd857d 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -287,6 +287,17 @@ xfs_buf_item_format_segment( */ base_size = offsetof(struct xfs_buf_log_format, blf_data_map) + (blfp->blf_map_size * sizeof(blfp->blf_data_map[0])); + + nvecs = 0; + first_bit = xfs_next_bit(blfp->blf_data_map, blfp->blf_map_size, 0); + if (!(bip->bli_flags & XFS_BLI_STALE) && first_bit == -1) { + /* + * If the map is not be dirty in the transaction, mark + * the size as zero and do not advance the vector pointer. + */ + goto out; + } + vecp->i_addr = blfp; vecp->i_len = base_size; vecp->i_type = XLOG_REG_TYPE_BFORMAT; @@ -301,15 +312,13 @@ xfs_buf_item_format_segment( */ trace_xfs_buf_item_format_stale(bip); ASSERT(blfp->blf_flags & XFS_BLF_CANCEL); - blfp->blf_size = nvecs; - return vecp; + goto out; } /* * Fill in an iovec for each set of contiguous chunks. */ - first_bit = xfs_next_bit(blfp->blf_data_map, blfp->blf_map_size, 0); - ASSERT(first_bit != -1); + last_bit = first_bit; nbits = 1; for (;;) { @@ -371,7 +380,8 @@ xfs_buf_item_format_segment( nbits++; } } - bip->__bli_format.blf_size = nvecs; +out: + blfp->blf_size = nvecs; return vecp; } -- cgit From c883d0c400fc8cdbd5bbe71e179c9e64ace58e86 Mon Sep 17 00:00:00 2001 From: Mark Tinguely Date: Tue, 4 Dec 2012 17:18:05 -0600 Subject: xfs: fix the multi-segment log buffer format Per Dave Chinner suggestion, this patch: 1) Corrects the detection of whether a multi-segment buffer is still tracking data. 2) Clears all the buffer log formats for a multi-segment buffer. Signed-off-by: Mark Tinguely Reviewed-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_buf_item.c | 13 ++++++++++--- fs/xfs/xfs_trans_buf.c | 7 +++++-- 2 files changed, 15 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index c48e60bd857d..77b09750e92c 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -611,7 +611,7 @@ xfs_buf_item_unlock( { struct xfs_buf_log_item *bip = BUF_ITEM(lip); struct xfs_buf *bp = bip->bli_buf; - int aborted; + int aborted, clean, i; uint hold; /* Clear the buffer's association with this transaction. */ @@ -654,8 +654,15 @@ xfs_buf_item_unlock( * If the buf item isn't tracking any data, free it, otherwise drop the * reference we hold to it. */ - if (xfs_bitmap_empty(bip->__bli_format.blf_data_map, - bip->__bli_format.blf_map_size)) + clean = 1; + for (i = 0; i < bip->bli_format_count; i++) { + if (!xfs_bitmap_empty(bip->bli_formats[i].blf_data_map, + bip->bli_formats[i].blf_map_size)) { + clean = 0; + break; + } + } + if (clean) xfs_buf_item_relse(bp); else atomic_dec(&bip->bli_refcount); diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c index f7510bf68284..3edf5dbee001 100644 --- a/fs/xfs/xfs_trans_buf.c +++ b/fs/xfs/xfs_trans_buf.c @@ -643,6 +643,7 @@ xfs_trans_binval( xfs_buf_t *bp) { xfs_buf_log_item_t *bip = bp->b_fspriv; + int i; ASSERT(bp->b_transp == tp); ASSERT(bip != NULL); @@ -670,8 +671,10 @@ xfs_trans_binval( bip->bli_flags &= ~(XFS_BLI_INODE_BUF | XFS_BLI_LOGGED | XFS_BLI_DIRTY); bip->__bli_format.blf_flags &= ~XFS_BLF_INODE_BUF; bip->__bli_format.blf_flags |= XFS_BLF_CANCEL; - memset((char *)(bip->__bli_format.blf_data_map), 0, - (bip->__bli_format.blf_map_size * sizeof(uint))); + for (i = 0; i < bip->bli_format_count; i++) { + memset(bip->bli_formats[i].blf_data_map, 0, + (bip->bli_formats[i].blf_map_size * sizeof(uint))); + } bip->bli_item.li_desc->lid_flags |= XFS_LID_DIRTY; tp->t_flags |= XFS_TRANS_DIRTY; } -- cgit From ec47eb6b0b450a4e82340b6de674104de3f0dc0a Mon Sep 17 00:00:00 2001 From: Mark Tinguely Date: Tue, 4 Dec 2012 17:18:06 -0600 Subject: xfs remove the XFS_TRANS_DEBUG routines Remove the XFS_TRANS_DEBUG routines. They are no longer appropriate and have not been used in years Signed-off-by: Mark Tinguely Reviewed-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_buf_item.c | 106 ----------------------------------------------- fs/xfs/xfs_buf_item.h | 14 ------- fs/xfs/xfs_inode.c | 6 --- fs/xfs/xfs_inode_item.c | 16 ------- fs/xfs/xfs_inode_item.h | 4 -- fs/xfs/xfs_trans_ail.c | 14 ------- fs/xfs/xfs_trans_inode.c | 41 ------------------ 7 files changed, 201 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 77b09750e92c..63c86c48014b 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -37,109 +37,6 @@ static inline struct xfs_buf_log_item *BUF_ITEM(struct xfs_log_item *lip) return container_of(lip, struct xfs_buf_log_item, bli_item); } - -#ifdef XFS_TRANS_DEBUG -/* - * This function uses an alternate strategy for tracking the bytes - * that the user requests to be logged. This can then be used - * in conjunction with the bli_orig array in the buf log item to - * catch bugs in our callers' code. - * - * We also double check the bits set in xfs_buf_item_log using a - * simple algorithm to check that every byte is accounted for. - */ -STATIC void -xfs_buf_item_log_debug( - xfs_buf_log_item_t *bip, - uint first, - uint last) -{ - uint x; - uint byte; - uint nbytes; - uint chunk_num; - uint word_num; - uint bit_num; - uint bit_set; - uint *wordp; - - ASSERT(bip->bli_logged != NULL); - byte = first; - nbytes = last - first + 1; - bfset(bip->bli_logged, first, nbytes); - for (x = 0; x < nbytes; x++) { - chunk_num = byte >> XFS_BLF_SHIFT; - word_num = chunk_num >> BIT_TO_WORD_SHIFT; - bit_num = chunk_num & (NBWORD - 1); - wordp = &(bip->__bli_format.blf_data_map[word_num]); - bit_set = *wordp & (1 << bit_num); - ASSERT(bit_set); - byte++; - } -} - -/* - * This function is called when we flush something into a buffer without - * logging it. This happens for things like inodes which are logged - * separately from the buffer. - */ -void -xfs_buf_item_flush_log_debug( - xfs_buf_t *bp, - uint first, - uint last) -{ - xfs_buf_log_item_t *bip = bp->b_fspriv; - uint nbytes; - - if (bip == NULL || (bip->bli_item.li_type != XFS_LI_BUF)) - return; - - ASSERT(bip->bli_logged != NULL); - nbytes = last - first + 1; - bfset(bip->bli_logged, first, nbytes); -} - -/* - * This function is called to verify that our callers have logged - * all the bytes that they changed. - * - * It does this by comparing the original copy of the buffer stored in - * the buf log item's bli_orig array to the current copy of the buffer - * and ensuring that all bytes which mismatch are set in the bli_logged - * array of the buf log item. - */ -STATIC void -xfs_buf_item_log_check( - xfs_buf_log_item_t *bip) -{ - char *orig; - char *buffer; - int x; - xfs_buf_t *bp; - - ASSERT(bip->bli_orig != NULL); - ASSERT(bip->bli_logged != NULL); - - bp = bip->bli_buf; - ASSERT(bp->b_length > 0); - ASSERT(bp->b_addr != NULL); - orig = bip->bli_orig; - buffer = bp->b_addr; - for (x = 0; x < BBTOB(bp->b_length); x++) { - if (orig[x] != buffer[x] && !btst(bip->bli_logged, x)) { - xfs_emerg(bp->b_mount, - "%s: bip %x buffer %x orig %x index %d", - __func__, bip, bp, orig, x); - ASSERT(0); - } - } -} -#else -#define xfs_buf_item_log_debug(x,y,z) -#define xfs_buf_item_log_check(x) -#endif - STATIC void xfs_buf_do_callbacks(struct xfs_buf *bp); /* @@ -429,7 +326,6 @@ xfs_buf_item_format( * Check to make sure everything is consistent. */ trace_xfs_buf_item_format(bip); - xfs_buf_item_log_check(bip); } /* @@ -915,8 +811,6 @@ xfs_buf_item_log_segment( mask = (1 << end_bit) - 1; *wordp |= mask; } - - xfs_buf_item_log_debug(bip, first, last); } /* diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h index 16def435944a..ee36c88ecfde 100644 --- a/fs/xfs/xfs_buf_item.h +++ b/fs/xfs/xfs_buf_item.h @@ -98,10 +98,6 @@ typedef struct xfs_buf_log_item { unsigned int bli_flags; /* misc flags */ unsigned int bli_recur; /* lock recursion count */ atomic_t bli_refcount; /* cnt of tp refs */ -#ifdef XFS_TRANS_DEBUG - char *bli_orig; /* original buffer copy */ - char *bli_logged; /* bytes logged (bitmap) */ -#endif int bli_format_count; /* count of headers */ struct xfs_buf_log_format *bli_formats; /* array of in-log header ptrs */ struct xfs_buf_log_format __bli_format; /* embedded in-log header */ @@ -117,16 +113,6 @@ void xfs_buf_attach_iodone(struct xfs_buf *, void xfs_buf_iodone_callbacks(struct xfs_buf *); void xfs_buf_iodone(struct xfs_buf *, struct xfs_log_item *); -#ifdef XFS_TRANS_DEBUG -void -xfs_buf_item_flush_log_debug( - struct xfs_buf *bp, - uint first, - uint last); -#else -#define xfs_buf_item_flush_log_debug(bp, first, last) -#endif - #endif /* __KERNEL__ */ #endif /* __XFS_BUF_ITEM_H__ */ diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 66282dcb821b..4f201656d2d9 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -2379,9 +2379,6 @@ xfs_iflush_fork( char *cp; xfs_ifork_t *ifp; xfs_mount_t *mp; -#ifdef XFS_TRANS_DEBUG - int first; -#endif static const short brootflag[2] = { XFS_ILOG_DBROOT, XFS_ILOG_ABROOT }; static const short dataflag[2] = @@ -2724,9 +2721,6 @@ xfs_iflush_int( xfs_inode_log_item_t *iip; xfs_dinode_t *dip; xfs_mount_t *mp; -#ifdef XFS_TRANS_DEBUG - int first; -#endif ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)); ASSERT(xfs_isiflocked(ip)); diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index d041d47d9d86..f034bd1652f0 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -269,17 +269,6 @@ xfs_inode_item_format( } else { ASSERT(!(iip->ili_fields & XFS_ILOG_DBROOT)); -#ifdef XFS_TRANS_DEBUG - if (iip->ili_root_size > 0) { - ASSERT(iip->ili_root_size == - ip->i_df.if_broot_bytes); - ASSERT(memcmp(iip->ili_orig_root, - ip->i_df.if_broot, - iip->ili_root_size) == 0); - } else { - ASSERT(ip->i_df.if_broot_bytes == 0); - } -#endif iip->ili_fields &= ~XFS_ILOG_DBROOT; } break; @@ -678,11 +667,6 @@ void xfs_inode_item_destroy( xfs_inode_t *ip) { -#ifdef XFS_TRANS_DEBUG - if (ip->i_itemp->ili_root_size != 0) { - kmem_free(ip->i_itemp->ili_orig_root); - } -#endif kmem_zone_free(xfs_ili_zone, ip->i_itemp); } diff --git a/fs/xfs/xfs_inode_item.h b/fs/xfs/xfs_inode_item.h index 376d4d0b2635..779812fb3d80 100644 --- a/fs/xfs/xfs_inode_item.h +++ b/fs/xfs/xfs_inode_item.h @@ -148,10 +148,6 @@ typedef struct xfs_inode_log_item { data exts */ struct xfs_bmbt_rec *ili_aextents_buf; /* array of logged attr exts */ -#ifdef XFS_TRANS_DEBUG - int ili_root_size; - char *ili_orig_root; -#endif xfs_inode_log_format_t ili_format; /* logged structure */ } xfs_inode_log_item_t; diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index 6011ee661339..0eda7254305f 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -55,20 +55,6 @@ xfs_ail_check( ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) >= 0); -#ifdef XFS_TRANS_DEBUG - /* - * Walk the list checking lsn ordering, and that every entry has the - * XFS_LI_IN_AIL flag set. This is really expensive, so only do it - * when specifically debugging the transaction subsystem. - */ - prev_lip = list_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); - list_for_each_entry(lip, &ailp->xa_ail, li_ail) { - if (&prev_lip->li_ail != &ailp->xa_ail) - ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); - ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); - prev_lip = lip; - } -#endif /* XFS_TRANS_DEBUG */ } #else /* !DEBUG */ #define xfs_ail_check(a,l) diff --git a/fs/xfs/xfs_trans_inode.c b/fs/xfs/xfs_trans_inode.c index d2eee20d5f5b..ac6d567704db 100644 --- a/fs/xfs/xfs_trans_inode.c +++ b/fs/xfs/xfs_trans_inode.c @@ -33,14 +33,6 @@ #include "xfs_inode_item.h" #include "xfs_trace.h" -#ifdef XFS_TRANS_DEBUG -STATIC void -xfs_trans_inode_broot_debug( - xfs_inode_t *ip); -#else -#define xfs_trans_inode_broot_debug(ip) -#endif - /* * Add a locked inode to the transaction. * @@ -67,8 +59,6 @@ xfs_trans_ijoin( * Get a log_item_desc to point at the new item. */ xfs_trans_add_item(tp, &iip->ili_item); - - xfs_trans_inode_broot_debug(ip); } /* @@ -135,34 +125,3 @@ xfs_trans_log_inode( flags |= ip->i_itemp->ili_last_fields; ip->i_itemp->ili_fields |= flags; } - -#ifdef XFS_TRANS_DEBUG -/* - * Keep track of the state of the inode btree root to make sure we - * log it properly. - */ -STATIC void -xfs_trans_inode_broot_debug( - xfs_inode_t *ip) -{ - xfs_inode_log_item_t *iip; - - ASSERT(ip->i_itemp != NULL); - iip = ip->i_itemp; - if (iip->ili_root_size != 0) { - ASSERT(iip->ili_orig_root != NULL); - kmem_free(iip->ili_orig_root); - iip->ili_root_size = 0; - iip->ili_orig_root = NULL; - } - if (ip->i_d.di_format == XFS_DINODE_FMT_BTREE) { - ASSERT((ip->i_df.if_broot != NULL) && - (ip->i_df.if_broot_bytes > 0)); - iip->ili_root_size = ip->i_df.if_broot_bytes; - iip->ili_orig_root = - (char*)kmem_alloc(iip->ili_root_size, KM_SLEEP); - memcpy(iip->ili_orig_root, (char*)(ip->i_df.if_broot), - iip->ili_root_size); - } -} -#endif -- cgit From f755503206ef705c20db622637d80a3e1b94a6f5 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Fri, 21 Dec 2012 10:45:17 -0500 Subject: xfs: remove int casts from debug dquot soft limit timer asserts The int casts here make it easy to trigger an assert with a large soft limit. For example, set a >4TB soft limit on an empty volume to reproduce a (0 > -x) comparison due to an overflow of d_blk_softlimit. Signed-off-by: Brian Foster Reviewed-by: Ben Myers Signed-off-by: Ben Myers --- fs/xfs/xfs_qm_syscalls.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 5f53e75409b8..8a59f8546552 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -784,11 +784,11 @@ xfs_qm_scall_getquota( (XFS_IS_OQUOTA_ENFORCED(mp) && (dst->d_flags & (FS_PROJ_QUOTA | FS_GROUP_QUOTA)))) && dst->d_id != 0) { - if (((int) dst->d_bcount > (int) dst->d_blk_softlimit) && + if ((dst->d_bcount > dst->d_blk_softlimit) && (dst->d_blk_softlimit > 0)) { ASSERT(dst->d_btimer != 0); } - if (((int) dst->d_icount > (int) dst->d_ino_softlimit) && + if ((dst->d_icount > dst->d_ino_softlimit) && (dst->d_ino_softlimit > 0)) { ASSERT(dst->d_itimer != 0); } -- cgit From 83a9ba00573904953c58565226f1d3bab9f10983 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Mon, 10 Dec 2012 14:49:15 -0600 Subject: xfs: don't zero structure members after a memset(0) Commit 408cc4e97a3ccd172d2d676e4b585badf439271b added memset(0, ...) to allocation args structures, so there is no need to explicitly set any of the fields to 0 after that. Signed-off-by: Eric Sandeen Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_alloc.c | 2 -- fs/xfs/xfs_bmap.c | 4 ---- fs/xfs/xfs_ialloc.c | 4 ---- 3 files changed, 10 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index 393055fe3aef..0ad23253e8b1 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -1925,8 +1925,6 @@ xfs_alloc_fix_freelist( targs.mp = mp; targs.agbp = agbp; targs.agno = args->agno; - targs.mod = targs.minleft = targs.wasdel = targs.userdata = - targs.minalignslop = 0; targs.alignment = targs.minlen = targs.prod = targs.isfl = 1; targs.type = XFS_ALLOCTYPE_THIS_AG; targs.pag = pag; diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 0e92d12765d2..c5077208e419 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -3099,8 +3099,6 @@ xfs_bmap_extents_to_btree( args.fsbno = *firstblock; } args.minlen = args.maxlen = args.prod = 1; - args.total = args.minleft = args.alignment = args.mod = args.isfl = - args.minalignslop = 0; args.wasdel = wasdel; *logflagsp = 0; if ((error = xfs_alloc_vextent(&args))) { @@ -3259,8 +3257,6 @@ xfs_bmap_local_to_extents( args.type = XFS_ALLOCTYPE_NEAR_BNO; } args.total = total; - args.mod = args.minleft = args.alignment = args.wasdel = - args.isfl = args.minalignslop = 0; args.minlen = args.maxlen = args.prod = 1; if ((error = xfs_alloc_vextent(&args))) goto done; diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index a815412eab80..515bf71ce01c 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -279,8 +279,6 @@ xfs_ialloc_ag_alloc( (args.agbno < be32_to_cpu(agi->agi_length)))) { args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno); args.type = XFS_ALLOCTYPE_THIS_BNO; - args.mod = args.total = args.wasdel = args.isfl = - args.userdata = args.minalignslop = 0; args.prod = 1; /* @@ -333,8 +331,6 @@ xfs_ialloc_ag_alloc( * Allocate a fixed-size extent of inodes. */ args.type = XFS_ALLOCTYPE_NEAR_BNO; - args.mod = args.total = args.wasdel = args.isfl = - args.userdata = args.minalignslop = 0; args.prod = 1; /* * Allow space for the inode btree to split. -- cgit From f11722834605a155022a8098ad7d8adacf44b22f Mon Sep 17 00:00:00 2001 From: David Teigland Date: Mon, 7 Jan 2013 11:03:42 -0600 Subject: dlm: avoid scanning unchanged toss lists Keep track of whether a toss list contains any shrinkable rsbs. If not, dlm_scand can avoid scanning the list for rsbs to shrink. Unnecessary scanning can otherwise waste a lot of time because the toss lists can contain a large number of rsbs that are non-shrinkable (directory records). Signed-off-by: David Teigland --- fs/dlm/dlm_internal.h | 3 +++ fs/dlm/lock.c | 15 +++++++++++++++ 2 files changed, 18 insertions(+) (limited to 'fs') diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index 77c0f70f8fe8..e7665c31f7b1 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -96,10 +96,13 @@ do { \ } +#define DLM_RTF_SHRINK 0x00000001 + struct dlm_rsbtable { struct rb_root keep; struct rb_root toss; spinlock_t lock; + uint32_t flags; }; diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index a579f30f237d..f7501651762d 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -1132,6 +1132,7 @@ static void toss_rsb(struct kref *kref) rb_erase(&r->res_hashnode, &ls->ls_rsbtbl[r->res_bucket].keep); rsb_insert(r, &ls->ls_rsbtbl[r->res_bucket].toss); r->res_toss_time = jiffies; + ls->ls_rsbtbl[r->res_bucket].flags |= DLM_RTF_SHRINK; if (r->res_lvbptr) { dlm_free_lvb(r->res_lvbptr); r->res_lvbptr = NULL; @@ -1659,11 +1660,18 @@ static void shrink_bucket(struct dlm_ls *ls, int b) char *name; int our_nodeid = dlm_our_nodeid(); int remote_count = 0; + int need_shrink = 0; int i, len, rv; memset(&ls->ls_remove_lens, 0, sizeof(int) * DLM_REMOVE_NAMES_MAX); spin_lock(&ls->ls_rsbtbl[b].lock); + + if (!(ls->ls_rsbtbl[b].flags & DLM_RTF_SHRINK)) { + spin_unlock(&ls->ls_rsbtbl[b].lock); + return; + } + for (n = rb_first(&ls->ls_rsbtbl[b].toss); n; n = next) { next = rb_next(n); r = rb_entry(n, struct dlm_rsb, res_hashnode); @@ -1679,6 +1687,8 @@ static void shrink_bucket(struct dlm_ls *ls, int b) continue; } + need_shrink = 1; + if (!time_after_eq(jiffies, r->res_toss_time + dlm_config.ci_toss_secs * HZ)) { continue; @@ -1710,6 +1720,11 @@ static void shrink_bucket(struct dlm_ls *ls, int b) rb_erase(&r->res_hashnode, &ls->ls_rsbtbl[b].toss); dlm_free_rsb(r); } + + if (need_shrink) + ls->ls_rsbtbl[b].flags |= DLM_RTF_SHRINK; + else + ls->ls_rsbtbl[b].flags &= ~DLM_RTF_SHRINK; spin_unlock(&ls->ls_rsbtbl[b].lock); /* -- cgit From 2c016dc2cbd160acf099bee8e4d5c056a8b691b1 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Wed, 26 Dec 2012 15:32:17 +0800 Subject: btrfs: fix comment typos Convert 'hepler' to 'helper'. Signed-off-by: Liu Bo Signed-off-by: Jiri Kosina --- fs/btrfs/extent-tree.c | 6 +++--- fs/btrfs/relocation.c | 4 ++-- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 06b2635073f3..604befdcc359 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -6482,7 +6482,7 @@ reada: } /* - * hepler to process tree block while walking down the tree. + * helper to process tree block while walking down the tree. * * when wc->stage == UPDATE_BACKREF, this function updates * back refs for pointers in the block. @@ -6557,7 +6557,7 @@ static noinline int walk_down_proc(struct btrfs_trans_handle *trans, } /* - * hepler to process tree block pointer. + * helper to process tree block pointer. * * when wc->stage == DROP_REFERENCE, this function checks * reference count of the block pointed to. if the block @@ -6695,7 +6695,7 @@ skip: } /* - * hepler to process tree block while walking up the tree. + * helper to process tree block while walking up the tree. * * when wc->stage == DROP_REFERENCE, this function drops * reference count on the block. diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 776f0aa128fc..c188e815a733 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -3474,7 +3474,7 @@ out: } /* - * hepler to find all tree blocks that reference a given data extent + * helper to find all tree blocks that reference a given data extent */ static noinline_for_stack int add_data_references(struct reloc_control *rc, @@ -3568,7 +3568,7 @@ int add_data_references(struct reloc_control *rc, } /* - * hepler to find next unprocessed extent + * helper to find next unprocessed extent */ static noinline_for_stack int find_next_extent(struct btrfs_trans_handle *trans, -- cgit From 8a168ca7074b463d0e19a9e9c15946db1afbddc4 Mon Sep 17 00:00:00 2001 From: Masanari Iida Date: Sat, 29 Dec 2012 02:00:09 +0900 Subject: treewide: Fix typo in various drivers Correct spelling typo in printk within various drivers. Signed-off-by: Masanari Iida Signed-off-by: Jiri Kosina --- fs/cifs/link.c | 2 +- fs/hpfs/inode.c | 2 +- fs/ocfs2/cluster/tcp.c | 2 +- fs/qnx6/inode.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/cifs/link.c b/fs/cifs/link.c index 51dc2fb6e854..9f6c4c45d21e 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -76,7 +76,7 @@ symlink_hash(unsigned int link_len, const char *link_str, u8 *md5_hash) } rc = crypto_shash_update(&sdescmd5->shash, link_str, link_len); if (rc) { - cERROR(1, "%s: Could not update iwth link_str", __func__); + cERROR(1, "%s: Could not update with link_str", __func__); goto symlink_hash_err; } rc = crypto_shash_final(&sdescmd5->shash, md5_hash); diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c index 804a9a842cbc..405ab77db39c 100644 --- a/fs/hpfs/inode.c +++ b/fs/hpfs/inode.c @@ -147,7 +147,7 @@ static void hpfs_write_inode_ea(struct inode *i, struct fnode *fnode) /*if (le32_to_cpu(fnode->acl_size_l) || le16_to_cpu(fnode->acl_size_s)) { Some unknown structures like ACL may be in fnode, we'd better not overwrite them - hpfs_error(i->i_sb, "fnode %08x has some unknown HPFS386 stuctures", i->i_ino); + hpfs_error(i->i_sb, "fnode %08x has some unknown HPFS386 structures", i->i_ino); } else*/ if (hpfs_sb(i->i_sb)->sb_eas >= 2) { __le32 ea; if (!uid_eq(i->i_uid, hpfs_sb(i->i_sb)->sb_uid) || hpfs_inode->i_ea_uid) { diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 1bfe8802cc1e..a1d83c58b296 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -870,7 +870,7 @@ int o2net_register_handler(u32 msg_type, u32 key, u32 max_len, /* we've had some trouble with handlers seemingly vanishing. */ mlog_bug_on_msg(o2net_handler_tree_lookup(msg_type, key, &p, &parent) == NULL, - "couldn't find handler we *just* registerd " + "couldn't find handler we *just* registered " "for type %u key %08x\n", msg_type, key); } write_unlock(&o2net_handler_lock); diff --git a/fs/qnx6/inode.c b/fs/qnx6/inode.c index b6addf560483..57199a52a351 100644 --- a/fs/qnx6/inode.c +++ b/fs/qnx6/inode.c @@ -285,7 +285,7 @@ static struct buffer_head *qnx6_check_first_superblock(struct super_block *s, if (fs32_to_cpu(sbi, sb->sb_magic) == QNX6_SUPER_MAGIC) { /* we got a big endian fs */ QNX6DEBUG((KERN_INFO "qnx6: fs got different" - " endianess.\n")); + " endianness.\n")); return bh; } else sbi->s_bytesex = BYTESEX_LE; -- cgit From 210b907acfb414650dcdae9b470eb2fccac7996b Mon Sep 17 00:00:00 2001 From: Wang Sheng-Hui Date: Mon, 24 Sep 2012 08:47:23 +0800 Subject: btrfs: remove unnecessary cur_trans set before goto loop in join_transaction In the big loop, cur_trans will be set fs_info->running_transaction before it's used. And after kmem_cache_free it and goto loop, it will be setup again. No need to setup it immediately after freed. Signed-off-by: Wang Sheng-Hui Signed-off-by: Jiri Kosina --- fs/btrfs/transaction.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 04bbfb1052eb..68704e74f0d3 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -111,7 +111,6 @@ loop: * to redo the trans_no_join checks above */ kmem_cache_free(btrfs_transaction_cachep, cur_trans); - cur_trans = fs_info->running_transaction; goto loop; } else if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { spin_unlock(&fs_info->trans_lock); -- cgit From 9f244e9cfd70c7c0f82d3c92ce772ab2a92d9f64 Mon Sep 17 00:00:00 2001 From: Seiji Aguchi Date: Fri, 11 Jan 2013 18:09:41 +0000 Subject: pstore: Avoid deadlock in panic and emergency-restart path [Issue] When pstore is in panic and emergency-restart paths, it may be blocked in those paths because it simply takes spin_lock. This is an example scenario which pstore may hang up in a panic path: - cpuA grabs psinfo->buf_lock - cpuB panics and calls smp_send_stop - smp_send_stop sends IRQ to cpuA - after 1 second, cpuB gives up on cpuA and sends an NMI instead - cpuA is now in an NMI handler while still holding buf_lock - cpuB is deadlocked This case may happen if a firmware has a bug and cpuA is stuck talking with it more than one second. Also, this is a similar scenario in an emergency-restart path: - cpuA grabs psinfo->buf_lock and stucks in a firmware - cpuB kicks emergency-restart via either sysrq-b or hangcheck timer. And then, cpuB is deadlocked by taking psinfo->buf_lock again. [Solution] This patch avoids the deadlocking issues in both panic and emergency_restart paths by introducing a function, is_non_blocking_path(), to check if a cpu can be blocked in current path. With this patch, pstore is not blocked even if another cpu has taken a spin_lock, in those paths by changing from spin_lock_irqsave to spin_trylock_irqsave. In addition, according to a comment of emergency_restart() in kernel/sys.c, spin_lock shouldn't be taken in an emergency_restart path to avoid deadlock. This patch fits the comment below. /** * emergency_restart - reboot the system * * Without shutting down any hardware or taking any locks * reboot the system. This is called when we know we are in * trouble so this is our best effort to reboot. This is * safe to call in interrupt context. */ void emergency_restart(void) Signed-off-by: Seiji Aguchi Acked-by: Don Zickus Signed-off-by: Tony Luck --- fs/pstore/platform.c | 35 +++++++++++++++++++++++++++++------ 1 file changed, 29 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c index 5ea2e77ff023..86d1038b5a12 100644 --- a/fs/pstore/platform.c +++ b/fs/pstore/platform.c @@ -96,6 +96,27 @@ static const char *get_reason_str(enum kmsg_dump_reason reason) } } +bool pstore_cannot_block_path(enum kmsg_dump_reason reason) +{ + /* + * In case of NMI path, pstore shouldn't be blocked + * regardless of reason. + */ + if (in_nmi()) + return true; + + switch (reason) { + /* In panic case, other cpus are stopped by smp_send_stop(). */ + case KMSG_DUMP_PANIC: + /* Emergency restart shouldn't be blocked by spin lock. */ + case KMSG_DUMP_EMERG: + return true; + default: + return false; + } +} +EXPORT_SYMBOL_GPL(pstore_cannot_block_path); + /* * callback from kmsg_dump. (s2,l2) has the most recently * written bytes, older bytes are in (s1,l1). Save as much @@ -114,10 +135,12 @@ static void pstore_dump(struct kmsg_dumper *dumper, why = get_reason_str(reason); - if (in_nmi()) { - is_locked = spin_trylock(&psinfo->buf_lock); - if (!is_locked) - pr_err("pstore dump routine blocked in NMI, may corrupt error record\n"); + if (pstore_cannot_block_path(reason)) { + is_locked = spin_trylock_irqsave(&psinfo->buf_lock, flags); + if (!is_locked) { + pr_err("pstore dump routine blocked in %s path, may corrupt error record\n" + , in_nmi() ? "NMI" : why); + } } else spin_lock_irqsave(&psinfo->buf_lock, flags); oopscount++; @@ -143,9 +166,9 @@ static void pstore_dump(struct kmsg_dumper *dumper, total += hsize + len; part++; } - if (in_nmi()) { + if (pstore_cannot_block_path(reason)) { if (is_locked) - spin_unlock(&psinfo->buf_lock); + spin_unlock_irqrestore(&psinfo->buf_lock, flags); } else spin_unlock_irqrestore(&psinfo->buf_lock, flags); } -- cgit From 1b6a78a52262495c26fe8ec58cc8f7a6f8425d7c Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 2 Oct 2012 11:19:19 -0700 Subject: fs/ceph: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: Sage Weil Signed-off-by: Kees Cook Acked-by: Sage Weil --- fs/ceph/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ceph/Kconfig b/fs/ceph/Kconfig index 9eb134ea6eb2..49bc78243db9 100644 --- a/fs/ceph/Kconfig +++ b/fs/ceph/Kconfig @@ -1,6 +1,6 @@ config CEPH_FS - tristate "Ceph distributed file system (EXPERIMENTAL)" - depends on INET && EXPERIMENTAL + tristate "Ceph distributed file system" + depends on INET select CEPH_LIB select LIBCRC32C select CRYPTO_AES -- cgit From 336d6d0323a58702caec260245376843d43f1945 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 2 Oct 2012 11:19:20 -0700 Subject: fs/ecryptfs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: Tyler Hicks CC: Dustin Kirkland Signed-off-by: Kees Cook Acked-by: Tyler Hicks --- fs/ecryptfs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/Kconfig b/fs/ecryptfs/Kconfig index cc16562654de..e15ef38c24fa 100644 --- a/fs/ecryptfs/Kconfig +++ b/fs/ecryptfs/Kconfig @@ -1,6 +1,6 @@ config ECRYPT_FS - tristate "eCrypt filesystem layer support (EXPERIMENTAL)" - depends on EXPERIMENTAL && KEYS && CRYPTO && (ENCRYPTED_KEYS || ENCRYPTED_KEYS=n) + tristate "eCrypt filesystem layer support" + depends on KEYS && CRYPTO && (ENCRYPTED_KEYS || ENCRYPTED_KEYS=n) select CRYPTO_ECB select CRYPTO_CBC select CRYPTO_MD5 -- cgit From f11cb2271f51d1c897a042d7ca4257f5deae6614 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 2 Oct 2012 11:19:26 -0700 Subject: fs/nilfs2: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: KONISHI Ryusuke Signed-off-by: Kees Cook Acked-by: Ryusuke Konishi --- fs/nilfs2/Kconfig | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/Kconfig b/fs/nilfs2/Kconfig index 251da07b2a1d..80da8eb27393 100644 --- a/fs/nilfs2/Kconfig +++ b/fs/nilfs2/Kconfig @@ -1,6 +1,5 @@ config NILFS2_FS - tristate "NILFS2 file system support (EXPERIMENTAL)" - depends on EXPERIMENTAL + tristate "NILFS2 file system support" select CRC32 help NILFS2 is a log-structured file system (LFS) supporting continuous -- cgit From d9777b8de415f2e699dcc4ce8825732ba8af732a Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 2 Oct 2012 11:19:27 -0700 Subject: fs/xfs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: Ben Myers CC: Alex Elder Signed-off-by: Kees Cook Acked-by: Ben Myers --- fs/xfs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/Kconfig b/fs/xfs/Kconfig index 5a7ffe54f5d5..cc33aaf219f1 100644 --- a/fs/xfs/Kconfig +++ b/fs/xfs/Kconfig @@ -70,8 +70,8 @@ config XFS_RT If unsure, say N. config XFS_DEBUG - bool "XFS Debugging support (EXPERIMENTAL)" - depends on XFS_FS && EXPERIMENTAL + bool "XFS Debugging support" + depends on XFS_FS help Say Y here to get an XFS build with many debugging features, including ASSERT checks, function wrappers around macros, -- cgit From 10ee27a06cc8eb57f83342a8eabcb75deb872d52 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 10 Jan 2013 13:47:57 +0800 Subject: vfs: re-implement writeback_inodes_sb(_nr)_if_idle() and rename them writeback_inodes_sb(_nr)_if_idle() is re-implemented by replacing down_read() with down_read_trylock() because - If ->s_umount is write locked, then the sb is not idle. That is writeback_inodes_sb(_nr)_if_idle() needn't wait for the lock. - writeback_inodes_sb(_nr)_if_idle() grabs s_umount lock when it want to start writeback, it may bring us deadlock problem when doing umount. In order to fix the problem, ext4 and btrfs implemented their own writeback functions instead of writeback_inodes_sb(_nr)_if_idle(), but it introduced the redundant code, it is better to implement a new writeback_inodes_sb(_nr)_if_idle(). The name of these two functions is cumbersome, so rename them to try_to_writeback_inodes_sb(_nr). This idea came from Christoph Hellwig. Some code is from the patch of Kamal Mostafa. Reviewed-by: Jan Kara Signed-off-by: Miao Xie Signed-off-by: Fengguang Wu --- fs/btrfs/extent-tree.c | 20 +++----------------- fs/ext4/inode.c | 8 ++------ fs/fs-writeback.c | 44 ++++++++++++++++++++------------------------ 3 files changed, 25 insertions(+), 47 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 521e9d4424f6..f31abb14e06f 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3689,20 +3689,6 @@ static int can_overcommit(struct btrfs_root *root, return 0; } -static int writeback_inodes_sb_nr_if_idle_safe(struct super_block *sb, - unsigned long nr_pages, - enum wb_reason reason) -{ - if (!writeback_in_progress(sb->s_bdi) && - down_read_trylock(&sb->s_umount)) { - writeback_inodes_sb_nr(sb, nr_pages, reason); - up_read(&sb->s_umount); - return 1; - } - - return 0; -} - /* * shrink metadata reservation for delalloc */ @@ -3735,9 +3721,9 @@ static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig, while (delalloc_bytes && loops < 3) { max_reclaim = min(delalloc_bytes, to_reclaim); nr_pages = max_reclaim >> PAGE_CACHE_SHIFT; - writeback_inodes_sb_nr_if_idle_safe(root->fs_info->sb, - nr_pages, - WB_REASON_FS_FREE_SPACE); + try_to_writeback_inodes_sb_nr(root->fs_info->sb, + nr_pages, + WB_REASON_FS_FREE_SPACE); /* * We need to wait for the async pages to actually start before diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index cbfe13bf5b2a..5f6eef71ff21 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2512,12 +2512,8 @@ static int ext4_nonda_switch(struct super_block *sb) /* * Start pushing delalloc when 1/2 of free blocks are dirty. */ - if (dirty_blocks && (free_blocks < 2 * dirty_blocks) && - !writeback_in_progress(sb->s_bdi) && - down_read_trylock(&sb->s_umount)) { - writeback_inodes_sb(sb, WB_REASON_FS_FREE_SPACE); - up_read(&sb->s_umount); - } + if (dirty_blocks && (free_blocks < 2 * dirty_blocks)) + try_to_writeback_inodes_sb(sb, WB_REASON_FS_FREE_SPACE); if (2 * free_blocks < 3 * dirty_blocks || free_blocks < (dirty_blocks + EXT4_FREECLUSTERS_WATERMARK)) { diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 310972b72a66..ad3cc46a743a 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -1332,47 +1332,43 @@ void writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) EXPORT_SYMBOL(writeback_inodes_sb); /** - * writeback_inodes_sb_if_idle - start writeback if none underway + * try_to_writeback_inodes_sb_nr - try to start writeback if none underway * @sb: the superblock - * @reason: reason why some writeback work was initiated + * @nr: the number of pages to write + * @reason: the reason of writeback * - * Invoke writeback_inodes_sb if no writeback is currently underway. + * Invoke writeback_inodes_sb_nr if no writeback is currently underway. * Returns 1 if writeback was started, 0 if not. */ -int writeback_inodes_sb_if_idle(struct super_block *sb, enum wb_reason reason) +int try_to_writeback_inodes_sb_nr(struct super_block *sb, + unsigned long nr, + enum wb_reason reason) { - if (!writeback_in_progress(sb->s_bdi)) { - down_read(&sb->s_umount); - writeback_inodes_sb(sb, reason); - up_read(&sb->s_umount); + if (writeback_in_progress(sb->s_bdi)) return 1; - } else + + if (!down_read_trylock(&sb->s_umount)) return 0; + + writeback_inodes_sb_nr(sb, nr, reason); + up_read(&sb->s_umount); + return 1; } -EXPORT_SYMBOL(writeback_inodes_sb_if_idle); +EXPORT_SYMBOL(try_to_writeback_inodes_sb_nr); /** - * writeback_inodes_sb_nr_if_idle - start writeback if none underway + * try_to_writeback_inodes_sb - try to start writeback if none underway * @sb: the superblock - * @nr: the number of pages to write * @reason: reason why some writeback work was initiated * - * Invoke writeback_inodes_sb if no writeback is currently underway. + * Implement by try_to_writeback_inodes_sb_nr() * Returns 1 if writeback was started, 0 if not. */ -int writeback_inodes_sb_nr_if_idle(struct super_block *sb, - unsigned long nr, - enum wb_reason reason) +int try_to_writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) { - if (!writeback_in_progress(sb->s_bdi)) { - down_read(&sb->s_umount); - writeback_inodes_sb_nr(sb, nr, reason); - up_read(&sb->s_umount); - return 1; - } else - return 0; + return try_to_writeback_inodes_sb_nr(sb, get_nr_dirty_pages(), reason); } -EXPORT_SYMBOL(writeback_inodes_sb_nr_if_idle); +EXPORT_SYMBOL(try_to_writeback_inodes_sb); /** * sync_inodes_sb - sync sb inode pages -- cgit From 860d21e2c585f7ee8a4ecc06f474fdc33c9474f4 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 12 Jan 2013 16:19:36 -0500 Subject: ext4: return ENOMEM if sb_getblk() fails The only reason for sb_getblk() failing is if it can't allocate the buffer_head. So ENOMEM is more appropriate than EIO. In addition, make sure that the file system is marked as being inconsistent if sb_getblk() fails. Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/ext4/extents.c | 25 ++++++++++++++----------- fs/ext4/indirect.c | 9 ++++++--- fs/ext4/inline.c | 2 +- fs/ext4/inode.c | 9 +++------ fs/ext4/mmp.c | 2 ++ fs/ext4/resize.c | 8 ++++---- fs/ext4/xattr.c | 3 ++- 7 files changed, 32 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 5ae1674ec12f..d42a8c49ad69 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -725,6 +725,7 @@ ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block, struct ext4_extent_header *eh; struct buffer_head *bh; short int depth, i, ppos = 0, alloc = 0; + int ret; eh = ext_inode_hdr(inode); depth = ext_depth(inode); @@ -752,12 +753,15 @@ ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block, path[ppos].p_ext = NULL; bh = sb_getblk(inode->i_sb, path[ppos].p_block); - if (unlikely(!bh)) + if (unlikely(!bh)) { + ret = -ENOMEM; goto err; + } if (!bh_uptodate_or_lock(bh)) { trace_ext4_ext_load_extent(inode, block, path[ppos].p_block); - if (bh_submit_read(bh) < 0) { + ret = bh_submit_read(bh); + if (ret < 0) { put_bh(bh); goto err; } @@ -768,13 +772,15 @@ ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block, put_bh(bh); EXT4_ERROR_INODE(inode, "ppos %d > depth %d", ppos, depth); + ret = -EIO; goto err; } path[ppos].p_bh = bh; path[ppos].p_hdr = eh; i--; - if (ext4_ext_check_block(inode, eh, i, bh)) + ret = ext4_ext_check_block(inode, eh, i, bh); + if (ret < 0) goto err; } @@ -796,7 +802,7 @@ err: ext4_ext_drop_refs(path); if (alloc) kfree(path); - return ERR_PTR(-EIO); + return ERR_PTR(ret); } /* @@ -951,7 +957,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, } bh = sb_getblk(inode->i_sb, newblock); if (!bh) { - err = -EIO; + err = -ENOMEM; goto cleanup; } lock_buffer(bh); @@ -1024,7 +1030,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, newblock = ablocks[--a]; bh = sb_getblk(inode->i_sb, newblock); if (!bh) { - err = -EIO; + err = -ENOMEM; goto cleanup; } lock_buffer(bh); @@ -1136,11 +1142,8 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, return err; bh = sb_getblk(inode->i_sb, newblock); - if (!bh) { - err = -EIO; - ext4_std_error(inode->i_sb, err); - return err; - } + if (!bh) + return -ENOMEM; lock_buffer(bh); err = ext4_journal_get_create_access(handle, bh); diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 20862f96e8ae..8d83d1e508e4 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -146,6 +146,7 @@ static Indirect *ext4_get_branch(struct inode *inode, int depth, struct super_block *sb = inode->i_sb; Indirect *p = chain; struct buffer_head *bh; + int ret = -EIO; *err = 0; /* i_data is not going away, no lock needed */ @@ -154,8 +155,10 @@ static Indirect *ext4_get_branch(struct inode *inode, int depth, goto no_block; while (--depth) { bh = sb_getblk(sb, le32_to_cpu(p->key)); - if (unlikely(!bh)) + if (unlikely(!bh)) { + ret = -ENOMEM; goto failure; + } if (!bh_uptodate_or_lock(bh)) { if (bh_submit_read(bh) < 0) { @@ -177,7 +180,7 @@ static Indirect *ext4_get_branch(struct inode *inode, int depth, return NULL; failure: - *err = -EIO; + *err = ret; no_block: return p; } @@ -471,7 +474,7 @@ static int ext4_alloc_branch(handle_t *handle, struct inode *inode, */ bh = sb_getblk(inode->i_sb, new_blocks[n-1]); if (unlikely(!bh)) { - err = -EIO; + err = -ENOMEM; goto failed; } diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index 387c47c6cda9..93a3408fc89b 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -1188,7 +1188,7 @@ static int ext4_convert_inline_data_nolock(handle_t *handle, data_bh = sb_getblk(inode->i_sb, map.m_pblk); if (!data_bh) { - error = -EIO; + error = -ENOMEM; goto out_restore; } diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index cbfe13bf5b2a..9ccc140b82d2 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -714,7 +714,7 @@ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode, bh = sb_getblk(inode->i_sb, map.m_pblk); if (!bh) { - *errp = -EIO; + *errp = -ENOMEM; return NULL; } if (map.m_flags & EXT4_MAP_NEW) { @@ -3660,11 +3660,8 @@ static int __ext4_get_inode_loc(struct inode *inode, iloc->offset = (inode_offset % inodes_per_block) * EXT4_INODE_SIZE(sb); bh = sb_getblk(sb, block); - if (!bh) { - EXT4_ERROR_INODE_BLOCK(inode, block, - "unable to read itable block"); - return -EIO; - } + if (!bh) + return -ENOMEM; if (!buffer_uptodate(bh)) { lock_buffer(bh); diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c index fe7c63f4717e..44734f1ca554 100644 --- a/fs/ext4/mmp.c +++ b/fs/ext4/mmp.c @@ -80,6 +80,8 @@ static int read_mmp_block(struct super_block *sb, struct buffer_head **bh, * is not blocked in the elevator. */ if (!*bh) *bh = sb_getblk(sb, mmp_block); + if (!*bh) + return -ENOMEM; if (*bh) { get_bh(*bh); lock_buffer(*bh); diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index d99387b89edd..02824dc2ff3b 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -334,7 +334,7 @@ static struct buffer_head *bclean(handle_t *handle, struct super_block *sb, bh = sb_getblk(sb, blk); if (!bh) - return ERR_PTR(-EIO); + return ERR_PTR(-ENOMEM); if ((err = ext4_journal_get_write_access(handle, bh))) { brelse(bh); bh = ERR_PTR(err); @@ -411,7 +411,7 @@ static int set_flexbg_block_bitmap(struct super_block *sb, handle_t *handle, bh = sb_getblk(sb, flex_gd->groups[group].block_bitmap); if (!bh) - return -EIO; + return -ENOMEM; err = ext4_journal_get_write_access(handle, bh); if (err) @@ -501,7 +501,7 @@ static int setup_new_flex_group_blocks(struct super_block *sb, gdb = sb_getblk(sb, block); if (!gdb) { - err = -EIO; + err = -ENOMEM; goto out; } @@ -1065,7 +1065,7 @@ static void update_backups(struct super_block *sb, int blk_off, char *data, bh = sb_getblk(sb, backup_block); if (!bh) { - err = -EIO; + err = -ENOMEM; break; } ext4_debug("update metadata backup %llu(+%llu)\n", diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 3a91ebc2b66f..07d684a4e523 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -887,16 +887,17 @@ inserted: new_bh = sb_getblk(sb, block); if (!new_bh) { + error = -ENOMEM; getblk_failed: ext4_free_blocks(handle, inode, NULL, block, 1, EXT4_FREE_BLOCKS_METADATA); - error = -EIO; goto cleanup; } lock_buffer(new_bh); error = ext4_journal_get_create_access(handle, new_bh); if (error) { unlock_buffer(new_bh); + error = -EIO; goto getblk_failed; } memcpy(new_bh->b_data, s->base, new_bh->b_size); -- cgit From aebf02430d25b6bd2b8542126fdcdb90e75a24b8 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Sat, 12 Jan 2013 16:28:47 -0500 Subject: ext4: use unlikely to improve the efficiency of the kernel Because the function 'sb_getblk' seldomly fails to return NULL value,it will be better to use 'unlikely' to optimize it. Signed-off-by: Wang Shilong Signed-off-by: "Theodore Ts'o" --- fs/ext4/extents.c | 6 +++--- fs/ext4/inode.c | 6 +++--- fs/ext4/mmp.c | 2 +- fs/ext4/resize.c | 10 +++++----- fs/ext4/xattr.c | 2 +- 5 files changed, 13 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index d42a8c49ad69..391e53a52e58 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -956,7 +956,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, goto cleanup; } bh = sb_getblk(inode->i_sb, newblock); - if (!bh) { + if (unlikely(!bh)) { err = -ENOMEM; goto cleanup; } @@ -1029,7 +1029,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, oldblock = newblock; newblock = ablocks[--a]; bh = sb_getblk(inode->i_sb, newblock); - if (!bh) { + if (unlikely(!bh)) { err = -ENOMEM; goto cleanup; } @@ -1142,7 +1142,7 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, return err; bh = sb_getblk(inode->i_sb, newblock); - if (!bh) + if (unlikely(!bh)) return -ENOMEM; lock_buffer(bh); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 9ccc140b82d2..93a7e8453a68 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -713,7 +713,7 @@ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode, return NULL; bh = sb_getblk(inode->i_sb, map.m_pblk); - if (!bh) { + if (unlikely(!bh)) { *errp = -ENOMEM; return NULL; } @@ -3660,7 +3660,7 @@ static int __ext4_get_inode_loc(struct inode *inode, iloc->offset = (inode_offset % inodes_per_block) * EXT4_INODE_SIZE(sb); bh = sb_getblk(sb, block); - if (!bh) + if (unlikely(!bh)) return -ENOMEM; if (!buffer_uptodate(bh)) { lock_buffer(bh); @@ -3693,7 +3693,7 @@ static int __ext4_get_inode_loc(struct inode *inode, /* Is the inode bitmap in cache? */ bitmap_bh = sb_getblk(sb, ext4_inode_bitmap(sb, gdp)); - if (!bitmap_bh) + if (unlikely(!bitmap_bh)) goto make_io; /* diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c index 44734f1ca554..f9b551561d2c 100644 --- a/fs/ext4/mmp.c +++ b/fs/ext4/mmp.c @@ -93,7 +93,7 @@ static int read_mmp_block(struct super_block *sb, struct buffer_head **bh, *bh = NULL; } } - if (!*bh) { + if (unlikely(!*bh)) { ext4_warning(sb, "Error while reading MMP block %llu", mmp_block); return -EIO; diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index 02824dc2ff3b..05f8d4502d42 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -333,7 +333,7 @@ static struct buffer_head *bclean(handle_t *handle, struct super_block *sb, int err; bh = sb_getblk(sb, blk); - if (!bh) + if (unlikely(!bh)) return ERR_PTR(-ENOMEM); if ((err = ext4_journal_get_write_access(handle, bh))) { brelse(bh); @@ -410,7 +410,7 @@ static int set_flexbg_block_bitmap(struct super_block *sb, handle_t *handle, return err; bh = sb_getblk(sb, flex_gd->groups[group].block_bitmap); - if (!bh) + if (unlikely(!bh)) return -ENOMEM; err = ext4_journal_get_write_access(handle, bh); @@ -500,7 +500,7 @@ static int setup_new_flex_group_blocks(struct super_block *sb, goto out; gdb = sb_getblk(sb, block); - if (!gdb) { + if (unlikely(!gdb)) { err = -ENOMEM; goto out; } @@ -1064,7 +1064,7 @@ static void update_backups(struct super_block *sb, int blk_off, char *data, ext4_bg_has_super(sb, group)); bh = sb_getblk(sb, backup_block); - if (!bh) { + if (unlikely(!bh)) { err = -ENOMEM; break; } @@ -1168,7 +1168,7 @@ static int ext4_add_new_descs(handle_t *handle, struct super_block *sb, static struct buffer_head *ext4_get_bitmap(struct super_block *sb, __u64 block) { struct buffer_head *bh = sb_getblk(sb, block); - if (!bh) + if (unlikely(!bh)) return NULL; if (!bh_uptodate_or_lock(bh)) { if (bh_submit_read(bh) < 0) { diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 07d684a4e523..c68990c392c7 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -886,7 +886,7 @@ inserted: (unsigned long long)block); new_bh = sb_getblk(sb, block); - if (!new_bh) { + if (unlikely(!new_bh)) { error = -ENOMEM; getblk_failed: ext4_free_blocks(handle, inode, NULL, block, 1, -- cgit From 15b49132fc972c63894592f218ea5a9a61b1a18f Mon Sep 17 00:00:00 2001 From: Eryu Guan Date: Sat, 12 Jan 2013 16:33:25 -0500 Subject: ext4: check bh in ext4_read_block_bitmap() Validate the bh pointer before using it, since ext4_read_block_bitmap_nowait() might return NULL. I've seen this in fsfuzz testing. EXT4-fs error (device loop0): ext4_read_block_bitmap_nowait:385: comm touch: Cannot get buffer for block bitmap - block_group = 0, block_bitmap = 3925999616 BUG: unable to handle kernel NULL pointer dereference at (null) IP: [] ext4_wait_block_bitmap+0x25/0xe0 ... Call Trace: [] ext4_read_block_bitmap+0x35/0x60 [] ext4_free_blocks+0x236/0xb80 [] ? __getblk+0x36/0x70 [] ? __find_get_block+0x8f/0x210 [] ? kmem_cache_free+0x33/0x140 [] ext4_xattr_release_block+0x1b5/0x1d0 [] ext4_xattr_delete_inode+0xbe/0x100 [] ext4_free_inode+0x7c/0x4d0 [] ? ext4_mark_inode_dirty+0x88/0x230 [] ext4_evict_inode+0x32c/0x490 [] evict+0xa7/0x1c0 [] iput_final+0xe3/0x170 [] iput+0x3e/0x50 [] ext4_add_nondir+0x4d/0x90 [] ext4_create+0xeb/0x170 [] vfs_create+0xac/0xd0 [] lookup_open+0x185/0x1c0 [] ? selinux_inode_permission+0xa9/0x170 [] do_last+0x2d4/0x7a0 [] path_openat+0xb3/0x480 [] ? handle_mm_fault+0x251/0x3b0 [] do_filp_open+0x49/0xa0 [] ? __alloc_fd+0xdd/0x150 [] do_sys_open+0x108/0x1f0 [] sys_open+0x21/0x30 [] system_call_fastpath+0x16/0x1b Also fix comment for ext4_read_block_bitmap_nowait() Signed-off-by: Eryu Guan Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/ext4/balloc.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index cf1821784a16..33938c120c85 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -358,7 +358,7 @@ void ext4_validate_block_bitmap(struct super_block *sb, } /** - * ext4_read_block_bitmap() + * ext4_read_block_bitmap_nowait() * @sb: super block * @block_group: given block group * @@ -457,6 +457,8 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group) struct buffer_head *bh; bh = ext4_read_block_bitmap_nowait(sb, block_group); + if (!bh) + return NULL; if (ext4_wait_block_bitmap(sb, block_group, bh)) { put_bh(bh); return NULL; -- cgit From 7f5118629f74b82bd4ba5e47415d1b4dcb940241 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sun, 13 Jan 2013 08:41:45 -0500 Subject: ext4: trigger the lazy inode table initialization after resize After we have finished extending the file system, we need to trigger a the lazy inode table thread to zero out the inode tables. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 2 ++ fs/ext4/ioctl.c | 9 +++++++++ fs/ext4/resize.c | 8 +++++--- fs/ext4/super.c | 21 +++++++++++---------- 4 files changed, 27 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 8462eb3c33aa..80246237f6d5 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2227,6 +2227,8 @@ extern int ext4_group_desc_csum_verify(struct super_block *sb, __u32 group, struct ext4_group_desc *gdp); extern void ext4_group_desc_csum_set(struct super_block *sb, __u32 group, struct ext4_group_desc *gdp); +extern int ext4_register_li_request(struct super_block *sb, + ext4_group_t first_not_zeroed); static inline int ext4_has_group_desc_csum(struct super_block *sb) { diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 5747f52f7c72..4784ac244fc6 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -313,6 +313,9 @@ mext_out: if (err == 0) err = err2; mnt_drop_write_file(filp); + if (!err && ext4_has_group_desc_csum(sb) && + test_opt(sb, INIT_INODE_TABLE)) + err = ext4_register_li_request(sb, input.group); group_add_out: ext4_resize_end(sb); return err; @@ -358,6 +361,7 @@ group_add_out: ext4_fsblk_t n_blocks_count; struct super_block *sb = inode->i_sb; int err = 0, err2 = 0; + ext4_group_t o_group = EXT4_SB(sb)->s_groups_count; if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_BIGALLOC)) { @@ -388,6 +392,11 @@ group_add_out: if (err == 0) err = err2; mnt_drop_write_file(filp); + if (!err && (o_group > EXT4_SB(sb)->s_groups_count) && + ext4_has_group_desc_csum(sb) && + test_opt(sb, INIT_INODE_TABLE)) + err = ext4_register_li_request(sb, o_group); + resizefs_out: ext4_resize_end(sb); return err; diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index 05f8d4502d42..8eefb636beb8 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -1506,10 +1506,12 @@ static int ext4_setup_next_flex_gd(struct super_block *sb, group_data[i].blocks_count = blocks_per_group; overhead = ext4_group_overhead_blocks(sb, group + i); group_data[i].free_blocks_count = blocks_per_group - overhead; - if (ext4_has_group_desc_csum(sb)) + if (ext4_has_group_desc_csum(sb)) { flex_gd->bg_flags[i] = EXT4_BG_BLOCK_UNINIT | EXT4_BG_INODE_UNINIT; - else + if (!test_opt(sb, INIT_INODE_TABLE)) + flex_gd->bg_flags[i] |= EXT4_BG_INODE_ZEROED; + } else flex_gd->bg_flags[i] = EXT4_BG_INODE_ZEROED; } @@ -1594,7 +1596,7 @@ int ext4_group_add(struct super_block *sb, struct ext4_new_group_data *input) err = ext4_alloc_flex_bg_array(sb, input->group + 1); if (err) - return err; + goto out; err = ext4_mb_alloc_groupinfo(sb, input->group + 1); if (err) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 3d4fb81bacd5..c014edd12648 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -2776,7 +2776,7 @@ static int ext4_run_li_request(struct ext4_li_request *elr) break; } - if (group == ngroups) + if (group >= ngroups) ret = 1; if (!ret) { @@ -3016,33 +3016,34 @@ static struct ext4_li_request *ext4_li_request_new(struct super_block *sb, return elr; } -static int ext4_register_li_request(struct super_block *sb, - ext4_group_t first_not_zeroed) +int ext4_register_li_request(struct super_block *sb, + ext4_group_t first_not_zeroed) { struct ext4_sb_info *sbi = EXT4_SB(sb); - struct ext4_li_request *elr; + struct ext4_li_request *elr = NULL; ext4_group_t ngroups = EXT4_SB(sb)->s_groups_count; int ret = 0; + mutex_lock(&ext4_li_mtx); if (sbi->s_li_request != NULL) { /* * Reset timeout so it can be computed again, because * s_li_wait_mult might have changed. */ sbi->s_li_request->lr_timeout = 0; - return 0; + goto out; } if (first_not_zeroed == ngroups || (sb->s_flags & MS_RDONLY) || !test_opt(sb, INIT_INODE_TABLE)) - return 0; + goto out; elr = ext4_li_request_new(sb, first_not_zeroed); - if (!elr) - return -ENOMEM; - - mutex_lock(&ext4_li_mtx); + if (!elr) { + ret = -ENOMEM; + goto out; + } if (NULL == ext4_li_info) { ret = ext4_li_info_new(); -- cgit From d4608632ec8f4ae3ffecdd343ede34e60eabc64f Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 10 Jan 2013 10:41:48 -0600 Subject: xfs: recalculate leaf entry pointer after compacting a dir2 block Dave Jones hit this assert when doing a compile on recent git, with CONFIG_XFS_DEBUG enabled: XFS: Assertion failed: (char *)dup - (char *)hdr == be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)), file: fs/xfs/xfs_dir2_data.c, line: 828 Upon further digging, the tag found by xfs_dir2_data_unused_tag_p(dup) contained "2" and not the proper offset, and I found that this value was changed after the memmoves under "Use a stale leaf for our new entry." in xfs_dir2_block_addname(), i.e. memmove(&blp[mid + 1], &blp[mid], (highstale - mid) * sizeof(*blp)); overwrote it. What has happened is that the previous call to xfs_dir2_block_compact() has rearranged things; it changes btp->count as well as the blp array. So after we make that call, we must recalculate the proper pointer to the leaf entries by making another call to xfs_dir2_block_leaf_p(). Dave provided a metadump image which led to a simple reproducer (create a particular filename in the affected directory) and this resolves the testcase as well as the bug on his live system. Thanks also to dchinner for looking at this one with me. Signed-off-by: Eric Sandeen Tested-by: Dave Jones Reviewed-by: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_dir2_block.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c index 7536faaa61e7..12afe07a91d7 100644 --- a/fs/xfs/xfs_dir2_block.c +++ b/fs/xfs/xfs_dir2_block.c @@ -355,10 +355,12 @@ xfs_dir2_block_addname( /* * If need to compact the leaf entries, do it now. */ - if (compact) + if (compact) { xfs_dir2_block_compact(tp, bp, hdr, btp, blp, &needlog, &lfloghigh, &lfloglow); - else if (btp->stale) { + /* recalculate blp post-compaction */ + blp = xfs_dir2_block_leaf_p(btp); + } else if (btp->stale) { /* * Set leaf logging boundaries to impossible state. * For the no-stale case they're set explicitly. -- cgit From a17164e54bf0e3c2cbc72c35b9f67c2873a122dd Mon Sep 17 00:00:00 2001 From: Abhijit Pawar Date: Wed, 9 Jan 2013 19:34:42 +0530 Subject: fs/xfs remove obsolete simple_strto This patch replaces usages of obsolete simple_strtoul with kstrtoint in xfs_args and suffix_strtoul. Signed-off-by: Abhijit Pawar Reviewed-by: Jie Liu Signed-off-by: Ben Myers --- fs/xfs/xfs_super.c | 29 +++++++++++++++++++---------- 1 file changed, 19 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index ab8839b26272..c407121873b4 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -139,9 +139,9 @@ static const match_table_t tokens = { STATIC unsigned long -suffix_strtoul(char *s, char **endp, unsigned int base) +suffix_kstrtoint(char *s, unsigned int base, int *res) { - int last, shift_left_factor = 0; + int last, shift_left_factor = 0, _res; char *value = s; last = strlen(value) - 1; @@ -158,7 +158,10 @@ suffix_strtoul(char *s, char **endp, unsigned int base) value[last] = '\0'; } - return simple_strtoul((const char *)s, endp, base) << shift_left_factor; + if (kstrtoint(s, base, &_res)) + return -EINVAL; + *res = _res << shift_left_factor; + return 0; } /* @@ -174,7 +177,7 @@ xfs_parseargs( char *options) { struct super_block *sb = mp->m_super; - char *this_char, *value, *eov; + char *this_char, *value; int dsunit = 0; int dswidth = 0; int iosize = 0; @@ -230,14 +233,16 @@ xfs_parseargs( this_char); return EINVAL; } - mp->m_logbufs = simple_strtoul(value, &eov, 10); + if (kstrtoint(value, 10, &mp->m_logbufs)) + return EINVAL; } else if (!strcmp(this_char, MNTOPT_LOGBSIZE)) { if (!value || !*value) { xfs_warn(mp, "%s option requires an argument", this_char); return EINVAL; } - mp->m_logbsize = suffix_strtoul(value, &eov, 10); + if (suffix_kstrtoint(value, 10, &mp->m_logbsize)) + return EINVAL; } else if (!strcmp(this_char, MNTOPT_LOGDEV)) { if (!value || !*value) { xfs_warn(mp, "%s option requires an argument", @@ -266,7 +271,8 @@ xfs_parseargs( this_char); return EINVAL; } - iosize = simple_strtoul(value, &eov, 10); + if (kstrtoint(value, 10, &iosize)) + return EINVAL; iosizelog = ffs(iosize) - 1; } else if (!strcmp(this_char, MNTOPT_ALLOCSIZE)) { if (!value || !*value) { @@ -274,7 +280,8 @@ xfs_parseargs( this_char); return EINVAL; } - iosize = suffix_strtoul(value, &eov, 10); + if (suffix_kstrtoint(value, 10, &iosize)) + return EINVAL; iosizelog = ffs(iosize) - 1; } else if (!strcmp(this_char, MNTOPT_GRPID) || !strcmp(this_char, MNTOPT_BSDGROUPS)) { @@ -296,14 +303,16 @@ xfs_parseargs( this_char); return EINVAL; } - dsunit = simple_strtoul(value, &eov, 10); + if (kstrtoint(value, 10, &dsunit)) + return EINVAL; } else if (!strcmp(this_char, MNTOPT_SWIDTH)) { if (!value || !*value) { xfs_warn(mp, "%s option requires an argument", this_char); return EINVAL; } - dswidth = simple_strtoul(value, &eov, 10); + if (kstrtoint(value, 10, &dswidth)) + return EINVAL; } else if (!strcmp(this_char, MNTOPT_32BITINODE)) { mp->m_flags |= XFS_MOUNT_SMALL_INUMS; } else if (!strcmp(this_char, MNTOPT_64BITINODE)) { -- cgit From 3a366e614d0837d9fc23f78cdb1a1186ebc3387f Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 11 Jan 2013 13:06:33 -0800 Subject: block: add missing block_bio_complete() tracepoint bio completion didn't kick block_bio_complete TP. Only dm was explicitly triggering the TP on IO completion. This makes block_bio_complete TP useless for tracers which want to know about bios, and all other bio based drivers skip generating blktrace completion events. This patch makes all bio completions via bio_endio() generate block_bio_complete TP. * Explicit trace_block_bio_complete() invocation removed from dm and the trace point is unexported. * @rq dropped from trace_block_bio_complete(). bios may fly around w/o queue associated. Verifying and accessing the assocaited queue belongs to TP probes. * blktrace now gets both request and bio completions. Make it ignore bio completions if request completion path is happening. This makes all bio based drivers generate blktrace completion events properly and makes the block_bio_complete TP actually useful. v2: With this change, block_bio_complete TP could be invoked on sg commands which have bio's with %NULL bi_bdev. Update TP assignment code to check whether bio->bi_bdev is %NULL before dereferencing. Signed-off-by: Tejun Heo Original-patch-by: Namhyung Kim Cc: Tejun Heo Cc: Steven Rostedt Cc: Alasdair Kergon Cc: dm-devel@redhat.com Cc: Neil Brown Signed-off-by: Jens Axboe --- fs/bio.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/bio.c b/fs/bio.c index b96fc6ce4855..bb5768f59b32 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -1428,6 +1428,8 @@ void bio_endio(struct bio *bio, int error) else if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) error = -EIO; + trace_block_bio_complete(bio, error); + if (bio->bi_end_io) bio->bi_end_io(bio, error); } -- cgit From f0059afd3e6e7aa1a0ffc23468b74c43d47660b8 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 11 Jan 2013 13:06:35 -0800 Subject: buffer: make touch_buffer() an exported function We want to add a trace point to touch_buffer() but macros and inline functions defined in header files can't have tracing points. Move touch_buffer() to fs/buffer.c and make it a proper function. The new exported function is also declared inline. As most uses of touch_buffer() are inside buffer.c with nilfs2 as the only other user, the effect of this change should be negligible. Signed-off-by: Tejun Heo Cc: Steven Rostedt Signed-off-by: Jens Axboe --- fs/buffer.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index c017a2dfb909..a8c2dfb68dcd 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -53,6 +53,12 @@ void init_buffer(struct buffer_head *bh, bh_end_io_t *handler, void *private) } EXPORT_SYMBOL(init_buffer); +inline void touch_buffer(struct buffer_head *bh) +{ + mark_page_accessed(bh->b_page); +} +EXPORT_SYMBOL(touch_buffer); + static int sleep_on_buffer(void *word) { io_schedule(); -- cgit From 5305cb830834549b9203ad4d009ad5483c5e293f Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 11 Jan 2013 13:06:36 -0800 Subject: block: add block_{touch|dirty}_buffer tracepoint The former is triggered from touch_buffer() and the latter mark_buffer_dirty(). This is part of tracepoint additions to improve visiblity into dirtying / writeback operations for io tracer and userland. v2: Transformed writeback_dirty_buffer to block_dirty_buffer and made it share TP definition with block_touch_buffer. Signed-off-by: Tejun Heo Cc: Fengguang Wu Signed-off-by: Jens Axboe --- fs/buffer.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index a8c2dfb68dcd..87ff335fbbe3 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -41,6 +41,7 @@ #include #include #include +#include static int fsync_buffers_list(spinlock_t *lock, struct list_head *list); @@ -55,6 +56,7 @@ EXPORT_SYMBOL(init_buffer); inline void touch_buffer(struct buffer_head *bh) { + trace_block_touch_buffer(bh); mark_page_accessed(bh->b_page); } EXPORT_SYMBOL(touch_buffer); @@ -1119,6 +1121,8 @@ void mark_buffer_dirty(struct buffer_head *bh) { WARN_ON_ONCE(!buffer_uptodate(bh)); + trace_block_dirty_buffer(bh); + /* * Very *carefully* optimize the it-is-already-dirty case. * -- cgit From 9fb0a7da0c528d9bd49b597aa63b1fe2216c7203 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 11 Jan 2013 13:06:37 -0800 Subject: writeback: add more tracepoints Add tracepoints for page dirtying, writeback_single_inode start, inode dirtying and writeback. For the latter two inode events, a pair of events are defined to denote start and end of the operations (the starting one has _start suffix and the one w/o suffix happens after the operation is complete). These inode ops are FS specific and can be non-trivial and having enclosing tracepoints is useful for external tracers. This is part of tracepoint additions to improve visiblity into dirtying / writeback operations for io tracer and userland. v2: writeback_dirty_inode[_start] TPs may be called for files on pseudo FSes w/ unregistered bdi. Check whether bdi->dev is %NULL before dereferencing. v3: buffer dirtying moved to a block TP. Signed-off-by: Tejun Heo Reviewed-by: Jan Kara Signed-off-by: Jens Axboe --- fs/fs-writeback.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 310972b72a66..359494ea1bde 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -318,8 +318,14 @@ static void queue_io(struct bdi_writeback *wb, struct wb_writeback_work *work) static int write_inode(struct inode *inode, struct writeback_control *wbc) { - if (inode->i_sb->s_op->write_inode && !is_bad_inode(inode)) - return inode->i_sb->s_op->write_inode(inode, wbc); + int ret; + + if (inode->i_sb->s_op->write_inode && !is_bad_inode(inode)) { + trace_writeback_write_inode_start(inode, wbc); + ret = inode->i_sb->s_op->write_inode(inode, wbc); + trace_writeback_write_inode(inode, wbc); + return ret; + } return 0; } @@ -450,6 +456,8 @@ __writeback_single_inode(struct inode *inode, struct writeback_control *wbc) WARN_ON(!(inode->i_state & I_SYNC)); + trace_writeback_single_inode_start(inode, wbc, nr_to_write); + ret = do_writepages(mapping, wbc); /* @@ -1150,8 +1158,12 @@ void __mark_inode_dirty(struct inode *inode, int flags) * dirty the inode itself */ if (flags & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) { + trace_writeback_dirty_inode_start(inode, flags); + if (sb->s_op->dirty_inode) sb->s_op->dirty_inode(inode, flags); + + trace_writeback_dirty_inode(inode, flags); } /* -- cgit From 3d251a5b9e2f09edcf25bbffe1fa308d0f648bf1 Mon Sep 17 00:00:00 2001 From: Akinobu Mita Date: Thu, 3 Jan 2013 21:19:09 +0900 Subject: UBIFS: rename random32() to prandom_u32() Use more preferable function name which implies using a pseudo-random number generator. Signed-off-by: Akinobu Mita Signed-off-by: Artem Bityutskiy --- fs/ubifs/debug.c | 8 ++++---- fs/ubifs/lpt_commit.c | 14 +++++++------- fs/ubifs/tnc_commit.c | 2 +- 3 files changed, 12 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index 12817ffc7345..7f60e900edff 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -2459,7 +2459,7 @@ error_dump: static inline int chance(unsigned int n, unsigned int out_of) { - return !!((random32() % out_of) + 1 <= n); + return !!((prandom_u32() % out_of) + 1 <= n); } @@ -2477,13 +2477,13 @@ static int power_cut_emulated(struct ubifs_info *c, int lnum, int write) if (chance(1, 2)) { d->pc_delay = 1; /* Fail withing 1 minute */ - delay = random32() % 60000; + delay = prandom_u32() % 60000; d->pc_timeout = jiffies; d->pc_timeout += msecs_to_jiffies(delay); ubifs_warn("failing after %lums", delay); } else { d->pc_delay = 2; - delay = random32() % 10000; + delay = prandom_u32() % 10000; /* Fail within 10000 operations */ d->pc_cnt_max = delay; ubifs_warn("failing after %lu calls", delay); @@ -2563,7 +2563,7 @@ static int corrupt_data(const struct ubifs_info *c, const void *buf, unsigned int from, to, ffs = chance(1, 2); unsigned char *p = (void *)buf; - from = random32() % (len + 1); + from = prandom_u32() % (len + 1); /* Corruption may only span one max. write unit */ to = min(len, ALIGN(from, c->max_write_size)); diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c index 9daaeef675dd..4b826abb1528 100644 --- a/fs/ubifs/lpt_commit.c +++ b/fs/ubifs/lpt_commit.c @@ -2007,28 +2007,28 @@ static int dbg_populate_lsave(struct ubifs_info *c) if (!dbg_is_chk_gen(c)) return 0; - if (random32() & 3) + if (prandom_u32() & 3) return 0; for (i = 0; i < c->lsave_cnt; i++) c->lsave[i] = c->main_first; list_for_each_entry(lprops, &c->empty_list, list) - c->lsave[random32() % c->lsave_cnt] = lprops->lnum; + c->lsave[prandom_u32() % c->lsave_cnt] = lprops->lnum; list_for_each_entry(lprops, &c->freeable_list, list) - c->lsave[random32() % c->lsave_cnt] = lprops->lnum; + c->lsave[prandom_u32() % c->lsave_cnt] = lprops->lnum; list_for_each_entry(lprops, &c->frdi_idx_list, list) - c->lsave[random32() % c->lsave_cnt] = lprops->lnum; + c->lsave[prandom_u32() % c->lsave_cnt] = lprops->lnum; heap = &c->lpt_heap[LPROPS_DIRTY_IDX - 1]; for (i = 0; i < heap->cnt; i++) - c->lsave[random32() % c->lsave_cnt] = heap->arr[i]->lnum; + c->lsave[prandom_u32() % c->lsave_cnt] = heap->arr[i]->lnum; heap = &c->lpt_heap[LPROPS_DIRTY - 1]; for (i = 0; i < heap->cnt; i++) - c->lsave[random32() % c->lsave_cnt] = heap->arr[i]->lnum; + c->lsave[prandom_u32() % c->lsave_cnt] = heap->arr[i]->lnum; heap = &c->lpt_heap[LPROPS_FREE - 1]; for (i = 0; i < heap->cnt; i++) - c->lsave[random32() % c->lsave_cnt] = heap->arr[i]->lnum; + c->lsave[prandom_u32() % c->lsave_cnt] = heap->arr[i]->lnum; return 1; } diff --git a/fs/ubifs/tnc_commit.c b/fs/ubifs/tnc_commit.c index 523bbad69c0c..52a6559275c4 100644 --- a/fs/ubifs/tnc_commit.c +++ b/fs/ubifs/tnc_commit.c @@ -683,7 +683,7 @@ static int alloc_idx_lebs(struct ubifs_info *c, int cnt) c->ilebs[c->ileb_cnt++] = lnum; dbg_cmt("LEB %d", lnum); } - if (dbg_is_chk_index(c) && !(random32() & 7)) + if (dbg_is_chk_index(c) && !(prandom_u32() & 7)) return -ENOSPC; return 0; } -- cgit From 1e817fb62cd185a2232ad4302579491805609489 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Mon, 19 Nov 2012 10:26:16 -0800 Subject: time: create __getnstimeofday for WARNless calls The pstore RAM backend can get called during resume, and must be defensive against a suspended time source. Expose getnstimeofday logic that returns an error instead of a WARN. This can be detected and the timestamp can be zeroed out. Reported-by: Doug Anderson Cc: John Stultz Cc: Anton Vorontsov Signed-off-by: Kees Cook Signed-off-by: John Stultz --- fs/pstore/ram.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c index 1a4f6da58eab..dacfe78aee7e 100644 --- a/fs/pstore/ram.c +++ b/fs/pstore/ram.c @@ -168,12 +168,16 @@ static ssize_t ramoops_pstore_read(u64 *id, enum pstore_type_id *type, static size_t ramoops_write_kmsg_hdr(struct persistent_ram_zone *prz) { char *hdr; - struct timeval timestamp; + struct timespec timestamp; size_t len; - do_gettimeofday(×tamp); + /* Report zeroed timestamp if called before timekeeping has resumed. */ + if (__getnstimeofday(×tamp)) { + timestamp.tv_sec = 0; + timestamp.tv_nsec = 0; + } hdr = kasprintf(GFP_ATOMIC, RAMOOPS_KERNMSG_HDR "%lu.%lu\n", - (long)timestamp.tv_sec, (long)timestamp.tv_usec); + (long)timestamp.tv_sec, (long)(timestamp.tv_nsec / 1000)); WARN_ON_ONCE(!hdr); len = hdr ? strlen(hdr) : 0; persistent_ram_write(prz, hdr, len); -- cgit From aeb4f20a02b4c984c48995ad54f40caf5ffa0705 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Wed, 16 Jan 2013 17:33:53 -0600 Subject: xfs: Do not return EFSCORRUPTED when filesystem probe finds no XFS magic 9802182 changed the return value from EWRONGFS (aka EINVAL) to EFSCORRUPTED which doesn't seem to be handled properly by the root filesystem probe. Signed-off-by: Eric Sandeen Tested-by: Sergei Trofimovich Reviewed-by: Ben Myers Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index da508463ff10..7d6df7c00c36 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -658,7 +658,7 @@ xfs_sb_quiet_read_verify( return; } /* quietly fail */ - xfs_buf_ioerror(bp, EFSCORRUPTED); + xfs_buf_ioerror(bp, EWRONGFS); } static void -- cgit From aaddea812cb0a2dc38b55ba557b68999bc2f6203 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Wed, 16 Jan 2013 20:21:26 -0500 Subject: ext4: add tracepoint in punching hole This patch adds a tracepoint in ext4_punch_hole. CC: Lukas Czerner Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 93a7e8453a68..5abf89c100b9 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3567,6 +3567,8 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length) return -EOPNOTSUPP; } + trace_ext4_punch_hole(inode, offset, length); + return ext4_ext_punch_hole(file, offset, length); } -- cgit From 4de80273a2ffc619f7fa20d85c37e6ba1bbae1ed Mon Sep 17 00:00:00 2001 From: Philippe De Muyter Date: Wed, 16 Jan 2013 22:55:48 +0100 Subject: fs/jfs: Fix typo in comment : 'how may' -> 'how many' Signed-off-by: Philippe De Muyter Signed-off-by: Jiri Kosina --- fs/jfs/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/jfs/super.c b/fs/jfs/super.c index 1a543be09c79..060ba638becb 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -154,7 +154,7 @@ static int jfs_statfs(struct dentry *dentry, struct kstatfs *buf) /* * If we really return the number of allocated & free inodes, some * applications will fail because they won't see enough free inodes. - * We'll try to calculate some guess as to how may inodes we can + * We'll try to calculate some guess as to how many inodes we can * really allocate * * buf->f_files = atomic_read(&imap->im_numinos); -- cgit From 79aec9844de339531f05b019644ccaf5dd777144 Mon Sep 17 00:00:00 2001 From: Sam Lang Date: Wed, 19 Dec 2012 09:44:23 -1000 Subject: ceph: Check for err on mds request in atomic_open The error returned by ceph_mdsc_do_request includes errors sending the request, errors on timeout, or any errors coming from the mds. If ceph_mdsc_do_request returns an error, the reply struct will most likely be bogus. We need to bail out and propogate the error instead of overwriting it. Signed-off-by: Sam Lang Reviewed-by: Sage Weil --- fs/ceph/file.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/ceph/file.c b/fs/ceph/file.c index d415096800a6..2c71cbd78332 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -243,6 +243,9 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry, err = ceph_mdsc_do_request(mdsc, (flags & (O_CREAT|O_TRUNC)) ? dir : NULL, req); + if (err) + goto out_err; + err = ceph_handle_snapdir(req, dentry, err); if (err == 0 && (flags & O_CREAT) && !req->r_reply_info.head->is_dentry) err = ceph_handle_notrace_create(dir, dentry); -- cgit From 6e8575faa8fa680d59404a4d58d12190667be815 Mon Sep 17 00:00:00 2001 From: Sam Lang Date: Fri, 28 Dec 2012 09:56:46 -0800 Subject: ceph: Check for created flag in response from mds The mds now sends back a created inode if the create request performed the create. If the file already existed, no inode is returned in the reply. This allows ceph to set the created flag in atomic_open so that permissions are properly checked in the case that the file wasn't created by the create call to the mds. To ensure compability with previous kernels, a feature for sending back the inode in the create reply was added, so that the mds will only send back the inode if the client indicates it supports the feature. Signed-off-by: Sam Lang Reviewed-by: Sage Weil --- fs/ceph/file.c | 3 +++ fs/ceph/mds_client.c | 33 +++++++++++++++++++++++++++++++-- fs/ceph/mds_client.h | 6 ++++++ 3 files changed, 40 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 2c71cbd78332..22b5b71b5401 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -266,6 +266,9 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry, err = finish_no_open(file, dn); } else { dout("atomic_open finish_open on dn %p\n", dn); + if (req->r_op == CEPH_MDS_OP_CREATE && req->r_reply_info.has_create_ino) { + *opened |= FILE_CREATED; + } err = finish_open(file, dentry, ceph_open, opened); } diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 9165eb8309eb..d95842036c8b 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -232,6 +232,30 @@ bad: return -EIO; } +/* + * parse create results + */ +static int parse_reply_info_create(void **p, void *end, + struct ceph_mds_reply_info_parsed *info, + int features) +{ + if (features & CEPH_FEATURE_REPLY_CREATE_INODE) { + if (*p == end) { + info->has_create_ino = false; + } else { + info->has_create_ino = true; + info->ino = ceph_decode_64(p); + } + } + + if (unlikely(*p != end)) + goto bad; + return 0; + +bad: + return -EIO; +} + /* * parse extra results */ @@ -241,8 +265,12 @@ static int parse_reply_info_extra(void **p, void *end, { if (info->head->op == CEPH_MDS_OP_GETFILELOCK) return parse_reply_info_filelock(p, end, info, features); - else + else if (info->head->op == CEPH_MDS_OP_READDIR) return parse_reply_info_dir(p, end, info, features); + else if (info->head->op == CEPH_MDS_OP_CREATE) + return parse_reply_info_create(p, end, info, features); + else + return -EIO; } /* @@ -2170,7 +2198,8 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg) mutex_lock(&req->r_fill_mutex); err = ceph_fill_trace(mdsc->fsc->sb, req, req->r_session); if (err == 0) { - if (result == 0 && req->r_op != CEPH_MDS_OP_GETFILELOCK && + if (result == 0 && (req->r_op == CEPH_MDS_OP_READDIR || + req->r_op == CEPH_MDS_OP_LSSNAP) && rinfo->dir_nr) ceph_readdir_prepopulate(req, req->r_session); ceph_unreserve_caps(mdsc, &req->r_caps_reservation); diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index dd26846dd71d..567f7c60354e 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -74,6 +74,12 @@ struct ceph_mds_reply_info_parsed { struct ceph_mds_reply_info_in *dir_in; u8 dir_complete, dir_end; }; + + /* for create results */ + struct { + bool has_create_ino; + u64 ino; + }; }; /* encoded blob describing snapshot contexts for certain -- cgit From 8a92a119b292012a9bd920b908c3e9f1c512291d Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Fri, 4 Jan 2013 14:28:07 +0800 Subject: ceph: move dirty inode to migrating list when clearing auth caps Signed-off-by: Yan, Zheng Reviewed-by: Sage Weil --- fs/ceph/caps.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index a1d9bb30c1bf..a9fe2d5784c9 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -611,8 +611,16 @@ retry: if (flags & CEPH_CAP_FLAG_AUTH) ci->i_auth_cap = cap; - else if (ci->i_auth_cap == cap) + else if (ci->i_auth_cap == cap) { ci->i_auth_cap = NULL; + spin_lock(&mdsc->cap_dirty_lock); + if (!list_empty(&ci->i_dirty_item)) { + dout(" moving %p to cap_dirty_migrating\n", inode); + list_move(&ci->i_dirty_item, + &mdsc->cap_dirty_migrating); + } + spin_unlock(&mdsc->cap_dirty_lock); + } dout("add_cap inode %p (%llx.%llx) cap %p %s now %s seq %d mds%d\n", inode, ceph_vinop(inode), cap, ceph_cap_string(issued), -- cgit From 395c312b9c535d57db122cbb5b7292223561d0b8 Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Fri, 4 Jan 2013 14:37:57 +0800 Subject: ceph: allow revoking duplicated caps issued by non-auth MDS Allow revoking duplicated caps issued by non-auth MDS if these caps are also issued by auth MDS. Signed-off-by: Yan, Zheng Reviewed-by: Sage Weil --- fs/ceph/caps.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index a9fe2d5784c9..76b19239c426 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1468,7 +1468,7 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags, struct ceph_mds_client *mdsc = fsc->mdsc; struct inode *inode = &ci->vfs_inode; struct ceph_cap *cap; - int file_wanted, used; + int file_wanted, used, cap_used; int took_snap_rwsem = 0; /* true if mdsc->snap_rwsem held */ int issued, implemented, want, retain, revoking, flushing = 0; int mds = -1; /* keep track of how far we've gone through i_caps list @@ -1571,9 +1571,14 @@ retry_locked: /* NOTE: no side-effects allowed, until we take s_mutex */ + cap_used = used; + if (ci->i_auth_cap && cap != ci->i_auth_cap) + cap_used &= ~ci->i_auth_cap->issued; + revoking = cap->implemented & ~cap->issued; - dout(" mds%d cap %p issued %s implemented %s revoking %s\n", + dout(" mds%d cap %p used %s issued %s implemented %s revoking %s\n", cap->mds, cap, ceph_cap_string(cap->issued), + ceph_cap_string(cap_used), ceph_cap_string(cap->implemented), ceph_cap_string(revoking)); @@ -1601,7 +1606,7 @@ retry_locked: } /* completed revocation? going down and there are no caps? */ - if (revoking && (revoking & used) == 0) { + if (revoking && (revoking & cap_used) == 0) { dout("completed revocation of %s\n", ceph_cap_string(cap->implemented & ~cap->issued)); goto ack; @@ -1678,8 +1683,8 @@ ack: sent++; /* __send_cap drops i_ceph_lock */ - delayed += __send_cap(mdsc, cap, CEPH_CAP_OP_UPDATE, used, want, - retain, flushing, NULL); + delayed += __send_cap(mdsc, cap, CEPH_CAP_OP_UPDATE, cap_used, + want, retain, flushing, NULL); goto retry; /* retake i_ceph_lock and restart our cap scan. */ } -- cgit From 66f58691c5c820283dd7e4d6fe8649033ed43ceb Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Fri, 4 Jan 2013 14:45:18 +0800 Subject: ceph: allocate cap_release message when receiving cap import When client wants to release an imported cap, it's possible there is no reserved cap_release message in corresponding mds session. so __queue_cap_release causes kernel panic. Signed-off-by: Yan, Zheng Reviewed-by: Sage Weil --- fs/ceph/caps.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 76b19239c426..40b5bbe63a39 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -2833,6 +2833,9 @@ void ceph_handle_caps(struct ceph_mds_session *session, dout(" mds%d seq %lld cap seq %u\n", session->s_mds, session->s_seq, (unsigned)seq); + if (op == CEPH_CAP_OP_IMPORT) + ceph_add_cap_releases(mdsc, session); + /* lookup ino */ inode = ceph_find_inode(sb, vino); ci = ceph_inode(inode); -- cgit From 390306c38dd43908f7f7730229999790a773d1d5 Mon Sep 17 00:00:00 2001 From: "Yan, Zheng" Date: Fri, 4 Jan 2013 15:30:10 +0800 Subject: ceph: check mds_wanted for imported cap The MDS may have incorrect wanted caps after importing caps. So the client should check the value mds has and send cap update if necessary. Signed-off-by: Yan, Zheng Reviewed-by: Sage Weil --- fs/ceph/caps.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 40b5bbe63a39..1e1e02055a2b 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -2429,7 +2429,9 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant, dout("mds wanted %s -> %s\n", ceph_cap_string(le32_to_cpu(grant->wanted)), ceph_cap_string(wanted)); - grant->wanted = cpu_to_le32(wanted); + /* imported cap may not have correct mds_wanted */ + if (le32_to_cpu(grant->op) == CEPH_CAP_OP_IMPORT) + check_caps = 1; } cap->seq = seq; -- cgit From fa5199648e273a5e3e80aca41c1eb53700438dc1 Mon Sep 17 00:00:00 2001 From: Simon Que Date: Thu, 17 Jan 2013 11:18:20 -0800 Subject: eCryptfs: initialize payload_len in keystore.c This is meant to remove a compiler warning. It should not make any functional change. payload_len should be initialized when it is passed to write_tag_64_packet() as a pointer. If that call fails, this function should return early, and payload_len won't be used. Signed-off-by: Simon Que Signed-off-by: Tyler Hicks --- fs/ecryptfs/keystore.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index 2333203a120b..6154cde3a052 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -1150,7 +1150,7 @@ decrypt_pki_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, struct ecryptfs_message *msg = NULL; char *auth_tok_sig; char *payload; - size_t payload_len; + size_t payload_len = 0; int rc; rc = ecryptfs_get_auth_tok_sig(&auth_tok_sig, auth_tok); -- cgit From bbcb03e3ce8c76a4ecf0ea7d365f9f0e913fc329 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Thu, 17 Jan 2013 11:36:10 -0800 Subject: eCryptfs: Fix -Wunused-but-set-variable warnings These two variables are no longer used and can be removed. Fixes warnings when building with W=1. Signed-off-by: Tyler Hicks --- fs/ecryptfs/dentry.c | 2 -- fs/ecryptfs/file.c | 2 -- 2 files changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/dentry.c b/fs/ecryptfs/dentry.c index 1b5d9af937df..bf12ba5dd223 100644 --- a/fs/ecryptfs/dentry.c +++ b/fs/ecryptfs/dentry.c @@ -45,14 +45,12 @@ static int ecryptfs_d_revalidate(struct dentry *dentry, unsigned int flags) { struct dentry *lower_dentry; - struct vfsmount *lower_mnt; int rc = 1; if (flags & LOOKUP_RCU) return -ECHILD; lower_dentry = ecryptfs_dentry_to_lower(dentry); - lower_mnt = ecryptfs_dentry_to_lower_mnt(dentry); if (!lower_dentry->d_op || !lower_dentry->d_op->d_revalidate) goto out; rc = lower_dentry->d_op->d_revalidate(lower_dentry, flags); diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index d45ba4568128..bfa52d2ef460 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -199,7 +199,6 @@ static int ecryptfs_open(struct inode *inode, struct file *file) struct dentry *ecryptfs_dentry = file->f_path.dentry; /* Private value of ecryptfs_dentry allocated in * ecryptfs_lookup() */ - struct dentry *lower_dentry; struct ecryptfs_file_info *file_info; mount_crypt_stat = &ecryptfs_superblock_to_private( @@ -222,7 +221,6 @@ static int ecryptfs_open(struct inode *inode, struct file *file) rc = -ENOMEM; goto out; } - lower_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry); crypt_stat = &ecryptfs_inode_to_private(inode)->crypt_stat; mutex_lock(&crypt_stat->cs_mutex); if (!(crypt_stat->flags & ECRYPTFS_POLICY_APPLIED)) { -- cgit From 111d61a25ec11c1837ac0fd81bf4b845d3327fb7 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Thu, 17 Jan 2013 12:19:35 -0800 Subject: eCryptfs: Fix -Wmissing-prototypes warnings Mark two inode operation fuctions as static. Fixes warnings when building with W=1. Signed-off-by: Tyler Hicks --- fs/ecryptfs/inode.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index cc7709e7c508..ddd961ba2cf9 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -999,8 +999,8 @@ out: return rc; } -int ecryptfs_getattr_link(struct vfsmount *mnt, struct dentry *dentry, - struct kstat *stat) +static int ecryptfs_getattr_link(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat) { struct ecryptfs_mount_crypt_stat *mount_crypt_stat; int rc = 0; @@ -1021,8 +1021,8 @@ int ecryptfs_getattr_link(struct vfsmount *mnt, struct dentry *dentry, return rc; } -int ecryptfs_getattr(struct vfsmount *mnt, struct dentry *dentry, - struct kstat *stat) +static int ecryptfs_getattr(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat) { struct kstat lower_stat; int rc; -- cgit From 6b8fbde41866306fa910754a59671b27aa5cfd37 Mon Sep 17 00:00:00 2001 From: Bin Wang Date: Thu, 6 Dec 2012 17:08:56 +0800 Subject: sysfs: Fixed a trailing white space error This patch removes the trailing white space in fs/sysfs/mount.c. Signed-off-by: Bin Wang Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/mount.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c index db940a9be045..8d924b5ec733 100644 --- a/fs/sysfs/mount.c +++ b/fs/sysfs/mount.c @@ -10,7 +10,7 @@ * Please see Documentation/filesystems/sysfs.txt for more information. */ -#define DEBUG +#define DEBUG #include #include -- cgit From 6ad58fa82db897b4422a873c01fa41f84b652502 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Wed, 12 Dec 2012 10:19:09 -0800 Subject: sysfs: Convert print_symbol to %pSR Use the new vsprintf extension to avoid any possible message interleaving. Signed-off-by: Joe Perches Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/file.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 602f56db0442..bd7e59ba9246 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -91,8 +91,8 @@ static int fill_read_buffer(struct dentry * dentry, struct sysfs_buffer * buffer * indicate truncated result or overflow in normal use cases. */ if (count >= (ssize_t)PAGE_SIZE) { - print_symbol("fill_read_buffer: %s returned bad count\n", - (unsigned long)ops->show); + printk("fill_read_buffer: %pSR returned bad count\n", + ops->show); /* Try to struggle along */ count = PAGE_SIZE - 1; } -- cgit From 1884bd4b1461d16d10ab4ab4d19725b4e0717f7a Mon Sep 17 00:00:00 2001 From: Sasha Levin Date: Thu, 20 Dec 2012 14:11:22 -0500 Subject: debugfs: remove redundant initialization of dentry We already initialize it to NULL when declaring it, no need to do that twice. Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman --- fs/debugfs/inode.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 153bb1e42e63..d329f9c55753 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -322,7 +322,6 @@ static struct dentry *__create_file(const char *name, umode_t mode, if (!parent) parent = debugfs_mount->mnt_root; - dentry = NULL; mutex_lock(&parent->d_inode->i_mutex); dentry = lookup_one_len(name, parent, strlen(name)); if (!IS_ERR(dentry)) { -- cgit From 595e0eb06756bdf332c76176980eded1ec50032a Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 17 Jan 2013 13:09:57 -0800 Subject: Revert "sysfs: Convert print_symbol to %pSR" This reverts commit 6ad58fa82db897b4422a873c01fa41f84b652502 as %pSR isn't in the tree yet. Cc: Joe Perches Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/file.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index bd7e59ba9246..602f56db0442 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -91,8 +91,8 @@ static int fill_read_buffer(struct dentry * dentry, struct sysfs_buffer * buffer * indicate truncated result or overflow in normal use cases. */ if (count >= (ssize_t)PAGE_SIZE) { - printk("fill_read_buffer: %pSR returned bad count\n", - ops->show); + print_symbol("fill_read_buffer: %s returned bad count\n", + (unsigned long)ops->show); /* Try to struggle along */ count = PAGE_SIZE - 1; } -- cgit From e8afad656cbcd06d02a7bacd4b318fa0e2907de0 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Wed, 14 Nov 2012 09:38:19 -0600 Subject: libceph: pass length to ceph_calc_file_object_mapping() ceph_calc_file_object_mapping() takes (among other things) a "file" offset and length, and based on the layout, determines the object number ("bno") backing the affected portion of the file's data and the offset into that object where the desired range begins. It also computes the size that should be used for the request--either the amount requested or something less if that would exceed the end of the object. This patch changes the input length parameter in this function so it is used only for input. That is, the argument will be passed by value rather than by address, so the value provided won't get updated by the function. The value would only get updated if the length would surpass the current object, and in that case the value it got updated to would be exactly that returned in *oxlen. Only one of the two callers is affected by this change. Update ceph_calc_raw_layout() so it records any updated value. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/ioctl.c b/fs/ceph/ioctl.c index 36549a46e311..3b22150d3e19 100644 --- a/fs/ceph/ioctl.c +++ b/fs/ceph/ioctl.c @@ -194,7 +194,7 @@ static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) return -EFAULT; down_read(&osdc->map_sem); - r = ceph_calc_file_object_mapping(&ci->i_layout, dl.file_offset, &len, + r = ceph_calc_file_object_mapping(&ci->i_layout, dl.file_offset, len, &dl.object_no, &dl.object_offset, &olen); if (r < 0) -- cgit From 667a9291c5b38318a47a2df730bc064e5d37d2a8 Mon Sep 17 00:00:00 2001 From: Thiago Farina Date: Mon, 12 Nov 2012 21:32:59 -0200 Subject: xfs: Remove boolean_t typedef completely. Since we are using C99 we have one builtin defined in include/linux/types.h, use that instead. v2: you missed one in fs/xfs/xfs_qm_bhv.c, cleaned up. -bpm Signed-off-by: Thiago Farina Reviewed-by: Ben Myers Signed-off-by: Ben Myers --- fs/xfs/xfs_log.c | 8 ++++---- fs/xfs/xfs_qm_bhv.c | 2 +- fs/xfs/xfs_qm_syscalls.c | 8 ++++---- fs/xfs/xfs_trans_dquot.c | 10 +++++----- fs/xfs/xfs_types.h | 1 - fs/xfs/xfs_vnodeops.c | 12 ++++++------ 6 files changed, 20 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 46bd9d52ab51..dafdae36f2fb 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -120,7 +120,7 @@ xlog_verify_iclog( struct xlog *log, struct xlog_in_core *iclog, int count, - boolean_t syncing); + bool syncing); STATIC void xlog_verify_tail_lsn( struct xlog *log, @@ -3611,7 +3611,7 @@ xlog_verify_iclog( struct xlog *log, struct xlog_in_core *iclog, int count, - boolean_t syncing) + bool syncing) { xlog_op_header_t *ophead; xlog_in_core_t *icptr; @@ -3659,7 +3659,7 @@ xlog_verify_iclog( /* clientid is only 1 byte */ field_offset = (__psint_t) ((xfs_caddr_t)&(ophead->oh_clientid) - base_ptr); - if (syncing == B_FALSE || (field_offset & 0x1ff)) { + if (!syncing || (field_offset & 0x1ff)) { clientid = ophead->oh_clientid; } else { idx = BTOBBT((xfs_caddr_t)&(ophead->oh_clientid) - iclog->ic_datap); @@ -3682,7 +3682,7 @@ xlog_verify_iclog( /* check length */ field_offset = (__psint_t) ((xfs_caddr_t)&(ophead->oh_len) - base_ptr); - if (syncing == B_FALSE || (field_offset & 0x1ff)) { + if (!syncing || (field_offset & 0x1ff)) { op_len = be32_to_cpu(ophead->oh_len); } else { idx = BTOBBT((__psint_t)&ophead->oh_len - diff --git a/fs/xfs/xfs_qm_bhv.c b/fs/xfs/xfs_qm_bhv.c index 6b39115bf145..2d02eac1c9a8 100644 --- a/fs/xfs/xfs_qm_bhv.c +++ b/fs/xfs/xfs_qm_bhv.c @@ -146,7 +146,7 @@ xfs_qm_newmount( * inode goes inactive and wants to free blocks, * or via xfs_log_mount_finish. */ - *needquotamount = B_TRUE; + *needquotamount = true; *quotaflags = mp->m_qflags; mp->m_qflags = 0; } diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 8a59f8546552..33d9c2b6f555 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -408,10 +408,10 @@ xfs_qm_scall_getqstat( { struct xfs_quotainfo *q = mp->m_quotainfo; struct xfs_inode *uip, *gip; - boolean_t tempuqip, tempgqip; + bool tempuqip, tempgqip; uip = gip = NULL; - tempuqip = tempgqip = B_FALSE; + tempuqip = tempgqip = false; memset(out, 0, sizeof(fs_quota_stat_t)); out->qs_version = FS_QSTAT_VERSION; @@ -434,12 +434,12 @@ xfs_qm_scall_getqstat( if (!uip && mp->m_sb.sb_uquotino != NULLFSINO) { if (xfs_iget(mp, NULL, mp->m_sb.sb_uquotino, 0, 0, &uip) == 0) - tempuqip = B_TRUE; + tempuqip = true; } if (!gip && mp->m_sb.sb_gquotino != NULLFSINO) { if (xfs_iget(mp, NULL, mp->m_sb.sb_gquotino, 0, 0, &gip) == 0) - tempgqip = B_TRUE; + tempgqip = true; } if (uip) { out->qs_uquota.qfs_nblks = uip->i_d.di_nblocks; diff --git a/fs/xfs/xfs_trans_dquot.c b/fs/xfs/xfs_trans_dquot.c index 0c7fa54f309e..642c2d6e1db1 100644 --- a/fs/xfs/xfs_trans_dquot.c +++ b/fs/xfs/xfs_trans_dquot.c @@ -516,7 +516,7 @@ xfs_trans_unreserve_and_mod_dquots( int i, j; xfs_dquot_t *dqp; xfs_dqtrx_t *qtrx, *qa; - boolean_t locked; + bool locked; if (!tp->t_dqinfo || !(tp->t_flags & XFS_TRANS_DQ_DIRTY)) return; @@ -537,17 +537,17 @@ xfs_trans_unreserve_and_mod_dquots( * about the number of blocks used field, or deltas. * Also we don't bother to zero the fields. */ - locked = B_FALSE; + locked = false; if (qtrx->qt_blk_res) { xfs_dqlock(dqp); - locked = B_TRUE; + locked = true; dqp->q_res_bcount -= (xfs_qcnt_t)qtrx->qt_blk_res; } if (qtrx->qt_ino_res) { if (!locked) { xfs_dqlock(dqp); - locked = B_TRUE; + locked = true; } dqp->q_res_icount -= (xfs_qcnt_t)qtrx->qt_ino_res; @@ -556,7 +556,7 @@ xfs_trans_unreserve_and_mod_dquots( if (qtrx->qt_rtblk_res) { if (!locked) { xfs_dqlock(dqp); - locked = B_TRUE; + locked = true; } dqp->q_res_rtbcount -= (xfs_qcnt_t)qtrx->qt_rtblk_res; diff --git a/fs/xfs/xfs_types.h b/fs/xfs/xfs_types.h index 7a41874f4c20..61ba1cfa974c 100644 --- a/fs/xfs/xfs_types.h +++ b/fs/xfs/xfs_types.h @@ -32,7 +32,6 @@ typedef unsigned int __uint32_t; typedef signed long long int __int64_t; typedef unsigned long long int __uint64_t; -typedef enum { B_FALSE,B_TRUE } boolean_t; typedef __uint32_t prid_t; /* project ID */ typedef __uint32_t inst_t; /* an instruction */ diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index d95f565a390e..77ad74834baa 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -725,7 +725,7 @@ xfs_create( int error; xfs_bmap_free_t free_list; xfs_fsblock_t first_block; - boolean_t unlock_dp_on_error = B_FALSE; + bool unlock_dp_on_error = false; uint cancel_flags; int committed; prid_t prid; @@ -794,7 +794,7 @@ xfs_create( } xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT); - unlock_dp_on_error = B_TRUE; + unlock_dp_on_error = true; xfs_bmap_init(&free_list, &first_block); @@ -830,7 +830,7 @@ xfs_create( * error path. */ xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL); - unlock_dp_on_error = B_FALSE; + unlock_dp_on_error = false; error = xfs_dir_createname(tp, dp, name, ip->i_ino, &first_block, &free_list, resblks ? @@ -1367,7 +1367,7 @@ xfs_symlink( int pathlen; xfs_bmap_free_t free_list; xfs_fsblock_t first_block; - boolean_t unlock_dp_on_error = B_FALSE; + bool unlock_dp_on_error = false; uint cancel_flags; int committed; xfs_fileoff_t first_fsb; @@ -1438,7 +1438,7 @@ xfs_symlink( } xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT); - unlock_dp_on_error = B_TRUE; + unlock_dp_on_error = true; /* * Check whether the directory allows new symlinks or not. @@ -1484,7 +1484,7 @@ xfs_symlink( * error path. */ xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL); - unlock_dp_on_error = B_FALSE; + unlock_dp_on_error = false; /* * Also attach the dquot(s) to it, if applicable. -- cgit From 9e96fe6df44425b69ed89f6ac20352cec1f127d7 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Thu, 17 Jan 2013 13:11:29 -0500 Subject: xfs: pull up stack_switch check into xfs_bmapi_write The stack_switch check currently occurs in __xfs_bmapi_allocate, which means the stack switch only occurs when xfs_bmapi_allocate() is called in a loop. Pull the check up before the loop in xfs_bmapi_write() such that the first iteration of the loop has consistent behavior. Signed-off-by: Brian Foster Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_bmap.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index c5077208e419..491f35e2dd26 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -4676,9 +4676,6 @@ __xfs_bmapi_allocate( return error; } - if (bma->flags & XFS_BMAPI_STACK_SWITCH) - bma->stack_switch = 1; - error = xfs_bmap_alloc(bma); if (error) return error; @@ -4952,6 +4949,9 @@ xfs_bmapi_write( bma.flist = flist; bma.firstblock = firstblock; + if (flags & XFS_BMAPI_STACK_SWITCH) + bma.stack_switch = 1; + while (bno < end && n < *nmap) { inhole = eof || bma.got.br_startoff > bno; wasdelay = !inhole && isnullstartblock(bma.got.br_startblock); -- cgit From 003fd6c8be14a348c56cb1d171605ab13fca906f Mon Sep 17 00:00:00 2001 From: Ben Myers Date: Fri, 18 Jan 2013 14:17:46 -0600 Subject: xfs: fix fs/xfs/xfs_log.c:1740:39: error: 'B_TRUE' undeclared Commit 667a9291c5b3 "xfs: Remove boolean_t typedef completely." didn't. Remove a stray B_TRUE that breaks CONFIG_XFS_DEBUG=y. Signed-off-by: Ben Myers Reported-by: Wu Fengguang Reviewed-by: Mark Tinguely --- fs/xfs/xfs_log.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index dafdae36f2fb..eec226f78a40 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -1737,7 +1737,7 @@ xlog_sync( ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1); ASSERT(XFS_BUF_ADDR(bp) + BTOBB(count) <= log->l_logBBsize); - xlog_verify_iclog(log, iclog, count, B_TRUE); + xlog_verify_iclog(log, iclog, count, true); /* account for log which doesn't start at block #0 */ XFS_BUF_SET_ADDR(bp, XFS_BUF_ADDR(bp) + log->l_logBBstart); -- cgit From 4f73bc4dd3e8563ef4109f293a092820dff66d92 Mon Sep 17 00:00:00 2001 From: Joe Millenbach Date: Thu, 17 Jan 2013 22:44:22 -0800 Subject: tty: Added a CONFIG_TTY option to allow removal of TTY The option allows you to remove TTY and compile without errors. This saves space on systems that won't support TTY interfaces anyway. bloat-o-meter output is below. The bulk of this patch consists of Kconfig changes adding "depends on TTY" to various serial devices and similar drivers that require the TTY layer. Ideally, these dependencies would occur on a common intermediate symbol such as SERIO, but most drivers "select SERIO" rather than "depends on SERIO", and "select" does not respect dependencies. bloat-o-meter output comparing our previous minimal to new minimal by removing TTY. The list is filtered to not show removed entries with awk '$3 != "-"' as the list was very long. add/remove: 0/226 grow/shrink: 2/14 up/down: 6/-35356 (-35350) function old new delta chr_dev_init 166 170 +4 allow_signal 80 82 +2 static.__warned 143 142 -1 disallow_signal 63 62 -1 __set_special_pids 95 94 -1 unregister_console 126 121 -5 start_kernel 546 541 -5 register_console 593 588 -5 copy_from_user 45 40 -5 sys_setsid 128 120 -8 sys_vhangup 32 19 -13 do_exit 1543 1526 -17 bitmap_zero 60 40 -20 arch_local_irq_save 137 117 -20 release_task 674 652 -22 static.spin_unlock_irqrestore 308 260 -48 Signed-off-by: Joe Millenbach Reviewed-by: Jamey Sharp Reviewed-by: Josh Triplett Signed-off-by: Greg Kroah-Hartman --- fs/proc/Makefile | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/proc/Makefile b/fs/proc/Makefile index 981b05601931..712f24db9600 100644 --- a/fs/proc/Makefile +++ b/fs/proc/Makefile @@ -8,7 +8,8 @@ proc-y := nommu.o task_nommu.o proc-$(CONFIG_MMU) := mmu.o task_mmu.o proc-y += inode.o root.o base.o generic.o array.o \ - proc_tty.o fd.o + fd.o +proc-$(CONFIG_TTY) += proc_tty.o proc-y += cmdline.o proc-y += consoles.o proc-y += cpuinfo.o -- cgit From 306a74920ba9ccf6b5f110f97c1cb6bb2caeff93 Mon Sep 17 00:00:00 2001 From: Guo Chao Date: Tue, 18 Dec 2012 16:56:39 +0800 Subject: ext3, ext4, ocfs2: remove unused macro NAMEI_RA_INDEX This macro, initially introduced by ext2 in v0.99.15, does not have any users from the beginning. It has been removed in later ext2 version but still remains in the code of ext3, ext4, ocfs2. Remove this macro there. Cc: Jan Kara Cc: linux-ext4@vger.kernel.org Cc: ocfs2-devel@oss.oracle.com Acked-by: Mark Fasheh Acked-by: "Theodore Ts'o" Signed-off-by: Guo Chao Signed-off-by: Jan Kara --- fs/ext3/namei.c | 1 - fs/ext4/namei.c | 1 - fs/ocfs2/dir.c | 1 - 3 files changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 890b8947c546..88f64eb1b6fa 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -36,7 +36,6 @@ #define NAMEI_RA_CHUNKS 2 #define NAMEI_RA_BLOCKS 4 #define NAMEI_RA_SIZE (NAMEI_RA_CHUNKS * NAMEI_RA_BLOCKS) -#define NAMEI_RA_INDEX(c,b) (((c) * NAMEI_RA_BLOCKS) + (b)) static struct buffer_head *ext3_append(handle_t *handle, struct inode *inode, diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index f9ed946a448e..cc9a512f3a8a 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -47,7 +47,6 @@ #define NAMEI_RA_CHUNKS 2 #define NAMEI_RA_BLOCKS 4 #define NAMEI_RA_SIZE (NAMEI_RA_CHUNKS * NAMEI_RA_BLOCKS) -#define NAMEI_RA_INDEX(c,b) (((c) * NAMEI_RA_BLOCKS) + (b)) static struct buffer_head *ext4_append(handle_t *handle, struct inode *inode, diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index 8fe4e2892ab9..fc121350d8cb 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -67,7 +67,6 @@ #define NAMEI_RA_CHUNKS 2 #define NAMEI_RA_BLOCKS 4 #define NAMEI_RA_SIZE (NAMEI_RA_CHUNKS * NAMEI_RA_BLOCKS) -#define NAMEI_RA_INDEX(c,b) (((c) * NAMEI_RA_BLOCKS) + (b)) static unsigned char ocfs2_filetype_table[] = { DT_UNKNOWN, DT_REG, DT_DIR, DT_CHR, DT_BLK, DT_FIFO, DT_SOCK, DT_LNK -- cgit From f56426ae4d4414c9c996567710dceecbdfc39acc Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 31 Dec 2012 12:38:36 +0100 Subject: ext3: Fix memory leak when quota options are specified multiple times When usrjquota or grpjquota mount options are specified several times, we leak memory storing the names. Free the memory correctly. Reported-by: Chen Gang Signed-off-by: Jan Kara --- fs/ext3/super.c | 51 ++++++++++++++++++++++++++++++--------------------- 1 file changed, 30 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 6e50223b3299..0926fe46ae3e 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -916,21 +916,24 @@ static int set_qf_name(struct super_block *sb, int qtype, substring_t *args) "Not enough memory for storing quotafile name"); return 0; } - if (sbi->s_qf_names[qtype] && - strcmp(sbi->s_qf_names[qtype], qname)) { - ext3_msg(sb, KERN_ERR, - "%s quota file already specified", QTYPE2NAME(qtype)); + if (sbi->s_qf_names[qtype]) { + int same = !strcmp(sbi->s_qf_names[qtype], qname); + kfree(qname); - return 0; + if (!same) { + ext3_msg(sb, KERN_ERR, + "%s quota file already specified", + QTYPE2NAME(qtype)); + } + return same; } - sbi->s_qf_names[qtype] = qname; - if (strchr(sbi->s_qf_names[qtype], '/')) { + if (strchr(qname, '/')) { ext3_msg(sb, KERN_ERR, "quotafile must be on filesystem root"); - kfree(sbi->s_qf_names[qtype]); - sbi->s_qf_names[qtype] = NULL; + kfree(qname); return 0; } + sbi->s_qf_names[qtype] = qname; set_opt(sbi->s_mount_opt, QUOTA); return 1; } @@ -945,11 +948,10 @@ static int clear_qf_name(struct super_block *sb, int qtype) { " when quota turned on"); return 0; } - /* - * The space will be released later when all options are confirmed - * to be correct - */ - sbi->s_qf_names[qtype] = NULL; + if (sbi->s_qf_names[qtype]) { + kfree(sbi->s_qf_names[qtype]); + sbi->s_qf_names[qtype] = NULL; + } return 1; } #endif @@ -2605,7 +2607,18 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data) #ifdef CONFIG_QUOTA old_opts.s_jquota_fmt = sbi->s_jquota_fmt; for (i = 0; i < MAXQUOTAS; i++) - old_opts.s_qf_names[i] = sbi->s_qf_names[i]; + if (sbi->s_qf_names[i]) { + old_opts.s_qf_names[i] = kstrdup(sbi->s_qf_names[i], + GFP_KERNEL); + if (!old_opts.s_qf_names[i]) { + int j; + + for (j = 0; j < i; j++) + kfree(old_opts.s_qf_names[j]); + return -ENOMEM; + } + } else + old_opts.s_qf_names[i] = NULL; #endif /* @@ -2698,9 +2711,7 @@ static int ext3_remount (struct super_block * sb, int * flags, char * data) #ifdef CONFIG_QUOTA /* Release old quota file names */ for (i = 0; i < MAXQUOTAS; i++) - if (old_opts.s_qf_names[i] && - old_opts.s_qf_names[i] != sbi->s_qf_names[i]) - kfree(old_opts.s_qf_names[i]); + kfree(old_opts.s_qf_names[i]); #endif if (enable_quota) dquot_resume(sb, -1); @@ -2714,9 +2725,7 @@ restore_opts: #ifdef CONFIG_QUOTA sbi->s_jquota_fmt = old_opts.s_jquota_fmt; for (i = 0; i < MAXQUOTAS; i++) { - if (sbi->s_qf_names[i] && - old_opts.s_qf_names[i] != sbi->s_qf_names[i]) - kfree(sbi->s_qf_names[i]); + kfree(sbi->s_qf_names[i]); sbi->s_qf_names[i] = old_opts.s_qf_names[i]; } #endif -- cgit From 8d8759eb488f9e88fa5f976c4fd7ed205661c872 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Sat, 12 Jan 2013 01:19:32 -0800 Subject: Ext2: free memory allocated and forget buffer head when io error happens Add a necessary check when an io error happens. If io error happens,free the memory allocated and forget buffer head. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext2/inode.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'fs') diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index 6363ac66fafa..c3881e56662e 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -495,6 +495,10 @@ static int ext2_alloc_branch(struct inode *inode, * parent to disk. */ bh = sb_getblk(inode->i_sb, new_blocks[n-1]); + if (unlikely(!bh)) { + err = -ENOMEM; + goto failed; + } branch[n].bh = bh; lock_buffer(bh); memset(bh->b_data, 0, blocksize); @@ -523,6 +527,14 @@ static int ext2_alloc_branch(struct inode *inode, } *blks = num; return err; + +failed: + for (i = 1; i < n; i++) + bforget(branch[i].bh); + for (i = 0; i < indirect_blks; i++) + ext2_free_blocks(inode, new_blocks[i], 1); + ext2_free_blocks(inode, new_blocks[i], num); + return err; } /** -- cgit From 61f43e6880dee5983999fe40bf96c1cf43740b4c Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Sat, 12 Jan 2013 01:22:33 -0800 Subject: Ext3: add necessary check in case IO error happens As we know io error may happen when the function 'sb_getblk' is called.Add necessary check for it The patch also fix a coding style problem. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext3/inode.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index b176d4253544..6e4f8a529fbc 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -676,6 +676,10 @@ static int ext3_alloc_branch(handle_t *handle, struct inode *inode, * parent to disk. */ bh = sb_getblk(inode->i_sb, new_blocks[n-1]); + if (unlikely(!bh)) { + err = -ENOMEM; + goto failed; + } branch[n].bh = bh; lock_buffer(bh); BUFFER_TRACE(bh, "call get_create_access"); @@ -717,7 +721,7 @@ failed: BUFFER_TRACE(branch[i].bh, "call journal_forget"); ext3_journal_forget(handle, branch[i].bh); } - for (i = 0; i Date: Sat, 12 Jan 2013 01:34:50 -0800 Subject: Ext2: use unlikely to improve the efficiency of the kernel Because the function 'sb_getblk' seldomly fails to return NULL value. It will be better to use unlikely to optimize it. Signed-off-by: Wang shilong Signed-off-by: Jan Kara --- fs/ext2/super.c | 2 +- fs/ext2/xattr.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext2/super.c b/fs/ext2/super.c index fa04d023177e..7f68c8114026 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -1500,7 +1500,7 @@ static ssize_t ext2_quota_write(struct super_block *sb, int type, bh = sb_bread(sb, tmp_bh.b_blocknr); else bh = sb_getblk(sb, tmp_bh.b_blocknr); - if (!bh) { + if (unlikely(!bh)) { err = -EIO; goto out; } diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c index b6754dbbce3c..06209ec46152 100644 --- a/fs/ext2/xattr.c +++ b/fs/ext2/xattr.c @@ -662,7 +662,7 @@ ext2_xattr_set2(struct inode *inode, struct buffer_head *old_bh, ea_idebug(inode, "creating block %d", block); new_bh = sb_getblk(sb, block); - if (!new_bh) { + if (unlikely(!new_bh)) { ext2_free_blocks(inode, block, 1); mark_inode_dirty(inode); error = -EIO; -- cgit From 1b7d76e9b1106f2be062f915b05d47658dd4fc63 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Sat, 12 Jan 2013 01:36:17 -0800 Subject: Ext3: use unlikely to improve the efficiency of the kernel Because the function 'sb_getblk' seldomly fails to return NULL value,it will be better to use unlikely to check it. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext3/inode.c | 6 +++--- fs/ext3/resize.c | 6 +++--- fs/ext3/xattr.c | 2 +- 3 files changed, 7 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 6e4f8a529fbc..d7df06839f6a 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1082,7 +1082,7 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode *inode, if (!err && buffer_mapped(&dummy)) { struct buffer_head *bh; bh = sb_getblk(inode->i_sb, dummy.b_blocknr); - if (!bh) { + if (unlikely(!bh)) { *errp = -EIO; goto err; } @@ -2733,7 +2733,7 @@ static int __ext3_get_inode_loc(struct inode *inode, return -EIO; bh = sb_getblk(inode->i_sb, block); - if (!bh) { + if (unlikely(!bh)) { ext3_error (inode->i_sb, "ext3_get_inode_loc", "unable to read inode block - " "inode=%lu, block="E3FSBLK, @@ -2787,7 +2787,7 @@ static int __ext3_get_inode_loc(struct inode *inode, bitmap_bh = sb_getblk(inode->i_sb, le32_to_cpu(desc->bg_inode_bitmap)); - if (!bitmap_bh) + if (unlikely(!bitmap_bh)) goto make_io; /* diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 0f814f3450de..704e8ce7d782 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -116,7 +116,7 @@ static struct buffer_head *bclean(handle_t *handle, struct super_block *sb, int err; bh = sb_getblk(sb, blk); - if (!bh) + if (unlikely(!bh)) return ERR_PTR(-EIO); if ((err = ext3_journal_get_write_access(handle, bh))) { brelse(bh); @@ -234,7 +234,7 @@ static int setup_new_group_blocks(struct super_block *sb, goto exit_bh; gdb = sb_getblk(sb, block); - if (!gdb) { + if (unlikely(!gdb)) { err = -EIO; goto exit_bh; } @@ -722,7 +722,7 @@ static void update_backups(struct super_block *sb, break; bh = sb_getblk(sb, group * bpg + blk_off); - if (!bh) { + if (unlikely(!bh)) { err = -EIO; break; } diff --git a/fs/ext3/xattr.c b/fs/ext3/xattr.c index d22ebb7a4f55..9f57470b1727 100644 --- a/fs/ext3/xattr.c +++ b/fs/ext3/xattr.c @@ -813,7 +813,7 @@ inserted: ea_idebug(inode, "creating block %d", block); new_bh = sb_getblk(sb, block); - if (!new_bh) { + if (unlikely(!new_bh)) { getblk_failed: ext3_free_blocks(handle, inode, block, 1); error = -EIO; -- cgit From ab6a773dbcbd2bba3ead8676ae21ce5adbbdc035 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Tue, 15 Jan 2013 21:19:06 -0800 Subject: Ext2: return ENOMEM rather than EIO if sb_getblk fails As the only reason that sb_getblks fails is that allocation fails. It will be better to use ENOMEM rather than EIO. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext2/xattr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c index 06209ec46152..2d7557db3ae8 100644 --- a/fs/ext2/xattr.c +++ b/fs/ext2/xattr.c @@ -665,7 +665,7 @@ ext2_xattr_set2(struct inode *inode, struct buffer_head *old_bh, if (unlikely(!new_bh)) { ext2_free_blocks(inode, block, 1); mark_inode_dirty(inode); - error = -EIO; + error = -ENOMEM; goto cleanup; } lock_buffer(new_bh); -- cgit From c04e88e271ab67de1409c3b4a4e80dbe13eac7b0 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Tue, 15 Jan 2013 21:20:01 -0800 Subject: Ext3: return ENOMEM rather than EIO if sb_getblk fails It will be better to use ENOMEM rather than EIO, because the only reason that sb_getblk fails is that allocation fails. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext3/inode.c | 4 ++-- fs/ext3/resize.c | 6 +++--- fs/ext3/xattr.c | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index d7df06839f6a..d512c4bc4ad7 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1083,7 +1083,7 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode *inode, struct buffer_head *bh; bh = sb_getblk(inode->i_sb, dummy.b_blocknr); if (unlikely(!bh)) { - *errp = -EIO; + *errp = -ENOMEM; goto err; } if (buffer_new(&dummy)) { @@ -2738,7 +2738,7 @@ static int __ext3_get_inode_loc(struct inode *inode, "unable to read inode block - " "inode=%lu, block="E3FSBLK, inode->i_ino, block); - return -EIO; + return -ENOMEM; } if (!buffer_uptodate(bh)) { lock_buffer(bh); diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 704e8ce7d782..27105655502c 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -117,7 +117,7 @@ static struct buffer_head *bclean(handle_t *handle, struct super_block *sb, bh = sb_getblk(sb, blk); if (unlikely(!bh)) - return ERR_PTR(-EIO); + return ERR_PTR(-ENOMEM); if ((err = ext3_journal_get_write_access(handle, bh))) { brelse(bh); bh = ERR_PTR(err); @@ -235,7 +235,7 @@ static int setup_new_group_blocks(struct super_block *sb, gdb = sb_getblk(sb, block); if (unlikely(!gdb)) { - err = -EIO; + err = -ENOMEM; goto exit_bh; } if ((err = ext3_journal_get_write_access(handle, gdb))) { @@ -723,7 +723,7 @@ static void update_backups(struct super_block *sb, bh = sb_getblk(sb, group * bpg + blk_off); if (unlikely(!bh)) { - err = -EIO; + err = -ENOMEM; break; } ext3_debug("update metadata backup %#04lx\n", diff --git a/fs/ext3/xattr.c b/fs/ext3/xattr.c index 9f57470b1727..b1fc96383e08 100644 --- a/fs/ext3/xattr.c +++ b/fs/ext3/xattr.c @@ -816,7 +816,7 @@ inserted: if (unlikely(!new_bh)) { getblk_failed: ext3_free_blocks(handle, inode, block, 1); - error = -EIO; + error = -ENOMEM; goto cleanup; } lock_buffer(new_bh); -- cgit From 9734c971aa6be6db61226b0046e080ca10383748 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 17 Jan 2013 22:11:38 +0100 Subject: udf: Write LVID to disk after opening / closing So far we just marked the buffer as dirty and left writing on flusher thread but especially on opening that opens possible race window where we could write other modified fs structures to disk before we mark filesystem as open. So sync LVID buffer to disk after opening and closing fs. Reported-by: Steve Nickel Signed-off-by: Jan Kara --- fs/udf/super.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/udf/super.c b/fs/udf/super.c index e9be396a558d..186adbf94b20 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -1866,6 +1866,8 @@ static void udf_open_lvid(struct super_block *sb) mark_buffer_dirty(bh); sbi->s_lvid_dirty = 0; mutex_unlock(&sbi->s_alloc_mutex); + /* Make opening of filesystem visible on the media immediately */ + sync_dirty_buffer(bh); } static void udf_close_lvid(struct super_block *sb) @@ -1906,6 +1908,8 @@ static void udf_close_lvid(struct super_block *sb) mark_buffer_dirty(bh); sbi->s_lvid_dirty = 0; mutex_unlock(&sbi->s_alloc_mutex); + /* Make closing of filesystem visible on the media immediately */ + sync_dirty_buffer(bh); } u64 lvid_get_unique_id(struct super_block *sb) -- cgit From c90419753306fa9a8fab5ab075c74b4d77b4f987 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:02 -0800 Subject: fs/9p: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: Eric Van Hensbergen CC: Ron Minnich CC: Latchesar Ionkov Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/9p/Kconfig | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/9p/Kconfig b/fs/9p/Kconfig index 0a93dc1cb4ac..55abfd62654a 100644 --- a/fs/9p/Kconfig +++ b/fs/9p/Kconfig @@ -11,8 +11,7 @@ config 9P_FS if 9P_FS config 9P_FSCACHE - bool "Enable 9P client caching support (EXPERIMENTAL)" - depends on EXPERIMENTAL + bool "Enable 9P client caching support" depends on 9P_FS=m && FSCACHE || 9P_FS=y && FSCACHE=y help Choose Y here to enable persistent, read-only local -- cgit From acd4fd07d4800e7c3b5243d1a260a9a84094ff50 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:03 -0800 Subject: fs/adfs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. Acked-by: Stuart Swales Cc: Russell King Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/adfs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/adfs/Kconfig b/fs/adfs/Kconfig index e55182a74605..c5a7787dd5e9 100644 --- a/fs/adfs/Kconfig +++ b/fs/adfs/Kconfig @@ -1,6 +1,6 @@ config ADFS_FS - tristate "ADFS file system support (EXPERIMENTAL)" - depends on BLOCK && EXPERIMENTAL + tristate "ADFS file system support" + depends on BLOCK help The Acorn Disc Filing System is the standard file system of the RiscOS operating system which runs on Acorn's ARM-based Risc PC -- cgit From 6d7a19fa747320e2bbe69736cfe9394492cb57f0 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:04 -0800 Subject: fs/affs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. Cc: Artem Bityutskiy Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/affs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/affs/Kconfig b/fs/affs/Kconfig index cfad9afb4762..a04d9e848d05 100644 --- a/fs/affs/Kconfig +++ b/fs/affs/Kconfig @@ -1,6 +1,6 @@ config AFFS_FS - tristate "Amiga FFS file system support (EXPERIMENTAL)" - depends on BLOCK && EXPERIMENTAL + tristate "Amiga FFS file system support" + depends on BLOCK help The Fast File System (FFS) is the common file system used on hard disks by Amiga(tm) systems since AmigaOS Version 1.3 (34.20). Say Y -- cgit From 8dbd5d6df380e8561d513e109516d8a98792fe45 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:05 -0800 Subject: fs/afs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: David Howells Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/afs/Kconfig | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/afs/Kconfig b/fs/afs/Kconfig index 8f975f25b486..ebba3b18e5da 100644 --- a/fs/afs/Kconfig +++ b/fs/afs/Kconfig @@ -1,6 +1,6 @@ config AFS_FS - tristate "Andrew File System support (AFS) (EXPERIMENTAL)" - depends on INET && EXPERIMENTAL + tristate "Andrew File System support (AFS)" + depends on INET select AF_RXRPC select DNS_RESOLVER help @@ -22,8 +22,7 @@ config AFS_DEBUG If unsure, say N. config AFS_FSCACHE - bool "Provide AFS client caching support (EXPERIMENTAL)" - depends on EXPERIMENTAL + bool "Provide AFS client caching support" depends on AFS_FS=m && FSCACHE || AFS_FS=y && FSCACHE=y help Say Y here if you want AFS data to be cached locally on disk through -- cgit From 5c8e0226e7363d9a15dda8a9307ac549baa531a0 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:06 -0800 Subject: fs/befs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/befs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/befs/Kconfig b/fs/befs/Kconfig index 7835d30f211f..edc5cc2aefad 100644 --- a/fs/befs/Kconfig +++ b/fs/befs/Kconfig @@ -1,6 +1,6 @@ config BEFS_FS - tristate "BeOS file system (BeFS) support (read only) (EXPERIMENTAL)" - depends on BLOCK && EXPERIMENTAL + tristate "BeOS file system (BeFS) support (read only)" + depends on BLOCK select NLS help The BeOS File System (BeFS) is the native file system of Be, Inc's -- cgit From d0e09c80f0391a9a5fe1b2a650eaaed363f85bff Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:07 -0800 Subject: fs/bfs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. Cc: "Tigran A. Aivazian" Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/bfs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/bfs/Kconfig b/fs/bfs/Kconfig index c2336c62024f..3728a6479c64 100644 --- a/fs/bfs/Kconfig +++ b/fs/bfs/Kconfig @@ -1,6 +1,6 @@ config BFS_FS - tristate "BFS file system support (EXPERIMENTAL)" - depends on BLOCK && EXPERIMENTAL + tristate "BFS file system support" + depends on BLOCK help Boot File System (BFS) is a file system used under SCO UnixWare to allow the bootloader access to the kernel image and other important -- cgit From 38db331b578005d32155bb6f6a80654ef127cff5 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:08 -0800 Subject: fs/btrfs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. Cc: Chris Mason Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/btrfs/Kconfig | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/Kconfig b/fs/btrfs/Kconfig index d33f01c08b60..ccd25ba7a9ac 100644 --- a/fs/btrfs/Kconfig +++ b/fs/btrfs/Kconfig @@ -1,6 +1,5 @@ config BTRFS_FS - tristate "Btrfs filesystem (EXPERIMENTAL) Unstable disk format" - depends on EXPERIMENTAL + tristate "Btrfs filesystem Unstable disk format" select LIBCRC32C select ZLIB_INFLATE select ZLIB_DEFLATE -- cgit From 00f3616b256c047c39d131772a0c504a4f1ba66f Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:09 -0800 Subject: fs/cifs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: Steve French CC: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/cifs/Kconfig | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index 21ff76c22a17..2906ee276408 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -155,14 +155,14 @@ config CIFS_DFS_UPCALL points. If unsure, say N. config CIFS_NFSD_EXPORT - bool "Allow nfsd to export CIFS file system (EXPERIMENTAL)" - depends on CIFS && EXPERIMENTAL && BROKEN + bool "Allow nfsd to export CIFS file system" + depends on CIFS && BROKEN help Allows NFS server to export a CIFS mounted share (nfsd over cifs) config CIFS_SMB2 - bool "SMB2 network file system support (EXPERIMENTAL)" - depends on CIFS && EXPERIMENTAL && INET + bool "SMB2 network file system support" + depends on CIFS && INET select NLS select KEYS select FSCACHE -- cgit From 462f16a5579319676f260c0bdde4e405c08c7ee3 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:10 -0800 Subject: fs/efs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. Cc: Neil Brown Cc: "J. Bruce Fields" Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/efs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/efs/Kconfig b/fs/efs/Kconfig index 6ebfc1c207a8..d020e3c30fea 100644 --- a/fs/efs/Kconfig +++ b/fs/efs/Kconfig @@ -1,6 +1,6 @@ config EFS_FS - tristate "EFS file system support (read only) (EXPERIMENTAL)" - depends on BLOCK && EXPERIMENTAL + tristate "EFS file system support (read only)" + depends on BLOCK help EFS is an older file system used for non-ISO9660 CD-ROMs and hard disk partitions by SGI's IRIX operating system (IRIX 6.0 and newer -- cgit From adae07485e164df62bc56535b089946e35074281 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:11 -0800 Subject: fs/hfs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. Cc: Artem Bityutskiy Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/hfs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/hfs/Kconfig b/fs/hfs/Kconfig index b77c5bc20f8a..998e3a6decf3 100644 --- a/fs/hfs/Kconfig +++ b/fs/hfs/Kconfig @@ -1,6 +1,6 @@ config HFS_FS - tristate "Apple Macintosh file system support (EXPERIMENTAL)" - depends on BLOCK && EXPERIMENTAL + tristate "Apple Macintosh file system support" + depends on BLOCK select NLS help If you say Y here, you will be able to mount Macintosh-formatted -- cgit From cf98c5e56807af7c793497bbd0446f71c383ffd5 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:12 -0800 Subject: fs/jffs2: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: David Woodhouse Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/jffs2/Kconfig | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/jffs2/Kconfig b/fs/jffs2/Kconfig index 6ae169cd8faa..d8bb6c411e96 100644 --- a/fs/jffs2/Kconfig +++ b/fs/jffs2/Kconfig @@ -50,8 +50,8 @@ config JFFS2_FS_WBUF_VERIFY write-buffer, and check for errors. config JFFS2_SUMMARY - bool "JFFS2 summary support (EXPERIMENTAL)" - depends on JFFS2_FS && EXPERIMENTAL + bool "JFFS2 summary support" + depends on JFFS2_FS default n help This feature makes it possible to use summary information @@ -63,8 +63,8 @@ config JFFS2_SUMMARY If unsure, say 'N'. config JFFS2_FS_XATTR - bool "JFFS2 XATTR support (EXPERIMENTAL)" - depends on JFFS2_FS && EXPERIMENTAL + bool "JFFS2 XATTR support" + depends on JFFS2_FS default n help Extended attributes are name:value pairs associated with inodes by @@ -173,7 +173,7 @@ config JFFS2_CMODE_PRIORITY successful one. config JFFS2_CMODE_SIZE - bool "size (EXPERIMENTAL)" + bool "size" help Tries all compressors and chooses the one which has the smallest result. -- cgit From 03edef0fea40ae003dd0d4cb337070913265e387 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:13 -0800 Subject: fs/logfs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: Joern Engel CC: Prasad Joshi Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/logfs/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/logfs/Kconfig b/fs/logfs/Kconfig index daf9a9b32dd3..09ed066c0221 100644 --- a/fs/logfs/Kconfig +++ b/fs/logfs/Kconfig @@ -1,6 +1,6 @@ config LOGFS - tristate "LogFS file system (EXPERIMENTAL)" - depends on (MTD || BLOCK) && EXPERIMENTAL + tristate "LogFS file system" + depends on (MTD || BLOCK) select ZLIB_INFLATE select ZLIB_DEFLATE select CRC32 -- cgit From f987c902575b5ee7397feb7d5568486b975a0589 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:14 -0800 Subject: fs/nfsd: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: "J. Bruce Fields" Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/nfsd/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig index 8df1ea4a6ff9..430b6872806f 100644 --- a/fs/nfsd/Kconfig +++ b/fs/nfsd/Kconfig @@ -65,8 +65,8 @@ config NFSD_V3_ACL If unsure, say N. config NFSD_V4 - bool "NFS server support for NFS version 4 (EXPERIMENTAL)" - depends on NFSD && PROC_FS && EXPERIMENTAL + bool "NFS server support for NFS version 4" + depends on NFSD && PROC_FS select NFSD_V3 select FS_POSIX_ACL select SUNRPC_GSS -- cgit From b7e17a10edf3638bb3e42621389448892a9cce6e Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 16 Jan 2013 18:54:15 -0800 Subject: fs/ufs: remove depends on CONFIG_EXPERIMENTAL The CONFIG_EXPERIMENTAL config item has not carried much meaning for a while now and is almost always enabled by default. As agreed during the Linux kernel summit, remove it from any "depends on" lines in Kconfigs. CC: Evgeniy Dushistov Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman --- fs/ufs/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ufs/Kconfig b/fs/ufs/Kconfig index e4f10a40768a..0bf6e16f8d79 100644 --- a/fs/ufs/Kconfig +++ b/fs/ufs/Kconfig @@ -29,7 +29,7 @@ config UFS_FS config UFS_FS_WRITE bool "UFS file system write support (DANGEROUS)" - depends on UFS_FS && EXPERIMENTAL + depends on UFS_FS help Say Y here if you want to try writing to UFS partitions. This is experimental, so you should back up your UFS partitions beforehand. -- cgit From 99600051b04bc4ec8bd4d16a8bf993ca54042db6 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sat, 19 Jan 2013 11:17:14 +0900 Subject: udf: add extent cache support in case of file reading This patch implements extent caching in case of file reading. While reading a file, currently, UDF reads metadata serially which takes a lot of time depending on the number of extents present in the file. Caching last accessd extent improves metadata read time. Instead of reading file metadata from start, now we read from the cached extent. This patch considerably improves the time spent by CPU in kernel mode. For example, while reading a 10.9 GB file using dd: Time before applying patch: 11677022208 bytes (10.9GB) copied, 1529.748921 seconds, 7.3MB/s real 25m 29.85s user 0m 12.41s sys 15m 34.75s Time after applying patch: 11677022208 bytes (10.9GB) copied, 1469.338231 seconds, 7.6MB/s real 24m 29.44s user 0m 15.73s sys 3m 27.61s [JK: Fix bh refcounting issues, simplify initialization] Signed-off-by: Namjae Jeon Signed-off-by: Ashish Sangwan Signed-off-by: Bonggil Bak Signed-off-by: Jan Kara --- fs/udf/inode.c | 86 ++++++++++++++++++++++++++++++++++++++++++++++++++++---- fs/udf/super.c | 2 ++ fs/udf/udf_i.h | 16 +++++++++++ fs/udf/udfdecl.h | 5 ---- 4 files changed, 98 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/udf/inode.c b/fs/udf/inode.c index cbae1ed0b7c1..7a12e48ad819 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -67,6 +67,74 @@ static void udf_update_extents(struct inode *, struct extent_position *); static int udf_get_block(struct inode *, sector_t, struct buffer_head *, int); +static void __udf_clear_extent_cache(struct inode *inode) +{ + struct udf_inode_info *iinfo = UDF_I(inode); + + if (iinfo->cached_extent.lstart != -1) { + brelse(iinfo->cached_extent.epos.bh); + iinfo->cached_extent.lstart = -1; + } +} + +/* Invalidate extent cache */ +static void udf_clear_extent_cache(struct inode *inode) +{ + struct udf_inode_info *iinfo = UDF_I(inode); + + spin_lock(&iinfo->i_extent_cache_lock); + __udf_clear_extent_cache(inode); + spin_unlock(&iinfo->i_extent_cache_lock); +} + +/* Return contents of extent cache */ +static int udf_read_extent_cache(struct inode *inode, loff_t bcount, + loff_t *lbcount, struct extent_position *pos) +{ + struct udf_inode_info *iinfo = UDF_I(inode); + int ret = 0; + + spin_lock(&iinfo->i_extent_cache_lock); + if ((iinfo->cached_extent.lstart <= bcount) && + (iinfo->cached_extent.lstart != -1)) { + /* Cache hit */ + *lbcount = iinfo->cached_extent.lstart; + memcpy(pos, &iinfo->cached_extent.epos, + sizeof(struct extent_position)); + if (pos->bh) + get_bh(pos->bh); + ret = 1; + } + spin_unlock(&iinfo->i_extent_cache_lock); + return ret; +} + +/* Add extent to extent cache */ +static void udf_update_extent_cache(struct inode *inode, loff_t estart, + struct extent_position *pos, int next_epos) +{ + struct udf_inode_info *iinfo = UDF_I(inode); + + spin_lock(&iinfo->i_extent_cache_lock); + /* Invalidate previously cached extent */ + __udf_clear_extent_cache(inode); + if (pos->bh) + get_bh(pos->bh); + memcpy(&iinfo->cached_extent.epos, pos, + sizeof(struct extent_position)); + iinfo->cached_extent.lstart = estart; + if (next_epos) + switch (iinfo->i_alloc_type) { + case ICBTAG_FLAG_AD_SHORT: + iinfo->cached_extent.epos.offset -= + sizeof(struct short_ad); + break; + case ICBTAG_FLAG_AD_LONG: + iinfo->cached_extent.epos.offset -= + sizeof(struct long_ad); + } + spin_unlock(&iinfo->i_extent_cache_lock); +} void udf_evict_inode(struct inode *inode) { @@ -90,6 +158,7 @@ void udf_evict_inode(struct inode *inode) } kfree(iinfo->i_ext.i_data); iinfo->i_ext.i_data = NULL; + udf_clear_extent_cache(inode); if (want_delete) { udf_free_inode(inode); } @@ -105,6 +174,7 @@ static void udf_write_failed(struct address_space *mapping, loff_t to) truncate_pagecache(inode, to, isize); if (iinfo->i_alloc_type != ICBTAG_FLAG_AD_IN_ICB) { down_write(&iinfo->i_data_sem); + udf_clear_extent_cache(inode); udf_truncate_extents(inode); up_write(&iinfo->i_data_sem); } @@ -372,7 +442,7 @@ static int udf_get_block(struct inode *inode, sector_t block, iinfo->i_next_alloc_goal++; } - + udf_clear_extent_cache(inode); phys = inode_getblk(inode, block, &err, &new); if (!phys) goto abort; @@ -1171,6 +1241,7 @@ set_size: } else { if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) { down_write(&iinfo->i_data_sem); + udf_clear_extent_cache(inode); memset(iinfo->i_ext.i_data + iinfo->i_lenEAttr + newsize, 0x00, bsize - newsize - udf_file_entry_alloc_offset(inode)); @@ -1184,6 +1255,7 @@ set_size: if (err) return err; down_write(&iinfo->i_data_sem); + udf_clear_extent_cache(inode); truncate_setsize(inode, newsize); udf_truncate_extents(inode); up_write(&iinfo->i_data_sem); @@ -2156,11 +2228,12 @@ int8_t inode_bmap(struct inode *inode, sector_t block, struct udf_inode_info *iinfo; iinfo = UDF_I(inode); - pos->offset = 0; - pos->block = iinfo->i_location; - pos->bh = NULL; + if (!udf_read_extent_cache(inode, bcount, &lbcount, pos)) { + pos->offset = 0; + pos->block = iinfo->i_location; + pos->bh = NULL; + } *elen = 0; - do { etype = udf_next_aext(inode, pos, eloc, elen, 1); if (etype == -1) { @@ -2170,7 +2243,8 @@ int8_t inode_bmap(struct inode *inode, sector_t block, } lbcount += *elen; } while (lbcount <= bcount); - + /* update extent cache */ + udf_update_extent_cache(inode, lbcount - *elen, pos, 1); *offset = (bcount + *elen - lbcount) >> blocksize_bits; return etype; diff --git a/fs/udf/super.c b/fs/udf/super.c index 186adbf94b20..da8ce9f14387 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -134,6 +134,8 @@ static struct inode *udf_alloc_inode(struct super_block *sb) ei->i_next_alloc_goal = 0; ei->i_strat4096 = 0; init_rwsem(&ei->i_data_sem); + ei->cached_extent.lstart = -1; + spin_lock_init(&ei->i_extent_cache_lock); return &ei->vfs_inode; } diff --git a/fs/udf/udf_i.h b/fs/udf/udf_i.h index bb8309dcd5c1..b5cd8ed2aa12 100644 --- a/fs/udf/udf_i.h +++ b/fs/udf/udf_i.h @@ -1,6 +1,19 @@ #ifndef _UDF_I_H #define _UDF_I_H +struct extent_position { + struct buffer_head *bh; + uint32_t offset; + struct kernel_lb_addr block; +}; + +struct udf_ext_cache { + /* Extent position */ + struct extent_position epos; + /* Start logical offset in bytes */ + loff_t lstart; +}; + /* * The i_data_sem and i_mutex serve for protection of allocation information * of a regular files and symlinks. This includes all extents belonging to @@ -35,6 +48,9 @@ struct udf_inode_info { __u8 *i_data; } i_ext; struct rw_semaphore i_data_sem; + struct udf_ext_cache cached_extent; + /* Spinlock for protecting extent cache */ + spinlock_t i_extent_cache_lock; struct inode vfs_inode; }; diff --git a/fs/udf/udfdecl.h b/fs/udf/udfdecl.h index de038da6f6bd..be7dabbbcb49 100644 --- a/fs/udf/udfdecl.h +++ b/fs/udf/udfdecl.h @@ -113,11 +113,6 @@ struct ustr { uint8_t u_len; }; -struct extent_position { - struct buffer_head *bh; - uint32_t offset; - struct kernel_lb_addr block; -}; /* super.c */ -- cgit From 84822d0b3bc5a74a4290727dd1ab4fc7dcd6a348 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 14 Dec 2012 17:57:50 -0500 Subject: nfsd4: simplify nfsd4_encode_fattr interface slightly It seems slightly simpler to make nfsd4_encode_fattr rather than its callers responsible for advancing the write pointer on success. (Also: the count == 0 check in the verify case looks superfluous. Running out of buffer space is really the only reason fattr encoding should fail with eresource.) Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 7 ++++--- fs/nfsd/nfs4xdr.c | 21 ++++++++------------- fs/nfsd/xdr4.h | 2 +- 3 files changed, 13 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 9d1c5dba2bbb..ae73175e6e68 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -993,14 +993,15 @@ _nfsd4_verify(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (!buf) return nfserr_jukebox; + p = buf; status = nfsd4_encode_fattr(&cstate->current_fh, cstate->current_fh.fh_export, - cstate->current_fh.fh_dentry, buf, - &count, verify->ve_bmval, + cstate->current_fh.fh_dentry, &p, + count, verify->ve_bmval, rqstp, 0); /* this means that nfsd4_encode_fattr() ran out of space */ - if (status == nfserr_resource && count == 0) + if (status == nfserr_resource) status = nfserr_not_same; if (status) goto out_kfree; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 0dc11586682f..fcb5bed99c33 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -2006,12 +2006,11 @@ static int get_parent_attributes(struct svc_export *exp, struct kstat *stat) * Note: @fhp can be NULL; in this case, we might have to compose the filehandle * ourselves. * - * @countp is the buffer size in _words_; upon successful return this becomes - * replaced with the number of words written. + * countp is the buffer size in _words_ */ __be32 nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, - struct dentry *dentry, __be32 *buffer, int *countp, u32 *bmval, + struct dentry *dentry, __be32 **buffer, int count, u32 *bmval, struct svc_rqst *rqstp, int ignore_crossmnt) { u32 bmval0 = bmval[0]; @@ -2020,12 +2019,12 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, struct kstat stat; struct svc_fh tempfh; struct kstatfs statfs; - int buflen = *countp << 2; + int buflen = count << 2; __be32 *attrlenp; u32 dummy; u64 dummy64; u32 rdattr_err = 0; - __be32 *p = buffer; + __be32 *p = *buffer; __be32 status; int err; int aclsupport = 0; @@ -2431,7 +2430,7 @@ out_acl: } *attrlenp = htonl((char *)p - (char *)attrlenp - 4); - *countp = p - buffer; + *buffer = p; status = nfs_ok; out: @@ -2443,7 +2442,6 @@ out_nfserr: status = nfserrno(err); goto out; out_resource: - *countp = 0; status = nfserr_resource; goto out; out_serverfault: @@ -2462,7 +2460,7 @@ static inline int attributes_need_mount(u32 *bmval) static __be32 nfsd4_encode_dirent_fattr(struct nfsd4_readdir *cd, - const char *name, int namlen, __be32 *p, int *buflen) + const char *name, int namlen, __be32 **p, int buflen) { struct svc_export *exp = cd->rd_fhp->fh_export; struct dentry *dentry; @@ -2568,10 +2566,9 @@ nfsd4_encode_dirent(void *ccdv, const char *name, int namlen, p = xdr_encode_hyper(p, NFS_OFFSET_MAX); /* offset of next entry */ p = xdr_encode_array(p, name, namlen); /* name length & name */ - nfserr = nfsd4_encode_dirent_fattr(cd, name, namlen, p, &buflen); + nfserr = nfsd4_encode_dirent_fattr(cd, name, namlen, &p, buflen); switch (nfserr) { case nfs_ok: - p += buflen; break; case nfserr_resource: nfserr = nfserr_toosmall; @@ -2698,10 +2695,8 @@ nfsd4_encode_getattr(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4 buflen = resp->end - resp->p - (COMPOUND_ERR_SLACK_SPACE >> 2); nfserr = nfsd4_encode_fattr(fhp, fhp->fh_export, fhp->fh_dentry, - resp->p, &buflen, getattr->ga_bmval, + &resp->p, buflen, getattr->ga_bmval, resp->rqstp, 0); - if (!nfserr) - resp->p += buflen; return nfserr; } diff --git a/fs/nfsd/xdr4.h b/fs/nfsd/xdr4.h index 0889bfb43dc9..546f8983ecf1 100644 --- a/fs/nfsd/xdr4.h +++ b/fs/nfsd/xdr4.h @@ -563,7 +563,7 @@ __be32 nfsd4_check_resp_size(struct nfsd4_compoundres *, u32); void nfsd4_encode_operation(struct nfsd4_compoundres *, struct nfsd4_op *); void nfsd4_encode_replay(struct nfsd4_compoundres *resp, struct nfsd4_op *op); __be32 nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, - struct dentry *dentry, __be32 *buffer, int *countp, + struct dentry *dentry, __be32 **buffer, int countp, u32 *bmval, struct svc_rqst *, int ignore_crossmnt); extern __be32 nfsd4_setclientid(struct svc_rqst *rqstp, struct nfsd4_compound_state *, -- cgit From 74b70dded311fa0e6e7529514b29bbb8e6bb1f3e Mon Sep 17 00:00:00 2001 From: Yanchuan Nian Date: Mon, 24 Dec 2012 18:11:27 +0800 Subject: nfsd: Pass correct slot number to nfsd4_put_drc_mem() In alloc_session(), numslots is the correct slot number used by the session. But the slot number passed to nfsd4_put_drc_mem() is the one from nfs client. Signed-off-by: Yanchuan Nian Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index ac8ed96c4199..29924a04cf3d 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -905,7 +905,7 @@ static struct nfsd4_session *alloc_session(struct nfsd4_channel_attrs *fchan, new = __alloc_session(slotsize, numslots); if (!new) { - nfsd4_put_drc_mem(slotsize, fchan->maxreqs); + nfsd4_put_drc_mem(slotsize, numslots); return NULL; } init_forechannel_attrs(&new->se_fchannel, fchan, numslots, slotsize, nn); -- cgit From 266533c6df7a4a4e2ebd0bfdd272f7eb7cf4b81f Mon Sep 17 00:00:00 2001 From: Yanchuan Nian Date: Mon, 24 Dec 2012 18:11:45 +0800 Subject: nfsd: Don't unlock the state while it's not locked In the procedure of CREATE_SESSION, the state is locked after alloc_conn_from_crses(). If the allocation fails, the function goes to "out_free_session", and then "out" where there is an unlock function. Signed-off-by: Yanchuan Nian Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 29924a04cf3d..cc41bf4bcab2 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1844,11 +1844,12 @@ nfsd4_create_session(struct svc_rqst *rqstp, /* cache solo and embedded create sessions under the state lock */ nfsd4_cache_create_session(cr_ses, cs_slot, status); -out: nfs4_unlock_state(); +out: dprintk("%s returns %d\n", __func__, ntohl(status)); return status; out_free_conn: + nfs4_unlock_state(); free_conn(conn); out_free_session: __free_session(new); -- cgit From bca0ec6511bb96bcb6cb247fd4100a4ea1d1e4f5 Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Wed, 9 Jan 2013 12:38:34 +0300 Subject: nfsd: fix unused "nn" variable warning in free_client() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If CONFIG_LOCKDEP is disabled, then there would be a warning like this: CC [M] fs/nfsd/nfs4state.o fs/nfsd/nfs4state.c: In function ‘free_client’: fs/nfsd/nfs4state.c:1051:19: warning: unused variable ‘nn’ [-Wunused-variable] So, let's add "maybe_unused" tag to this variable. Reported-by: Toralf Förster Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index cc41bf4bcab2..4db46aa387b7 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1048,7 +1048,7 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name) static inline void free_client(struct nfs4_client *clp) { - struct nfsd_net *nn = net_generic(clp->net, nfsd_net_id); + struct nfsd_net __maybe_unused *nn = net_generic(clp->net, nfsd_net_id); lockdep_assert_held(&nn->client_lock); while (!list_empty(&clp->cl_sessions)) { -- cgit From ff89be87c70247ffe3a72271e02eb7765cdd12c4 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Wed, 23 Jan 2013 18:25:01 -0500 Subject: nfsd4: require version 4 when enabling or disabling minorversion The current code will allow silly things like: echo "+2 +3 +4 +7.1">/proc/fs/nfsd/versions Reported-by: Fan Chaoting Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 74934284d9a7..65889ec1d28e 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -534,7 +534,7 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) else num = simple_strtol(vers, &minorp, 0); if (*minorp == '.') { - if (num < 4) + if (num != 4) return -EINVAL; minor = simple_strtoul(minorp+1, NULL, 0); if (minor == 0) -- cgit From 0b05b18381eea98c9c9ada95629bf659a88c9374 Mon Sep 17 00:00:00 2001 From: "Anand V. Avati" Date: Sun, 19 Aug 2012 08:53:23 -0400 Subject: fuse: implement NFS-like readdirplus support This patch implements readdirplus support in FUSE, similar to NFS. The payload returned in the readdirplus call contains 'fuse_entry_out' structure thereby providing all the necessary inputs for 'faking' a lookup() operation on the spot. If the dentry and inode already existed (for e.g. in a re-run of ls -l) then just the inode attributes timeout and dentry timeout are refreshed. With a simple client->network->server implementation of a FUSE based filesystem, the following performance observations were made: Test: Performing a filesystem crawl over 20,000 files with sh# time ls -lR /mnt Without readdirplus: Run 1: 18.1s Run 2: 16.0s Run 3: 16.2s With readdirplus: Run 1: 4.1s Run 2: 3.8s Run 3: 3.8s The performance improvement is significant as it avoided 20,000 upcalls calls (lookup). Cache consistency is no worse than what already is. Signed-off-by: Anand V. Avati Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 19 +++++++ fs/fuse/dir.c | 160 +++++++++++++++++++++++++++++++++++++++++++++++++++++-- fs/fuse/fuse_i.h | 6 +++ fs/fuse/inode.c | 5 +- 4 files changed, 185 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index e83351aa5bad..05c3eec298f2 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -491,6 +491,25 @@ void fuse_request_send_background_locked(struct fuse_conn *fc, fuse_request_send_nowait_locked(fc, req); } +void fuse_force_forget(struct file *file, u64 nodeid) +{ + struct inode *inode = file->f_path.dentry->d_inode; + struct fuse_conn *fc = get_fuse_conn(inode); + struct fuse_req *req; + struct fuse_forget_in inarg; + + memset(&inarg, 0, sizeof(inarg)); + inarg.nlookup = 1; + req = fuse_get_req_nofail(fc, file); + req->in.h.opcode = FUSE_FORGET; + req->in.h.nodeid = nodeid; + req->in.numargs = 1; + req->in.args[0].size = sizeof(inarg); + req->in.args[0].value = &inarg; + req->isreply = 0; + fuse_request_send_nowait(fc, req); +} + /* * Lock the request. Up to the next unlock_request() there mustn't be * anything that could cause a page-fault. If the request was already diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index b7c09f9eb40c..dcc1e522c7d4 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1155,6 +1155,143 @@ static int parse_dirfile(char *buf, size_t nbytes, struct file *file, return 0; } +static int fuse_direntplus_link(struct file *file, + struct fuse_direntplus *direntplus, + u64 attr_version) +{ + int err; + struct fuse_entry_out *o = &direntplus->entry_out; + struct fuse_dirent *dirent = &direntplus->dirent; + struct dentry *parent = file->f_path.dentry; + struct qstr name = QSTR_INIT(dirent->name, dirent->namelen); + struct dentry *dentry; + struct dentry *alias; + struct inode *dir = parent->d_inode; + struct fuse_conn *fc; + struct inode *inode; + + if (!o->nodeid) { + /* + * Unlike in the case of fuse_lookup, zero nodeid does not mean + * ENOENT. Instead, it only means the userspace filesystem did + * not want to return attributes/handle for this entry. + * + * So do nothing. + */ + return 0; + } + + if (name.name[0] == '.') { + /* + * We could potentially refresh the attributes of the directory + * and its parent? + */ + if (name.len == 1) + return 0; + if (name.name[1] == '.' && name.len == 2) + return 0; + } + fc = get_fuse_conn(dir); + + name.hash = full_name_hash(name.name, name.len); + dentry = d_lookup(parent, &name); + if (dentry && dentry->d_inode) { + inode = dentry->d_inode; + if (get_node_id(inode) == o->nodeid) { + struct fuse_inode *fi; + fi = get_fuse_inode(inode); + spin_lock(&fc->lock); + fi->nlookup++; + spin_unlock(&fc->lock); + + /* + * The other branch to 'found' comes via fuse_iget() + * which bumps nlookup inside + */ + goto found; + } + err = d_invalidate(dentry); + if (err) + goto out; + dput(dentry); + dentry = NULL; + } + + dentry = d_alloc(parent, &name); + err = -ENOMEM; + if (!dentry) + goto out; + + inode = fuse_iget(dir->i_sb, o->nodeid, o->generation, + &o->attr, entry_attr_timeout(o), attr_version); + if (!inode) + goto out; + + alias = d_materialise_unique(dentry, inode); + err = PTR_ERR(alias); + if (IS_ERR(alias)) + goto out; + if (alias) { + dput(dentry); + dentry = alias; + } + +found: + fuse_change_attributes(inode, &o->attr, entry_attr_timeout(o), + attr_version); + + fuse_change_entry_timeout(dentry, o); + + err = 0; +out: + if (dentry) + dput(dentry); + return err; +} + +static int parse_dirplusfile(char *buf, size_t nbytes, struct file *file, + void *dstbuf, filldir_t filldir, u64 attr_version) +{ + struct fuse_direntplus *direntplus; + struct fuse_dirent *dirent; + size_t reclen; + int over = 0; + int ret; + + while (nbytes >= FUSE_NAME_OFFSET_DIRENTPLUS) { + direntplus = (struct fuse_direntplus *) buf; + dirent = &direntplus->dirent; + reclen = FUSE_DIRENTPLUS_SIZE(direntplus); + + if (!dirent->namelen || dirent->namelen > FUSE_NAME_MAX) + return -EIO; + if (reclen > nbytes) + break; + + if (!over) { + /* We fill entries into dstbuf only as much as + it can hold. But we still continue iterating + over remaining entries to link them. If not, + we need to send a FORGET for each of those + which we did not link. + */ + over = filldir(dstbuf, dirent->name, dirent->namelen, + file->f_pos, dirent->ino, + dirent->type); + file->f_pos = dirent->off; + } + + buf += reclen; + nbytes -= reclen; + + ret = fuse_direntplus_link(file, direntplus, attr_version); + if (ret) + fuse_force_forget(file, direntplus->entry_out.nodeid); + } + + return 0; +} + static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) { int err; @@ -1163,6 +1300,7 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) struct inode *inode = file->f_path.dentry->d_inode; struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_req *req; + u64 attr_version = 0; if (is_bad_inode(inode)) return -EIO; @@ -1179,14 +1317,28 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) req->out.argpages = 1; req->num_pages = 1; req->pages[0] = page; - fuse_read_fill(req, file, file->f_pos, PAGE_SIZE, FUSE_READDIR); + if (fc->do_readdirplus) { + attr_version = fuse_get_attr_version(fc); + fuse_read_fill(req, file, file->f_pos, PAGE_SIZE, + FUSE_READDIRPLUS); + } else { + fuse_read_fill(req, file, file->f_pos, PAGE_SIZE, + FUSE_READDIR); + } fuse_request_send(fc, req); nbytes = req->out.args[0].size; err = req->out.h.error; fuse_put_request(fc, req); - if (!err) - err = parse_dirfile(page_address(page), nbytes, file, dstbuf, - filldir); + if (!err) { + if (fc->do_readdirplus) { + err = parse_dirplusfile(page_address(page), nbytes, + file, dstbuf, filldir, + attr_version); + } else { + err = parse_dirfile(page_address(page), nbytes, file, + dstbuf, filldir); + } + } __free_page(page); fuse_invalidate_attr(inode); /* atime changed */ diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index e105a53fc72d..5c5055306d3c 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -487,6 +487,9 @@ struct fuse_conn { /** Use enhanced/automatic page cache invalidation. */ unsigned auto_inval_data:1; + /** Does the filesystem support readdir-plus? */ + unsigned do_readdirplus:1; + /** The number of requests waiting for completion */ atomic_t num_waiting; @@ -578,6 +581,9 @@ void fuse_queue_forget(struct fuse_conn *fc, struct fuse_forget_link *forget, struct fuse_forget_link *fuse_alloc_forget(void); +/* Used by READDIRPLUS */ +void fuse_force_forget(struct file *file, u64 nodeid); + /** * Initialize READ or READDIR request */ diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 73ca6b72beaf..6f7d5746bf52 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -863,6 +863,8 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) fc->dont_mask = 1; if (arg->flags & FUSE_AUTO_INVAL_DATA) fc->auto_inval_data = 1; + if (arg->flags & FUSE_DO_READDIRPLUS) + fc->do_readdirplus = 1; } else { ra_pages = fc->max_read / PAGE_CACHE_SIZE; fc->no_lock = 1; @@ -889,7 +891,8 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) arg->flags |= FUSE_ASYNC_READ | FUSE_POSIX_LOCKS | FUSE_ATOMIC_O_TRUNC | FUSE_EXPORT_SUPPORT | FUSE_BIG_WRITES | FUSE_DONT_MASK | FUSE_SPLICE_WRITE | FUSE_SPLICE_MOVE | FUSE_SPLICE_READ | - FUSE_FLOCK_LOCKS | FUSE_IOCTL_DIR | FUSE_AUTO_INVAL_DATA; + FUSE_FLOCK_LOCKS | FUSE_IOCTL_DIR | FUSE_AUTO_INVAL_DATA | + FUSE_DO_READDIRPLUS; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); -- cgit From 4250c0668ea10a19f3d37b1733f54ce6c8a37234 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:48:07 +0400 Subject: fuse: general infrastructure for pages[] of variable size The patch removes inline array of FUSE_MAX_PAGES_PER_REQ page pointers from fuse_req. Instead of that, req->pages may now point either to small inline array or to an array allocated dynamically. This essentially means that all callers of fuse_request_alloc[_nofs] should pass the number of pages needed explicitly. The patch doesn't make any logic changes. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 47 ++++++++++++++++++++++++++++++++++------------- fs/fuse/file.c | 4 ++-- fs/fuse/fuse_i.h | 15 ++++++++++++--- fs/fuse/inode.c | 4 ++-- 4 files changed, 50 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 05c3eec298f2..af37ae138252 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -34,34 +34,55 @@ static struct fuse_conn *fuse_get_conn(struct file *file) return file->private_data; } -static void fuse_request_init(struct fuse_req *req) +static void fuse_request_init(struct fuse_req *req, struct page **pages, + unsigned npages) { memset(req, 0, sizeof(*req)); + memset(pages, 0, sizeof(*pages) * npages); INIT_LIST_HEAD(&req->list); INIT_LIST_HEAD(&req->intr_entry); init_waitqueue_head(&req->waitq); atomic_set(&req->count, 1); + req->pages = pages; + req->max_pages = npages; } -struct fuse_req *fuse_request_alloc(void) +static struct fuse_req *__fuse_request_alloc(unsigned npages, gfp_t flags) { - struct fuse_req *req = kmem_cache_alloc(fuse_req_cachep, GFP_KERNEL); - if (req) - fuse_request_init(req); + struct fuse_req *req = kmem_cache_alloc(fuse_req_cachep, flags); + if (req) { + struct page **pages; + + if (npages <= FUSE_REQ_INLINE_PAGES) + pages = req->inline_pages; + else + pages = kmalloc(sizeof(struct page *) * npages, flags); + + if (!pages) { + kmem_cache_free(fuse_req_cachep, req); + return NULL; + } + + fuse_request_init(req, pages, npages); + } return req; } + +struct fuse_req *fuse_request_alloc(unsigned npages) +{ + return __fuse_request_alloc(npages, GFP_KERNEL); +} EXPORT_SYMBOL_GPL(fuse_request_alloc); -struct fuse_req *fuse_request_alloc_nofs(void) +struct fuse_req *fuse_request_alloc_nofs(unsigned npages) { - struct fuse_req *req = kmem_cache_alloc(fuse_req_cachep, GFP_NOFS); - if (req) - fuse_request_init(req); - return req; + return __fuse_request_alloc(npages, GFP_NOFS); } void fuse_request_free(struct fuse_req *req) { + if (req->pages != req->inline_pages) + kfree(req->pages); kmem_cache_free(fuse_req_cachep, req); } @@ -116,7 +137,7 @@ struct fuse_req *fuse_get_req(struct fuse_conn *fc) if (!fc->connected) goto out; - req = fuse_request_alloc(); + req = fuse_request_alloc(FUSE_MAX_PAGES_PER_REQ); err = -ENOMEM; if (!req) goto out; @@ -165,7 +186,7 @@ static void put_reserved_req(struct fuse_conn *fc, struct fuse_req *req) struct fuse_file *ff = file->private_data; spin_lock(&fc->lock); - fuse_request_init(req); + fuse_request_init(req, req->pages, req->max_pages); BUG_ON(ff->reserved_req); ff->reserved_req = req; wake_up_all(&fc->reserved_req_waitq); @@ -192,7 +213,7 @@ struct fuse_req *fuse_get_req_nofail(struct fuse_conn *fc, struct file *file) atomic_inc(&fc->num_waiting); wait_event(fc->blocked_waitq, !fc->blocked); - req = fuse_request_alloc(); + req = fuse_request_alloc(FUSE_MAX_PAGES_PER_REQ); if (!req) req = get_reserved_req(fc, file); diff --git a/fs/fuse/file.c b/fs/fuse/file.c index f3ab824fa302..2565f635c04f 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -57,7 +57,7 @@ struct fuse_file *fuse_file_alloc(struct fuse_conn *fc) return NULL; ff->fc = fc; - ff->reserved_req = fuse_request_alloc(); + ff->reserved_req = fuse_request_alloc(0); if (unlikely(!ff->reserved_req)) { kfree(ff); return NULL; @@ -1272,7 +1272,7 @@ static int fuse_writepage_locked(struct page *page) set_page_writeback(page); - req = fuse_request_alloc_nofs(); + req = fuse_request_alloc_nofs(1); if (!req) goto err; diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 5c5055306d3c..0c5b9310f930 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -44,6 +44,9 @@ doing the mount will be allowed to access the filesystem */ #define FUSE_ALLOW_OTHER (1 << 1) +/** Number of page pointers embedded in fuse_req */ +#define FUSE_REQ_INLINE_PAGES 1 + /** List of active connections */ extern struct list_head fuse_conn_list; @@ -291,7 +294,13 @@ struct fuse_req { } misc; /** page vector */ - struct page *pages[FUSE_MAX_PAGES_PER_REQ]; + struct page **pages; + + /** size of the 'pages' array */ + unsigned max_pages; + + /** inline page vector */ + struct page *inline_pages[FUSE_REQ_INLINE_PAGES]; /** number of pages in vector */ unsigned num_pages; @@ -664,9 +673,9 @@ void fuse_ctl_cleanup(void); /** * Allocate a request */ -struct fuse_req *fuse_request_alloc(void); +struct fuse_req *fuse_request_alloc(unsigned npages); -struct fuse_req *fuse_request_alloc_nofs(void); +struct fuse_req *fuse_request_alloc_nofs(unsigned npages); /** * Free a request diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 6f7d5746bf52..9a937f0239e8 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -1037,12 +1037,12 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) /* only now - we want root dentry with NULL ->d_op */ sb->s_d_op = &fuse_dentry_operations; - init_req = fuse_request_alloc(); + init_req = fuse_request_alloc(0); if (!init_req) goto err_put_root; if (is_bdev) { - fc->destroy_req = fuse_request_alloc(); + fc->destroy_req = fuse_request_alloc(0); if (!fc->destroy_req) goto err_free_init_req; } -- cgit From b111c8c0e3e5e780ae0758fc4c1c376a7c9d5997 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:48:30 +0400 Subject: fuse: categorize fuse_get_req() The patch categorizes all fuse_get_req() invocations into two categories: - fuse_get_req_nopages(fc) - when caller doesn't care about req->pages - fuse_get_req(fc, n) - when caller need n page pointers (n > 0) Adding fuse_get_req_nopages() helps to avoid numerous fuse_get_req(fc, 0) scattered over code. Now it's clear from the first glance when a caller need fuse_req with page pointers. The patch doesn't make any logic changes. In multi-page case, it silly allocates array of FUSE_MAX_PAGES_PER_REQ page pointers. This will be amended by future patches. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/cuse.c | 2 +- fs/fuse/dev.c | 13 +++++++------ fs/fuse/dir.c | 38 +++++++++++++++++++------------------- fs/fuse/file.c | 30 +++++++++++++++--------------- fs/fuse/fuse_i.h | 17 ++++++++++++++--- fs/fuse/inode.c | 2 +- 6 files changed, 57 insertions(+), 45 deletions(-) (limited to 'fs') diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c index e397b675b029..5cc838f320d4 100644 --- a/fs/fuse/cuse.c +++ b/fs/fuse/cuse.c @@ -419,7 +419,7 @@ static int cuse_send_init(struct cuse_conn *cc) BUILD_BUG_ON(CUSE_INIT_INFO_MAX > PAGE_SIZE); - req = fuse_get_req(fc); + req = fuse_get_req(fc, 1); if (IS_ERR(req)) { rc = PTR_ERR(req); goto err; diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index af37ae138252..ff5e8bed5d88 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -118,7 +118,7 @@ static void fuse_req_init_context(struct fuse_req *req) req->in.h.pid = current->pid; } -struct fuse_req *fuse_get_req(struct fuse_conn *fc) +struct fuse_req *fuse_get_req(struct fuse_conn *fc, unsigned npages) { struct fuse_req *req; sigset_t oldset; @@ -137,7 +137,7 @@ struct fuse_req *fuse_get_req(struct fuse_conn *fc) if (!fc->connected) goto out; - req = fuse_request_alloc(FUSE_MAX_PAGES_PER_REQ); + req = fuse_request_alloc(npages); err = -ENOMEM; if (!req) goto out; @@ -207,13 +207,14 @@ static void put_reserved_req(struct fuse_conn *fc, struct fuse_req *req) * filesystem should not have it's own file open. If deadlock is * intentional, it can still be broken by "aborting" the filesystem. */ -struct fuse_req *fuse_get_req_nofail(struct fuse_conn *fc, struct file *file) +struct fuse_req *fuse_get_req_nofail_nopages(struct fuse_conn *fc, + struct file *file) { struct fuse_req *req; atomic_inc(&fc->num_waiting); wait_event(fc->blocked_waitq, !fc->blocked); - req = fuse_request_alloc(FUSE_MAX_PAGES_PER_REQ); + req = fuse_request_alloc(0); if (!req) req = get_reserved_req(fc, file); @@ -521,7 +522,7 @@ void fuse_force_forget(struct file *file, u64 nodeid) memset(&inarg, 0, sizeof(inarg)); inarg.nlookup = 1; - req = fuse_get_req_nofail(fc, file); + req = fuse_get_req_nofail_nopages(fc, file); req->in.h.opcode = FUSE_FORGET; req->in.h.nodeid = nodeid; req->in.numargs = 1; @@ -1577,7 +1578,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, unsigned int offset; size_t total_len = 0; - req = fuse_get_req(fc); + req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); if (IS_ERR(req)) return PTR_ERR(req); diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index dcc1e522c7d4..d04bcc5cccd0 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -178,7 +178,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, unsigned int flags) return -ECHILD; fc = get_fuse_conn(inode); - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return 0; @@ -271,7 +271,7 @@ int fuse_lookup_name(struct super_block *sb, u64 nodeid, struct qstr *name, if (name->len > FUSE_NAME_MAX) goto out; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); err = PTR_ERR(req); if (IS_ERR(req)) goto out; @@ -391,7 +391,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, if (!forget) goto out_err; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); err = PTR_ERR(req); if (IS_ERR(req)) goto out_put_forget_req; @@ -592,7 +592,7 @@ static int fuse_mknod(struct inode *dir, struct dentry *entry, umode_t mode, { struct fuse_mknod_in inarg; struct fuse_conn *fc = get_fuse_conn(dir); - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -623,7 +623,7 @@ static int fuse_mkdir(struct inode *dir, struct dentry *entry, umode_t mode) { struct fuse_mkdir_in inarg; struct fuse_conn *fc = get_fuse_conn(dir); - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -647,7 +647,7 @@ static int fuse_symlink(struct inode *dir, struct dentry *entry, { struct fuse_conn *fc = get_fuse_conn(dir); unsigned len = strlen(link) + 1; - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -664,7 +664,7 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry) { int err; struct fuse_conn *fc = get_fuse_conn(dir); - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -696,7 +696,7 @@ static int fuse_rmdir(struct inode *dir, struct dentry *entry) { int err; struct fuse_conn *fc = get_fuse_conn(dir); - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -723,7 +723,7 @@ static int fuse_rename(struct inode *olddir, struct dentry *oldent, int err; struct fuse_rename_in inarg; struct fuse_conn *fc = get_fuse_conn(olddir); - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -776,7 +776,7 @@ static int fuse_link(struct dentry *entry, struct inode *newdir, struct fuse_link_in inarg; struct inode *inode = entry->d_inode; struct fuse_conn *fc = get_fuse_conn(inode); - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -848,7 +848,7 @@ static int fuse_do_getattr(struct inode *inode, struct kstat *stat, struct fuse_req *req; u64 attr_version; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1029,7 +1029,7 @@ static int fuse_access(struct inode *inode, int mask) if (fc->no_access) return 0; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1305,7 +1305,7 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) if (is_bad_inode(inode)) return -EIO; - req = fuse_get_req(fc); + req = fuse_get_req(fc, 1); if (IS_ERR(req)) return PTR_ERR(req); @@ -1349,7 +1349,7 @@ static char *read_link(struct dentry *dentry) { struct inode *inode = dentry->d_inode; struct fuse_conn *fc = get_fuse_conn(inode); - struct fuse_req *req = fuse_get_req(fc); + struct fuse_req *req = fuse_get_req_nopages(fc); char *link; if (IS_ERR(req)) @@ -1562,7 +1562,7 @@ static int fuse_do_setattr(struct dentry *entry, struct iattr *attr, if (attr->ia_valid & ATTR_SIZE) is_truncate = true; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1670,7 +1670,7 @@ static int fuse_setxattr(struct dentry *entry, const char *name, if (fc->no_setxattr) return -EOPNOTSUPP; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1709,7 +1709,7 @@ static ssize_t fuse_getxattr(struct dentry *entry, const char *name, if (fc->no_getxattr) return -EOPNOTSUPP; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1761,7 +1761,7 @@ static ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size) if (fc->no_listxattr) return -EOPNOTSUPP; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1806,7 +1806,7 @@ static int fuse_removexattr(struct dentry *entry, const char *name) if (fc->no_removexattr) return -EOPNOTSUPP; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 2565f635c04f..882877125c1d 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -25,7 +25,7 @@ static int fuse_send_open(struct fuse_conn *fc, u64 nodeid, struct file *file, struct fuse_req *req; int err; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -368,7 +368,7 @@ static int fuse_flush(struct file *file, fl_owner_t id) if (fc->no_flush) return 0; - req = fuse_get_req_nofail(fc, file); + req = fuse_get_req_nofail_nopages(fc, file); memset(&inarg, 0, sizeof(inarg)); inarg.fh = ff->fh; inarg.lock_owner = fuse_lock_owner_id(fc, id); @@ -436,7 +436,7 @@ int fuse_fsync_common(struct file *file, loff_t start, loff_t end, fuse_sync_writes(inode); - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) { err = PTR_ERR(req); goto out; @@ -544,7 +544,7 @@ static int fuse_readpage(struct file *file, struct page *page) */ fuse_wait_on_page_writeback(inode, page->index); - req = fuse_get_req(fc); + req = fuse_get_req(fc, 1); err = PTR_ERR(req); if (IS_ERR(req)) goto out; @@ -657,7 +657,7 @@ static int fuse_readpages_fill(void *_data, struct page *page) (req->num_pages + 1) * PAGE_CACHE_SIZE > fc->max_read || req->pages[req->num_pages - 1]->index + 1 != page->index)) { fuse_send_readpages(req, data->file); - data->req = req = fuse_get_req(fc); + data->req = req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); if (IS_ERR(req)) { unlock_page(page); return PTR_ERR(req); @@ -683,7 +683,7 @@ static int fuse_readpages(struct file *file, struct address_space *mapping, data.file = file; data.inode = inode; - data.req = fuse_get_req(fc); + data.req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); err = PTR_ERR(data.req); if (IS_ERR(data.req)) goto out; @@ -890,7 +890,7 @@ static ssize_t fuse_perform_write(struct file *file, struct fuse_req *req; ssize_t count; - req = fuse_get_req(fc); + req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); if (IS_ERR(req)) { err = PTR_ERR(req); break; @@ -1072,7 +1072,7 @@ ssize_t fuse_direct_io(struct file *file, const char __user *buf, ssize_t res = 0; struct fuse_req *req; - req = fuse_get_req(fc); + req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); if (IS_ERR(req)) return PTR_ERR(req); @@ -1108,7 +1108,7 @@ ssize_t fuse_direct_io(struct file *file, const char __user *buf, break; if (count) { fuse_put_request(fc, req); - req = fuse_get_req(fc); + req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); if (IS_ERR(req)) break; } @@ -1471,7 +1471,7 @@ static int fuse_getlk(struct file *file, struct file_lock *fl) struct fuse_lk_out outarg; int err; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1506,7 +1506,7 @@ static int fuse_setlk(struct file *file, struct file_lock *fl, int flock) if (fl->fl_flags & FL_CLOSE) return 0; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); @@ -1575,7 +1575,7 @@ static sector_t fuse_bmap(struct address_space *mapping, sector_t block) if (!inode->i_sb->s_bdev || fc->no_bmap) return 0; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return 0; @@ -1873,7 +1873,7 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, num_pages++; } - req = fuse_get_req(fc); + req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); if (IS_ERR(req)) { err = PTR_ERR(req); req = NULL; @@ -2076,7 +2076,7 @@ unsigned fuse_file_poll(struct file *file, poll_table *wait) fuse_register_polled_file(fc, ff); } - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return POLLERR; @@ -2194,7 +2194,7 @@ static long fuse_file_fallocate(struct file *file, int mode, loff_t offset, if (fc->no_fallocate) return -EOPNOTSUPP; - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 0c5b9310f930..5b21e6ab9e75 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -683,14 +683,25 @@ struct fuse_req *fuse_request_alloc_nofs(unsigned npages); void fuse_request_free(struct fuse_req *req); /** - * Get a request, may fail with -ENOMEM + * Get a request, may fail with -ENOMEM, + * caller should specify # elements in req->pages[] explicitly */ -struct fuse_req *fuse_get_req(struct fuse_conn *fc); +struct fuse_req *fuse_get_req(struct fuse_conn *fc, unsigned npages); + +/** + * Get a request, may fail with -ENOMEM, + * useful for callers who doesn't use req->pages[] + */ +static inline struct fuse_req *fuse_get_req_nopages(struct fuse_conn *fc) +{ + return fuse_get_req(fc, 0); +} /** * Gets a requests for a file operation, always succeeds */ -struct fuse_req *fuse_get_req_nofail(struct fuse_conn *fc, struct file *file); +struct fuse_req *fuse_get_req_nofail_nopages(struct fuse_conn *fc, + struct file *file); /** * Decrement reference count of a request. If count goes to zero free diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 9a937f0239e8..9d95a5a3d55c 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -413,7 +413,7 @@ static int fuse_statfs(struct dentry *dentry, struct kstatfs *buf) return 0; } - req = fuse_get_req(fc); + req = fuse_get_req_nopages(fc); if (IS_ERR(req)) return PTR_ERR(req); -- cgit From 4d53dc99baf139e4fa0d395f7658032cc2eb3297 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:48:42 +0400 Subject: fuse: rework fuse_retrieve() The patch reworks fuse_retrieve() to allocate only so many page pointers as needed. The core part of the patch is the following calculation: num_pages = (num + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; (thanks Miklos for formula). All other changes are mostly shuffling lines. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index ff5e8bed5d88..28d9792de4a2 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -1577,13 +1577,24 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, unsigned int num; unsigned int offset; size_t total_len = 0; + int num_pages; - req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); + offset = outarg->offset & ~PAGE_CACHE_MASK; + file_size = i_size_read(inode); + + num = outarg->size; + if (outarg->offset > file_size) + num = 0; + else if (outarg->offset + num > file_size) + num = file_size - outarg->offset; + + num_pages = (num + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; + num_pages = min(num_pages, FUSE_MAX_PAGES_PER_REQ); + + req = fuse_get_req(fc, num_pages); if (IS_ERR(req)) return PTR_ERR(req); - offset = outarg->offset & ~PAGE_CACHE_MASK; - req->in.h.opcode = FUSE_NOTIFY_REPLY; req->in.h.nodeid = outarg->nodeid; req->in.numargs = 2; @@ -1592,14 +1603,8 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, req->end = fuse_retrieve_end; index = outarg->offset >> PAGE_CACHE_SHIFT; - file_size = i_size_read(inode); - num = outarg->size; - if (outarg->offset > file_size) - num = 0; - else if (outarg->offset + num > file_size) - num = file_size - outarg->offset; - while (num && req->num_pages < FUSE_MAX_PAGES_PER_REQ) { + while (num && req->num_pages < num_pages) { struct page *page; unsigned int this_num; -- cgit From f8dbdf81821b5ab4c5e86e7b2bd7edb892c159c2 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:48:51 +0400 Subject: fuse: rework fuse_readpages() The patch uses 'nr_pages' argument of fuse_readpages() as heuristics for the number of page pointers to allocate. This can be improved further by taking in consideration fc->max_read and gaps between page indices, but it's not clear whether it's worthy or not. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 882877125c1d..5fd06bae1790 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -641,6 +641,7 @@ struct fuse_fill_data { struct fuse_req *req; struct file *file; struct inode *inode; + unsigned nr_pages; }; static int fuse_readpages_fill(void *_data, struct page *page) @@ -656,16 +657,25 @@ static int fuse_readpages_fill(void *_data, struct page *page) (req->num_pages == FUSE_MAX_PAGES_PER_REQ || (req->num_pages + 1) * PAGE_CACHE_SIZE > fc->max_read || req->pages[req->num_pages - 1]->index + 1 != page->index)) { + int nr_alloc = min_t(unsigned, data->nr_pages, + FUSE_MAX_PAGES_PER_REQ); fuse_send_readpages(req, data->file); - data->req = req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); + data->req = req = fuse_get_req(fc, nr_alloc); if (IS_ERR(req)) { unlock_page(page); return PTR_ERR(req); } } + + if (WARN_ON(req->num_pages >= req->max_pages)) { + fuse_put_request(fc, req); + return -EIO; + } + page_cache_get(page); req->pages[req->num_pages] = page; req->num_pages++; + data->nr_pages--; return 0; } @@ -676,6 +686,7 @@ static int fuse_readpages(struct file *file, struct address_space *mapping, struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_fill_data data; int err; + int nr_alloc = min_t(unsigned, nr_pages, FUSE_MAX_PAGES_PER_REQ); err = -EIO; if (is_bad_inode(inode)) @@ -683,7 +694,8 @@ static int fuse_readpages(struct file *file, struct address_space *mapping, data.file = file; data.inode = inode; - data.req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); + data.req = fuse_get_req(fc, nr_alloc); + data.nr_pages = nr_pages; err = PTR_ERR(data.req); if (IS_ERR(data.req)) goto out; -- cgit From d07f09f509fb21482096e1975f160b694c0edf84 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:49:00 +0400 Subject: fuse: rework fuse_perform_write() The patch allocates as many page pointers in fuse_req as needed to cover interval [pos .. pos+len-1]. Inline helper fuse_wr_pages() is introduced to hide this cumbersome arithmetic. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 5fd06bae1790..b9972502f43a 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -881,11 +881,19 @@ static ssize_t fuse_fill_write_pages(struct fuse_req *req, if (!fc->big_writes) break; } while (iov_iter_count(ii) && count < fc->max_write && - req->num_pages < FUSE_MAX_PAGES_PER_REQ && offset == 0); + req->num_pages < req->max_pages && offset == 0); return count > 0 ? count : err; } +static inline unsigned fuse_wr_pages(loff_t pos, size_t len) +{ + return min_t(unsigned, + ((pos + len - 1) >> PAGE_CACHE_SHIFT) - + (pos >> PAGE_CACHE_SHIFT) + 1, + FUSE_MAX_PAGES_PER_REQ); +} + static ssize_t fuse_perform_write(struct file *file, struct address_space *mapping, struct iov_iter *ii, loff_t pos) @@ -901,8 +909,9 @@ static ssize_t fuse_perform_write(struct file *file, do { struct fuse_req *req; ssize_t count; + unsigned nr_pages = fuse_wr_pages(pos, iov_iter_count(ii)); - req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); + req = fuse_get_req(fc, nr_pages); if (IS_ERR(req)) { err = PTR_ERR(req); break; -- cgit From 54b966702dafe396b6f4e609f222b8e0fdb4d7a4 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:49:13 +0400 Subject: fuse: rework fuse_do_ioctl() fuse_do_ioctl() already calculates the number of pages it's going to use. It is stored in 'num_pages' variable. So the patch simply uses it for allocating fuse_req. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index b9972502f43a..fdb5b33198aa 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1894,7 +1894,7 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, num_pages++; } - req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); + req = fuse_get_req(fc, num_pages); if (IS_ERR(req)) { err = PTR_ERR(req); req = NULL; -- cgit From b2430d7567a376b3685627ca7e9d712f6f27d49b Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:49:24 +0400 Subject: fuse: add per-page descriptor to fuse_req The ability to save page pointers along with lengths and offsets in fuse_req will be useful to cover several iovec-s with a single fuse_req. Per-request page_offset is removed because anybody who need it can use req->page_descs[0].offset instead. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 28 ++++++++++++++++++++-------- fs/fuse/file.c | 10 +++++----- fs/fuse/fuse_i.h | 15 ++++++++++++--- 3 files changed, 37 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 28d9792de4a2..db4af8f3886a 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -35,15 +35,18 @@ static struct fuse_conn *fuse_get_conn(struct file *file) } static void fuse_request_init(struct fuse_req *req, struct page **pages, + struct fuse_page_desc *page_descs, unsigned npages) { memset(req, 0, sizeof(*req)); memset(pages, 0, sizeof(*pages) * npages); + memset(page_descs, 0, sizeof(*page_descs) * npages); INIT_LIST_HEAD(&req->list); INIT_LIST_HEAD(&req->intr_entry); init_waitqueue_head(&req->waitq); atomic_set(&req->count, 1); req->pages = pages; + req->page_descs = page_descs; req->max_pages = npages; } @@ -52,18 +55,25 @@ static struct fuse_req *__fuse_request_alloc(unsigned npages, gfp_t flags) struct fuse_req *req = kmem_cache_alloc(fuse_req_cachep, flags); if (req) { struct page **pages; + struct fuse_page_desc *page_descs; - if (npages <= FUSE_REQ_INLINE_PAGES) + if (npages <= FUSE_REQ_INLINE_PAGES) { pages = req->inline_pages; - else + page_descs = req->inline_page_descs; + } else { pages = kmalloc(sizeof(struct page *) * npages, flags); + page_descs = kmalloc(sizeof(struct fuse_page_desc) * + npages, flags); + } - if (!pages) { + if (!pages || !page_descs) { + kfree(pages); + kfree(page_descs); kmem_cache_free(fuse_req_cachep, req); return NULL; } - fuse_request_init(req, pages, npages); + fuse_request_init(req, pages, page_descs, npages); } return req; } @@ -81,8 +91,10 @@ struct fuse_req *fuse_request_alloc_nofs(unsigned npages) void fuse_request_free(struct fuse_req *req) { - if (req->pages != req->inline_pages) + if (req->pages != req->inline_pages) { kfree(req->pages); + kfree(req->page_descs); + } kmem_cache_free(fuse_req_cachep, req); } @@ -186,7 +198,7 @@ static void put_reserved_req(struct fuse_conn *fc, struct fuse_req *req) struct fuse_file *ff = file->private_data; spin_lock(&fc->lock); - fuse_request_init(req, req->pages, req->max_pages); + fuse_request_init(req, req->pages, req->page_descs, req->max_pages); BUG_ON(ff->reserved_req); ff->reserved_req = req; wake_up_all(&fc->reserved_req_waitq); @@ -891,7 +903,7 @@ static int fuse_copy_pages(struct fuse_copy_state *cs, unsigned nbytes, { unsigned i; struct fuse_req *req = cs->req; - unsigned offset = req->page_offset; + unsigned offset = req->page_descs[0].offset; unsigned count = min(nbytes, (unsigned) PAGE_SIZE - offset); for (i = 0; i < req->num_pages && (nbytes || zeroing); i++) { @@ -1599,7 +1611,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, req->in.h.nodeid = outarg->nodeid; req->in.numargs = 2; req->in.argpages = 1; - req->page_offset = offset; + req->page_descs[0].offset = offset; req->end = fuse_retrieve_end; index = outarg->offset >> PAGE_CACHE_SHIFT; diff --git a/fs/fuse/file.c b/fs/fuse/file.c index fdb5b33198aa..2b6f08ac62c3 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -798,7 +798,7 @@ static size_t fuse_send_write_pages(struct fuse_req *req, struct file *file, res = fuse_send_write(req, file, pos, count, NULL); - offset = req->page_offset; + offset = req->page_descs[0].offset; count = res; for (i = 0; i < req->num_pages; i++) { struct page *page = req->pages[i]; @@ -829,7 +829,7 @@ static ssize_t fuse_fill_write_pages(struct fuse_req *req, int err; req->in.argpages = 1; - req->page_offset = offset; + req->page_descs[0].offset = offset; do { size_t tmp; @@ -1070,14 +1070,14 @@ static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, return npages; req->num_pages = npages; - req->page_offset = offset; + req->page_descs[0].offset = offset; if (write) req->in.argpages = 1; else req->out.argpages = 1; - nbytes = (req->num_pages << PAGE_SHIFT) - req->page_offset; + nbytes = (req->num_pages << PAGE_SHIFT) - req->page_descs[0].offset; *nbytesp = min(*nbytesp, nbytes); return 0; @@ -1314,7 +1314,7 @@ static int fuse_writepage_locked(struct page *page) req->in.argpages = 1; req->num_pages = 1; req->pages[0] = tmp_page; - req->page_offset = 0; + req->page_descs[0].offset = 0; req->end = fuse_writepage_end; req->inode = inode; diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 5b21e6ab9e75..70cef60afe0e 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -203,6 +203,12 @@ struct fuse_out { struct fuse_arg args[3]; }; +/** FUSE page descriptor */ +struct fuse_page_desc { + unsigned int length; + unsigned int offset; +}; + /** The request state */ enum fuse_req_state { FUSE_REQ_INIT = 0, @@ -296,18 +302,21 @@ struct fuse_req { /** page vector */ struct page **pages; + /** page-descriptor vector */ + struct fuse_page_desc *page_descs; + /** size of the 'pages' array */ unsigned max_pages; /** inline page vector */ struct page *inline_pages[FUSE_REQ_INLINE_PAGES]; + /** inline page-descriptor vector */ + struct fuse_page_desc inline_page_descs[FUSE_REQ_INLINE_PAGES]; + /** number of pages in vector */ unsigned num_pages; - /** offset of data on first page */ - unsigned page_offset; - /** File used in the request (or NULL) */ struct fuse_file *ff; -- cgit From 85f40aec887110ae6bbefa87988def4606a3d583 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:49:33 +0400 Subject: fuse: use req->page_descs[] for argpages cases Previously, anyone who set flag 'argpages' only filled req->pages[] and set per-request page_offset. This patch re-works all cases where argpages=1 to fill req->page_descs[] properly. Having req->page_descs[] filled properly allows to re-work fuse_copy_pages() to copy page fragments described by req->page_descs[]. This will be useful for next patches optimizing direct_IO. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/cuse.c | 1 + fs/fuse/dev.c | 7 +++---- fs/fuse/dir.c | 1 + fs/fuse/file.c | 20 ++++++++++++++++++++ 4 files changed, 25 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c index 5cc838f320d4..fb9dcfdf040b 100644 --- a/fs/fuse/cuse.c +++ b/fs/fuse/cuse.c @@ -449,6 +449,7 @@ static int cuse_send_init(struct cuse_conn *cc) req->out.argvar = 1; req->out.argpages = 1; req->pages[0] = page; + req->page_descs[0].length = req->out.args[1].size; req->num_pages = 1; req->end = cuse_process_init_reply; fuse_request_send_background(fc, req); diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index db4af8f3886a..cbae09e5a491 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -903,11 +903,11 @@ static int fuse_copy_pages(struct fuse_copy_state *cs, unsigned nbytes, { unsigned i; struct fuse_req *req = cs->req; - unsigned offset = req->page_descs[0].offset; - unsigned count = min(nbytes, (unsigned) PAGE_SIZE - offset); for (i = 0; i < req->num_pages && (nbytes || zeroing); i++) { int err; + unsigned offset = req->page_descs[i].offset; + unsigned count = min(nbytes, req->page_descs[i].length); err = fuse_copy_page(cs, &req->pages[i], offset, count, zeroing); @@ -915,8 +915,6 @@ static int fuse_copy_pages(struct fuse_copy_state *cs, unsigned nbytes, return err; nbytes -= count; - count = min(nbytes, (unsigned) PAGE_SIZE); - offset = 0; } return 0; } @@ -1626,6 +1624,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, this_num = min_t(unsigned, num, PAGE_CACHE_SIZE - offset); req->pages[req->num_pages] = page; + req->page_descs[req->num_pages].length = this_num; req->num_pages++; offset = 0; diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index d04bcc5cccd0..ed8f8c55ce10 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1317,6 +1317,7 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) req->out.argpages = 1; req->num_pages = 1; req->pages[0] = page; + req->page_descs[0].length = PAGE_SIZE; if (fc->do_readdirplus) { attr_version = fuse_get_attr_version(fc); fuse_read_fill(req, file, file->f_pos, PAGE_SIZE, diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 2b6f08ac62c3..3384a200b3ee 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -555,6 +555,7 @@ static int fuse_readpage(struct file *file, struct page *page) req->out.argpages = 1; req->num_pages = 1; req->pages[0] = page; + req->page_descs[0].length = count; num_read = fuse_send_read(req, file, pos, count, NULL); err = req->out.h.error; fuse_put_request(fc, req); @@ -674,6 +675,7 @@ static int fuse_readpages_fill(void *_data, struct page *page) page_cache_get(page); req->pages[req->num_pages] = page; + req->page_descs[req->num_pages].length = PAGE_SIZE; req->num_pages++; data->nr_pages--; return 0; @@ -869,6 +871,7 @@ static ssize_t fuse_fill_write_pages(struct fuse_req *req, err = 0; req->pages[req->num_pages] = page; + req->page_descs[req->num_pages].length = tmp; req->num_pages++; iov_iter_advance(ii, tmp); @@ -1044,6 +1047,15 @@ static void fuse_release_user_pages(struct fuse_req *req, int write) } } +static inline void fuse_page_descs_length_init(struct fuse_req *req) +{ + int i; + + for (i = 0; i < req->num_pages; i++) + req->page_descs[i].length = PAGE_SIZE - + req->page_descs[i].offset; +} + static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, size_t *nbytesp, int write) { @@ -1071,6 +1083,7 @@ static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, req->num_pages = npages; req->page_descs[0].offset = offset; + fuse_page_descs_length_init(req); if (write) req->in.argpages = 1; @@ -1078,6 +1091,11 @@ static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, req->out.argpages = 1; nbytes = (req->num_pages << PAGE_SHIFT) - req->page_descs[0].offset; + + if (*nbytesp < nbytes) + req->page_descs[req->num_pages - 1].length -= + nbytes - *nbytesp; + *nbytesp = min(*nbytesp, nbytes); return 0; @@ -1315,6 +1333,7 @@ static int fuse_writepage_locked(struct page *page) req->num_pages = 1; req->pages[0] = tmp_page; req->page_descs[0].offset = 0; + req->page_descs[0].length = PAGE_SIZE; req->end = fuse_writepage_end; req->inode = inode; @@ -1902,6 +1921,7 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, } memcpy(req->pages, pages, sizeof(req->pages[0]) * num_pages); req->num_pages = num_pages; + fuse_page_descs_length_init(req); /* okay, let's send it to the client */ req->in.h.opcode = FUSE_IOCTL; -- cgit From b98d023a24496bf7d538c549e5426b1173c6f55d Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:50:15 +0400 Subject: fuse: pass iov[] to fuse_get_user_pages() The patch makes preliminary work for the next patch optimizing scatter-gather direct IO. The idea is to allow fuse_get_user_pages() to pack as many iov-s to each fuse request as possible. So, here we only rework all related call-paths to carry iov[] from fuse_direct_IO() to fuse_get_user_pages(). Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 103 +++++++++++++++++++++++++++------------------------------ 1 file changed, 49 insertions(+), 54 deletions(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 3384a200b3ee..542ad97b103a 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1056,14 +1056,18 @@ static inline void fuse_page_descs_length_init(struct fuse_req *req) req->page_descs[i].offset; } -static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, +static int fuse_get_user_pages(struct fuse_req *req, struct iov_iter *ii, size_t *nbytesp, int write) { size_t nbytes = *nbytesp; - unsigned long user_addr = (unsigned long) buf; - unsigned offset = user_addr & ~PAGE_MASK; + size_t frag_size = min(iov_iter_single_seg_count(ii), nbytes); + unsigned long user_addr; + unsigned offset; int npages; + user_addr = (unsigned long)ii->iov->iov_base + ii->iov_offset; + offset = user_addr & ~PAGE_MASK; + /* Special case for kernel I/O: can copy directly into the buffer */ if (segment_eq(get_fs(), KERNEL_DS)) { if (write) @@ -1071,10 +1075,12 @@ static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, else req->out.args[0].value = (void *) user_addr; + iov_iter_advance(ii, frag_size); + *nbytesp = frag_size; return 0; } - nbytes = min_t(size_t, nbytes, FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT); + nbytes = min_t(size_t, frag_size, FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT); npages = (nbytes + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; npages = clamp(npages, 1, FUSE_MAX_PAGES_PER_REQ); npages = get_user_pages_fast(user_addr, npages, !write, req->pages); @@ -1092,17 +1098,19 @@ static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, nbytes = (req->num_pages << PAGE_SHIFT) - req->page_descs[0].offset; - if (*nbytesp < nbytes) + if (frag_size < nbytes) req->page_descs[req->num_pages - 1].length -= - nbytes - *nbytesp; + nbytes - frag_size; - *nbytesp = min(*nbytesp, nbytes); + *nbytesp = min(frag_size, nbytes); + iov_iter_advance(ii, *nbytesp); return 0; } -ssize_t fuse_direct_io(struct file *file, const char __user *buf, - size_t count, loff_t *ppos, int write) +static ssize_t __fuse_direct_io(struct file *file, const struct iovec *iov, + unsigned long nr_segs, size_t count, + loff_t *ppos, int write) { struct fuse_file *ff = file->private_data; struct fuse_conn *fc = ff->fc; @@ -1110,6 +1118,9 @@ ssize_t fuse_direct_io(struct file *file, const char __user *buf, loff_t pos = *ppos; ssize_t res = 0; struct fuse_req *req; + struct iov_iter ii; + + iov_iter_init(&ii, iov, nr_segs, count, 0); req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); if (IS_ERR(req)) @@ -1119,7 +1130,7 @@ ssize_t fuse_direct_io(struct file *file, const char __user *buf, size_t nres; fl_owner_t owner = current->files; size_t nbytes = min(count, nmax); - int err = fuse_get_user_pages(req, buf, &nbytes, write); + int err = fuse_get_user_pages(req, &ii, &nbytes, write); if (err) { res = err; break; @@ -1142,7 +1153,6 @@ ssize_t fuse_direct_io(struct file *file, const char __user *buf, count -= nres; res += nres; pos += nres; - buf += nres; if (nres != nbytes) break; if (count) { @@ -1159,10 +1169,17 @@ ssize_t fuse_direct_io(struct file *file, const char __user *buf, return res; } + +ssize_t fuse_direct_io(struct file *file, const char __user *buf, + size_t count, loff_t *ppos, int write) +{ + struct iovec iov = { .iov_base = (void *)buf, .iov_len = count }; + return __fuse_direct_io(file, &iov, 1, count, ppos, write); +} EXPORT_SYMBOL_GPL(fuse_direct_io); -static ssize_t fuse_direct_read(struct file *file, char __user *buf, - size_t count, loff_t *ppos) +static ssize_t __fuse_direct_read(struct file *file, const struct iovec *iov, + unsigned long nr_segs, loff_t *ppos) { ssize_t res; struct inode *inode = file->f_path.dentry->d_inode; @@ -1170,22 +1187,31 @@ static ssize_t fuse_direct_read(struct file *file, char __user *buf, if (is_bad_inode(inode)) return -EIO; - res = fuse_direct_io(file, buf, count, ppos, 0); + res = __fuse_direct_io(file, iov, nr_segs, iov_length(iov, nr_segs), + ppos, 0); fuse_invalidate_attr(inode); return res; } -static ssize_t __fuse_direct_write(struct file *file, const char __user *buf, - size_t count, loff_t *ppos) +static ssize_t fuse_direct_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + struct iovec iov = { .iov_base = (void *)buf, .iov_len = count }; + return __fuse_direct_read(file, &iov, 1, ppos); +} + +static ssize_t __fuse_direct_write(struct file *file, const struct iovec *iov, + unsigned long nr_segs, loff_t *ppos) { struct inode *inode = file->f_path.dentry->d_inode; + size_t count = iov_length(iov, nr_segs); ssize_t res; res = generic_write_checks(file, ppos, &count, 0); if (!res) { - res = fuse_direct_io(file, buf, count, ppos, 1); + res = __fuse_direct_io(file, iov, nr_segs, count, ppos, 1); if (res > 0) fuse_write_update_size(inode, *ppos); } @@ -1198,6 +1224,7 @@ static ssize_t __fuse_direct_write(struct file *file, const char __user *buf, static ssize_t fuse_direct_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { + struct iovec iov = { .iov_base = (void *)buf, .iov_len = count }; struct inode *inode = file->f_path.dentry->d_inode; ssize_t res; @@ -1206,7 +1233,7 @@ static ssize_t fuse_direct_write(struct file *file, const char __user *buf, /* Don't allow parallel writes to the same file */ mutex_lock(&inode->i_mutex); - res = __fuse_direct_write(file, buf, count, ppos); + res = __fuse_direct_write(file, &iov, 1, ppos); mutex_unlock(&inode->i_mutex); return res; @@ -2167,41 +2194,6 @@ int fuse_notify_poll_wakeup(struct fuse_conn *fc, return 0; } -static ssize_t fuse_loop_dio(struct file *filp, const struct iovec *iov, - unsigned long nr_segs, loff_t *ppos, int rw) -{ - const struct iovec *vector = iov; - ssize_t ret = 0; - - while (nr_segs > 0) { - void __user *base; - size_t len; - ssize_t nr; - - base = vector->iov_base; - len = vector->iov_len; - vector++; - nr_segs--; - - if (rw == WRITE) - nr = __fuse_direct_write(filp, base, len, ppos); - else - nr = fuse_direct_read(filp, base, len, ppos); - - if (nr < 0) { - if (!ret) - ret = nr; - break; - } - ret += nr; - if (nr != len) - break; - } - - return ret; -} - - static ssize_t fuse_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t offset, unsigned long nr_segs) @@ -2213,7 +2205,10 @@ fuse_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, file = iocb->ki_filp; pos = offset; - ret = fuse_loop_dio(file, iov, nr_segs, &pos, rw); + if (rw == WRITE) + ret = __fuse_direct_write(file, iov, nr_segs, &pos); + else + ret = __fuse_direct_read(file, iov, nr_segs, &pos); return ret; } -- cgit From 7c190c8b9c0dd373cdd4d96e63306ec6e1a7115d Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:50:29 +0400 Subject: fuse: optimize fuse_get_user_pages() Let fuse_get_user_pages() pack as many iov-s to a single fuse_req as possible. This is very beneficial in case of iov[] consisting of many iov-s of relatively small sizes (e.g. PAGE_SIZE). Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 79 +++++++++++++++++++++++++++++++++++++--------------------- 1 file changed, 51 insertions(+), 28 deletions(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 542ad97b103a..b2aa6c21e209 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1047,29 +1047,37 @@ static void fuse_release_user_pages(struct fuse_req *req, int write) } } -static inline void fuse_page_descs_length_init(struct fuse_req *req) +static inline void fuse_page_descs_length_init(struct fuse_req *req, + unsigned index, unsigned nr_pages) { int i; - for (i = 0; i < req->num_pages; i++) + for (i = index; i < index + nr_pages; i++) req->page_descs[i].length = PAGE_SIZE - req->page_descs[i].offset; } +static inline unsigned long fuse_get_user_addr(const struct iov_iter *ii) +{ + return (unsigned long)ii->iov->iov_base + ii->iov_offset; +} + +static inline size_t fuse_get_frag_size(const struct iov_iter *ii, + size_t max_size) +{ + return min(iov_iter_single_seg_count(ii), max_size); +} + static int fuse_get_user_pages(struct fuse_req *req, struct iov_iter *ii, size_t *nbytesp, int write) { - size_t nbytes = *nbytesp; - size_t frag_size = min(iov_iter_single_seg_count(ii), nbytes); - unsigned long user_addr; - unsigned offset; - int npages; - - user_addr = (unsigned long)ii->iov->iov_base + ii->iov_offset; - offset = user_addr & ~PAGE_MASK; + size_t nbytes = 0; /* # bytes already packed in req */ /* Special case for kernel I/O: can copy directly into the buffer */ if (segment_eq(get_fs(), KERNEL_DS)) { + unsigned long user_addr = fuse_get_user_addr(ii); + size_t frag_size = fuse_get_frag_size(ii, *nbytesp); + if (write) req->in.args[1].value = (void *) user_addr; else @@ -1080,30 +1088,45 @@ static int fuse_get_user_pages(struct fuse_req *req, struct iov_iter *ii, return 0; } - nbytes = min_t(size_t, frag_size, FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT); - npages = (nbytes + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; - npages = clamp(npages, 1, FUSE_MAX_PAGES_PER_REQ); - npages = get_user_pages_fast(user_addr, npages, !write, req->pages); - if (npages < 0) - return npages; + while (nbytes < *nbytesp && req->num_pages < FUSE_MAX_PAGES_PER_REQ) { + unsigned npages; + unsigned long user_addr = fuse_get_user_addr(ii); + unsigned offset = user_addr & ~PAGE_MASK; + size_t frag_size = fuse_get_frag_size(ii, *nbytesp - nbytes); + int ret; - req->num_pages = npages; - req->page_descs[0].offset = offset; - fuse_page_descs_length_init(req); + unsigned n = FUSE_MAX_PAGES_PER_REQ - req->num_pages; + frag_size = min_t(size_t, frag_size, n << PAGE_SHIFT); + + npages = (frag_size + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; + npages = clamp(npages, 1U, n); + + ret = get_user_pages_fast(user_addr, npages, !write, + &req->pages[req->num_pages]); + if (ret < 0) + return ret; + + npages = ret; + frag_size = min_t(size_t, frag_size, + (npages << PAGE_SHIFT) - offset); + iov_iter_advance(ii, frag_size); + + req->page_descs[req->num_pages].offset = offset; + fuse_page_descs_length_init(req, req->num_pages, npages); + + req->num_pages += npages; + req->page_descs[req->num_pages - 1].length -= + (npages << PAGE_SHIFT) - offset - frag_size; + + nbytes += frag_size; + } if (write) req->in.argpages = 1; else req->out.argpages = 1; - nbytes = (req->num_pages << PAGE_SHIFT) - req->page_descs[0].offset; - - if (frag_size < nbytes) - req->page_descs[req->num_pages - 1].length -= - nbytes - frag_size; - - *nbytesp = min(frag_size, nbytes); - iov_iter_advance(ii, *nbytesp); + *nbytesp = nbytes; return 0; } @@ -1948,7 +1971,7 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, } memcpy(req->pages, pages, sizeof(req->pages[0]) * num_pages); req->num_pages = num_pages; - fuse_page_descs_length_init(req); + fuse_page_descs_length_init(req, 0, req->num_pages); /* okay, let's send it to the client */ req->in.h.opcode = FUSE_IOCTL; -- cgit From 5565a9d884327ac45d49041f1b846dac273e110c Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Fri, 26 Oct 2012 19:50:36 +0400 Subject: fuse: optimize __fuse_direct_io() __fuse_direct_io() allocates fuse-requests by calling fuse_get_req(fc, n). The patch calculates 'n' based on iov[] array. This is useful because allocating FUSE_MAX_PAGES_PER_REQ page pointers and descriptors for each fuse request would be waste of memory in case of iov-s of smaller size. Signed-off-by: Maxim Patlasov Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index b2aa6c21e209..68e10d43bd3f 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1088,14 +1088,14 @@ static int fuse_get_user_pages(struct fuse_req *req, struct iov_iter *ii, return 0; } - while (nbytes < *nbytesp && req->num_pages < FUSE_MAX_PAGES_PER_REQ) { + while (nbytes < *nbytesp && req->num_pages < req->max_pages) { unsigned npages; unsigned long user_addr = fuse_get_user_addr(ii); unsigned offset = user_addr & ~PAGE_MASK; size_t frag_size = fuse_get_frag_size(ii, *nbytesp - nbytes); int ret; - unsigned n = FUSE_MAX_PAGES_PER_REQ - req->num_pages; + unsigned n = req->max_pages - req->num_pages; frag_size = min_t(size_t, frag_size, n << PAGE_SHIFT); npages = (frag_size + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; @@ -1131,6 +1131,23 @@ static int fuse_get_user_pages(struct fuse_req *req, struct iov_iter *ii, return 0; } +static inline int fuse_iter_npages(const struct iov_iter *ii_p) +{ + struct iov_iter ii = *ii_p; + int npages = 0; + + while (iov_iter_count(&ii) && npages < FUSE_MAX_PAGES_PER_REQ) { + unsigned long user_addr = fuse_get_user_addr(&ii); + unsigned offset = user_addr & ~PAGE_MASK; + size_t frag_size = iov_iter_single_seg_count(&ii); + + npages += (frag_size + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; + iov_iter_advance(&ii, frag_size); + } + + return min(npages, FUSE_MAX_PAGES_PER_REQ); +} + static ssize_t __fuse_direct_io(struct file *file, const struct iovec *iov, unsigned long nr_segs, size_t count, loff_t *ppos, int write) @@ -1145,7 +1162,7 @@ static ssize_t __fuse_direct_io(struct file *file, const struct iovec *iov, iov_iter_init(&ii, iov, nr_segs, count, 0); - req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); + req = fuse_get_req(fc, fuse_iter_npages(&ii)); if (IS_ERR(req)) return PTR_ERR(req); @@ -1180,7 +1197,7 @@ static ssize_t __fuse_direct_io(struct file *file, const struct iovec *iov, break; if (count) { fuse_put_request(fc, req); - req = fuse_get_req(fc, FUSE_MAX_PAGES_PER_REQ); + req = fuse_get_req(fc, fuse_iter_npages(&ii)); if (IS_ERR(req)) break; } -- cgit From fb05f41f5f96f7423c53da4d87913fb44fd0565d Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Sat, 10 Nov 2012 16:55:56 +0100 Subject: fuse: cleanup fuse_direct_io() Fix the following sparse warnings: fs/fuse/file.c:1216:43: warning: cast removes address space of expression fs/fuse/file.c:1216:43: warning: incorrect type in initializer (different address spaces) fs/fuse/file.c:1216:43: expected void [noderef] *iov_base fs/fuse/file.c:1216:43: got void * fs/fuse/file.c:1241:43: warning: cast removes address space of expression fs/fuse/file.c:1241:43: warning: incorrect type in initializer (different address spaces) fs/fuse/file.c:1241:43: expected void [noderef] *iov_base fs/fuse/file.c:1241:43: got void * fs/fuse/file.c:1267:43: warning: cast removes address space of expression fs/fuse/file.c:1267:43: warning: incorrect type in initializer (different address spaces) fs/fuse/file.c:1267:43: expected void [noderef] *iov_base fs/fuse/file.c:1267:43: got void * Signed-off-by: Miklos Szeredi --- fs/fuse/cuse.c | 7 +++++-- fs/fuse/file.c | 23 ++++++++--------------- fs/fuse/fuse_i.h | 5 +++-- 3 files changed, 16 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c index fb9dcfdf040b..6f96a8def147 100644 --- a/fs/fuse/cuse.c +++ b/fs/fuse/cuse.c @@ -91,19 +91,22 @@ static ssize_t cuse_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { loff_t pos = 0; + struct iovec iov = { .iov_base = buf, .iov_len = count }; - return fuse_direct_io(file, buf, count, &pos, 0); + return fuse_direct_io(file, &iov, 1, count, &pos, 0); } static ssize_t cuse_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { loff_t pos = 0; + struct iovec iov = { .iov_base = (void __user *)buf, .iov_len = count }; + /* * No locking or generic_write_checks(), the server is * responsible for locking and sanity checks. */ - return fuse_direct_io(file, buf, count, &pos, 1); + return fuse_direct_io(file, &iov, 1, count, &pos, 1); } static int cuse_open(struct inode *inode, struct file *file) diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 68e10d43bd3f..28bc9c672196 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1148,9 +1148,9 @@ static inline int fuse_iter_npages(const struct iov_iter *ii_p) return min(npages, FUSE_MAX_PAGES_PER_REQ); } -static ssize_t __fuse_direct_io(struct file *file, const struct iovec *iov, - unsigned long nr_segs, size_t count, - loff_t *ppos, int write) +ssize_t fuse_direct_io(struct file *file, const struct iovec *iov, + unsigned long nr_segs, size_t count, loff_t *ppos, + int write) { struct fuse_file *ff = file->private_data; struct fuse_conn *fc = ff->fc; @@ -1209,13 +1209,6 @@ static ssize_t __fuse_direct_io(struct file *file, const struct iovec *iov, return res; } - -ssize_t fuse_direct_io(struct file *file, const char __user *buf, - size_t count, loff_t *ppos, int write) -{ - struct iovec iov = { .iov_base = (void *)buf, .iov_len = count }; - return __fuse_direct_io(file, &iov, 1, count, ppos, write); -} EXPORT_SYMBOL_GPL(fuse_direct_io); static ssize_t __fuse_direct_read(struct file *file, const struct iovec *iov, @@ -1227,8 +1220,8 @@ static ssize_t __fuse_direct_read(struct file *file, const struct iovec *iov, if (is_bad_inode(inode)) return -EIO; - res = __fuse_direct_io(file, iov, nr_segs, iov_length(iov, nr_segs), - ppos, 0); + res = fuse_direct_io(file, iov, nr_segs, iov_length(iov, nr_segs), + ppos, 0); fuse_invalidate_attr(inode); @@ -1238,7 +1231,7 @@ static ssize_t __fuse_direct_read(struct file *file, const struct iovec *iov, static ssize_t fuse_direct_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct iovec iov = { .iov_base = (void *)buf, .iov_len = count }; + struct iovec iov = { .iov_base = buf, .iov_len = count }; return __fuse_direct_read(file, &iov, 1, ppos); } @@ -1251,7 +1244,7 @@ static ssize_t __fuse_direct_write(struct file *file, const struct iovec *iov, res = generic_write_checks(file, ppos, &count, 0); if (!res) { - res = __fuse_direct_io(file, iov, nr_segs, count, ppos, 1); + res = fuse_direct_io(file, iov, nr_segs, count, ppos, 1); if (res > 0) fuse_write_update_size(inode, *ppos); } @@ -1264,7 +1257,7 @@ static ssize_t __fuse_direct_write(struct file *file, const struct iovec *iov, static ssize_t fuse_direct_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { - struct iovec iov = { .iov_base = (void *)buf, .iov_len = count }; + struct iovec iov = { .iov_base = (void __user *)buf, .iov_len = count }; struct inode *inode = file->f_path.dentry->d_inode; ssize_t res; diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 70cef60afe0e..13befcd29c5b 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -811,8 +811,9 @@ int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, int fuse_do_open(struct fuse_conn *fc, u64 nodeid, struct file *file, bool isdir); -ssize_t fuse_direct_io(struct file *file, const char __user *buf, - size_t count, loff_t *ppos, int write); +ssize_t fuse_direct_io(struct file *file, const struct iovec *iov, + unsigned long nr_segs, size_t count, loff_t *ppos, + int write); long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, unsigned int flags); long fuse_ioctl_common(struct file *file, unsigned int cmd, -- cgit From 10616b806d1d7835b1d23b8d75ef638f92cb98b6 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 21 Jan 2013 23:53:52 +1100 Subject: xfs: fix _xfs_buf_find oops on blocks beyond the filesystem end When _xfs_buf_find is passed an out of range address, it will fail to find a relevant struct xfs_perag and oops with a null dereference. This can happen when trying to walk a filesystem with a metadata inode that has a partially corrupted extent map (i.e. the block number returned is corrupt, but is otherwise intact) and we try to read from the corrupted block address. In this case, just fail the lookup. If it is readahead being issued, it will simply not be done, but if it is real read that fails we will get an error being reported. Ideally this case should result in an EFSCORRUPTED error being reported, but we cannot return an error through xfs_buf_read() or xfs_buf_get() so this lookup failure may result in ENOMEM or EIO errors being reported instead. Signed-off-by: Dave Chinner Reviewed-by: Brian Foster Reviewed-by: Ben Myers Signed-off-by: Ben Myers --- fs/xfs/xfs_buf.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'fs') diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 56d1614760cf..689d72655ea6 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -487,6 +487,7 @@ _xfs_buf_find( struct rb_node *parent; xfs_buf_t *bp; xfs_daddr_t blkno = map[0].bm_bn; + xfs_daddr_t eofs; int numblks = 0; int i; @@ -498,6 +499,23 @@ _xfs_buf_find( ASSERT(!(numbytes < (1 << btp->bt_sshift))); ASSERT(!(BBTOB(blkno) & (xfs_off_t)btp->bt_smask)); + /* + * Corrupted block numbers can get through to here, unfortunately, so we + * have to check that the buffer falls within the filesystem bounds. + */ + eofs = XFS_FSB_TO_BB(btp->bt_mount, btp->bt_mount->m_sb.sb_dblocks); + if (blkno >= eofs) { + /* + * XXX (dgc): we should really be returning EFSCORRUPTED here, + * but none of the higher level infrastructure supports + * returning a specific error on buffer lookup failures. + */ + xfs_alert(btp->bt_mount, + "%s: Block out of range: block 0x%llx, EOFS 0x%llx ", + __func__, blkno, eofs); + return NULL; + } + /* get tree root */ pag = xfs_perag_get(btp->bt_mount, xfs_daddr_to_agno(btp->bt_mount, blkno)); -- cgit From 4d559a3bcb7383f34334092af07e68fb60910684 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 21 Jan 2013 23:53:54 +1100 Subject: xfs: limit speculative prealloc near ENOSPC thresholds There is a window on small filesytsems where specualtive preallocation can be larger than that ENOSPC throttling thresholds, resulting in specualtive preallocation trying to reserve more space than there is space available. This causes immediate ENOSPC to be triggered, prealloc to be turned off and flushing to occur. One the next write (i.e. next 4k page), we do exactly the same thing, and so effective drive into synchronous 4k writes by triggering ENOSPC flushing on every page while in the window between the prealloc size and the ENOSPC prealloc throttle threshold. Fix this by checking to see if the prealloc size would consume all free space, and throttle it appropriately to avoid premature ENOSPC... Signed-off-by: Dave Chinner Reviewed-by: Brian Foster Signed-off-by: Ben Myers --- fs/xfs/xfs_iomap.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index add06b4e9a63..364818eef40e 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -351,6 +351,15 @@ xfs_iomap_prealloc_size( } if (shift) alloc_blocks >>= shift; + + /* + * If we are still trying to allocate more space than is + * available, squash the prealloc hard. This can happen if we + * have a large file on a small filesystem and the above + * lowspace thresholds are smaller than MAXEXTLEN. + */ + while (alloc_blocks >= freesp) + alloc_blocks >>= 4; } if (alloc_blocks < mp->m_writeio_blocks) -- cgit From 72ba74508b2857e71d65fc93f0d6b684492fc740 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Thu, 24 Jan 2013 23:24:54 -0500 Subject: ext4: release sysfs kobject when failing to enable quotas on mount In addition, print the error returned from ext4_enable_quotas() Signed-off-by: "Theodore Ts'o" Reviewed-by: Carlos Maiolino Cc: stable@vger.kernel.org --- fs/ext4/super.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index c014edd12648..3ac306064b28 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4009,7 +4009,7 @@ no_journal: !(sb->s_flags & MS_RDONLY)) { err = ext4_enable_quotas(sb); if (err) - goto failed_mount7; + goto failed_mount8; } #endif /* CONFIG_QUOTA */ @@ -4036,6 +4036,10 @@ cantfind_ext4: ext4_msg(sb, KERN_ERR, "VFS: Can't find ext4 filesystem"); goto failed_mount; +#ifdef CONFIG_QUOTA +failed_mount8: + kobject_del(&sbi->s_kobj); +#endif failed_mount7: ext4_unregister_li_request(sb); failed_mount6: @@ -5006,9 +5010,9 @@ static int ext4_enable_quotas(struct super_block *sb) DQUOT_USAGE_ENABLED); if (err) { ext4_warning(sb, - "Failed to enable quota (type=%d) " - "tracking. Please run e2fsck to fix.", - type); + "Failed to enable quota tracking " + "(type=%d, err=%d). Please run " + "e2fsck to fix.", type, err); return err; } } -- cgit From 03dafb5f59bd31b3f590329e95434203f0ca6661 Mon Sep 17 00:00:00 2001 From: Chen Gang Date: Thu, 24 Jan 2013 23:24:58 -0500 Subject: ext4: fix memory leak when quota options are specified multiple times When usrjquota or grpjquota mount options are specified several times, we leak memory storing the names. Free the memory correctly. Signed-off-by: Chen Gang Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/super.c | 52 +++++++++++++++++++++++++++++----------------------- 1 file changed, 29 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 3ac306064b28..3b04d15f303d 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1337,6 +1337,7 @@ static int set_qf_name(struct super_block *sb, int qtype, substring_t *args) { struct ext4_sb_info *sbi = EXT4_SB(sb); char *qname; + int ret = -1; if (sb_any_quota_loaded(sb) && !sbi->s_qf_names[qtype]) { @@ -1351,23 +1352,26 @@ static int set_qf_name(struct super_block *sb, int qtype, substring_t *args) "Not enough memory for storing quotafile name"); return -1; } - if (sbi->s_qf_names[qtype] && - strcmp(sbi->s_qf_names[qtype], qname)) { - ext4_msg(sb, KERN_ERR, - "%s quota file already specified", QTYPE2NAME(qtype)); - kfree(qname); - return -1; + if (sbi->s_qf_names[qtype]) { + if (strcmp(sbi->s_qf_names[qtype], qname) == 0) + ret = 1; + else + ext4_msg(sb, KERN_ERR, + "%s quota file already specified", + QTYPE2NAME(qtype)); + goto errout; } - sbi->s_qf_names[qtype] = qname; - if (strchr(sbi->s_qf_names[qtype], '/')) { + if (strchr(qname, '/')) { ext4_msg(sb, KERN_ERR, "quotafile must be on filesystem root"); - kfree(sbi->s_qf_names[qtype]); - sbi->s_qf_names[qtype] = NULL; - return -1; + goto errout; } + sbi->s_qf_names[qtype] = qname; set_opt(sb, QUOTA); return 1; +errout: + kfree(qname); + return ret; } static int clear_qf_name(struct super_block *sb, int qtype) @@ -1381,10 +1385,7 @@ static int clear_qf_name(struct super_block *sb, int qtype) " when quota turned on"); return -1; } - /* - * The space will be released later when all options are confirmed - * to be correct - */ + kfree(sbi->s_qf_names[qtype]); sbi->s_qf_names[qtype] = NULL; return 1; } @@ -4593,7 +4594,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO; int err = 0; #ifdef CONFIG_QUOTA - int i; + int i, j; #endif char *orig_data = kstrdup(data, GFP_KERNEL); @@ -4609,7 +4610,16 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) #ifdef CONFIG_QUOTA old_opts.s_jquota_fmt = sbi->s_jquota_fmt; for (i = 0; i < MAXQUOTAS; i++) - old_opts.s_qf_names[i] = sbi->s_qf_names[i]; + if (sbi->s_qf_names[i]) { + old_opts.s_qf_names[i] = kstrdup(sbi->s_qf_names[i], + GFP_KERNEL); + if (!old_opts.s_qf_names[i]) { + for (j = 0; j < i; j++) + kfree(old_opts.s_qf_names[j]); + return -ENOMEM; + } + } else + old_opts.s_qf_names[i] = NULL; #endif if (sbi->s_journal && sbi->s_journal->j_task->io_context) journal_ioprio = sbi->s_journal->j_task->io_context->ioprio; @@ -4742,9 +4752,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) #ifdef CONFIG_QUOTA /* Release old quota file names */ for (i = 0; i < MAXQUOTAS; i++) - if (old_opts.s_qf_names[i] && - old_opts.s_qf_names[i] != sbi->s_qf_names[i]) - kfree(old_opts.s_qf_names[i]); + kfree(old_opts.s_qf_names[i]); if (enable_quota) { if (sb_any_quota_suspended(sb)) dquot_resume(sb, -1); @@ -4773,9 +4781,7 @@ restore_opts: #ifdef CONFIG_QUOTA sbi->s_jquota_fmt = old_opts.s_jquota_fmt; for (i = 0; i < MAXQUOTAS; i++) { - if (sbi->s_qf_names[i] && - old_opts.s_qf_names[i] != sbi->s_qf_names[i]) - kfree(sbi->s_qf_names[i]); + kfree(sbi->s_qf_names[i]); sbi->s_qf_names[i] = old_opts.s_qf_names[i]; } #endif -- cgit From 0bb8f3d6ae621945e6fa2102aa894f72b76a023e Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 25 Jan 2013 21:51:13 +0100 Subject: sysfs: Functions for adding/removing symlinks to/from attribute groups The most convenient way to expose ACPI power resources lists of a device is to put symbolic links to sysfs directories representing those resources into special attribute groups in the device's sysfs directory. For this purpose, it is necessary to be able to add symbolic links to attribute groups. For this reason, add sysfs helper functions for adding/removing symbolic links to/from attribute groups, sysfs_add_link_to_group() and sysfs_remove_link_from_group(), respectively. This change set includes a build fix from David Rientjes. Signed-off-by: Rafael J. Wysocki Acked-by: Greg Kroah-Hartman --- fs/sysfs/group.c | 42 ++++++++++++++++++++++++++++++++++++++++++ fs/sysfs/symlink.c | 45 ++++++++++++++++++++++++++++++++------------- fs/sysfs/sysfs.h | 2 ++ 3 files changed, 76 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/group.c b/fs/sysfs/group.c index 2df555c66d57..aec3d5c98c94 100644 --- a/fs/sysfs/group.c +++ b/fs/sysfs/group.c @@ -205,6 +205,48 @@ void sysfs_unmerge_group(struct kobject *kobj, } EXPORT_SYMBOL_GPL(sysfs_unmerge_group); +/** + * sysfs_add_link_to_group - add a symlink to an attribute group. + * @kobj: The kobject containing the group. + * @group_name: The name of the group. + * @target: The target kobject of the symlink to create. + * @link_name: The name of the symlink to create. + */ +int sysfs_add_link_to_group(struct kobject *kobj, const char *group_name, + struct kobject *target, const char *link_name) +{ + struct sysfs_dirent *dir_sd; + int error = 0; + + dir_sd = sysfs_get_dirent(kobj->sd, NULL, group_name); + if (!dir_sd) + return -ENOENT; + + error = sysfs_create_link_sd(dir_sd, target, link_name); + sysfs_put(dir_sd); + + return error; +} +EXPORT_SYMBOL_GPL(sysfs_add_link_to_group); + +/** + * sysfs_remove_link_from_group - remove a symlink from an attribute group. + * @kobj: The kobject containing the group. + * @group_name: The name of the group. + * @link_name: The name of the symlink to remove. + */ +void sysfs_remove_link_from_group(struct kobject *kobj, const char *group_name, + const char *link_name) +{ + struct sysfs_dirent *dir_sd; + + dir_sd = sysfs_get_dirent(kobj->sd, NULL, group_name); + if (dir_sd) { + sysfs_hash_and_remove(dir_sd, NULL, link_name); + sysfs_put(dir_sd); + } +} +EXPORT_SYMBOL_GPL(sysfs_remove_link_from_group); EXPORT_SYMBOL_GPL(sysfs_create_group); EXPORT_SYMBOL_GPL(sysfs_update_group); diff --git a/fs/sysfs/symlink.c b/fs/sysfs/symlink.c index 3c9eb5624f5e..8c940df97a52 100644 --- a/fs/sysfs/symlink.c +++ b/fs/sysfs/symlink.c @@ -21,26 +21,17 @@ #include "sysfs.h" -static int sysfs_do_create_link(struct kobject *kobj, struct kobject *target, - const char *name, int warn) +static int sysfs_do_create_link_sd(struct sysfs_dirent *parent_sd, + struct kobject *target, + const char *name, int warn) { - struct sysfs_dirent *parent_sd = NULL; struct sysfs_dirent *target_sd = NULL; struct sysfs_dirent *sd = NULL; struct sysfs_addrm_cxt acxt; enum kobj_ns_type ns_type; int error; - BUG_ON(!name); - - if (!kobj) - parent_sd = &sysfs_root; - else - parent_sd = kobj->sd; - - error = -EFAULT; - if (!parent_sd) - goto out_put; + BUG_ON(!name || !parent_sd); /* target->sd can go away beneath us but is protected with * sysfs_assoc_lock. Fetch target_sd from it. @@ -95,6 +86,34 @@ static int sysfs_do_create_link(struct kobject *kobj, struct kobject *target, return error; } +/** + * sysfs_create_link_sd - create symlink to a given object. + * @sd: directory we're creating the link in. + * @target: object we're pointing to. + * @name: name of the symlink. + */ +int sysfs_create_link_sd(struct sysfs_dirent *sd, struct kobject *target, + const char *name) +{ + return sysfs_do_create_link_sd(sd, target, name, 1); +} + +static int sysfs_do_create_link(struct kobject *kobj, struct kobject *target, + const char *name, int warn) +{ + struct sysfs_dirent *parent_sd = NULL; + + if (!kobj) + parent_sd = &sysfs_root; + else + parent_sd = kobj->sd; + + if (!parent_sd) + return -EFAULT; + + return sysfs_do_create_link_sd(parent_sd, target, name, warn); +} + /** * sysfs_create_link - create symlink between two objects. * @kobj: object whose directory we're creating the link in. diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index d73c0932bbd6..d1e4043eb0c3 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -240,3 +240,5 @@ void unmap_bin_file(struct sysfs_dirent *attr_sd); * symlink.c */ extern const struct inode_operations sysfs_symlink_inode_operations; +int sysfs_create_link_sd(struct sysfs_dirent *sd, struct kobject *target, + const char *name); -- cgit From 3b19034d4f4554e39ca244fb28962bbf2ccba046 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 21 Jan 2013 23:53:55 +1100 Subject: xfs: fix shutdown hang on invalid inode during create When the new inode verify in xfs_iread() fails, the create transaction is aborted and a shutdown occurs. The subsequent unmount then hangs in xfs_wait_buftarg() on a buffer that has an elevated hold count. Debug showed that it was an AGI buffer getting stuck: [ 22.576147] XFS (vdb): buffer 0x2/0x1, hold 0x2 stuck [ 22.976213] XFS (vdb): buffer 0x2/0x1, hold 0x2 stuck [ 23.376206] XFS (vdb): buffer 0x2/0x1, hold 0x2 stuck [ 23.776325] XFS (vdb): buffer 0x2/0x1, hold 0x2 stuck The trace of this buffer leading up to the shutdown (trimmed for brevity) looks like: xfs_buf_init: bno 0x2 nblks 0x1 hold 1 caller xfs_buf_get_map xfs_buf_get: bno 0x2 len 0x200 hold 1 caller xfs_buf_read_map xfs_buf_read: bno 0x2 len 0x200 hold 1 caller xfs_trans_read_buf_map xfs_buf_iorequest: bno 0x2 nblks 0x1 hold 1 caller _xfs_buf_read xfs_buf_hold: bno 0x2 nblks 0x1 hold 1 caller xfs_buf_iorequest xfs_buf_rele: bno 0x2 nblks 0x1 hold 2 caller xfs_buf_iorequest xfs_buf_iowait: bno 0x2 nblks 0x1 hold 1 caller _xfs_buf_read xfs_buf_ioerror: bno 0x2 len 0x200 hold 1 caller xfs_buf_bio_end_io xfs_buf_iodone: bno 0x2 nblks 0x1 hold 1 caller _xfs_buf_ioend xfs_buf_iowait_done: bno 0x2 nblks 0x1 hold 1 caller _xfs_buf_read xfs_buf_hold: bno 0x2 nblks 0x1 hold 1 caller xfs_buf_item_init xfs_trans_read_buf: bno 0x2 len 0x200 hold 2 recur 0 refcount 1 xfs_trans_brelse: bno 0x2 len 0x200 hold 2 recur 0 refcount 1 xfs_buf_item_relse: bno 0x2 nblks 0x1 hold 2 caller xfs_trans_brelse xfs_buf_rele: bno 0x2 nblks 0x1 hold 2 caller xfs_buf_item_relse xfs_buf_unlock: bno 0x2 nblks 0x1 hold 1 caller xfs_trans_brelse xfs_buf_rele: bno 0x2 nblks 0x1 hold 1 caller xfs_trans_brelse xfs_buf_trylock: bno 0x2 nblks 0x1 hold 2 caller _xfs_buf_find xfs_buf_find: bno 0x2 len 0x200 hold 2 caller xfs_buf_get_map xfs_buf_get: bno 0x2 len 0x200 hold 2 caller xfs_buf_read_map xfs_buf_read: bno 0x2 len 0x200 hold 2 caller xfs_trans_read_buf_map xfs_buf_hold: bno 0x2 nblks 0x1 hold 2 caller xfs_buf_item_init xfs_trans_read_buf: bno 0x2 len 0x200 hold 3 recur 0 refcount 1 xfs_trans_log_buf: bno 0x2 len 0x200 hold 3 recur 0 refcount 1 xfs_buf_item_unlock: bno 0x2 len 0x200 hold 3 flags DIRTY liflags ABORTED xfs_buf_unlock: bno 0x2 nblks 0x1 hold 3 caller xfs_buf_item_unlock xfs_buf_rele: bno 0x2 nblks 0x1 hold 3 caller xfs_buf_item_unlock And that is the AGI buffer from cold cache read into memory to transaction abort. You can see at transaction abort the bli is dirty and only has a single reference. The item is not pinned, and it's not in the AIL. Hence the only reference to it is this transaction. The problem is that the xfs_buf_item_unlock() call is dropping the last reference to the xfs_buf_log_item attached to the buffer (which holds a reference to the buffer), but it is not freeing the xfs_buf_log_item. Hence nothing will ever release the buffer, and the unmount hangs waiting for this reference to go away. The fix is simple - xfs_buf_item_unlock needs to detect the last reference going away in this case and free the xfs_buf_log_item to release the reference it holds on the buffer. Signed-off-by: Dave Chinner Reviewed-by: Ben Myers Signed-off-by: Ben Myers --- fs/xfs/xfs_buf.c | 2 ++ fs/xfs/xfs_buf_item.c | 12 ++++++++++-- fs/xfs/xfs_trace.h | 1 + 3 files changed, 13 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 689d72655ea6..fbbb9eb92e32 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -1505,6 +1505,8 @@ restart: while (!list_empty(&btp->bt_lru)) { bp = list_first_entry(&btp->bt_lru, struct xfs_buf, b_lru); if (atomic_read(&bp->b_hold) > 1) { + trace_xfs_buf_wait_buftarg(bp, _RET_IP_); + list_move_tail(&bp->b_lru, &btp->bt_lru); spin_unlock(&btp->bt_lru_lock); delay(100); goto restart; diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 63c86c48014b..9c4c050e4270 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -548,7 +548,10 @@ xfs_buf_item_unlock( /* * If the buf item isn't tracking any data, free it, otherwise drop the - * reference we hold to it. + * reference we hold to it. If we are aborting the transaction, this may + * be the only reference to the buf item, so we free it anyway + * regardless of whether it is dirty or not. A dirty abort implies a + * shutdown, anyway. */ clean = 1; for (i = 0; i < bip->bli_format_count; i++) { @@ -560,7 +563,12 @@ xfs_buf_item_unlock( } if (clean) xfs_buf_item_relse(bp); - else + else if (aborted) { + if (atomic_dec_and_test(&bip->bli_refcount)) { + ASSERT(XFS_FORCED_SHUTDOWN(lip->li_mountp)); + xfs_buf_item_relse(bp); + } + } else atomic_dec(&bip->bli_refcount); if (!hold) diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 2e137d4a85ae..16a812977eab 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -341,6 +341,7 @@ DEFINE_BUF_EVENT(xfs_buf_item_relse); DEFINE_BUF_EVENT(xfs_buf_item_iodone); DEFINE_BUF_EVENT(xfs_buf_item_iodone_async); DEFINE_BUF_EVENT(xfs_buf_error_relse); +DEFINE_BUF_EVENT(xfs_buf_wait_buftarg); DEFINE_BUF_EVENT(xfs_trans_read_buf_io); DEFINE_BUF_EVENT(xfs_trans_read_buf_shut); -- cgit From ced55f38d6bde7c10a14ea51c2edcd51a98575e3 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 23 Jan 2013 13:56:18 +0100 Subject: xfs: Fix possible use-after-free with AIO Running AIO is pinning inode in memory using file reference. Once AIO is completed using aio_complete(), file reference is put and inode can be freed from memory. So we have to be sure that calling aio_complete() is the last thing we do with the inode. CC: xfs@oss.sgi.com CC: Ben Myers CC: stable@vger.kernel.org Signed-off-by: Jan Kara Reviewed-by: Ben Myers Signed-off-by: Ben Myers --- fs/xfs/xfs_aops.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 4111a40ebe1a..5f707e537171 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -86,11 +86,11 @@ xfs_destroy_ioend( } if (ioend->io_iocb) { + inode_dio_done(ioend->io_inode); if (ioend->io_isasync) { aio_complete(ioend->io_iocb, ioend->io_error ? ioend->io_error : ioend->io_result, 0); } - inode_dio_done(ioend->io_inode); } mempool_free(ioend, xfs_ioend_pool); -- cgit From ec2aa8e8dd7d35524a68c89b9e7ceb07fb002561 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 20 Aug 2012 17:28:58 -0700 Subject: userns: Allow the userns root to mount of devpts - The context in which devpts is mounted has no effect on the creation of ptys as the /dev/ptmx interface has been used by unprivileged users for many years. - Only support unprivileged mounts in combination with the newinstance option to ensure that mounting of /dev/pts in a user namespace will not allow the options of an existing mount of devpts to be modified. - Create /dev/pts/ptmx as the root user in the user namespace that mounts devpts so that it's permissions to be changed. Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- fs/devpts/inode.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'fs') diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 472e6befc54d..073d30b9d1ac 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -243,6 +243,13 @@ static int mknod_ptmx(struct super_block *sb) struct dentry *root = sb->s_root; struct pts_fs_info *fsi = DEVPTS_SB(sb); struct pts_mount_opts *opts = &fsi->mount_opts; + kuid_t root_uid; + kgid_t root_gid; + + root_uid = make_kuid(current_user_ns(), 0); + root_gid = make_kgid(current_user_ns(), 0); + if (!uid_valid(root_uid) || !gid_valid(root_gid)) + return -EINVAL; mutex_lock(&root->d_inode->i_mutex); @@ -273,6 +280,8 @@ static int mknod_ptmx(struct super_block *sb) mode = S_IFCHR|opts->ptmxmode; init_special_inode(inode, mode, MKDEV(TTYAUX_MAJOR, 2)); + inode->i_uid = root_uid; + inode->i_gid = root_gid; d_add(dentry, inode); @@ -438,6 +447,12 @@ static struct dentry *devpts_mount(struct file_system_type *fs_type, if (error) return ERR_PTR(error); + /* Require newinstance for all user namespace mounts to ensure + * the mount options are not changed. + */ + if ((current_user_ns() != &init_user_ns) && !opts.newinstance) + return ERR_PTR(-EINVAL); + if (opts.newinstance) s = sget(fs_type, NULL, set_anon_super, flags, NULL); else @@ -491,6 +506,9 @@ static struct file_system_type devpts_fs_type = { .name = "devpts", .mount = devpts_mount, .kill_sb = devpts_kill_sb, +#ifdef CONFIG_DEVPTS_MULTIPLE_INSTANCES + .fs_flags = FS_USERNS_MOUNT | FS_USERNS_DEV_MOUNT, +#endif }; /* -- cgit From b3c6761d9b5cc82deea7416d11c2118f6380e99c Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 25 Jan 2013 16:37:44 -0800 Subject: userns: Allow the userns root to mount ramfs. There is no backing store to ramfs and file creation rules are the same as for any other filesystem so it is semantically safe to allow unprivileged users to mount it. The memory control group successfully limits how much memory ramfs can consume on any system that cares about a user namespace root using ramfs to exhaust memory the memory control group can be deployed. Signed-off-by: "Eric W. Biederman" --- fs/ramfs/inode.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ramfs/inode.c b/fs/ramfs/inode.c index eab8c09d3801..c24f1e10b946 100644 --- a/fs/ramfs/inode.c +++ b/fs/ramfs/inode.c @@ -260,6 +260,7 @@ static struct file_system_type ramfs_fs_type = { .name = "ramfs", .mount = ramfs_mount, .kill_sb = ramfs_kill_sb, + .fs_flags = FS_USERNS_MOUNT, }; static struct file_system_type rootfs_fs_type = { .name = "rootfs", -- cgit From 6fac4829ce0ef9b7f24369086ce5f0e9f38d37bc Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 13 Nov 2012 14:20:55 +0100 Subject: cputime: Use accessors to read task cputime stats This is in preparation for the full dynticks feature. While remotely reading the cputime of a task running in a full dynticks CPU, we'll need to do some extra-computation. This way we can account the time it spent tickless in userspace since its last cputime snapshot. Signed-off-by: Frederic Weisbecker Cc: Andrew Morton Cc: Ingo Molnar Cc: Li Zhong Cc: Namhyung Kim Cc: Paul E. McKenney Cc: Paul Gortmaker Cc: Peter Zijlstra Cc: Steven Rostedt Cc: Thomas Gleixner --- fs/binfmt_elf.c | 8 ++++++-- fs/binfmt_elf_fdpic.c | 7 +++++-- fs/proc/array.c | 4 ++-- 3 files changed, 13 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 0c42cdbabecf..49d0b43458b7 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include #include @@ -1320,8 +1321,11 @@ static void fill_prstatus(struct elf_prstatus *prstatus, cputime_to_timeval(cputime.utime, &prstatus->pr_utime); cputime_to_timeval(cputime.stime, &prstatus->pr_stime); } else { - cputime_to_timeval(p->utime, &prstatus->pr_utime); - cputime_to_timeval(p->stime, &prstatus->pr_stime); + cputime_t utime, stime; + + task_cputime(p, &utime, &stime); + cputime_to_timeval(utime, &prstatus->pr_utime); + cputime_to_timeval(stime, &prstatus->pr_stime); } cputime_to_timeval(p->signal->cutime, &prstatus->pr_cutime); cputime_to_timeval(p->signal->cstime, &prstatus->pr_cstime); diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index dc84732e554f..cb240dd3b402 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -1375,8 +1375,11 @@ static void fill_prstatus(struct elf_prstatus *prstatus, cputime_to_timeval(cputime.utime, &prstatus->pr_utime); cputime_to_timeval(cputime.stime, &prstatus->pr_stime); } else { - cputime_to_timeval(p->utime, &prstatus->pr_utime); - cputime_to_timeval(p->stime, &prstatus->pr_stime); + cputime_t utime, stime; + + task_cputime(p, &utime, &stime); + cputime_to_timeval(utime, &prstatus->pr_utime); + cputime_to_timeval(stime, &prstatus->pr_stime); } cputime_to_timeval(p->signal->cutime, &prstatus->pr_cutime); cputime_to_timeval(p->signal->cstime, &prstatus->pr_cstime); diff --git a/fs/proc/array.c b/fs/proc/array.c index 6a91e6ffbcbd..f7ed9ee46eb9 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -449,7 +449,7 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns, do { min_flt += t->min_flt; maj_flt += t->maj_flt; - gtime += t->gtime; + gtime += task_gtime(t); t = next_thread(t); } while (t != task); @@ -472,7 +472,7 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns, min_flt = task->min_flt; maj_flt = task->maj_flt; task_cputime_adjusted(task, &utime, &stime); - gtime = task->gtime; + gtime = task_gtime(task); } /* scale priority and nice values from timeslices to -20..20 */ -- cgit From dee972b967ae111ad5705733de17a3bfc4632311 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 16 Jan 2013 15:05:44 -0500 Subject: NFS: Fix error reporting in nfs_xdev_mount Currently, nfs_xdev_mount converts all errors from clone_server() to ENOMEM, which can then leak to userspace (for instance to 'mount'). Fix that. Also ensure that if nfs_fs_mount_common() returns an error, we don't dprintk(0)... The regression originated in commit 3d176e3fe4f6dc379b252bf43e2e146a8f7caf01 (NFS: Use nfs_fs_mount_common() for xdev mounts) Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org [>= 3.5] --- fs/nfs/super.c | 22 +++++++++------------- 1 file changed, 9 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 2e7e8c878e5d..b056b1628722 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2589,27 +2589,23 @@ nfs_xdev_mount(struct file_system_type *fs_type, int flags, struct nfs_server *server; struct dentry *mntroot = ERR_PTR(-ENOMEM); struct nfs_subversion *nfs_mod = NFS_SB(data->sb)->nfs_client->cl_nfs_mod; - int error; - dprintk("--> nfs_xdev_mount_common()\n"); + dprintk("--> nfs_xdev_mount()\n"); mount_info.mntfh = mount_info.cloned->fh; /* create a new volume representation */ server = nfs_mod->rpc_ops->clone_server(NFS_SB(data->sb), data->fh, data->fattr, data->authflavor); - if (IS_ERR(server)) { - error = PTR_ERR(server); - goto out_err; - } - mntroot = nfs_fs_mount_common(server, flags, dev_name, &mount_info, nfs_mod); - dprintk("<-- nfs_xdev_mount_common() = 0\n"); -out: - return mntroot; + if (IS_ERR(server)) + mntroot = ERR_CAST(server); + else + mntroot = nfs_fs_mount_common(server, flags, + dev_name, &mount_info, nfs_mod); -out_err: - dprintk("<-- nfs_xdev_mount_common() = %d [error]\n", error); - goto out; + dprintk("<-- nfs_xdev_mount() = %ld\n", + IS_ERR(mntroot) ? PTR_ERR(mntroot) : 0L); + return mntroot; } #if IS_ENABLED(CONFIG_NFS_V4) -- cgit From 4ae19c2dd713edb7b8ad3d4ab9d234ed5dcb6b98 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 18 Jan 2013 22:41:53 -0500 Subject: NFSv4: Fix NFSv4 reference counting for trunked sessions The reference counting in nfs4_init_client assumes wongly that it is safe for nfs4_discover_server_trunking() to return a pointer to a nfs_client prior to bumping the reference count. Signed-off-by: Trond Myklebust Cc: Chuck Lever Cc: Ben Greear Cc: stable@vger.kernel.org [>=3.7] --- fs/nfs/nfs4client.c | 31 +++++++++++++++---------------- 1 file changed, 15 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index acc347268124..65a290a73065 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -236,11 +236,10 @@ struct nfs_client *nfs4_init_client(struct nfs_client *clp, error = nfs4_discover_server_trunking(clp, &old); if (error < 0) goto error; + nfs_put_client(clp); if (clp != old) { clp->cl_preserve_clid = true; - nfs_put_client(clp); clp = old; - atomic_inc(&clp->cl_count); } return clp; @@ -306,7 +305,7 @@ int nfs40_walk_client_list(struct nfs_client *new, .clientid = new->cl_clientid, .confirm = new->cl_confirm, }; - int status; + int status = -NFS4ERR_STALE_CLIENTID; spin_lock(&nn->nfs_client_lock); list_for_each_entry_safe(pos, n, &nn->nfs_client_list, cl_share_link) { @@ -332,28 +331,28 @@ int nfs40_walk_client_list(struct nfs_client *new, if (prev) nfs_put_client(prev); + prev = pos; status = nfs4_proc_setclientid_confirm(pos, &clid, cred); - if (status == 0) { + switch (status) { + case -NFS4ERR_STALE_CLIENTID: + break; + case 0: nfs4_swap_callback_idents(pos, new); - nfs_put_client(pos); + prev = NULL; *result = pos; dprintk("NFS: <-- %s using nfs_client = %p ({%d})\n", __func__, pos, atomic_read(&pos->cl_count)); - return 0; - } - if (status != -NFS4ERR_STALE_CLIENTID) { - nfs_put_client(pos); - dprintk("NFS: <-- %s status = %d, no result\n", - __func__, status); - return status; + default: + goto out; } spin_lock(&nn->nfs_client_lock); - prev = pos; } + spin_unlock(&nn->nfs_client_lock); +out: /* * No matching nfs_client found. This should be impossible, * because the new nfs_client has already been added to @@ -363,9 +362,8 @@ int nfs40_walk_client_list(struct nfs_client *new, */ if (prev) nfs_put_client(prev); - spin_unlock(&nn->nfs_client_lock); - pr_err("NFS: %s Error: no matching nfs_client found\n", __func__); - return -NFS4ERR_STALE_CLIENTID; + dprintk("NFS: <-- %s status = %d\n", __func__, status); + return status; } #ifdef CONFIG_NFS_V4_1 @@ -473,6 +471,7 @@ int nfs41_walk_client_list(struct nfs_client *new, if (!nfs4_match_serverowners(pos, new)) continue; + atomic_inc(&pos->cl_count); spin_unlock(&nn->nfs_client_lock); dprintk("NFS: <-- %s using nfs_client = %p ({%d})\n", __func__, pos, atomic_read(&pos->cl_count)); -- cgit From 202c312dba7d95b96493b412c606163a0cd83984 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 18 Jan 2013 22:56:23 -0500 Subject: NFSv4: Fix NFSv4 trunking discovery If walking the list in nfs4[01]_walk_client_list fails, then the most likely explanation is that the server dropped the clientid before we actually managed to confirm it. As long as our nfs_client is the very last one in the list to be tested, the caller can be assured that this is the case when the final return value is NFS4ERR_STALE_CLIENTID. Reported-by: Ben Greear Signed-off-by: Trond Myklebust Cc: Chuck Lever Cc: stable@vger.kernel.org [>=3.7] Tested-by: Ben Greear --- fs/nfs/nfs4client.c | 26 +++++++------------------- fs/nfs/nfs4state.c | 8 ++------ 2 files changed, 9 insertions(+), 25 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index 65a290a73065..2f21f17fb16c 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -352,14 +352,8 @@ int nfs40_walk_client_list(struct nfs_client *new, } spin_unlock(&nn->nfs_client_lock); + /* No match found. The server lost our clientid */ out: - /* - * No matching nfs_client found. This should be impossible, - * because the new nfs_client has already been added to - * nfs_client_list by nfs_get_client(). - * - * Don't BUG(), since the caller is holding a mutex. - */ if (prev) nfs_put_client(prev); dprintk("NFS: <-- %s status = %d\n", __func__, status); @@ -430,7 +424,7 @@ int nfs41_walk_client_list(struct nfs_client *new, { struct nfs_net *nn = net_generic(new->cl_net, nfs_net_id); struct nfs_client *pos, *n, *prev = NULL; - int error; + int status = -NFS4ERR_STALE_CLIENTID; spin_lock(&nn->nfs_client_lock); list_for_each_entry_safe(pos, n, &nn->nfs_client_list, cl_share_link) { @@ -446,8 +440,8 @@ int nfs41_walk_client_list(struct nfs_client *new, nfs_put_client(prev); prev = pos; - error = nfs_wait_client_init_complete(pos); - if (error < 0) { + status = nfs_wait_client_init_complete(pos); + if (status < 0) { nfs_put_client(pos); spin_lock(&nn->nfs_client_lock); continue; @@ -480,16 +474,10 @@ int nfs41_walk_client_list(struct nfs_client *new, return 0; } - /* - * No matching nfs_client found. This should be impossible, - * because the new nfs_client has already been added to - * nfs_client_list by nfs_get_client(). - * - * Don't BUG(), since the caller is holding a mutex. - */ + /* No matching nfs_client found. */ spin_unlock(&nn->nfs_client_lock); - pr_err("NFS: %s Error: no matching nfs_client found\n", __func__); - return -NFS4ERR_STALE_CLIENTID; + dprintk("NFS: <-- %s status = %d\n", __func__, status); + return status; } #endif /* CONFIG_NFS_V4_1 */ diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 9448c579d41a..f72561ca6895 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -136,16 +136,11 @@ int nfs40_discover_server_trunking(struct nfs_client *clp, clp->cl_confirm = clid.confirm; status = nfs40_walk_client_list(clp, result, cred); - switch (status) { - case -NFS4ERR_STALE_CLIENTID: - set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); - case 0: + if (status == 0) { /* Sustain the lease, even if it's empty. If the clientid4 * goes stale it's of no use for trunking discovery. */ nfs4_schedule_state_renewal(*result); - break; } - out: return status; } @@ -1863,6 +1858,7 @@ again: case -ETIMEDOUT: case -EAGAIN: ssleep(1); + case -NFS4ERR_STALE_CLIENTID: dprintk("NFS: %s after status %d, retrying\n", __func__, status); goto again; -- cgit From 65436ec0c8e344d9b23302b686e418f2a7b7cf7b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 18 Jan 2013 23:01:43 -0500 Subject: NFSv4.1: Ensure that nfs41_walk_client_list() does start lease recovery We do need to start the lease recovery thread prior to waiting for the client initialisation to complete in NFSv4.1. Signed-off-by: Trond Myklebust Cc: Chuck Lever Cc: Ben Greear Cc: stable@vger.kernel.org [>=3.7] --- fs/nfs/nfs4client.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index 2f21f17fb16c..2e9779b58b7a 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -440,14 +440,17 @@ int nfs41_walk_client_list(struct nfs_client *new, nfs_put_client(prev); prev = pos; + nfs4_schedule_lease_recovery(pos); status = nfs_wait_client_init_complete(pos); if (status < 0) { nfs_put_client(pos); spin_lock(&nn->nfs_client_lock); continue; } - + status = pos->cl_cons_state; spin_lock(&nn->nfs_client_lock); + if (status < 0) + continue; } if (pos->rpc_ops != new->rpc_ops) -- cgit From 8bad6fc813a3a5300f51369c39d315679fd88c72 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 28 Jan 2013 09:21:37 -0500 Subject: ext4: add punching hole support for non-extent-mapped files This patch add supports for indirect file support punching hole. It is almost the same as ext4_ext_punch_hole. First, we invalidate all pages between this hole, and then we try to deallocate all blocks of this hole. A recursive function is used to handle deallocation of blocks. In this function, it iterates over the entries in inode's i_blocks or indirect blocks, and try to free the block for each one of them. After applying this patch, xfstest #255 will not pass w/o extent because indirect-based file doesn't support unwritten extents. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 1 + fs/ext4/extents.c | 14 ++-- fs/ext4/indirect.c | 240 +++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/ext4/inode.c | 6 +- 4 files changed, 250 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 80246237f6d5..ca9294f6b730 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2103,6 +2103,7 @@ extern ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb, extern int ext4_ind_calc_metadata_amount(struct inode *inode, sector_t lblock); extern int ext4_ind_trans_blocks(struct inode *inode, int nrblocks, int chunk); extern void ext4_ind_truncate(struct inode *inode); +extern int ext4_ind_punch_hole(struct file *file, loff_t offset, loff_t length); /* ioctl.c */ extern long ext4_ioctl(struct file *, unsigned int, unsigned long); diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 391e53a52e58..566c8f3789e1 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4400,13 +4400,6 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) struct ext4_map_blocks map; unsigned int credits, blkbits = inode->i_blkbits; - /* - * currently supporting (pre)allocate mode for extent-based - * files _only_ - */ - if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) - return -EOPNOTSUPP; - /* Return error if mode is not supported */ if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) return -EOPNOTSUPP; @@ -4418,6 +4411,13 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) if (ret) return ret; + /* + * currently supporting (pre)allocate mode for extent-based + * files _only_ + */ + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) + return -EOPNOTSUPP; + trace_ext4_fallocate_enter(inode, offset, len, mode); map.m_lblk = offset >> blkbits; /* diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 8d83d1e508e4..bdd20231e66c 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -1518,3 +1518,243 @@ out_stop: trace_ext4_truncate_exit(inode); } +static int free_hole_blocks(handle_t *handle, struct inode *inode, + struct buffer_head *parent_bh, __le32 *i_data, + int level, ext4_lblk_t first, + ext4_lblk_t count, int max) +{ + struct buffer_head *bh = NULL; + int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); + int ret = 0; + int i, inc; + ext4_lblk_t offset; + __le32 blk; + + inc = 1 << ((EXT4_BLOCK_SIZE_BITS(inode->i_sb) - 2) * level); + for (i = 0, offset = 0; i < max; i++, i_data++, offset += inc) { + if (offset >= count + first) + break; + if (*i_data == 0 || (offset + inc) <= first) + continue; + blk = *i_data; + if (level > 0) { + ext4_lblk_t first2; + bh = sb_bread(inode->i_sb, blk); + if (!bh) { + EXT4_ERROR_INODE_BLOCK(inode, blk, + "Read failure"); + return -EIO; + } + first2 = (first > offset) ? first - offset : 0; + ret = free_hole_blocks(handle, inode, bh, + (__le32 *)bh->b_data, level - 1, + first2, count - offset, + inode->i_sb->s_blocksize >> 2); + if (ret) { + brelse(bh); + goto err; + } + } + if (level == 0 || + (bh && all_zeroes((__le32 *)bh->b_data, + (__le32 *)bh->b_data + addr_per_block))) { + ext4_free_data(handle, inode, parent_bh, &blk, &blk+1); + *i_data = 0; + } + brelse(bh); + bh = NULL; + } + +err: + return ret; +} + +static int ext4_free_hole_blocks(handle_t *handle, struct inode *inode, + ext4_lblk_t first, ext4_lblk_t stop) +{ + int addr_per_block = EXT4_ADDR_PER_BLOCK(inode->i_sb); + int level, ret = 0; + int num = EXT4_NDIR_BLOCKS; + ext4_lblk_t count, max = EXT4_NDIR_BLOCKS; + __le32 *i_data = EXT4_I(inode)->i_data; + + count = stop - first; + for (level = 0; level < 4; level++, max *= addr_per_block) { + if (first < max) { + ret = free_hole_blocks(handle, inode, NULL, i_data, + level, first, count, num); + if (ret) + goto err; + if (count > max - first) + count -= max - first; + else + break; + first = 0; + } else { + first -= max; + } + i_data += num; + if (level == 0) { + num = 1; + max = 1; + } + } + +err: + return ret; +} + +int ext4_ind_punch_hole(struct file *file, loff_t offset, loff_t length) +{ + struct inode *inode = file->f_path.dentry->d_inode; + struct super_block *sb = inode->i_sb; + ext4_lblk_t first_block, stop_block; + struct address_space *mapping = inode->i_mapping; + handle_t *handle = NULL; + loff_t first_page, last_page, page_len; + loff_t first_page_offset, last_page_offset; + int err = 0; + + /* + * Write out all dirty pages to avoid race conditions + * Then release them. + */ + if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { + err = filemap_write_and_wait_range(mapping, + offset, offset + length - 1); + if (err) + return err; + } + + mutex_lock(&inode->i_mutex); + /* It's not possible punch hole on append only file */ + if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) { + err = -EPERM; + goto out_mutex; + } + if (IS_SWAPFILE(inode)) { + err = -ETXTBSY; + goto out_mutex; + } + + /* No need to punch hole beyond i_size */ + if (offset >= inode->i_size) + goto out_mutex; + + /* + * If the hole extents beyond i_size, set the hole + * to end after the page that contains i_size + */ + if (offset + length > inode->i_size) { + length = inode->i_size + + PAGE_CACHE_SIZE - (inode->i_size & (PAGE_CACHE_SIZE - 1)) - + offset; + } + + first_page = (offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + last_page = (offset + length) >> PAGE_CACHE_SHIFT; + + first_page_offset = first_page << PAGE_CACHE_SHIFT; + last_page_offset = last_page << PAGE_CACHE_SHIFT; + + /* Now release the pages */ + if (last_page_offset > first_page_offset) { + truncate_pagecache_range(inode, first_page_offset, + last_page_offset - 1); + } + + /* Wait all existing dio works, newcomers will block on i_mutex */ + inode_dio_wait(inode); + + handle = start_transaction(inode); + if (IS_ERR(handle)) + goto out_mutex; + + /* + * Now we need to zero out the non-page-aligned data in the + * pages at the start and tail of the hole, and unmap the buffer + * heads for the block aligned regions of the page that were + * completely zerod. + */ + if (first_page > last_page) { + /* + * If the file space being truncated is contained within a page + * just zero out and unmap the middle of that page + */ + err = ext4_discard_partial_page_buffers(handle, + mapping, offset, length, 0); + if (err) + goto out; + } else { + /* + * Zero out and unmap the paritial page that contains + * the start of the hole + */ + page_len = first_page_offset - offset; + if (page_len > 0) { + err = ext4_discard_partial_page_buffers(handle, mapping, + offset, page_len, 0); + if (err) + goto out; + } + + /* + * Zero out and unmap the partial page that contains + * the end of the hole + */ + page_len = offset + length - last_page_offset; + if (page_len > 0) { + err = ext4_discard_partial_page_buffers(handle, mapping, + last_page_offset, page_len, 0); + if (err) + goto out; + } + } + + /* + * If i_size contained in the last page, we need to + * unmap and zero the paritial page after i_size + */ + if (inode->i_size >> PAGE_CACHE_SHIFT == last_page && + inode->i_size % PAGE_CACHE_SIZE != 0) { + page_len = PAGE_CACHE_SIZE - + (inode->i_size & (PAGE_CACHE_SIZE - 1)); + if (page_len > 0) { + err = ext4_discard_partial_page_buffers(handle, + mapping, inode->i_size, page_len, 0); + if (err) + goto out; + } + } + + first_block = (offset + sb->s_blocksize - 1) >> + EXT4_BLOCK_SIZE_BITS(sb); + stop_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb); + + if (first_block >= stop_block) + goto out; + + down_write(&EXT4_I(inode)->i_data_sem); + ext4_discard_preallocations(inode); + + err = ext4_es_remove_extent(inode, first_block, + stop_block - first_block); + err = ext4_free_hole_blocks(handle, inode, first_block, stop_block); + + ext4_discard_preallocations(inode); + + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + + up_write(&EXT4_I(inode)->i_data_sem); + +out: + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + +out_mutex: + mutex_unlock(&inode->i_mutex); + + return err; +} diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 5abf89c100b9..80683bf0df1e 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3557,10 +3557,8 @@ int ext4_punch_hole(struct file *file, loff_t offset, loff_t length) if (!S_ISREG(inode->i_mode)) return -EOPNOTSUPP; - if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { - /* TODO: Add support for non extent hole punching */ - return -EOPNOTSUPP; - } + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + return ext4_ind_punch_hole(file, offset, length); if (EXT4_SB(inode->i_sb)->s_cluster_ratio > 1) { /* TODO: Add support for bigalloc file systems */ -- cgit From 36ade451a5d736e61ac8302b64aacc5acb5e440f Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 09:30:52 -0500 Subject: ext4: Always use ext4_bio_write_page() for writeout Currently we sometimes used block_write_full_page() and sometimes ext4_bio_write_page() for writeback (depending on mount options and call path). Let's always use ext4_bio_write_page() to simplify things a bit. Reviewed-by: Zheng Liu Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 2 - fs/ext4/inode.c | 129 ++++++------------------------------------------------ fs/ext4/page-io.c | 2 - fs/ext4/super.c | 7 +-- 4 files changed, 16 insertions(+), 124 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index ca9294f6b730..0ccda0c9682e 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -215,7 +215,6 @@ typedef struct ext4_io_end { struct list_head list; /* per-file finished IO list */ struct inode *inode; /* file being written to */ unsigned int flag; /* unwritten or not */ - struct page *page; /* for writepage() path */ loff_t offset; /* offset in the file */ ssize_t size; /* size of the extent */ struct work_struct work; /* data work queue */ @@ -985,7 +984,6 @@ struct ext4_inode_info { #define EXT4_MOUNT_DIOREAD_NOLOCK 0x400000 /* Enable support for dio read nolocking */ #define EXT4_MOUNT_JOURNAL_CHECKSUM 0x800000 /* Journal checksums */ #define EXT4_MOUNT_JOURNAL_ASYNC_COMMIT 0x1000000 /* Journal Async Commit */ -#define EXT4_MOUNT_MBLK_IO_SUBMIT 0x4000000 /* multi-block io submits */ #define EXT4_MOUNT_DELALLOC 0x8000000 /* Delalloc support */ #define EXT4_MOUNT_DATA_ERR_ABORT 0x10000000 /* Abort on file data write */ #define EXT4_MOUNT_BLOCK_VALIDITY 0x20000000 /* Block validity checking */ diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 80683bf0df1e..82f934282a1f 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -134,8 +134,6 @@ static inline int ext4_begin_ordered_truncate(struct inode *inode, static void ext4_invalidatepage(struct page *page, unsigned long offset); static int noalloc_get_block_write(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create); -static int ext4_set_bh_endio(struct buffer_head *bh, struct inode *inode); -static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate); static int __ext4_journalled_writepage(struct page *page, unsigned int len); static int ext4_bh_delay_or_unwritten(handle_t *handle, struct buffer_head *bh); static int ext4_discard_partial_page_buffers_no_lock(handle_t *handle, @@ -808,11 +806,10 @@ int ext4_walk_page_buffers(handle_t *handle, * and the commit_write(). So doing the jbd2_journal_start at the start of * prepare_write() is the right place. * - * Also, this function can nest inside ext4_writepage() -> - * block_write_full_page(). In that case, we *know* that ext4_writepage() - * has generated enough buffer credits to do the whole page. So we won't - * block on the journal in that case, which is good, because the caller may - * be PF_MEMALLOC. + * Also, this function can nest inside ext4_writepage(). In that case, we + * *know* that ext4_writepage() has generated enough buffer credits to do the + * whole page. So we won't block on the journal in that case, which is good, + * because the caller may be PF_MEMALLOC. * * By accident, ext4 can be reentered when a transaction is open via * quota file writes. If we were to commit the transaction while thus @@ -1463,18 +1460,9 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, */ if (unlikely(journal_data && PageChecked(page))) err = __ext4_journalled_writepage(page, len); - else if (test_opt(inode->i_sb, MBLK_IO_SUBMIT)) + else err = ext4_bio_write_page(&io_submit, page, len, mpd->wbc); - else if (buffer_uninit(page_bufs)) { - ext4_set_bh_endio(page_bufs, inode); - err = block_write_full_page_endio(page, - noalloc_get_block_write, - mpd->wbc, ext4_end_io_buffer_write); - } else - err = block_write_full_page(page, - noalloc_get_block_write, mpd->wbc); - if (!err) mpd->pages_written++; /* @@ -1891,16 +1879,16 @@ int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, } /* - * This function is used as a standard get_block_t calback function - * when there is no desire to allocate any blocks. It is used as a - * callback function for block_write_begin() and block_write_full_page(). - * These functions should only try to map a single block at a time. + * This function is used as a standard get_block_t calback function when there + * is no desire to allocate any blocks. It is used as a callback function for + * block_write_begin(). These functions should only try to map a single block + * at a time. * * Since this function doesn't do block allocations even if the caller * requests it by passing in create=1, it is critically important that * any caller checks to make sure that any buffer heads are returned * by this function are either all already mapped or marked for - * delayed allocation before calling block_write_full_page(). Otherwise, + * delayed allocation before calling ext4_bio_write_page(). Otherwise, * b_blocknr could be left unitialized, and the page write functions will * be taken by surprise. */ @@ -2040,6 +2028,7 @@ static int ext4_writepage(struct page *page, unsigned int len; struct buffer_head *page_bufs = NULL; struct inode *inode = page->mapping->host; + struct ext4_io_submit io_submit; trace_ext4_writepage(page); size = i_size_read(inode); @@ -2089,14 +2078,9 @@ static int ext4_writepage(struct page *page, */ return __ext4_journalled_writepage(page, len); - if (buffer_uninit(page_bufs)) { - ext4_set_bh_endio(page_bufs, inode); - ret = block_write_full_page_endio(page, noalloc_get_block_write, - wbc, ext4_end_io_buffer_write); - } else - ret = block_write_full_page(page, noalloc_get_block_write, - wbc); - + memset(&io_submit, 0, sizeof(io_submit)); + ret = ext4_bio_write_page(&io_submit, page, len, wbc); + ext4_io_submit(&io_submit); return ret; } @@ -2858,36 +2842,10 @@ ext4_readpages(struct file *file, struct address_space *mapping, return mpage_readpages(mapping, pages, nr_pages, ext4_get_block); } -static void ext4_invalidatepage_free_endio(struct page *page, unsigned long offset) -{ - struct buffer_head *head, *bh; - unsigned int curr_off = 0; - - if (!page_has_buffers(page)) - return; - head = bh = page_buffers(page); - do { - if (offset <= curr_off && test_clear_buffer_uninit(bh) - && bh->b_private) { - ext4_free_io_end(bh->b_private); - bh->b_private = NULL; - bh->b_end_io = NULL; - } - curr_off = curr_off + bh->b_size; - bh = bh->b_this_page; - } while (bh != head); -} - static void ext4_invalidatepage(struct page *page, unsigned long offset) { trace_ext4_invalidatepage(page, offset); - /* - * free any io_end structure allocated for buffers to be discarded - */ - if (ext4_should_dioread_nolock(page->mapping->host)) - ext4_invalidatepage_free_endio(page, offset); - /* No journalling happens on data buffers when this function is used */ WARN_ON(page_has_buffers(page) && buffer_jbd(page_buffers(page))); @@ -2993,65 +2951,6 @@ out: ext4_add_complete_io(io_end); } -static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate) -{ - ext4_io_end_t *io_end = bh->b_private; - struct inode *inode; - - if (!test_clear_buffer_uninit(bh) || !io_end) - goto out; - - if (!(io_end->inode->i_sb->s_flags & MS_ACTIVE)) { - ext4_msg(io_end->inode->i_sb, KERN_INFO, - "sb umounted, discard end_io request for inode %lu", - io_end->inode->i_ino); - ext4_free_io_end(io_end); - goto out; - } - - /* - * It may be over-defensive here to check EXT4_IO_END_UNWRITTEN now, - * but being more careful is always safe for the future change. - */ - inode = io_end->inode; - ext4_set_io_unwritten_flag(inode, io_end); - ext4_add_complete_io(io_end); -out: - bh->b_private = NULL; - bh->b_end_io = NULL; - clear_buffer_uninit(bh); - end_buffer_async_write(bh, uptodate); -} - -static int ext4_set_bh_endio(struct buffer_head *bh, struct inode *inode) -{ - ext4_io_end_t *io_end; - struct page *page = bh->b_page; - loff_t offset = (sector_t)page->index << PAGE_CACHE_SHIFT; - size_t size = bh->b_size; - -retry: - io_end = ext4_init_io_end(inode, GFP_ATOMIC); - if (!io_end) { - pr_warn_ratelimited("%s: allocation fail\n", __func__); - schedule(); - goto retry; - } - io_end->offset = offset; - io_end->size = size; - /* - * We need to hold a reference to the page to make sure it - * doesn't get evicted before ext4_end_io_work() has a chance - * to convert the extent from written to unwritten. - */ - io_end->page = page; - get_page(io_end->page); - - bh->b_private = io_end; - bh->b_end_io = ext4_end_io_buffer_write; - return 0; -} - /* * For ext4 extent files, ext4 will do direct-io write to holes, * preallocated extents, and those write extend the file, no need to diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 0016fbca2a40..ddb3d401543c 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -73,8 +73,6 @@ void ext4_free_io_end(ext4_io_end_t *io) BUG_ON(!list_empty(&io->list)); BUG_ON(io->flag & EXT4_IO_END_UNWRITTEN); - if (io->page) - put_page(io->page); for (i = 0; i < io->num_io_pages; i++) put_io_page(io->pages[i]); io->num_io_pages = 0; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 3b04d15f303d..d5d336bd2180 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1280,8 +1280,8 @@ static const match_table_t tokens = { {Opt_stripe, "stripe=%u"}, {Opt_delalloc, "delalloc"}, {Opt_nodelalloc, "nodelalloc"}, - {Opt_mblk_io_submit, "mblk_io_submit"}, - {Opt_nomblk_io_submit, "nomblk_io_submit"}, + {Opt_removed, "mblk_io_submit"}, + {Opt_removed, "nomblk_io_submit"}, {Opt_block_validity, "block_validity"}, {Opt_noblock_validity, "noblock_validity"}, {Opt_inode_readahead_blks, "inode_readahead_blks=%u"}, @@ -1415,8 +1415,6 @@ static const struct mount_opts { {Opt_bsd_df, EXT4_MOUNT_MINIX_DF, MOPT_CLEAR}, {Opt_grpid, EXT4_MOUNT_GRPID, MOPT_SET}, {Opt_nogrpid, EXT4_MOUNT_GRPID, MOPT_CLEAR}, - {Opt_mblk_io_submit, EXT4_MOUNT_MBLK_IO_SUBMIT, MOPT_SET}, - {Opt_nomblk_io_submit, EXT4_MOUNT_MBLK_IO_SUBMIT, MOPT_CLEAR}, {Opt_block_validity, EXT4_MOUNT_BLOCK_VALIDITY, MOPT_SET}, {Opt_noblock_validity, EXT4_MOUNT_BLOCK_VALIDITY, MOPT_CLEAR}, {Opt_dioread_nolock, EXT4_MOUNT_DIOREAD_NOLOCK, MOPT_SET}, @@ -3381,7 +3379,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) #ifdef CONFIG_EXT4_FS_POSIX_ACL set_opt(sb, POSIX_ACL); #endif - set_opt(sb, MBLK_IO_SUBMIT); if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_DATA) set_opt(sb, JOURNAL_DATA); else if ((def_mount_opts & EXT4_DEFM_JMODE) == EXT4_DEFM_JMODE_ORDERED) -- cgit From 1ae48a6354a364413d372df1525d523a3fb4fb8c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 09:32:54 -0500 Subject: ext4: use redirty_page_for_writepage() in ext4_bio_write_page() When we cannot write a page we should use redirty_page_for_writepage() instead of plain set_page_dirty(). That tells writeback code we have problems, redirties only the page (redirtying buffers is not needed), and updates mm accounting of failed page writes. Also move clearing of buffer dirty flag after io_submit_add_bh(). At that moment we are sure buffer will be going to disk. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/page-io.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index ddb3d401543c..05795f10e55a 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -23,6 +23,7 @@ #include #include #include +#include #include "ext4_jbd2.h" #include "xattr.h" @@ -434,7 +435,7 @@ int ext4_bio_write_page(struct ext4_io_submit *io, io_page = kmem_cache_alloc(io_page_cachep, GFP_NOFS); if (!io_page) { - set_page_dirty(page); + redirty_page_for_writepage(wbc, page); unlock_page(page); return -ENOMEM; } @@ -466,7 +467,6 @@ int ext4_bio_write_page(struct ext4_io_submit *io, set_buffer_uptodate(bh); continue; } - clear_buffer_dirty(bh); ret = io_submit_add_bh(io, io_page, inode, wbc, bh); if (ret) { /* @@ -474,9 +474,10 @@ int ext4_bio_write_page(struct ext4_io_submit *io, * we can do but mark the page as dirty, and * better luck next time. */ - set_page_dirty(page); + redirty_page_for_writepage(wbc, page); break; } + clear_buffer_dirty(bh); } unlock_page(page); /* -- cgit From fe089c77f1466c74f0f19ad2475b1630216b8b19 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 09:38:49 -0500 Subject: ext4: remove __ext4_journalled_writepage() from mpage_da_submit_io() We don't support delayed allocation in data=journal mode. So checking for it in mpage_da_submit_io() doesn't make really sence. If we ever decide to extend delayed allocation support to data=journal mode, adding __ext4_journalled_writepage() call will be the least of problems we have to solve. Most likely we'd have to implement separate writepages call anyways because we don't have transaction credits for writing more than a single page so mapping of page buffers would have to be done differently. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 13 ++----------- 1 file changed, 2 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 82f934282a1f..cbbf58327a29 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1354,7 +1354,6 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, loff_t size = i_size_read(inode); unsigned int len, block_start; struct buffer_head *bh, *page_bufs = NULL; - int journal_data = ext4_should_journal_data(inode); sector_t pblock = 0, cur_logical = 0; struct ext4_io_submit io_submit; @@ -1453,16 +1452,8 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, block_commit_write(page, 0, len); clear_page_dirty_for_io(page); - /* - * Delalloc doesn't support data journalling, - * but eventually maybe we'll lift this - * restriction. - */ - if (unlikely(journal_data && PageChecked(page))) - err = __ext4_journalled_writepage(page, len); - else - err = ext4_bio_write_page(&io_submit, page, - len, mpd->wbc); + err = ext4_bio_write_page(&io_submit, page, len, + mpd->wbc); if (!err) mpd->pages_written++; /* -- cgit From 84c17543ab5685d950da73209df0ecda26e72d3b Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 09:43:46 -0500 Subject: ext4: move work from io_end to inode It does not make much sense to have struct work in ext4_io_end_t because we always use it for only one ext4_io_end_t per inode (the first one in the i_completed_io list). So just move the structure to inode itself. This also allows for a small simplification in processing io_end structures. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 6 +++--- fs/ext4/page-io.c | 33 +++++++++------------------------ fs/ext4/super.c | 1 + 3 files changed, 13 insertions(+), 27 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 0ccda0c9682e..d93393eb5f2d 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -194,8 +194,7 @@ struct mpage_da_data { */ #define EXT4_IO_END_UNWRITTEN 0x0001 #define EXT4_IO_END_ERROR 0x0002 -#define EXT4_IO_END_QUEUED 0x0004 -#define EXT4_IO_END_DIRECT 0x0008 +#define EXT4_IO_END_DIRECT 0x0004 struct ext4_io_page { struct page *p_page; @@ -217,7 +216,6 @@ typedef struct ext4_io_end { unsigned int flag; /* unwritten or not */ loff_t offset; /* offset in the file */ ssize_t size; /* size of the extent */ - struct work_struct work; /* data work queue */ struct kiocb *iocb; /* iocb struct for AIO */ int result; /* error value for AIO */ int num_io_pages; /* for writepages() */ @@ -929,6 +927,7 @@ struct ext4_inode_info { spinlock_t i_completed_io_lock; atomic_t i_ioend_count; /* Number of outstanding io_end structs */ atomic_t i_unwritten; /* Nr. of inflight conversions pending */ + struct work_struct i_unwritten_work; /* deferred extent conversion */ spinlock_t i_block_reservation_lock; @@ -2538,6 +2537,7 @@ extern void ext4_exit_pageio(void); extern void ext4_ioend_wait(struct inode *); extern void ext4_free_io_end(ext4_io_end_t *io); extern ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags); +extern void ext4_end_io_work(struct work_struct *work); extern void ext4_io_submit(struct ext4_io_submit *io); extern int ext4_bio_write_page(struct ext4_io_submit *io, struct page *page, diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 05795f10e55a..a0290176ee75 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -151,16 +151,13 @@ void ext4_add_complete_io(ext4_io_end_t *io_end) wq = EXT4_SB(io_end->inode->i_sb)->dio_unwritten_wq; spin_lock_irqsave(&ei->i_completed_io_lock, flags); - if (list_empty(&ei->i_completed_io_list)) { - io_end->flag |= EXT4_IO_END_QUEUED; - queue_work(wq, &io_end->work); - } + if (list_empty(&ei->i_completed_io_list)) + queue_work(wq, &ei->i_unwritten_work); list_add_tail(&io_end->list, &ei->i_completed_io_list); spin_unlock_irqrestore(&ei->i_completed_io_lock, flags); } -static int ext4_do_flush_completed_IO(struct inode *inode, - ext4_io_end_t *work_io) +static int ext4_do_flush_completed_IO(struct inode *inode) { ext4_io_end_t *io; struct list_head unwritten, complete, to_free; @@ -191,19 +188,7 @@ static int ext4_do_flush_completed_IO(struct inode *inode, while (!list_empty(&complete)) { io = list_entry(complete.next, ext4_io_end_t, list); io->flag &= ~EXT4_IO_END_UNWRITTEN; - /* end_io context can not be destroyed now because it still - * used by queued worker. Worker thread will destroy it later */ - if (io->flag & EXT4_IO_END_QUEUED) - list_del_init(&io->list); - else - list_move(&io->list, &to_free); - } - /* If we are called from worker context, it is time to clear queued - * flag, and destroy it's end_io if it was converted already */ - if (work_io) { - work_io->flag &= ~EXT4_IO_END_QUEUED; - if (!(work_io->flag & EXT4_IO_END_UNWRITTEN)) - list_add_tail(&work_io->list, &to_free); + list_move(&io->list, &to_free); } spin_unlock_irqrestore(&ei->i_completed_io_lock, flags); @@ -218,10 +203,11 @@ static int ext4_do_flush_completed_IO(struct inode *inode, /* * work on completed aio dio IO, to convert unwritten extents to extents */ -static void ext4_end_io_work(struct work_struct *work) +void ext4_end_io_work(struct work_struct *work) { - ext4_io_end_t *io = container_of(work, ext4_io_end_t, work); - ext4_do_flush_completed_IO(io->inode, io); + struct ext4_inode_info *ei = container_of(work, struct ext4_inode_info, + i_unwritten_work); + ext4_do_flush_completed_IO(&ei->vfs_inode); } int ext4_flush_unwritten_io(struct inode *inode) @@ -229,7 +215,7 @@ int ext4_flush_unwritten_io(struct inode *inode) int ret; WARN_ON_ONCE(!mutex_is_locked(&inode->i_mutex) && !(inode->i_state & I_FREEING)); - ret = ext4_do_flush_completed_IO(inode, NULL); + ret = ext4_do_flush_completed_IO(inode); ext4_unwritten_wait(inode); return ret; } @@ -240,7 +226,6 @@ ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags) if (io) { atomic_inc(&EXT4_I(inode)->i_ioend_count); io->inode = inode; - INIT_WORK(&io->work, ext4_end_io_work); INIT_LIST_HEAD(&io->list); } return io; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index d5d336bd2180..dc0fb7b942cf 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -960,6 +960,7 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) ei->i_datasync_tid = 0; atomic_set(&ei->i_ioend_count, 0); atomic_set(&ei->i_unwritten, 0); + INIT_WORK(&ei->i_unwritten_work, ext4_end_io_work); return &ei->vfs_inode; } -- cgit From 002bd7fa3ac7441bdb36df67b2c64bc8c1be5360 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 09:49:15 -0500 Subject: ext4: simplify list handling in ext4_do_flush_completed_IO() The function splices i_completed_io_list to its private list first. From that moment on we don't need any lock for working with io_end structures because all io_end structure on the list are only our own. So we can remove the other two lists in the function and free io_end immediately after we are done with it. CC: Dmitry Monakhov Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/page-io.c | 18 +----------------- 1 file changed, 1 insertion(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index a0290176ee75..3fb385cd9670 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -160,14 +160,11 @@ void ext4_add_complete_io(ext4_io_end_t *io_end) static int ext4_do_flush_completed_IO(struct inode *inode) { ext4_io_end_t *io; - struct list_head unwritten, complete, to_free; + struct list_head unwritten; unsigned long flags; struct ext4_inode_info *ei = EXT4_I(inode); int err, ret = 0; - INIT_LIST_HEAD(&complete); - INIT_LIST_HEAD(&to_free); - spin_lock_irqsave(&ei->i_completed_io_lock, flags); dump_completed_IO(inode); list_replace_init(&ei->i_completed_io_list, &unwritten); @@ -181,20 +178,7 @@ static int ext4_do_flush_completed_IO(struct inode *inode) err = ext4_end_io(io); if (unlikely(!ret && err)) ret = err; - - list_add_tail(&io->list, &complete); - } - spin_lock_irqsave(&ei->i_completed_io_lock, flags); - while (!list_empty(&complete)) { - io = list_entry(complete.next, ext4_io_end_t, list); io->flag &= ~EXT4_IO_END_UNWRITTEN; - list_move(&io->list, &to_free); - } - spin_unlock_irqrestore(&ei->i_completed_io_lock, flags); - - while (!list_empty(&to_free)) { - io = list_entry(to_free.next, ext4_io_end_t, list); - list_del_init(&io->list); ext4_free_io_end(io); } return ret; -- cgit From f8bec37037aceb126d695c021cf4dc93b7238d47 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 12:55:08 -0500 Subject: ext4: dirty page has always buffers attached ext4_writepage(), write_cache_pages_da(), and mpage_da_submit_io() doesn't have to deal with the case when page doesn't have buffers. We attach buffers to a page in ->write_begin() and ->page_mkwrite() which covers all places where a page can become dirty. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 147 +++++++++++++++----------------------------------------- 1 file changed, 38 insertions(+), 109 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index cbbf58327a29..8a89cbbf0f1a 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -132,8 +132,6 @@ static inline int ext4_begin_ordered_truncate(struct inode *inode, } static void ext4_invalidatepage(struct page *page, unsigned long offset); -static int noalloc_get_block_write(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create); static int __ext4_journalled_writepage(struct page *page, unsigned int len); static int ext4_bh_delay_or_unwritten(handle_t *handle, struct buffer_head *bh); static int ext4_discard_partial_page_buffers_no_lock(handle_t *handle, @@ -1374,7 +1372,7 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, if (nr_pages == 0) break; for (i = 0; i < nr_pages; i++) { - int commit_write = 0, skip_page = 0; + int skip_page = 0; struct page *page = pvec.pages[i]; index = page->index; @@ -1396,27 +1394,9 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, BUG_ON(!PageLocked(page)); BUG_ON(PageWriteback(page)); - /* - * If the page does not have buffers (for - * whatever reason), try to create them using - * __block_write_begin. If this fails, - * skip the page and move on. - */ - if (!page_has_buffers(page)) { - if (__block_write_begin(page, 0, len, - noalloc_get_block_write)) { - skip_page: - unlock_page(page); - continue; - } - commit_write = 1; - } - bh = page_bufs = page_buffers(page); block_start = 0; do { - if (!bh) - goto skip_page; if (map && (cur_logical >= map->m_lblk) && (cur_logical <= (map->m_lblk + (map->m_len - 1)))) { @@ -1444,12 +1424,10 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, pblock++; } while (bh != page_bufs); - if (skip_page) - goto skip_page; - - if (commit_write) - /* mark the buffer_heads as dirty & uptodate */ - block_commit_write(page, 0, len); + if (skip_page) { + unlock_page(page); + continue; + } clear_page_dirty_for_io(page); err = ext4_bio_write_page(&io_submit, page, len, @@ -1869,27 +1847,6 @@ int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, return 0; } -/* - * This function is used as a standard get_block_t calback function when there - * is no desire to allocate any blocks. It is used as a callback function for - * block_write_begin(). These functions should only try to map a single block - * at a time. - * - * Since this function doesn't do block allocations even if the caller - * requests it by passing in create=1, it is critically important that - * any caller checks to make sure that any buffer heads are returned - * by this function are either all already mapped or marked for - * delayed allocation before calling ext4_bio_write_page(). Otherwise, - * b_blocknr could be left unitialized, and the page write functions will - * be taken by surprise. - */ -static int noalloc_get_block_write(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create) -{ - BUG_ON(bh_result->b_size != inode->i_sb->s_blocksize); - return _ext4_get_block(inode, iblock, bh_result, 0); -} - static int bget_one(handle_t *handle, struct buffer_head *bh) { get_bh(bh); @@ -2014,7 +1971,7 @@ out: static int ext4_writepage(struct page *page, struct writeback_control *wbc) { - int ret = 0, commit_write = 0; + int ret = 0; loff_t size; unsigned int len; struct buffer_head *page_bufs = NULL; @@ -2028,21 +1985,6 @@ static int ext4_writepage(struct page *page, else len = PAGE_CACHE_SIZE; - /* - * If the page does not have buffers (for whatever reason), - * try to create them using __block_write_begin. If this - * fails, redirty the page and move on. - */ - if (!page_has_buffers(page)) { - if (__block_write_begin(page, 0, len, - noalloc_get_block_write)) { - redirty_page: - redirty_page_for_writepage(wbc, page); - unlock_page(page); - return 0; - } - commit_write = 1; - } page_bufs = page_buffers(page); if (ext4_walk_page_buffers(NULL, page_bufs, 0, len, NULL, ext4_bh_delay_or_unwritten)) { @@ -2056,11 +1998,10 @@ static int ext4_writepage(struct page *page, */ WARN_ON_ONCE((current->flags & (PF_MEMALLOC|PF_KSWAPD)) == PF_MEMALLOC); - goto redirty_page; + redirty_page_for_writepage(wbc, page); + unlock_page(page); + return 0; } - if (commit_write) - /* now mark the buffer_heads as dirty and uptodate */ - block_commit_write(page, 0, len); if (PageChecked(page) && ext4_should_journal_data(inode)) /* @@ -2203,51 +2144,39 @@ static int write_cache_pages_da(handle_t *handle, logical = (sector_t) page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits); - if (!page_has_buffers(page)) { - mpage_add_bh_to_extent(mpd, logical, - PAGE_CACHE_SIZE, - (1 << BH_Dirty) | (1 << BH_Uptodate)); - if (mpd->io_done) - goto ret_extent_tail; - } else { + /* Add all dirty buffers to mpd */ + head = page_buffers(page); + bh = head; + do { + BUG_ON(buffer_locked(bh)); /* - * Page with regular buffer heads, - * just add all dirty ones + * We need to try to allocate unmapped blocks + * in the same page. Otherwise we won't make + * progress with the page in ext4_writepage */ - head = page_buffers(page); - bh = head; - do { - BUG_ON(buffer_locked(bh)); + if (ext4_bh_delay_or_unwritten(NULL, bh)) { + mpage_add_bh_to_extent(mpd, logical, + bh->b_size, + bh->b_state); + if (mpd->io_done) + goto ret_extent_tail; + } else if (buffer_dirty(bh) && + buffer_mapped(bh)) { /* - * We need to try to allocate - * unmapped blocks in the same page. - * Otherwise we won't make progress - * with the page in ext4_writepage + * mapped dirty buffer. We need to + * update the b_state because we look + * at b_state in mpage_da_map_blocks. + * We don't update b_size because if we + * find an unmapped buffer_head later + * we need to use the b_state flag of + * that buffer_head. */ - if (ext4_bh_delay_or_unwritten(NULL, bh)) { - mpage_add_bh_to_extent(mpd, logical, - bh->b_size, - bh->b_state); - if (mpd->io_done) - goto ret_extent_tail; - } else if (buffer_dirty(bh) && (buffer_mapped(bh))) { - /* - * mapped dirty buffer. We need - * to update the b_state - * because we look at b_state - * in mpage_da_map_blocks. We - * don't update b_size because - * if we find an unmapped - * buffer_head later we need to - * use the b_state flag of that - * buffer_head. - */ - if (mpd->b_size == 0) - mpd->b_state = bh->b_state & BH_FLAGS; - } - logical++; - } while ((bh = bh->b_this_page) != head); - } + if (mpd->b_size == 0) + mpd->b_state = + bh->b_state & BH_FLAGS; + } + logical++; + } while ((bh = bh->b_this_page) != head); if (nr_to_write > 0) { nr_to_write--; -- cgit From b6a8e62f8b0aec7607c947ba0d37d30fef65440f Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 13:06:48 -0500 Subject: ext4: simplify mpage_add_bh_to_extent() The argument b_size of mpage_add_bh_to_extent() was bogus since it was always == blocksize (which we can easily derive from inode->i_blkbits). Also second branch of condition: if (nrblocks >= EXT4_MAX_TRANS_DATA) { } else if ((nrblocks + (b_size >> mpd->inode->i_blkbits)) > EXT4_MAX_TRANS_DATA) { } was never taken because (b_size >> mpd->inode->i_blkbits) == 1. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 29 +++++++++-------------------- 1 file changed, 9 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 8a89cbbf0f1a..6824cb1bd1bb 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1647,16 +1647,16 @@ submit_io: * * @mpd->lbh - extent of blocks * @logical - logical number of the block in the file - * @bh - bh of the block (used to access block's state) + * @b_state - b_state of the buffer head added * * the function is used to collect contig. blocks in same state */ -static void mpage_add_bh_to_extent(struct mpage_da_data *mpd, - sector_t logical, size_t b_size, +static void mpage_add_bh_to_extent(struct mpage_da_data *mpd, sector_t logical, unsigned long b_state) { sector_t next; - int nrblocks = mpd->b_size >> mpd->inode->i_blkbits; + int blkbits = mpd->inode->i_blkbits; + int nrblocks = mpd->b_size >> blkbits; /* * XXX Don't go larger than mballoc is willing to allocate @@ -1664,11 +1664,11 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd, * mpage_da_submit_io() into this function and then call * ext4_map_blocks() multiple times in a loop */ - if (nrblocks >= 8*1024*1024/mpd->inode->i_sb->s_blocksize) + if (nrblocks >= (8*1024*1024 >> blkbits)) goto flush_it; - /* check if thereserved journal credits might overflow */ - if (!(ext4_test_inode_flag(mpd->inode, EXT4_INODE_EXTENTS))) { + /* check if the reserved journal credits might overflow */ + if (!ext4_test_inode_flag(mpd->inode, EXT4_INODE_EXTENTS)) { if (nrblocks >= EXT4_MAX_TRANS_DATA) { /* * With non-extent format we are limited by the journal @@ -1677,16 +1677,6 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd, * nrblocks. So limit nrblocks. */ goto flush_it; - } else if ((nrblocks + (b_size >> mpd->inode->i_blkbits)) > - EXT4_MAX_TRANS_DATA) { - /* - * Adding the new buffer_head would make it cross the - * allowed limit for which we have journal credit - * reserved. So limit the new bh->b_size - */ - b_size = (EXT4_MAX_TRANS_DATA - nrblocks) << - mpd->inode->i_blkbits; - /* we will do mpage_da_submit_io in the next loop */ } } /* @@ -1694,7 +1684,7 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd, */ if (mpd->b_size == 0) { mpd->b_blocknr = logical; - mpd->b_size = b_size; + mpd->b_size = 1 << blkbits; mpd->b_state = b_state & BH_FLAGS; return; } @@ -1704,7 +1694,7 @@ static void mpage_add_bh_to_extent(struct mpage_da_data *mpd, * Can we merge the block to our big extent? */ if (logical == next && (b_state & BH_FLAGS) == mpd->b_state) { - mpd->b_size += b_size; + mpd->b_size += 1 << blkbits; return; } @@ -2156,7 +2146,6 @@ static int write_cache_pages_da(handle_t *handle, */ if (ext4_bh_delay_or_unwritten(NULL, bh)) { mpage_add_bh_to_extent(mpd, logical, - bh->b_size, bh->b_state); if (mpd->io_done) goto ret_extent_tail; -- cgit From 2729423cf2ef3ac51d040f24a5ddd84c1b2acc70 Mon Sep 17 00:00:00 2001 From: Torsten Kaiser Date: Sun, 20 Jan 2013 10:24:49 +0100 Subject: xfs: Fix xfs_swap_extents() after removal of xfs_flushinval_pages() Commit fb59581404ab7ec5075299065c22cb211a9262a9 removed xfs_flushinval_pages() and changed its callers to use filemap_write_and_wait() and truncate_pagecache_range() directly. But in xfs_swap_extents() this change accidental switched the argument for 'tip' to 'ip'. This patch switches it back to 'tip' Signed-off-by: Torsten Kaiser Reviewed-by: Ben Myers Signed-off-by: Ben Myers --- fs/xfs/xfs_dfrag.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index d0e9c74d3d96..a8bd26b82ecb 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -246,10 +246,10 @@ xfs_swap_extents( goto out_unlock; } - error = -filemap_write_and_wait(VFS_I(ip)->i_mapping); + error = -filemap_write_and_wait(VFS_I(tip)->i_mapping); if (error) goto out_unlock; - truncate_pagecache_range(VFS_I(ip), 0, -1); + truncate_pagecache_range(VFS_I(tip), 0, -1); /* Verify O_DIRECT for ftmp */ if (VN_CACHED(VFS_I(tip)) != 0) { -- cgit From a07c48ad5be5cbf32a2741f10e6fbf4bbfcaa362 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 26 Jan 2013 10:48:42 +0300 Subject: eCryptfs: remove unneeded checks in virt_to_scatterlist() This is always called with a valid "sg" pointer. My static checker complains because the call to sg_init_table() dereferences "sg" before we reach the checks. Signed-off-by: Dan Carpenter Signed-off-by: Tyler Hicks --- fs/ecryptfs/crypto.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index a7b0c2dfb3db..d5c25db4398f 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -301,17 +301,14 @@ int virt_to_scatterlist(const void *addr, int size, struct scatterlist *sg, while (size > 0 && i < sg_size) { pg = virt_to_page(addr); offset = offset_in_page(addr); - if (sg) - sg_set_page(&sg[i], pg, 0, offset); + sg_set_page(&sg[i], pg, 0, offset); remainder_of_page = PAGE_CACHE_SIZE - offset; if (size >= remainder_of_page) { - if (sg) - sg[i].length = remainder_of_page; + sg[i].length = remainder_of_page; addr += remainder_of_page; size -= remainder_of_page; } else { - if (sg) - sg[i].length = size; + sg[i].length = size; addr += size; size = 0; } -- cgit From 8a850c3fb8d0f204eabc1a32b502f47d3c16eac4 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 20:53:28 -0500 Subject: ext4: Make ext4_bio_writepage() handle unprepared buffers So far ext4_bio_writepage() unconditionally cleared dirty bit on all buffers underlying the page. That implicitely assumes we can write all buffers. So far that is true because callers call into ext4_bio_writepage() make sure all buffers in the page are mapped but: a) it's a data corruption bug waiting to happen b) in data=ordered mode when blocksize < pagesize we do need to write pages that may have only some of dirty buffers mapped. So change ext4_bio_writepage() to skip buffers that cannot be written without clearing their dirty bit. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/page-io.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 3fb385cd9670..0290bf85f97e 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -350,14 +350,6 @@ static int io_submit_add_bh(struct ext4_io_submit *io, unmap_underlying_metadata(bh->b_bdev, bh->b_blocknr); } - if (!buffer_mapped(bh) || buffer_delay(bh)) { - if (!buffer_mapped(bh)) - clear_buffer_dirty(bh); - if (io->io_bio) - ext4_io_submit(io); - return 0; - } - if (io->io_bio && bh->b_blocknr != io->io_next_block) { submit_and_retry: ext4_io_submit(io); @@ -436,6 +428,15 @@ int ext4_bio_write_page(struct ext4_io_submit *io, set_buffer_uptodate(bh); continue; } + if (!buffer_dirty(bh) || buffer_delay(bh) || + !buffer_mapped(bh) || buffer_unwritten(bh)) { + /* A hole? We can safely clear the dirty bit */ + if (!buffer_mapped(bh)) + clear_buffer_dirty(bh); + if (io->io_bio) + ext4_io_submit(io); + continue; + } ret = io_submit_add_bh(io, io_page, inode, wbc, bh); if (ret) { /* -- cgit From fe386132f6731d02a45c380be0a3d339e6446cb5 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 28 Jan 2013 21:06:42 -0500 Subject: ext4: fix ext4_writepage() to achieve data=ordered guarantees So far ext4_writepage() skipped writing pages that had any delayed or unwritten buffers attached. When blocksize < pagesize this breaks data=ordered mode guarantees as we can have a page with one freshly allocated buffer whose allocation is part of the committing transaction and another buffer in the page which is delayed or unwritten. So fix this problem by calling ext4_bio_writepage() anyway. It will submit mapped buffers and leave others alone. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 30 ++++++++++++++++++------------ 1 file changed, 18 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 6824cb1bd1bb..86bf43d6dfcd 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1976,21 +1976,27 @@ static int ext4_writepage(struct page *page, len = PAGE_CACHE_SIZE; page_bufs = page_buffers(page); + /* + * We cannot do block allocation or other extent handling in this + * function. If there are buffers needing that, we have to redirty + * the page. But we may reach here when we do a journal commit via + * journal_submit_inode_data_buffers() and in that case we must write + * allocated buffers to achieve data=ordered mode guarantees. + */ if (ext4_walk_page_buffers(NULL, page_bufs, 0, len, NULL, ext4_bh_delay_or_unwritten)) { - /* - * We don't want to do block allocation, so redirty - * the page and return. We may reach here when we do - * a journal commit via journal_submit_inode_data_buffers. - * We can also reach here via shrink_page_list but it - * should never be for direct reclaim so warn if that - * happens - */ - WARN_ON_ONCE((current->flags & (PF_MEMALLOC|PF_KSWAPD)) == - PF_MEMALLOC); redirty_page_for_writepage(wbc, page); - unlock_page(page); - return 0; + if (current->flags & PF_MEMALLOC) { + /* + * For memory cleaning there's no point in writing only + * some buffers. So just bail out. Warn if we came here + * from direct reclaim. + */ + WARN_ON_ONCE((current->flags & (PF_MEMALLOC|PF_KSWAPD)) + == PF_MEMALLOC); + unlock_page(page); + return 0; + } } if (PageChecked(page) && ext4_should_journal_data(inode)) -- cgit From cfa7275482414fa87c9e51dd7b9d4d5d3f7a7fed Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Mon, 28 Jan 2013 21:14:11 -0500 Subject: ext4: remove unused variable flags Remove unused variable flags from dump_completed_IO(). The code is only exercised when EXT4FS_DEBUG is defined. Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" Reviewed-by: Zheng Liu --- fs/ext4/page-io.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 0290bf85f97e..5d8c66948e1b 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -118,7 +118,6 @@ static void dump_completed_IO(struct inode *inode) #ifdef EXT4FS_DEBUG struct list_head *cur, *before, *after; ext4_io_end_t *io, *io0, *io1; - unsigned long flags; if (list_empty(&EXT4_I(inode)->i_completed_io_list)) { ext4_debug("inode %lu completed_io list is empty\n", -- cgit From b06acd38a44127b382fa53e49878f7a2b70af6f2 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Mon, 28 Jan 2013 21:21:12 -0500 Subject: ext4: remove explicit WARN_ON when ext4_map_blocks() fails In two places we call WARN_ON() before we print out the debug message, however we agreed that the WARN_ON() is unnecessary at those places so remove them. Also use ext4_warning() instead of ext4_msg() and printk(). Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" --- fs/ext4/extents.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 566c8f3789e1..db55c62e1f01 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4462,11 +4462,11 @@ retry: ret = ext4_map_blocks(handle, inode, &map, flags); if (ret <= 0) { #ifdef EXT4FS_DEBUG - WARN_ON(ret <= 0); - printk(KERN_ERR "%s: ext4_ext_map_blocks " - "returned error inode#%lu, block=%u, " - "max_blocks=%u", __func__, - inode->i_ino, map.m_lblk, max_blocks); + ext4_warning(inode->i_sb, + "inode #%lu: block %u: len %u: " + "ext4_ext_map_blocks returned %d", + inode->i_ino, map.m_lblk, + map.m_len, ret); #endif ext4_mark_inode_dirty(handle, inode); ret2 = ext4_journal_stop(handle); @@ -4539,14 +4539,12 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset, } ret = ext4_map_blocks(handle, inode, &map, EXT4_GET_BLOCKS_IO_CONVERT_EXT); - if (ret <= 0) { - WARN_ON(ret <= 0); - ext4_msg(inode->i_sb, KERN_ERR, - "%s:%d: inode #%lu: block %u: len %u: " - "ext4_ext_map_blocks returned %d", - __func__, __LINE__, inode->i_ino, map.m_lblk, - map.m_len, ret); - } + if (ret <= 0) + ext4_warning(inode->i_sb, + "inode #%lu: block %u: len %u: " + "ext4_ext_map_blocks returned %d", + inode->i_ino, map.m_lblk, + map.m_len, ret); ext4_mark_inode_dirty(handle, inode); ret2 = ext4_journal_stop(handle); if (ret <= 0 || ret2 ) -- cgit From d5ac77730516028f3ceda825abefac9a1153b138 Mon Sep 17 00:00:00 2001 From: Guo Chao Date: Mon, 28 Jan 2013 21:23:24 -0500 Subject: ext4: release buffer when checksum failed Commit b0336e8d (ext4: calculate and verify checksums of directory leaf blocks) and commit dbe89444 (ext4: Calculate and verify checksums for htree nodes) forget to release buffer when checksum failed, at some places. Signed-off-by: Guo Chao Signed-off-by: "Theodore Ts'o" Reviewed-by: Darrick J. Wong --- fs/ext4/dir.c | 1 + fs/ext4/namei.c | 12 ++++++++++-- 2 files changed, 11 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index 80a28b297279..3882fbc5e215 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -185,6 +185,7 @@ static int ext4_readdir(struct file *filp, "at offset %llu", (unsigned long long)filp->f_pos); filp->f_pos += sb->s_blocksize - offset; + brelse(bh); continue; } set_buffer_verified(bh); diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index f9ed946a448e..99813db159d9 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -837,6 +837,7 @@ static int ext4_htree_next_block(struct inode *dir, __u32 hash, !ext4_dx_csum_verify(dir, (struct ext4_dir_entry *)bh->b_data)) { ext4_warning(dir->i_sb, "Node failed checksum"); + brelse(bh); return -EIO; } set_buffer_verified(bh); @@ -877,8 +878,11 @@ static int htree_dirblock_to_tree(struct file *dir_file, } if (!buffer_verified(bh) && - !ext4_dirent_csum_verify(dir, (struct ext4_dir_entry *)bh->b_data)) + !ext4_dirent_csum_verify(dir, + (struct ext4_dir_entry *)bh->b_data)) { + brelse(bh); return -EIO; + } set_buffer_verified(bh); de = (struct ext4_dir_entry_2 *) bh->b_data; @@ -1929,8 +1933,10 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, } if (!buffer_verified(bh) && !ext4_dirent_csum_verify(dir, - (struct ext4_dir_entry *)bh->b_data)) + (struct ext4_dir_entry *)bh->b_data)) { + brelse(bh); return -EIO; + } set_buffer_verified(bh); retval = add_dirent_to_buf(handle, dentry, inode, NULL, bh); if (retval != -ENOSPC) { @@ -2492,6 +2498,7 @@ static int empty_dir(struct inode *inode) (struct ext4_dir_entry *)bh->b_data)) { EXT4_ERROR_INODE(inode, "checksum error reading directory " "lblock 0"); + brelse(bh); return -EIO; } set_buffer_verified(bh); @@ -2536,6 +2543,7 @@ static int empty_dir(struct inode *inode) (struct ext4_dir_entry *)bh->b_data)) { EXT4_ERROR_INODE(inode, "checksum error " "reading directory lblock 0"); + brelse(bh); return -EIO; } set_buffer_verified(bh); -- cgit From 2bbbee2a68a726deeac7da5ae7dd79b00301d6fd Mon Sep 17 00:00:00 2001 From: Guo Chao Date: Mon, 28 Jan 2013 21:26:44 -0500 Subject: ext4: remove unused variable in add_dirent_to_buf() After commit 978fef9 (create __ext4_insert_dentry for dir entry insertion), 'reclen' is not used anymore. Signed-off-by: Guo Chao Signed-off-by: "Theodore Ts'o" Reviewed-by: Darrick J. Wong --- fs/ext4/namei.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 99813db159d9..8bc01a3dcaca 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1703,7 +1703,6 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, const char *name = dentry->d_name.name; int namelen = dentry->d_name.len; unsigned int blocksize = dir->i_sb->s_blocksize; - unsigned short reclen; int csum_size = 0; int err; @@ -1711,7 +1710,6 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) csum_size = sizeof(struct ext4_dir_entry_tail); - reclen = EXT4_DIR_REC_LEN(namelen); if (!de) { err = ext4_find_dest_de(dir, inode, bh, bh->b_data, blocksize - csum_size, -- cgit From 41be871f747f64e076b09a68ae82a643e2ffb215 Mon Sep 17 00:00:00 2001 From: Guo Chao Date: Mon, 28 Jan 2013 21:33:28 -0500 Subject: ext4: remove useless assignment in dx_probe() Signed-off-by: Guo Chao Signed-off-by: "Theodore Ts'o" --- fs/ext4/namei.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 8bc01a3dcaca..f4b95870870e 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -714,7 +714,7 @@ dx_probe(const struct qstr *d_name, struct inode *dir, *err = ERR_BAD_DX_DIR; goto fail2; } - at = entries = ((struct dx_node *) bh->b_data)->entries; + entries = ((struct dx_node *) bh->b_data)->entries; if (!buffer_verified(bh) && !ext4_dx_csum_verify(dir, -- cgit From b1deefc99e668348f7c785c6ece5f6ff4c6d8b5c Mon Sep 17 00:00:00 2001 From: Guo Chao Date: Mon, 28 Jan 2013 21:41:02 -0500 Subject: ext4: remove unnecessary NULL pointer check brelse() and ext4_journal_force_commit() are both inlined and able to handle NULL. Signed-off-by: Guo Chao Reviewed-by: Darrick J. Wong Signed-off-by: "Theodore Ts'o" --- fs/ext4/namei.c | 3 +-- fs/ext4/super.c | 6 +----- 2 files changed, 2 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index f4b95870870e..34ed624d2c56 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2110,8 +2110,7 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, journal_error: ext4_std_error(dir->i_sb, err); cleanup: - if (bh) - brelse(bh); + brelse(bh); dx_release(frames); return err; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index dc0fb7b942cf..aac52738eb60 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4480,16 +4480,12 @@ static void ext4_clear_journal_err(struct super_block *sb, int ext4_force_commit(struct super_block *sb) { journal_t *journal; - int ret = 0; if (sb->s_flags & MS_RDONLY) return 0; journal = EXT4_SB(sb)->s_journal; - if (journal) - ret = ext4_journal_force_commit(journal); - - return ret; + return ext4_journal_force_commit(journal); } static int ext4_sync_fs(struct super_block *sb, int wait) -- cgit From 2a005855938235771b6960206d7581a4933cf6b7 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 14 Dec 2012 12:28:30 +0000 Subject: GFS2: Separate LRU scanning from shrinker This breaks out the LRU scanning function from the shrinker in preparation for adding other callers to the LRU scanner. Signed-off-by: Steven Whitehouse --- fs/gfs2/glock.c | 28 +++++++++++++++++----------- 1 file changed, 17 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 992c5c0cb504..3ad8fd36f8da 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -1376,23 +1376,19 @@ void gfs2_glock_complete(struct gfs2_glock *gl, int ret) gfs2_glock_put(gl); } +/** + * gfs2_scan_glock_lru - Scan the LRU looking for locks to demote + * @nr: The number of entries to scan + * + */ -static int gfs2_shrink_glock_memory(struct shrinker *shrink, - struct shrink_control *sc) +static void gfs2_scan_glock_lru(int nr) { struct gfs2_glock *gl; int may_demote; int nr_skipped = 0; - int nr = sc->nr_to_scan; - gfp_t gfp_mask = sc->gfp_mask; LIST_HEAD(skipped); - if (nr == 0) - goto out; - - if (!(gfp_mask & __GFP_FS)) - return -1; - spin_lock(&lru_lock); while(nr && !list_empty(&lru_list)) { gl = list_entry(lru_list.next, struct gfs2_glock, gl_lru); @@ -1425,7 +1421,17 @@ static int gfs2_shrink_glock_memory(struct shrinker *shrink, list_splice(&skipped, &lru_list); atomic_add(nr_skipped, &lru_count); spin_unlock(&lru_lock); -out: +} + +static int gfs2_shrink_glock_memory(struct shrinker *shrink, + struct shrink_control *sc) +{ + if (sc->nr_to_scan) { + if (!(sc->gfp_mask & __GFP_FS)) + return -1; + gfs2_scan_glock_lru(sc->nr_to_scan); + } + return (atomic_read(&lru_count) / 100) * sysctl_vfs_cache_pressure; } -- cgit From 75f2b879aeb3158daf6aa95d135803f7e71df06b Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 14 Dec 2012 12:29:56 +0000 Subject: GFS2: Merge revoke adding functions This moves the lo_add function for revokes into trans.c, removing a function call and making the code easier to read. Signed-off-by: Steven Whitehouse --- fs/gfs2/lops.c | 15 --------------- fs/gfs2/trans.c | 10 +++++++++- 2 files changed, 9 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c index 9ceccb1595a3..9c8074223461 100644 --- a/fs/gfs2/lops.c +++ b/fs/gfs2/lops.c @@ -600,20 +600,6 @@ static void buf_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass) jd->jd_jid, sdp->sd_replayed_blocks, sdp->sd_found_blocks); } -static void revoke_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) -{ - struct gfs2_glock *gl = bd->bd_gl; - struct gfs2_trans *tr; - - tr = current->journal_info; - tr->tr_touched = 1; - tr->tr_num_revoke++; - sdp->sd_log_num_revoke++; - atomic_inc(&gl->gl_revokes); - set_bit(GLF_LFLUSH, &gl->gl_flags); - list_add(&bd->bd_list, &sdp->sd_log_le_revoke); -} - static void revoke_lo_before_commit(struct gfs2_sbd *sdp) { struct gfs2_meta_header *mh; @@ -895,7 +881,6 @@ const struct gfs2_log_operations gfs2_buf_lops = { }; const struct gfs2_log_operations gfs2_revoke_lops = { - .lo_add = revoke_lo_add, .lo_before_commit = revoke_lo_before_commit, .lo_after_commit = revoke_lo_after_commit, .lo_before_scan = revoke_lo_before_scan, diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c index 413627072f36..6f3ddbcb5cc9 100644 --- a/fs/gfs2/trans.c +++ b/fs/gfs2/trans.c @@ -175,11 +175,19 @@ void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta) void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) { + struct gfs2_glock *gl = bd->bd_gl; + struct gfs2_trans *tr = current->journal_info; + BUG_ON(!list_empty(&bd->bd_list)); BUG_ON(!list_empty(&bd->bd_ail_st_list)); BUG_ON(!list_empty(&bd->bd_ail_gl_list)); lops_init_le(bd, &gfs2_revoke_lops); - lops_add(sdp, bd); + tr->tr_touched = 1; + tr->tr_num_revoke++; + sdp->sd_log_num_revoke++; + atomic_inc(&gl->gl_revokes); + set_bit(GLF_LFLUSH, &gl->gl_flags); + list_add(&bd->bd_list, &sdp->sd_log_le_revoke); } void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len) -- cgit From 350a9b0a7269c6f4b99eca34d0e773c3c2072520 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 14 Dec 2012 12:36:02 +0000 Subject: GFS2: Split gfs2_trans_add_bh() into two There is little common content in gfs2_trans_add_bh() between the data and meta classes by the time that the functions which it calls are taken into account. The intent here is to split this into two separate functions. Stage one is to introduce gfs2_trans_add_data() and gfs2_trans_add_meta() and update the callers accordingly. Later patches will then pull in the content of gfs2_trans_add_bh() and its dependent functions in order to clean up the code in this area. Signed-off-by: Steven Whitehouse --- fs/gfs2/aops.c | 4 ++-- fs/gfs2/bmap.c | 24 ++++++++++++------------ fs/gfs2/dir.c | 30 +++++++++++++++--------------- fs/gfs2/file.c | 4 ++-- fs/gfs2/inode.c | 8 ++++---- fs/gfs2/quota.c | 4 ++-- fs/gfs2/rgrp.c | 18 +++++++++--------- fs/gfs2/super.c | 8 ++++---- fs/gfs2/trans.c | 12 +++++++++++- fs/gfs2/trans.h | 3 ++- fs/gfs2/xattr.c | 36 ++++++++++++++++++------------------ 11 files changed, 81 insertions(+), 70 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 30de4f2a2ea9..92340dd23bba 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -51,7 +51,7 @@ static void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, continue; if (gfs2_is_jdata(ip)) set_buffer_uptodate(bh); - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_data(ip->i_gl, bh); } } @@ -852,7 +852,7 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, goto failed; } - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); if (gfs2_is_stuffed(ip)) return gfs2_stuffed_write_end(inode, dibh, pos, len, copied, page); diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index a68e91bcef3d..7a8627569a25 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -93,7 +93,7 @@ static int gfs2_unstuffer_page(struct gfs2_inode *ip, struct buffer_head *dibh, if (!gfs2_is_jdata(ip)) mark_buffer_dirty(bh); if (!gfs2_is_writeback(ip)) - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_data(ip->i_gl, bh); if (release) { unlock_page(page); @@ -153,7 +153,7 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) /* Set up the pointer to the new block */ - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); di = (struct gfs2_dinode *)dibh->b_data; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); @@ -405,7 +405,7 @@ static inline __be64 *gfs2_indirect_init(struct metapath *mp, BUG_ON(i < 1); BUG_ON(mp->mp_bh[i] != NULL); mp->mp_bh[i] = gfs2_meta_new(gl, bn); - gfs2_trans_add_bh(gl, mp->mp_bh[i], 1); + gfs2_trans_add_meta(gl, mp->mp_bh[i]); gfs2_metatype_set(mp->mp_bh[i], GFS2_METATYPE_IN, GFS2_FORMAT_IN); gfs2_buffer_clear_tail(mp->mp_bh[i], sizeof(struct gfs2_meta_header)); ptr += offset; @@ -468,7 +468,7 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, BUG_ON(sheight < 1); BUG_ON(dibh == NULL); - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); if (height == sheight) { struct buffer_head *bh; @@ -544,7 +544,7 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, /* Branching from existing tree */ case ALLOC_GROW_DEPTH: if (i > 1 && i < height) - gfs2_trans_add_bh(ip->i_gl, mp->mp_bh[i-1], 1); + gfs2_trans_add_meta(ip->i_gl, mp->mp_bh[i-1]); for (; i < height && n > 0; i++, n--) gfs2_indirect_init(mp, ip->i_gl, i, mp->mp_list[i-1], bn++); @@ -556,7 +556,7 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock, case ALLOC_DATA: BUG_ON(n > dblks); BUG_ON(mp->mp_bh[end_of_metadata] == NULL); - gfs2_trans_add_bh(ip->i_gl, mp->mp_bh[end_of_metadata], 1); + gfs2_trans_add_meta(ip->i_gl, mp->mp_bh[end_of_metadata]); dblks = n; ptr = metapointer(end_of_metadata, mp); dblock = bn; @@ -796,8 +796,8 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, down_write(&ip->i_rw_mutex); - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); + gfs2_trans_add_meta(ip->i_gl, bh); bstart = 0; blen = 0; @@ -981,7 +981,7 @@ static int gfs2_block_truncate_page(struct address_space *mapping, loff_t from) } if (!gfs2_is_writeback(ip)) - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_data(ip->i_gl, bh); zero_user(page, offset, length); mark_buffer_dirty(bh); @@ -1046,7 +1046,7 @@ static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) if (error) goto out; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); if (gfs2_is_stuffed(ip)) { gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode) + newsize); @@ -1141,7 +1141,7 @@ static int trunc_end(struct gfs2_inode *ip) ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; ip->i_diskflags &= ~GFS2_DIF_TRUNC_IN_PROG; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); @@ -1246,7 +1246,7 @@ static int do_grow(struct inode *inode, u64 size) i_size_write(inode, size); ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 9a35670fdc38..7179478e5a28 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -93,7 +93,7 @@ int gfs2_dir_get_new_buffer(struct gfs2_inode *ip, u64 block, struct buffer_head *bh; bh = gfs2_meta_new(ip->i_gl, block); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_metatype_set(bh, GFS2_METATYPE_JD, GFS2_FORMAT_JD); gfs2_buffer_clear_tail(bh, sizeof(struct gfs2_meta_header)); *bhp = bh; @@ -127,7 +127,7 @@ static int gfs2_dir_write_stuffed(struct gfs2_inode *ip, const char *buf, if (error) return error; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); memcpy(dibh->b_data + offset + sizeof(struct gfs2_dinode), buf, size); if (ip->i_inode.i_size < offset + size) i_size_write(&ip->i_inode, offset + size); @@ -209,7 +209,7 @@ static int gfs2_dir_write_data(struct gfs2_inode *ip, const char *buf, if (error) goto fail; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); memcpy(bh->b_data + o, buf, amount); brelse(bh); @@ -231,7 +231,7 @@ out: i_size_write(&ip->i_inode, offset + copied); ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); @@ -647,7 +647,7 @@ static void dirent_del(struct gfs2_inode *dip, struct buffer_head *bh, return; } - gfs2_trans_add_bh(dip->i_gl, bh, 1); + gfs2_trans_add_meta(dip->i_gl, bh); /* If there is no prev entry, this is the first entry in the block. The de_rec_len is already as big as it needs to be. Just zero @@ -690,7 +690,7 @@ static struct gfs2_dirent *gfs2_init_dirent(struct inode *inode, offset = GFS2_DIRENT_SIZE(be16_to_cpu(dent->de_name_len)); totlen = be16_to_cpu(dent->de_rec_len); BUG_ON(offset + name->len > totlen); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); ndent = (struct gfs2_dirent *)((char *)dent + offset); dent->de_rec_len = cpu_to_be16(offset); gfs2_qstr2dirent(name, totlen - offset, ndent); @@ -831,7 +831,7 @@ static struct gfs2_leaf *new_leaf(struct inode *inode, struct buffer_head **pbh, return NULL; gfs2_trans_add_unrevoke(GFS2_SB(inode), bn, 1); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_metatype_set(bh, GFS2_METATYPE_LF, GFS2_FORMAT_LF); leaf = (struct gfs2_leaf *)bh->b_data; leaf->lf_depth = cpu_to_be16(depth); @@ -916,7 +916,7 @@ static int dir_make_exhash(struct inode *inode) /* We're done with the new leaf block, now setup the new hash table. */ - gfs2_trans_add_bh(dip->i_gl, dibh, 1); + gfs2_trans_add_meta(dip->i_gl, dibh); gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); lp = (__be64 *)(dibh->b_data + sizeof(struct gfs2_dinode)); @@ -976,7 +976,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) return 1; /* can't split */ } - gfs2_trans_add_bh(dip->i_gl, obh, 1); + gfs2_trans_add_meta(dip->i_gl, obh); nleaf = new_leaf(inode, &nbh, be16_to_cpu(oleaf->lf_depth) + 1); if (!nleaf) { @@ -1069,7 +1069,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) error = gfs2_meta_inode_buffer(dip, &dibh); if (!gfs2_assert_withdraw(GFS2_SB(&dip->i_inode), !error)) { - gfs2_trans_add_bh(dip->i_gl, dibh, 1); + gfs2_trans_add_meta(dip->i_gl, dibh); gfs2_add_inode_blocks(&dip->i_inode, 1); gfs2_dinode_out(dip, dibh->b_data); brelse(dibh); @@ -1622,7 +1622,7 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) return error; } while(1); - gfs2_trans_add_bh(ip->i_gl, obh, 1); + gfs2_trans_add_meta(ip->i_gl, obh); leaf = new_leaf(inode, &bh, be16_to_cpu(oleaf->lf_depth)); if (!leaf) { @@ -1636,7 +1636,7 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) error = gfs2_meta_inode_buffer(ip, &bh); if (error) return error; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_add_inode_blocks(&ip->i_inode, 1); gfs2_dinode_out(ip, bh->b_data); brelse(bh); @@ -1795,7 +1795,7 @@ int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, if (IS_ERR(dent)) return PTR_ERR(dent); - gfs2_trans_add_bh(dip->i_gl, bh, 1); + gfs2_trans_add_meta(dip->i_gl, bh); gfs2_inum_out(nip, dent); dent->de_type = cpu_to_be16(new_type); @@ -1804,7 +1804,7 @@ int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, error = gfs2_meta_inode_buffer(dip, &bh); if (error) return error; - gfs2_trans_add_bh(dip->i_gl, bh, 1); + gfs2_trans_add_meta(dip->i_gl, bh); } dip->i_inode.i_mtime = dip->i_inode.i_ctime = CURRENT_TIME; @@ -1917,7 +1917,7 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, if (error) goto out_end_trans; - gfs2_trans_add_bh(dip->i_gl, dibh, 1); + gfs2_trans_add_meta(dip->i_gl, dibh); /* On the last dealloc, make this a regular file in case we crash. (We don't want to free these blocks a second time.) */ if (last_dealloc) diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 991ab2d484dd..06b7092a3f25 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -276,7 +276,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) error = gfs2_meta_inode_buffer(ip, &bh); if (error) goto out_trans_end; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); ip->i_diskflags = new_flags; gfs2_dinode_out(ip, bh->b_data); brelse(bh); @@ -709,7 +709,7 @@ static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len, if (unlikely(error)) return error; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); if (gfs2_is_stuffed(ip)) { error = gfs2_unstuff_dinode(ip, NULL); diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 2b6f5698ef18..db048a8ab6a8 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -447,7 +447,7 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_inode *ip, struct timespec tv = CURRENT_TIME; dibh = gfs2_meta_new(ip->i_gl, ip->i_no_addr); - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_metatype_set(dibh, GFS2_METATYPE_DI, GFS2_FORMAT_DI); gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); di = (struct gfs2_dinode *)dibh->b_data; @@ -584,7 +584,7 @@ static int link_dinode(struct gfs2_inode *dip, const struct qstr *name, if (error) goto fail_end_trans; set_nlink(&ip->i_inode, S_ISDIR(ip->i_inode.i_mode) ? 2 : 1); - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); return 0; @@ -931,7 +931,7 @@ static int gfs2_link(struct dentry *old_dentry, struct inode *dir, if (error) goto out_brelse; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); inc_nlink(&ip->i_inode); ip->i_inode.i_ctime = CURRENT_TIME; ihold(inode); @@ -1412,7 +1412,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry, if (error) goto out_end_trans; ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index ae55e248c3b7..06122d09c0d1 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -590,7 +590,7 @@ static void do_qc(struct gfs2_quota_data *qd, s64 change) s64 x; mutex_lock(&sdp->sd_quota_mutex); - gfs2_trans_add_bh(ip->i_gl, qd->qd_bh, 1); + gfs2_trans_add_meta(ip->i_gl, qd->qd_bh); if (!test_bit(QDF_CHANGE, &qd->qd_flags)) { qc->qc_change = 0; @@ -726,7 +726,7 @@ get_a_page: goto unlock_out; } - gfs2_trans_add_bh(ip->i_gl, bh, 0); + gfs2_trans_add_meta(ip->i_gl, bh); kaddr = kmap_atomic(page); if (offset + sizeof(struct gfs2_quota) > PAGE_CACHE_SIZE) diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index b7eff078fe90..52c2aeaf45ce 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -1323,7 +1323,7 @@ int gfs2_fitrim(struct file *filp, void __user *argp) if (ret == 0) { bh = rgd->rd_bits[0].bi_bh; rgd->rd_flags |= GFS2_RGF_TRIMMED; - gfs2_trans_add_bh(rgd->rd_gl, bh, 1); + gfs2_trans_add_meta(rgd->rd_gl, bh); gfs2_rgrp_out(rgd, bh->b_data); gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, bh->b_data); gfs2_trans_end(sdp); @@ -1968,14 +1968,14 @@ static void gfs2_alloc_extent(const struct gfs2_rbm *rbm, bool dinode, *n = 1; block = gfs2_rbm_to_block(rbm); - gfs2_trans_add_bh(rbm->rgd->rd_gl, rbm->bi->bi_bh, 1); + gfs2_trans_add_meta(rbm->rgd->rd_gl, rbm->bi->bi_bh); gfs2_setbit(rbm, true, dinode ? GFS2_BLKST_DINODE : GFS2_BLKST_USED); block++; while (*n < elen) { ret = gfs2_rbm_from_block(&pos, block); if (ret || gfs2_testbit(&pos) != GFS2_BLKST_FREE) break; - gfs2_trans_add_bh(pos.rgd->rd_gl, pos.bi->bi_bh, 1); + gfs2_trans_add_meta(pos.rgd->rd_gl, pos.bi->bi_bh); gfs2_setbit(&pos, true, GFS2_BLKST_USED); (*n)++; block++; @@ -2014,7 +2014,7 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, rbm.bi->bi_bh->b_data + rbm.bi->bi_offset, rbm.bi->bi_len); } - gfs2_trans_add_bh(rbm.rgd->rd_gl, rbm.bi->bi_bh, 1); + gfs2_trans_add_meta(rbm.rgd->rd_gl, rbm.bi->bi_bh); gfs2_setbit(&rbm, false, new_state); } @@ -2157,7 +2157,7 @@ int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *nblocks, if (error == 0) { struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); di->di_goal_meta = di->di_goal_data = cpu_to_be64(ip->i_goal); brelse(dibh); @@ -2176,7 +2176,7 @@ int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *nblocks, *generation = rbm.rgd->rd_igeneration++; } - gfs2_trans_add_bh(rbm.rgd->rd_gl, rbm.rgd->rd_bits[0].bi_bh, 1); + gfs2_trans_add_meta(rbm.rgd->rd_gl, rbm.rgd->rd_bits[0].bi_bh); gfs2_rgrp_out(rbm.rgd, rbm.rgd->rd_bits[0].bi_bh->b_data); gfs2_rgrp_ondisk2lvb(rbm.rgd->rd_rgl, rbm.rgd->rd_bits[0].bi_bh->b_data); @@ -2223,7 +2223,7 @@ void __gfs2_free_blocks(struct gfs2_inode *ip, u64 bstart, u32 blen, int meta) trace_gfs2_block_alloc(ip, rgd, bstart, blen, GFS2_BLKST_FREE); rgd->rd_free += blen; rgd->rd_flags &= ~GFS2_RGF_TRIMMED; - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); + gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data); @@ -2260,7 +2260,7 @@ void gfs2_unlink_di(struct inode *inode) if (!rgd) return; trace_gfs2_block_alloc(ip, rgd, blkno, 1, GFS2_BLKST_UNLINKED); - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); + gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data); update_rgrp_lvb_unlinked(rgd, 1); @@ -2281,7 +2281,7 @@ static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno) rgd->rd_dinodes--; rgd->rd_free++; - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); + gfs2_trans_add_meta(rgd->rd_gl, rgd->rd_bits[0].bi_bh); gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_rgrp_ondisk2lvb(rgd->rd_rgl, rgd->rd_bits[0].bi_bh->b_data); update_rgrp_lvb_unlinked(rgd, -1); diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index d6488674d916..4dfda4c94634 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -500,7 +500,7 @@ void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free, if (error) return; - gfs2_trans_add_bh(l_ip->i_gl, l_bh, 1); + gfs2_trans_add_meta(l_ip->i_gl, l_bh); spin_lock(&sdp->sd_statfs_spin); l_sc->sc_total += total; @@ -528,7 +528,7 @@ void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master; struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local; - gfs2_trans_add_bh(l_ip->i_gl, l_bh, 1); + gfs2_trans_add_meta(l_ip->i_gl, l_bh); spin_lock(&sdp->sd_statfs_spin); m_sc->sc_total += l_sc->sc_total; @@ -539,7 +539,7 @@ void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, 0, sizeof(struct gfs2_statfs_change)); spin_unlock(&sdp->sd_statfs_spin); - gfs2_trans_add_bh(m_ip->i_gl, m_bh, 1); + gfs2_trans_add_meta(m_ip->i_gl, m_bh); gfs2_statfs_change_out(m_sc, m_bh->b_data + sizeof(struct gfs2_dinode)); } @@ -824,7 +824,7 @@ static void gfs2_dirty_inode(struct inode *inode, int flags) ret = gfs2_meta_inode_buffer(ip, &bh); if (ret == 0) { - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_dinode_out(ip, bh->b_data); brelse(bh); } diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c index 6f3ddbcb5cc9..cb795ac754ea 100644 --- a/fs/gfs2/trans.c +++ b/fs/gfs2/trans.c @@ -150,7 +150,7 @@ void gfs2_trans_end(struct gfs2_sbd *sdp) * */ -void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta) +static void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta) { struct gfs2_sbd *sdp = gl->gl_sbd; struct gfs2_bufdata *bd; @@ -173,6 +173,16 @@ void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta) unlock_buffer(bh); } +void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) +{ + return gfs2_trans_add_bh(gl, bh, 0); +} + +void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh) +{ + return gfs2_trans_add_bh(gl, bh, 1); +} + void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) { struct gfs2_glock *gl = bd->bd_gl; diff --git a/fs/gfs2/trans.h b/fs/gfs2/trans.h index bf2ae9aeee7a..1e6e7da25a17 100644 --- a/fs/gfs2/trans.h +++ b/fs/gfs2/trans.h @@ -39,7 +39,8 @@ extern int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks, unsigned int revokes); extern void gfs2_trans_end(struct gfs2_sbd *sdp); -extern void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta); +extern void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh); +extern void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh); extern void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd); extern void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len); diff --git a/fs/gfs2/xattr.c b/fs/gfs2/xattr.c index 76c144b3c9bb..cbb46c2baa69 100644 --- a/fs/gfs2/xattr.c +++ b/fs/gfs2/xattr.c @@ -270,7 +270,7 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, if (error) goto out_gunlock; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); dataptrs = GFS2_EA2DATAPTRS(ea); for (x = 0; x < ea->ea_num_ptrs; x++, dataptrs++) { @@ -309,7 +309,7 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -509,7 +509,7 @@ static int gfs2_iter_unstuffed(struct gfs2_inode *ip, struct gfs2_ea_header *ea, } if (din) { - gfs2_trans_add_bh(ip->i_gl, bh[x], 1); + gfs2_trans_add_meta(ip->i_gl, bh[x]); memcpy(pos, din, cp_size); din += sdp->sd_jbsize; } @@ -629,7 +629,7 @@ static int ea_alloc_blk(struct gfs2_inode *ip, struct buffer_head **bhp) return error; gfs2_trans_add_unrevoke(sdp, block, 1); *bhp = gfs2_meta_new(ip->i_gl, block); - gfs2_trans_add_bh(ip->i_gl, *bhp, 1); + gfs2_trans_add_meta(ip->i_gl, *bhp); gfs2_metatype_set(*bhp, GFS2_METATYPE_EA, GFS2_FORMAT_EA); gfs2_buffer_clear_tail(*bhp, sizeof(struct gfs2_meta_header)); @@ -691,7 +691,7 @@ static int ea_write(struct gfs2_inode *ip, struct gfs2_ea_header *ea, return error; gfs2_trans_add_unrevoke(sdp, block, 1); bh = gfs2_meta_new(ip->i_gl, block); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); gfs2_metatype_set(bh, GFS2_METATYPE_ED, GFS2_FORMAT_ED); gfs2_add_inode_blocks(&ip->i_inode, 1); @@ -751,7 +751,7 @@ static int ea_alloc_skeleton(struct gfs2_inode *ip, struct gfs2_ea_request *er, error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -834,7 +834,7 @@ static void ea_set_remove_stuffed(struct gfs2_inode *ip, struct gfs2_ea_header *prev = el->el_prev; u32 len; - gfs2_trans_add_bh(ip->i_gl, el->el_bh, 1); + gfs2_trans_add_meta(ip->i_gl, el->el_bh); if (!prev || !GFS2_EA_IS_STUFFED(ea)) { ea->ea_type = GFS2_EATYPE_UNUSED; @@ -872,7 +872,7 @@ static int ea_set_simple_noalloc(struct gfs2_inode *ip, struct buffer_head *bh, if (error) return error; - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_meta(ip->i_gl, bh); if (es->ea_split) ea = ea_split_ea(ea); @@ -886,7 +886,7 @@ static int ea_set_simple_noalloc(struct gfs2_inode *ip, struct buffer_head *bh, if (error) goto out; ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); out: @@ -901,7 +901,7 @@ static int ea_set_simple_alloc(struct gfs2_inode *ip, struct gfs2_ea_header *ea = es->es_ea; int error; - gfs2_trans_add_bh(ip->i_gl, es->es_bh, 1); + gfs2_trans_add_meta(ip->i_gl, es->es_bh); if (es->ea_split) ea = ea_split_ea(ea); @@ -997,7 +997,7 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er, goto out; } - gfs2_trans_add_bh(ip->i_gl, indbh, 1); + gfs2_trans_add_meta(ip->i_gl, indbh); } else { u64 blk; unsigned int n = 1; @@ -1006,7 +1006,7 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er, return error; gfs2_trans_add_unrevoke(sdp, blk, 1); indbh = gfs2_meta_new(ip->i_gl, blk); - gfs2_trans_add_bh(ip->i_gl, indbh, 1); + gfs2_trans_add_meta(ip->i_gl, indbh); gfs2_metatype_set(indbh, GFS2_METATYPE_IN, GFS2_FORMAT_IN); gfs2_buffer_clear_tail(indbh, mh_size); @@ -1092,7 +1092,7 @@ static int ea_remove_stuffed(struct gfs2_inode *ip, struct gfs2_ea_location *el) if (error) return error; - gfs2_trans_add_bh(ip->i_gl, el->el_bh, 1); + gfs2_trans_add_meta(ip->i_gl, el->el_bh); if (prev) { u32 len; @@ -1109,7 +1109,7 @@ static int ea_remove_stuffed(struct gfs2_inode *ip, struct gfs2_ea_location *el) error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -1265,7 +1265,7 @@ int gfs2_xattr_acl_chmod(struct gfs2_inode *ip, struct iattr *attr, char *data) if (GFS2_EA_IS_STUFFED(el.el_ea)) { error = gfs2_trans_begin(sdp, RES_DINODE + RES_EATTR, 0); if (error == 0) { - gfs2_trans_add_bh(ip->i_gl, el.el_bh, 1); + gfs2_trans_add_meta(ip->i_gl, el.el_bh); memcpy(GFS2_EA2DATA(el.el_ea), data, GFS2_EA_DATA_LEN(el.el_ea)); } @@ -1352,7 +1352,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) if (error) goto out_gunlock; - gfs2_trans_add_bh(ip->i_gl, indbh, 1); + gfs2_trans_add_meta(ip->i_gl, indbh); eablk = (__be64 *)(indbh->b_data + sizeof(struct gfs2_meta_header)); bstart = 0; @@ -1384,7 +1384,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } @@ -1434,7 +1434,7 @@ static int ea_dealloc_block(struct gfs2_inode *ip) error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { - gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_trans_add_meta(ip->i_gl, dibh); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); } -- cgit From 767f433f346959d6a09b85478eb5db133ab25c6f Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 14 Dec 2012 12:52:14 +0000 Subject: GFS2: Copy gfs2_trans_add_bh into new data/meta functions This patch copies the body of gfs2_trans_add_bh into the two newly added gfs2_trans_add_data and gfs2_trans_add_meta functions. We can then move the .lo_add functions from lops.c into trans.c and call them directly. As a result of this, we no longer need to use the .lo_add functions at all, so that is removed from the log operations structure. Signed-off-by: Steven Whitehouse --- fs/gfs2/incore.h | 1 - fs/gfs2/lops.c | 68 +---------------------------------------- fs/gfs2/lops.h | 7 +---- fs/gfs2/trans.c | 92 ++++++++++++++++++++++++++++++++++++++++++++++++++------ 4 files changed, 84 insertions(+), 84 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index c373a24fedd9..5d129ab89733 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -52,7 +52,6 @@ struct gfs2_log_header_host { */ struct gfs2_log_operations { - void (*lo_add) (struct gfs2_sbd *sdp, struct gfs2_bufdata *bd); void (*lo_before_commit) (struct gfs2_sbd *sdp); void (*lo_after_commit) (struct gfs2_sbd *sdp, struct gfs2_ail *ai); void (*lo_before_scan) (struct gfs2_jdesc *jd, diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c index 9c8074223461..a5055977a214 100644 --- a/fs/gfs2/lops.c +++ b/fs/gfs2/lops.c @@ -37,7 +37,7 @@ * * The log lock must be held when calling this function */ -static void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) +void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) { struct gfs2_bufdata *bd; @@ -388,32 +388,6 @@ static struct page *gfs2_get_log_desc(struct gfs2_sbd *sdp, u32 ld_type, return page; } -static void buf_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) -{ - struct gfs2_meta_header *mh; - struct gfs2_trans *tr; - - tr = current->journal_info; - tr->tr_touched = 1; - if (!list_empty(&bd->bd_list)) - return; - set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); - set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); - mh = (struct gfs2_meta_header *)bd->bd_bh->b_data; - if (unlikely(mh->mh_magic != cpu_to_be32(GFS2_MAGIC))) { - printk(KERN_ERR - "Attempting to add uninitialised block to journal (inplace block=%lld)\n", - (unsigned long long)bd->bd_bh->b_blocknr); - BUG(); - } - gfs2_pin(sdp, bd->bd_bh); - mh->__pad0 = cpu_to_be64(0); - mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid); - sdp->sd_log_num_buf++; - list_add(&bd->bd_list, &sdp->sd_log_le_buf); - tr->tr_num_buf_new++; -} - static void gfs2_check_magic(struct buffer_head *bh) { void *kaddr; @@ -734,44 +708,6 @@ static void revoke_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass) gfs2_revoke_clean(sdp); } -/** - * databuf_lo_add - Add a databuf to the transaction. - * - * This is used in two distinct cases: - * i) In ordered write mode - * We put the data buffer on a list so that we can ensure that its - * synced to disk at the right time - * ii) In journaled data mode - * We need to journal the data block in the same way as metadata in - * the functions above. The difference is that here we have a tag - * which is two __be64's being the block number (as per meta data) - * and a flag which says whether the data block needs escaping or - * not. This means we need a new log entry for each 251 or so data - * blocks, which isn't an enormous overhead but twice as much as - * for normal metadata blocks. - */ -static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) -{ - struct gfs2_trans *tr = current->journal_info; - struct address_space *mapping = bd->bd_bh->b_page->mapping; - struct gfs2_inode *ip = GFS2_I(mapping->host); - - if (tr) - tr->tr_touched = 1; - if (!list_empty(&bd->bd_list)) - return; - set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); - set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); - if (gfs2_is_jdata(ip)) { - gfs2_pin(sdp, bd->bd_bh); - tr->tr_num_databuf_new++; - sdp->sd_log_num_databuf++; - list_add_tail(&bd->bd_list, &sdp->sd_log_le_databuf); - } else { - list_add_tail(&bd->bd_list, &sdp->sd_log_le_ordered); - } -} - /** * databuf_lo_before_commit - Scan the data buffers, writing as we go * @@ -871,7 +807,6 @@ static void databuf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai) const struct gfs2_log_operations gfs2_buf_lops = { - .lo_add = buf_lo_add, .lo_before_commit = buf_lo_before_commit, .lo_after_commit = buf_lo_after_commit, .lo_before_scan = buf_lo_before_scan, @@ -894,7 +829,6 @@ const struct gfs2_log_operations gfs2_rg_lops = { }; const struct gfs2_log_operations gfs2_databuf_lops = { - .lo_add = databuf_lo_add, .lo_before_commit = databuf_lo_before_commit, .lo_after_commit = databuf_lo_after_commit, .lo_scan_elements = databuf_lo_scan_elements, diff --git a/fs/gfs2/lops.h b/fs/gfs2/lops.h index 954a330585f4..d85b3767ee08 100644 --- a/fs/gfs2/lops.h +++ b/fs/gfs2/lops.h @@ -29,6 +29,7 @@ extern const struct gfs2_log_operations gfs2_databuf_lops; extern const struct gfs2_log_operations *gfs2_log_ops[]; extern void gfs2_log_write_page(struct gfs2_sbd *sdp, struct page *page); extern void gfs2_log_flush_bio(struct gfs2_sbd *sdp, int rw); +extern void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh); static inline unsigned int buf_limit(struct gfs2_sbd *sdp) { @@ -53,12 +54,6 @@ static inline void lops_init_le(struct gfs2_bufdata *bd, bd->bd_ops = lops; } -static inline void lops_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) -{ - if (bd->bd_ops->lo_add) - bd->bd_ops->lo_add(sdp, bd); -} - static inline void lops_before_commit(struct gfs2_sbd *sdp) { int x; diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c index cb795ac754ea..1fbd57eafa48 100644 --- a/fs/gfs2/trans.c +++ b/fs/gfs2/trans.c @@ -18,6 +18,7 @@ #include "gfs2.h" #include "incore.h" #include "glock.h" +#include "inode.h" #include "log.h" #include "lops.h" #include "meta_io.h" @@ -143,15 +144,46 @@ void gfs2_trans_end(struct gfs2_sbd *sdp) } /** - * gfs2_trans_add_bh - Add a to-be-modified buffer to the current transaction - * @gl: the glock the buffer belongs to - * @bh: The buffer to add - * @meta: True in the case of adding metadata + * databuf_lo_add - Add a databuf to the transaction. * + * This is used in two distinct cases: + * i) In ordered write mode + * We put the data buffer on a list so that we can ensure that its + * synced to disk at the right time + * ii) In journaled data mode + * We need to journal the data block in the same way as metadata in + * the functions above. The difference is that here we have a tag + * which is two __be64's being the block number (as per meta data) + * and a flag which says whether the data block needs escaping or + * not. This means we need a new log entry for each 251 or so data + * blocks, which isn't an enormous overhead but twice as much as + * for normal metadata blocks. */ +static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) +{ + struct gfs2_trans *tr = current->journal_info; + struct address_space *mapping = bd->bd_bh->b_page->mapping; + struct gfs2_inode *ip = GFS2_I(mapping->host); + + if (tr) + tr->tr_touched = 1; + if (!list_empty(&bd->bd_list)) + return; + set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); + set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); + if (gfs2_is_jdata(ip)) { + gfs2_pin(sdp, bd->bd_bh); + tr->tr_num_databuf_new++; + sdp->sd_log_num_databuf++; + list_add_tail(&bd->bd_list, &sdp->sd_log_le_databuf); + } else { + list_add_tail(&bd->bd_list, &sdp->sd_log_le_ordered); + } +} -static void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta) +void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) { + struct gfs2_sbd *sdp = gl->gl_sbd; struct gfs2_bufdata *bd; @@ -163,24 +195,64 @@ static void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int else { gfs2_log_unlock(sdp); unlock_buffer(bh); - gfs2_attach_bufdata(gl, bh, meta); + gfs2_attach_bufdata(gl, bh, 0); bd = bh->b_private; lock_buffer(bh); gfs2_log_lock(sdp); } - lops_add(sdp, bd); + databuf_lo_add(sdp, bd); gfs2_log_unlock(sdp); unlock_buffer(bh); } -void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) +static void meta_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) { - return gfs2_trans_add_bh(gl, bh, 0); + struct gfs2_meta_header *mh; + struct gfs2_trans *tr; + + tr = current->journal_info; + tr->tr_touched = 1; + if (!list_empty(&bd->bd_list)) + return; + set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); + set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); + mh = (struct gfs2_meta_header *)bd->bd_bh->b_data; + if (unlikely(mh->mh_magic != cpu_to_be32(GFS2_MAGIC))) { + printk(KERN_ERR + "Attempting to add uninitialised block to journal (inplace block=%lld)\n", + (unsigned long long)bd->bd_bh->b_blocknr); + BUG(); + } + gfs2_pin(sdp, bd->bd_bh); + mh->__pad0 = cpu_to_be64(0); + mh->mh_jid = cpu_to_be32(sdp->sd_jdesc->jd_jid); + sdp->sd_log_num_buf++; + list_add(&bd->bd_list, &sdp->sd_log_le_buf); + tr->tr_num_buf_new++; } void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh) { - return gfs2_trans_add_bh(gl, bh, 1); + + struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_bufdata *bd; + + lock_buffer(bh); + gfs2_log_lock(sdp); + bd = bh->b_private; + if (bd) + gfs2_assert(sdp, bd->bd_gl == gl); + else { + gfs2_log_unlock(sdp); + unlock_buffer(bh); + gfs2_attach_bufdata(gl, bh, 1); + bd = bh->b_private; + lock_buffer(bh); + gfs2_log_lock(sdp); + } + meta_lo_add(sdp, bd); + gfs2_log_unlock(sdp); + unlock_buffer(bh); } void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) -- cgit From c76c4d96bdd89027306cebc80eb3397286d8da66 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 14 Dec 2012 17:54:21 +0000 Subject: GFS2: Merge gfs2_attach_bufdata() into trans.c The locking in gfs2_attach_bufdata() was type specific (data/meta) which made the function rather confusing. This patch moves the core of gfs2_attach_bufdata() into trans.c renaming it gfs2_alloc_bufdata() and moving the locking into gfs2_trans_add_data()/gfs2_trans_add_meta() As a result all of the locking related to adding data and metadata to the journal is now in these two functions. This should help to clarify what is going on, and give us some opportunities to simplify in some cases. Signed-off-by: Steven Whitehouse --- fs/gfs2/lops.h | 7 ------- fs/gfs2/meta_io.c | 35 ----------------------------------- fs/gfs2/meta_io.h | 3 --- fs/gfs2/trans.c | 37 ++++++++++++++++++++++++++----------- 4 files changed, 26 insertions(+), 56 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/lops.h b/fs/gfs2/lops.h index d85b3767ee08..ba77b7da8325 100644 --- a/fs/gfs2/lops.h +++ b/fs/gfs2/lops.h @@ -47,13 +47,6 @@ static inline unsigned int databuf_limit(struct gfs2_sbd *sdp) return limit; } -static inline void lops_init_le(struct gfs2_bufdata *bd, - const struct gfs2_log_operations *lops) -{ - INIT_LIST_HEAD(&bd->bd_list); - bd->bd_ops = lops; -} - static inline void lops_before_commit(struct gfs2_sbd *sdp) { int x; diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c index 22255d96b27e..b059bbb5059e 100644 --- a/fs/gfs2/meta_io.c +++ b/fs/gfs2/meta_io.c @@ -271,41 +271,6 @@ int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh) return 0; } -/** - * gfs2_attach_bufdata - attach a struct gfs2_bufdata structure to a buffer - * @gl: the glock the buffer belongs to - * @bh: The buffer to be attached to - * @meta: Flag to indicate whether its metadata or not - */ - -void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh, - int meta) -{ - struct gfs2_bufdata *bd; - - if (meta) - lock_page(bh->b_page); - - if (bh->b_private) { - if (meta) - unlock_page(bh->b_page); - return; - } - - bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS | __GFP_NOFAIL); - bd->bd_bh = bh; - bd->bd_gl = gl; - - if (meta) - lops_init_le(bd, &gfs2_buf_lops); - else - lops_init_le(bd, &gfs2_databuf_lops); - bh->b_private = bd; - - if (meta) - unlock_page(bh->b_page); -} - void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int meta) { struct address_space *mapping = bh->b_page->mapping; diff --git a/fs/gfs2/meta_io.h b/fs/gfs2/meta_io.h index c30973b07a7c..0d4c843b6f8e 100644 --- a/fs/gfs2/meta_io.h +++ b/fs/gfs2/meta_io.h @@ -56,9 +56,6 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int gfs2_meta_wait(struct gfs2_sbd *sdp, struct buffer_head *bh); struct buffer_head *gfs2_getbuf(struct gfs2_glock *gl, u64 blkno, int create); -void gfs2_attach_bufdata(struct gfs2_glock *gl, struct buffer_head *bh, - int meta); - void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int meta); diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c index 1fbd57eafa48..14dbf6d3cdc0 100644 --- a/fs/gfs2/trans.c +++ b/fs/gfs2/trans.c @@ -143,6 +143,21 @@ void gfs2_trans_end(struct gfs2_sbd *sdp) sb_end_intwrite(sdp->sd_vfs); } +static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl, + struct buffer_head *bh, + const struct gfs2_log_operations *lops) +{ + struct gfs2_bufdata *bd; + + bd = kmem_cache_zalloc(gfs2_bufdata_cachep, GFP_NOFS | __GFP_NOFAIL); + bd->bd_bh = bh; + bd->bd_gl = gl; + bd->bd_ops = lops; + INIT_LIST_HEAD(&bd->bd_list); + bh->b_private = bd; + return bd; +} + /** * databuf_lo_add - Add a databuf to the transaction. * @@ -190,16 +205,15 @@ void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) lock_buffer(bh); gfs2_log_lock(sdp); bd = bh->b_private; - if (bd) - gfs2_assert(sdp, bd->bd_gl == gl); - else { + if (bd == NULL) { gfs2_log_unlock(sdp); unlock_buffer(bh); - gfs2_attach_bufdata(gl, bh, 0); - bd = bh->b_private; + if (bh->b_private == NULL) + bd = gfs2_alloc_bufdata(gl, bh, &gfs2_databuf_lops); lock_buffer(bh); gfs2_log_lock(sdp); } + gfs2_assert(sdp, bd->bd_gl == gl); databuf_lo_add(sdp, bd); gfs2_log_unlock(sdp); unlock_buffer(bh); @@ -240,16 +254,17 @@ void gfs2_trans_add_meta(struct gfs2_glock *gl, struct buffer_head *bh) lock_buffer(bh); gfs2_log_lock(sdp); bd = bh->b_private; - if (bd) - gfs2_assert(sdp, bd->bd_gl == gl); - else { + if (bd == NULL) { gfs2_log_unlock(sdp); unlock_buffer(bh); - gfs2_attach_bufdata(gl, bh, 1); - bd = bh->b_private; + lock_page(bh->b_page); + if (bh->b_private == NULL) + bd = gfs2_alloc_bufdata(gl, bh, &gfs2_buf_lops); + unlock_page(bh->b_page); lock_buffer(bh); gfs2_log_lock(sdp); } + gfs2_assert(sdp, bd->bd_gl == gl); meta_lo_add(sdp, bd); gfs2_log_unlock(sdp); unlock_buffer(bh); @@ -263,7 +278,7 @@ void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) BUG_ON(!list_empty(&bd->bd_list)); BUG_ON(!list_empty(&bd->bd_ail_st_list)); BUG_ON(!list_empty(&bd->bd_ail_gl_list)); - lops_init_le(bd, &gfs2_revoke_lops); + bd->bd_ops = &gfs2_revoke_lops; tr->tr_touched = 1; tr->tr_num_revoke++; sdp->sd_log_num_revoke++; -- cgit From d564053f074634e7a966359dc97d26900fa5f52d Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 11 Jan 2013 10:49:34 +0000 Subject: GFS2: Clean up freeze code The freeze code has not been looked at a lot recently. Upstream has moved on, and this is an attempt to catch us back up again. There is a vfs level interface for the freeze code which can be called from our (obsolete, but kept for backward compatibility purposes) sysfs freeze interface. This means freezing this way vs. doing it from the ioctl should now work in identical fashion. As a result of this, the freeze function is only called once and we can drop our own special purpose code for counting the number of freezes. Signed-off-by: Steven Whitehouse --- fs/gfs2/incore.h | 3 --- fs/gfs2/ops_fstype.c | 2 -- fs/gfs2/super.c | 61 ++++------------------------------------------------ fs/gfs2/super.h | 3 --- fs/gfs2/sys.c | 24 ++++++++++----------- 5 files changed, 15 insertions(+), 78 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 5d129ab89733..19750bcb1ce7 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -757,10 +757,7 @@ struct gfs2_sbd { unsigned int sd_replayed_blocks; /* For quiescing the filesystem */ - struct gfs2_holder sd_freeze_gh; - struct mutex sd_freeze_lock; - unsigned int sd_freeze_count; char sd_fsname[GFS2_FSNAME_LEN]; char sd_table_name[GFS2_FSNAME_LEN]; diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 0e3554edb8f2..5f5aba529fb1 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -115,8 +115,6 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) INIT_LIST_HEAD(&sdp->sd_revoke_list); - mutex_init(&sdp->sd_freeze_lock); - return sdp; } diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 4dfda4c94634..c075b62aef59 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -663,54 +663,6 @@ out: return error; } -/** - * gfs2_freeze_fs - freezes the file system - * @sdp: the file system - * - * This function flushes data and meta data for all machines by - * acquiring the transaction log exclusively. All journals are - * ensured to be in a clean state as well. - * - * Returns: errno - */ - -int gfs2_freeze_fs(struct gfs2_sbd *sdp) -{ - int error = 0; - - mutex_lock(&sdp->sd_freeze_lock); - - if (!sdp->sd_freeze_count++) { - error = gfs2_lock_fs_check_clean(sdp, &sdp->sd_freeze_gh); - if (error) - sdp->sd_freeze_count--; - } - - mutex_unlock(&sdp->sd_freeze_lock); - - return error; -} - -/** - * gfs2_unfreeze_fs - unfreezes the file system - * @sdp: the file system - * - * This function allows the file system to proceed by unlocking - * the exclusively held transaction lock. Other GFS2 nodes are - * now free to acquire the lock shared and go on with their lives. - * - */ - -void gfs2_unfreeze_fs(struct gfs2_sbd *sdp) -{ - mutex_lock(&sdp->sd_freeze_lock); - - if (sdp->sd_freeze_count && !--sdp->sd_freeze_count) - gfs2_glock_dq_uninit(&sdp->sd_freeze_gh); - - mutex_unlock(&sdp->sd_freeze_lock); -} - void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) { struct gfs2_dinode *str = buf; @@ -888,13 +840,6 @@ static void gfs2_put_super(struct super_block *sb) int error; struct gfs2_jdesc *jd; - /* Unfreeze the filesystem, if we need to */ - - mutex_lock(&sdp->sd_freeze_lock); - if (sdp->sd_freeze_count) - gfs2_glock_dq_uninit(&sdp->sd_freeze_gh); - mutex_unlock(&sdp->sd_freeze_lock); - /* No more recovery requests */ set_bit(SDF_NORECOVERY, &sdp->sd_flags); smp_mb(); @@ -985,7 +930,7 @@ static int gfs2_freeze(struct super_block *sb) return -EINVAL; for (;;) { - error = gfs2_freeze_fs(sdp); + error = gfs2_lock_fs_check_clean(sdp, &sdp->sd_freeze_gh); if (!error) break; @@ -1013,7 +958,9 @@ static int gfs2_freeze(struct super_block *sb) static int gfs2_unfreeze(struct super_block *sb) { - gfs2_unfreeze_fs(sb->s_fs_info); + struct gfs2_sbd *sdp = sb->s_fs_info; + + gfs2_glock_dq_uninit(&sdp->sd_freeze_gh); return 0; } diff --git a/fs/gfs2/super.h b/fs/gfs2/super.h index a0464680af0b..90e3322ffa10 100644 --- a/fs/gfs2/super.h +++ b/fs/gfs2/super.h @@ -46,9 +46,6 @@ extern void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, struct buffer_head *l_bh); extern int gfs2_statfs_sync(struct super_block *sb, int type); -extern int gfs2_freeze_fs(struct gfs2_sbd *sdp); -extern void gfs2_unfreeze_fs(struct gfs2_sbd *sdp); - extern struct file_system_type gfs2_fs_type; extern struct file_system_type gfs2meta_fs_type; extern const struct export_operations gfs2_export_ops; diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 8056b7b7238e..462e84142759 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -91,19 +91,15 @@ static ssize_t uuid_show(struct gfs2_sbd *sdp, char *buf) static ssize_t freeze_show(struct gfs2_sbd *sdp, char *buf) { - unsigned int count; - - mutex_lock(&sdp->sd_freeze_lock); - count = sdp->sd_freeze_count; - mutex_unlock(&sdp->sd_freeze_lock); + struct super_block *sb = sdp->sd_vfs; + int frozen = (sb->s_writers.frozen == SB_UNFROZEN) ? 0 : 1; - return snprintf(buf, PAGE_SIZE, "%u\n", count); + return snprintf(buf, PAGE_SIZE, "%u\n", frozen); } static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { - ssize_t ret = len; - int error = 0; + int error; int n = simple_strtol(buf, NULL, 0); if (!capable(CAP_SYS_ADMIN)) @@ -111,19 +107,21 @@ static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len) switch (n) { case 0: - gfs2_unfreeze_fs(sdp); + error = thaw_super(sdp->sd_vfs); break; case 1: - error = gfs2_freeze_fs(sdp); + error = freeze_super(sdp->sd_vfs); break; default: - ret = -EINVAL; + return -EINVAL; } - if (error) + if (error) { fs_warn(sdp, "freeze %d error %d", n, error); + return error; + } - return ret; + return len; } static ssize_t withdraw_show(struct gfs2_sbd *sdp, char *buf) -- cgit From 4513899092b3254b3539f92a65d2839afa1d50f6 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Mon, 28 Jan 2013 09:30:07 +0000 Subject: GFS2: Use ->writepages for ordered writes Instead of using a list of buffers to write ahead of the journal flush, this now uses a list of inodes and calls ->writepages via filemap_fdatawrite() in order to achieve the same thing. For most use cases this results in a shorter ordered write list, as well as much larger i/os being issued. The ordered write list is sorted by inode number before writing in order to retain the disk block ordering between inodes as per the previous code. The previous ordered write code used to conflict in its assumptions about how to write out the disk blocks with mpage_writepages() so that with this updated version we can also use mpage_writepages() for GFS2's ordered write, writepages implementation. So we will also send larger i/os from writeback too. Signed-off-by: Steven Whitehouse --- fs/gfs2/aops.c | 13 +++++---- fs/gfs2/bmap.c | 2 ++ fs/gfs2/incore.h | 3 +++ fs/gfs2/log.c | 76 +++++++++++++++++++++++++--------------------------- fs/gfs2/log.h | 12 +++++++++ fs/gfs2/ops_fstype.c | 1 + fs/gfs2/super.c | 1 + fs/gfs2/trans.c | 41 +++++++++++++--------------- 8 files changed, 79 insertions(+), 70 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 92340dd23bba..24f414f0ce61 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -230,16 +230,14 @@ out_ignore: } /** - * gfs2_writeback_writepages - Write a bunch of dirty pages back to disk + * gfs2_writepages - Write a bunch of dirty pages back to disk * @mapping: The mapping to write * @wbc: Write-back control * - * For the data=writeback case we can already ignore buffer heads - * and write whole extents at once. This is a big reduction in the - * number of I/O requests we send and the bmap calls we make in this case. + * Used for both ordered and writeback modes. */ -static int gfs2_writeback_writepages(struct address_space *mapping, - struct writeback_control *wbc) +static int gfs2_writepages(struct address_space *mapping, + struct writeback_control *wbc) { return mpage_writepages(mapping, wbc, gfs2_get_block_noalloc); } @@ -1102,7 +1100,7 @@ cannot_release: static const struct address_space_operations gfs2_writeback_aops = { .writepage = gfs2_writeback_writepage, - .writepages = gfs2_writeback_writepages, + .writepages = gfs2_writepages, .readpage = gfs2_readpage, .readpages = gfs2_readpages, .write_begin = gfs2_write_begin, @@ -1118,6 +1116,7 @@ static const struct address_space_operations gfs2_writeback_aops = { static const struct address_space_operations gfs2_ordered_aops = { .writepage = gfs2_ordered_writepage, + .writepages = gfs2_writepages, .readpage = gfs2_readpage, .readpages = gfs2_readpages, .write_begin = gfs2_write_begin, diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index 7a8627569a25..d29d7793b211 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -22,6 +22,7 @@ #include "meta_io.h" #include "quota.h" #include "rgrp.h" +#include "log.h" #include "super.h" #include "trans.h" #include "dir.h" @@ -1137,6 +1138,7 @@ static int trunc_end(struct gfs2_inode *ip) ip->i_height = 0; ip->i_goal = ip->i_no_addr; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); + gfs2_ordered_del_inode(ip); } ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; ip->i_diskflags &= ~GFS2_DIF_TRUNC_IN_PROG; diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 19750bcb1ce7..1533cf8b4269 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -340,6 +340,7 @@ enum { GIF_QD_LOCKED = 1, GIF_ALLOC_FAILED = 2, GIF_SW_PAGED = 3, + GIF_ORDERED = 4, }; struct gfs2_inode { @@ -356,6 +357,7 @@ struct gfs2_inode { struct gfs2_rgrpd *i_rgd; u64 i_goal; /* goal block for allocations */ struct rw_semaphore i_rw_mutex; + struct list_head i_ordered; struct list_head i_trunc_list; __be64 *i_hash_cache; u32 i_entries; @@ -722,6 +724,7 @@ struct gfs2_sbd { struct list_head sd_log_le_revoke; struct list_head sd_log_le_databuf; struct list_head sd_log_le_ordered; + spinlock_t sd_ordered_lock; atomic_t sd_log_thresh1; atomic_t sd_log_thresh2; diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index f4beeb9c81c1..9a2ca8be7647 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -482,70 +482,66 @@ static void log_flush_wait(struct gfs2_sbd *sdp) } } -static int bd_cmp(void *priv, struct list_head *a, struct list_head *b) +static int ip_cmp(void *priv, struct list_head *a, struct list_head *b) { - struct gfs2_bufdata *bda, *bdb; + struct gfs2_inode *ipa, *ipb; - bda = list_entry(a, struct gfs2_bufdata, bd_list); - bdb = list_entry(b, struct gfs2_bufdata, bd_list); + ipa = list_entry(a, struct gfs2_inode, i_ordered); + ipb = list_entry(b, struct gfs2_inode, i_ordered); - if (bda->bd_bh->b_blocknr < bdb->bd_bh->b_blocknr) + if (ipa->i_no_addr < ipb->i_no_addr) return -1; - if (bda->bd_bh->b_blocknr > bdb->bd_bh->b_blocknr) + if (ipa->i_no_addr > ipb->i_no_addr) return 1; return 0; } static void gfs2_ordered_write(struct gfs2_sbd *sdp) { - struct gfs2_bufdata *bd; - struct buffer_head *bh; + struct gfs2_inode *ip; LIST_HEAD(written); - gfs2_log_lock(sdp); - list_sort(NULL, &sdp->sd_log_le_ordered, &bd_cmp); + spin_lock(&sdp->sd_ordered_lock); + list_sort(NULL, &sdp->sd_log_le_ordered, &ip_cmp); while (!list_empty(&sdp->sd_log_le_ordered)) { - bd = list_entry(sdp->sd_log_le_ordered.next, struct gfs2_bufdata, bd_list); - list_move(&bd->bd_list, &written); - bh = bd->bd_bh; - if (!buffer_dirty(bh)) + ip = list_entry(sdp->sd_log_le_ordered.next, struct gfs2_inode, i_ordered); + list_move(&ip->i_ordered, &written); + if (ip->i_inode.i_mapping->nrpages == 0) continue; - get_bh(bh); - gfs2_log_unlock(sdp); - lock_buffer(bh); - if (buffer_mapped(bh) && test_clear_buffer_dirty(bh)) { - bh->b_end_io = end_buffer_write_sync; - submit_bh(WRITE_SYNC, bh); - } else { - unlock_buffer(bh); - brelse(bh); - } - gfs2_log_lock(sdp); + spin_unlock(&sdp->sd_ordered_lock); + filemap_fdatawrite(ip->i_inode.i_mapping); + spin_lock(&sdp->sd_ordered_lock); } list_splice(&written, &sdp->sd_log_le_ordered); - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ordered_lock); } static void gfs2_ordered_wait(struct gfs2_sbd *sdp) { - struct gfs2_bufdata *bd; - struct buffer_head *bh; + struct gfs2_inode *ip; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ordered_lock); while (!list_empty(&sdp->sd_log_le_ordered)) { - bd = list_entry(sdp->sd_log_le_ordered.prev, struct gfs2_bufdata, bd_list); - bh = bd->bd_bh; - if (buffer_locked(bh)) { - get_bh(bh); - gfs2_log_unlock(sdp); - wait_on_buffer(bh); - brelse(bh); - gfs2_log_lock(sdp); + ip = list_entry(sdp->sd_log_le_ordered.next, struct gfs2_inode, i_ordered); + list_del(&ip->i_ordered); + WARN_ON(!test_and_clear_bit(GIF_ORDERED, &ip->i_flags)); + if (ip->i_inode.i_mapping->nrpages == 0) continue; - } - list_del_init(&bd->bd_list); + spin_unlock(&sdp->sd_ordered_lock); + filemap_fdatawait(ip->i_inode.i_mapping); + spin_lock(&sdp->sd_ordered_lock); } - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ordered_lock); +} + +void gfs2_ordered_del_inode(struct gfs2_inode *ip) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + + spin_lock(&sdp->sd_ordered_lock); + if (test_and_clear_bit(GIF_ORDERED, &ip->i_flags)) + list_del(&ip->i_ordered); + spin_unlock(&sdp->sd_ordered_lock); } /** diff --git a/fs/gfs2/log.h b/fs/gfs2/log.h index 3fd5215ea25f..3566f35915e0 100644 --- a/fs/gfs2/log.h +++ b/fs/gfs2/log.h @@ -48,6 +48,18 @@ static inline void gfs2_log_pointers_init(struct gfs2_sbd *sdp, sdp->sd_log_head = sdp->sd_log_tail = value; } +static inline void gfs2_ordered_add_inode(struct gfs2_inode *ip) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + + if (!test_bit(GIF_ORDERED, &ip->i_flags)) { + spin_lock(&sdp->sd_ordered_lock); + if (!test_and_set_bit(GIF_ORDERED, &ip->i_flags)) + list_add(&ip->i_ordered, &sdp->sd_log_le_ordered); + spin_unlock(&sdp->sd_ordered_lock); + } +} +extern void gfs2_ordered_del_inode(struct gfs2_inode *ip); extern unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct, unsigned int ssize); diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 5f5aba529fb1..e063f22d9e4c 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -102,6 +102,7 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) INIT_LIST_HEAD(&sdp->sd_log_le_revoke); INIT_LIST_HEAD(&sdp->sd_log_le_databuf); INIT_LIST_HEAD(&sdp->sd_log_le_ordered); + spin_lock_init(&sdp->sd_ordered_lock); init_waitqueue_head(&sdp->sd_log_waitq); init_waitqueue_head(&sdp->sd_logd_waitq); diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index c075b62aef59..a3b40eeaa6e2 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -1524,6 +1524,7 @@ out: /* Case 3 starts here */ truncate_inode_pages(&inode->i_data, 0); gfs2_rs_delete(ip); + gfs2_ordered_del_inode(ip); clear_inode(inode); gfs2_dir_hash_inval(ip); ip->i_gl->gl_object = NULL; diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c index 14dbf6d3cdc0..88162fae27a5 100644 --- a/fs/gfs2/trans.c +++ b/fs/gfs2/trans.c @@ -159,7 +159,9 @@ static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl, } /** - * databuf_lo_add - Add a databuf to the transaction. + * gfs2_trans_add_data - Add a databuf to the transaction. + * @gl: The inode glock associated with the buffer + * @bh: The buffer to add * * This is used in two distinct cases: * i) In ordered write mode @@ -174,33 +176,18 @@ static struct gfs2_bufdata *gfs2_alloc_bufdata(struct gfs2_glock *gl, * blocks, which isn't an enormous overhead but twice as much as * for normal metadata blocks. */ -static void databuf_lo_add(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) +void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) { struct gfs2_trans *tr = current->journal_info; - struct address_space *mapping = bd->bd_bh->b_page->mapping; + struct gfs2_sbd *sdp = gl->gl_sbd; + struct address_space *mapping = bh->b_page->mapping; struct gfs2_inode *ip = GFS2_I(mapping->host); + struct gfs2_bufdata *bd; - if (tr) - tr->tr_touched = 1; - if (!list_empty(&bd->bd_list)) + if (!gfs2_is_jdata(ip)) { + gfs2_ordered_add_inode(ip); return; - set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); - set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); - if (gfs2_is_jdata(ip)) { - gfs2_pin(sdp, bd->bd_bh); - tr->tr_num_databuf_new++; - sdp->sd_log_num_databuf++; - list_add_tail(&bd->bd_list, &sdp->sd_log_le_databuf); - } else { - list_add_tail(&bd->bd_list, &sdp->sd_log_le_ordered); } -} - -void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) -{ - - struct gfs2_sbd *sdp = gl->gl_sbd; - struct gfs2_bufdata *bd; lock_buffer(bh); gfs2_log_lock(sdp); @@ -214,7 +201,15 @@ void gfs2_trans_add_data(struct gfs2_glock *gl, struct buffer_head *bh) gfs2_log_lock(sdp); } gfs2_assert(sdp, bd->bd_gl == gl); - databuf_lo_add(sdp, bd); + tr->tr_touched = 1; + if (list_empty(&bd->bd_list)) { + set_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); + set_bit(GLF_DIRTY, &bd->bd_gl->gl_flags); + gfs2_pin(sdp, bd->bd_bh); + tr->tr_num_databuf_new++; + sdp->sd_log_num_databuf++; + list_add_tail(&bd->bd_list, &sdp->sd_log_le_databuf); + } gfs2_log_unlock(sdp); unlock_buffer(bh); } -- cgit From 885c91f74662404dc179b8775494df383479311c Mon Sep 17 00:00:00 2001 From: majianpeng Date: Tue, 29 Jan 2013 13:16:01 +0800 Subject: nfsd: Fix memleak in svc_export_put In func svc_export_parse, the uuid which used kmemdup to alloc will be changed in func export_update.So the later kfree don't free this memory. And it can't be free in func svc_export_parse because other place still used.So put this operation in func svc_export_put. Signed-off-by: Jianpeng Ma Signed-off-by: J. Bruce Fields --- fs/nfsd/export.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index a3946cf13fc8..45159ee952e6 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -315,6 +315,7 @@ static void svc_export_put(struct kref *ref) path_put(&exp->ex_path); auth_domain_put(exp->ex_client); nfsd4_fslocs_free(&exp->ex_fslocs); + kfree(exp->ex_uuid); kfree(exp); } -- cgit From 091e26dfc156aeb3b73bc5c5f277e433ad39331c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 29 Jan 2013 22:48:17 -0500 Subject: ext4: fix possible use-after-free with AIO Running AIO is pinning inode in memory using file reference. Once AIO is completed using aio_complete(), file reference is put and inode can be freed from memory. So we have to be sure that calling aio_complete() is the last thing we do with the inode. CC: stable@vger.kernel.org Reviewed-by: Carlos Maiolino Acked-by: Jeff Moyer Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 2 +- fs/ext4/page-io.c | 9 ++++----- 2 files changed, 5 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 86bf43d6dfcd..07d9defeaf8c 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2850,9 +2850,9 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, if (!(io_end->flag & EXT4_IO_END_UNWRITTEN)) { ext4_free_io_end(io_end); out: + inode_dio_done(inode); if (is_async) aio_complete(iocb, ret, 0); - inode_dio_done(inode); return; } diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 5d8c66948e1b..809b31003ecc 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -102,14 +102,13 @@ static int ext4_end_io(ext4_io_end_t *io) "(inode %lu, offset %llu, size %zd, error %d)", inode->i_ino, offset, size, ret); } - if (io->iocb) - aio_complete(io->iocb, io->result, 0); - - if (io->flag & EXT4_IO_END_DIRECT) - inode_dio_done(inode); /* Wake up anyone waiting on unwritten extent conversion */ if (atomic_dec_and_test(&EXT4_I(inode)->i_unwritten)) wake_up_all(ext4_ioend_wq(inode)); + if (io->flag & EXT4_IO_END_DIRECT) + inode_dio_done(inode); + if (io->iocb) + aio_complete(io->iocb, io->result, 0); return ret; } -- cgit From e7b04ac00ee273cb2c699ed14139fc072add4097 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Wed, 30 Jan 2013 00:39:28 -0500 Subject: jbd2: don't wake kjournald unnecessarily Don't send an extra wakeup to kjournald in the case where we already have the proper target in j_commit_request, i.e. that transaction has already been requested for commit. commit deeeaf13 "jbd2: fix fsync() tid wraparound bug" changed the logic leading to a wakeup, but it caused some extra wakeups which were found to lead to a measurable performance regression. Signed-off-by: Eric Sandeen [tytso@mit.edu: reworked check to make it clearer] Signed-off-by: "Theodore Ts'o" --- fs/jbd2/journal.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index dbf41f9452db..1a80e3146a59 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -513,6 +513,10 @@ int __jbd2_log_space_left(journal_t *journal) */ int __jbd2_log_start_commit(journal_t *journal, tid_t target) { + /* Return if the txn has already requested to be committed */ + if (journal->j_commit_request == target) + return 0; + /* * The only transaction we can possibly wait upon is the * currently running transaction (if it exists). Otherwise, -- cgit From ab225417825963b6dc66be7ea80f94ac1378dfdf Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 22 Jan 2013 00:17:06 -0500 Subject: NFS: Don't silently fail setattr() requests on mountpoints Ensure that any setattr and getattr requests for junctions and/or mountpoints are sent to the server. Ever since commit 0ec26fd0698 (vfs: automount should ignore LOOKUP_FOLLOW), we have silently dropped any setattr requests to a server-side mountpoint. For referrals, we have silently dropped both getattr and setattr requests. This patch restores the original behaviour for setattr on mountpoints, and tries to do the same for referrals, provided that we have a filehandle... Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/nfs/namespace.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) (limited to 'fs') diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index dd057bc6b65b..fc8dc20fdeb9 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -177,11 +177,31 @@ out_nofree: return mnt; } +static int +nfs_namespace_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) +{ + if (NFS_FH(dentry->d_inode)->size != 0) + return nfs_getattr(mnt, dentry, stat); + generic_fillattr(dentry->d_inode, stat); + return 0; +} + +static int +nfs_namespace_setattr(struct dentry *dentry, struct iattr *attr) +{ + if (NFS_FH(dentry->d_inode)->size != 0) + return nfs_setattr(dentry, attr); + return -EACCES; +} + const struct inode_operations nfs_mountpoint_inode_operations = { .getattr = nfs_getattr, + .setattr = nfs_setattr, }; const struct inode_operations nfs_referral_inode_operations = { + .getattr = nfs_namespace_getattr, + .setattr = nfs_namespace_setattr, }; static void nfs_expire_automounts(struct work_struct *work) -- cgit From c489ee290bdbbace6bb63ebe6ebd4dd605819495 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 30 Jan 2013 13:04:10 -0500 Subject: NFSv4.1: Handle NFS4ERR_DELAY when resetting the NFSv4.1 session NFS4ERR_DELAY is a legal reply when we call DESTROY_SESSION. It usually means that the server is busy handling an unfinished RPC request. Just sleep for a second and then retry. We also need to be able to handle the NFS4ERR_BACK_CHAN_BUSY return value. If the NFS server has outstanding callbacks, we just want to similarly sleep & retry. Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/nfs/nfs4state.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index f72561ca6895..e61f68d5ef21 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -2018,8 +2018,18 @@ static int nfs4_reset_session(struct nfs_client *clp) nfs4_begin_drain_session(clp); cred = nfs4_get_exchange_id_cred(clp); status = nfs4_proc_destroy_session(clp->cl_session, cred); - if (status && status != -NFS4ERR_BADSESSION && - status != -NFS4ERR_DEADSESSION) { + switch (status) { + case 0: + case -NFS4ERR_BADSESSION: + case -NFS4ERR_DEADSESSION: + break; + case -NFS4ERR_BACK_CHAN_BUSY: + case -NFS4ERR_DELAY: + set_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state); + status = 0; + ssleep(1); + goto out; + default: status = nfs4_recovery_handle_error(clp, status); goto out; } -- cgit From c2132c1bc73d9a279cec148f74ea709c960b3d89 Mon Sep 17 00:00:00 2001 From: Anatol Pomozov Date: Mon, 14 Jan 2013 22:30:00 -0800 Subject: Do not use RCU for current process credentials Commit c69e8d9c0 added rcu lock to fuse/dir.c It was assuming that 'task' is some other process but in fact this parameter always equals to 'current'. Inline this parameter to make it more readable and remove RCU lock as it is not needed when access current process credentials. Signed-off-by: Anatol Pomozov Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 22 +++++++++------------- fs/fuse/file.c | 2 +- fs/fuse/fuse_i.h | 4 ++-- fs/fuse/inode.c | 2 +- 4 files changed, 13 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index ed8f8c55ce10..aa0b6ade0e68 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -985,7 +985,7 @@ int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, /* * Calling into a user-controlled filesystem gives the filesystem - * daemon ptrace-like capabilities over the requester process. This + * daemon ptrace-like capabilities over the current process. This * means, that the filesystem daemon is able to record the exact * filesystem operations performed, and can also control the behavior * of the requester process in otherwise impossible ways. For example @@ -996,27 +996,23 @@ int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, * for which the owner of the mount has ptrace privilege. This * excludes processes started by other users, suid or sgid processes. */ -int fuse_allow_task(struct fuse_conn *fc, struct task_struct *task) +int fuse_allow_current_process(struct fuse_conn *fc) { const struct cred *cred; - int ret; if (fc->flags & FUSE_ALLOW_OTHER) return 1; - rcu_read_lock(); - ret = 0; - cred = __task_cred(task); + cred = current_cred(); if (uid_eq(cred->euid, fc->user_id) && uid_eq(cred->suid, fc->user_id) && uid_eq(cred->uid, fc->user_id) && gid_eq(cred->egid, fc->group_id) && gid_eq(cred->sgid, fc->group_id) && gid_eq(cred->gid, fc->group_id)) - ret = 1; - rcu_read_unlock(); + return 1; - return ret; + return 0; } static int fuse_access(struct inode *inode, int mask) @@ -1077,7 +1073,7 @@ static int fuse_permission(struct inode *inode, int mask) bool refreshed = false; int err = 0; - if (!fuse_allow_task(fc, current)) + if (!fuse_allow_current_process(fc)) return -EACCES; /* @@ -1544,7 +1540,7 @@ static int fuse_do_setattr(struct dentry *entry, struct iattr *attr, loff_t oldsize; int err; - if (!fuse_allow_task(fc, current)) + if (!fuse_allow_current_process(fc)) return -EACCES; if (!(fc->flags & FUSE_DEFAULT_PERMISSIONS)) @@ -1653,7 +1649,7 @@ static int fuse_getattr(struct vfsmount *mnt, struct dentry *entry, struct inode *inode = entry->d_inode; struct fuse_conn *fc = get_fuse_conn(inode); - if (!fuse_allow_task(fc, current)) + if (!fuse_allow_current_process(fc)) return -EACCES; return fuse_update_attributes(inode, stat, NULL, NULL); @@ -1756,7 +1752,7 @@ static ssize_t fuse_listxattr(struct dentry *entry, char *list, size_t size) struct fuse_getxattr_out outarg; ssize_t ret; - if (!fuse_allow_task(fc, current)) + if (!fuse_allow_current_process(fc)) return -EACCES; if (fc->no_listxattr) diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 28bc9c672196..a010585b0a74 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -2082,7 +2082,7 @@ long fuse_ioctl_common(struct file *file, unsigned int cmd, struct inode *inode = file->f_dentry->d_inode; struct fuse_conn *fc = get_fuse_conn(inode); - if (!fuse_allow_task(fc, current)) + if (!fuse_allow_current_process(fc)) return -EACCES; if (is_bad_inode(inode)) diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 13befcd29c5b..af51c146a9ae 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -774,9 +774,9 @@ void fuse_ctl_remove_conn(struct fuse_conn *fc); int fuse_valid_type(int m); /** - * Is task allowed to perform filesystem operation? + * Is current process allowed to perform filesystem operation? */ -int fuse_allow_task(struct fuse_conn *fc, struct task_struct *task); +int fuse_allow_current_process(struct fuse_conn *fc); u64 fuse_lock_owner_id(struct fuse_conn *fc, fl_owner_t id); diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 9d95a5a3d55c..79b70deb7cd6 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -408,7 +408,7 @@ static int fuse_statfs(struct dentry *dentry, struct kstatfs *buf) struct fuse_statfs_out outarg; int err; - if (!fuse_allow_task(fc, current)) { + if (!fuse_allow_current_process(fc)) { buf->f_type = FUSE_SUPER_MAGIC; return 0; } -- cgit From 4582a4ab2a0e7218449fb2e895d0aae9ea753c94 Mon Sep 17 00:00:00 2001 From: Feng Shuo Date: Tue, 15 Jan 2013 11:23:28 +0800 Subject: FUSE: Adapt readdirplus to application usage patterns Use the same adaptive readdirplus mechanism as NFS: http://permalink.gmane.org/gmane.linux.nfs/49299 If the user space implementation wants to disable readdirplus temporarily, it could just return ENOTSUPP. Then kernel will recall it with readdir. Signed-off-by: Feng Shuo Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 31 ++++++++++++++++++++++++++++--- fs/fuse/fuse_i.h | 9 +++++++++ fs/fuse/inode.c | 1 + 3 files changed, 38 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index aa0b6ade0e68..dc5e64893375 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -14,6 +14,27 @@ #include #include +static bool fuse_use_readdirplus(struct inode *dir, struct file *filp) +{ + struct fuse_conn *fc = get_fuse_conn(dir); + struct fuse_inode *fi = get_fuse_inode(dir); + + if (!fc->do_readdirplus) + return false; + if (test_and_clear_bit(FUSE_I_ADVISE_RDPLUS, &fi->state)) + return true; + if (filp->f_pos == 0) + return true; + return false; +} + +static void fuse_advise_use_readdirplus(struct inode *dir) +{ + struct fuse_inode *fi = get_fuse_inode(dir); + + set_bit(FUSE_I_ADVISE_RDPLUS, &fi->state); +} + #if BITS_PER_LONG >= 64 static inline void fuse_dentry_settime(struct dentry *entry, u64 time) { @@ -219,6 +240,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, unsigned int flags) attr_version); fuse_change_entry_timeout(entry, &outarg); } + fuse_advise_use_readdirplus(inode); return 1; } @@ -355,6 +377,7 @@ static struct dentry *fuse_lookup(struct inode *dir, struct dentry *entry, else fuse_invalidate_entry_cache(entry); + fuse_advise_use_readdirplus(dir); return newent; out_iput: @@ -1290,7 +1313,7 @@ static int parse_dirplusfile(char *buf, size_t nbytes, struct file *file, static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) { - int err; + int plus, err; size_t nbytes; struct page *page; struct inode *inode = file->f_path.dentry->d_inode; @@ -1310,11 +1333,13 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) fuse_put_request(fc, req); return -ENOMEM; } + + plus = fuse_use_readdirplus(inode, file); req->out.argpages = 1; req->num_pages = 1; req->pages[0] = page; req->page_descs[0].length = PAGE_SIZE; - if (fc->do_readdirplus) { + if (plus) { attr_version = fuse_get_attr_version(fc); fuse_read_fill(req, file, file->f_pos, PAGE_SIZE, FUSE_READDIRPLUS); @@ -1327,7 +1352,7 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) err = req->out.h.error; fuse_put_request(fc, req); if (!err) { - if (fc->do_readdirplus) { + if (plus) { err = parse_dirplusfile(page_address(page), nbytes, file, dstbuf, filldir, attr_version); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index af51c146a9ae..fc55dd33c1e2 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -106,6 +106,15 @@ struct fuse_inode { /** List of writepage requestst (pending or sent) */ struct list_head writepages; + + /** Miscellaneous bits describing inode state */ + unsigned long state; +}; + +/** FUSE inode state bits */ +enum { + /** Advise readdirplus */ + FUSE_I_ADVISE_RDPLUS, }; struct fuse_conn; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 79b70deb7cd6..9876a87255fe 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -92,6 +92,7 @@ static struct inode *fuse_alloc_inode(struct super_block *sb) fi->attr_version = 0; fi->writectr = 0; fi->orig_ino = 0; + fi->state = 0; INIT_LIST_HEAD(&fi->write_files); INIT_LIST_HEAD(&fi->queued_writes); INIT_LIST_HEAD(&fi->writepages); -- cgit From 322b2b9032f4beba6f1c4158852a5a5b9ab841d7 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 11 Jan 2013 16:39:51 -0500 Subject: Revert "NFS: add nfs_sb_deactive_async to avoid deadlock" This reverts commit 324d003b0cd82151adbaecefef57b73f7959a469. The deadlock turned out to be caused by a workqueue limitation that has now been worked around in the RPC code (see comment in rpc_free_task). Signed-off-by: Trond Myklebust --- fs/nfs/inode.c | 5 +---- fs/nfs/internal.h | 1 - fs/nfs/nfs4proc.c | 2 +- fs/nfs/super.c | 49 ------------------------------------------------- fs/nfs/unlink.c | 2 +- 5 files changed, 3 insertions(+), 56 deletions(-) (limited to 'fs') diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index ebeb94ce1b0b..6acc73c80d7f 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -694,10 +694,7 @@ static void __put_nfs_open_context(struct nfs_open_context *ctx, int is_sync) if (ctx->cred != NULL) put_rpccred(ctx->cred); dput(ctx->dentry); - if (is_sync) - nfs_sb_deactive(sb); - else - nfs_sb_deactive_async(sb); + nfs_sb_deactive(sb); kfree(ctx->mdsthreshold); kfree(ctx); } diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index f0e6c7df1a07..541c9ebdbc5a 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -329,7 +329,6 @@ extern int __init register_nfs_fs(void); extern void __exit unregister_nfs_fs(void); extern void nfs_sb_active(struct super_block *sb); extern void nfs_sb_deactive(struct super_block *sb); -extern void nfs_sb_deactive_async(struct super_block *sb); /* namespace.c */ #define NFS_PATH_CANONICAL 1 diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index cf747ef86650..62fa4001e47a 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2088,7 +2088,7 @@ static void nfs4_free_closedata(void *data) nfs4_put_open_state(calldata->state); nfs_free_seqid(calldata->arg.seqid); nfs4_put_state_owner(sp); - nfs_sb_deactive_async(sb); + nfs_sb_deactive(sb); kfree(calldata); } diff --git a/fs/nfs/super.c b/fs/nfs/super.c index b056b1628722..befbae0cce41 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -54,7 +54,6 @@ #include #include #include -#include #include @@ -418,54 +417,6 @@ void nfs_sb_deactive(struct super_block *sb) } EXPORT_SYMBOL_GPL(nfs_sb_deactive); -static int nfs_deactivate_super_async_work(void *ptr) -{ - struct super_block *sb = ptr; - - deactivate_super(sb); - module_put_and_exit(0); - return 0; -} - -/* - * same effect as deactivate_super, but will do final unmount in kthread - * context - */ -static void nfs_deactivate_super_async(struct super_block *sb) -{ - struct task_struct *task; - char buf[INET6_ADDRSTRLEN + 1]; - struct nfs_server *server = NFS_SB(sb); - struct nfs_client *clp = server->nfs_client; - - if (!atomic_add_unless(&sb->s_active, -1, 1)) { - rcu_read_lock(); - snprintf(buf, sizeof(buf), - rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR)); - rcu_read_unlock(); - - __module_get(THIS_MODULE); - task = kthread_run(nfs_deactivate_super_async_work, sb, - "%s-deactivate-super", buf); - if (IS_ERR(task)) { - pr_err("%s: kthread_run: %ld\n", - __func__, PTR_ERR(task)); - /* make synchronous call and hope for the best */ - deactivate_super(sb); - module_put(THIS_MODULE); - } - } -} - -void nfs_sb_deactive_async(struct super_block *sb) -{ - struct nfs_server *server = NFS_SB(sb); - - if (atomic_dec_and_test(&server->active)) - nfs_deactivate_super_async(sb); -} -EXPORT_SYMBOL_GPL(nfs_sb_deactive_async); - /* * Deliver file system statistics to userspace */ diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index 3f79c77153b8..13cea637eff8 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -95,7 +95,7 @@ static void nfs_async_unlink_release(void *calldata) nfs_dec_sillycount(data->dir); nfs_free_unlinkdata(data); - nfs_sb_deactive_async(sb); + nfs_sb_deactive(sb); } static void nfs_unlink_prepare(struct rpc_task *task, void *calldata) -- cgit From 3c911608085bf2d5a0822c418129f96a2a89d1b5 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:55:02 +0000 Subject: btrfs: don't try to notify udev about missing devices If we remove a missing device, bdev is null, and if we send that off to btrfs_kobject_uevent we'll panic. Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 5cce6aa74012..485a5423e3c6 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1556,7 +1556,8 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) ret = 0; /* Notify udev that device has changed */ - btrfs_kobject_uevent(bdev, KOBJ_CHANGE); + if (bdev) + btrfs_kobject_uevent(bdev, KOBJ_CHANGE); error_brelse: brelse(bh); -- cgit From 64a167011bcabc1e855658387c8a4464b71f3138 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 15 Jul 2009 23:29:37 +0100 Subject: Btrfs: add rw argument to merge_bio_hook() We'll want to merge writes so they can fill a full RAID[56] stripe, but not necessarily reads. Signed-off-by: David Woodhouse Signed-off-by: Chris Mason --- fs/btrfs/compression.c | 4 ++-- fs/btrfs/ctree.h | 6 +++--- fs/btrfs/extent_io.c | 6 +++--- fs/btrfs/extent_io.h | 2 +- fs/btrfs/inode.c | 4 ++-- 5 files changed, 11 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index 94ab2f80e7e3..15b94089abc4 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -372,7 +372,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start, page = compressed_pages[pg_index]; page->mapping = inode->i_mapping; if (bio->bi_size) - ret = io_tree->ops->merge_bio_hook(page, 0, + ret = io_tree->ops->merge_bio_hook(WRITE, page, 0, PAGE_CACHE_SIZE, bio, 0); else @@ -655,7 +655,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, page->index = em_start >> PAGE_CACHE_SHIFT; if (comp_bio->bi_size) - ret = tree->ops->merge_bio_hook(page, 0, + ret = tree->ops->merge_bio_hook(READ, page, 0, PAGE_CACHE_SIZE, comp_bio, 0); else diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 547b7b05727f..0ab51be6879f 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3459,9 +3459,9 @@ int btrfs_writepages(struct address_space *mapping, struct writeback_control *wbc); int btrfs_create_subvol_root(struct btrfs_trans_handle *trans, struct btrfs_root *new_root, u64 new_dirid); -int btrfs_merge_bio_hook(struct page *page, unsigned long offset, - size_t size, struct bio *bio, unsigned long bio_flags); - +int btrfs_merge_bio_hook(int rw, struct page *page, unsigned long offset, + size_t size, struct bio *bio, + unsigned long bio_flags); int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf); int btrfs_readpage(struct file *file, struct page *page); void btrfs_evict_inode(struct inode *inode); diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 1b319df29eee..86ecca48c604 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -2489,13 +2489,13 @@ static int __must_check submit_one_bio(int rw, struct bio *bio, return ret; } -static int merge_bio(struct extent_io_tree *tree, struct page *page, +static int merge_bio(int rw, struct extent_io_tree *tree, struct page *page, unsigned long offset, size_t size, struct bio *bio, unsigned long bio_flags) { int ret = 0; if (tree->ops && tree->ops->merge_bio_hook) - ret = tree->ops->merge_bio_hook(page, offset, size, bio, + ret = tree->ops->merge_bio_hook(rw, page, offset, size, bio, bio_flags); BUG_ON(ret < 0); return ret; @@ -2530,7 +2530,7 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree, sector; if (prev_bio_flags != bio_flags || !contig || - merge_bio(tree, page, offset, page_size, bio, bio_flags) || + merge_bio(rw, tree, page, offset, page_size, bio, bio_flags) || bio_add_page(bio, page, page_size, offset) < page_size) { ret = submit_one_bio(rw, bio, mirror_num, prev_bio_flags); diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index 2eacfabd3263..b14b36a80eba 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -72,7 +72,7 @@ struct extent_io_ops { int (*writepage_start_hook)(struct page *page, u64 start, u64 end); int (*writepage_io_hook)(struct page *page, u64 start, u64 end); extent_submit_bio_hook_t *submit_bio_hook; - int (*merge_bio_hook)(struct page *page, unsigned long offset, + int (*merge_bio_hook)(int rw, struct page *page, unsigned long offset, size_t size, struct bio *bio, unsigned long bio_flags); int (*readpage_io_hook)(struct page *page, u64 start, u64 end); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 67ed24ae86bb..1b98c4ce3c6f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1566,7 +1566,7 @@ static void btrfs_clear_bit_hook(struct inode *inode, * extent_io.c merge_bio_hook, this must check the chunk tree to make sure * we don't create bios that span stripes or chunks */ -int btrfs_merge_bio_hook(struct page *page, unsigned long offset, +int btrfs_merge_bio_hook(int rw, struct page *page, unsigned long offset, size_t size, struct bio *bio, unsigned long bio_flags) { @@ -1581,7 +1581,7 @@ int btrfs_merge_bio_hook(struct page *page, unsigned long offset, length = bio->bi_size; map_length = length; - ret = btrfs_map_block(root->fs_info, READ, logical, + ret = btrfs_map_block(root->fs_info, rw, logical, &map_length, NULL, 0); /* Will always return 0 with map_multi == NULL */ BUG_ON(ret < 0); -- cgit From 53b381b3abeb86f12787a6c40fee9b2f71edc23b Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Tue, 29 Jan 2013 18:40:14 -0500 Subject: Btrfs: RAID5 and RAID6 This builds on David Woodhouse's original Btrfs raid5/6 implementation. The code has changed quite a bit, blame Chris Mason for any bugs. Read/modify/write is done after the higher levels of the filesystem have prepared a given bio. This means the higher layers are not responsible for building full stripes, and they don't need to query for the topology of the extents that may get allocated during delayed allocation runs. It also means different files can easily share the same stripe. But, it does expose us to incorrect parity if we crash or lose power while doing a read/modify/write cycle. This will be addressed in a later commit. Scrub is unable to repair crc errors on raid5/6 chunks. Discard does not work on raid5/6 (yet) The stripe size is fixed at 64KiB per disk. This will be tunable in a later commit. Signed-off-by: Chris Mason --- fs/btrfs/Kconfig | 2 + fs/btrfs/Makefile | 2 +- fs/btrfs/ctree.h | 35 +- fs/btrfs/disk-io.c | 62 +- fs/btrfs/disk-io.h | 7 + fs/btrfs/extent-tree.c | 88 ++- fs/btrfs/extent_io.c | 18 +- fs/btrfs/free-space-cache.c | 50 +- fs/btrfs/inode.c | 18 +- fs/btrfs/raid56.c | 1647 +++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/raid56.h | 51 ++ fs/btrfs/scrub.c | 8 + fs/btrfs/transaction.c | 3 + fs/btrfs/volumes.c | 385 ++++++++-- fs/btrfs/volumes.h | 9 +- 15 files changed, 2283 insertions(+), 102 deletions(-) create mode 100644 fs/btrfs/raid56.c create mode 100644 fs/btrfs/raid56.h (limited to 'fs') diff --git a/fs/btrfs/Kconfig b/fs/btrfs/Kconfig index d33f01c08b60..4f5dc93fa2f8 100644 --- a/fs/btrfs/Kconfig +++ b/fs/btrfs/Kconfig @@ -6,6 +6,8 @@ config BTRFS_FS select ZLIB_DEFLATE select LZO_COMPRESS select LZO_DECOMPRESS + select RAID6_PQ + help Btrfs is a new filesystem with extents, writable snapshotting, support for multiple devices and many more features. diff --git a/fs/btrfs/Makefile b/fs/btrfs/Makefile index 7df3e0f0ee51..3932224f99e9 100644 --- a/fs/btrfs/Makefile +++ b/fs/btrfs/Makefile @@ -8,7 +8,7 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \ extent_io.o volumes.o async-thread.o ioctl.o locking.o orphan.o \ export.o tree-log.o free-space-cache.o zlib.o lzo.o \ compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o \ - reada.o backref.o ulist.o qgroup.o send.o dev-replace.o + reada.o backref.o ulist.o qgroup.o send.o dev-replace.o raid56.o btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 0ab51be6879f..0cce3aafbd62 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -502,6 +502,7 @@ struct btrfs_super_block { #define BTRFS_FEATURE_INCOMPAT_BIG_METADATA (1ULL << 5) #define BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF (1ULL << 6) +#define BTRFS_FEATURE_INCOMPAT_RAID56 (1ULL << 7) #define BTRFS_FEATURE_COMPAT_SUPP 0ULL #define BTRFS_FEATURE_COMPAT_RO_SUPP 0ULL @@ -511,6 +512,7 @@ struct btrfs_super_block { BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS | \ BTRFS_FEATURE_INCOMPAT_BIG_METADATA | \ BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO | \ + BTRFS_FEATURE_INCOMPAT_RAID56 | \ BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF) /* @@ -952,8 +954,10 @@ struct btrfs_dev_replace_item { #define BTRFS_BLOCK_GROUP_RAID1 (1ULL << 4) #define BTRFS_BLOCK_GROUP_DUP (1ULL << 5) #define BTRFS_BLOCK_GROUP_RAID10 (1ULL << 6) +#define BTRFS_BLOCK_GROUP_RAID5 (1 << 7) +#define BTRFS_BLOCK_GROUP_RAID6 (1 << 8) #define BTRFS_BLOCK_GROUP_RESERVED BTRFS_AVAIL_ALLOC_BIT_SINGLE -#define BTRFS_NR_RAID_TYPES 5 +#define BTRFS_NR_RAID_TYPES 7 #define BTRFS_BLOCK_GROUP_TYPE_MASK (BTRFS_BLOCK_GROUP_DATA | \ BTRFS_BLOCK_GROUP_SYSTEM | \ @@ -961,6 +965,8 @@ struct btrfs_dev_replace_item { #define BTRFS_BLOCK_GROUP_PROFILE_MASK (BTRFS_BLOCK_GROUP_RAID0 | \ BTRFS_BLOCK_GROUP_RAID1 | \ + BTRFS_BLOCK_GROUP_RAID5 | \ + BTRFS_BLOCK_GROUP_RAID6 | \ BTRFS_BLOCK_GROUP_DUP | \ BTRFS_BLOCK_GROUP_RAID10) /* @@ -1185,6 +1191,10 @@ struct btrfs_block_group_cache { u64 flags; u64 sectorsize; u64 cache_generation; + + /* for raid56, this is a full stripe, without parity */ + unsigned long full_stripe_len; + unsigned int ro:1; unsigned int dirty:1; unsigned int iref:1; @@ -1225,6 +1235,20 @@ struct seq_list { u64 seq; }; +/* used by the raid56 code to lock stripes for read/modify/write */ +struct btrfs_stripe_hash { + struct list_head hash_list; + wait_queue_head_t wait; + spinlock_t lock; +}; + +/* used by the raid56 code to lock stripes for read/modify/write */ +struct btrfs_stripe_hash_table { + struct btrfs_stripe_hash *table; +}; + +#define BTRFS_STRIPE_HASH_TABLE_BITS 11 + /* fs_info */ struct reloc_control; struct btrfs_device; @@ -1307,6 +1331,13 @@ struct btrfs_fs_info { struct mutex cleaner_mutex; struct mutex chunk_mutex; struct mutex volume_mutex; + + /* this is used during read/modify/write to make sure + * no two ios are trying to mod the same stripe at the same + * time + */ + struct btrfs_stripe_hash_table *stripe_hash_table; + /* * this protects the ordered operations list only while we are * processing all of the entries on it. This way we make @@ -1395,6 +1426,8 @@ struct btrfs_fs_info { struct btrfs_workers flush_workers; struct btrfs_workers endio_workers; struct btrfs_workers endio_meta_workers; + struct btrfs_workers endio_raid56_workers; + struct btrfs_workers rmw_workers; struct btrfs_workers endio_meta_write_workers; struct btrfs_workers endio_write_workers; struct btrfs_workers endio_freespace_worker; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 65f03670a952..e9fa7b4d18e3 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -46,6 +46,7 @@ #include "check-integrity.h" #include "rcu-string.h" #include "dev-replace.h" +#include "raid56.h" #ifdef CONFIG_X86 #include @@ -639,8 +640,15 @@ err: btree_readahead_hook(root, eb, eb->start, ret); } - if (ret) + if (ret) { + /* + * our io error hook is going to dec the io pages + * again, we have to make sure it has something + * to decrement + */ + atomic_inc(&eb->io_pages); clear_extent_buffer_uptodate(eb); + } free_extent_buffer(eb); out: return ret; @@ -654,6 +662,7 @@ static int btree_io_failed_hook(struct page *page, int failed_mirror) eb = (struct extent_buffer *)page->private; set_bit(EXTENT_BUFFER_IOERR, &eb->bflags); eb->read_mirror = failed_mirror; + atomic_dec(&eb->io_pages); if (test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags)) btree_readahead_hook(root, eb, eb->start, -EIO); return -EIO; /* we fixed nothing */ @@ -670,17 +679,23 @@ static void end_workqueue_bio(struct bio *bio, int err) end_io_wq->work.flags = 0; if (bio->bi_rw & REQ_WRITE) { - if (end_io_wq->metadata == 1) + if (end_io_wq->metadata == BTRFS_WQ_ENDIO_METADATA) btrfs_queue_worker(&fs_info->endio_meta_write_workers, &end_io_wq->work); - else if (end_io_wq->metadata == 2) + else if (end_io_wq->metadata == BTRFS_WQ_ENDIO_FREE_SPACE) btrfs_queue_worker(&fs_info->endio_freespace_worker, &end_io_wq->work); + else if (end_io_wq->metadata == BTRFS_WQ_ENDIO_RAID56) + btrfs_queue_worker(&fs_info->endio_raid56_workers, + &end_io_wq->work); else btrfs_queue_worker(&fs_info->endio_write_workers, &end_io_wq->work); } else { - if (end_io_wq->metadata) + if (end_io_wq->metadata == BTRFS_WQ_ENDIO_RAID56) + btrfs_queue_worker(&fs_info->endio_raid56_workers, + &end_io_wq->work); + else if (end_io_wq->metadata) btrfs_queue_worker(&fs_info->endio_meta_workers, &end_io_wq->work); else @@ -695,6 +710,7 @@ static void end_workqueue_bio(struct bio *bio, int err) * 0 - if data * 1 - if normal metadta * 2 - if writing to the free space cache area + * 3 - raid parity work */ int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio, int metadata) @@ -2165,6 +2181,12 @@ int open_ctree(struct super_block *sb, init_waitqueue_head(&fs_info->transaction_blocked_wait); init_waitqueue_head(&fs_info->async_submit_wait); + ret = btrfs_alloc_stripe_hash_table(fs_info); + if (ret) { + err = -ENOMEM; + goto fail_alloc; + } + __setup_root(4096, 4096, 4096, 4096, tree_root, fs_info, BTRFS_ROOT_TREE_OBJECTID); @@ -2332,6 +2354,12 @@ int open_ctree(struct super_block *sb, btrfs_init_workers(&fs_info->endio_meta_write_workers, "endio-meta-write", fs_info->thread_pool_size, &fs_info->generic_worker); + btrfs_init_workers(&fs_info->endio_raid56_workers, + "endio-raid56", fs_info->thread_pool_size, + &fs_info->generic_worker); + btrfs_init_workers(&fs_info->rmw_workers, + "rmw", fs_info->thread_pool_size, + &fs_info->generic_worker); btrfs_init_workers(&fs_info->endio_write_workers, "endio-write", fs_info->thread_pool_size, &fs_info->generic_worker); @@ -2350,6 +2378,8 @@ int open_ctree(struct super_block *sb, */ fs_info->endio_workers.idle_thresh = 4; fs_info->endio_meta_workers.idle_thresh = 4; + fs_info->endio_raid56_workers.idle_thresh = 4; + fs_info->rmw_workers.idle_thresh = 2; fs_info->endio_write_workers.idle_thresh = 2; fs_info->endio_meta_write_workers.idle_thresh = 2; @@ -2366,6 +2396,8 @@ int open_ctree(struct super_block *sb, ret |= btrfs_start_workers(&fs_info->fixup_workers); ret |= btrfs_start_workers(&fs_info->endio_workers); ret |= btrfs_start_workers(&fs_info->endio_meta_workers); + ret |= btrfs_start_workers(&fs_info->rmw_workers); + ret |= btrfs_start_workers(&fs_info->endio_raid56_workers); ret |= btrfs_start_workers(&fs_info->endio_meta_write_workers); ret |= btrfs_start_workers(&fs_info->endio_write_workers); ret |= btrfs_start_workers(&fs_info->endio_freespace_worker); @@ -2710,6 +2742,8 @@ fail_sb_buffer: btrfs_stop_workers(&fs_info->workers); btrfs_stop_workers(&fs_info->endio_workers); btrfs_stop_workers(&fs_info->endio_meta_workers); + btrfs_stop_workers(&fs_info->endio_raid56_workers); + btrfs_stop_workers(&fs_info->rmw_workers); btrfs_stop_workers(&fs_info->endio_meta_write_workers); btrfs_stop_workers(&fs_info->endio_write_workers); btrfs_stop_workers(&fs_info->endio_freespace_worker); @@ -2728,6 +2762,7 @@ fail_bdi: fail_srcu: cleanup_srcu_struct(&fs_info->subvol_srcu); fail: + btrfs_free_stripe_hash_table(fs_info); btrfs_close_devices(fs_info->fs_devices); return err; @@ -3076,11 +3111,16 @@ int btrfs_calc_num_tolerated_disk_barrier_failures( ((flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) == 0))) num_tolerated_disk_barrier_failures = 0; - else if (num_tolerated_disk_barrier_failures > 1 - && - (flags & (BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_RAID10))) - num_tolerated_disk_barrier_failures = 1; + else if (num_tolerated_disk_barrier_failures > 1) { + if (flags & (BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID10)) { + num_tolerated_disk_barrier_failures = 1; + } else if (flags & + BTRFS_BLOCK_GROUP_RAID5) { + num_tolerated_disk_barrier_failures = 2; + } + } } } up_read(&sinfo->groups_sem); @@ -3384,6 +3424,8 @@ int close_ctree(struct btrfs_root *root) btrfs_stop_workers(&fs_info->workers); btrfs_stop_workers(&fs_info->endio_workers); btrfs_stop_workers(&fs_info->endio_meta_workers); + btrfs_stop_workers(&fs_info->endio_raid56_workers); + btrfs_stop_workers(&fs_info->rmw_workers); btrfs_stop_workers(&fs_info->endio_meta_write_workers); btrfs_stop_workers(&fs_info->endio_write_workers); btrfs_stop_workers(&fs_info->endio_freespace_worker); @@ -3404,6 +3446,8 @@ int close_ctree(struct btrfs_root *root) bdi_destroy(&fs_info->bdi); cleanup_srcu_struct(&fs_info->subvol_srcu); + btrfs_free_stripe_hash_table(fs_info); + return 0; } diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 305c33efb0e3..034d7dc552b2 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -25,6 +25,13 @@ #define BTRFS_SUPER_MIRROR_MAX 3 #define BTRFS_SUPER_MIRROR_SHIFT 12 +enum { + BTRFS_WQ_ENDIO_DATA = 0, + BTRFS_WQ_ENDIO_METADATA = 1, + BTRFS_WQ_ENDIO_FREE_SPACE = 2, + BTRFS_WQ_ENDIO_RAID56 = 3, +}; + static inline u64 btrfs_sb_offset(int mirror) { u64 start = 16 * 1024; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index d133edfcd449..3345f68fc64b 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -31,6 +31,7 @@ #include "print-tree.h" #include "transaction.h" #include "volumes.h" +#include "raid56.h" #include "locking.h" #include "free-space-cache.h" #include "math.h" @@ -1852,6 +1853,8 @@ static int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr, *actual_bytes = discarded_bytes; + if (ret == -EOPNOTSUPP) + ret = 0; return ret; } @@ -3276,6 +3279,7 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) u64 num_devices = root->fs_info->fs_devices->rw_devices + root->fs_info->fs_devices->missing_devices; u64 target; + u64 tmp; /* * see if restripe for this chunk_type is in progress, if so @@ -3292,30 +3296,32 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) } spin_unlock(&root->fs_info->balance_lock); + /* First, mask out the RAID levels which aren't possible */ if (num_devices == 1) - flags &= ~(BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID0); + flags &= ~(BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID0 | + BTRFS_BLOCK_GROUP_RAID5); + if (num_devices < 3) + flags &= ~BTRFS_BLOCK_GROUP_RAID6; if (num_devices < 4) flags &= ~BTRFS_BLOCK_GROUP_RAID10; - if ((flags & BTRFS_BLOCK_GROUP_DUP) && - (flags & (BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_RAID10))) { - flags &= ~BTRFS_BLOCK_GROUP_DUP; - } - - if ((flags & BTRFS_BLOCK_GROUP_RAID1) && - (flags & BTRFS_BLOCK_GROUP_RAID10)) { - flags &= ~BTRFS_BLOCK_GROUP_RAID1; - } + tmp = flags & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID0 | + BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6 | BTRFS_BLOCK_GROUP_RAID10); + flags &= ~tmp; - if ((flags & BTRFS_BLOCK_GROUP_RAID0) && - ((flags & BTRFS_BLOCK_GROUP_RAID1) | - (flags & BTRFS_BLOCK_GROUP_RAID10) | - (flags & BTRFS_BLOCK_GROUP_DUP))) { - flags &= ~BTRFS_BLOCK_GROUP_RAID0; - } + if (tmp & BTRFS_BLOCK_GROUP_RAID6) + tmp = BTRFS_BLOCK_GROUP_RAID6; + else if (tmp & BTRFS_BLOCK_GROUP_RAID5) + tmp = BTRFS_BLOCK_GROUP_RAID5; + else if (tmp & BTRFS_BLOCK_GROUP_RAID10) + tmp = BTRFS_BLOCK_GROUP_RAID10; + else if (tmp & BTRFS_BLOCK_GROUP_RAID1) + tmp = BTRFS_BLOCK_GROUP_RAID1; + else if (tmp & BTRFS_BLOCK_GROUP_RAID0) + tmp = BTRFS_BLOCK_GROUP_RAID0; - return extended_to_chunk(flags); + return extended_to_chunk(flags | tmp); } static u64 get_alloc_profile(struct btrfs_root *root, u64 flags) @@ -3333,6 +3339,7 @@ static u64 get_alloc_profile(struct btrfs_root *root, u64 flags) u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data) { u64 flags; + u64 ret; if (data) flags = BTRFS_BLOCK_GROUP_DATA; @@ -3341,7 +3348,8 @@ u64 btrfs_get_alloc_profile(struct btrfs_root *root, int data) else flags = BTRFS_BLOCK_GROUP_METADATA; - return get_alloc_profile(root, flags); + ret = get_alloc_profile(root, flags); + return ret; } /* @@ -3516,8 +3524,10 @@ static u64 get_system_chunk_thresh(struct btrfs_root *root, u64 type) { u64 num_dev; - if (type & BTRFS_BLOCK_GROUP_RAID10 || - type & BTRFS_BLOCK_GROUP_RAID0) + if (type & (BTRFS_BLOCK_GROUP_RAID10 | + BTRFS_BLOCK_GROUP_RAID0 | + BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6)) num_dev = root->fs_info->fs_devices->rw_devices; else if (type & BTRFS_BLOCK_GROUP_RAID1) num_dev = 2; @@ -3667,7 +3677,9 @@ static int can_overcommit(struct btrfs_root *root, /* * If we have dup, raid1 or raid10 then only half of the free - * space is actually useable. + * space is actually useable. For raid56, the space info used + * doesn't include the parity drive, so we don't have to + * change the math */ if (profile & (BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 | @@ -5455,10 +5467,14 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, return ret; } -static u64 stripe_align(struct btrfs_root *root, u64 val) +static u64 stripe_align(struct btrfs_root *root, + struct btrfs_block_group_cache *cache, + u64 val, u64 num_bytes) { - u64 mask = ((u64)root->stripesize - 1); - u64 ret = (val + mask) & ~mask; + u64 mask; + u64 ret; + mask = ((u64)root->stripesize - 1); + ret = (val + mask) & ~mask; return ret; } @@ -5519,9 +5535,12 @@ int __get_raid_index(u64 flags) index = 2; else if (flags & BTRFS_BLOCK_GROUP_RAID0) index = 3; + else if (flags & BTRFS_BLOCK_GROUP_RAID5) + index = 5; + else if (flags & BTRFS_BLOCK_GROUP_RAID6) + index = 6; else - index = 4; - + index = 4; /* BTRFS_BLOCK_GROUP_SINGLE */ return index; } @@ -5665,6 +5684,8 @@ search: if (!block_group_bits(block_group, data)) { u64 extra = BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6 | BTRFS_BLOCK_GROUP_RAID10; /* @@ -5835,7 +5856,8 @@ unclustered_alloc: goto loop; } checks: - search_start = stripe_align(root, offset); + search_start = stripe_align(root, used_block_group, + offset, num_bytes); /* move on to the next group */ if (search_start + num_bytes > @@ -7203,6 +7225,7 @@ static u64 update_block_group_flags(struct btrfs_root *root, u64 flags) root->fs_info->fs_devices->missing_devices; stripped = BTRFS_BLOCK_GROUP_RAID0 | + BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6 | BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10; if (num_devices == 1) { @@ -7754,7 +7777,9 @@ int btrfs_read_block_groups(struct btrfs_root *root) btrfs_release_path(path); cache->flags = btrfs_block_group_flags(&cache->item); cache->sectorsize = root->sectorsize; - + cache->full_stripe_len = btrfs_full_stripe_len(root, + &root->fs_info->mapping_tree, + found_key.objectid); btrfs_init_free_space_ctl(cache); /* @@ -7808,6 +7833,8 @@ int btrfs_read_block_groups(struct btrfs_root *root) if (!(get_alloc_profile(root, space_info->flags) & (BTRFS_BLOCK_GROUP_RAID10 | BTRFS_BLOCK_GROUP_RAID1 | + BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6 | BTRFS_BLOCK_GROUP_DUP))) continue; /* @@ -7883,6 +7910,9 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY; cache->sectorsize = root->sectorsize; cache->fs_info = root->fs_info; + cache->full_stripe_len = btrfs_full_stripe_len(root, + &root->fs_info->mapping_tree, + chunk_offset); atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 86ecca48c604..3b9fb478b0d1 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -1895,13 +1895,11 @@ static int free_io_failure(struct inode *inode, struct io_failure_record *rec, if (ret) err = ret; - if (did_repair) { - ret = clear_extent_bits(&BTRFS_I(inode)->io_tree, rec->start, - rec->start + rec->len - 1, - EXTENT_DAMAGED, GFP_NOFS); - if (ret && !err) - err = ret; - } + ret = clear_extent_bits(&BTRFS_I(inode)->io_tree, rec->start, + rec->start + rec->len - 1, + EXTENT_DAMAGED, GFP_NOFS); + if (ret && !err) + err = ret; kfree(rec); return err; @@ -1932,10 +1930,15 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start, u64 map_length = 0; u64 sector; struct btrfs_bio *bbio = NULL; + struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; int ret; BUG_ON(!mirror_num); + /* we can't repair anything in raid56 yet */ + if (btrfs_is_parity_mirror(map_tree, logical, length, mirror_num)) + return 0; + bio = bio_alloc(GFP_NOFS, 1); if (!bio) return -EIO; @@ -2052,6 +2055,7 @@ static int clean_io_failure(u64 start, struct page *page) failrec->failed_mirror); did_repair = !ret; } + ret = 0; } out: diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 59ea2e4349c9..62020b7f7036 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -1463,10 +1463,14 @@ static int search_bitmap(struct btrfs_free_space_ctl *ctl, } static struct btrfs_free_space * -find_free_space(struct btrfs_free_space_ctl *ctl, u64 *offset, u64 *bytes) +find_free_space(struct btrfs_free_space_ctl *ctl, u64 *offset, u64 *bytes, + unsigned long align) { struct btrfs_free_space *entry; struct rb_node *node; + u64 ctl_off; + u64 tmp; + u64 align_off; int ret; if (!ctl->free_space_offset.rb_node) @@ -1481,15 +1485,34 @@ find_free_space(struct btrfs_free_space_ctl *ctl, u64 *offset, u64 *bytes) if (entry->bytes < *bytes) continue; + /* make sure the space returned is big enough + * to match our requested alignment + */ + if (*bytes >= align) { + ctl_off = entry->offset - ctl->start; + tmp = ctl_off + align - 1;; + do_div(tmp, align); + tmp = tmp * align + ctl->start; + align_off = tmp - entry->offset; + } else { + align_off = 0; + tmp = entry->offset; + } + + if (entry->bytes < *bytes + align_off) + continue; + if (entry->bitmap) { - ret = search_bitmap(ctl, entry, offset, bytes); - if (!ret) + ret = search_bitmap(ctl, entry, &tmp, bytes); + if (!ret) { + *offset = tmp; return entry; + } continue; } - *offset = entry->offset; - *bytes = entry->bytes; + *offset = tmp; + *bytes = entry->bytes - align_off; return entry; } @@ -2091,9 +2114,12 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, struct btrfs_free_space *entry = NULL; u64 bytes_search = bytes + empty_size; u64 ret = 0; + u64 align_gap = 0; + u64 align_gap_len = 0; spin_lock(&ctl->tree_lock); - entry = find_free_space(ctl, &offset, &bytes_search); + entry = find_free_space(ctl, &offset, &bytes_search, + block_group->full_stripe_len); if (!entry) goto out; @@ -2103,9 +2129,15 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, if (!entry->bytes) free_bitmap(ctl, entry); } else { + unlink_free_space(ctl, entry); - entry->offset += bytes; - entry->bytes -= bytes; + align_gap_len = offset - entry->offset; + align_gap = entry->offset; + + entry->offset = offset + bytes; + WARN_ON(entry->bytes < bytes + align_gap_len); + + entry->bytes -= bytes + align_gap_len; if (!entry->bytes) kmem_cache_free(btrfs_free_space_cachep, entry); else @@ -2115,6 +2147,8 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, out: spin_unlock(&ctl->tree_lock); + if (align_gap_len) + __btrfs_add_free_space(ctl, align_gap, align_gap_len); return ret; } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 1b98c4ce3c6f..6f4e41dca970 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -39,6 +39,7 @@ #include #include #include +#include #include "compat.h" #include "ctree.h" #include "disk-io.h" @@ -6386,19 +6387,24 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, int async_submit = 0; map_length = orig_bio->bi_size; - ret = btrfs_map_block(root->fs_info, READ, start_sector << 9, + ret = btrfs_map_block(root->fs_info, rw, start_sector << 9, &map_length, NULL, 0); if (ret) { bio_put(orig_bio); return -EIO; } - if (map_length >= orig_bio->bi_size) { bio = orig_bio; goto submit; } - async_submit = 1; + /* async crcs make it difficult to collect full stripe writes. */ + if (btrfs_get_alloc_profile(root, 1) & + (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) + async_submit = 0; + else + async_submit = 1; + bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS); if (!bio) return -ENOMEM; @@ -6440,7 +6446,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, bio->bi_end_io = btrfs_end_dio_bio; map_length = orig_bio->bi_size; - ret = btrfs_map_block(root->fs_info, READ, + ret = btrfs_map_block(root->fs_info, rw, start_sector << 9, &map_length, NULL, 0); if (ret) { @@ -6583,15 +6589,17 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; + ssize_t ret; if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iov, offset, nr_segs)) return 0; - return __blockdev_direct_IO(rw, iocb, inode, + ret = __blockdev_direct_IO(rw, iocb, inode, BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, iov, offset, nr_segs, btrfs_get_blocks_direct, NULL, btrfs_submit_direct, 0); + return ret; } #define BTRFS_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c new file mode 100644 index 000000000000..d02510f34936 --- /dev/null +++ b/fs/btrfs/raid56.c @@ -0,0 +1,1647 @@ +/* + * Copyright (C) 2012 Fusion-io All rights reserved. + * Copyright (C) 2012 Intel Corp. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License v2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public + * License along with this program; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 021110-1307, USA. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "compat.h" +#include "ctree.h" +#include "extent_map.h" +#include "disk-io.h" +#include "transaction.h" +#include "print-tree.h" +#include "volumes.h" +#include "raid56.h" +#include "async-thread.h" +#include "check-integrity.h" +#include "rcu-string.h" + +/* set when additional merges to this rbio are not allowed */ +#define RBIO_RMW_LOCKED_BIT 1 + +struct btrfs_raid_bio { + struct btrfs_fs_info *fs_info; + struct btrfs_bio *bbio; + + /* + * logical block numbers for the start of each stripe + * The last one or two are p/q. These are sorted, + * so raid_map[0] is the start of our full stripe + */ + u64 *raid_map; + + /* while we're doing rmw on a stripe + * we put it into a hash table so we can + * lock the stripe and merge more rbios + * into it. + */ + struct list_head hash_list; + + /* + * for scheduling work in the helper threads + */ + struct btrfs_work work; + + /* + * bio list and bio_list_lock are used + * to add more bios into the stripe + * in hopes of avoiding the full rmw + */ + struct bio_list bio_list; + spinlock_t bio_list_lock; + + /* + * also protected by the bio_list_lock, the + * stripe locking code uses plug_list to hand off + * the stripe lock to the next pending IO + */ + struct list_head plug_list; + + /* + * flags that tell us if it is safe to + * merge with this bio + */ + unsigned long flags; + + /* size of each individual stripe on disk */ + int stripe_len; + + /* number of data stripes (no p/q) */ + int nr_data; + + /* + * set if we're doing a parity rebuild + * for a read from higher up, which is handled + * differently from a parity rebuild as part of + * rmw + */ + int read_rebuild; + + /* first bad stripe */ + int faila; + + /* second bad stripe (for raid6 use) */ + int failb; + + /* + * number of pages needed to represent the full + * stripe + */ + int nr_pages; + + /* + * size of all the bios in the bio_list. This + * helps us decide if the rbio maps to a full + * stripe or not + */ + int bio_list_bytes; + + atomic_t refs; + + /* + * these are two arrays of pointers. We allocate the + * rbio big enough to hold them both and setup their + * locations when the rbio is allocated + */ + + /* pointers to pages that we allocated for + * reading/writing stripes directly from the disk (including P/Q) + */ + struct page **stripe_pages; + + /* + * pointers to the pages in the bio_list. Stored + * here for faster lookup + */ + struct page **bio_pages; +}; + +static int __raid56_parity_recover(struct btrfs_raid_bio *rbio); +static noinline void finish_rmw(struct btrfs_raid_bio *rbio); +static void rmw_work(struct btrfs_work *work); +static void read_rebuild_work(struct btrfs_work *work); +static void async_rmw_stripe(struct btrfs_raid_bio *rbio); +static void async_read_rebuild(struct btrfs_raid_bio *rbio); +static int fail_bio_stripe(struct btrfs_raid_bio *rbio, struct bio *bio); +static int fail_rbio_index(struct btrfs_raid_bio *rbio, int failed); +static void __free_raid_bio(struct btrfs_raid_bio *rbio); +static void index_rbio_pages(struct btrfs_raid_bio *rbio); +static int alloc_rbio_pages(struct btrfs_raid_bio *rbio); + +/* + * the stripe hash table is used for locking, and to collect + * bios in hopes of making a full stripe + */ +int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info) +{ + struct btrfs_stripe_hash_table *table; + struct btrfs_stripe_hash_table *x; + struct btrfs_stripe_hash *cur; + struct btrfs_stripe_hash *h; + int num_entries = 1 << BTRFS_STRIPE_HASH_TABLE_BITS; + int i; + + if (info->stripe_hash_table) + return 0; + + table = kzalloc(sizeof(*table) + sizeof(*h) * num_entries, GFP_NOFS); + if (!table) + return -ENOMEM; + + table->table = (void *)(table + 1); + h = table->table; + + for (i = 0; i < num_entries; i++) { + cur = h + i; + INIT_LIST_HEAD(&cur->hash_list); + spin_lock_init(&cur->lock); + init_waitqueue_head(&cur->wait); + } + + x = cmpxchg(&info->stripe_hash_table, NULL, table); + if (x) + kfree(x); + return 0; +} + +/* + * we hash on the first logical address of the stripe + */ +static int rbio_bucket(struct btrfs_raid_bio *rbio) +{ + u64 num = rbio->raid_map[0]; + + /* + * we shift down quite a bit. We're using byte + * addressing, and most of the lower bits are zeros. + * This tends to upset hash_64, and it consistently + * returns just one or two different values. + * + * shifting off the lower bits fixes things. + */ + return hash_64(num >> 16, BTRFS_STRIPE_HASH_TABLE_BITS); +} + +/* + * merging means we take the bio_list from the victim and + * splice it into the destination. The victim should + * be discarded afterwards. + * + * must be called with dest->rbio_list_lock held + */ +static void merge_rbio(struct btrfs_raid_bio *dest, + struct btrfs_raid_bio *victim) +{ + bio_list_merge(&dest->bio_list, &victim->bio_list); + dest->bio_list_bytes += victim->bio_list_bytes; + bio_list_init(&victim->bio_list); +} + +/* + * free the hash table used by unmount + */ +void btrfs_free_stripe_hash_table(struct btrfs_fs_info *info) +{ + if (!info->stripe_hash_table) + return; + kfree(info->stripe_hash_table); + info->stripe_hash_table = NULL; +} + +/* + * helper function to run the xor_blocks api. It is only + * able to do MAX_XOR_BLOCKS at a time, so we need to + * loop through. + */ +static void run_xor(void **pages, int src_cnt, ssize_t len) +{ + int src_off = 0; + int xor_src_cnt = 0; + void *dest = pages[src_cnt]; + + while(src_cnt > 0) { + xor_src_cnt = min(src_cnt, MAX_XOR_BLOCKS); + xor_blocks(xor_src_cnt, len, dest, pages + src_off); + + src_cnt -= xor_src_cnt; + src_off += xor_src_cnt; + } +} + +/* + * returns true if the bio list inside this rbio + * covers an entire stripe (no rmw required). + * Must be called with the bio list lock held, or + * at a time when you know it is impossible to add + * new bios into the list + */ +static int __rbio_is_full(struct btrfs_raid_bio *rbio) +{ + unsigned long size = rbio->bio_list_bytes; + int ret = 1; + + if (size != rbio->nr_data * rbio->stripe_len) + ret = 0; + + BUG_ON(size > rbio->nr_data * rbio->stripe_len); + return ret; +} + +static int rbio_is_full(struct btrfs_raid_bio *rbio) +{ + unsigned long flags; + int ret; + + spin_lock_irqsave(&rbio->bio_list_lock, flags); + ret = __rbio_is_full(rbio); + spin_unlock_irqrestore(&rbio->bio_list_lock, flags); + return ret; +} + +/* + * returns 1 if it is safe to merge two rbios together. + * The merging is safe if the two rbios correspond to + * the same stripe and if they are both going in the same + * direction (read vs write), and if neither one is + * locked for final IO + * + * The caller is responsible for locking such that + * rmw_locked is safe to test + */ +static int rbio_can_merge(struct btrfs_raid_bio *last, + struct btrfs_raid_bio *cur) +{ + if (test_bit(RBIO_RMW_LOCKED_BIT, &last->flags) || + test_bit(RBIO_RMW_LOCKED_BIT, &cur->flags)) + return 0; + + if (last->raid_map[0] != + cur->raid_map[0]) + return 0; + + /* reads can't merge with writes */ + if (last->read_rebuild != + cur->read_rebuild) { + return 0; + } + + return 1; +} + +/* + * helper to index into the pstripe + */ +static struct page *rbio_pstripe_page(struct btrfs_raid_bio *rbio, int index) +{ + index += (rbio->nr_data * rbio->stripe_len) >> PAGE_CACHE_SHIFT; + return rbio->stripe_pages[index]; +} + +/* + * helper to index into the qstripe, returns null + * if there is no qstripe + */ +static struct page *rbio_qstripe_page(struct btrfs_raid_bio *rbio, int index) +{ + if (rbio->nr_data + 1 == rbio->bbio->num_stripes) + return NULL; + + index += ((rbio->nr_data + 1) * rbio->stripe_len) >> + PAGE_CACHE_SHIFT; + return rbio->stripe_pages[index]; +} + +/* + * The first stripe in the table for a logical address + * has the lock. rbios are added in one of three ways: + * + * 1) Nobody has the stripe locked yet. The rbio is given + * the lock and 0 is returned. The caller must start the IO + * themselves. + * + * 2) Someone has the stripe locked, but we're able to merge + * with the lock owner. The rbio is freed and the IO will + * start automatically along with the existing rbio. 1 is returned. + * + * 3) Someone has the stripe locked, but we're not able to merge. + * The rbio is added to the lock owner's plug list, or merged into + * an rbio already on the plug list. When the lock owner unlocks, + * the next rbio on the list is run and the IO is started automatically. + * 1 is returned + * + * If we return 0, the caller still owns the rbio and must continue with + * IO submission. If we return 1, the caller must assume the rbio has + * already been freed. + */ +static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) +{ + int bucket = rbio_bucket(rbio); + struct btrfs_stripe_hash *h = rbio->fs_info->stripe_hash_table->table + bucket; + struct btrfs_raid_bio *cur; + struct btrfs_raid_bio *pending; + unsigned long flags; + DEFINE_WAIT(wait); + struct btrfs_raid_bio *freeit = NULL; + int ret = 0; + int walk = 0; + + spin_lock_irqsave(&h->lock, flags); + list_for_each_entry(cur, &h->hash_list, hash_list) { + walk++; + if (cur->raid_map[0] == rbio->raid_map[0]) { + spin_lock(&cur->bio_list_lock); + + /* can we merge into the lock owner? */ + if (rbio_can_merge(cur, rbio)) { + merge_rbio(cur, rbio); + spin_unlock(&cur->bio_list_lock); + freeit = rbio; + ret = 1; + goto out; + } + + /* + * we couldn't merge with the running + * rbio, see if we can merge with the + * pending ones. We don't have to + * check for rmw_locked because there + * is no way they are inside finish_rmw + * right now + */ + list_for_each_entry(pending, &cur->plug_list, + plug_list) { + if (rbio_can_merge(pending, rbio)) { + merge_rbio(pending, rbio); + spin_unlock(&cur->bio_list_lock); + freeit = rbio; + ret = 1; + goto out; + } + } + + /* no merging, put us on the tail of the plug list, + * our rbio will be started with the currently + * running rbio unlocks + */ + list_add_tail(&rbio->plug_list, &cur->plug_list); + spin_unlock(&cur->bio_list_lock); + ret = 1; + goto out; + } + } + + atomic_inc(&rbio->refs); + list_add(&rbio->hash_list, &h->hash_list); +out: + spin_unlock_irqrestore(&h->lock, flags); + if (freeit) + __free_raid_bio(freeit); + return ret; +} + +/* + * called as rmw or parity rebuild is completed. If the plug list has more + * rbios waiting for this stripe, the next one on the list will be started + */ +static noinline void unlock_stripe(struct btrfs_raid_bio *rbio) +{ + int bucket; + struct btrfs_stripe_hash *h; + unsigned long flags; + + bucket = rbio_bucket(rbio); + h = rbio->fs_info->stripe_hash_table->table + bucket; + + spin_lock_irqsave(&h->lock, flags); + spin_lock(&rbio->bio_list_lock); + + if (!list_empty(&rbio->hash_list)) { + + list_del_init(&rbio->hash_list); + atomic_dec(&rbio->refs); + + /* + * we use the plug list to hold all the rbios + * waiting for the chance to lock this stripe. + * hand the lock over to one of them. + */ + if (!list_empty(&rbio->plug_list)) { + struct btrfs_raid_bio *next; + struct list_head *head = rbio->plug_list.next; + + next = list_entry(head, struct btrfs_raid_bio, + plug_list); + + list_del_init(&rbio->plug_list); + + list_add(&next->hash_list, &h->hash_list); + atomic_inc(&next->refs); + spin_unlock(&rbio->bio_list_lock); + spin_unlock_irqrestore(&h->lock, flags); + + if (next->read_rebuild) + async_read_rebuild(next); + else + async_rmw_stripe(next); + + goto done_nolock; + + } else if (waitqueue_active(&h->wait)) { + spin_unlock(&rbio->bio_list_lock); + spin_unlock_irqrestore(&h->lock, flags); + wake_up(&h->wait); + goto done_nolock; + } + } + spin_unlock(&rbio->bio_list_lock); + spin_unlock_irqrestore(&h->lock, flags); + +done_nolock: + return; +} + +static void __free_raid_bio(struct btrfs_raid_bio *rbio) +{ + int i; + + WARN_ON(atomic_read(&rbio->refs) < 0); + if (!atomic_dec_and_test(&rbio->refs)) + return; + + WARN_ON(!list_empty(&rbio->hash_list)); + WARN_ON(!bio_list_empty(&rbio->bio_list)); + + for (i = 0; i < rbio->nr_pages; i++) { + if (rbio->stripe_pages[i]) { + __free_page(rbio->stripe_pages[i]); + rbio->stripe_pages[i] = NULL; + } + } + kfree(rbio->raid_map); + kfree(rbio->bbio); + kfree(rbio); +} + +static void free_raid_bio(struct btrfs_raid_bio *rbio) +{ + unlock_stripe(rbio); + __free_raid_bio(rbio); +} + +/* + * this frees the rbio and runs through all the bios in the + * bio_list and calls end_io on them + */ +static void rbio_orig_end_io(struct btrfs_raid_bio *rbio, int err, int uptodate) +{ + struct bio *cur = bio_list_get(&rbio->bio_list); + struct bio *next; + free_raid_bio(rbio); + + while (cur) { + next = cur->bi_next; + cur->bi_next = NULL; + if (uptodate) + set_bit(BIO_UPTODATE, &cur->bi_flags); + bio_endio(cur, err); + cur = next; + } +} + +/* + * end io function used by finish_rmw. When we finally + * get here, we've written a full stripe + */ +static void raid_write_end_io(struct bio *bio, int err) +{ + struct btrfs_raid_bio *rbio = bio->bi_private; + + if (err) + fail_bio_stripe(rbio, bio); + + bio_put(bio); + + if (!atomic_dec_and_test(&rbio->bbio->stripes_pending)) + return; + + err = 0; + + /* OK, we have read all the stripes we need to. */ + if (atomic_read(&rbio->bbio->error) > rbio->bbio->max_errors) + err = -EIO; + + rbio_orig_end_io(rbio, err, 0); + return; +} + +/* + * the read/modify/write code wants to use the original bio for + * any pages it included, and then use the rbio for everything + * else. This function decides if a given index (stripe number) + * and page number in that stripe fall inside the original bio + * or the rbio. + * + * if you set bio_list_only, you'll get a NULL back for any ranges + * that are outside the bio_list + * + * This doesn't take any refs on anything, you get a bare page pointer + * and the caller must bump refs as required. + * + * You must call index_rbio_pages once before you can trust + * the answers from this function. + */ +static struct page *page_in_rbio(struct btrfs_raid_bio *rbio, + int index, int pagenr, int bio_list_only) +{ + int chunk_page; + struct page *p = NULL; + + chunk_page = index * (rbio->stripe_len >> PAGE_SHIFT) + pagenr; + + spin_lock_irq(&rbio->bio_list_lock); + p = rbio->bio_pages[chunk_page]; + spin_unlock_irq(&rbio->bio_list_lock); + + if (p || bio_list_only) + return p; + + return rbio->stripe_pages[chunk_page]; +} + +/* + * number of pages we need for the entire stripe across all the + * drives + */ +static unsigned long rbio_nr_pages(unsigned long stripe_len, int nr_stripes) +{ + unsigned long nr = stripe_len * nr_stripes; + return (nr + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; +} + +/* + * allocation and initial setup for the btrfs_raid_bio. Not + * this does not allocate any pages for rbio->pages. + */ +static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, + struct btrfs_bio *bbio, u64 *raid_map, + u64 stripe_len) +{ + struct btrfs_raid_bio *rbio; + int nr_data = 0; + int num_pages = rbio_nr_pages(stripe_len, bbio->num_stripes); + void *p; + + rbio = kzalloc(sizeof(*rbio) + num_pages * sizeof(struct page *) * 2, + GFP_NOFS); + if (!rbio) { + kfree(raid_map); + kfree(bbio); + return ERR_PTR(-ENOMEM); + } + + bio_list_init(&rbio->bio_list); + INIT_LIST_HEAD(&rbio->plug_list); + spin_lock_init(&rbio->bio_list_lock); + INIT_LIST_HEAD(&rbio->hash_list); + rbio->bbio = bbio; + rbio->raid_map = raid_map; + rbio->fs_info = root->fs_info; + rbio->stripe_len = stripe_len; + rbio->nr_pages = num_pages; + rbio->faila = -1; + rbio->failb = -1; + atomic_set(&rbio->refs, 1); + + /* + * the stripe_pages and bio_pages array point to the extra + * memory we allocated past the end of the rbio + */ + p = rbio + 1; + rbio->stripe_pages = p; + rbio->bio_pages = p + sizeof(struct page *) * num_pages; + + if (raid_map[bbio->num_stripes - 1] == RAID6_Q_STRIPE) + nr_data = bbio->num_stripes - 2; + else + nr_data = bbio->num_stripes - 1; + + rbio->nr_data = nr_data; + return rbio; +} + +/* allocate pages for all the stripes in the bio, including parity */ +static int alloc_rbio_pages(struct btrfs_raid_bio *rbio) +{ + int i; + struct page *page; + + for (i = 0; i < rbio->nr_pages; i++) { + if (rbio->stripe_pages[i]) + continue; + page = alloc_page(GFP_NOFS | __GFP_HIGHMEM); + if (!page) + return -ENOMEM; + rbio->stripe_pages[i] = page; + ClearPageUptodate(page); + } + return 0; +} + +/* allocate pages for just the p/q stripes */ +static int alloc_rbio_parity_pages(struct btrfs_raid_bio *rbio) +{ + int i; + struct page *page; + + i = (rbio->nr_data * rbio->stripe_len) >> PAGE_CACHE_SHIFT; + + for (; i < rbio->nr_pages; i++) { + if (rbio->stripe_pages[i]) + continue; + page = alloc_page(GFP_NOFS | __GFP_HIGHMEM); + if (!page) + return -ENOMEM; + rbio->stripe_pages[i] = page; + } + return 0; +} + +/* + * add a single page from a specific stripe into our list of bios for IO + * this will try to merge into existing bios if possible, and returns + * zero if all went well. + */ +int rbio_add_io_page(struct btrfs_raid_bio *rbio, + struct bio_list *bio_list, + struct page *page, + int stripe_nr, + unsigned long page_index, + unsigned long bio_max_len) +{ + struct bio *last = bio_list->tail; + u64 last_end = 0; + int ret; + struct bio *bio; + struct btrfs_bio_stripe *stripe; + u64 disk_start; + + stripe = &rbio->bbio->stripes[stripe_nr]; + disk_start = stripe->physical + (page_index << PAGE_CACHE_SHIFT); + + /* if the device is missing, just fail this stripe */ + if (!stripe->dev->bdev) + return fail_rbio_index(rbio, stripe_nr); + + /* see if we can add this page onto our existing bio */ + if (last) { + last_end = (u64)last->bi_sector << 9; + last_end += last->bi_size; + + /* + * we can't merge these if they are from different + * devices or if they are not contiguous + */ + if (last_end == disk_start && stripe->dev->bdev && + test_bit(BIO_UPTODATE, &last->bi_flags) && + last->bi_bdev == stripe->dev->bdev) { + ret = bio_add_page(last, page, PAGE_CACHE_SIZE, 0); + if (ret == PAGE_CACHE_SIZE) + return 0; + } + } + + /* put a new bio on the list */ + bio = bio_alloc(GFP_NOFS, bio_max_len >> PAGE_SHIFT?:1); + if (!bio) + return -ENOMEM; + + bio->bi_size = 0; + bio->bi_bdev = stripe->dev->bdev; + bio->bi_sector = disk_start >> 9; + set_bit(BIO_UPTODATE, &bio->bi_flags); + + bio_add_page(bio, page, PAGE_CACHE_SIZE, 0); + bio_list_add(bio_list, bio); + return 0; +} + +/* + * while we're doing the read/modify/write cycle, we could + * have errors in reading pages off the disk. This checks + * for errors and if we're not able to read the page it'll + * trigger parity reconstruction. The rmw will be finished + * after we've reconstructed the failed stripes + */ +static void validate_rbio_for_rmw(struct btrfs_raid_bio *rbio) +{ + if (rbio->faila >= 0 || rbio->failb >= 0) { + BUG_ON(rbio->faila == rbio->bbio->num_stripes - 1); + __raid56_parity_recover(rbio); + } else { + finish_rmw(rbio); + } +} + +/* + * these are just the pages from the rbio array, not from anything + * the FS sent down to us + */ +static struct page *rbio_stripe_page(struct btrfs_raid_bio *rbio, int stripe, int page) +{ + int index; + index = stripe * (rbio->stripe_len >> PAGE_CACHE_SHIFT); + index += page; + return rbio->stripe_pages[index]; +} + +/* + * helper function to walk our bio list and populate the bio_pages array with + * the result. This seems expensive, but it is faster than constantly + * searching through the bio list as we setup the IO in finish_rmw or stripe + * reconstruction. + * + * This must be called before you trust the answers from page_in_rbio + */ +static void index_rbio_pages(struct btrfs_raid_bio *rbio) +{ + struct bio *bio; + u64 start; + unsigned long stripe_offset; + unsigned long page_index; + struct page *p; + int i; + + spin_lock_irq(&rbio->bio_list_lock); + bio_list_for_each(bio, &rbio->bio_list) { + start = (u64)bio->bi_sector << 9; + stripe_offset = start - rbio->raid_map[0]; + page_index = stripe_offset >> PAGE_CACHE_SHIFT; + + for (i = 0; i < bio->bi_vcnt; i++) { + p = bio->bi_io_vec[i].bv_page; + rbio->bio_pages[page_index + i] = p; + } + } + spin_unlock_irq(&rbio->bio_list_lock); +} + +/* + * this is called from one of two situations. We either + * have a full stripe from the higher layers, or we've read all + * the missing bits off disk. + * + * This will calculate the parity and then send down any + * changed blocks. + */ +static noinline void finish_rmw(struct btrfs_raid_bio *rbio) +{ + struct btrfs_bio *bbio = rbio->bbio; + void *pointers[bbio->num_stripes]; + int stripe_len = rbio->stripe_len; + int nr_data = rbio->nr_data; + int stripe; + int pagenr; + int p_stripe = -1; + int q_stripe = -1; + struct bio_list bio_list; + struct bio *bio; + int pages_per_stripe = stripe_len >> PAGE_CACHE_SHIFT; + int ret; + + bio_list_init(&bio_list); + + if (bbio->num_stripes - rbio->nr_data == 1) { + p_stripe = bbio->num_stripes - 1; + } else if (bbio->num_stripes - rbio->nr_data == 2) { + p_stripe = bbio->num_stripes - 2; + q_stripe = bbio->num_stripes - 1; + } else { + BUG(); + } + + /* at this point we either have a full stripe, + * or we've read the full stripe from the drive. + * recalculate the parity and write the new results. + * + * We're not allowed to add any new bios to the + * bio list here, anyone else that wants to + * change this stripe needs to do their own rmw. + */ + spin_lock_irq(&rbio->bio_list_lock); + set_bit(RBIO_RMW_LOCKED_BIT, &rbio->flags); + spin_unlock_irq(&rbio->bio_list_lock); + + atomic_set(&rbio->bbio->error, 0); + + /* + * now that we've set rmw_locked, run through the + * bio list one last time and map the page pointers + */ + index_rbio_pages(rbio); + + for (pagenr = 0; pagenr < pages_per_stripe; pagenr++) { + struct page *p; + /* first collect one page from each data stripe */ + for (stripe = 0; stripe < nr_data; stripe++) { + p = page_in_rbio(rbio, stripe, pagenr, 0); + pointers[stripe] = kmap(p); + } + + /* then add the parity stripe */ + p = rbio_pstripe_page(rbio, pagenr); + SetPageUptodate(p); + pointers[stripe++] = kmap(p); + + if (q_stripe != -1) { + + /* + * raid6, add the qstripe and call the + * library function to fill in our p/q + */ + p = rbio_qstripe_page(rbio, pagenr); + SetPageUptodate(p); + pointers[stripe++] = kmap(p); + + raid6_call.gen_syndrome(bbio->num_stripes, PAGE_SIZE, + pointers); + } else { + /* raid5 */ + memcpy(pointers[nr_data], pointers[0], PAGE_SIZE); + run_xor(pointers + 1, nr_data - 1, PAGE_CACHE_SIZE); + } + + + for (stripe = 0; stripe < bbio->num_stripes; stripe++) + kunmap(page_in_rbio(rbio, stripe, pagenr, 0)); + } + + /* + * time to start writing. Make bios for everything from the + * higher layers (the bio_list in our rbio) and our p/q. Ignore + * everything else. + */ + for (stripe = 0; stripe < bbio->num_stripes; stripe++) { + for (pagenr = 0; pagenr < pages_per_stripe; pagenr++) { + struct page *page; + if (stripe < rbio->nr_data) { + page = page_in_rbio(rbio, stripe, pagenr, 1); + if (!page) + continue; + } else { + page = rbio_stripe_page(rbio, stripe, pagenr); + } + + ret = rbio_add_io_page(rbio, &bio_list, + page, stripe, pagenr, rbio->stripe_len); + if (ret) + goto cleanup; + } + } + + atomic_set(&bbio->stripes_pending, bio_list_size(&bio_list)); + BUG_ON(atomic_read(&bbio->stripes_pending) == 0); + + while (1) { + bio = bio_list_pop(&bio_list); + if (!bio) + break; + + bio->bi_private = rbio; + bio->bi_end_io = raid_write_end_io; + BUG_ON(!test_bit(BIO_UPTODATE, &bio->bi_flags)); + submit_bio(WRITE, bio); + } + return; + +cleanup: + rbio_orig_end_io(rbio, -EIO, 0); +} + +/* + * helper to find the stripe number for a given bio. Used to figure out which + * stripe has failed. This expects the bio to correspond to a physical disk, + * so it looks up based on physical sector numbers. + */ +static int find_bio_stripe(struct btrfs_raid_bio *rbio, + struct bio *bio) +{ + u64 physical = bio->bi_sector; + u64 stripe_start; + int i; + struct btrfs_bio_stripe *stripe; + + physical <<= 9; + + for (i = 0; i < rbio->bbio->num_stripes; i++) { + stripe = &rbio->bbio->stripes[i]; + stripe_start = stripe->physical; + if (physical >= stripe_start && + physical < stripe_start + rbio->stripe_len) { + return i; + } + } + return -1; +} + +/* + * helper to find the stripe number for a given + * bio (before mapping). Used to figure out which stripe has + * failed. This looks up based on logical block numbers. + */ +static int find_logical_bio_stripe(struct btrfs_raid_bio *rbio, + struct bio *bio) +{ + u64 logical = bio->bi_sector; + u64 stripe_start; + int i; + + logical <<= 9; + + for (i = 0; i < rbio->nr_data; i++) { + stripe_start = rbio->raid_map[i]; + if (logical >= stripe_start && + logical < stripe_start + rbio->stripe_len) { + return i; + } + } + return -1; +} + +/* + * returns -EIO if we had too many failures + */ +static int fail_rbio_index(struct btrfs_raid_bio *rbio, int failed) +{ + unsigned long flags; + int ret = 0; + + spin_lock_irqsave(&rbio->bio_list_lock, flags); + + /* we already know this stripe is bad, move on */ + if (rbio->faila == failed || rbio->failb == failed) + goto out; + + if (rbio->faila == -1) { + /* first failure on this rbio */ + rbio->faila = failed; + atomic_inc(&rbio->bbio->error); + } else if (rbio->failb == -1) { + /* second failure on this rbio */ + rbio->failb = failed; + atomic_inc(&rbio->bbio->error); + } else { + ret = -EIO; + } +out: + spin_unlock_irqrestore(&rbio->bio_list_lock, flags); + + return ret; +} + +/* + * helper to fail a stripe based on a physical disk + * bio. + */ +static int fail_bio_stripe(struct btrfs_raid_bio *rbio, + struct bio *bio) +{ + int failed = find_bio_stripe(rbio, bio); + + if (failed < 0) + return -EIO; + + return fail_rbio_index(rbio, failed); +} + +/* + * this sets each page in the bio uptodate. It should only be used on private + * rbio pages, nothing that comes in from the higher layers + */ +static void set_bio_pages_uptodate(struct bio *bio) +{ + int i; + struct page *p; + + for (i = 0; i < bio->bi_vcnt; i++) { + p = bio->bi_io_vec[i].bv_page; + SetPageUptodate(p); + } +} + +/* + * end io for the read phase of the rmw cycle. All the bios here are physical + * stripe bios we've read from the disk so we can recalculate the parity of the + * stripe. + * + * This will usually kick off finish_rmw once all the bios are read in, but it + * may trigger parity reconstruction if we had any errors along the way + */ +static void raid_rmw_end_io(struct bio *bio, int err) +{ + struct btrfs_raid_bio *rbio = bio->bi_private; + + if (err) + fail_bio_stripe(rbio, bio); + else + set_bio_pages_uptodate(bio); + + bio_put(bio); + + if (!atomic_dec_and_test(&rbio->bbio->stripes_pending)) + return; + + err = 0; + if (atomic_read(&rbio->bbio->error) > rbio->bbio->max_errors) + goto cleanup; + + /* + * this will normally call finish_rmw to start our write + * but if there are any failed stripes we'll reconstruct + * from parity first + */ + validate_rbio_for_rmw(rbio); + return; + +cleanup: + + rbio_orig_end_io(rbio, -EIO, 0); +} + +static void async_rmw_stripe(struct btrfs_raid_bio *rbio) +{ + rbio->work.flags = 0; + rbio->work.func = rmw_work; + + btrfs_queue_worker(&rbio->fs_info->rmw_workers, + &rbio->work); +} + +static void async_read_rebuild(struct btrfs_raid_bio *rbio) +{ + rbio->work.flags = 0; + rbio->work.func = read_rebuild_work; + + btrfs_queue_worker(&rbio->fs_info->rmw_workers, + &rbio->work); +} + +/* + * the stripe must be locked by the caller. It will + * unlock after all the writes are done + */ +static int raid56_rmw_stripe(struct btrfs_raid_bio *rbio) +{ + int bios_to_read = 0; + struct btrfs_bio *bbio = rbio->bbio; + struct bio_list bio_list; + int ret; + int nr_pages = (rbio->stripe_len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + int pagenr; + int stripe; + struct bio *bio; + + bio_list_init(&bio_list); + + ret = alloc_rbio_pages(rbio); + if (ret) + goto cleanup; + + index_rbio_pages(rbio); + + atomic_set(&rbio->bbio->error, 0); + /* + * build a list of bios to read all the missing parts of this + * stripe + */ + for (stripe = 0; stripe < rbio->nr_data; stripe++) { + for (pagenr = 0; pagenr < nr_pages; pagenr++) { + struct page *page; + /* + * we want to find all the pages missing from + * the rbio and read them from the disk. If + * page_in_rbio finds a page in the bio list + * we don't need to read it off the stripe. + */ + page = page_in_rbio(rbio, stripe, pagenr, 1); + if (page) + continue; + + page = rbio_stripe_page(rbio, stripe, pagenr); + ret = rbio_add_io_page(rbio, &bio_list, page, + stripe, pagenr, rbio->stripe_len); + if (ret) + goto cleanup; + } + } + + bios_to_read = bio_list_size(&bio_list); + if (!bios_to_read) { + /* + * this can happen if others have merged with + * us, it means there is nothing left to read. + * But if there are missing devices it may not be + * safe to do the full stripe write yet. + */ + goto finish; + } + + /* + * the bbio may be freed once we submit the last bio. Make sure + * not to touch it after that + */ + atomic_set(&bbio->stripes_pending, bios_to_read); + while (1) { + bio = bio_list_pop(&bio_list); + if (!bio) + break; + + bio->bi_private = rbio; + bio->bi_end_io = raid_rmw_end_io; + + btrfs_bio_wq_end_io(rbio->fs_info, bio, + BTRFS_WQ_ENDIO_RAID56); + + BUG_ON(!test_bit(BIO_UPTODATE, &bio->bi_flags)); + submit_bio(READ, bio); + } + /* the actual write will happen once the reads are done */ + return 0; + +cleanup: + rbio_orig_end_io(rbio, -EIO, 0); + return -EIO; + +finish: + validate_rbio_for_rmw(rbio); + return 0; +} + +/* + * if the upper layers pass in a full stripe, we thank them by only allocating + * enough pages to hold the parity, and sending it all down quickly. + */ +static int full_stripe_write(struct btrfs_raid_bio *rbio) +{ + int ret; + + ret = alloc_rbio_parity_pages(rbio); + if (ret) + return ret; + + ret = lock_stripe_add(rbio); + if (ret == 0) + finish_rmw(rbio); + return 0; +} + +/* + * partial stripe writes get handed over to async helpers. + * We're really hoping to merge a few more writes into this + * rbio before calculating new parity + */ +static int partial_stripe_write(struct btrfs_raid_bio *rbio) +{ + int ret; + + ret = lock_stripe_add(rbio); + if (ret == 0) + async_rmw_stripe(rbio); + return 0; +} + +/* + * sometimes while we were reading from the drive to + * recalculate parity, enough new bios come into create + * a full stripe. So we do a check here to see if we can + * go directly to finish_rmw + */ +static int __raid56_parity_write(struct btrfs_raid_bio *rbio) +{ + /* head off into rmw land if we don't have a full stripe */ + if (!rbio_is_full(rbio)) + return partial_stripe_write(rbio); + return full_stripe_write(rbio); +} + +/* + * our main entry point for writes from the rest of the FS. + */ +int raid56_parity_write(struct btrfs_root *root, struct bio *bio, + struct btrfs_bio *bbio, u64 *raid_map, + u64 stripe_len) +{ + struct btrfs_raid_bio *rbio; + + rbio = alloc_rbio(root, bbio, raid_map, stripe_len); + if (IS_ERR(rbio)) { + kfree(raid_map); + kfree(bbio); + return PTR_ERR(rbio); + } + bio_list_add(&rbio->bio_list, bio); + rbio->bio_list_bytes = bio->bi_size; + return __raid56_parity_write(rbio); +} + +/* + * all parity reconstruction happens here. We've read in everything + * we can find from the drives and this does the heavy lifting of + * sorting the good from the bad. + */ +static void __raid_recover_end_io(struct btrfs_raid_bio *rbio) +{ + int pagenr, stripe; + void **pointers; + int faila = -1, failb = -1; + int nr_pages = (rbio->stripe_len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + struct page *page; + int err; + int i; + + pointers = kzalloc(rbio->bbio->num_stripes * sizeof(void *), + GFP_NOFS); + if (!pointers) { + err = -ENOMEM; + goto cleanup_io; + } + + faila = rbio->faila; + failb = rbio->failb; + + if (rbio->read_rebuild) { + spin_lock_irq(&rbio->bio_list_lock); + set_bit(RBIO_RMW_LOCKED_BIT, &rbio->flags); + spin_unlock_irq(&rbio->bio_list_lock); + } + + index_rbio_pages(rbio); + + for (pagenr = 0; pagenr < nr_pages; pagenr++) { + /* setup our array of pointers with pages + * from each stripe + */ + for (stripe = 0; stripe < rbio->bbio->num_stripes; stripe++) { + /* + * if we're rebuilding a read, we have to use + * pages from the bio list + */ + if (rbio->read_rebuild && + (stripe == faila || stripe == failb)) { + page = page_in_rbio(rbio, stripe, pagenr, 0); + } else { + page = rbio_stripe_page(rbio, stripe, pagenr); + } + pointers[stripe] = kmap(page); + } + + /* all raid6 handling here */ + if (rbio->raid_map[rbio->bbio->num_stripes - 1] == + RAID6_Q_STRIPE) { + + /* + * single failure, rebuild from parity raid5 + * style + */ + if (failb < 0) { + if (faila == rbio->nr_data) { + /* + * Just the P stripe has failed, without + * a bad data or Q stripe. + * TODO, we should redo the xor here. + */ + err = -EIO; + goto cleanup; + } + /* + * a single failure in raid6 is rebuilt + * in the pstripe code below + */ + goto pstripe; + } + + /* make sure our ps and qs are in order */ + if (faila > failb) { + int tmp = failb; + failb = faila; + faila = tmp; + } + + /* if the q stripe is failed, do a pstripe reconstruction + * from the xors. + * If both the q stripe and the P stripe are failed, we're + * here due to a crc mismatch and we can't give them the + * data they want + */ + if (rbio->raid_map[failb] == RAID6_Q_STRIPE) { + if (rbio->raid_map[faila] == RAID5_P_STRIPE) { + err = -EIO; + goto cleanup; + } + /* + * otherwise we have one bad data stripe and + * a good P stripe. raid5! + */ + goto pstripe; + } + + if (rbio->raid_map[failb] == RAID5_P_STRIPE) { + raid6_datap_recov(rbio->bbio->num_stripes, + PAGE_SIZE, faila, pointers); + } else { + raid6_2data_recov(rbio->bbio->num_stripes, + PAGE_SIZE, faila, failb, + pointers); + } + } else { + void *p; + + /* rebuild from P stripe here (raid5 or raid6) */ + BUG_ON(failb != -1); +pstripe: + /* Copy parity block into failed block to start with */ + memcpy(pointers[faila], + pointers[rbio->nr_data], + PAGE_CACHE_SIZE); + + /* rearrange the pointer array */ + p = pointers[faila]; + for (stripe = faila; stripe < rbio->nr_data - 1; stripe++) + pointers[stripe] = pointers[stripe + 1]; + pointers[rbio->nr_data - 1] = p; + + /* xor in the rest */ + run_xor(pointers, rbio->nr_data - 1, PAGE_CACHE_SIZE); + } + /* if we're doing this rebuild as part of an rmw, go through + * and set all of our private rbio pages in the + * failed stripes as uptodate. This way finish_rmw will + * know they can be trusted. If this was a read reconstruction, + * other endio functions will fiddle the uptodate bits + */ + if (!rbio->read_rebuild) { + for (i = 0; i < nr_pages; i++) { + if (faila != -1) { + page = rbio_stripe_page(rbio, faila, i); + SetPageUptodate(page); + } + if (failb != -1) { + page = rbio_stripe_page(rbio, failb, i); + SetPageUptodate(page); + } + } + } + for (stripe = 0; stripe < rbio->bbio->num_stripes; stripe++) { + /* + * if we're rebuilding a read, we have to use + * pages from the bio list + */ + if (rbio->read_rebuild && + (stripe == faila || stripe == failb)) { + page = page_in_rbio(rbio, stripe, pagenr, 0); + } else { + page = rbio_stripe_page(rbio, stripe, pagenr); + } + kunmap(page); + } + } + + err = 0; +cleanup: + kfree(pointers); + +cleanup_io: + + if (rbio->read_rebuild) { + rbio_orig_end_io(rbio, err, err == 0); + } else if (err == 0) { + rbio->faila = -1; + rbio->failb = -1; + finish_rmw(rbio); + } else { + rbio_orig_end_io(rbio, err, 0); + } +} + +/* + * This is called only for stripes we've read from disk to + * reconstruct the parity. + */ +static void raid_recover_end_io(struct bio *bio, int err) +{ + struct btrfs_raid_bio *rbio = bio->bi_private; + + /* + * we only read stripe pages off the disk, set them + * up to date if there were no errors + */ + if (err) + fail_bio_stripe(rbio, bio); + else + set_bio_pages_uptodate(bio); + bio_put(bio); + + if (!atomic_dec_and_test(&rbio->bbio->stripes_pending)) + return; + + if (atomic_read(&rbio->bbio->error) > rbio->bbio->max_errors) + rbio_orig_end_io(rbio, -EIO, 0); + else + __raid_recover_end_io(rbio); +} + +/* + * reads everything we need off the disk to reconstruct + * the parity. endio handlers trigger final reconstruction + * when the IO is done. + * + * This is used both for reads from the higher layers and for + * parity construction required to finish a rmw cycle. + */ +static int __raid56_parity_recover(struct btrfs_raid_bio *rbio) +{ + int bios_to_read = 0; + struct btrfs_bio *bbio = rbio->bbio; + struct bio_list bio_list; + int ret; + int nr_pages = (rbio->stripe_len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + int pagenr; + int stripe; + struct bio *bio; + + bio_list_init(&bio_list); + + ret = alloc_rbio_pages(rbio); + if (ret) + goto cleanup; + + atomic_set(&rbio->bbio->error, 0); + + /* + * read everything that hasn't failed. + */ + for (stripe = 0; stripe < bbio->num_stripes; stripe++) { + if (rbio->faila == stripe || + rbio->failb == stripe) + continue; + + for (pagenr = 0; pagenr < nr_pages; pagenr++) { + struct page *p; + + /* + * the rmw code may have already read this + * page in + */ + p = rbio_stripe_page(rbio, stripe, pagenr); + if (PageUptodate(p)) + continue; + + ret = rbio_add_io_page(rbio, &bio_list, + rbio_stripe_page(rbio, stripe, pagenr), + stripe, pagenr, rbio->stripe_len); + if (ret < 0) + goto cleanup; + } + } + + bios_to_read = bio_list_size(&bio_list); + if (!bios_to_read) { + /* + * we might have no bios to read just because the pages + * were up to date, or we might have no bios to read because + * the devices were gone. + */ + if (atomic_read(&rbio->bbio->error) <= rbio->bbio->max_errors) { + __raid_recover_end_io(rbio); + goto out; + } else { + goto cleanup; + } + } + + /* + * the bbio may be freed once we submit the last bio. Make sure + * not to touch it after that + */ + atomic_set(&bbio->stripes_pending, bios_to_read); + while (1) { + bio = bio_list_pop(&bio_list); + if (!bio) + break; + + bio->bi_private = rbio; + bio->bi_end_io = raid_recover_end_io; + + btrfs_bio_wq_end_io(rbio->fs_info, bio, + BTRFS_WQ_ENDIO_RAID56); + + BUG_ON(!test_bit(BIO_UPTODATE, &bio->bi_flags)); + submit_bio(READ, bio); + } +out: + return 0; + +cleanup: + if (rbio->read_rebuild) + rbio_orig_end_io(rbio, -EIO, 0); + return -EIO; +} + +/* + * the main entry point for reads from the higher layers. This + * is really only called when the normal read path had a failure, + * so we assume the bio they send down corresponds to a failed part + * of the drive. + */ +int raid56_parity_recover(struct btrfs_root *root, struct bio *bio, + struct btrfs_bio *bbio, u64 *raid_map, + u64 stripe_len, int mirror_num) +{ + struct btrfs_raid_bio *rbio; + int ret; + + rbio = alloc_rbio(root, bbio, raid_map, stripe_len); + if (IS_ERR(rbio)) { + return PTR_ERR(rbio); + } + + rbio->read_rebuild = 1; + bio_list_add(&rbio->bio_list, bio); + rbio->bio_list_bytes = bio->bi_size; + + rbio->faila = find_logical_bio_stripe(rbio, bio); + if (rbio->faila == -1) { + BUG(); + kfree(rbio); + return -EIO; + } + + /* + * reconstruct from the q stripe if they are + * asking for mirror 3 + */ + if (mirror_num == 3) + rbio->failb = bbio->num_stripes - 2; + + ret = lock_stripe_add(rbio); + + /* + * __raid56_parity_recover will end the bio with + * any errors it hits. We don't want to return + * its error value up the stack because our caller + * will end up calling bio_endio with any nonzero + * return + */ + if (ret == 0) + __raid56_parity_recover(rbio); + /* + * our rbio has been added to the list of + * rbios that will be handled after the + * currently lock owner is done + */ + return 0; + +} + +static void rmw_work(struct btrfs_work *work) +{ + struct btrfs_raid_bio *rbio; + + rbio = container_of(work, struct btrfs_raid_bio, work); + raid56_rmw_stripe(rbio); +} + +static void read_rebuild_work(struct btrfs_work *work) +{ + struct btrfs_raid_bio *rbio; + + rbio = container_of(work, struct btrfs_raid_bio, work); + __raid56_parity_recover(rbio); +} diff --git a/fs/btrfs/raid56.h b/fs/btrfs/raid56.h new file mode 100644 index 000000000000..ea5d73bfdfbe --- /dev/null +++ b/fs/btrfs/raid56.h @@ -0,0 +1,51 @@ +/* + * Copyright (C) 2012 Fusion-io All rights reserved. + * Copyright (C) 2012 Intel Corp. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License v2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public + * License along with this program; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 021110-1307, USA. + */ + +#ifndef __BTRFS_RAID56__ +#define __BTRFS_RAID56__ +static inline int nr_parity_stripes(struct map_lookup *map) +{ + if (map->type & BTRFS_BLOCK_GROUP_RAID5) + return 1; + else if (map->type & BTRFS_BLOCK_GROUP_RAID6) + return 2; + else + return 0; +} + +static inline int nr_data_stripes(struct map_lookup *map) +{ + return map->num_stripes - nr_parity_stripes(map); +} +#define RAID5_P_STRIPE ((u64)-2) +#define RAID6_Q_STRIPE ((u64)-1) + +#define is_parity_stripe(x) (((x) == RAID5_P_STRIPE) || \ + ((x) == RAID6_Q_STRIPE)) + +int raid56_parity_recover(struct btrfs_root *root, struct bio *bio, + struct btrfs_bio *bbio, u64 *raid_map, + u64 stripe_len, int mirror_num); +int raid56_parity_write(struct btrfs_root *root, struct bio *bio, + struct btrfs_bio *bbio, u64 *raid_map, + u64 stripe_len); + +int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info); +void btrfs_free_stripe_hash_table(struct btrfs_fs_info *info); +#endif diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index bdbb94f245c9..bc35ed4238b8 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -28,6 +28,7 @@ #include "dev-replace.h" #include "check-integrity.h" #include "rcu-string.h" +#include "raid56.h" /* * This is only the first step towards a full-features scrub. It reads all @@ -2246,6 +2247,13 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, struct btrfs_device *extent_dev; int extent_mirror_num; + if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6)) { + if (num >= nr_data_stripes(map)) { + return 0; + } + } + nstripes = length; offset = 0; do_div(nstripes, map->stripe_len); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 87fac9a21ea5..a065dec0e330 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -686,7 +686,9 @@ int btrfs_write_marked_extents(struct btrfs_root *root, struct extent_state *cached_state = NULL; u64 start = 0; u64 end; + struct blk_plug plug; + blk_start_plug(&plug); while (!find_first_extent_bit(dirty_pages, start, &start, &end, mark, &cached_state)) { convert_extent_bit(dirty_pages, start, end, EXTENT_NEED_WAIT, @@ -700,6 +702,7 @@ int btrfs_write_marked_extents(struct btrfs_root *root, } if (err) werr = err; + blk_finish_plug(&plug); return werr; } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 485a5423e3c6..c372264b85bf 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -25,6 +25,8 @@ #include #include #include +#include +#include #include "compat.h" #include "ctree.h" #include "extent_map.h" @@ -32,6 +34,7 @@ #include "transaction.h" #include "print-tree.h" #include "volumes.h" +#include "raid56.h" #include "async-thread.h" #include "check-integrity.h" #include "rcu-string.h" @@ -1389,6 +1392,14 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) } btrfs_dev_replace_unlock(&root->fs_info->dev_replace); + if ((all_avail & (BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6) && num_devices <= 3)) { + printk(KERN_ERR "btrfs: unable to go below three devices " + "on raid5 or raid6\n"); + ret = -EINVAL; + goto out; + } + if ((all_avail & BTRFS_BLOCK_GROUP_RAID10) && num_devices <= 4) { printk(KERN_ERR "btrfs: unable to go below four devices " "on raid10\n"); @@ -1403,6 +1414,21 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) goto out; } + if ((all_avail & BTRFS_BLOCK_GROUP_RAID5) && + root->fs_info->fs_devices->rw_devices <= 2) { + printk(KERN_ERR "btrfs: unable to go below two " + "devices on raid5\n"); + ret = -EINVAL; + goto out; + } + if ((all_avail & BTRFS_BLOCK_GROUP_RAID6) && + root->fs_info->fs_devices->rw_devices <= 3) { + printk(KERN_ERR "btrfs: unable to go below three " + "devices on raid6\n"); + ret = -EINVAL; + goto out; + } + if (strcmp(device_path, "missing") == 0) { struct list_head *devices; struct btrfs_device *tmp; @@ -2657,11 +2683,15 @@ static int chunk_drange_filter(struct extent_buffer *leaf, return 0; if (btrfs_chunk_type(leaf, chunk) & (BTRFS_BLOCK_GROUP_DUP | - BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10)) - factor = 2; - else - factor = 1; - factor = num_stripes / factor; + BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10)) { + factor = num_stripes / 2; + } else if (btrfs_chunk_type(leaf, chunk) & BTRFS_BLOCK_GROUP_RAID5) { + factor = num_stripes - 1; + } else if (btrfs_chunk_type(leaf, chunk) & BTRFS_BLOCK_GROUP_RAID6) { + factor = num_stripes - 2; + } else { + factor = num_stripes; + } for (i = 0; i < num_stripes; i++) { stripe = btrfs_stripe_nr(chunk, i); @@ -2976,6 +3006,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl, int mixed = 0; int ret; u64 num_devices; + int cancel = 0; if (btrfs_fs_closing(fs_info) || atomic_read(&fs_info->balance_pause_req) || @@ -3018,7 +3049,9 @@ int btrfs_balance(struct btrfs_balance_control *bctl, allowed |= (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID1); else allowed |= (BTRFS_BLOCK_GROUP_RAID0 | BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_RAID10); + BTRFS_BLOCK_GROUP_RAID10 | + BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6); if ((bctl->data.flags & BTRFS_BALANCE_ARGS_CONVERT) && (!alloc_profile_is_valid(bctl->data.target, 1) || @@ -3058,7 +3091,10 @@ int btrfs_balance(struct btrfs_balance_control *bctl, /* allow to reduce meta or sys integrity only if force set */ allowed = BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_RAID10; + BTRFS_BLOCK_GROUP_RAID10 | + BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6; + if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) && (fs_info->avail_system_alloc_bits & allowed) && !(bctl->sys.target & allowed)) || @@ -3124,15 +3160,17 @@ int btrfs_balance(struct btrfs_balance_control *bctl, } if ((ret && ret != -ECANCELED && ret != -ENOSPC) || - balance_need_close(fs_info)) { - __cancel_balance(fs_info); - } + balance_need_close(fs_info)) + cancel = 1; if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) { fs_info->num_tolerated_disk_barrier_failures = btrfs_calc_num_tolerated_disk_barrier_failures(fs_info); } + if (cancel) + __cancel_balance(fs_info); + wake_up(&fs_info->balance_wait_q); return ret; @@ -3493,13 +3531,45 @@ static int btrfs_cmp_device_info(const void *a, const void *b) } struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = { + /* + * sub_stripes info for map, + * dev_stripes -- stripes per dev, 2 for DUP, 1 other wise + * devs_max -- max devices per stripe, 0 for unlimited + * devs_min -- min devices per stripe + * devs_increment -- ndevs must be a multiple of this + * ncopies -- how many copies of the data we have + */ { 2, 1, 0, 4, 2, 2 /* raid10 */ }, { 1, 1, 2, 2, 2, 2 /* raid1 */ }, { 1, 2, 1, 1, 1, 2 /* dup */ }, { 1, 1, 0, 2, 1, 1 /* raid0 */ }, { 1, 1, 0, 1, 1, 1 /* single */ }, + { 1, 1, 0, 2, 1, 2 /* raid5 */ }, + { 1, 1, 0, 3, 1, 3 /* raid6 */ }, }; +static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target) +{ + /* TODO allow them to set a preferred stripe size */ + return 64 * 1024; +} + +static void check_raid56_incompat_flag(struct btrfs_fs_info *info, u64 type) +{ + u64 features; + + if (!(type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6))) + return; + + features = btrfs_super_incompat_flags(info->super_copy); + if (features & BTRFS_FEATURE_INCOMPAT_RAID56) + return; + + features |= BTRFS_FEATURE_INCOMPAT_RAID56; + btrfs_set_super_incompat_flags(info->super_copy, features); + printk(KERN_INFO "btrfs: setting RAID5/6 feature flag\n"); +} + static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, struct btrfs_root *extent_root, struct map_lookup **map_ret, @@ -3515,6 +3585,8 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, struct btrfs_device_info *devices_info = NULL; u64 total_avail; int num_stripes; /* total number of stripes to allocate */ + int data_stripes; /* number of stripes that count for + block group size */ int sub_stripes; /* sub_stripes info for map */ int dev_stripes; /* stripes per dev */ int devs_max; /* max devs to use */ @@ -3526,6 +3598,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, u64 max_chunk_size; u64 stripe_size; u64 num_bytes; + u64 raid_stripe_len = BTRFS_STRIPE_LEN; int ndevs; int i; int j; @@ -3651,16 +3724,31 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, stripe_size = devices_info[ndevs-1].max_avail; num_stripes = ndevs * dev_stripes; + /* + * this will have to be fixed for RAID1 and RAID10 over + * more drives + */ + data_stripes = num_stripes / ncopies; + if (stripe_size * ndevs > max_chunk_size * ncopies) { stripe_size = max_chunk_size * ncopies; do_div(stripe_size, ndevs); } - + if (type & BTRFS_BLOCK_GROUP_RAID5) { + raid_stripe_len = find_raid56_stripe_len(ndevs - 1, + btrfs_super_stripesize(info->super_copy)); + data_stripes = num_stripes - 1; + } + if (type & BTRFS_BLOCK_GROUP_RAID6) { + raid_stripe_len = find_raid56_stripe_len(ndevs - 2, + btrfs_super_stripesize(info->super_copy)); + data_stripes = num_stripes - 2; + } do_div(stripe_size, dev_stripes); /* align to BTRFS_STRIPE_LEN */ - do_div(stripe_size, BTRFS_STRIPE_LEN); - stripe_size *= BTRFS_STRIPE_LEN; + do_div(stripe_size, raid_stripe_len); + stripe_size *= raid_stripe_len; map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS); if (!map) { @@ -3678,14 +3766,14 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, } } map->sector_size = extent_root->sectorsize; - map->stripe_len = BTRFS_STRIPE_LEN; - map->io_align = BTRFS_STRIPE_LEN; - map->io_width = BTRFS_STRIPE_LEN; + map->stripe_len = raid_stripe_len; + map->io_align = raid_stripe_len; + map->io_width = raid_stripe_len; map->type = type; map->sub_stripes = sub_stripes; *map_ret = map; - num_bytes = stripe_size * (num_stripes / ncopies); + num_bytes = stripe_size * data_stripes; *stripe_size_out = stripe_size; *num_bytes_out = num_bytes; @@ -3734,6 +3822,8 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, } } + check_raid56_incompat_flag(extent_root->fs_info, type); + kfree(devices_info); return 0; @@ -4003,6 +4093,10 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len) ret = map->num_stripes; else if (map->type & BTRFS_BLOCK_GROUP_RAID10) ret = map->sub_stripes; + else if (map->type & BTRFS_BLOCK_GROUP_RAID5) + ret = 2; + else if (map->type & BTRFS_BLOCK_GROUP_RAID6) + ret = 3; else ret = 1; free_extent_map(em); @@ -4015,6 +4109,52 @@ int btrfs_num_copies(struct btrfs_fs_info *fs_info, u64 logical, u64 len) return ret; } +unsigned long btrfs_full_stripe_len(struct btrfs_root *root, + struct btrfs_mapping_tree *map_tree, + u64 logical) +{ + struct extent_map *em; + struct map_lookup *map; + struct extent_map_tree *em_tree = &map_tree->map_tree; + unsigned long len = root->sectorsize; + + read_lock(&em_tree->lock); + em = lookup_extent_mapping(em_tree, logical, len); + read_unlock(&em_tree->lock); + BUG_ON(!em); + + BUG_ON(em->start > logical || em->start + em->len < logical); + map = (struct map_lookup *)em->bdev; + if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6)) { + len = map->stripe_len * nr_data_stripes(map); + } + free_extent_map(em); + return len; +} + +int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree, + u64 logical, u64 len, int mirror_num) +{ + struct extent_map *em; + struct map_lookup *map; + struct extent_map_tree *em_tree = &map_tree->map_tree; + int ret = 0; + + read_lock(&em_tree->lock); + em = lookup_extent_mapping(em_tree, logical, len); + read_unlock(&em_tree->lock); + BUG_ON(!em); + + BUG_ON(em->start > logical || em->start + em->len < logical); + map = (struct map_lookup *)em->bdev; + if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6)) + ret = 1; + free_extent_map(em); + return ret; +} + static int find_live_mirror(struct btrfs_fs_info *fs_info, struct map_lookup *map, int first, int num, int optimal, int dev_replace_is_ongoing) @@ -4052,10 +4192,39 @@ static int find_live_mirror(struct btrfs_fs_info *fs_info, return optimal; } +static inline int parity_smaller(u64 a, u64 b) +{ + return a > b; +} + +/* Bubble-sort the stripe set to put the parity/syndrome stripes last */ +static void sort_parity_stripes(struct btrfs_bio *bbio, u64 *raid_map) +{ + struct btrfs_bio_stripe s; + int i; + u64 l; + int again = 1; + + while (again) { + again = 0; + for (i = 0; i < bbio->num_stripes - 1; i++) { + if (parity_smaller(raid_map[i], raid_map[i+1])) { + s = bbio->stripes[i]; + l = raid_map[i]; + bbio->stripes[i] = bbio->stripes[i+1]; + raid_map[i] = raid_map[i+1]; + bbio->stripes[i+1] = s; + raid_map[i+1] = l; + again = 1; + } + } + } +} + static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, u64 logical, u64 *length, struct btrfs_bio **bbio_ret, - int mirror_num) + int mirror_num, u64 **raid_map_ret) { struct extent_map *em; struct map_lookup *map; @@ -4067,6 +4236,8 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, u64 stripe_nr; u64 stripe_nr_orig; u64 stripe_nr_end; + u64 stripe_len; + u64 *raid_map = NULL; int stripe_index; int i; int ret = 0; @@ -4078,6 +4249,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, int num_alloc_stripes; int patch_the_first_stripe_for_dev_replace = 0; u64 physical_to_patch_in_first_stripe = 0; + u64 raid56_full_stripe_start = (u64)-1; read_lock(&em_tree->lock); em = lookup_extent_mapping(em_tree, logical, *length); @@ -4094,29 +4266,63 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, map = (struct map_lookup *)em->bdev; offset = logical - em->start; + if (mirror_num > map->num_stripes) + mirror_num = 0; + + stripe_len = map->stripe_len; stripe_nr = offset; /* * stripe_nr counts the total number of stripes we have to stride * to get to this block */ - do_div(stripe_nr, map->stripe_len); + do_div(stripe_nr, stripe_len); - stripe_offset = stripe_nr * map->stripe_len; + stripe_offset = stripe_nr * stripe_len; BUG_ON(offset < stripe_offset); /* stripe_offset is the offset of this block in its stripe*/ stripe_offset = offset - stripe_offset; - if (rw & REQ_DISCARD) + /* if we're here for raid56, we need to know the stripe aligned start */ + if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { + unsigned long full_stripe_len = stripe_len * nr_data_stripes(map); + raid56_full_stripe_start = offset; + + /* allow a write of a full stripe, but make sure we don't + * allow straddling of stripes + */ + do_div(raid56_full_stripe_start, full_stripe_len); + raid56_full_stripe_start *= full_stripe_len; + } + + if (rw & REQ_DISCARD) { + /* we don't discard raid56 yet */ + if (map->type & + (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { + ret = -EOPNOTSUPP; + goto out; + } *length = min_t(u64, em->len - offset, *length); - else if (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) { - /* we limit the length of each bio to what fits in a stripe */ - *length = min_t(u64, em->len - offset, - map->stripe_len - stripe_offset); + } else if (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) { + u64 max_len; + /* For writes to RAID[56], allow a full stripeset across all disks. + For other RAID types and for RAID[56] reads, just allow a single + stripe (on a single disk). */ + if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6) && + (rw & REQ_WRITE)) { + max_len = stripe_len * nr_data_stripes(map) - + (offset - raid56_full_stripe_start); + } else { + /* we limit the length of each bio to what fits in a stripe */ + max_len = stripe_len - stripe_offset; + } + *length = min_t(u64, em->len - offset, max_len); } else { *length = em->len - offset; } + /* This is for when we're called from btrfs_merge_bio_hook() and all + it cares about is the length */ if (!bbio_ret) goto out; @@ -4149,7 +4355,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, u64 physical_of_found = 0; ret = __btrfs_map_block(fs_info, REQ_GET_READ_MIRRORS, - logical, &tmp_length, &tmp_bbio, 0); + logical, &tmp_length, &tmp_bbio, 0, NULL); if (ret) { WARN_ON(tmp_bbio != NULL); goto out; @@ -4215,6 +4421,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, do_div(stripe_nr_end, map->stripe_len); stripe_end_offset = stripe_nr_end * map->stripe_len - (offset + *length); + if (map->type & BTRFS_BLOCK_GROUP_RAID0) { if (rw & REQ_DISCARD) num_stripes = min_t(u64, map->num_stripes, @@ -4265,6 +4472,65 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, dev_replace_is_ongoing); mirror_num = stripe_index - old_stripe_index + 1; } + + } else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6)) { + u64 tmp; + + if (bbio_ret && ((rw & REQ_WRITE) || mirror_num > 1) + && raid_map_ret) { + int i, rot; + + /* push stripe_nr back to the start of the full stripe */ + stripe_nr = raid56_full_stripe_start; + do_div(stripe_nr, stripe_len); + + stripe_index = do_div(stripe_nr, nr_data_stripes(map)); + + /* RAID[56] write or recovery. Return all stripes */ + num_stripes = map->num_stripes; + max_errors = nr_parity_stripes(map); + + raid_map = kmalloc(sizeof(u64) * num_stripes, + GFP_NOFS); + if (!raid_map) { + ret = -ENOMEM; + goto out; + } + + /* Work out the disk rotation on this stripe-set */ + tmp = stripe_nr; + rot = do_div(tmp, num_stripes); + + /* Fill in the logical address of each stripe */ + tmp = stripe_nr * nr_data_stripes(map); + for (i = 0; i < nr_data_stripes(map); i++) + raid_map[(i+rot) % num_stripes] = + em->start + (tmp + i) * map->stripe_len; + + raid_map[(i+rot) % map->num_stripes] = RAID5_P_STRIPE; + if (map->type & BTRFS_BLOCK_GROUP_RAID6) + raid_map[(i+rot+1) % num_stripes] = + RAID6_Q_STRIPE; + + *length = map->stripe_len; + stripe_index = 0; + stripe_offset = 0; + } else { + /* + * Mirror #0 or #1 means the original data block. + * Mirror #2 is RAID5 parity block. + * Mirror #3 is RAID6 Q block. + */ + stripe_index = do_div(stripe_nr, nr_data_stripes(map)); + if (mirror_num > 1) + stripe_index = nr_data_stripes(map) + + mirror_num - 2; + + /* We distribute the parity blocks across stripes */ + tmp = stripe_nr + stripe_index; + stripe_index = do_div(tmp, map->num_stripes); + } } else { /* * after this do_div call, stripe_nr is the number of stripes @@ -4373,8 +4639,11 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, if (rw & (REQ_WRITE | REQ_GET_READ_MIRRORS)) { if (map->type & (BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10 | + BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_DUP)) { max_errors = 1; + } else if (map->type & BTRFS_BLOCK_GROUP_RAID6) { + max_errors = 2; } } @@ -4475,6 +4744,10 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, bbio->stripes[0].physical = physical_to_patch_in_first_stripe; bbio->mirror_num = map->num_stripes + 1; } + if (raid_map) { + sort_parity_stripes(bbio, raid_map); + *raid_map_ret = raid_map; + } out: if (dev_replace_is_ongoing) btrfs_dev_replace_unlock(dev_replace); @@ -4487,7 +4760,7 @@ int btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, struct btrfs_bio **bbio_ret, int mirror_num) { return __btrfs_map_block(fs_info, rw, logical, length, bbio_ret, - mirror_num); + mirror_num, NULL); } int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, @@ -4501,6 +4774,7 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, u64 bytenr; u64 length; u64 stripe_nr; + u64 rmap_len; int i, j, nr = 0; read_lock(&em_tree->lock); @@ -4511,10 +4785,17 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, map = (struct map_lookup *)em->bdev; length = em->len; + rmap_len = map->stripe_len; + if (map->type & BTRFS_BLOCK_GROUP_RAID10) do_div(length, map->num_stripes / map->sub_stripes); else if (map->type & BTRFS_BLOCK_GROUP_RAID0) do_div(length, map->num_stripes); + else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | + BTRFS_BLOCK_GROUP_RAID6)) { + do_div(length, nr_data_stripes(map)); + rmap_len = map->stripe_len * nr_data_stripes(map); + } buf = kzalloc(sizeof(u64) * map->num_stripes, GFP_NOFS); BUG_ON(!buf); /* -ENOMEM */ @@ -4534,8 +4815,11 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, do_div(stripe_nr, map->sub_stripes); } else if (map->type & BTRFS_BLOCK_GROUP_RAID0) { stripe_nr = stripe_nr * map->num_stripes + i; - } - bytenr = chunk_start + stripe_nr * map->stripe_len; + } /* else if RAID[56], multiply by nr_data_stripes(). + * Alternatively, just use rmap_len below instead of + * map->stripe_len */ + + bytenr = chunk_start + stripe_nr * rmap_len; WARN_ON(nr >= map->num_stripes); for (j = 0; j < nr; j++) { if (buf[j] == bytenr) @@ -4549,7 +4833,7 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, *logical = buf; *naddrs = nr; - *stripe_len = map->stripe_len; + *stripe_len = rmap_len; free_extent_map(em); return 0; @@ -4623,7 +4907,7 @@ static void btrfs_end_bio(struct bio *bio, int err) bio->bi_bdev = (struct block_device *) (unsigned long)bbio->mirror_num; /* only send an error to the higher layers if it is - * beyond the tolerance of the multi-bio + * beyond the tolerance of the btrfs bio */ if (atomic_read(&bbio->error) > bbio->max_errors) { err = -EIO; @@ -4657,13 +4941,18 @@ struct async_sched { * This will add one bio to the pending list for a device and make sure * the work struct is scheduled. */ -static noinline void schedule_bio(struct btrfs_root *root, +noinline void btrfs_schedule_bio(struct btrfs_root *root, struct btrfs_device *device, int rw, struct bio *bio) { int should_queue = 1; struct btrfs_pending_bios *pending_bios; + if (device->missing || !device->bdev) { + bio_endio(bio, -EIO); + return; + } + /* don't bother with additional async steps for reads, right now */ if (!(rw & REQ_WRITE)) { bio_get(bio); @@ -4761,7 +5050,7 @@ static void submit_stripe_bio(struct btrfs_root *root, struct btrfs_bio *bbio, #endif bio->bi_bdev = dev->bdev; if (async) - schedule_bio(root, dev, rw, bio); + btrfs_schedule_bio(root, dev, rw, bio); else btrfsic_submit_bio(rw, bio); } @@ -4820,6 +5109,7 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, u64 logical = (u64)bio->bi_sector << 9; u64 length = 0; u64 map_length; + u64 *raid_map = NULL; int ret; int dev_nr = 0; int total_devs = 1; @@ -4828,12 +5118,30 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, length = bio->bi_size; map_length = length; - ret = btrfs_map_block(root->fs_info, rw, logical, &map_length, &bbio, - mirror_num); - if (ret) + ret = __btrfs_map_block(root->fs_info, rw, logical, &map_length, &bbio, + mirror_num, &raid_map); + if (ret) /* -ENOMEM */ return ret; total_devs = bbio->num_stripes; + bbio->orig_bio = first_bio; + bbio->private = first_bio->bi_private; + bbio->end_io = first_bio->bi_end_io; + atomic_set(&bbio->stripes_pending, bbio->num_stripes); + + if (raid_map) { + /* In this case, map_length has been set to the length of + a single stripe; not the whole write */ + if (rw & WRITE) { + return raid56_parity_write(root, bio, bbio, + raid_map, map_length); + } else { + return raid56_parity_recover(root, bio, bbio, + raid_map, map_length, + mirror_num); + } + } + if (map_length < length) { printk(KERN_CRIT "btrfs: mapping failed logical %llu bio len %llu " "len %llu\n", (unsigned long long)logical, @@ -4842,11 +5150,6 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, BUG(); } - bbio->orig_bio = first_bio; - bbio->private = first_bio->bi_private; - bbio->end_io = first_bio->bi_end_io; - atomic_set(&bbio->stripes_pending, bbio->num_stripes); - while (dev_nr < total_devs) { dev = bbio->stripes[dev_nr].dev; if (!dev || !dev->bdev || (rw & WRITE && !dev->writeable)) { diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index d3c3939ac751..0c2b856ecd98 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -321,7 +321,14 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, void btrfs_init_dev_replace_tgtdev_for_resume(struct btrfs_fs_info *fs_info, struct btrfs_device *tgtdev); int btrfs_scratch_superblock(struct btrfs_device *device); - +void btrfs_schedule_bio(struct btrfs_root *root, + struct btrfs_device *device, + int rw, struct bio *bio); +int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree, + u64 logical, u64 len, int mirror_num); +unsigned long btrfs_full_stripe_len(struct btrfs_root *root, + struct btrfs_mapping_tree *map_tree, + u64 logical); static inline void btrfs_dev_stat_inc(struct btrfs_device *dev, int index) { -- cgit From 4ae10b3a133e1147f3c818fe2ebaf005b217b7bf Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Thu, 31 Jan 2013 14:42:09 -0500 Subject: Btrfs: Add a stripe cache to raid56 The stripe cache allows us to avoid extra read/modify/write cycles by caching the pages we read off the disk. Pages are cached when: * They are read in during a read/modify/write cycle * They are written during a read/modify/write cycle * They are involved in a parity rebuild Pages are not cached if we're doing a full stripe write. We're assuming that a full stripe write won't be followed by another partial stripe write any time soon. This provides a substantial boost in performance for workloads that synchronously modify adjacent offsets in the file, and for the parity rebuild use case in general. The size of the stripe cache isn't tunable (yet) and is set at 1024 entries. Example on flash: dd if=/dev/zero of=/mnt/xxx bs=4K oflag=direct Without the stripe cache -- 2.1MB/s With the stripe cache 21MB/s Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 5 +- fs/btrfs/raid56.c | 327 ++++++++++++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 324 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 0cce3aafbd62..e3a4fd70f55a 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1244,7 +1244,10 @@ struct btrfs_stripe_hash { /* used by the raid56 code to lock stripes for read/modify/write */ struct btrfs_stripe_hash_table { - struct btrfs_stripe_hash *table; + struct list_head stripe_cache; + spinlock_t cache_lock; + int cache_size; + struct btrfs_stripe_hash table[]; }; #define BTRFS_STRIPE_HASH_TABLE_BITS 11 diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c index d02510f34936..7ccddca9ee71 100644 --- a/fs/btrfs/raid56.c +++ b/fs/btrfs/raid56.c @@ -47,6 +47,20 @@ /* set when additional merges to this rbio are not allowed */ #define RBIO_RMW_LOCKED_BIT 1 +/* + * set when this rbio is sitting in the hash, but it is just a cache + * of past RMW + */ +#define RBIO_CACHE_BIT 2 + +/* + * set when it is safe to trust the stripe_pages for caching + */ +#define RBIO_CACHE_READY_BIT 3 + + +#define RBIO_CACHE_SIZE 1024 + struct btrfs_raid_bio { struct btrfs_fs_info *fs_info; struct btrfs_bio *bbio; @@ -65,6 +79,11 @@ struct btrfs_raid_bio { */ struct list_head hash_list; + /* + * LRU list for the stripe cache + */ + struct list_head stripe_cache; + /* * for scheduling work in the helper threads */ @@ -176,7 +195,9 @@ int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info) if (!table) return -ENOMEM; - table->table = (void *)(table + 1); + spin_lock_init(&table->cache_lock); + INIT_LIST_HEAD(&table->stripe_cache); + h = table->table; for (i = 0; i < num_entries; i++) { @@ -192,6 +213,42 @@ int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info) return 0; } +/* + * caching an rbio means to copy anything from the + * bio_pages array into the stripe_pages array. We + * use the page uptodate bit in the stripe cache array + * to indicate if it has valid data + * + * once the caching is done, we set the cache ready + * bit. + */ +static void cache_rbio_pages(struct btrfs_raid_bio *rbio) +{ + int i; + char *s; + char *d; + int ret; + + ret = alloc_rbio_pages(rbio); + if (ret) + return; + + for (i = 0; i < rbio->nr_pages; i++) { + if (!rbio->bio_pages[i]) + continue; + + s = kmap(rbio->bio_pages[i]); + d = kmap(rbio->stripe_pages[i]); + + memcpy(d, s, PAGE_CACHE_SIZE); + + kunmap(rbio->bio_pages[i]); + kunmap(rbio->stripe_pages[i]); + SetPageUptodate(rbio->stripe_pages[i]); + } + set_bit(RBIO_CACHE_READY_BIT, &rbio->flags); +} + /* * we hash on the first logical address of the stripe */ @@ -210,6 +267,34 @@ static int rbio_bucket(struct btrfs_raid_bio *rbio) return hash_64(num >> 16, BTRFS_STRIPE_HASH_TABLE_BITS); } +/* + * stealing an rbio means taking all the uptodate pages from the stripe + * array in the source rbio and putting them into the destination rbio + */ +static void steal_rbio(struct btrfs_raid_bio *src, struct btrfs_raid_bio *dest) +{ + int i; + struct page *s; + struct page *d; + + if (!test_bit(RBIO_CACHE_READY_BIT, &src->flags)) + return; + + for (i = 0; i < dest->nr_pages; i++) { + s = src->stripe_pages[i]; + if (!s || !PageUptodate(s)) { + continue; + } + + d = dest->stripe_pages[i]; + if (d) + __free_page(d); + + dest->stripe_pages[i] = s; + src->stripe_pages[i] = NULL; + } +} + /* * merging means we take the bio_list from the victim and * splice it into the destination. The victim should @@ -226,16 +311,170 @@ static void merge_rbio(struct btrfs_raid_bio *dest, } /* - * free the hash table used by unmount + * used to prune items that are in the cache. The caller + * must hold the hash table lock. + */ +static void __remove_rbio_from_cache(struct btrfs_raid_bio *rbio) +{ + int bucket = rbio_bucket(rbio); + struct btrfs_stripe_hash_table *table; + struct btrfs_stripe_hash *h; + int freeit = 0; + + /* + * check the bit again under the hash table lock. + */ + if (!test_bit(RBIO_CACHE_BIT, &rbio->flags)) + return; + + table = rbio->fs_info->stripe_hash_table; + h = table->table + bucket; + + /* hold the lock for the bucket because we may be + * removing it from the hash table + */ + spin_lock(&h->lock); + + /* + * hold the lock for the bio list because we need + * to make sure the bio list is empty + */ + spin_lock(&rbio->bio_list_lock); + + if (test_and_clear_bit(RBIO_CACHE_BIT, &rbio->flags)) { + list_del_init(&rbio->stripe_cache); + table->cache_size -= 1; + freeit = 1; + + /* if the bio list isn't empty, this rbio is + * still involved in an IO. We take it out + * of the cache list, and drop the ref that + * was held for the list. + * + * If the bio_list was empty, we also remove + * the rbio from the hash_table, and drop + * the corresponding ref + */ + if (bio_list_empty(&rbio->bio_list)) { + if (!list_empty(&rbio->hash_list)) { + list_del_init(&rbio->hash_list); + atomic_dec(&rbio->refs); + BUG_ON(!list_empty(&rbio->plug_list)); + } + } + } + + spin_unlock(&rbio->bio_list_lock); + spin_unlock(&h->lock); + + if (freeit) + __free_raid_bio(rbio); +} + +/* + * prune a given rbio from the cache + */ +static void remove_rbio_from_cache(struct btrfs_raid_bio *rbio) +{ + struct btrfs_stripe_hash_table *table; + unsigned long flags; + + if (!test_bit(RBIO_CACHE_BIT, &rbio->flags)) + return; + + table = rbio->fs_info->stripe_hash_table; + + spin_lock_irqsave(&table->cache_lock, flags); + __remove_rbio_from_cache(rbio); + spin_unlock_irqrestore(&table->cache_lock, flags); +} + +/* + * remove everything in the cache + */ +void btrfs_clear_rbio_cache(struct btrfs_fs_info *info) +{ + struct btrfs_stripe_hash_table *table; + unsigned long flags; + struct btrfs_raid_bio *rbio; + + table = info->stripe_hash_table; + + spin_lock_irqsave(&table->cache_lock, flags); + while (!list_empty(&table->stripe_cache)) { + rbio = list_entry(table->stripe_cache.next, + struct btrfs_raid_bio, + stripe_cache); + __remove_rbio_from_cache(rbio); + } + spin_unlock_irqrestore(&table->cache_lock, flags); +} + +/* + * remove all cached entries and free the hash table + * used by unmount */ void btrfs_free_stripe_hash_table(struct btrfs_fs_info *info) { if (!info->stripe_hash_table) return; + btrfs_clear_rbio_cache(info); kfree(info->stripe_hash_table); info->stripe_hash_table = NULL; } +/* + * insert an rbio into the stripe cache. It + * must have already been prepared by calling + * cache_rbio_pages + * + * If this rbio was already cached, it gets + * moved to the front of the lru. + * + * If the size of the rbio cache is too big, we + * prune an item. + */ +static void cache_rbio(struct btrfs_raid_bio *rbio) +{ + struct btrfs_stripe_hash_table *table; + unsigned long flags; + + if (!test_bit(RBIO_CACHE_READY_BIT, &rbio->flags)) + return; + + table = rbio->fs_info->stripe_hash_table; + + spin_lock_irqsave(&table->cache_lock, flags); + spin_lock(&rbio->bio_list_lock); + + /* bump our ref if we were not in the list before */ + if (!test_and_set_bit(RBIO_CACHE_BIT, &rbio->flags)) + atomic_inc(&rbio->refs); + + if (!list_empty(&rbio->stripe_cache)){ + list_move(&rbio->stripe_cache, &table->stripe_cache); + } else { + list_add(&rbio->stripe_cache, &table->stripe_cache); + table->cache_size += 1; + } + + spin_unlock(&rbio->bio_list_lock); + + if (table->cache_size > RBIO_CACHE_SIZE) { + struct btrfs_raid_bio *found; + + found = list_entry(table->stripe_cache.prev, + struct btrfs_raid_bio, + stripe_cache); + + if (found != rbio) + __remove_rbio_from_cache(found); + } + + spin_unlock_irqrestore(&table->cache_lock, flags); + return; +} + /* * helper function to run the xor_blocks api. It is only * able to do MAX_XOR_BLOCKS at a time, so we need to @@ -303,6 +542,17 @@ static int rbio_can_merge(struct btrfs_raid_bio *last, test_bit(RBIO_RMW_LOCKED_BIT, &cur->flags)) return 0; + /* + * we can't merge with cached rbios, since the + * idea is that when we merge the destination + * rbio is going to run our IO for us. We can + * steal from cached rbio's though, other functions + * handle that. + */ + if (test_bit(RBIO_CACHE_BIT, &last->flags) || + test_bit(RBIO_CACHE_BIT, &cur->flags)) + return 0; + if (last->raid_map[0] != cur->raid_map[0]) return 0; @@ -370,6 +620,7 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) unsigned long flags; DEFINE_WAIT(wait); struct btrfs_raid_bio *freeit = NULL; + struct btrfs_raid_bio *cache_drop = NULL; int ret = 0; int walk = 0; @@ -379,6 +630,21 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) if (cur->raid_map[0] == rbio->raid_map[0]) { spin_lock(&cur->bio_list_lock); + /* can we steal this cached rbio's pages? */ + if (bio_list_empty(&cur->bio_list) && + list_empty(&cur->plug_list) && + test_bit(RBIO_CACHE_BIT, &cur->flags) && + !test_bit(RBIO_RMW_LOCKED_BIT, &cur->flags)) { + list_del_init(&cur->hash_list); + atomic_dec(&cur->refs); + + steal_rbio(cur, rbio); + cache_drop = cur; + spin_unlock(&cur->bio_list_lock); + + goto lockit; + } + /* can we merge into the lock owner? */ if (rbio_can_merge(cur, rbio)) { merge_rbio(cur, rbio); @@ -388,6 +654,7 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) goto out; } + /* * we couldn't merge with the running * rbio, see if we can merge with the @@ -417,11 +684,13 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) goto out; } } - +lockit: atomic_inc(&rbio->refs); list_add(&rbio->hash_list, &h->hash_list); out: spin_unlock_irqrestore(&h->lock, flags); + if (cache_drop) + remove_rbio_from_cache(cache_drop); if (freeit) __free_raid_bio(freeit); return ret; @@ -436,14 +705,30 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio) int bucket; struct btrfs_stripe_hash *h; unsigned long flags; + int keep_cache = 0; bucket = rbio_bucket(rbio); h = rbio->fs_info->stripe_hash_table->table + bucket; + if (list_empty(&rbio->plug_list)) + cache_rbio(rbio); + spin_lock_irqsave(&h->lock, flags); spin_lock(&rbio->bio_list_lock); if (!list_empty(&rbio->hash_list)) { + /* + * if we're still cached and there is no other IO + * to perform, just leave this rbio here for others + * to steal from later + */ + if (list_empty(&rbio->plug_list) && + test_bit(RBIO_CACHE_BIT, &rbio->flags)) { + keep_cache = 1; + clear_bit(RBIO_RMW_LOCKED_BIT, &rbio->flags); + BUG_ON(!bio_list_empty(&rbio->bio_list)); + goto done; + } list_del_init(&rbio->hash_list); atomic_dec(&rbio->refs); @@ -469,11 +754,12 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio) if (next->read_rebuild) async_read_rebuild(next); - else + else { + steal_rbio(rbio, next); async_rmw_stripe(next); + } goto done_nolock; - } else if (waitqueue_active(&h->wait)) { spin_unlock(&rbio->bio_list_lock); spin_unlock_irqrestore(&h->lock, flags); @@ -481,11 +767,13 @@ static noinline void unlock_stripe(struct btrfs_raid_bio *rbio) goto done_nolock; } } +done: spin_unlock(&rbio->bio_list_lock); spin_unlock_irqrestore(&h->lock, flags); done_nolock: - return; + if (!keep_cache) + remove_rbio_from_cache(rbio); } static void __free_raid_bio(struct btrfs_raid_bio *rbio) @@ -496,6 +784,7 @@ static void __free_raid_bio(struct btrfs_raid_bio *rbio) if (!atomic_dec_and_test(&rbio->refs)) return; + WARN_ON(!list_empty(&rbio->stripe_cache)); WARN_ON(!list_empty(&rbio->hash_list)); WARN_ON(!bio_list_empty(&rbio->bio_list)); @@ -630,6 +919,7 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, bio_list_init(&rbio->bio_list); INIT_LIST_HEAD(&rbio->plug_list); spin_lock_init(&rbio->bio_list_lock); + INIT_LIST_HEAD(&rbio->stripe_cache); INIT_LIST_HEAD(&rbio->hash_list); rbio->bbio = bbio; rbio->raid_map = raid_map; @@ -864,8 +1154,17 @@ static noinline void finish_rmw(struct btrfs_raid_bio *rbio) /* * now that we've set rmw_locked, run through the * bio list one last time and map the page pointers + * + * We don't cache full rbios because we're assuming + * the higher layers are unlikely to use this area of + * the disk again soon. If they do use it again, + * hopefully they will send another full bio. */ index_rbio_pages(rbio); + if (!rbio_is_full(rbio)) + cache_rbio_pages(rbio); + else + clear_bit(RBIO_CACHE_READY_BIT, &rbio->flags); for (pagenr = 0; pagenr < pages_per_stripe; pagenr++) { struct page *p; @@ -1155,6 +1454,13 @@ static int raid56_rmw_stripe(struct btrfs_raid_bio *rbio) continue; page = rbio_stripe_page(rbio, stripe, pagenr); + /* + * the bio cache may have handed us an uptodate + * page. If so, be happy and use it + */ + if (PageUptodate(page)) + continue; + ret = rbio_add_io_page(rbio, &bio_list, page, stripe, pagenr, rbio->stripe_len); if (ret) @@ -1440,6 +1746,11 @@ cleanup: cleanup_io: if (rbio->read_rebuild) { + if (err == 0) + cache_rbio_pages(rbio); + else + clear_bit(RBIO_CACHE_READY_BIT, &rbio->flags); + rbio_orig_end_io(rbio, err, err == 0); } else if (err == 0) { rbio->faila = -1; @@ -1505,7 +1816,9 @@ static int __raid56_parity_recover(struct btrfs_raid_bio *rbio) atomic_set(&rbio->bbio->error, 0); /* - * read everything that hasn't failed. + * read everything that hasn't failed. Thanks to the + * stripe cache, it is possible that some or all of these + * pages are going to be uptodate. */ for (stripe = 0; stripe < bbio->num_stripes; stripe++) { if (rbio->faila == stripe || -- cgit From 6ac0f4884eaad28441c6e0f9d1400a08c2149049 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Thu, 31 Jan 2013 14:42:28 -0500 Subject: Btrfs: add a plugging callback to raid56 writes Buffered writes and DIRECT_IO writes will often break up big contiguous changes to the file into sub-stripe writes. This adds a plugging callback to gather those smaller writes full stripe writes. Example on flash: fio job to do 64K writes in batches of 3 (which makes a full stripe): With plugging: 450MB/s Without plugging: 220MB/s Signed-off-by: Chris Mason --- fs/btrfs/raid56.c | 128 ++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 124 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c index 7ccddca9ee71..e34e568534d9 100644 --- a/fs/btrfs/raid56.c +++ b/fs/btrfs/raid56.c @@ -97,9 +97,10 @@ struct btrfs_raid_bio { struct bio_list bio_list; spinlock_t bio_list_lock; - /* - * also protected by the bio_list_lock, the - * stripe locking code uses plug_list to hand off + /* also protected by the bio_list_lock, the + * plug list is used by the plugging code + * to collect partial bios while plugged. The + * stripe locking code also uses it to hand off * the stripe lock to the next pending IO */ struct list_head plug_list; @@ -1557,6 +1558,103 @@ static int __raid56_parity_write(struct btrfs_raid_bio *rbio) return full_stripe_write(rbio); } +/* + * We use plugging call backs to collect full stripes. + * Any time we get a partial stripe write while plugged + * we collect it into a list. When the unplug comes down, + * we sort the list by logical block number and merge + * everything we can into the same rbios + */ +struct btrfs_plug_cb { + struct blk_plug_cb cb; + struct btrfs_fs_info *info; + struct list_head rbio_list; + struct btrfs_work work; +}; + +/* + * rbios on the plug list are sorted for easier merging. + */ +static int plug_cmp(void *priv, struct list_head *a, struct list_head *b) +{ + struct btrfs_raid_bio *ra = container_of(a, struct btrfs_raid_bio, + plug_list); + struct btrfs_raid_bio *rb = container_of(b, struct btrfs_raid_bio, + plug_list); + u64 a_sector = ra->bio_list.head->bi_sector; + u64 b_sector = rb->bio_list.head->bi_sector; + + if (a_sector < b_sector) + return -1; + if (a_sector > b_sector) + return 1; + return 0; +} + +static void run_plug(struct btrfs_plug_cb *plug) +{ + struct btrfs_raid_bio *cur; + struct btrfs_raid_bio *last = NULL; + + /* + * sort our plug list then try to merge + * everything we can in hopes of creating full + * stripes. + */ + list_sort(NULL, &plug->rbio_list, plug_cmp); + while (!list_empty(&plug->rbio_list)) { + cur = list_entry(plug->rbio_list.next, + struct btrfs_raid_bio, plug_list); + list_del_init(&cur->plug_list); + + if (rbio_is_full(cur)) { + /* we have a full stripe, send it down */ + full_stripe_write(cur); + continue; + } + if (last) { + if (rbio_can_merge(last, cur)) { + merge_rbio(last, cur); + __free_raid_bio(cur); + continue; + + } + __raid56_parity_write(last); + } + last = cur; + } + if (last) { + __raid56_parity_write(last); + } + kfree(plug); +} + +/* + * if the unplug comes from schedule, we have to push the + * work off to a helper thread + */ +static void unplug_work(struct btrfs_work *work) +{ + struct btrfs_plug_cb *plug; + plug = container_of(work, struct btrfs_plug_cb, work); + run_plug(plug); +} + +static void btrfs_raid_unplug(struct blk_plug_cb *cb, bool from_schedule) +{ + struct btrfs_plug_cb *plug; + plug = container_of(cb, struct btrfs_plug_cb, cb); + + if (from_schedule) { + plug->work.flags = 0; + plug->work.func = unplug_work; + btrfs_queue_worker(&plug->info->rmw_workers, + &plug->work); + return; + } + run_plug(plug); +} + /* * our main entry point for writes from the rest of the FS. */ @@ -1565,6 +1663,8 @@ int raid56_parity_write(struct btrfs_root *root, struct bio *bio, u64 stripe_len) { struct btrfs_raid_bio *rbio; + struct btrfs_plug_cb *plug = NULL; + struct blk_plug_cb *cb; rbio = alloc_rbio(root, bbio, raid_map, stripe_len); if (IS_ERR(rbio)) { @@ -1574,7 +1674,27 @@ int raid56_parity_write(struct btrfs_root *root, struct bio *bio, } bio_list_add(&rbio->bio_list, bio); rbio->bio_list_bytes = bio->bi_size; - return __raid56_parity_write(rbio); + + /* + * don't plug on full rbios, just get them out the door + * as quickly as we can + */ + if (rbio_is_full(rbio)) + return full_stripe_write(rbio); + + cb = blk_check_plugged(btrfs_raid_unplug, root->fs_info, + sizeof(*plug)); + if (cb) { + plug = container_of(cb, struct btrfs_plug_cb, cb); + if (!plug->info) { + plug->info = root->fs_info; + INIT_LIST_HEAD(&plug->rbio_list); + } + list_add_tail(&rbio->plug_list, &plug->rbio_list); + } else { + return __raid56_parity_write(rbio); + } + return 0; } /* -- cgit From 8de972b4faa4be9b2a3c53103e18d86092a5da45 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 4 Jan 2013 15:39:43 -0500 Subject: Btrfs: fix cluster alignment for mount -o ssd With the new raid56 code, we want to make sure we're properly aligning our allocation clusters with -o ssd Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 3345f68fc64b..f13402104c96 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5715,6 +5715,7 @@ have_block_group: * lets look there */ if (last_ptr) { + unsigned long aligned_cluster; /* * the refill lock keeps out other * people trying to start a new cluster @@ -5781,11 +5782,15 @@ refill_cluster: goto unclustered_alloc; } + aligned_cluster = max_t(unsigned long, + empty_cluster + empty_size, + block_group->full_stripe_len); + /* allocate a cluster in this block group */ ret = btrfs_find_space_cluster(trans, root, block_group, last_ptr, search_start, num_bytes, - empty_cluster + empty_size); + aligned_cluster); if (ret == 0) { /* * now pull our allocation out of this -- cgit From 242e18c7c1a8ff3aa05c9fbb6e0bb427511152a6 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 29 Jan 2013 17:49:37 -0500 Subject: Btrfs: reduce lock contention on extent buffer locks The extent buffers have a refs_lock which we use to make coordinate freeing the extent buffer with operations on the radix tree. On tree roots and other extent buffers that very cache hot, this can be highly contended. These are also the extent buffers that are basically pinned in memory. This commit adds code to cmpxchg our way through the ref modifications, and as long as the result of the reference change is still pinned in ram, we skip the expensive spinlock. Signed-off-by: Chris Mason --- fs/btrfs/extent_io.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 3b9fb478b0d1..907ed0025dd4 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -4184,6 +4184,7 @@ static inline void btrfs_release_extent_buffer(struct extent_buffer *eb) static void check_buffer_tree_ref(struct extent_buffer *eb) { + int refs; /* the ref bit is tricky. We have to make sure it is set * if we have the buffer dirty. Otherwise the * code to free a buffer can end up dropping a dirty @@ -4204,6 +4205,10 @@ static void check_buffer_tree_ref(struct extent_buffer *eb) * So bump the ref count first, then set the bit. If someone * beat us to it, drop the ref we added. */ + refs = atomic_read(&eb->refs); + if (refs >= 2 && test_bit(EXTENT_BUFFER_TREE_REF, &eb->bflags)) + return; + spin_lock(&eb->refs_lock); if (!test_and_set_bit(EXTENT_BUFFER_TREE_REF, &eb->bflags)) atomic_inc(&eb->refs); @@ -4405,9 +4410,20 @@ static int release_extent_buffer(struct extent_buffer *eb, gfp_t mask) void free_extent_buffer(struct extent_buffer *eb) { + int refs; + int old; if (!eb) return; + while (1) { + refs = atomic_read(&eb->refs); + if (refs <= 3) + break; + old = atomic_cmpxchg(&eb->refs, refs, refs - 1); + if (old == refs) + return; + } + spin_lock(&eb->refs_lock); if (atomic_read(&eb->refs) == 2 && test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags)) -- cgit From bb721703aa551e98dc5c7fb259cf90343408baf2 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 29 Jan 2013 18:44:12 -0500 Subject: Btrfs: reduce CPU contention while waiting for delayed extent operations We batch up operations to the extent allocation tree, which allows us to deal with the recursive nature of using the extent allocation tree to allocate extents to the extent allocation tree. It also provides a mechanism to sort and collect extent operations, which makes it much more efficient to record extents that are close together. The delayed extent operations must all be finished before the running transaction commits, so we have code to make sure and run a few of the batched operations when closing our transaction handles. This creates a great deal of contention for the locks in the delayed extent operation tree, and also contention for the lock on the extent allocation tree itself. All the extra contention just slows down the operations and doesn't get things done any faster. This commit changes things to use a wait queue instead. As procs want to run the delayed operations, one of them races in and gets permission to hit the tree, and the others step back and wait for progress to be made. Signed-off-by: Chris Mason --- fs/btrfs/delayed-ref.h | 9 ++++++++ fs/btrfs/extent-tree.c | 60 ++++++++++++++++++++++++++++++++++++++++++++++---- fs/btrfs/transaction.c | 6 ++++- 3 files changed, 70 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index c9d703693df0..23bdeb8502a6 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -131,6 +131,15 @@ struct btrfs_delayed_ref_root { /* total number of head nodes ready for processing */ unsigned long num_heads_ready; + /* + * bumped when someone is making progress on the delayed + * refs, so that other procs know they are just adding to + * contention intead of helping + */ + atomic_t procs_running_refs; + atomic_t ref_seq; + wait_queue_head_t wait; + /* * set when the tree is flushing before a transaction commit, * used by the throttling code to decide if new updates need diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index f13402104c96..87b0e856b6d0 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2438,6 +2438,16 @@ int btrfs_delayed_refs_qgroup_accounting(struct btrfs_trans_handle *trans, return ret; } +static int refs_newer(struct btrfs_delayed_ref_root *delayed_refs, int seq, + int count) +{ + int val = atomic_read(&delayed_refs->ref_seq); + + if (val < seq || val >= seq + count) + return 1; + return 0; +} + /* * this starts processing the delayed reference count updates and * extent insertions we have queued up so far. count can be @@ -2472,6 +2482,44 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, delayed_refs = &trans->transaction->delayed_refs; INIT_LIST_HEAD(&cluster); + if (count == 0) { + count = delayed_refs->num_entries * 2; + run_most = 1; + } + + if (!run_all && !run_most) { + int old; + int seq = atomic_read(&delayed_refs->ref_seq); + +progress: + old = atomic_cmpxchg(&delayed_refs->procs_running_refs, 0, 1); + if (old) { + DEFINE_WAIT(__wait); + if (delayed_refs->num_entries < 16348) + return 0; + + prepare_to_wait(&delayed_refs->wait, &__wait, + TASK_UNINTERRUPTIBLE); + + old = atomic_cmpxchg(&delayed_refs->procs_running_refs, 0, 1); + if (old) { + schedule(); + finish_wait(&delayed_refs->wait, &__wait); + + if (!refs_newer(delayed_refs, seq, 256)) + goto progress; + else + return 0; + } else { + finish_wait(&delayed_refs->wait, &__wait); + goto again; + } + } + + } else { + atomic_inc(&delayed_refs->procs_running_refs); + } + again: loops = 0; spin_lock(&delayed_refs->lock); @@ -2480,10 +2528,6 @@ again: delayed_refs->run_delayed_start = find_middle(&delayed_refs->root); #endif - if (count == 0) { - count = delayed_refs->num_entries * 2; - run_most = 1; - } while (1) { if (!(run_all || run_most) && delayed_refs->num_heads_ready < 64) @@ -2505,9 +2549,12 @@ again: if (ret < 0) { spin_unlock(&delayed_refs->lock); btrfs_abort_transaction(trans, root, ret); + atomic_dec(&delayed_refs->procs_running_refs); return ret; } + atomic_add(ret, &delayed_refs->ref_seq); + count -= min_t(unsigned long, ret, count); if (count == 0) @@ -2576,6 +2623,11 @@ again: goto again; } out: + atomic_dec(&delayed_refs->procs_running_refs); + smp_mb(); + if (waitqueue_active(&delayed_refs->wait)) + wake_up(&delayed_refs->wait); + spin_unlock(&delayed_refs->lock); assert_qgroups_uptodate(trans); return 0; diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index a065dec0e330..1e7f176bd211 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -156,6 +156,9 @@ loop: spin_lock_init(&cur_trans->commit_lock); spin_lock_init(&cur_trans->delayed_refs.lock); + atomic_set(&cur_trans->delayed_refs.procs_running_refs, 0); + atomic_set(&cur_trans->delayed_refs.ref_seq, 0); + init_waitqueue_head(&cur_trans->delayed_refs.wait); INIT_LIST_HEAD(&cur_trans->pending_snapshots); list_add_tail(&cur_trans->list, &fs_info->trans_list); @@ -577,7 +580,7 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, if (!list_empty(&trans->new_bgs)) btrfs_create_pending_block_groups(trans, root); - while (count < 2) { + while (count < 1) { unsigned long cur = trans->delayed_ref_updates; trans->delayed_ref_updates = 0; if (cur && @@ -589,6 +592,7 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, } count++; } + btrfs_trans_release_metadata(trans, root); trans->block_rsv = NULL; -- cgit From 4f3b57832ba39223c6f8823d07b9fb206e282ced Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:25:35 +0800 Subject: xfs: add a helper to figure out the space log reservation per item Add a new helper xfs_calc_buf_res() to calcuate out the transaction space reservations per item. xfs_buf_log_overhead() is used to figure out the extra space for struct xfs_buf_log_format that gets written into the log for every buffer as well as a log opheader, i.e. struct xlog_op_header. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_trans.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) (limited to 'fs') diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 06ed520a767f..6c601ea77a9e 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -37,14 +37,45 @@ #include "xfs_extent_busy.h" #include "xfs_bmap.h" #include "xfs_quota.h" +#include "xfs_qm.h" #include "xfs_trans_priv.h" #include "xfs_trans_space.h" #include "xfs_inode_item.h" +#include "xfs_log_priv.h" +#include "xfs_buf_item.h" #include "xfs_trace.h" kmem_zone_t *xfs_trans_zone; kmem_zone_t *xfs_log_item_desc_zone; +/* + * A buffer has a format structure overhead in the log in addition + * to the data, so we need to take this into account when reserving + * space in a transaction for a buffer. Round the space required up + * to a multiple of 128 bytes so that we don't change the historical + * reservation that has been used for this overhead. + */ +STATIC uint +xfs_buf_log_overhead(void) +{ + return round_up(sizeof(struct xlog_op_header) + + sizeof(struct xfs_buf_log_format), 128); +} + +/* + * Calculate out transaction log reservation per item in bytes. + * + * The nbufs argument is used to indicate the number of items that + * will be changed in a transaction. size is used to tell how many + * bytes should be reserved per item. + */ +STATIC uint +xfs_calc_buf_res( + uint nbufs, + uint size) +{ + return nbufs * (size + xfs_buf_log_overhead()); +} /* * Various log reservation values. -- cgit From 4506a519f2a76775e6c236f44d50e9f79a7d3434 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 1 Feb 2013 20:36:03 +0000 Subject: GFS2: Split glock lru processing into two parts The intent here is to split the processing of the glock lru list into two parts, so that the selection of glocks and the disposal are separate functions. The plan is then, that further updates can then be made to these functions in the future to improve the selection of glocks and also the efficiency of glock disposal. The new feature which this patch brings is sorting the glocks to be disposed of into glock number (and thus also disk block number) order. Not all glocks will need i/o in order to dispose of them, but some will, and at least we'll generate mostly disk block order i/o now. Signed-off-by: Steven Whitehouse --- fs/gfs2/glock.c | 90 ++++++++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 67 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 3ad8fd36f8da..cf3515546739 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -30,6 +30,7 @@ #include #include #include +#include #include "gfs2.h" #include "incore.h" @@ -1376,50 +1377,93 @@ void gfs2_glock_complete(struct gfs2_glock *gl, int ret) gfs2_glock_put(gl); } +static int glock_cmp(void *priv, struct list_head *a, struct list_head *b) +{ + struct gfs2_glock *gla, *glb; + + gla = list_entry(a, struct gfs2_glock, gl_lru); + glb = list_entry(b, struct gfs2_glock, gl_lru); + + if (gla->gl_name.ln_number > glb->gl_name.ln_number) + return 1; + if (gla->gl_name.ln_number < glb->gl_name.ln_number) + return -1; + + return 0; +} + +/** + * gfs2_dispose_glock_lru - Demote a list of glocks + * @list: The list to dispose of + * + * Disposing of glocks may involve disk accesses, so that here we sort + * the glocks by number (i.e. disk location of the inodes) so that if + * there are any such accesses, they'll be sent in order (mostly). + * + * Must be called under the lru_lock, but may drop and retake this + * lock. While the lru_lock is dropped, entries may vanish from the + * list, but no new entries will appear on the list (since it is + * private) + */ + +static void gfs2_dispose_glock_lru(struct list_head *list) +__releases(&lru_lock) +__acquires(&lru_lock) +{ + struct gfs2_glock *gl; + + list_sort(NULL, list, glock_cmp); + + while(!list_empty(list)) { + gl = list_entry(list->next, struct gfs2_glock, gl_lru); + list_del_init(&gl->gl_lru); + clear_bit(GLF_LRU, &gl->gl_flags); + gfs2_glock_hold(gl); + spin_unlock(&lru_lock); + spin_lock(&gl->gl_spin); + if (demote_ok(gl)) + handle_callback(gl, LM_ST_UNLOCKED, 0); + WARN_ON(!test_and_clear_bit(GLF_LOCK, &gl->gl_flags)); + smp_mb__after_clear_bit(); + if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) + gfs2_glock_put_nolock(gl); + spin_unlock(&gl->gl_spin); + spin_lock(&lru_lock); + } +} + /** * gfs2_scan_glock_lru - Scan the LRU looking for locks to demote * @nr: The number of entries to scan * + * This function selects the entries on the LRU which are able to + * be demoted, and then kicks off the process by calling + * gfs2_dispose_glock_lru() above. */ static void gfs2_scan_glock_lru(int nr) { struct gfs2_glock *gl; - int may_demote; - int nr_skipped = 0; LIST_HEAD(skipped); + LIST_HEAD(dispose); spin_lock(&lru_lock); while(nr && !list_empty(&lru_list)) { gl = list_entry(lru_list.next, struct gfs2_glock, gl_lru); - list_del_init(&gl->gl_lru); - clear_bit(GLF_LRU, &gl->gl_flags); - atomic_dec(&lru_count); /* Test for being demotable */ if (!test_and_set_bit(GLF_LOCK, &gl->gl_flags)) { - gfs2_glock_hold(gl); - spin_unlock(&lru_lock); - spin_lock(&gl->gl_spin); - may_demote = demote_ok(gl); - if (may_demote) { - handle_callback(gl, LM_ST_UNLOCKED, 0); - nr--; - } - clear_bit(GLF_LOCK, &gl->gl_flags); - smp_mb__after_clear_bit(); - if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) - gfs2_glock_put_nolock(gl); - spin_unlock(&gl->gl_spin); - spin_lock(&lru_lock); + list_move(&gl->gl_lru, &dispose); + atomic_dec(&lru_count); + nr--; continue; } - nr_skipped++; - list_add(&gl->gl_lru, &skipped); - set_bit(GLF_LRU, &gl->gl_flags); + + list_move(&gl->gl_lru, &skipped); } list_splice(&skipped, &lru_list); - atomic_add(nr_skipped, &lru_count); + if (!list_empty(&dispose)) + gfs2_dispose_glock_lru(&dispose); spin_unlock(&lru_lock); } -- cgit From d2b47cfb26fe06002b8011707baac71a9ae8166f Mon Sep 17 00:00:00 2001 From: Bob Peterson Date: Fri, 1 Feb 2013 12:03:02 -0500 Subject: GFS2: Get a block reservation before resizing a file This patch allocates a block reservation structure before growing or shrinking a file. Without this structure, the grow or shink code can reference the bad pointer. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse --- fs/gfs2/bmap.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index d29d7793b211..df686d13a7d2 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -1288,6 +1288,10 @@ int gfs2_setattr_size(struct inode *inode, u64 newsize) inode_dio_wait(inode); + ret = gfs2_rs_alloc(GFS2_I(inode)); + if (ret) + return ret; + oldsize = inode->i_size; if (newsize >= oldsize) return do_grow(inode, newsize); -- cgit From 5b292ae3a951a58e32119d73c7ac8f5bec7395a3 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Fri, 1 Feb 2013 14:39:29 -0600 Subject: xfs: make use of xfs_calc_buf_res() in xfs_trans.c Refining the existing reservations with xfs_calc_buf_res() in xfs_trans.c Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_trans.c | 240 +++++++++++++++++++++++++---------------------------- 1 file changed, 112 insertions(+), 128 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 6c601ea77a9e..a81625c75a91 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -116,18 +116,15 @@ xfs_calc_write_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((mp->m_sb.sb_inodesize + - XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) + - 2 * mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 2) + - 128 * (4 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + - XFS_ALLOCFREE_LOG_COUNT(mp, 2))), - (2 * mp->m_sb.sb_sectsize + - 2 * mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 2) + - 128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))); + MAX((xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK), + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 2), + XFS_FSB_TO_B(mp, 1))), + (xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 2), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -148,18 +145,17 @@ xfs_calc_itruncate_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((mp->m_sb.sb_inodesize + - XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1) + - 128 * (2 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK))), - (4 * mp->m_sb.sb_sectsize + - 4 * mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 4) + - 128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4)) + - 128 * 5 + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); + MAX((xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + 1, + XFS_FSB_TO_B(mp, 1))), + (xfs_calc_buf_res(9, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 4), + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(5, 0) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(2 + XFS_IALLOC_BLOCKS(mp) + + mp->m_in_maxlevels, 0))); } /* @@ -179,14 +175,12 @@ xfs_calc_rename_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((4 * mp->m_sb.sb_inodesize + - 2 * XFS_DIROP_LOG_RES(mp) + - 128 * (4 + 2 * XFS_DIROP_LOG_COUNT(mp))), - (3 * mp->m_sb.sb_sectsize + - 3 * mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 3) + - 128 * (7 + XFS_ALLOCFREE_LOG_COUNT(mp, 3)))); + MAX((xfs_calc_buf_res(4, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(2 * XFS_DIROP_LOG_COUNT(mp), + XFS_FSB_TO_B(mp, 1))), + (xfs_calc_buf_res(7, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 3), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -206,15 +200,12 @@ xfs_calc_link_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((mp->m_sb.sb_inodesize + - mp->m_sb.sb_inodesize + - XFS_DIROP_LOG_RES(mp) + - 128 * (2 + XFS_DIROP_LOG_COUNT(mp))), - (mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); + MAX((xfs_calc_buf_res(2, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), + XFS_FSB_TO_B(mp, 1))), + (xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -234,15 +225,12 @@ xfs_calc_remove_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((mp->m_sb.sb_inodesize + - mp->m_sb.sb_inodesize + - XFS_DIROP_LOG_RES(mp) + - 128 * (2 + XFS_DIROP_LOG_COUNT(mp))), - (2 * mp->m_sb.sb_sectsize + - 2 * mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 2) + - 128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))); + MAX((xfs_calc_buf_res(2, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), + XFS_FSB_TO_B(mp, 1))), + (xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 2), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -264,18 +252,18 @@ xfs_calc_symlink_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((mp->m_sb.sb_inodesize + - mp->m_sb.sb_inodesize + - XFS_FSB_TO_B(mp, 1) + - XFS_DIROP_LOG_RES(mp) + - 1024 + - 128 * (4 + XFS_DIROP_LOG_COUNT(mp))), - (2 * mp->m_sb.sb_sectsize + - XFS_FSB_TO_B(mp, XFS_IALLOC_BLOCKS(mp)) + - XFS_FSB_TO_B(mp, mp->m_in_maxlevels) + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); + MAX((xfs_calc_buf_res(2, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(1, XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(1, 1024)), + (xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_IALLOC_BLOCKS(mp), + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(mp->m_in_maxlevels, + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -298,18 +286,19 @@ xfs_calc_create_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((mp->m_sb.sb_inodesize + - mp->m_sb.sb_inodesize + + MAX((xfs_calc_buf_res(2, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) + + (uint)XFS_FSB_TO_B(mp, 1) + + xfs_calc_buf_res(XFS_DIROP_LOG_COUNT(mp), + XFS_FSB_TO_B(mp, 1))), + (xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) + mp->m_sb.sb_sectsize + - XFS_FSB_TO_B(mp, 1) + - XFS_DIROP_LOG_RES(mp) + - 128 * (3 + XFS_DIROP_LOG_COUNT(mp))), - (3 * mp->m_sb.sb_sectsize + - XFS_FSB_TO_B(mp, XFS_IALLOC_BLOCKS(mp)) + - XFS_FSB_TO_B(mp, mp->m_in_maxlevels) + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + - XFS_ALLOCFREE_LOG_COUNT(mp, 1)))); + xfs_calc_buf_res(XFS_IALLOC_BLOCKS(mp), + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(mp->m_in_maxlevels, + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -337,16 +326,16 @@ xfs_calc_ifree_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - mp->m_sb.sb_inodesize + - mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_FSB_TO_B(mp, 1) + + xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(1, XFS_FSB_TO_B(mp, 1)) + MAX((__uint16_t)XFS_FSB_TO_B(mp, 1), XFS_INODE_CLUSTER_SIZE(mp)) + - 128 * 5 + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (2 + XFS_IALLOC_BLOCKS(mp) + mp->m_in_maxlevels + - XFS_ALLOCFREE_LOG_COUNT(mp, 1)); + xfs_calc_buf_res(1, 0) + + xfs_calc_buf_res(2 + XFS_IALLOC_BLOCKS(mp) + + mp->m_in_maxlevels, 0) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)); } /* @@ -374,9 +363,9 @@ STATIC uint xfs_calc_growdata_reservation( struct xfs_mount *mp) { - return mp->m_sb.sb_sectsize * 3 + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (3 + XFS_ALLOCFREE_LOG_COUNT(mp, 1)); + return xfs_calc_buf_res(3, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)); } /* @@ -393,12 +382,12 @@ STATIC uint xfs_calc_growrtalloc_reservation( struct xfs_mount *mp) { - return 2 * mp->m_sb.sb_sectsize + - XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK)) + - mp->m_sb.sb_inodesize + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (3 + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) + - XFS_ALLOCFREE_LOG_COUNT(mp, 1)); + return xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK), + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)); } /* @@ -410,7 +399,7 @@ STATIC uint xfs_calc_growrtzero_reservation( struct xfs_mount *mp) { - return mp->m_sb.sb_blocksize + 128; + return xfs_calc_buf_res(1, mp->m_sb.sb_blocksize); } /* @@ -427,11 +416,10 @@ STATIC uint xfs_calc_growrtfree_reservation( struct xfs_mount *mp) { - return mp->m_sb.sb_sectsize + - 2 * mp->m_sb.sb_inodesize + - mp->m_sb.sb_blocksize + - mp->m_rsumsize + - 128 * 5; + return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(2, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(1, mp->m_sb.sb_blocksize) + + xfs_calc_buf_res(1, mp->m_rsumsize); } /* @@ -442,7 +430,7 @@ STATIC uint xfs_calc_swrite_reservation( struct xfs_mount *mp) { - return mp->m_sb.sb_inodesize + 128; + return xfs_calc_buf_res(1, mp->m_sb.sb_inodesize); } /* @@ -452,7 +440,7 @@ xfs_calc_swrite_reservation( STATIC uint xfs_calc_writeid_reservation(xfs_mount_t *mp) { - return mp->m_sb.sb_inodesize + 128; + return xfs_calc_buf_res(1, mp->m_sb.sb_inodesize); } /* @@ -468,13 +456,13 @@ xfs_calc_addafork_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - mp->m_sb.sb_inodesize + - mp->m_sb.sb_sectsize * 2 + - mp->m_dirblksize + - XFS_FSB_TO_B(mp, XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1) + - XFS_ALLOCFREE_LOG_RES(mp, 1) + - 128 * (4 + XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1 + - XFS_ALLOCFREE_LOG_COUNT(mp, 1)); + xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(1, mp->m_dirblksize) + + xfs_calc_buf_res(XFS_DAENTER_BMAP1B(mp, XFS_DATA_FORK) + 1, + XFS_FSB_TO_B(mp, 1)) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 1), + XFS_FSB_TO_B(mp, 1)); } /* @@ -492,14 +480,12 @@ STATIC uint xfs_calc_attrinval_reservation( struct xfs_mount *mp) { - return MAX((mp->m_sb.sb_inodesize + - XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) + - 128 * (1 + XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))), - (4 * mp->m_sb.sb_sectsize + - 4 * mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 4) + - 128 * (9 + XFS_ALLOCFREE_LOG_COUNT(mp, 4)))); + return MAX((xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK), + XFS_FSB_TO_B(mp, 1))), + (xfs_calc_buf_res(9, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 4), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -517,10 +503,9 @@ xfs_calc_attrset_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - mp->m_sb.sb_inodesize + - mp->m_sb.sb_sectsize + - XFS_FSB_TO_B(mp, XFS_DA_NODE_MAXDEPTH) + - 128 * (2 + XFS_DA_NODE_MAXDEPTH); + xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_DA_NODE_MAXDEPTH, XFS_FSB_TO_B(mp, 1)); } /* @@ -539,16 +524,15 @@ xfs_calc_attrrm_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + - MAX((mp->m_sb.sb_inodesize + - XFS_FSB_TO_B(mp, XFS_DA_NODE_MAXDEPTH) + - XFS_FSB_TO_B(mp, XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) + - 128 * (1 + XFS_DA_NODE_MAXDEPTH + - XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK))), - (2 * mp->m_sb.sb_sectsize + - 2 * mp->m_sb.sb_sectsize + - mp->m_sb.sb_sectsize + - XFS_ALLOCFREE_LOG_RES(mp, 2) + - 128 * (5 + XFS_ALLOCFREE_LOG_COUNT(mp, 2)))); + MAX((xfs_calc_buf_res(1, mp->m_sb.sb_inodesize) + + xfs_calc_buf_res(XFS_DA_NODE_MAXDEPTH, + XFS_FSB_TO_B(mp, 1)) + + (uint)XFS_FSB_TO_B(mp, + XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK), 0)), + (xfs_calc_buf_res(5, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_ALLOCFREE_LOG_COUNT(mp, 2), + XFS_FSB_TO_B(mp, 1)))); } /* @@ -558,7 +542,7 @@ STATIC uint xfs_calc_clear_agi_bucket_reservation( struct xfs_mount *mp) { - return mp->m_sb.sb_sectsize + 128; + return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize); } /* -- cgit From b0c10b983a3e5cc35f239999df1b8bad1ba5b8f6 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:26:16 +0800 Subject: xfs: calculate XFS_TRANS_QM_SBCHANGE space log reservation at mount time The transaction log space for clearing/reseting the quota flags is calculated out at runtime, this patch can figure it out at mount time. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.c | 4 ++-- fs/xfs/xfs_mount.h | 1 + fs/xfs/xfs_trans.c | 12 ++++++++++++ fs/xfs/xfs_trans.h | 2 +- 4 files changed, 16 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 7d6df7c00c36..0db40205778a 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -1109,8 +1109,8 @@ xfs_mount_reset_sbqflags( return 0; tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE); - error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0, - XFS_DEFAULT_LOG_COUNT); + error = xfs_trans_reserve(tp, 0, XFS_QM_SBCHANGE_LOG_RES(mp), + 0, 0, XFS_DEFAULT_LOG_COUNT); if (error) { xfs_trans_cancel(tp, 0); xfs_alert(mp, "%s: Superblock update failed!", __func__); diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index bab8314507e4..4e9aa9524840 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -40,6 +40,7 @@ typedef struct xfs_trans_reservations { uint tr_growrtalloc; /* grow realtime allocations */ uint tr_growrtzero; /* grow realtime zeroing */ uint tr_growrtfree; /* grow realtime freeing */ + uint tr_qm_sbchange; /* change quota flags */ } xfs_trans_reservations_t; #ifndef __KERNEL__ diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index a81625c75a91..6ba62c0e24d1 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -545,6 +545,17 @@ xfs_calc_clear_agi_bucket_reservation( return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize); } +/* + * Clearing the quotaflags in the superblock. + * the super block for changing quota flags: sector size + */ +STATIC uint +xfs_calc_qm_sbchange_reservation( + struct xfs_mount *mp) +{ + return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize); +} + /* * Initialize the precomputed transaction reservation values * in the mount structure. @@ -576,6 +587,7 @@ xfs_trans_init( resp->tr_growrtalloc = xfs_calc_growrtalloc_reservation(mp); resp->tr_growrtzero = xfs_calc_growrtzero_reservation(mp); resp->tr_growrtfree = xfs_calc_growrtfree_reservation(mp); + resp->tr_qm_sbchange = xfs_calc_qm_sbchange_reservation(mp); } /* diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index c6c0601abd7a..a690f2803594 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -262,7 +262,7 @@ struct xfs_log_item_desc { (128 * (ext + (ext * XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))))) #define XFS_ATTRRM_LOG_RES(mp) ((mp)->m_reservations.tr_attrrm) #define XFS_CLEAR_AGI_BUCKET_LOG_RES(mp) ((mp)->m_reservations.tr_clearagi) - +#define XFS_QM_SBCHANGE_LOG_RES(mp) ((mp)->m_reservations.tr_qm_sbchange) /* * Various log count values. -- cgit From f910a8c620e21fdb53384ad741ff9e4995cc6e67 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:26:34 +0800 Subject: xfs: calculate xfs_qm_write_sb_changes() space log reservation at mount time For the transaction that write the incore superblock changes of quota flags to disk, it would reserve the same log space to clear/reset quota flags transaction, hence we can use XFS_TRANS_SBCHANGE_LOG_RES() for it as well. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_qm.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 60eff4763156..e5b5cf973781 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -1584,10 +1584,9 @@ xfs_qm_write_sb_changes( int error; tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SBCHANGE); - if ((error = xfs_trans_reserve(tp, 0, - mp->m_sb.sb_sectsize + 128, 0, - 0, - XFS_DEFAULT_LOG_COUNT))) { + error = xfs_trans_reserve(tp, 0, XFS_QM_SBCHANGE_LOG_RES(mp), + 0, 0, XFS_DEFAULT_LOG_COUNT); + if (error) { xfs_trans_cancel(tp, 0); return error; } -- cgit From f0f2df94faca43fd26f85af7e83df240777c8c37 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:26:49 +0800 Subject: xfs: calcuate XFS_TRANS_QM_SETQLIM space log reservation at mount time For adjusting quota limits transactions, we calculate out the log space reservation at runtime, this patch does it at mount time. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.h | 1 + fs/xfs/xfs_qm_syscalls.c | 5 +++-- fs/xfs/xfs_trans.c | 12 ++++++++++++ fs/xfs/xfs_trans.h | 1 + 4 files changed, 17 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 4e9aa9524840..9baa7e0fbd45 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -41,6 +41,7 @@ typedef struct xfs_trans_reservations { uint tr_growrtzero; /* grow realtime zeroing */ uint tr_growrtfree; /* grow realtime freeing */ uint tr_qm_sbchange; /* change quota flags */ + uint tr_qm_setqlim; /* adjust quota limits */ } xfs_trans_reservations_t; #ifndef __KERNEL__ diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 33d9c2b6f555..1476b6fd4ed5 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -490,8 +490,9 @@ xfs_qm_scall_setqlim( return 0; tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SETQLIM); - if ((error = xfs_trans_reserve(tp, 0, sizeof(xfs_disk_dquot_t) + 128, - 0, 0, XFS_DEFAULT_LOG_COUNT))) { + error = xfs_trans_reserve(tp, 0, XFS_QM_SETQLIM_LOG_RES(mp), + 0, 0, XFS_DEFAULT_LOG_COUNT); + if (error) { xfs_trans_cancel(tp, 0); return (error); } diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 6ba62c0e24d1..1838850de145 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -556,6 +556,17 @@ xfs_calc_qm_sbchange_reservation( return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize); } +/* + * Adjusting quota limits. + * the xfs_disk_dquot_t: sizeof(struct xfs_disk_dquot) + */ +STATIC uint +xfs_calc_qm_setqlim_reservation( + struct xfs_mount *mp) +{ + return xfs_calc_buf_res(1, sizeof(struct xfs_disk_dquot)); +} + /* * Initialize the precomputed transaction reservation values * in the mount structure. @@ -588,6 +599,7 @@ xfs_trans_init( resp->tr_growrtzero = xfs_calc_growrtzero_reservation(mp); resp->tr_growrtfree = xfs_calc_growrtfree_reservation(mp); resp->tr_qm_sbchange = xfs_calc_qm_sbchange_reservation(mp); + resp->tr_qm_setqlim = xfs_calc_qm_setqlim_reservation(mp); } /* diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index a690f2803594..93a0ec73c6fe 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -263,6 +263,7 @@ struct xfs_log_item_desc { #define XFS_ATTRRM_LOG_RES(mp) ((mp)->m_reservations.tr_attrrm) #define XFS_CLEAR_AGI_BUCKET_LOG_RES(mp) ((mp)->m_reservations.tr_clearagi) #define XFS_QM_SBCHANGE_LOG_RES(mp) ((mp)->m_reservations.tr_qm_sbchange) +#define XFS_QM_SETQLIM_LOG_RES(mp) ((mp)->m_reservations.tr_qm_setqlim) /* * Various log count values. -- cgit From 4800104438a4467ffa5ae1e51d5a59c0f64e5f9a Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:04 +0800 Subject: xfs: calculate XFS_TRANS_QM_DQALLOC space log reservation at mount time The disk quota allocation log space reservation is calcuated at runtime, this patch does it at mount time. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_dquot.c | 12 +++--------- fs/xfs/xfs_mount.h | 1 + fs/xfs/xfs_trans.c | 15 +++++++++++++++ fs/xfs/xfs_trans.h | 1 + 4 files changed, 20 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c index 9e1bf5294c91..8025eb23ad72 100644 --- a/fs/xfs/xfs_dquot.c +++ b/fs/xfs/xfs_dquot.c @@ -612,15 +612,9 @@ xfs_qm_dqread( if (flags & XFS_QMOPT_DQALLOC) { tp = xfs_trans_alloc(mp, XFS_TRANS_QM_DQALLOC); error = xfs_trans_reserve(tp, XFS_QM_DQALLOC_SPACE_RES(mp), - XFS_WRITE_LOG_RES(mp) + - /* - * Round the chunklen up to the next multiple - * of 128 (buf log item chunk size)). - */ - BBTOB(mp->m_quotainfo->qi_dqchunklen) - 1 + 128, - 0, - XFS_TRANS_PERM_LOG_RES, - XFS_WRITE_LOG_COUNT); + XFS_QM_DQALLOC_LOG_RES(mp), 0, + XFS_TRANS_PERM_LOG_RES, + XFS_WRITE_LOG_COUNT); if (error) goto error1; cancelflags = XFS_TRANS_RELEASE_LOG_RES; diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 9baa7e0fbd45..fc500c6c8800 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -42,6 +42,7 @@ typedef struct xfs_trans_reservations { uint tr_growrtfree; /* grow realtime freeing */ uint tr_qm_sbchange; /* change quota flags */ uint tr_qm_setqlim; /* adjust quota limits */ + uint tr_qm_dqalloc; /* allocate quota on disk */ } xfs_trans_reservations_t; #ifndef __KERNEL__ diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 1838850de145..6dd2c043efcd 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -567,6 +567,20 @@ xfs_calc_qm_setqlim_reservation( return xfs_calc_buf_res(1, sizeof(struct xfs_disk_dquot)); } +/* + * Allocating quota on disk if needed. + * the write transaction log space: XFS_WRITE_LOG_RES(mp) + * the unit of quota allocation: one system block size + */ +STATIC uint +xfs_calc_qm_dqalloc_reservation( + struct xfs_mount *mp) +{ + return XFS_WRITE_LOG_RES(mp) + + xfs_calc_buf_res(1, + XFS_FSB_TO_B(mp, XFS_DQUOT_CLUSTER_SIZE_FSB) - 1); +} + /* * Initialize the precomputed transaction reservation values * in the mount structure. @@ -600,6 +614,7 @@ xfs_trans_init( resp->tr_growrtfree = xfs_calc_growrtfree_reservation(mp); resp->tr_qm_sbchange = xfs_calc_qm_sbchange_reservation(mp); resp->tr_qm_setqlim = xfs_calc_qm_setqlim_reservation(mp); + resp->tr_qm_dqalloc = xfs_calc_qm_dqalloc_reservation(mp); } /* diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index 93a0ec73c6fe..04575db806d4 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -264,6 +264,7 @@ struct xfs_log_item_desc { #define XFS_CLEAR_AGI_BUCKET_LOG_RES(mp) ((mp)->m_reservations.tr_clearagi) #define XFS_QM_SBCHANGE_LOG_RES(mp) ((mp)->m_reservations.tr_qm_sbchange) #define XFS_QM_SETQLIM_LOG_RES(mp) ((mp)->m_reservations.tr_qm_setqlim) +#define XFS_QM_DQALLOC_LOG_RES(mp) ((mp)->m_reservations.tr_qm_dqalloc) /* * Various log count values. -- cgit From a1bd9557544d59140c4ac87fe405069b9e1aaf99 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:15 +0800 Subject: xfs: calculate XFS_TRANS_QM_QUOTAOFF space log reservation at mount time Convert the calculation of quota off transaction log space reservation from runtime to mount time. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.h | 1 + fs/xfs/xfs_qm_syscalls.c | 10 +++------- fs/xfs/xfs_trans.c | 14 ++++++++++++++ fs/xfs/xfs_trans.h | 1 + 4 files changed, 19 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index fc500c6c8800..4f5e148ffd1c 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -43,6 +43,7 @@ typedef struct xfs_trans_reservations { uint tr_qm_sbchange; /* change quota flags */ uint tr_qm_setqlim; /* adjust quota limits */ uint tr_qm_dqalloc; /* allocate quota on disk */ + uint tr_qm_quotaoff; /* turn quota off */ } xfs_trans_reservations_t; #ifndef __KERNEL__ diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 1476b6fd4ed5..4605f8914cb4 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -672,14 +672,10 @@ xfs_qm_log_quotaoff( uint oldsbqflag=0; tp = xfs_trans_alloc(mp, XFS_TRANS_QM_QUOTAOFF); - if ((error = xfs_trans_reserve(tp, 0, - sizeof(xfs_qoff_logitem_t) * 2 + - mp->m_sb.sb_sectsize + 128, - 0, - 0, - XFS_DEFAULT_LOG_COUNT))) { + error = xfs_trans_reserve(tp, 0, XFS_QM_QUOTAOFF_LOG_RES(mp), + 0, 0, XFS_DEFAULT_LOG_COUNT); + if (error) goto error0; - } qoffi = xfs_trans_get_qoff_item(tp, NULL, flags & XFS_ALL_QUOTA_ACCT); xfs_trans_log_quotaoff_item(tp, qoffi); diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 6dd2c043efcd..a557c82e58f0 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -581,6 +581,19 @@ xfs_calc_qm_dqalloc_reservation( XFS_FSB_TO_B(mp, XFS_DQUOT_CLUSTER_SIZE_FSB) - 1); } +/* + * Turning off quotas. + * the xfs_qoff_logitem_t: sizeof(struct xfs_qoff_logitem) * 2 + * the superblock for the quota flags: sector size + */ +STATIC uint +xfs_calc_qm_quotaoff_reservation( + struct xfs_mount *mp) +{ + return sizeof(struct xfs_qoff_logitem) * 2 + + xfs_calc_buf_res(1, mp->m_sb.sb_sectsize); +} + /* * Initialize the precomputed transaction reservation values * in the mount structure. @@ -615,6 +628,7 @@ xfs_trans_init( resp->tr_qm_sbchange = xfs_calc_qm_sbchange_reservation(mp); resp->tr_qm_setqlim = xfs_calc_qm_setqlim_reservation(mp); resp->tr_qm_dqalloc = xfs_calc_qm_dqalloc_reservation(mp); + resp->tr_qm_quotaoff = xfs_calc_qm_quotaoff_reservation(mp); } /* diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index 04575db806d4..1e103dad17d1 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -265,6 +265,7 @@ struct xfs_log_item_desc { #define XFS_QM_SBCHANGE_LOG_RES(mp) ((mp)->m_reservations.tr_qm_sbchange) #define XFS_QM_SETQLIM_LOG_RES(mp) ((mp)->m_reservations.tr_qm_setqlim) #define XFS_QM_DQALLOC_LOG_RES(mp) ((mp)->m_reservations.tr_qm_dqalloc) +#define XFS_QM_QUOTAOFF_LOG_RES(mp) ((mp)->m_reservations.tr_qm_quotaoff) /* * Various log count values. -- cgit From 762d7ba657ed4a0934b4da7dcef058012f252e0f Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:21 +0800 Subject: xfs: calculate XFS_TRANS_QM_QUOTAOFF_END space log reservation at mount time Convert the calculation for end of quotaoff log space reservation from runtime to mount time. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.h | 1 + fs/xfs/xfs_qm_syscalls.c | 5 +++-- fs/xfs/xfs_trans.c | 12 ++++++++++++ fs/xfs/xfs_trans.h | 1 + 4 files changed, 17 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 4f5e148ffd1c..b46009dc7b13 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -44,6 +44,7 @@ typedef struct xfs_trans_reservations { uint tr_qm_setqlim; /* adjust quota limits */ uint tr_qm_dqalloc; /* allocate quota on disk */ uint tr_qm_quotaoff; /* turn quota off */ + uint tr_qm_equotaoff;/* end of turn quota off */ } xfs_trans_reservations_t; #ifndef __KERNEL__ diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 4605f8914cb4..cf9a34051e07 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -639,8 +639,9 @@ xfs_qm_log_quotaoff_end( tp = xfs_trans_alloc(mp, XFS_TRANS_QM_QUOTAOFF_END); - if ((error = xfs_trans_reserve(tp, 0, sizeof(xfs_qoff_logitem_t) * 2, - 0, 0, XFS_DEFAULT_LOG_COUNT))) { + error = xfs_trans_reserve(tp, 0, XFS_QM_QUOTAOFF_END_LOG_RES(mp), + 0, 0, XFS_DEFAULT_LOG_COUNT); + if (error) { xfs_trans_cancel(tp, 0); return (error); } diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index a557c82e58f0..84643fa5e465 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -594,6 +594,17 @@ xfs_calc_qm_quotaoff_reservation( xfs_calc_buf_res(1, mp->m_sb.sb_sectsize); } +/* + * End of turning off quotas. + * the xfs_qoff_logitem_t: sizeof(struct xfs_qoff_logitem) * 2 + */ +STATIC uint +xfs_calc_qm_quotaoff_end_reservation( + struct xfs_mount *mp) +{ + return sizeof(struct xfs_qoff_logitem) * 2; +} + /* * Initialize the precomputed transaction reservation values * in the mount structure. @@ -629,6 +640,7 @@ xfs_trans_init( resp->tr_qm_setqlim = xfs_calc_qm_setqlim_reservation(mp); resp->tr_qm_dqalloc = xfs_calc_qm_dqalloc_reservation(mp); resp->tr_qm_quotaoff = xfs_calc_qm_quotaoff_reservation(mp); + resp->tr_qm_equotaoff = xfs_calc_qm_quotaoff_end_reservation(mp); } /* diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index 1e103dad17d1..57339e404cf6 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -266,6 +266,7 @@ struct xfs_log_item_desc { #define XFS_QM_SETQLIM_LOG_RES(mp) ((mp)->m_reservations.tr_qm_setqlim) #define XFS_QM_DQALLOC_LOG_RES(mp) ((mp)->m_reservations.tr_qm_dqalloc) #define XFS_QM_QUOTAOFF_LOG_RES(mp) ((mp)->m_reservations.tr_qm_quotaoff) +#define XFS_QM_QUOTAOFF_END_LOG_RES(mp) ((mp)->m_reservations.tr_qm_equotaoff) /* * Various log count values. -- cgit From a7bd794a0f489a66ad595f2bcab0eac8f232e409 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:25 +0800 Subject: xfs: introduce XFS_SB_LOG_RES() for transactions that modify sb on disk Introduce a new transaction space reservation XFS_SB_LOG_RES() for those transactions that need to modify the superblock on disk. Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.h | 1 + fs/xfs/xfs_trans.c | 12 ++++++++++++ fs/xfs/xfs_trans.h | 1 + 3 files changed, 14 insertions(+) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index b46009dc7b13..2e30a9a84f9f 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -45,6 +45,7 @@ typedef struct xfs_trans_reservations { uint tr_qm_dqalloc; /* allocate quota on disk */ uint tr_qm_quotaoff; /* turn quota off */ uint tr_qm_equotaoff;/* end of turn quota off */ + uint tr_sb; /* modify superblock */ } xfs_trans_reservations_t; #ifndef __KERNEL__ diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 84643fa5e465..72da2aa08421 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -605,6 +605,17 @@ xfs_calc_qm_quotaoff_end_reservation( return sizeof(struct xfs_qoff_logitem) * 2; } +/* + * Syncing the incore super block changes to disk. + * the super block to reflect the changes: sector size + */ +STATIC uint +xfs_calc_sb_reservation( + struct xfs_mount *mp) +{ + return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize); +} + /* * Initialize the precomputed transaction reservation values * in the mount structure. @@ -641,6 +652,7 @@ xfs_trans_init( resp->tr_qm_dqalloc = xfs_calc_qm_dqalloc_reservation(mp); resp->tr_qm_quotaoff = xfs_calc_qm_quotaoff_reservation(mp); resp->tr_qm_equotaoff = xfs_calc_qm_quotaoff_end_reservation(mp); + resp->tr_sb = xfs_calc_sb_reservation(mp); } /* diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index 57339e404cf6..d06919e28904 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -267,6 +267,7 @@ struct xfs_log_item_desc { #define XFS_QM_DQALLOC_LOG_RES(mp) ((mp)->m_reservations.tr_qm_dqalloc) #define XFS_QM_QUOTAOFF_LOG_RES(mp) ((mp)->m_reservations.tr_qm_quotaoff) #define XFS_QM_QUOTAOFF_END_LOG_RES(mp) ((mp)->m_reservations.tr_qm_equotaoff) +#define XFS_SB_LOG_RES(mp) ((mp)->m_reservations.tr_sb) /* * Various log count values. -- cgit From e457274b60a3e3046d9451a199826281ce92023d Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:31 +0800 Subject: xfs: make use of XFS_SB_LOG_RES() at xfs_log_sbcount() Make use of XFS_SB_LOG_RES() at xfs_log_sbcount(). Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 0db40205778a..a67046191824 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -1583,8 +1583,8 @@ xfs_log_sbcount(xfs_mount_t *mp) return 0; tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_COUNT, KM_SLEEP); - error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0, - XFS_DEFAULT_LOG_COUNT); + error = xfs_trans_reserve(tp, 0, XFS_SB_LOG_RES(mp), 0, 0, + XFS_DEFAULT_LOG_COUNT); if (error) { xfs_trans_cancel(tp, 0); return error; -- cgit From 5166ab06558cfe4a3745252a91ad6b495ba49290 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:39 +0800 Subject: xfs: make use of XFS_SB_LOG_RES() at xfs_mount_log_sb() Make use of XFS_SB_LOG_RES() at xfs_mount_log_sb(). Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_mount.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index a67046191824..3806088a8f77 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -1945,8 +1945,8 @@ xfs_mount_log_sb( XFS_SB_VERSIONNUM)); tp = xfs_trans_alloc(mp, XFS_TRANS_SB_UNIT); - error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0, - XFS_DEFAULT_LOG_COUNT); + error = xfs_trans_reserve(tp, 0, XFS_SB_LOG_RES(mp), 0, 0, + XFS_DEFAULT_LOG_COUNT); if (error) { xfs_trans_cancel(tp, 0); return error; -- cgit From 762c585b187cde93bbb8cadd3cb3871bbae1b75a Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:46 +0800 Subject: xfs: make use of XFS_SB_LOG_RES() at xfs_fs_log_dummy() Make use of XFS_SB_LOG_RES() at xfs_fs_log_dummy(). Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_fsops.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 94eaeedc5498..2866b8c78b7a 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -709,8 +709,8 @@ xfs_fs_log_dummy( int error; tp = _xfs_trans_alloc(mp, XFS_TRANS_DUMMY1, KM_SLEEP); - error = xfs_trans_reserve(tp, 0, mp->m_sb.sb_sectsize + 128, 0, 0, - XFS_DEFAULT_LOG_COUNT); + error = xfs_trans_reserve(tp, 0, XFS_SB_LOG_RES(mp), 0, 0, + XFS_DEFAULT_LOG_COUNT); if (error) { xfs_trans_cancel(tp, 0); return error; -- cgit From a21cd503678286c56b1d0cca1c99349a4aa042f4 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Mon, 28 Jan 2013 21:27:53 +0800 Subject: xfs: refactor space log reservation for XFS_TRANS_ATTR_SET Currently, we calculate the attribute set transaction log space reservation at runtime in two parts: 1) XFS_ATTRSET_LOG_RES() which is calcuated out at mount time. 2) ((ext * (mp)->m_sb.sb_sectsize) + \ (ext * XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))) + \ (128 * (ext + (ext * XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK)))))) which is calculated out at runtime since it depend on the given extent length in blocks. This patch renamed XFS_ATTRSET_LOG_RES(mp) to XFS_ATTRSETM_LOG_RES(mp) to indicate that it is figured out at mount time. Introduce XFS_ATTRSETRT_LOG_RES(mp) which would be used to calculate out the unit of the log space reservation for one block. In this way, the total runtime space for the given extent length can be figured out by: XFS_ATTRSETM_LOG_RES(mp) + XFS_ATTRSETRT_LOG_RES(mp) * ext Signed-off-by: Jie Liu CC: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_attr.c | 9 ++++++--- fs/xfs/xfs_mount.h | 3 ++- fs/xfs/xfs_trans.c | 28 ++++++++++++++++++++++++---- fs/xfs/xfs_trans.h | 11 ++++------- 4 files changed, 36 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index aaf472532b3c..888683844d98 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -300,9 +300,12 @@ xfs_attr_set_int( if (rsvd) args.trans->t_flags |= XFS_TRANS_RESERVE; - if ((error = xfs_trans_reserve(args.trans, args.total, - XFS_ATTRSET_LOG_RES(mp, args.total), 0, - XFS_TRANS_PERM_LOG_RES, XFS_ATTRSET_LOG_COUNT))) { + error = xfs_trans_reserve(args.trans, args.total, + XFS_ATTRSETM_LOG_RES(mp) + + XFS_ATTRSETRT_LOG_RES(mp) * args.total, + 0, XFS_TRANS_PERM_LOG_RES, + XFS_ATTRSET_LOG_COUNT); + if (error) { xfs_trans_cancel(args.trans, 0); return(error); } diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 2e30a9a84f9f..bc907061d392 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -34,7 +34,8 @@ typedef struct xfs_trans_reservations { uint tr_addafork; /* cvt inode to attributed trans */ uint tr_writeid; /* write setuid/setgid file */ uint tr_attrinval; /* attr fork buffer invalidation */ - uint tr_attrset; /* set/create an attribute */ + uint tr_attrsetm; /* set/create an attribute at mount time */ + uint tr_attrsetrt; /* set/create an attribute at runtime */ uint tr_attrrm; /* remove an attribute */ uint tr_clearagi; /* clear bad agi unlinked ino bucket */ uint tr_growrtalloc; /* grow realtime allocations */ diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 72da2aa08421..2fd7c1ff1d21 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -489,17 +489,18 @@ xfs_calc_attrinval_reservation( } /* - * Setting an attribute. + * Setting an attribute at mount time. * the inode getting the attribute * the superblock for allocations * the agfs extents are allocated from * the attribute btree * max depth * the inode allocation btree * Since attribute transaction space is dependent on the size of the attribute, - * the calculation is done partially at mount time and partially at runtime. + * the calculation is done partially at mount time and partially at runtime(see + * below). */ STATIC uint -xfs_calc_attrset_reservation( +xfs_calc_attrsetm_reservation( struct xfs_mount *mp) { return XFS_DQUOT_LOGRES(mp) + @@ -508,6 +509,24 @@ xfs_calc_attrset_reservation( xfs_calc_buf_res(XFS_DA_NODE_MAXDEPTH, XFS_FSB_TO_B(mp, 1)); } +/* + * Setting an attribute at runtime, transaction space unit per block. + * the superblock for allocations: sector size + * the inode bmap btree could join or split: max depth * block size + * Since the runtime attribute transaction space is dependent on the total + * blocks needed for the 1st bmap, here we calculate out the space unit for + * one block so that the caller could figure out the total space according + * to the attibute extent length in blocks by: ext * XFS_ATTRSETRT_LOG_RES(mp). + */ +STATIC uint +xfs_calc_attrsetrt_reservation( + struct xfs_mount *mp) +{ + return xfs_calc_buf_res(1, mp->m_sb.sb_sectsize) + + xfs_calc_buf_res(XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK), + XFS_FSB_TO_B(mp, 1)); +} + /* * Removing an attribute. * the inode: inode size @@ -641,7 +660,8 @@ xfs_trans_init( resp->tr_writeid = xfs_calc_writeid_reservation(mp); resp->tr_addafork = xfs_calc_addafork_reservation(mp); resp->tr_attrinval = xfs_calc_attrinval_reservation(mp); - resp->tr_attrset = xfs_calc_attrset_reservation(mp); + resp->tr_attrsetm = xfs_calc_attrsetm_reservation(mp); + resp->tr_attrsetrt = xfs_calc_attrsetrt_reservation(mp); resp->tr_attrrm = xfs_calc_attrrm_reservation(mp); resp->tr_clearagi = xfs_calc_clear_agi_bucket_reservation(mp); resp->tr_growrtalloc = xfs_calc_growrtalloc_reservation(mp); diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index d06919e28904..cd29f6171021 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -252,15 +252,12 @@ struct xfs_log_item_desc { * as long as SWRITE logs the entire inode core */ #define XFS_FSYNC_TS_LOG_RES(mp) ((mp)->m_reservations.tr_swrite) -#define XFS_WRITEID_LOG_RES(mp) ((mp)->m_reservations.tr_swrite) +#define XFS_WRITEID_LOG_RES(mp) ((mp)->m_reservations.tr_swrite) #define XFS_ADDAFORK_LOG_RES(mp) ((mp)->m_reservations.tr_addafork) #define XFS_ATTRINVAL_LOG_RES(mp) ((mp)->m_reservations.tr_attrinval) -#define XFS_ATTRSET_LOG_RES(mp, ext) \ - ((mp)->m_reservations.tr_attrset + \ - (ext * (mp)->m_sb.sb_sectsize) + \ - (ext * XFS_FSB_TO_B((mp), XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))) + \ - (128 * (ext + (ext * XFS_BM_MAXLEVELS(mp, XFS_ATTR_FORK))))) -#define XFS_ATTRRM_LOG_RES(mp) ((mp)->m_reservations.tr_attrrm) +#define XFS_ATTRSETM_LOG_RES(mp) ((mp)->m_reservations.tr_attrsetm) +#define XFS_ATTRSETRT_LOG_RES(mp) ((mp)->m_reservations.tr_attrsetrt) +#define XFS_ATTRRM_LOG_RES(mp) ((mp)->m_reservations.tr_attrrm) #define XFS_CLEAR_AGI_BUCKET_LOG_RES(mp) ((mp)->m_reservations.tr_clearagi) #define XFS_QM_SBCHANGE_LOG_RES(mp) ((mp)->m_reservations.tr_qm_sbchange) #define XFS_QM_SETQLIM_LOG_RES(mp) ((mp)->m_reservations.tr_qm_setqlim) -- cgit From 524c19ebc961799b1ec126e4b063b941a70275e5 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Fri, 1 Feb 2013 20:07:21 -0500 Subject: ext4: use WARN in ext4_alloc_blocks Use WARN rather than printk followed by WARN_ON(1), for conciseness. A simplified version of the semantic patch that makes this transformation is as follows: (http://coccinelle.lip6.fr/) // @@ expression list es; @@ -printk( +WARN(1, es); -WARN_ON(1); // Signed-off-by: Julia Lawall Signed-off-by: "Theodore Ts'o" --- fs/ext4/indirect.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index bdd20231e66c..193281098fb6 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -358,9 +358,8 @@ static int ext4_alloc_blocks(handle_t *handle, struct inode *inode, * for the first direct block */ new_blocks[index] = current_block; - printk(KERN_INFO "%s returned more blocks than " + WARN(1, KERN_INFO "%s returned more blocks than " "requested\n", __func__); - WARN_ON(1); break; } } -- cgit From 87e698734b9e618276c797092ccdd91da292d10e Mon Sep 17 00:00:00 2001 From: Akria Fujita Date: Fri, 1 Feb 2013 20:52:46 -0500 Subject: ext4: fix smatch warning in move_extent.c's mext_replace_branches() Commit 2147b1a6a48 resulted in a new smatch warning: > fs/ext4/move_extent.c:693 mext_replace_branches() > warn: variable dereferenced before check 'dext' (see line 683) Fix this by adding a check to make sure dext is non-NULL before we derefrence it. Signed-off-by: Akria Fujita [ modified by tytso to make sure an ext4_error is called ] Signed-off-by: "Theodore Ts'o" --- fs/ext4/move_extent.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index d9cc5ee42f53..e4cdb5188f34 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -681,6 +681,8 @@ mext_replace_branches(handle_t *handle, struct inode *orig_inode, depth = ext_depth(donor_inode); dext = donor_path[depth].p_ext; + if (unlikely(!dext)) + goto missing_donor_extent; tmp_dext = *dext; *err = mext_calc_swap_extents(&tmp_dext, &tmp_oext, orig_off, @@ -691,7 +693,8 @@ mext_replace_branches(handle_t *handle, struct inode *orig_inode, /* Loop for the donor extents */ while (1) { /* The extent for donor must be found. */ - if (!dext) { + if (unlikely(!dext)) { + missing_donor_extent: EXT4_ERROR_INODE(donor_inode, "The extent for donor must be found"); *err = -EIO; -- cgit From f1167009711032b0d747ec89a632a626c901a1ad Mon Sep 17 00:00:00 2001 From: Niu Yawei Date: Fri, 1 Feb 2013 21:31:27 -0500 Subject: ext4: fix race in ext4_mb_add_n_trim() In ext4_mb_add_n_trim(), lg_prealloc_lock should be taken when changing the lg_prealloc_list. Signed-off-by: Niu Yawei Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/ext4/mballoc.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 1bf6fe785c4f..061727acd990 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4136,7 +4136,7 @@ static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac) /* The max size of hash table is PREALLOC_TB_SIZE */ order = PREALLOC_TB_SIZE - 1; /* Add the prealloc space to lg */ - rcu_read_lock(); + spin_lock(&lg->lg_prealloc_lock); list_for_each_entry_rcu(tmp_pa, &lg->lg_prealloc_list[order], pa_inode_list) { spin_lock(&tmp_pa->pa_lock); @@ -4160,12 +4160,12 @@ static void ext4_mb_add_n_trim(struct ext4_allocation_context *ac) if (!added) list_add_tail_rcu(&pa->pa_inode_list, &lg->lg_prealloc_list[order]); - rcu_read_unlock(); + spin_unlock(&lg->lg_prealloc_lock); /* Now trim the list to be not more than 8 elements */ if (lg_prealloc_count > 8) { ext4_mb_discard_lg_preallocations(sb, lg, - order, lg_prealloc_count); + order, lg_prealloc_count); return; } return ; -- cgit From 0e79537d30f9dd66ccef70f1c7571594088e30be Mon Sep 17 00:00:00 2001 From: Cong Ding Date: Fri, 1 Feb 2013 22:33:21 -0500 Subject: ext4: reduce one "if" comparison in ext4_dirhash() It is unnecessary to check i<4 after the loop; just do it before the break. Signed-off-by: Cong Ding Signed-off-by: "Theodore Ts'o" --- fs/ext4/hash.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/hash.c b/fs/ext4/hash.c index fa8e4911d354..3d586f02883e 100644 --- a/fs/ext4/hash.c +++ b/fs/ext4/hash.c @@ -155,11 +155,11 @@ int ext4fs_dirhash(const char *name, int len, struct dx_hash_info *hinfo) /* Check to see if the seed is all zero's */ if (hinfo->seed) { for (i = 0; i < 4; i++) { - if (hinfo->seed[i]) + if (hinfo->seed[i]) { + memcpy(buf, hinfo->seed, sizeof(buf)); break; + } } - if (i < 4) - memcpy(buf, hinfo->seed, sizeof(buf)); } switch (hinfo->hash_version) { -- cgit From 0efb3b23002fea3ed996597783939617e50523ec Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sat, 2 Feb 2013 22:52:19 -0500 Subject: ext4: move several mount options to standard handling loop Several mount option (resuid, resgid, journal_dev, journal_ioprio) are currently handled before we enter standard option handling loop. I don't see a reason for this so move them to normal handling loop to make things more regular. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/super.c | 69 +++++++++++++++++++++++++++++++-------------------------- 1 file changed, 38 insertions(+), 31 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index aac52738eb60..69e5e70cfdee 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1444,6 +1444,10 @@ static const struct mount_opts { {Opt_inode_readahead_blks, 0, MOPT_GTE0}, {Opt_init_itable, 0, MOPT_GTE0}, {Opt_stripe, 0, MOPT_GTE0}, + {Opt_resuid, 0, MOPT_GTE0}, + {Opt_resgid, 0, MOPT_GTE0}, + {Opt_journal_dev, 0, MOPT_GTE0}, + {Opt_journal_ioprio, 0, MOPT_GTE0}, {Opt_data_journal, EXT4_MOUNT_JOURNAL_DATA, MOPT_DATAJ}, {Opt_data_ordered, EXT4_MOUNT_ORDERED_DATA, MOPT_DATAJ}, {Opt_data_writeback, EXT4_MOUNT_WRITEBACK_DATA, MOPT_DATAJ}, @@ -1496,8 +1500,6 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, else if (token == Opt_offgrpjquota) return clear_qf_name(sb, GRPQUOTA); #endif - if (args->from && match_int(args, &arg)) - return -1; switch (token) { case Opt_noacl: case Opt_nouser_xattr: @@ -1509,46 +1511,19 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, ext4_msg(sb, KERN_WARNING, "Ignoring removed %s option", opt); return 1; - case Opt_resuid: - uid = make_kuid(current_user_ns(), arg); - if (!uid_valid(uid)) { - ext4_msg(sb, KERN_ERR, "Invalid uid value %d", arg); - return -1; - } - sbi->s_resuid = uid; - return 1; - case Opt_resgid: - gid = make_kgid(current_user_ns(), arg); - if (!gid_valid(gid)) { - ext4_msg(sb, KERN_ERR, "Invalid gid value %d", arg); - return -1; - } - sbi->s_resgid = gid; - return 1; case Opt_abort: sbi->s_mount_flags |= EXT4_MF_FS_ABORTED; return 1; case Opt_i_version: sb->s_flags |= MS_I_VERSION; return 1; - case Opt_journal_dev: - if (is_remount) { - ext4_msg(sb, KERN_ERR, - "Cannot specify journal on remount"); - return -1; - } - *journal_devnum = arg; - return 1; - case Opt_journal_ioprio: - if (arg < 0 || arg > 7) - return -1; - *journal_ioprio = IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, arg); - return 1; } for (m = ext4_mount_opts; m->token != Opt_err; m++) { if (token != m->token) continue; + if (args->from && match_int(args, &arg)) + return -1; if (args->from && (m->flags & MOPT_GTE0) && (arg < 0)) return -1; if (m->flags & MOPT_EXPLICIT) @@ -1592,6 +1567,38 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, sbi->s_max_dir_size_kb = arg; } else if (token == Opt_stripe) { sbi->s_stripe = arg; + } else if (token == Opt_resuid) { + uid = make_kuid(current_user_ns(), arg); + if (!uid_valid(uid)) { + ext4_msg(sb, KERN_ERR, + "Invalid uid value %d", arg); + return -1; + } + sbi->s_resuid = uid; + } else if (token == Opt_resgid) { + gid = make_kgid(current_user_ns(), arg); + if (!gid_valid(gid)) { + ext4_msg(sb, KERN_ERR, + "Invalid gid value %d", arg); + return -1; + } + sbi->s_resgid = gid; + } else if (token == Opt_journal_dev) { + if (is_remount) { + ext4_msg(sb, KERN_ERR, + "Cannot specify journal on remount"); + return -1; + } + *journal_devnum = arg; + } else if (token == Opt_journal_ioprio) { + if (arg > 7) { + ext4_msg(sb, KERN_ERR, + "Invalid journal IO priority" + " (must be 0-7)"); + return -1; + } + *journal_ioprio = + IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, arg); } else if (m->flags & MOPT_DATAJ) { if (is_remount) { if (!sbi->s_journal) -- cgit From 5f3633e36b8a67c02d2d10cde5f056f9e5da49a1 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sat, 2 Feb 2013 23:09:36 -0500 Subject: ext4: make mount option parsing loop more logical The loop looking for correct mount option entry is more logical if it is written rewritten as an empty loop looking for correct option entry and then code handling the option. It also saves one level of indentation for a lot of code so we can join a couple of split lines. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/super.c | 230 ++++++++++++++++++++++++++++---------------------------- 1 file changed, 113 insertions(+), 117 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 69e5e70cfdee..9b36c11baad9 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1508,8 +1508,7 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, case Opt_sb: return 1; /* handled by get_sb_block() */ case Opt_removed: - ext4_msg(sb, KERN_WARNING, - "Ignoring removed %s option", opt); + ext4_msg(sb, KERN_WARNING, "Ignoring removed %s option", opt); return 1; case Opt_abort: sbi->s_mount_flags |= EXT4_MF_FS_ABORTED; @@ -1519,132 +1518,129 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, return 1; } - for (m = ext4_mount_opts; m->token != Opt_err; m++) { - if (token != m->token) - continue; - if (args->from && match_int(args, &arg)) + for (m = ext4_mount_opts; m->token != Opt_err; m++) + if (token == m->token) + break; + + if (m->token == Opt_err) { + ext4_msg(sb, KERN_ERR, "Unrecognized mount option \"%s\" " + "or missing value", opt); + return -1; + } + + if (args->from && match_int(args, &arg)) + return -1; + if (args->from && (m->flags & MOPT_GTE0) && (arg < 0)) + return -1; + if (m->flags & MOPT_EXPLICIT) + set_opt2(sb, EXPLICIT_DELALLOC); + if (m->flags & MOPT_CLEAR_ERR) + clear_opt(sb, ERRORS_MASK); + if (token == Opt_noquota && sb_any_quota_loaded(sb)) { + ext4_msg(sb, KERN_ERR, "Cannot change quota " + "options when quota turned on"); + return -1; + } + + if (m->flags & MOPT_NOSUPPORT) { + ext4_msg(sb, KERN_ERR, "%s option not supported", opt); + } else if (token == Opt_commit) { + if (arg == 0) + arg = JBD2_DEFAULT_MAX_COMMIT_AGE; + sbi->s_commit_interval = HZ * arg; + } else if (token == Opt_max_batch_time) { + if (arg == 0) + arg = EXT4_DEF_MAX_BATCH_TIME; + sbi->s_max_batch_time = arg; + } else if (token == Opt_min_batch_time) { + sbi->s_min_batch_time = arg; + } else if (token == Opt_inode_readahead_blks) { + if (arg > (1 << 30)) return -1; - if (args->from && (m->flags & MOPT_GTE0) && (arg < 0)) + if (arg && !is_power_of_2(arg)) { + ext4_msg(sb, KERN_ERR, "EXT4-fs: inode_readahead_blks" + " must be a power of 2"); return -1; - if (m->flags & MOPT_EXPLICIT) - set_opt2(sb, EXPLICIT_DELALLOC); - if (m->flags & MOPT_CLEAR_ERR) - clear_opt(sb, ERRORS_MASK); - if (token == Opt_noquota && sb_any_quota_loaded(sb)) { - ext4_msg(sb, KERN_ERR, "Cannot change quota " - "options when quota turned on"); + } + sbi->s_inode_readahead_blks = arg; + } else if (token == Opt_init_itable) { + set_opt(sb, INIT_INODE_TABLE); + if (!args->from) + arg = EXT4_DEF_LI_WAIT_MULT; + sbi->s_li_wait_mult = arg; + } else if (token == Opt_max_dir_size_kb) { + sbi->s_max_dir_size_kb = arg; + } else if (token == Opt_stripe) { + sbi->s_stripe = arg; + } else if (token == Opt_resuid) { + uid = make_kuid(current_user_ns(), arg); + if (!uid_valid(uid)) { + ext4_msg(sb, KERN_ERR, "Invalid uid value %d", arg); return -1; } - - if (m->flags & MOPT_NOSUPPORT) { - ext4_msg(sb, KERN_ERR, "%s option not supported", opt); - } else if (token == Opt_commit) { - if (arg == 0) - arg = JBD2_DEFAULT_MAX_COMMIT_AGE; - sbi->s_commit_interval = HZ * arg; - } else if (token == Opt_max_batch_time) { - if (arg == 0) - arg = EXT4_DEF_MAX_BATCH_TIME; - sbi->s_max_batch_time = arg; - } else if (token == Opt_min_batch_time) { - sbi->s_min_batch_time = arg; - } else if (token == Opt_inode_readahead_blks) { - if (arg > (1 << 30)) - return -1; - if (arg && !is_power_of_2(arg)) { - ext4_msg(sb, KERN_ERR, - "EXT4-fs: inode_readahead_blks" - " must be a power of 2"); - return -1; - } - sbi->s_inode_readahead_blks = arg; - } else if (token == Opt_init_itable) { - set_opt(sb, INIT_INODE_TABLE); - if (!args->from) - arg = EXT4_DEF_LI_WAIT_MULT; - sbi->s_li_wait_mult = arg; - } else if (token == Opt_max_dir_size_kb) { - sbi->s_max_dir_size_kb = arg; - } else if (token == Opt_stripe) { - sbi->s_stripe = arg; - } else if (token == Opt_resuid) { - uid = make_kuid(current_user_ns(), arg); - if (!uid_valid(uid)) { - ext4_msg(sb, KERN_ERR, - "Invalid uid value %d", arg); - return -1; - } - sbi->s_resuid = uid; - } else if (token == Opt_resgid) { - gid = make_kgid(current_user_ns(), arg); - if (!gid_valid(gid)) { - ext4_msg(sb, KERN_ERR, - "Invalid gid value %d", arg); - return -1; - } - sbi->s_resgid = gid; - } else if (token == Opt_journal_dev) { - if (is_remount) { - ext4_msg(sb, KERN_ERR, - "Cannot specify journal on remount"); - return -1; - } - *journal_devnum = arg; - } else if (token == Opt_journal_ioprio) { - if (arg > 7) { + sbi->s_resuid = uid; + } else if (token == Opt_resgid) { + gid = make_kgid(current_user_ns(), arg); + if (!gid_valid(gid)) { + ext4_msg(sb, KERN_ERR, "Invalid gid value %d", arg); + return -1; + } + sbi->s_resgid = gid; + } else if (token == Opt_journal_dev) { + if (is_remount) { + ext4_msg(sb, KERN_ERR, + "Cannot specify journal on remount"); + return -1; + } + *journal_devnum = arg; + } else if (token == Opt_journal_ioprio) { + if (arg > 7) { + ext4_msg(sb, KERN_ERR, "Invalid journal IO priority" + " (must be 0-7)"); + return -1; + } + *journal_ioprio = + IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, arg); + } else if (m->flags & MOPT_DATAJ) { + if (is_remount) { + if (!sbi->s_journal) + ext4_msg(sb, KERN_WARNING, "Remounting file system with no journal so ignoring journalled data option"); + else if (test_opt(sb, DATA_FLAGS) != m->mount_opt) { ext4_msg(sb, KERN_ERR, - "Invalid journal IO priority" - " (must be 0-7)"); - return -1; - } - *journal_ioprio = - IOPRIO_PRIO_VALUE(IOPRIO_CLASS_BE, arg); - } else if (m->flags & MOPT_DATAJ) { - if (is_remount) { - if (!sbi->s_journal) - ext4_msg(sb, KERN_WARNING, "Remounting file system with no journal so ignoring journalled data option"); - else if (test_opt(sb, DATA_FLAGS) != - m->mount_opt) { - ext4_msg(sb, KERN_ERR, "Cannot change data mode on remount"); - return -1; - } - } else { - clear_opt(sb, DATA_FLAGS); - sbi->s_mount_opt |= m->mount_opt; - } -#ifdef CONFIG_QUOTA - } else if (m->flags & MOPT_QFMT) { - if (sb_any_quota_loaded(sb) && - sbi->s_jquota_fmt != m->mount_opt) { - ext4_msg(sb, KERN_ERR, "Cannot " - "change journaled quota options " - "when quota turned on"); return -1; } - sbi->s_jquota_fmt = m->mount_opt; -#endif } else { - if (!args->from) - arg = 1; - if (m->flags & MOPT_CLEAR) - arg = !arg; - else if (unlikely(!(m->flags & MOPT_SET))) { - ext4_msg(sb, KERN_WARNING, - "buggy handling of option %s", opt); - WARN_ON(1); - return -1; - } - if (arg != 0) - sbi->s_mount_opt |= m->mount_opt; - else - sbi->s_mount_opt &= ~m->mount_opt; + clear_opt(sb, DATA_FLAGS); + sbi->s_mount_opt |= m->mount_opt; } - return 1; +#ifdef CONFIG_QUOTA + } else if (m->flags & MOPT_QFMT) { + if (sb_any_quota_loaded(sb) && + sbi->s_jquota_fmt != m->mount_opt) { + ext4_msg(sb, KERN_ERR, "Cannot change journaled " + "quota options when quota turned on"); + return -1; + } + sbi->s_jquota_fmt = m->mount_opt; +#endif + } else { + if (!args->from) + arg = 1; + if (m->flags & MOPT_CLEAR) + arg = !arg; + else if (unlikely(!(m->flags & MOPT_SET))) { + ext4_msg(sb, KERN_WARNING, + "buggy handling of option %s", opt); + WARN_ON(1); + return -1; + } + if (arg != 0) + sbi->s_mount_opt |= m->mount_opt; + else + sbi->s_mount_opt &= ~m->mount_opt; } - ext4_msg(sb, KERN_ERR, "Unrecognized mount option \"%s\" " - "or missing value", opt); - return -1; + return 1; } static int parse_options(char *options, struct super_block *sb, -- cgit From e33e60eaed5353c9e6863124ad1081a38640db4b Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sat, 2 Feb 2013 23:14:31 -0500 Subject: ext4: print error when argument of inode_readahead_blk is invalid If argument of inode_readahead_blk is too big, we just bail out without printing any error. Fix this since it could confuse users. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/super.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 9b36c11baad9..b68f4b671bc7 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1555,11 +1555,10 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, } else if (token == Opt_min_batch_time) { sbi->s_min_batch_time = arg; } else if (token == Opt_inode_readahead_blks) { - if (arg > (1 << 30)) - return -1; - if (arg && !is_power_of_2(arg)) { - ext4_msg(sb, KERN_ERR, "EXT4-fs: inode_readahead_blks" - " must be a power of 2"); + if (arg && (arg > (1 << 30) || !is_power_of_2(arg))) { + ext4_msg(sb, KERN_ERR, + "EXT4-fs: inode_readahead_blks must be " + "0 or a power of 2 smaller than 2^31"); return -1; } sbi->s_inode_readahead_blks = arg; -- cgit From 8dc0aa8cf0f7b51e6c7c342e6f1e61520fb94222 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 2 Feb 2013 23:38:39 -0500 Subject: ext4: check incompatible mount options while mounting ext2/3 Check for incompatible mount options when using the ext4 file system driver to mount ext2 or ext3 file systems. Signed-off-by: "Theodore Ts'o" --- fs/ext4/super.c | 47 +++++++++++++++++++++++++++++++++++------------ 1 file changed, 35 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index b68f4b671bc7..2e1f94704b1f 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1406,6 +1406,9 @@ static int clear_qf_name(struct super_block *sb, int qtype) #define MOPT_QFMT MOPT_NOSUPPORT #endif #define MOPT_DATAJ 0x0080 +#define MOPT_NO_EXT2 0x0100 +#define MOPT_NO_EXT3 0x0200 +#define MOPT_EXT4_ONLY (MOPT_NO_EXT2 | MOPT_NO_EXT3) static const struct mount_opts { int token; @@ -1418,21 +1421,29 @@ static const struct mount_opts { {Opt_nogrpid, EXT4_MOUNT_GRPID, MOPT_CLEAR}, {Opt_block_validity, EXT4_MOUNT_BLOCK_VALIDITY, MOPT_SET}, {Opt_noblock_validity, EXT4_MOUNT_BLOCK_VALIDITY, MOPT_CLEAR}, - {Opt_dioread_nolock, EXT4_MOUNT_DIOREAD_NOLOCK, MOPT_SET}, - {Opt_dioread_lock, EXT4_MOUNT_DIOREAD_NOLOCK, MOPT_CLEAR}, + {Opt_dioread_nolock, EXT4_MOUNT_DIOREAD_NOLOCK, + MOPT_EXT4_ONLY | MOPT_SET}, + {Opt_dioread_lock, EXT4_MOUNT_DIOREAD_NOLOCK, + MOPT_EXT4_ONLY | MOPT_CLEAR}, {Opt_discard, EXT4_MOUNT_DISCARD, MOPT_SET}, {Opt_nodiscard, EXT4_MOUNT_DISCARD, MOPT_CLEAR}, - {Opt_delalloc, EXT4_MOUNT_DELALLOC, MOPT_SET | MOPT_EXPLICIT}, - {Opt_nodelalloc, EXT4_MOUNT_DELALLOC, MOPT_CLEAR | MOPT_EXPLICIT}, - {Opt_journal_checksum, EXT4_MOUNT_JOURNAL_CHECKSUM, MOPT_SET}, + {Opt_delalloc, EXT4_MOUNT_DELALLOC, + MOPT_EXT4_ONLY | MOPT_SET | MOPT_EXPLICIT}, + {Opt_nodelalloc, EXT4_MOUNT_DELALLOC, + MOPT_EXT4_ONLY | MOPT_CLEAR | MOPT_EXPLICIT}, + {Opt_journal_checksum, EXT4_MOUNT_JOURNAL_CHECKSUM, + MOPT_EXT4_ONLY | MOPT_SET}, {Opt_journal_async_commit, (EXT4_MOUNT_JOURNAL_ASYNC_COMMIT | - EXT4_MOUNT_JOURNAL_CHECKSUM), MOPT_SET}, - {Opt_noload, EXT4_MOUNT_NOLOAD, MOPT_SET}, + EXT4_MOUNT_JOURNAL_CHECKSUM), + MOPT_EXT4_ONLY | MOPT_SET}, + {Opt_noload, EXT4_MOUNT_NOLOAD, MOPT_NO_EXT2 | MOPT_SET}, {Opt_err_panic, EXT4_MOUNT_ERRORS_PANIC, MOPT_SET | MOPT_CLEAR_ERR}, {Opt_err_ro, EXT4_MOUNT_ERRORS_RO, MOPT_SET | MOPT_CLEAR_ERR}, {Opt_err_cont, EXT4_MOUNT_ERRORS_CONT, MOPT_SET | MOPT_CLEAR_ERR}, - {Opt_data_err_abort, EXT4_MOUNT_DATA_ERR_ABORT, MOPT_SET}, - {Opt_data_err_ignore, EXT4_MOUNT_DATA_ERR_ABORT, MOPT_CLEAR}, + {Opt_data_err_abort, EXT4_MOUNT_DATA_ERR_ABORT, + MOPT_NO_EXT2 | MOPT_SET}, + {Opt_data_err_ignore, EXT4_MOUNT_DATA_ERR_ABORT, + MOPT_NO_EXT2 | MOPT_CLEAR}, {Opt_barrier, EXT4_MOUNT_BARRIER, MOPT_SET}, {Opt_nobarrier, EXT4_MOUNT_BARRIER, MOPT_CLEAR}, {Opt_noauto_da_alloc, EXT4_MOUNT_NO_AUTO_DA_ALLOC, MOPT_SET}, @@ -1448,9 +1459,10 @@ static const struct mount_opts { {Opt_resgid, 0, MOPT_GTE0}, {Opt_journal_dev, 0, MOPT_GTE0}, {Opt_journal_ioprio, 0, MOPT_GTE0}, - {Opt_data_journal, EXT4_MOUNT_JOURNAL_DATA, MOPT_DATAJ}, - {Opt_data_ordered, EXT4_MOUNT_ORDERED_DATA, MOPT_DATAJ}, - {Opt_data_writeback, EXT4_MOUNT_WRITEBACK_DATA, MOPT_DATAJ}, + {Opt_data_journal, EXT4_MOUNT_JOURNAL_DATA, MOPT_NO_EXT2 | MOPT_DATAJ}, + {Opt_data_ordered, EXT4_MOUNT_ORDERED_DATA, MOPT_NO_EXT2 | MOPT_DATAJ}, + {Opt_data_writeback, EXT4_MOUNT_WRITEBACK_DATA, + MOPT_NO_EXT2 | MOPT_DATAJ}, {Opt_user_xattr, EXT4_MOUNT_XATTR_USER, MOPT_SET}, {Opt_nouser_xattr, EXT4_MOUNT_XATTR_USER, MOPT_CLEAR}, #ifdef CONFIG_EXT4_FS_POSIX_ACL @@ -1528,6 +1540,17 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, return -1; } + if ((m->flags & MOPT_NO_EXT2) && IS_EXT2_SB(sb)) { + ext4_msg(sb, KERN_ERR, + "Mount option \"%s\" incompatible with ext2", opt); + return -1; + } + if ((m->flags & MOPT_NO_EXT3) && IS_EXT3_SB(sb)) { + ext4_msg(sb, KERN_ERR, + "Mount option \"%s\" incompatible with ext3", opt); + return -1; + } + if (args->from && match_int(args, &arg)) return -1; if (args->from && (m->flags & MOPT_GTE0) && (arg < 0)) -- cgit From f482e1b4a4abf926507a4c1c6317acd3e7aa61b7 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 25 Dec 2012 20:24:38 -0500 Subject: switch compat_sys_open* to COMPAT_SYSCALL_DEFINE --- fs/compat.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/compat.c b/fs/compat.c index 015e1e1f87c6..f9e2fe3794fa 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1278,8 +1278,7 @@ compat_sys_vmsplice(int fd, const struct compat_iovec __user *iov32, * Exactly like fs/open.c:sys_open(), except that it doesn't set the * O_LARGEFILE flag. */ -asmlinkage long -compat_sys_open(const char __user *filename, int flags, umode_t mode) +COMPAT_SYSCALL_DEFINE3(open, const char __user *, filename, int, flags, umode_t, mode) { return do_sys_open(AT_FDCWD, filename, flags, mode); } @@ -1288,8 +1287,7 @@ compat_sys_open(const char __user *filename, int flags, umode_t mode) * Exactly like fs/open.c:sys_openat(), except that it doesn't set the * O_LARGEFILE flag. */ -asmlinkage long -compat_sys_openat(unsigned int dfd, const char __user *filename, int flags, umode_t mode) +COMPAT_SYSCALL_DEFINE4(openat, int, dfd, const char __user *, filename, int, flags, umode_t, mode) { return do_sys_open(dfd, filename, flags, mode); } @@ -1785,9 +1783,8 @@ asmlinkage long compat_sys_timerfd_gettime(int ufd, * Exactly like fs/open.c:sys_open_by_handle_at(), except that it * doesn't set the O_LARGEFILE flag. */ -asmlinkage long -compat_sys_open_by_handle_at(int mountdirfd, - struct file_handle __user *handle, int flags) +COMPAT_SYSCALL_DEFINE3(open_by_handle_at, int, mountdirfd, + struct file_handle __user *, handle, int, flags) { return do_handle_open(mountdirfd, handle, flags); } -- cgit From 9d94b9e2f354f79461aa674e75b0926d0e768db6 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 27 Dec 2012 16:52:33 -0500 Subject: switch timerfd compat syscalls to COMPAT_SYSCALL_DEFINE ... and move them over to fs/timerfd.c. Cleaner and easier that way... Signed-off-by: Al Viro --- fs/compat.c | 41 ----------------------------- fs/timerfd.c | 85 ++++++++++++++++++++++++++++++++++++++++++++++-------------- 2 files changed, 66 insertions(+), 60 deletions(-) (limited to 'fs') diff --git a/fs/compat.c b/fs/compat.c index f9e2fe3794fa..fe40fde29111 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1737,47 +1737,6 @@ asmlinkage long compat_sys_signalfd(int ufd, } #endif /* CONFIG_SIGNALFD */ -#ifdef CONFIG_TIMERFD - -asmlinkage long compat_sys_timerfd_settime(int ufd, int flags, - const struct compat_itimerspec __user *utmr, - struct compat_itimerspec __user *otmr) -{ - int error; - struct itimerspec t; - struct itimerspec __user *ut; - - if (get_compat_itimerspec(&t, utmr)) - return -EFAULT; - ut = compat_alloc_user_space(2 * sizeof(struct itimerspec)); - if (copy_to_user(&ut[0], &t, sizeof(t))) - return -EFAULT; - error = sys_timerfd_settime(ufd, flags, &ut[0], &ut[1]); - if (!error && otmr) - error = (copy_from_user(&t, &ut[1], sizeof(struct itimerspec)) || - put_compat_itimerspec(otmr, &t)) ? -EFAULT: 0; - - return error; -} - -asmlinkage long compat_sys_timerfd_gettime(int ufd, - struct compat_itimerspec __user *otmr) -{ - int error; - struct itimerspec t; - struct itimerspec __user *ut; - - ut = compat_alloc_user_space(sizeof(struct itimerspec)); - error = sys_timerfd_gettime(ufd, ut); - if (!error) - error = (copy_from_user(&t, ut, sizeof(struct itimerspec)) || - put_compat_itimerspec(otmr, &t)) ? -EFAULT: 0; - - return error; -} - -#endif /* CONFIG_TIMERFD */ - #ifdef CONFIG_FHANDLE /* * Exactly like fs/open.c:sys_open_by_handle_at(), except that it diff --git a/fs/timerfd.c b/fs/timerfd.c index d03822bbf190..0e606b12a59d 100644 --- a/fs/timerfd.c +++ b/fs/timerfd.c @@ -22,6 +22,7 @@ #include #include #include +#include #include struct timerfd_ctx { @@ -278,21 +279,17 @@ SYSCALL_DEFINE2(timerfd_create, int, clockid, int, flags) return ufd; } -SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, - const struct itimerspec __user *, utmr, - struct itimerspec __user *, otmr) +static int do_timerfd_settime(int ufd, int flags, + const struct itimerspec *new, + struct itimerspec *old) { struct fd f; struct timerfd_ctx *ctx; - struct itimerspec ktmr, kotmr; int ret; - if (copy_from_user(&ktmr, utmr, sizeof(ktmr))) - return -EFAULT; - if ((flags & ~TFD_SETTIME_FLAGS) || - !timespec_valid(&ktmr.it_value) || - !timespec_valid(&ktmr.it_interval)) + !timespec_valid(&new->it_value) || + !timespec_valid(&new->it_interval)) return -EINVAL; ret = timerfd_fget(ufd, &f); @@ -323,27 +320,23 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, if (ctx->expired && ctx->tintv.tv64) hrtimer_forward_now(&ctx->tmr, ctx->tintv); - kotmr.it_value = ktime_to_timespec(timerfd_get_remaining(ctx)); - kotmr.it_interval = ktime_to_timespec(ctx->tintv); + old->it_value = ktime_to_timespec(timerfd_get_remaining(ctx)); + old->it_interval = ktime_to_timespec(ctx->tintv); /* * Re-program the timer to the new value ... */ - ret = timerfd_setup(ctx, flags, &ktmr); + ret = timerfd_setup(ctx, flags, new); spin_unlock_irq(&ctx->wqh.lock); fdput(f); - if (otmr && copy_to_user(otmr, &kotmr, sizeof(kotmr))) - return -EFAULT; - return ret; } -SYSCALL_DEFINE2(timerfd_gettime, int, ufd, struct itimerspec __user *, otmr) +static int do_timerfd_gettime(int ufd, struct itimerspec *t) { struct fd f; struct timerfd_ctx *ctx; - struct itimerspec kotmr; int ret = timerfd_fget(ufd, &f); if (ret) return ret; @@ -356,11 +349,65 @@ SYSCALL_DEFINE2(timerfd_gettime, int, ufd, struct itimerspec __user *, otmr) hrtimer_forward_now(&ctx->tmr, ctx->tintv) - 1; hrtimer_restart(&ctx->tmr); } - kotmr.it_value = ktime_to_timespec(timerfd_get_remaining(ctx)); - kotmr.it_interval = ktime_to_timespec(ctx->tintv); + t->it_value = ktime_to_timespec(timerfd_get_remaining(ctx)); + t->it_interval = ktime_to_timespec(ctx->tintv); spin_unlock_irq(&ctx->wqh.lock); fdput(f); + return 0; +} + +SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, + const struct itimerspec __user *, utmr, + struct itimerspec __user *, otmr) +{ + struct itimerspec new, old; + int ret; + + if (copy_from_user(&new, utmr, sizeof(new))) + return -EFAULT; + ret = do_timerfd_settime(ufd, flags, &new, &old); + if (ret) + return ret; + if (otmr && copy_to_user(otmr, &old, sizeof(old))) + return -EFAULT; + + return ret; +} +SYSCALL_DEFINE2(timerfd_gettime, int, ufd, struct itimerspec __user *, otmr) +{ + struct itimerspec kotmr; + int ret = do_timerfd_gettime(ufd, &kotmr); + if (ret) + return ret; return copy_to_user(otmr, &kotmr, sizeof(kotmr)) ? -EFAULT: 0; } +#ifdef COMPAT +COMPAT_SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, + const struct itimerspec __user *, utmr, + struct itimerspec __user *, otmr) +{ + struct itimerspec new, old; + int ret; + + if (get_compat_itimerspec(&new, utmr)) + return -EFAULT; + ret = do_timerfd_settime(ufd, flags, &new, &old); + if (ret) + return ret; + if (otmr && put_compat_itimerspec(otmr, &old)) + return -EFAULT; + return ret; +} + +COMPAT_SYSCALL_DEFINE2(timerfd_gettime, int, ufd, + struct itimerspec __user *, otmr) +{ + struct itimerspec kotmr; + int ret = do_timerfd_gettime(ufd, &kotmr); + if (ret) + return ret; + return put_compat_itimerspec(otmr, &t) ? -EFAULT: 0; +} +#endif -- cgit From 2928f0d0c5ebd6c9605c0d98207a44376387c298 Mon Sep 17 00:00:00 2001 From: Adam Thomas Date: Sat, 2 Feb 2013 22:32:31 +0000 Subject: UBIFS: fix use of freed ubifs_orphan objects The last orphan in the cnext list has its cnext set to NULL. Because of that, ubifs_delete_orphan assumes that it is not on the cnext list and frees it immediately instead of adding it to the dnext list. The freed orphan is later modified by write_orph_node. This can cause various inconsistencies including directory entries that cannot be removed and this error: UBIFS error (pid 20685): layout_cnodes: LPT out of space at LEB 14:129009 needing 17, done_ltab 1, done_lsave 1 This is a regression introduced by "7074e5eb UBIFS: remove invalid reference to list iterator variable". This change adds an explicit flag to ubifs_orphan indicating whether it is pending commit. Signed-off-by: Adam Thomas Reviewed-by: Adrian Hunter Cc: stable@vger.kernel.org # v3.6+ Signed-off-by: Artem Bityutskiy --- fs/ubifs/orphan.c | 7 ++++++- fs/ubifs/ubifs.h | 4 +++- 2 files changed, 9 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c index 769701ccb5c9..8534d9c6492f 100644 --- a/fs/ubifs/orphan.c +++ b/fs/ubifs/orphan.c @@ -132,7 +132,7 @@ void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum) (unsigned long)inum); return; } - if (o->cnext) { + if (o->cmt) { o->dnext = c->orph_dnext; c->orph_dnext = o; spin_unlock(&c->orphan_lock); @@ -172,7 +172,9 @@ int ubifs_orphan_start_commit(struct ubifs_info *c) last = &c->orph_cnext; list_for_each_entry(orphan, &c->orph_new, new_list) { ubifs_assert(orphan->new); + ubifs_assert(!orphan->cmt); orphan->new = 0; + orphan->cmt = 1; *last = orphan; last = &orphan->cnext; } @@ -299,7 +301,9 @@ static int write_orph_node(struct ubifs_info *c, int atomic) cnext = c->orph_cnext; for (i = 0; i < cnt; i++) { orphan = cnext; + ubifs_assert(orphan->cmt); orph->inos[i] = cpu_to_le64(orphan->inum); + orphan->cmt = 0; cnext = orphan->cnext; orphan->cnext = NULL; } @@ -378,6 +382,7 @@ static int consolidate(struct ubifs_info *c) list_for_each_entry(orphan, &c->orph_list, list) { if (orphan->new) continue; + orphan->cmt = 1; *last = orphan; last = &orphan->cnext; cnt += 1; diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index d133c276fe05..c16fff7271d3 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -904,6 +904,7 @@ struct ubifs_budget_req { * @dnext: next orphan to delete * @inum: inode number * @new: %1 => added since the last commit, otherwise %0 + * @cmt: %1 => commit pending, otherwise %0 */ struct ubifs_orphan { struct rb_node rb; @@ -912,7 +913,8 @@ struct ubifs_orphan { struct ubifs_orphan *cnext; struct ubifs_orphan *dnext; ino_t inum; - int new; + unsigned new:1; + unsigned cmt:1; }; /** -- cgit From 8afd500cb52a5d00bab4525dd5a560d199f979b9 Mon Sep 17 00:00:00 2001 From: Adam Thomas Date: Sat, 2 Feb 2013 22:35:08 +0000 Subject: UBIFS: fix double free of ubifs_orphan objects The last orphan in the dnext list has its dnext set to NULL. Because of that, ubifs_delete_orphan assumes that it is not on the dnext list and frees it immediately instead ignoring it as a second delete. The orphan is later freed again by erase_deleted. This change adds an explicit flag to ubifs_orphan indicating whether it is pending delete. Signed-off-by: Adam Thomas Signed-off-by: Artem Bityutskiy Cc: stable@vger.kernel.org --- fs/ubifs/orphan.c | 5 ++++- fs/ubifs/ubifs.h | 2 ++ 2 files changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c index 8534d9c6492f..ba32da3fe08a 100644 --- a/fs/ubifs/orphan.c +++ b/fs/ubifs/orphan.c @@ -126,13 +126,14 @@ void ubifs_delete_orphan(struct ubifs_info *c, ino_t inum) else if (inum > o->inum) p = p->rb_right; else { - if (o->dnext) { + if (o->del) { spin_unlock(&c->orphan_lock); dbg_gen("deleted twice ino %lu", (unsigned long)inum); return; } if (o->cmt) { + o->del = 1; o->dnext = c->orph_dnext; c->orph_dnext = o; spin_unlock(&c->orphan_lock); @@ -447,6 +448,7 @@ static void erase_deleted(struct ubifs_info *c) orphan = dnext; dnext = orphan->dnext; ubifs_assert(!orphan->new); + ubifs_assert(orphan->del); rb_erase(&orphan->rb, &c->orph_tree); list_del(&orphan->list); c->tot_orphans -= 1; @@ -536,6 +538,7 @@ static int insert_dead_orphan(struct ubifs_info *c, ino_t inum) rb_link_node(&orphan->rb, parent, p); rb_insert_color(&orphan->rb, &c->orph_tree); list_add_tail(&orphan->list, &c->orph_list); + orphan->del = 1; orphan->dnext = c->orph_dnext; c->orph_dnext = orphan; dbg_mnt("ino %lu, new %d, tot %d", (unsigned long)inum, diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index c16fff7271d3..b2babce4d70f 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -905,6 +905,7 @@ struct ubifs_budget_req { * @inum: inode number * @new: %1 => added since the last commit, otherwise %0 * @cmt: %1 => commit pending, otherwise %0 + * @del: %1 => delete pending, otherwise %0 */ struct ubifs_orphan { struct rb_node rb; @@ -915,6 +916,7 @@ struct ubifs_orphan { ino_t inum; unsigned new:1; unsigned cmt:1; + unsigned del:1; }; /** -- cgit From 7b9e8522a65886d8ae168547a67c3617b6ba83f1 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:07 -0500 Subject: nfsd: fix IPv6 address handling in the DRC Currently, it only stores the first 16 bytes of any address. struct sockaddr_in6 is 28 bytes however, so we're currently ignoring the last 12 bytes of the address. Expand the c_addr field to a sockaddr_in6, and cast it to a sockaddr_in as necessary. Also fix the comparitor to use the existing RPC helpers for this. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/cache.h | 6 +++++- fs/nfsd/nfscache.c | 7 +++++-- 2 files changed, 10 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/cache.h b/fs/nfsd/cache.h index 93cc9d34c459..2cac76c63b97 100644 --- a/fs/nfsd/cache.h +++ b/fs/nfsd/cache.h @@ -12,6 +12,10 @@ /* * Representation of a reply cache entry. + * + * Note that we use a sockaddr_in6 to hold the address instead of the more + * typical sockaddr_storage. This is for space reasons, since sockaddr_storage + * is much larger than a sockaddr_in6. */ struct svc_cacherep { struct hlist_node c_hash; @@ -20,7 +24,7 @@ struct svc_cacherep { unsigned char c_state, /* unused, inprog, done */ c_type, /* status, buffer */ c_secure : 1; /* req came from port < 1024 */ - struct sockaddr_in c_addr; + struct sockaddr_in6 c_addr; __be32 c_xid; u32 c_prot; u32 c_proc; diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 2cbac34a55da..5dd9ec2a177f 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -9,6 +9,7 @@ */ #include +#include #include "nfsd.h" #include "cache.h" @@ -146,7 +147,8 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) xid == rp->c_xid && proc == rp->c_proc && proto == rp->c_prot && vers == rp->c_vers && time_before(jiffies, rp->c_timestamp + 120*HZ) && - memcmp((char*)&rqstp->rq_addr, (char*)&rp->c_addr, sizeof(rp->c_addr))==0) { + rpc_cmp_addr(svc_addr(rqstp), (struct sockaddr *)&rp->c_addr) && + rpc_get_port(svc_addr(rqstp)) == rpc_get_port((struct sockaddr *)&rp->c_addr)) { nfsdstats.rchits++; goto found_entry; } @@ -183,7 +185,8 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) rp->c_state = RC_INPROG; rp->c_xid = xid; rp->c_proc = proc; - memcpy(&rp->c_addr, svc_addr_in(rqstp), sizeof(rp->c_addr)); + rpc_copy_addr((struct sockaddr *)&rp->c_addr, svc_addr(rqstp)); + rpc_set_port((struct sockaddr *)&rp->c_addr, rpc_get_port(svc_addr(rqstp))); rp->c_prot = proto; rp->c_vers = vers; rp->c_timestamp = jiffies; -- cgit From 6dc8889589678c228a675b9f01721318a31a8f5b Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:08 -0500 Subject: nfsd: remove unneeded spinlock in nfsd_cache_update The locking rules for cache entries say that locking the cache_lock isn't needed if you're just touching the current entry. Earlier in this function we set rp->c_state to RC_UNUSED without any locking, so I believe it's ok to do the same here. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 5dd9ec2a177f..972c14a033ee 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -286,9 +286,7 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) cachv = &rp->c_replvec; cachv->iov_base = kmalloc(len << 2, GFP_KERNEL); if (!cachv->iov_base) { - spin_lock(&cache_lock); rp->c_state = RC_UNUSED; - spin_unlock(&cache_lock); return; } cachv->iov_len = len << 2; -- cgit From 09662d58d5a2d75c8c29558dda4fc5134ef14b25 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:09 -0500 Subject: nfsd: get rid of RC_INTR The reply cache code never returns this status. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/cache.h | 3 +-- fs/nfsd/nfssvc.c | 1 - 2 files changed, 1 insertion(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/cache.h b/fs/nfsd/cache.h index 2cac76c63b97..f8c6df8649dc 100644 --- a/fs/nfsd/cache.h +++ b/fs/nfsd/cache.h @@ -50,8 +50,7 @@ enum { enum { RC_DROPIT, RC_REPLY, - RC_DOIT, - RC_INTR + RC_DOIT }; /* diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index cee62ab9d4a3..40cb1cbaba4e 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -652,7 +652,6 @@ nfsd_dispatch(struct svc_rqst *rqstp, __be32 *statp) /* Check whether we have this call in the cache. */ switch (nfsd_cache_lookup(rqstp)) { - case RC_INTR: case RC_DROPIT: return 0; case RC_REPLY: -- cgit From 8a8bc40d9ba0890f88dbf7a7c8fa81ddc77c08e3 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:10 -0500 Subject: nfsd: create a dedicated slabcache for DRC entries Currently we use kmalloc() which wastes a little bit of memory on each allocation since it's a power of 2 allocator. Since we're allocating a 1024 of these now, and may need even more later, let's create a new slabcache for them. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 972c14a033ee..4aad9e4a6161 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -26,6 +26,7 @@ static struct hlist_head * cache_hash; static struct list_head lru_head; static int cache_disabled = 1; +static struct kmem_cache *drc_slab; /* * Calculate the hash index from an XID. @@ -51,10 +52,15 @@ int nfsd_reply_cache_init(void) struct svc_cacherep *rp; int i; + drc_slab = kmem_cache_create("nfsd_drc", sizeof(struct svc_cacherep), + 0, 0, NULL); + if (!drc_slab) + goto out_nomem; + INIT_LIST_HEAD(&lru_head); i = CACHESIZE; while (i) { - rp = kmalloc(sizeof(*rp), GFP_KERNEL); + rp = kmem_cache_alloc(drc_slab, GFP_KERNEL); if (!rp) goto out_nomem; list_add(&rp->c_lru, &lru_head); @@ -85,13 +91,18 @@ void nfsd_reply_cache_shutdown(void) if (rp->c_state == RC_DONE && rp->c_type == RC_REPLBUFF) kfree(rp->c_replvec.iov_base); list_del(&rp->c_lru); - kfree(rp); + kmem_cache_free(drc_slab, rp); } cache_disabled = 1; kfree (cache_hash); cache_hash = NULL; + + if (drc_slab) { + kmem_cache_destroy(drc_slab); + drc_slab = NULL; + } } /* -- cgit From f09841fdfad8dd95593725dfa70388f55b170947 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:11 -0500 Subject: nfsd: add alloc and free functions for DRC entries Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 36 +++++++++++++++++++++++++++--------- 1 file changed, 27 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 4aad9e4a6161..363bc6188155 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -47,10 +47,34 @@ static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *vec); */ static DEFINE_SPINLOCK(cache_lock); -int nfsd_reply_cache_init(void) +static struct svc_cacherep * +nfsd_reply_cache_alloc(void) { struct svc_cacherep *rp; + + rp = kmem_cache_alloc(drc_slab, GFP_KERNEL); + if (rp) { + rp->c_state = RC_UNUSED; + rp->c_type = RC_NOCACHE; + INIT_LIST_HEAD(&rp->c_lru); + INIT_HLIST_NODE(&rp->c_hash); + } + return rp; +} + +static void +nfsd_reply_cache_free_locked(struct svc_cacherep *rp) +{ + if (rp->c_state == RC_DONE && rp->c_type == RC_REPLBUFF) + kfree(rp->c_replvec.iov_base); + list_del(&rp->c_lru); + kmem_cache_free(drc_slab, rp); +} + +int nfsd_reply_cache_init(void) +{ int i; + struct svc_cacherep *rp; drc_slab = kmem_cache_create("nfsd_drc", sizeof(struct svc_cacherep), 0, 0, NULL); @@ -60,13 +84,10 @@ int nfsd_reply_cache_init(void) INIT_LIST_HEAD(&lru_head); i = CACHESIZE; while (i) { - rp = kmem_cache_alloc(drc_slab, GFP_KERNEL); + rp = nfsd_reply_cache_alloc(); if (!rp) goto out_nomem; list_add(&rp->c_lru, &lru_head); - rp->c_state = RC_UNUSED; - rp->c_type = RC_NOCACHE; - INIT_HLIST_NODE(&rp->c_hash); i--; } @@ -88,10 +109,7 @@ void nfsd_reply_cache_shutdown(void) while (!list_empty(&lru_head)) { rp = list_entry(lru_head.next, struct svc_cacherep, c_lru); - if (rp->c_state == RC_DONE && rp->c_type == RC_REPLBUFF) - kfree(rp->c_replvec.iov_base); - list_del(&rp->c_lru); - kmem_cache_free(drc_slab, rp); + nfsd_reply_cache_free_locked(rp); } cache_disabled = 1; -- cgit From 25e6b8b0e1a247747db5275b1b6b362f5acf2245 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:12 -0500 Subject: nfsd: remove redundant test from nfsd_reply_cache_free Entries can only get a c_type of RC_REPLBUFF iff they are RC_DONE. Therefore the test for RC_DONE isn't necessary here. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 363bc6188155..2cdc4be84553 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -65,7 +65,7 @@ nfsd_reply_cache_alloc(void) static void nfsd_reply_cache_free_locked(struct svc_cacherep *rp) { - if (rp->c_state == RC_DONE && rp->c_type == RC_REPLBUFF) + if (rp->c_type == RC_REPLBUFF) kfree(rp->c_replvec.iov_base); list_del(&rp->c_lru); kmem_cache_free(drc_slab, rp); -- cgit From d1a0774de6cb908f5ba7806d09aaf86bb03fa182 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:13 -0500 Subject: nfsd: clean up and clarify the cache expiration code Add a preprocessor constant for the expiry time of cache entries, and move the test for an expired entry into a function. Note that the current code does not test for RC_INPROG. It just assumes that it won't take more than 2 minutes to fill out an in-progress entry. I'm not sure how valid that assumption is though, so let's just ensure that we never consider an RC_INPROG entry to be expired. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/cache.h | 3 +++ fs/nfsd/nfscache.c | 9 ++++++++- 2 files changed, 11 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/cache.h b/fs/nfsd/cache.h index f8c6df8649dc..9c7232b45103 100644 --- a/fs/nfsd/cache.h +++ b/fs/nfsd/cache.h @@ -70,6 +70,9 @@ enum { */ #define RC_DELAY (HZ/5) +/* Cache entries expire after this time period */ +#define RC_EXPIRE (120 * HZ) + int nfsd_reply_cache_init(void); void nfsd_reply_cache_shutdown(void); int nfsd_cache_lookup(struct svc_rqst *); diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 2cdc4be84553..634b8566aaef 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -142,6 +142,13 @@ hash_refile(struct svc_cacherep *rp) hlist_add_head(&rp->c_hash, cache_hash + request_hash(rp->c_xid)); } +static inline bool +nfsd_cache_entry_expired(struct svc_cacherep *rp) +{ + return rp->c_state != RC_INPROG && + time_after(jiffies, rp->c_timestamp + RC_EXPIRE); +} + /* * Try to find an entry matching the current call in the cache. When none * is found, we grab the oldest unlocked entry off the LRU list. @@ -175,7 +182,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) if (rp->c_state != RC_UNUSED && xid == rp->c_xid && proc == rp->c_proc && proto == rp->c_prot && vers == rp->c_vers && - time_before(jiffies, rp->c_timestamp + 120*HZ) && + !nfsd_cache_entry_expired(rp) && rpc_cmp_addr(svc_addr(rqstp), (struct sockaddr *)&rp->c_addr) && rpc_get_port(svc_addr(rqstp)) == rpc_get_port((struct sockaddr *)&rp->c_addr)) { nfsdstats.rchits++; -- cgit From a4a3ec3291249c7db0e03d8b8188ef259b2873da Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 28 Jan 2013 14:41:14 -0500 Subject: nfsd: break out hashtable search into separate function Later, we'll need more than one call site for this, so break it out into a new function. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 46 +++++++++++++++++++++++++++++++++------------- 1 file changed, 33 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 634b8566aaef..b89e7c8b10ce 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -149,6 +149,35 @@ nfsd_cache_entry_expired(struct svc_cacherep *rp) time_after(jiffies, rp->c_timestamp + RC_EXPIRE); } +/* + * Search the request hash for an entry that matches the given rqstp. + * Must be called with cache_lock held. Returns the found entry or + * NULL on failure. + */ +static struct svc_cacherep * +nfsd_cache_search(struct svc_rqst *rqstp) +{ + struct svc_cacherep *rp; + struct hlist_node *hn; + struct hlist_head *rh; + __be32 xid = rqstp->rq_xid; + u32 proto = rqstp->rq_prot, + vers = rqstp->rq_vers, + proc = rqstp->rq_proc; + + rh = &cache_hash[request_hash(xid)]; + hlist_for_each_entry(rp, hn, rh, c_hash) { + if (rp->c_state != RC_UNUSED && + xid == rp->c_xid && proc == rp->c_proc && + proto == rp->c_prot && vers == rp->c_vers && + !nfsd_cache_entry_expired(rp) && + rpc_cmp_addr(svc_addr(rqstp), (struct sockaddr *)&rp->c_addr) && + rpc_get_port(svc_addr(rqstp)) == rpc_get_port((struct sockaddr *)&rp->c_addr)) + return rp; + } + return NULL; +} + /* * Try to find an entry matching the current call in the cache. When none * is found, we grab the oldest unlocked entry off the LRU list. @@ -157,8 +186,6 @@ nfsd_cache_entry_expired(struct svc_cacherep *rp) int nfsd_cache_lookup(struct svc_rqst *rqstp) { - struct hlist_node *hn; - struct hlist_head *rh; struct svc_cacherep *rp; __be32 xid = rqstp->rq_xid; u32 proto = rqstp->rq_prot, @@ -177,17 +204,10 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) spin_lock(&cache_lock); rtn = RC_DOIT; - rh = &cache_hash[request_hash(xid)]; - hlist_for_each_entry(rp, hn, rh, c_hash) { - if (rp->c_state != RC_UNUSED && - xid == rp->c_xid && proc == rp->c_proc && - proto == rp->c_prot && vers == rp->c_vers && - !nfsd_cache_entry_expired(rp) && - rpc_cmp_addr(svc_addr(rqstp), (struct sockaddr *)&rp->c_addr) && - rpc_get_port(svc_addr(rqstp)) == rpc_get_port((struct sockaddr *)&rp->c_addr)) { - nfsdstats.rchits++; - goto found_entry; - } + rp = nfsd_cache_search(rqstp); + if (rp) { + nfsdstats.rchits++; + goto found_entry; } nfsdstats.rcmisses++; -- cgit From 2eeb9b2abc83a7fd02063c7558b8e729f5a3ae00 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Sat, 2 Feb 2013 05:55:08 -0500 Subject: nfsd: initialize the exp->ex_uuid field in svc_export_init commit 885c91f7466 in Bruce's tree was causing oopses for me: general protection fault: 0000 [#1] SMP Modules linked in: nfsd(OF) nfs_acl(OF) auth_rpcgss(OF) lockd(OF) sunrpc(OF) kvm_amd kvm microcode i2c_piix4 virtio_net virtio_balloon cirrus drm_kms_helper ttm drm virtio_blk i2c_core CPU 0 Pid: 564, comm: exportfs Tainted: GF O 3.8.0-0.rc5.git2.1.fc19.x86_64 #1 Bochs Bochs RIP: 0010:[] [] kfree+0x49/0x280 RSP: 0018:ffff88007a3d7c50 EFLAGS: 00010203 RAX: 01adaf8dadadad80 RBX: 6b6b6b6b6b6b6b6b RCX: 0000000000000001 RDX: ffffffff7fffffff RSI: 0000000000000000 RDI: 6b6b6b6b6b6b6b6b RBP: ffff88007a3d7c80 R08: 6b6b6b6b6b6b6b6b R09: 0000000000000000 R10: 0000000000000018 R11: 0000000000000000 R12: ffff88006a117b50 R13: ffffffffa01a589c R14: ffff8800631b0f50 R15: 01ad998dadadad80 FS: 00007fcaa3616740(0000) GS:ffff88007fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00007f5d84b6fdd8 CR3: 0000000064db4000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process exportfs (pid: 564, threadinfo ffff88007a3d6000, task ffff88006af28000) Stack: ffff88007a3d7c80 ffff88006a117b68 ffff88006a117b50 0000000000000000 ffff8800631b0f50 ffff88006a117b50 ffff88007a3d7ca0 ffffffffa01a589c ffff880036be1148 ffff88007a3d7cf8 ffff88007a3d7e28 ffffffffa01a6a98 Call Trace: [] svc_export_put+0x5c/0x70 [nfsd] [] svc_export_parse+0x328/0x7e0 [nfsd] [] cache_do_downcall+0x57/0x70 [sunrpc] [] cache_downcall+0x7e/0x100 [sunrpc] [] cache_write_procfs+0x58/0x90 [sunrpc] [] ? cache_downcall+0x100/0x100 [sunrpc] [] proc_reg_write+0x75/0xb0 [] vfs_write+0x9f/0x170 [] sys_write+0x49/0xa0 [] system_call_fastpath+0x16/0x1b Code: 66 66 66 90 48 83 fb 10 0f 86 c3 00 00 00 48 89 df 49 bf 00 00 00 00 00 ea ff ff e8 f2 12 ea ff 48 c1 e8 0c 48 c1 e0 06 49 01 c7 <49> 8b 07 f6 c4 80 0f 85 1d 02 00 00 49 8b 07 a8 80 0f 84 ee 01 RIP [] kfree+0x49/0x280 RSP I think Majianpeng's patch is correct, but incomplete. In order for it to be safe to free the ex_uuid unconditionally in svc_export_put, we need to make sure it's initialized to NULL in the init routine. Cc: majianpeng Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/export.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index 45159ee952e6..8e9df45b3d3c 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -671,6 +671,7 @@ static void svc_export_init(struct cache_head *cnew, struct cache_head *citem) new->ex_fslocs.locations = NULL; new->ex_fslocs.locations_count = 0; new->ex_fslocs.migrated = 0; + new->ex_uuid = NULL; new->cd = item->cd; } -- cgit From 6a4e922c3db06f7da27e072729e047185c2fba66 Mon Sep 17 00:00:00 2001 From: Eric Wong Date: Mon, 4 Feb 2013 13:04:44 +0000 Subject: fuse: avoid out-of-scope stack access The all pointers within fuse_req must point to valid memory once fuse_force_forget() returns. This bug appeared in "fuse: implement NFS-like readdirplus support" and was never in any official Linux release. I tested the fuse_force_forget() code path by injecting to fake -ENOMEM and verified the FORGET operation was called properly in userspace. Signed-off-by: Eric Wong Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index cbae09e5a491..e9bdec0b16d9 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -440,9 +440,8 @@ __acquires(fc->lock) } } -void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) +static void __fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) { - req->isreply = 1; spin_lock(&fc->lock); if (!fc->connected) req->out.h.error = -ENOTCONN; @@ -459,6 +458,12 @@ void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) } spin_unlock(&fc->lock); } + +void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) +{ + req->isreply = 1; + __fuse_request_send(fc, req); +} EXPORT_SYMBOL_GPL(fuse_request_send); static void fuse_request_send_nowait_locked(struct fuse_conn *fc, @@ -541,7 +546,9 @@ void fuse_force_forget(struct file *file, u64 nodeid) req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; req->isreply = 0; - fuse_request_send_nowait(fc, req); + __fuse_request_send(fc, req); + /* ignore errors */ + fuse_put_request(fc, req); } /* -- cgit From dfca7cebc2679f3d129f8e680a8f199a7ad16e38 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Mon, 4 Feb 2013 15:57:42 +0100 Subject: fuse: don't WARN when nlink is zero drop_nlink() warns if nlink is already zero. This is triggerable by a buggy userspace filesystem. The cure, I think, is worse than the disease so disable the warning. Reported-by: Tero Roponen Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index dc5e64893375..2b112d978e9f 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -705,7 +705,14 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry) spin_lock(&fc->lock); fi->attr_version = ++fc->attr_version; - drop_nlink(inode); + /* + * If i_nlink == 0 then unlink doesn't make sense, yet this can + * happen if userspace filesystem is careless. It would be + * difficult to enforce correct nlink usage so just ignore this + * condition here + */ + if (inode->i_nlink > 0) + drop_nlink(inode); spin_unlock(&fc->lock); fuse_invalidate_attr(inode); fuse_invalidate_attr(dir); -- cgit From 0415d291022543d83ee799e9ffee08d856bca6e8 Mon Sep 17 00:00:00 2001 From: Enke Chen Date: Mon, 4 Feb 2013 16:14:32 +0100 Subject: fuse: send poll events commit 626cf23660 "poll: add poll_requested_events()..." enabled us to send the requested events to the filesystem. Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index a010585b0a74..c8071768b950 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -2167,6 +2167,7 @@ unsigned fuse_file_poll(struct file *file, poll_table *wait) return DEFAULT_POLLMASK; poll_wait(file, &ff->poll_wait, wait); + inarg.events = (__u32)poll_requested_events(wait); /* * Ask for notification iff there's someone waiting for it. -- cgit From 40ae3487628235e5f1eb27542cca0cdb6e5dbe16 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 4 Feb 2013 15:08:40 -0500 Subject: ext4: optimize mballoc for large allocations The ext4 block allocator only maintains buddy bitmaps for chunks which are less than or equal to one quarter of a block group. That is, for a file aystem with a 1k blocksize, and where the number of blocks in a block group is 8192 blocks, the largest chunk size tracked by buddy bitmaps is 2048 blocks. For a file system with a 4k blocksize, and where the number of blocks in a block group is 32768 blocks, the largest chunk size tracked by buddy bitmaps is 8192 blocks. To work around this code, mballoc.c before this commit would truncate allocation requests to the number of blocks in a block group minus 10. Why 10? Aside from being a completely arbitrary number, it avoids block allocation to be a power of two larger than 25% of the block group. If you try to explicitly fallocate 50% of the block group size, this will demonstrate the problem; the block allocation code will scan the all of the blocks in the file system with cr==0 (since the request is for a natural power of two), but then completely fail for all blocks groups, since the buddy bitmaps don't track chunk sizes of 50% of the block group. To fix this, in these we use ext4_mb_complex_scan_group() instead of ext4_mb_simple_scan_group(). Signed-off-by: "Theodore Ts'o" Cc: Andreas Dilger --- fs/ext4/mballoc.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 061727acd990..e350885aec30 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -1884,15 +1884,19 @@ static int ext4_mb_good_group(struct ext4_allocation_context *ac, case 0: BUG_ON(ac->ac_2order == 0); - if (grp->bb_largest_free_order < ac->ac_2order) - return 0; - /* Avoid using the first bg of a flexgroup for data files */ if ((ac->ac_flags & EXT4_MB_HINT_DATA) && (flex_size >= EXT4_FLEX_SIZE_DIR_ALLOC_SCHEME) && ((group % flex_size) == 0)) return 0; + if ((ac->ac_2order > ac->ac_sb->s_blocksize_bits+1) || + (free / fragments) >= ac->ac_g_ex.fe_len) + return 1; + + if (grp->bb_largest_free_order < ac->ac_2order) + return 0; + return 1; case 1: if ((free / fragments) >= ac->ac_g_ex.fe_len) @@ -2007,7 +2011,7 @@ repeat: } ac->ac_groups_scanned++; - if (cr == 0) + if (cr == 0 && ac->ac_2order < sb->s_blocksize_bits+2) ext4_mb_simple_scan_group(ac, &e4b); else if (cr == 1 && sbi->s_stripe && !(ac->ac_g_ex.fe_len % sbi->s_stripe)) @@ -4005,8 +4009,8 @@ ext4_mb_initialize_context(struct ext4_allocation_context *ac, len = ar->len; /* just a dirty hack to filter too big requests */ - if (len >= EXT4_CLUSTERS_PER_GROUP(sb) - 10) - len = EXT4_CLUSTERS_PER_GROUP(sb) - 10; + if (len >= EXT4_CLUSTERS_PER_GROUP(sb)) + len = EXT4_CLUSTERS_PER_GROUP(sb); /* start searching from the goal */ goal = ar->goal; -- cgit From d4b0bcf32b946590afd29e202d6a399b84fe6c67 Mon Sep 17 00:00:00 2001 From: David Teigland Date: Mon, 4 Feb 2013 15:31:22 -0600 Subject: dlm: check the write size from user Return EINVAL from write if the size is larger than allowed. Do this before allocating kernel memory for the bogus size, which could lead to OOM. Reported-by: Sasha Levin Tested-by: Jana Saout Signed-off-by: David Teigland --- fs/dlm/user.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/dlm/user.c b/fs/dlm/user.c index 7ff49852b0cb..911649a47dd5 100644 --- a/fs/dlm/user.c +++ b/fs/dlm/user.c @@ -503,11 +503,11 @@ static ssize_t device_write(struct file *file, const char __user *buf, #endif return -EINVAL; -#ifdef CONFIG_COMPAT - if (count > sizeof(struct dlm_write_request32) + DLM_RESNAME_MAXLEN) -#else + /* + * can't compare against COMPAT/dlm_write_request32 because + * we don't yet know if is64bit is zero + */ if (count > sizeof(struct dlm_write_request) + DLM_RESNAME_MAXLEN) -#endif return -EINVAL; kbuf = kzalloc(count + 1, GFP_NOFS); -- cgit From 56c2548b2d24de440868885e4c36c985eca2d083 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 08:18:00 -0500 Subject: nfsd: always move DRC entries to the end of LRU list when updating timestamp ...otherwise, we end up with the list ordering wrong. Currently, it's not a problem since we skip RC_INPROG entries, but keeping the ordering strict will be necessary for a later patch that adds a cache cleaner. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index b89e7c8b10ce..9d80dfa7778e 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -129,6 +129,7 @@ void nfsd_reply_cache_shutdown(void) static void lru_put_end(struct svc_cacherep *rp) { + rp->c_timestamp = jiffies; list_move_tail(&rp->c_lru, &lru_head); } @@ -245,9 +246,9 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) rpc_set_port((struct sockaddr *)&rp->c_addr, rpc_get_port(svc_addr(rqstp))); rp->c_prot = proto; rp->c_vers = vers; - rp->c_timestamp = jiffies; hash_refile(rp); + lru_put_end(rp); /* release any buffer */ if (rp->c_type == RC_REPLBUFF) { @@ -262,7 +263,6 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) found_entry: /* We found a matching entry which is either in progress or done. */ age = jiffies - rp->c_timestamp; - rp->c_timestamp = jiffies; lru_put_end(rp); rtn = RC_DROPIT; @@ -354,7 +354,6 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) rp->c_secure = rqstp->rq_secure; rp->c_type = cachetype; rp->c_state = RC_DONE; - rp->c_timestamp = jiffies; spin_unlock(&cache_lock); return; } -- cgit From 0ee0bf7ee5b55f232b645c4af0b0c37d4e115a32 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 08:18:01 -0500 Subject: nfsd: track the number of DRC entries in the cache Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 9d80dfa7778e..c0c58471eb45 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -27,6 +27,7 @@ static struct hlist_head * cache_hash; static struct list_head lru_head; static int cache_disabled = 1; static struct kmem_cache *drc_slab; +static unsigned int num_drc_entries; /* * Calculate the hash index from an XID. @@ -68,6 +69,7 @@ nfsd_reply_cache_free_locked(struct svc_cacherep *rp) if (rp->c_type == RC_REPLBUFF) kfree(rp->c_replvec.iov_base); list_del(&rp->c_lru); + --num_drc_entries; kmem_cache_free(drc_slab, rp); } @@ -83,10 +85,12 @@ int nfsd_reply_cache_init(void) INIT_LIST_HEAD(&lru_head); i = CACHESIZE; + num_drc_entries = 0; while (i) { rp = nfsd_reply_cache_alloc(); if (!rp) goto out_nomem; + ++num_drc_entries; list_add(&rp->c_lru, &lru_head); i--; } -- cgit From 0338dd157282c19696d3c32614a748d0ba814b12 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 08:18:02 -0500 Subject: nfsd: dynamically allocate DRC entries The existing code keeps a fixed-size cache of 1024 entries. This is much too small for a busy server, and wastes memory on an idle one. This patch changes the code to dynamically allocate and free these cache entries. A cap on the number of entries is retained, but it's much larger than the existing value and now scales with the amount of low memory in the machine. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 125 ++++++++++++++++++++++++++++++++--------------------- 1 file changed, 75 insertions(+), 50 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index c0c58471eb45..d213e6e69e46 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -10,17 +10,13 @@ #include #include +#include #include "nfsd.h" #include "cache.h" -/* Size of reply cache. Common values are: - * 4.3BSD: 128 - * 4.4BSD: 256 - * Solaris2: 1024 - * DEC Unix: 512-4096 - */ -#define CACHESIZE 1024 +#define NFSDDBG_FACILITY NFSDDBG_REPCACHE + #define HASHSIZE 64 static struct hlist_head * cache_hash; @@ -28,6 +24,7 @@ static struct list_head lru_head; static int cache_disabled = 1; static struct kmem_cache *drc_slab; static unsigned int num_drc_entries; +static unsigned int max_drc_entries; /* * Calculate the hash index from an XID. @@ -48,6 +45,34 @@ static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *vec); */ static DEFINE_SPINLOCK(cache_lock); +/* + * Put a cap on the size of the DRC based on the amount of available + * low memory in the machine. + * + * 64MB: 8192 + * 128MB: 11585 + * 256MB: 16384 + * 512MB: 23170 + * 1GB: 32768 + * 2GB: 46340 + * 4GB: 65536 + * 8GB: 92681 + * 16GB: 131072 + * + * ...with a hard cap of 256k entries. In the worst case, each entry will be + * ~1k, so the above numbers should give a rough max of the amount of memory + * used in k. + */ +static unsigned int +nfsd_cache_size_limit(void) +{ + unsigned int limit; + unsigned long low_pages = totalram_pages - totalhigh_pages; + + limit = (16 * int_sqrt(low_pages)) << (PAGE_SHIFT-10); + return min_t(unsigned int, limit, 256*1024); +} + static struct svc_cacherep * nfsd_reply_cache_alloc(void) { @@ -68,6 +93,7 @@ nfsd_reply_cache_free_locked(struct svc_cacherep *rp) { if (rp->c_type == RC_REPLBUFF) kfree(rp->c_replvec.iov_base); + hlist_del(&rp->c_hash); list_del(&rp->c_lru); --num_drc_entries; kmem_cache_free(drc_slab, rp); @@ -75,30 +101,18 @@ nfsd_reply_cache_free_locked(struct svc_cacherep *rp) int nfsd_reply_cache_init(void) { - int i; - struct svc_cacherep *rp; - drc_slab = kmem_cache_create("nfsd_drc", sizeof(struct svc_cacherep), 0, 0, NULL); if (!drc_slab) goto out_nomem; - INIT_LIST_HEAD(&lru_head); - i = CACHESIZE; - num_drc_entries = 0; - while (i) { - rp = nfsd_reply_cache_alloc(); - if (!rp) - goto out_nomem; - ++num_drc_entries; - list_add(&rp->c_lru, &lru_head); - i--; - } - - cache_hash = kcalloc (HASHSIZE, sizeof(struct hlist_head), GFP_KERNEL); + cache_hash = kcalloc(HASHSIZE, sizeof(struct hlist_head), GFP_KERNEL); if (!cache_hash) goto out_nomem; + INIT_LIST_HEAD(&lru_head); + max_drc_entries = nfsd_cache_size_limit(); + num_drc_entries = 0; cache_disabled = 0; return 0; out_nomem: @@ -191,7 +205,7 @@ nfsd_cache_search(struct svc_rqst *rqstp) int nfsd_cache_lookup(struct svc_rqst *rqstp) { - struct svc_cacherep *rp; + struct svc_cacherep *rp, *found; __be32 xid = rqstp->rq_xid; u32 proto = rqstp->rq_prot, vers = rqstp->rq_vers, @@ -210,38 +224,48 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) rtn = RC_DOIT; rp = nfsd_cache_search(rqstp); - if (rp) { - nfsdstats.rchits++; + if (rp) goto found_entry; + + /* Try to use the first entry on the LRU */ + if (!list_empty(&lru_head)) { + rp = list_first_entry(&lru_head, struct svc_cacherep, c_lru); + if (nfsd_cache_entry_expired(rp) || + num_drc_entries >= max_drc_entries) + goto setup_entry; } - nfsdstats.rcmisses++; - /* This loop shouldn't take more than a few iterations normally */ - { - int safe = 0; - list_for_each_entry(rp, &lru_head, c_lru) { - if (rp->c_state != RC_INPROG) - break; - if (safe++ > CACHESIZE) { - printk("nfsd: loop in repcache LRU list\n"); - cache_disabled = 1; - goto out; - } + spin_unlock(&cache_lock); + rp = nfsd_reply_cache_alloc(); + if (!rp) { + dprintk("nfsd: unable to allocate DRC entry!\n"); + return RC_DOIT; } + spin_lock(&cache_lock); + ++num_drc_entries; + + /* + * Must search again just in case someone inserted one + * after we dropped the lock above. + */ + found = nfsd_cache_search(rqstp); + if (found) { + nfsd_reply_cache_free_locked(rp); + rp = found; + goto found_entry; } - /* All entries on the LRU are in-progress. This should not happen */ - if (&rp->c_lru == &lru_head) { - static int complaints; - - printk(KERN_WARNING "nfsd: all repcache entries locked!\n"); - if (++complaints > 5) { - printk(KERN_WARNING "nfsd: disabling repcache.\n"); - cache_disabled = 1; - } - goto out; - } + /* + * We're keeping the one we just allocated. Are we now over the + * limit? Prune one off the tip of the LRU in trade for the one we + * just allocated if so. + */ + if (num_drc_entries >= max_drc_entries) + nfsd_reply_cache_free_locked(list_first_entry(&lru_head, + struct svc_cacherep, c_lru)); +setup_entry: + nfsdstats.rcmisses++; rqstp->rq_cacherep = rp; rp->c_state = RC_INPROG; rp->c_xid = xid; @@ -265,6 +289,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) return rtn; found_entry: + nfsdstats.rchits++; /* We found a matching entry which is either in progress or done. */ age = jiffies - rp->c_timestamp; lru_put_end(rp); @@ -295,7 +320,7 @@ found_entry: break; default: printk(KERN_WARNING "nfsd: bad repcache type %d\n", rp->c_type); - rp->c_state = RC_UNUSED; + nfsd_reply_cache_free_locked(rp); } goto out; -- cgit From 13cc8a78e89db0469e67ac9b3ae466b661af93fa Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 08:18:03 -0500 Subject: nfsd: remove the cache_disabled flag With the change to dynamically allocate entries, the cache is never disabled on the fly. Remove this flag. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index d213e6e69e46..69d29d4ea579 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -21,7 +21,6 @@ static struct hlist_head * cache_hash; static struct list_head lru_head; -static int cache_disabled = 1; static struct kmem_cache *drc_slab; static unsigned int num_drc_entries; static unsigned int max_drc_entries; @@ -113,7 +112,6 @@ int nfsd_reply_cache_init(void) INIT_LIST_HEAD(&lru_head); max_drc_entries = nfsd_cache_size_limit(); num_drc_entries = 0; - cache_disabled = 0; return 0; out_nomem: printk(KERN_ERR "nfsd: failed to allocate reply cache\n"); @@ -130,8 +128,6 @@ void nfsd_reply_cache_shutdown(void) nfsd_reply_cache_free_locked(rp); } - cache_disabled = 1; - kfree (cache_hash); cache_hash = NULL; @@ -215,7 +211,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) int rtn; rqstp->rq_cacherep = NULL; - if (cache_disabled || type == RC_NOCACHE) { + if (type == RC_NOCACHE) { nfsdstats.rcnocache++; return RC_DOIT; } @@ -345,11 +341,11 @@ found_entry: void nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) { - struct svc_cacherep *rp; + struct svc_cacherep *rp = rqstp->rq_cacherep; struct kvec *resv = &rqstp->rq_res.head[0], *cachv; int len; - if (!(rp = rqstp->rq_cacherep) || cache_disabled) + if (!rp) return; len = resv->iov_len - ((char*)statp - (char*)resv->iov_base); -- cgit From 2c6b691c05bf77c4bc7c9f1a9b6d93a160928421 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 08:18:04 -0500 Subject: nfsd: when updating an entry with RC_NOCACHE, just free it There's no need to keep entries around that we're declaring RC_NOCACHE. Ditto if there's a problem with the entry. With this change too, there's no need to test for RC_UNUSED in the search function. If the entry's in the hash table then it's either INPROG or DONE. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 69d29d4ea579..e8ea785e295d 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -98,6 +98,14 @@ nfsd_reply_cache_free_locked(struct svc_cacherep *rp) kmem_cache_free(drc_slab, rp); } +static void +nfsd_reply_cache_free(struct svc_cacherep *rp) +{ + spin_lock(&cache_lock); + nfsd_reply_cache_free_locked(rp); + spin_unlock(&cache_lock); +} + int nfsd_reply_cache_init(void) { drc_slab = kmem_cache_create("nfsd_drc", sizeof(struct svc_cacherep), @@ -182,8 +190,7 @@ nfsd_cache_search(struct svc_rqst *rqstp) rh = &cache_hash[request_hash(xid)]; hlist_for_each_entry(rp, hn, rh, c_hash) { - if (rp->c_state != RC_UNUSED && - xid == rp->c_xid && proc == rp->c_proc && + if (xid == rp->c_xid && proc == rp->c_proc && proto == rp->c_prot && vers == rp->c_vers && !nfsd_cache_entry_expired(rp) && rpc_cmp_addr(svc_addr(rqstp), (struct sockaddr *)&rp->c_addr) && @@ -353,7 +360,7 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) /* Don't cache excessive amounts of data and XDR failures */ if (!statp || len > (256 >> 2)) { - rp->c_state = RC_UNUSED; + nfsd_reply_cache_free(rp); return; } @@ -367,12 +374,15 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) cachv = &rp->c_replvec; cachv->iov_base = kmalloc(len << 2, GFP_KERNEL); if (!cachv->iov_base) { - rp->c_state = RC_UNUSED; + nfsd_reply_cache_free(rp); return; } cachv->iov_len = len << 2; memcpy(cachv->iov_base, statp, len << 2); break; + case RC_NOCACHE: + nfsd_reply_cache_free(rp); + return; } spin_lock(&cache_lock); lru_put_end(rp); -- cgit From aca8a23de60c705e2458b2c6731ad59aa0717f83 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 08:18:05 -0500 Subject: nfsd: add recurring workqueue job to clean the cache It's not sufficient to only clean the cache when requests come in. What if we have a flurry of activity and then the server goes idle? Add a workqueue job that will clean the cache every RC_EXPIRE period. Care is taken to only run this when we expect to have entries expiring. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 50 +++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 47 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index e8ea785e295d..d7b088bee684 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -36,6 +36,7 @@ static inline u32 request_hash(u32 xid) } static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *vec); +static void cache_cleaner_func(struct work_struct *unused); /* * locking for the reply cache: @@ -43,6 +44,7 @@ static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *vec); * Otherwise, it when accessing _prev or _next, the lock must be held. */ static DEFINE_SPINLOCK(cache_lock); +static DECLARE_DELAYED_WORK(cache_cleaner, cache_cleaner_func); /* * Put a cap on the size of the DRC based on the amount of available @@ -131,6 +133,8 @@ void nfsd_reply_cache_shutdown(void) { struct svc_cacherep *rp; + cancel_delayed_work_sync(&cache_cleaner); + while (!list_empty(&lru_head)) { rp = list_entry(lru_head.next, struct svc_cacherep, c_lru); nfsd_reply_cache_free_locked(rp); @@ -146,13 +150,15 @@ void nfsd_reply_cache_shutdown(void) } /* - * Move cache entry to end of LRU list + * Move cache entry to end of LRU list, and queue the cleaner to run if it's + * not already scheduled. */ static void lru_put_end(struct svc_cacherep *rp) { rp->c_timestamp = jiffies; list_move_tail(&rp->c_lru, &lru_head); + schedule_delayed_work(&cache_cleaner, RC_EXPIRE); } /* @@ -172,6 +178,42 @@ nfsd_cache_entry_expired(struct svc_cacherep *rp) time_after(jiffies, rp->c_timestamp + RC_EXPIRE); } +/* + * Walk the LRU list and prune off entries that are older than RC_EXPIRE. + * Also prune the oldest ones when the total exceeds the max number of entries. + */ +static void +prune_cache_entries(void) +{ + struct svc_cacherep *rp, *tmp; + + list_for_each_entry_safe(rp, tmp, &lru_head, c_lru) { + if (!nfsd_cache_entry_expired(rp) && + num_drc_entries <= max_drc_entries) + break; + nfsd_reply_cache_free_locked(rp); + } + + /* + * Conditionally rearm the job. If we cleaned out the list, then + * cancel any pending run (since there won't be any work to do). + * Otherwise, we rearm the job or modify the existing one to run in + * RC_EXPIRE since we just ran the pruner. + */ + if (list_empty(&lru_head)) + cancel_delayed_work(&cache_cleaner); + else + mod_delayed_work(system_wq, &cache_cleaner, RC_EXPIRE); +} + +static void +cache_cleaner_func(struct work_struct *unused) +{ + spin_lock(&cache_lock); + prune_cache_entries(); + spin_unlock(&cache_lock); +} + /* * Search the request hash for an entry that matches the given rqstp. * Must be called with cache_lock held. Returns the found entry or @@ -192,7 +234,6 @@ nfsd_cache_search(struct svc_rqst *rqstp) hlist_for_each_entry(rp, hn, rh, c_hash) { if (xid == rp->c_xid && proc == rp->c_proc && proto == rp->c_prot && vers == rp->c_vers && - !nfsd_cache_entry_expired(rp) && rpc_cmp_addr(svc_addr(rqstp), (struct sockaddr *)&rp->c_addr) && rpc_get_port(svc_addr(rqstp)) == rpc_get_port((struct sockaddr *)&rp->c_addr)) return rp; @@ -234,8 +275,11 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) if (!list_empty(&lru_head)) { rp = list_first_entry(&lru_head, struct svc_cacherep, c_lru); if (nfsd_cache_entry_expired(rp) || - num_drc_entries >= max_drc_entries) + num_drc_entries >= max_drc_entries) { + lru_put_end(rp); + prune_cache_entries(); goto setup_entry; + } } spin_unlock(&cache_lock); -- cgit From b4e7f2c9450968303d24b48fec9da8abebcc0036 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 08:18:06 -0500 Subject: nfsd: register a shrinker for DRC cache entries Since we dynamically allocate them now, allow the system to call us up to release them if it gets low on memory. Since these entries aren't replaceable, only free ones that are expired or that are over the cap. The the seeks value is set to '1' however to indicate that freeing the these entries is low-cost. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index d7b088bee684..d16a5d6435f2 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -37,6 +37,13 @@ static inline u32 request_hash(u32 xid) static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *vec); static void cache_cleaner_func(struct work_struct *unused); +static int nfsd_reply_cache_shrink(struct shrinker *shrink, + struct shrink_control *sc); + +struct shrinker nfsd_reply_cache_shrinker = { + .shrink = nfsd_reply_cache_shrink, + .seeks = 1, +}; /* * locking for the reply cache: @@ -110,6 +117,7 @@ nfsd_reply_cache_free(struct svc_cacherep *rp) int nfsd_reply_cache_init(void) { + register_shrinker(&nfsd_reply_cache_shrinker); drc_slab = kmem_cache_create("nfsd_drc", sizeof(struct svc_cacherep), 0, 0, NULL); if (!drc_slab) @@ -133,6 +141,7 @@ void nfsd_reply_cache_shutdown(void) { struct svc_cacherep *rp; + unregister_shrinker(&nfsd_reply_cache_shrinker); cancel_delayed_work_sync(&cache_cleaner); while (!list_empty(&lru_head)) { @@ -214,6 +223,20 @@ cache_cleaner_func(struct work_struct *unused) spin_unlock(&cache_lock); } +static int +nfsd_reply_cache_shrink(struct shrinker *shrink, struct shrink_control *sc) +{ + unsigned int num; + + spin_lock(&cache_lock); + if (sc->nr_to_scan) + prune_cache_entries(); + num = num_drc_entries; + spin_unlock(&cache_lock); + + return num; +} + /* * Search the request hash for an entry that matches the given rqstp. * Must be called with cache_lock held. Returns the found entry or -- cgit From a9bae189542e71f91e61a4428adf6e5a7dfe8063 Mon Sep 17 00:00:00 2001 From: Vyacheslav Dubeyko Date: Mon, 4 Feb 2013 14:28:41 -0800 Subject: nilfs2: fix fix very long mount time issue There exists a situation when GC can work in background alone without any other filesystem activity during significant time. The nilfs_clean_segments() method calls nilfs_segctor_construct() that updates superblocks in the case of NILFS_SC_SUPER_ROOT and THE_NILFS_DISCONTINUED flags are set. But when GC is working alone the nilfs_clean_segments() is called with unset THE_NILFS_DISCONTINUED flag. As a result, the update of superblocks doesn't occurred all this time and in the case of SPOR superblocks keep very old values of last super root placement. SYMPTOMS: Trying to mount a NILFS2 volume after SPOR in such environment ends with very long mounting time (it can achieve about several hours in some cases). REPRODUCING PATH: 1. It needs to use external USB HDD, disable automount and doesn't make any additional filesystem activity on the NILFS2 volume. 2. Generate temporary file with size about 100 - 500 GB (for example, dd if=/dev/zero of= bs=1073741824 count=200). The size of file defines duration of GC working. 3. Then it needs to delete file. 4. Start GC manually by means of command "nilfs-clean -p 0". When you start GC by means of such way then, at the end, superblocks is updated by once. So, for simulation of SPOR, it needs to wait sometime (15 - 40 minutes) and simply switch off USB HDD manually. 5. Switch on USB HDD again and try to mount NILFS2 volume. As a result, NILFS2 volume will mount during very long time. REPRODUCIBILITY: 100% FIX: This patch adds checking that superblocks need to update and set THE_NILFS_DISCONTINUED flag before nilfs_clean_segments() call. Reported-by: Sergey Alexandrov Signed-off-by: Vyacheslav Dubeyko Tested-by: Vyacheslav Dubeyko Acked-by: Ryusuke Konishi Tested-by: Ryusuke Konishi Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/nilfs2/ioctl.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index fdb180769485..f3859354e41a 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -664,8 +664,11 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, if (ret < 0) printk(KERN_ERR "NILFS: GC failed during preparation: " "cannot read source blocks: err=%d\n", ret); - else + else { + if (nilfs_sb_need_update(nilfs)) + set_nilfs_discontinued(nilfs); ret = nilfs_clean_segments(inode->i_sb, argv, kbufs); + } nilfs_remove_all_gcinodes(nilfs); clear_nilfs_gc_running(nilfs); -- cgit From 2d32b29a1c2830f7c42caa8258c714acd983961f Mon Sep 17 00:00:00 2001 From: majianpeng Date: Tue, 29 Jan 2013 13:16:06 +0800 Subject: nfsd: Fix memleak When free nfs-client, it must free the ->cl_stateids. Cc: stable@kernel.org Signed-off-by: Jianpeng Ma Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 4db46aa387b7..a6637de7ae13 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1060,6 +1060,8 @@ free_client(struct nfs4_client *clp) } free_svc_cred(&clp->cl_cred); kfree(clp->cl_name.data); + idr_remove_all(&clp->cl_stateids); + idr_destroy(&clp->cl_stateids); kfree(clp); } -- cgit From 3abdb6071250b896b9d5b6e0c310d6e95666b4d7 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 3 Feb 2013 12:23:01 -0500 Subject: nfsd4: simplify idr allocation We don't really need to preallocate at all; just allocate and initialize everything at once, but leave the sc_type field initially 0 to prevent finding the stateid till it's fully initialized. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 81 ++++++++++++++++++++++++++++++++++------------------- 1 file changed, 52 insertions(+), 29 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index a6637de7ae13..27c77a0b423f 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -261,33 +261,46 @@ static inline int get_new_stid(struct nfs4_stid *stid) return new_stid; } -static void init_stid(struct nfs4_stid *stid, struct nfs4_client *cl, unsigned char type) +static struct nfs4_stid *nfs4_alloc_stid(struct nfs4_client *cl, struct +kmem_cache *slab) { - stateid_t *s = &stid->sc_stateid; + struct idr *stateids = &cl->cl_stateids; + static int min_stateid = 0; + struct nfs4_stid *stid; int new_id; - stid->sc_type = type; + stid = kmem_cache_alloc(slab, GFP_KERNEL); + if (!stid) + return NULL; + + if (!idr_pre_get(stateids, GFP_KERNEL)) + goto out_free; + if (idr_get_new_above(stateids, stid, min_stateid, &new_id)) + goto out_free; stid->sc_client = cl; - s->si_opaque.so_clid = cl->cl_clientid; - new_id = get_new_stid(stid); - s->si_opaque.so_id = (u32)new_id; + stid->sc_type = 0; + stid->sc_stateid.si_opaque.so_id = new_id; + stid->sc_stateid.si_opaque.so_clid = cl->cl_clientid; /* Will be incremented before return to client: */ - s->si_generation = 0; -} + stid->sc_stateid.si_generation = 0; -static struct nfs4_stid *nfs4_alloc_stid(struct nfs4_client *cl, struct kmem_cache *slab) -{ - struct idr *stateids = &cl->cl_stateids; - - if (!idr_pre_get(stateids, GFP_KERNEL)) - return NULL; /* - * Note: if we fail here (or any time between now and the time - * we actually get the new idr), we won't need to undo the idr - * preallocation, since the idr code caps the number of - * preallocated entries. + * It shouldn't be a problem to reuse an opaque stateid value. + * I don't think it is for 4.1. But with 4.0 I worry that, for + * example, a stray write retransmission could be accepted by + * the server when it should have been rejected. Therefore, + * adopt a trick from the sctp code to attempt to maximize the + * amount of time until an id is reused, by ensuring they always + * "increase" (mod INT_MAX): */ - return kmem_cache_alloc(slab, GFP_KERNEL); + + min_stateid = new_id+1; + if (min_stateid == INT_MAX) + min_stateid = 0; + return stid; +out_free: + kfree(stid); + return NULL; } static struct nfs4_ol_stateid * nfs4_alloc_stateid(struct nfs4_client *clp) @@ -316,7 +329,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_ol_stateid *stp, struct sv dp = delegstateid(nfs4_alloc_stid(clp, deleg_slab)); if (dp == NULL) return dp; - init_stid(&dp->dl_stid, clp, NFS4_DELEG_STID); + dp->dl_stid.sc_type = NFS4_DELEG_STID; /* * delegation seqid's are never incremented. The 4.1 special * meaning of seqid 0 isn't meaningful, really, but let's avoid @@ -337,13 +350,21 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_ol_stateid *stp, struct sv return dp; } +void free_stid(struct nfs4_stid *s, struct kmem_cache *slab) +{ + struct idr *stateids = &s->sc_client->cl_stateids; + + idr_remove(stateids, s->sc_stateid.si_opaque.so_id); + kmem_cache_free(slab, s); +} + void nfs4_put_delegation(struct nfs4_delegation *dp) { if (atomic_dec_and_test(&dp->dl_count)) { dprintk("NFSD: freeing dp %p\n",dp); put_nfs4_file(dp->dl_file); - kmem_cache_free(deleg_slab, dp); + free_stid(&dp->dl_stid, deleg_slab); num_delegations--; } } @@ -360,9 +381,7 @@ static void nfs4_put_deleg_lease(struct nfs4_file *fp) static void unhash_stid(struct nfs4_stid *s) { - struct idr *stateids = &s->sc_client->cl_stateids; - - idr_remove(stateids, s->sc_stateid.si_opaque.so_id); + s->sc_type = 0; } /* Called under the state lock. */ @@ -519,7 +538,7 @@ static void close_generic_stateid(struct nfs4_ol_stateid *stp) static void free_generic_stateid(struct nfs4_ol_stateid *stp) { - kmem_cache_free(stateid_slab, stp); + free_stid(&stp->st_stid, stateid_slab); } static void release_lock_stateid(struct nfs4_ol_stateid *stp) @@ -1260,7 +1279,12 @@ static void gen_confirm(struct nfs4_client *clp) static struct nfs4_stid *find_stateid(struct nfs4_client *cl, stateid_t *t) { - return idr_find(&cl->cl_stateids, t->si_opaque.so_id); + struct nfs4_stid *ret; + + ret = idr_find(&cl->cl_stateids, t->si_opaque.so_id); + if (!ret || !ret->sc_type) + return NULL; + return ret; } static struct nfs4_stid *find_stateid_by_type(struct nfs4_client *cl, stateid_t *t, char typemask) @@ -2446,9 +2470,8 @@ alloc_init_open_stateowner(unsigned int strhashval, struct nfs4_client *clp, str static void init_open_stateid(struct nfs4_ol_stateid *stp, struct nfs4_file *fp, struct nfsd4_open *open) { struct nfs4_openowner *oo = open->op_openowner; - struct nfs4_client *clp = oo->oo_owner.so_client; - init_stid(&stp->st_stid, clp, NFS4_OPEN_STID); + stp->st_stid.sc_type = NFS4_OPEN_STID; INIT_LIST_HEAD(&stp->st_lockowners); list_add(&stp->st_perstateowner, &oo->oo_owner.so_stateids); list_add(&stp->st_perfile, &fp->fi_stateids); @@ -4034,7 +4057,7 @@ alloc_init_lock_stateid(struct nfs4_lockowner *lo, struct nfs4_file *fp, struct stp = nfs4_alloc_stateid(clp); if (stp == NULL) return NULL; - init_stid(&stp->st_stid, clp, NFS4_LOCK_STID); + stp->st_stid.sc_type = NFS4_LOCK_STID; list_add(&stp->st_perfile, &fp->fi_stateids); list_add(&stp->st_perstateowner, &lo->lo_owner.so_stateids); stp->st_stateowner = &lo->lo_owner; -- cgit From 5976687a2b3d1969f02aba16b80ad3ed79be6ad3 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 12:50:00 -0500 Subject: sunrpc: move address copy/cmp/convert routines and prototypes from clnt.h to addr.h These routines are used by server and client code, so having them in a separate header would be best. Signed-off-by: Jeff Layton Acked-by: Trond Myklebust Signed-off-by: J. Bruce Fields --- fs/lockd/clntlock.c | 2 +- fs/lockd/host.c | 1 + fs/lockd/mon.c | 1 + fs/lockd/svcsubs.c | 2 +- fs/nfs/dns_resolve.c | 2 ++ fs/nfs/nfs4client.c | 1 + fs/nfs/nfs4filelayoutdev.c | 1 + fs/nfs/nfs4namespace.c | 1 + fs/nfs/super.c | 1 + fs/nfsd/fault_inject.c | 2 +- fs/nfsd/nfs4state.c | 2 +- fs/nfsd/nfscache.c | 2 +- fs/nfsd/nfsctl.c | 2 +- 13 files changed, 14 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index ca0a08001449..4885b53e56bd 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -11,7 +11,7 @@ #include #include #include -#include +#include #include #include #include diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 0e17090c310f..764c4d2ed804 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 3c2cfc683631..1812f026960c 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -12,6 +12,7 @@ #include #include +#include #include #include #include diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index 0deb5f6c9dd4..8064435e8bef 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -13,7 +13,7 @@ #include #include #include -#include +#include #include #include #include diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index ca4b11ec87a2..62c8c4127219 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -10,6 +10,7 @@ #include #include +#include #include #include "dns_resolve.h" @@ -42,6 +43,7 @@ EXPORT_SYMBOL_GPL(nfs_dns_resolve_name); #include #include #include +#include #include #include #include diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index acc347268124..e5f2fad14ff8 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -6,6 +6,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/nfs/nfs4filelayoutdev.c b/fs/nfs/nfs4filelayoutdev.c index b720064bcd7f..1fe284f01f8b 100644 --- a/fs/nfs/nfs4filelayoutdev.c +++ b/fs/nfs/nfs4filelayoutdev.c @@ -31,6 +31,7 @@ #include #include #include +#include #include "internal.h" #include "nfs4session.h" diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c index 1e09eb78543b..0dd766079e1c 100644 --- a/fs/nfs/nfs4namespace.c +++ b/fs/nfs/nfs4namespace.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include #include "internal.h" diff --git a/fs/nfs/super.c b/fs/nfs/super.c index c25cadf8f8c4..3250b41eb562 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/nfsd/fault_inject.c b/fs/nfsd/fault_inject.c index e761ee95617f..247c00ccdb0f 100644 --- a/fs/nfsd/fault_inject.c +++ b/fs/nfsd/fault_inject.c @@ -9,7 +9,7 @@ #include #include #include -#include +#include #include #include "state.h" diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 27c77a0b423f..60a2282905e0 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -40,7 +40,7 @@ #include #include #include -#include +#include #include "xdr4.h" #include "vfs.h" #include "current_stateid.h" diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index d16a5d6435f2..f7544698e6e6 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -9,7 +9,7 @@ */ #include -#include +#include #include #include "nfsd.h" diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 65889ec1d28e..29c3f0d25469 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -10,7 +10,7 @@ #include #include -#include +#include #include #include #include -- cgit From 10e78e3a8aec820f8f6344dfd4d465f5f84399e6 Mon Sep 17 00:00:00 2001 From: Tomasz Torcz Date: Tue, 5 Feb 2013 09:38:53 -0500 Subject: Btrfs: select XOR_BLOCKS in Kconfig The Btrfs raid56 uses the generic xor helpers. Signed-off-by: Chris Mason --- fs/btrfs/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/Kconfig b/fs/btrfs/Kconfig index 4f5dc93fa2f8..5f583c8a36d0 100644 --- a/fs/btrfs/Kconfig +++ b/fs/btrfs/Kconfig @@ -7,6 +7,7 @@ config BTRFS_FS select LZO_COMPRESS select LZO_DECOMPRESS select RAID6_PQ + select XOR_BLOCKS help Btrfs is a new filesystem with extents, writable snapshotting, -- cgit From 1f0905ec156eec8f12cd593bc564551770319720 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 5 Feb 2013 10:01:42 -0500 Subject: Btrfs: remove conflicting check for minimum number of devices in raid56 The device removal code was incorrectly checking against two different limits for raid5 and raid6. Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index c372264b85bf..77620f2d8af9 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1392,14 +1392,6 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) } btrfs_dev_replace_unlock(&root->fs_info->dev_replace); - if ((all_avail & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6) && num_devices <= 3)) { - printk(KERN_ERR "btrfs: unable to go below three devices " - "on raid5 or raid6\n"); - ret = -EINVAL; - goto out; - } - if ((all_avail & BTRFS_BLOCK_GROUP_RAID10) && num_devices <= 4) { printk(KERN_ERR "btrfs: unable to go below four devices " "on raid10\n"); -- cgit From 89b1f39eb4189de745fae554b0d614d87c8d5c63 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 5 Feb 2013 13:59:56 +0100 Subject: udf: Fix bitmap overflow on large filesystems with small block size For large UDF filesystems with 512-byte blocks the number of necessary bitmap blocks is larger than 2^16 so s_nr_groups in udf_bitmap overflows (the number will overflow for filesystems larger than 128 GB with 512-byte blocks). That results in ENOSPC errors despite the filesystem has plenty of free space. Fix the problem by changing s_nr_groups' type to 'int'. That is enough even for filesystems 2^32 blocks (UDF maximum) and 512-byte blocksize. Reported-and-tested-by: v10lator@myway.de Signed-off-by: Jan Kara --- fs/udf/udf_sb.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/udf/udf_sb.h b/fs/udf/udf_sb.h index 5f027227f085..8d1c9d4f439e 100644 --- a/fs/udf/udf_sb.h +++ b/fs/udf/udf_sb.h @@ -82,7 +82,7 @@ struct udf_virtual_data { struct udf_bitmap { __u32 s_extLength; __u32 s_extPosition; - __u16 s_nr_groups; + int s_nr_groups; struct buffer_head **s_block_bitmap; }; -- cgit From c60305b578674eefe333198c7476dba2178a9082 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 5 Feb 2013 14:08:57 +0100 Subject: udf: Make s_block_bitmap standard array struct udf_bitmap has array of buffer pointers attached to it. The code unnecessarily used s_block_bitmap as a pointer to the array instead of the standard trick of using 0 length array in the declaration. Change that to make code more readable and actually shrink the structure by one pointer. Signed-off-by: Jan Kara --- fs/udf/super.c | 1 - fs/udf/udf_sb.h | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/udf/super.c b/fs/udf/super.c index da8ce9f14387..f8830803d389 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -1023,7 +1023,6 @@ static struct udf_bitmap *udf_sb_alloc_bitmap(struct super_block *sb, u32 index) if (bitmap == NULL) return NULL; - bitmap->s_block_bitmap = (struct buffer_head **)(bitmap + 1); bitmap->s_nr_groups = nr_groups; return bitmap; } diff --git a/fs/udf/udf_sb.h b/fs/udf/udf_sb.h index 8d1c9d4f439e..4f7ddb796991 100644 --- a/fs/udf/udf_sb.h +++ b/fs/udf/udf_sb.h @@ -83,7 +83,7 @@ struct udf_bitmap { __u32 s_extLength; __u32 s_extPosition; int s_nr_groups; - struct buffer_head **s_block_bitmap; + struct buffer_head *s_block_bitmap[0]; }; struct udf_part_map { -- cgit From 288be96de66fa7c09f2cf00a7793db5b4bac4213 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 5 Feb 2013 13:58:29 +0100 Subject: udf: Remove unused s_extLength from udf_bitmap s_extLength was assigned to but the value was never really used. So just remove the field. Signed-off-by: Jan Kara --- fs/udf/super.c | 4 ---- fs/udf/udf_sb.h | 1 - 2 files changed, 5 deletions(-) (limited to 'fs') diff --git a/fs/udf/super.c b/fs/udf/super.c index f8830803d389..bc5b30a819e8 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -1080,8 +1080,6 @@ static int udf_fill_partdesc_info(struct super_block *sb, if (!bitmap) return 1; map->s_uspace.s_bitmap = bitmap; - bitmap->s_extLength = le32_to_cpu( - phd->unallocSpaceBitmap.extLength); bitmap->s_extPosition = le32_to_cpu( phd->unallocSpaceBitmap.extPosition); map->s_partition_flags |= UDF_PART_FLAG_UNALLOC_BITMAP; @@ -1116,8 +1114,6 @@ static int udf_fill_partdesc_info(struct super_block *sb, if (!bitmap) return 1; map->s_fspace.s_bitmap = bitmap; - bitmap->s_extLength = le32_to_cpu( - phd->freedSpaceBitmap.extLength); bitmap->s_extPosition = le32_to_cpu( phd->freedSpaceBitmap.extPosition); map->s_partition_flags |= UDF_PART_FLAG_FREED_BITMAP; diff --git a/fs/udf/udf_sb.h b/fs/udf/udf_sb.h index 4f7ddb796991..ed401e94aa8c 100644 --- a/fs/udf/udf_sb.h +++ b/fs/udf/udf_sb.h @@ -80,7 +80,6 @@ struct udf_virtual_data { }; struct udf_bitmap { - __u32 s_extLength; __u32 s_extPosition; int s_nr_groups; struct buffer_head *s_block_bitmap[0]; -- cgit From 222c81dc3874b4fe98371be665d0447a36447653 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 28 Jan 2013 09:45:20 -0500 Subject: Btrfs: do not merge logged extents if we've removed them from the tree You can run into this problem where if somebody is fsyncing and writing out the existing extents you will have removed the extent map from the em tree, but it's still valid for the current fsync so we go ahead and write it. The problem is we unconditionally try to merge it back into the em tree, but if we've removed it from the em tree that will cause use after free problems. Fix this to only merge if we are still a part of the tree. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/extent_map.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index ed88f5ee4bea..9759911dd34e 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -289,7 +289,8 @@ out: void clear_em_logging(struct extent_map_tree *tree, struct extent_map *em) { clear_bit(EXTENT_FLAG_LOGGING, &em->flags); - try_merge_map(tree, em); + if (em->in_tree) + try_merge_map(tree, em); } /** -- cgit From 0a3404dcff29a7589e8d6ce8c36f97c5411f85b4 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Mon, 28 Jan 2013 12:34:55 +0000 Subject: Btrfs: fix wrong sync_writers decrement in btrfs_file_aio_write() If the checks at the beginning of btrfs_file_aio_write() fail, we needn't decrease ->sync_writers, because we have not increased it. Fix it. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/file.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 841cfe3be0e0..a902faab7161 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1595,9 +1595,10 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, if (err < 0 && num_written > 0) num_written = err; } -out: + if (sync) atomic_dec(&BTRFS_I(inode)->sync_writers); +out: sb_end_write(inode->i_sb); current->backing_dev_info = NULL; return num_written ? num_written : err; -- cgit From 843fcf35733164076a77ad833c72c32da8228ad0 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Mon, 28 Jan 2013 12:36:22 +0000 Subject: Btrfs: fix missing release of the space/qgroup reservation in start_transaction() When we fail to start a transaction, we need to release the reserved free space and qgroup space, fix it. Signed-off-by: Miao Xie Reviewed-by: Jan Schmidt Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 27 +++++++++++++++++++-------- 1 file changed, 19 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index f15494699f3b..fc03aa60b684 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -333,12 +333,14 @@ start_transaction(struct btrfs_root *root, u64 num_items, int type, &root->fs_info->trans_block_rsv, num_bytes, flush); if (ret) - return ERR_PTR(ret); + goto reserve_fail; } again: h = kmem_cache_alloc(btrfs_trans_handle_cachep, GFP_NOFS); - if (!h) - return ERR_PTR(-ENOMEM); + if (!h) { + ret = -ENOMEM; + goto alloc_fail; + } /* * If we are JOIN_NOLOCK we're already committing a transaction and @@ -365,11 +367,7 @@ again: if (ret < 0) { /* We must get the transaction if we are JOIN_NOLOCK. */ BUG_ON(type == TRANS_JOIN_NOLOCK); - - if (type < TRANS_JOIN_NOLOCK) - sb_end_intwrite(root->fs_info->sb); - kmem_cache_free(btrfs_trans_handle_cachep, h); - return ERR_PTR(ret); + goto join_fail; } cur_trans = root->fs_info->running_transaction; @@ -410,6 +408,19 @@ got_it: if (!current->journal_info && type != TRANS_USERSPACE) current->journal_info = h; return h; + +join_fail: + if (type < TRANS_JOIN_NOLOCK) + sb_end_intwrite(root->fs_info->sb); + kmem_cache_free(btrfs_trans_handle_cachep, h); +alloc_fail: + if (num_bytes) + btrfs_block_rsv_release(root, &root->fs_info->trans_block_rsv, + num_bytes); +reserve_fail: + if (qgroup_reserved) + btrfs_qgroup_free(root, qgroup_reserved); + return ERR_PTR(ret); } struct btrfs_trans_handle *btrfs_start_transaction(struct btrfs_root *root, -- cgit From 6f1c36055f96e80031c7fdda3fd5be826b8d7782 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Tue, 29 Jan 2013 03:22:10 +0000 Subject: Btrfs: fix race between snapshot deletion and getting inode While running snapshot testscript created by Mitch and David, the race between autodefrag and snapshot deletion can lead to corruption of dead_root list so that we can get crash on btrfs_clean_old_snapshots(). And besides autodefrag, scrub also does the same thing, ie. read root first and get inode. Here is the story(take autodefrag as an example): (1) when we delete a snapshot or subvolume, it will set its root's refs to zero and do a iput() on its own inode, and if this inode happens to be the only active in-meory one in root's inode rbtree, it will add itself to the global dead_roots list for later cleanup. (2) after (1), the autodefrag thread may read another inode for defrag and the inode is just in the deleted snapshot/subvolume, but all of these are without checking if the root is still valid(refs > 0). So the end up result is adding the deleted snapshot/subvolume's root to the global dead_roots list AGAIN. Fortunately, we already have a srcu lock to avoid the race, ie. subvol_srcu. So all we need to do is to take the lock to protect 'read root and get inode', since we synchronize to wait for the rcu grace period before adding something to the global dead_roots list. Reported-by: Mitch Harder Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/file.c | 22 ++++++++++++++++++---- fs/btrfs/scrub.c | 25 ++++++++++++++++++++----- 2 files changed, 38 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index a902faab7161..b06d289f998f 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -293,15 +293,24 @@ static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info, struct btrfs_key key; struct btrfs_ioctl_defrag_range_args range; int num_defrag; + int index; + int ret; /* get the inode */ key.objectid = defrag->root; btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY); key.offset = (u64)-1; + + index = srcu_read_lock(&fs_info->subvol_srcu); + inode_root = btrfs_read_fs_root_no_name(fs_info, &key); if (IS_ERR(inode_root)) { - kmem_cache_free(btrfs_inode_defrag_cachep, defrag); - return PTR_ERR(inode_root); + ret = PTR_ERR(inode_root); + goto cleanup; + } + if (btrfs_root_refs(&inode_root->root_item) == 0) { + ret = -ENOENT; + goto cleanup; } key.objectid = defrag->ino; @@ -309,9 +318,10 @@ static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info, key.offset = 0; inode = btrfs_iget(fs_info->sb, &key, inode_root, NULL); if (IS_ERR(inode)) { - kmem_cache_free(btrfs_inode_defrag_cachep, defrag); - return PTR_ERR(inode); + ret = PTR_ERR(inode); + goto cleanup; } + srcu_read_unlock(&fs_info->subvol_srcu, index); /* do a chunk of defrag */ clear_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags); @@ -346,6 +356,10 @@ static int __btrfs_run_defrag_inode(struct btrfs_fs_info *fs_info, iput(inode); return 0; +cleanup: + srcu_read_unlock(&fs_info->subvol_srcu, index); + kmem_cache_free(btrfs_inode_defrag_cachep, defrag); + return ret; } /* diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index bdbb94f245c9..67783e03d121 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -580,20 +580,29 @@ static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx) int corrected = 0; struct btrfs_key key; struct inode *inode = NULL; + struct btrfs_fs_info *fs_info; u64 end = offset + PAGE_SIZE - 1; struct btrfs_root *local_root; + int srcu_index; key.objectid = root; key.type = BTRFS_ROOT_ITEM_KEY; key.offset = (u64)-1; - local_root = btrfs_read_fs_root_no_name(fixup->root->fs_info, &key); - if (IS_ERR(local_root)) + + fs_info = fixup->root->fs_info; + srcu_index = srcu_read_lock(&fs_info->subvol_srcu); + + local_root = btrfs_read_fs_root_no_name(fs_info, &key); + if (IS_ERR(local_root)) { + srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); return PTR_ERR(local_root); + } key.type = BTRFS_INODE_ITEM_KEY; key.objectid = inum; key.offset = 0; - inode = btrfs_iget(fixup->root->fs_info->sb, &key, local_root, NULL); + inode = btrfs_iget(fs_info->sb, &key, local_root, NULL); + srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); if (IS_ERR(inode)) return PTR_ERR(inode); @@ -606,7 +615,6 @@ static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx) } if (PageUptodate(page)) { - struct btrfs_fs_info *fs_info; if (PageDirty(page)) { /* * we need to write the data to the defect sector. the @@ -3180,18 +3188,25 @@ static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root, void *ctx) u64 physical_for_dev_replace; u64 len; struct btrfs_fs_info *fs_info = nocow_ctx->sctx->dev_root->fs_info; + int srcu_index; key.objectid = root; key.type = BTRFS_ROOT_ITEM_KEY; key.offset = (u64)-1; + + srcu_index = srcu_read_lock(&fs_info->subvol_srcu); + local_root = btrfs_read_fs_root_no_name(fs_info, &key); - if (IS_ERR(local_root)) + if (IS_ERR(local_root)) { + srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); return PTR_ERR(local_root); + } key.type = BTRFS_INODE_ITEM_KEY; key.objectid = inum; key.offset = 0; inode = btrfs_iget(fs_info->sb, &key, local_root, NULL); + srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); if (IS_ERR(inode)) return PTR_ERR(inode); -- cgit From 5d1f40202bad12d4c70a2d40a420b30d23a72b1a Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 30 Jan 2013 14:17:31 -0500 Subject: Btrfs: fix missing i_size update If we have an ordered extent before the ordered extent we are currently completing that is after the current disk_i_size we will put our i_size update into that ordered extent so that we do not expose stale data. The problem is that if our disk i_size is updated past the previous ordered extent we won't update the i_size with the pending i_size update. So check the pending i_size update and if its above the current disk i_size we need to go ahead and try to update. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ordered-data.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index f10731297040..fc6840b53d9d 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -836,9 +836,16 @@ int btrfs_ordered_update_i_size(struct inode *inode, u64 offset, * if the disk i_size is already at the inode->i_size, or * this ordered extent is inside the disk i_size, we're done */ - if (disk_i_size == i_size || offset <= disk_i_size) { + if (disk_i_size == i_size) + goto out; + + /* + * We still need to update disk_i_size if outstanding_isize is greater + * than disk_i_size. + */ + if (offset <= disk_i_size && + (!ordered || ordered->outstanding_isize <= disk_i_size)) goto out; - } /* * walk backward from this ordered extent to disk_i_size. -- cgit From 59fe4f41976f6331b695ff049296d082cf621823 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 30 Jan 2013 14:31:31 -0500 Subject: Btrfs: fix possible stale data exposure We specifically do not update the disk i_size if there are ordered extents outstanding for any area between the current disk_i_size and our ordered extent so that we do not expose stale data. The problem is the check we have only checks if the ordered extent starts at or after the current disk_i_size, which doesn't take into account an ordered extent that starts before the current disk_i_size and ends past the disk_i_size. Fix this by checking if the extent ends past the disk_i_size. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ordered-data.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index fc6840b53d9d..e5ed56729607 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -877,7 +877,7 @@ int btrfs_ordered_update_i_size(struct inode *inode, u64 offset, break; if (test->file_offset >= i_size) break; - if (test->file_offset >= disk_i_size) { + if (entry_end(test) > disk_i_size) { /* * we don't update disk_i_size now, so record this * undealt i_size. Or we will not know the real -- cgit From 98783e453c1084527388ec1a7f6367cd6aabbe63 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Wed, 6 Feb 2013 14:14:26 +0800 Subject: Ext2: remove the overhead check about sb in the function ext2_new_blocks It can be guranteed that inode->i_sb should not be null in vfs. So here the check about it is overhead. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext2/balloc.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c index 2616d0ea5c5c..ea88181932df 100644 --- a/fs/ext2/balloc.c +++ b/fs/ext2/balloc.c @@ -1239,10 +1239,6 @@ ext2_fsblk_t ext2_new_blocks(struct inode *inode, ext2_fsblk_t goal, *errp = -ENOSPC; sb = inode->i_sb; - if (!sb) { - printk("ext2_new_blocks: nonexistent device"); - return 0; - } /* * Check quota for allocation of this block. -- cgit From eb6b88d92c6df083dd09a8c471011e3788dfd7c6 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Sun, 27 Jan 2013 23:26:00 -0700 Subject: Btrfs: fix EDQUOT handling in btrfs_delalloc_reserve_metadata When btrfs_qgroup_reserve returned a failure, we were missing a counter operation for BTRFS_I(inode)->outstanding_extents++, leading to warning messages about outstanding extents and space_info->bytes_may_use != 0. Additionally, the error handling code didn't take into account that we dropped the inode lock which might require more cleanup. Luckily, all the cleanup code we need is already there and can be shared with reserve_metadata_bytes, which is exactly what this patch does. Reported-by: Lev Vainblat Signed-off-by: Jan Schmidt Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 22 ++++++++++------------ 1 file changed, 10 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 85b8454d9608..61da9d0bb805 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4534,7 +4534,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) unsigned nr_extents = 0; int extra_reserve = 0; enum btrfs_reserve_flush_enum flush = BTRFS_RESERVE_FLUSH_ALL; - int ret; + int ret = 0; bool delalloc_lock = true; /* If we are a free space inode we need to not flush since we will be in @@ -4579,20 +4579,18 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) csum_bytes = BTRFS_I(inode)->csum_bytes; spin_unlock(&BTRFS_I(inode)->lock); - if (root->fs_info->quota_enabled) { + if (root->fs_info->quota_enabled) ret = btrfs_qgroup_reserve(root, num_bytes + nr_extents * root->leafsize); - if (ret) { - spin_lock(&BTRFS_I(inode)->lock); - calc_csum_metadata_size(inode, num_bytes, 0); - spin_unlock(&BTRFS_I(inode)->lock); - if (delalloc_lock) - mutex_unlock(&BTRFS_I(inode)->delalloc_mutex); - return ret; - } - } - ret = reserve_metadata_bytes(root, block_rsv, to_reserve, flush); + /* + * ret != 0 here means the qgroup reservation failed, we go straight to + * the shared error handling then. + */ + if (ret == 0) + ret = reserve_metadata_bytes(root, block_rsv, + to_reserve, flush); + if (ret) { u64 to_free = 0; unsigned dropped; -- cgit From 1a65e24b0bb7dde48cac4a2bf74d5558f9e32ba7 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Wed, 6 Feb 2013 12:06:02 -0500 Subject: Btrfs: move d_instantiate outside the transaction during mksubvol Dave Sterba triggered a lockdep complaint about lock ordering between the sb_internal lock and the cleaner semaphore. btrfs_lookup_dentry() checks for orphans if we're looking up the inode for a subvolume, and subvolume creation is triggering the lookup with a transaction running. This commit moves the d_instantiate after the transaction closes. Signed-off-by: Chris Mason --- fs/btrfs/ioctl.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index afbf3ac2079d..a31cd931d36e 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -515,7 +515,6 @@ static noinline int create_subvol(struct btrfs_root *root, BUG_ON(ret); - d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry)); fail: if (async_transid) { *async_transid = trans->transid; @@ -525,6 +524,10 @@ fail: } if (err && !ret) ret = err; + + if (!ret) + d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry)); + return ret; } -- cgit From 9fff24aa2c5c504aadead1ff9599e813604c2e53 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 6 Feb 2013 22:30:23 -0500 Subject: jbd2: track request delay statistics Track the delay between when we first request that the commit begin and when it actually begins, so we can see how much of a gap exists. In theory, this should just be the remaining scheduling quantuum of the thread which requested the commit (assuming it was not a synchronous operation which triggered the commit request) plus scheduling overhead; however, it's possible that real time processes might get in the way of letting the kjournald thread from executing. Signed-off-by: "Theodore Ts'o" --- fs/jbd2/commit.c | 8 ++++++++ fs/jbd2/journal.c | 12 +++++++++--- fs/jbd2/transaction.c | 1 + 3 files changed, 18 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 3091d42992f0..750c70148eff 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -435,7 +435,12 @@ void jbd2_journal_commit_transaction(journal_t *journal) trace_jbd2_commit_locking(journal, commit_transaction); stats.run.rs_wait = commit_transaction->t_max_wait; + stats.run.rs_request_delay = 0; stats.run.rs_locked = jiffies; + if (commit_transaction->t_requested) + stats.run.rs_request_delay = + jbd2_time_diff(commit_transaction->t_requested, + stats.run.rs_locked); stats.run.rs_running = jbd2_time_diff(commit_transaction->t_start, stats.run.rs_locked); @@ -1116,7 +1121,10 @@ restart_loop: */ spin_lock(&journal->j_history_lock); journal->j_stats.ts_tid++; + if (commit_transaction->t_requested) + journal->j_stats.ts_requested++; journal->j_stats.run.rs_wait += stats.run.rs_wait; + journal->j_stats.run.rs_request_delay += stats.run.rs_request_delay; journal->j_stats.run.rs_running += stats.run.rs_running; journal->j_stats.run.rs_locked += stats.run.rs_locked; journal->j_stats.run.rs_flushing += stats.run.rs_flushing; diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 1a80e3146a59..4ba2e81e35ac 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -533,6 +533,7 @@ int __jbd2_log_start_commit(journal_t *journal, tid_t target) jbd_debug(1, "JBD2: requesting commit %d/%d\n", journal->j_commit_request, journal->j_commit_sequence); + journal->j_running_transaction->t_requested = jiffies; wake_up(&journal->j_wait_commit); return 1; } else if (!tid_geq(journal->j_commit_request, target)) @@ -898,13 +899,18 @@ static int jbd2_seq_info_show(struct seq_file *seq, void *v) if (v != SEQ_START_TOKEN) return 0; - seq_printf(seq, "%lu transaction, each up to %u blocks\n", - s->stats->ts_tid, - s->journal->j_max_transaction_buffers); + seq_printf(seq, "%lu transactions (%lu requested), " + "each up to %u blocks\n", + s->stats->ts_tid, s->stats->ts_requested, + s->journal->j_max_transaction_buffers); if (s->stats->ts_tid == 0) return 0; seq_printf(seq, "average: \n %ums waiting for transaction\n", jiffies_to_msecs(s->stats->run.rs_wait / s->stats->ts_tid)); + seq_printf(seq, " %ums request delay\n", + (s->stats->ts_requested == 0) ? 0 : + jiffies_to_msecs(s->stats->run.rs_request_delay / + s->stats->ts_requested)); seq_printf(seq, " %ums running transaction\n", jiffies_to_msecs(s->stats->run.rs_running / s->stats->ts_tid)); seq_printf(seq, " %ums transaction was being locked\n", diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index df9f29760efa..735609e2d636 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -100,6 +100,7 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction) journal->j_running_transaction = transaction; transaction->t_max_wait = 0; transaction->t_start = jiffies; + transaction->t_requested = 0; return transaction; } -- cgit From 634734b63ac39e137a1c623ba74f3e062b6577db Mon Sep 17 00:00:00 2001 From: Eric Wong Date: Wed, 6 Feb 2013 22:29:01 +0000 Subject: fuse: allow control of adaptive readdirplus use For some filesystems (e.g. GlusterFS), the cost of performing a normal readdir and readdirplus are identical. Since adaptively using readdirplus has no benefit for those systems, give users/filesystems the option to control adaptive readdirplus use. v2 of this patch incorporates Miklos's suggestion to simplify the code, as well as improving consistency of macro names and documentation. Signed-off-by: Eric Wong Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 2 ++ fs/fuse/fuse_i.h | 5 ++++- fs/fuse/inode.c | 4 +++- 3 files changed, 9 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 2b112d978e9f..85065221a58a 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -21,6 +21,8 @@ static bool fuse_use_readdirplus(struct inode *dir, struct file *filp) if (!fc->do_readdirplus) return false; + if (!fc->readdirplus_auto) + return true; if (test_and_clear_bit(FUSE_I_ADVISE_RDPLUS, &fi->state)) return true; if (filp->f_pos == 0) diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index fc55dd33c1e2..6aeba864f070 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -514,9 +514,12 @@ struct fuse_conn { /** Use enhanced/automatic page cache invalidation. */ unsigned auto_inval_data:1; - /** Does the filesystem support readdir-plus? */ + /** Does the filesystem support readdirplus? */ unsigned do_readdirplus:1; + /** Does the filesystem want adaptive readdirplus? */ + unsigned readdirplus_auto:1; + /** The number of requests waiting for completion */ atomic_t num_waiting; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 9876a87255fe..01353ed75750 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -866,6 +866,8 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) fc->auto_inval_data = 1; if (arg->flags & FUSE_DO_READDIRPLUS) fc->do_readdirplus = 1; + if (arg->flags & FUSE_READDIRPLUS_AUTO) + fc->readdirplus_auto = 1; } else { ra_pages = fc->max_read / PAGE_CACHE_SIZE; fc->no_lock = 1; @@ -893,7 +895,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) FUSE_EXPORT_SUPPORT | FUSE_BIG_WRITES | FUSE_DONT_MASK | FUSE_SPLICE_WRITE | FUSE_SPLICE_MOVE | FUSE_SPLICE_READ | FUSE_FLOCK_LOCKS | FUSE_IOCTL_DIR | FUSE_AUTO_INVAL_DATA | - FUSE_DO_READDIRPLUS; + FUSE_DO_READDIRPLUS | FUSE_READDIRPLUS_AUTO; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); -- cgit From 311f08acde635e4e5ccea9b9d8c856cc2e0ced95 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Mon, 4 Feb 2013 10:13:11 -0600 Subject: xfs: memory barrier before wake_up_bit() In xfs_ifunlock() there is a call to wake_up_bit() after clearing the flush lock on the xfs inode. This is not guaranteed to be safe, as noted in the comments above wake_up_bit() beginning with: In order for this to function properly, as it uses waitqueue_active() internally, some kind of memory barrier must be done prior to calling this. Signed-off-by: Alex Elder Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_inode.h | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 22baf6ea4fac..237e7f6f2ab3 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -419,6 +419,7 @@ static inline void xfs_iflock(struct xfs_inode *ip) static inline void xfs_ifunlock(struct xfs_inode *ip) { xfs_iflags_clear(ip, XFS_IFLOCK); + smp_mb(); wake_up_bit(&ip->i_flags, __XFS_IFLOCK_BIT); } -- cgit From 8e3dffc651cb668e1ff4d8b89cc1c3dde7540d3b Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Thu, 7 Feb 2013 22:57:53 +0800 Subject: Ext2: mark inode dirty after the function dquot_free_block_nodirty is called We should mark inode dirty after the function dquot_free_block_nodirty is called.Besides,add a check whether it is necessary to call dquot_free_block_nodirty functon. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext2/balloc.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c index ea88181932df..132da4c0692f 100644 --- a/fs/ext2/balloc.c +++ b/fs/ext2/balloc.c @@ -568,8 +568,11 @@ do_more: } error_return: brelse(bitmap_bh); - release_blocks(sb, freed); - dquot_free_block_nodirty(inode, freed); + if (freed) { + release_blocks(sb, freed); + dquot_free_block_nodirty(inode, freed); + mark_inode_dirty(inode); + } } /** @@ -1412,9 +1415,11 @@ allocated: *errp = 0; brelse(bitmap_bh); - dquot_free_block_nodirty(inode, *count-num); - mark_inode_dirty(inode); - *count = num; + if (num < *count) { + dquot_free_block_nodirty(inode, *count-num); + mark_inode_dirty(inode); + *count = num; + } return ret_block; io_error: -- cgit From 712ddc52ffa1f5324cd8f682d9e5b047b91f34d3 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Thu, 7 Feb 2013 22:57:54 +0800 Subject: Ext2: remove the static function release_blocks to optimize the kernel Because the static function 'release_blocks' is only called when releasing blocks,it will be more simple and efficient to call the function 'percpu_counter_add' directly. Signed-off-by: Wang Shilong Signed-off-by: Jan Kara --- fs/ext2/balloc.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c index 132da4c0692f..9f9992b37924 100644 --- a/fs/ext2/balloc.c +++ b/fs/ext2/balloc.c @@ -159,15 +159,6 @@ read_block_bitmap(struct super_block *sb, unsigned int block_group) return bh; } -static void release_blocks(struct super_block *sb, int count) -{ - if (count) { - struct ext2_sb_info *sbi = EXT2_SB(sb); - - percpu_counter_add(&sbi->s_freeblocks_counter, count); - } -} - static void group_adjust_blocks(struct super_block *sb, int group_no, struct ext2_group_desc *desc, struct buffer_head *bh, int count) { @@ -569,7 +560,7 @@ do_more: error_return: brelse(bitmap_bh); if (freed) { - release_blocks(sb, freed); + percpu_counter_add(&sbi->s_freeblocks_counter, freed); dquot_free_block_nodirty(inode, freed); mark_inode_dirty(inode); } -- cgit From 8bd75c77b7c6a3954140dd2e20346aef3efe4a35 Mon Sep 17 00:00:00 2001 From: Clark Williams Date: Thu, 7 Feb 2013 09:47:07 -0600 Subject: sched/rt: Move rt specific bits into new header file Move rt scheduler definitions out of include/linux/sched.h into new file include/linux/sched/rt.h Signed-off-by: Clark Williams Cc: Peter Zijlstra Cc: Steven Rostedt Link: http://lkml.kernel.org/r/20130207094707.7b9f825f@riff.lan Signed-off-by: Ingo Molnar --- fs/select.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/select.c b/fs/select.c index 2ef72d965036..8c1c96c27062 100644 --- a/fs/select.c +++ b/fs/select.c @@ -26,6 +26,7 @@ #include #include #include +#include #include -- cgit From 53dc9a67769d0a9733adb5156adfc07edcbc1ea3 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 25 Jan 2013 16:01:21 -0500 Subject: f2fs: init_dent_inode() should take qstr for one thing, it doesn't (and shouldn't) use anything else from dentry; for another, on some call chains the dentry is fake and should be eliminated completely. Signed-off-by: Al Viro --- fs/f2fs/dir.c | 10 +++++----- fs/f2fs/f2fs.h | 2 +- fs/f2fs/node.c | 2 +- 3 files changed, 7 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 951ed52748f6..51332291b4bd 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -265,7 +265,7 @@ void f2fs_set_link(struct inode *dir, struct f2fs_dir_entry *de, mutex_unlock_op(sbi, DENTRY_OPS); } -void init_dent_inode(struct dentry *dentry, struct page *ipage) +void init_dent_inode(const struct qstr *name, struct page *ipage) { struct f2fs_node *rn; @@ -274,10 +274,10 @@ void init_dent_inode(struct dentry *dentry, struct page *ipage) wait_on_page_writeback(ipage); - /* copy dentry info. to this inode page */ + /* copy name info. to this inode page */ rn = (struct f2fs_node *)page_address(ipage); - rn->i.i_namelen = cpu_to_le32(dentry->d_name.len); - memcpy(rn->i.i_name, dentry->d_name.name, dentry->d_name.len); + rn->i.i_namelen = cpu_to_le32(name->len); + memcpy(rn->i.i_name, name->name, name->len); set_page_dirty(ipage); } @@ -310,7 +310,7 @@ static int init_inode_metadata(struct inode *inode, struct dentry *dentry) if (IS_ERR(ipage)) return PTR_ERR(ipage); set_cold_node(inode, ipage); - init_dent_inode(dentry, ipage); + init_dent_inode(&dentry->d_name, ipage); f2fs_put_page(ipage, 1); } if (is_inode_flag_set(F2FS_I(inode), FI_INC_LINK)) { diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 13c6dfbb7183..732862596311 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -867,7 +867,7 @@ struct f2fs_dir_entry *f2fs_parent_dir(struct inode *, struct page **); ino_t f2fs_inode_by_name(struct inode *, struct qstr *); void f2fs_set_link(struct inode *, struct f2fs_dir_entry *, struct page *, struct inode *); -void init_dent_inode(struct dentry *, struct page *); +void init_dent_inode(const struct qstr *, struct page *); int f2fs_add_link(struct dentry *, struct inode *); void f2fs_delete_entry(struct f2fs_dir_entry *, struct page *, struct inode *); int f2fs_make_empty(struct inode *, struct inode *); diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 5066bfd256c9..5caa94676f60 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -790,7 +790,7 @@ int new_inode_page(struct inode *inode, struct dentry *dentry) set_new_dnode(&dn, inode, NULL, NULL, inode->i_ino); mutex_lock_op(sbi, NODE_NEW); page = new_node_page(&dn, 0); - init_dent_inode(dentry, page); + init_dent_inode(&dentry->d_name, page); mutex_unlock_op(sbi, NODE_NEW); if (IS_ERR(page)) return PTR_ERR(page); -- cgit From c004363dd6aa89f1ccbebd694f261f86db0c840a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 25 Jan 2013 16:04:58 -0500 Subject: f2fs: switch new_inode_page() from dentry to qstr Signed-off-by: Al Viro --- fs/f2fs/dir.c | 2 +- fs/f2fs/f2fs.h | 2 +- fs/f2fs/node.c | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 51332291b4bd..ca7948a2770d 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -287,7 +287,7 @@ static int init_inode_metadata(struct inode *inode, struct dentry *dentry) if (is_inode_flag_set(F2FS_I(inode), FI_NEW_INODE)) { int err; - err = new_inode_page(inode, dentry); + err = new_inode_page(inode, &dentry->d_name); if (err) return err; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 732862596311..bfdc10741ff1 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -894,7 +894,7 @@ void get_node_info(struct f2fs_sb_info *, nid_t, struct node_info *); int get_dnode_of_data(struct dnode_of_data *, pgoff_t, int); int truncate_inode_blocks(struct inode *, pgoff_t); int remove_inode_page(struct inode *); -int new_inode_page(struct inode *, struct dentry *); +int new_inode_page(struct inode *, const struct qstr *); struct page *new_node_page(struct dnode_of_data *, unsigned int); void ra_node_page(struct f2fs_sb_info *, nid_t); struct page *get_node_page(struct f2fs_sb_info *, pgoff_t); diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 5caa94676f60..6625ca819716 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -780,7 +780,7 @@ int remove_inode_page(struct inode *inode) return 0; } -int new_inode_page(struct inode *inode, struct dentry *dentry) +int new_inode_page(struct inode *inode, const struct qstr *name) { struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb); struct page *page; @@ -790,7 +790,7 @@ int new_inode_page(struct inode *inode, struct dentry *dentry) set_new_dnode(&dn, inode, NULL, NULL, inode->i_ino); mutex_lock_op(sbi, NODE_NEW); page = new_node_page(&dn, 0); - init_dent_inode(&dentry->d_name, page); + init_dent_inode(name, page); mutex_unlock_op(sbi, NODE_NEW); if (IS_ERR(page)) return PTR_ERR(page); -- cgit From 69f24eac55725859a89c440ee2d19f36fa09e8fc Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 25 Jan 2013 16:08:53 -0500 Subject: f2fs: switch init_inode_metadata() to passing parent and name separately ... sure, it's tempting to just pass dentry. Except that we don't _have_ anything resembling a real dentry on one of the paths to it. Signed-off-by: Al Viro --- fs/f2fs/dir.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index ca7948a2770d..208a804180d6 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -281,13 +281,12 @@ void init_dent_inode(const struct qstr *name, struct page *ipage) set_page_dirty(ipage); } -static int init_inode_metadata(struct inode *inode, struct dentry *dentry) +static int init_inode_metadata(struct inode *inode, + struct inode *dir, const struct qstr *name) { - struct inode *dir = dentry->d_parent->d_inode; - if (is_inode_flag_set(F2FS_I(inode), FI_NEW_INODE)) { int err; - err = new_inode_page(inode, &dentry->d_name); + err = new_inode_page(inode, name); if (err) return err; @@ -310,7 +309,7 @@ static int init_inode_metadata(struct inode *inode, struct dentry *dentry) if (IS_ERR(ipage)) return PTR_ERR(ipage); set_cold_node(inode, ipage); - init_dent_inode(&dentry->d_name, ipage); + init_dent_inode(name, ipage); f2fs_put_page(ipage, 1); } if (is_inode_flag_set(F2FS_I(inode), FI_INC_LINK)) { @@ -433,7 +432,7 @@ start: ++level; goto start; add_dentry: - err = init_inode_metadata(inode, dentry); + err = init_inode_metadata(inode, dir, &dentry->d_name); if (err) goto fail; -- cgit From b7f7a5e0be94d13875a1c6c9aa65eeb11a46fc1b Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 25 Jan 2013 16:15:43 -0500 Subject: f2fs: get rid of fake on-stack dentries those should never be used for a lot of reasons... Signed-off-by: Al Viro --- fs/f2fs/dir.c | 12 +++++------- fs/f2fs/f2fs.h | 8 +++++++- fs/f2fs/recovery.c | 12 +++++------- 3 files changed, 17 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 208a804180d6..47df9252ad6a 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -370,7 +370,7 @@ next: goto next; } -int f2fs_add_link(struct dentry *dentry, struct inode *inode) +int __f2fs_add_link(struct inode *dir, const struct qstr *name, struct inode *inode) { unsigned int bit_pos; unsigned int level; @@ -379,17 +379,15 @@ int f2fs_add_link(struct dentry *dentry, struct inode *inode) f2fs_hash_t dentry_hash; struct f2fs_dir_entry *de; unsigned int nbucket, nblock; - struct inode *dir = dentry->d_parent->d_inode; struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb); - const char *name = dentry->d_name.name; - size_t namelen = dentry->d_name.len; + size_t namelen = name->len; struct page *dentry_page = NULL; struct f2fs_dentry_block *dentry_blk = NULL; int slots = GET_DENTRY_SLOTS(namelen); int err = 0; int i; - dentry_hash = f2fs_dentry_hash(name, dentry->d_name.len); + dentry_hash = f2fs_dentry_hash(name->name, name->len); level = 0; current_depth = F2FS_I(dir)->i_current_depth; if (F2FS_I(dir)->chash == dentry_hash) { @@ -432,7 +430,7 @@ start: ++level; goto start; add_dentry: - err = init_inode_metadata(inode, dir, &dentry->d_name); + err = init_inode_metadata(inode, dir, name); if (err) goto fail; @@ -441,7 +439,7 @@ add_dentry: de = &dentry_blk->dentry[bit_pos]; de->hash_code = dentry_hash; de->name_len = cpu_to_le16(namelen); - memcpy(dentry_blk->filename[bit_pos], name, namelen); + memcpy(dentry_blk->filename[bit_pos], name->name, name->len); de->ino = cpu_to_le32(inode->i_ino); set_de_type(de, inode); for (i = 0; i < slots; i++) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index bfdc10741ff1..6895ecc351ed 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -868,11 +868,17 @@ ino_t f2fs_inode_by_name(struct inode *, struct qstr *); void f2fs_set_link(struct inode *, struct f2fs_dir_entry *, struct page *, struct inode *); void init_dent_inode(const struct qstr *, struct page *); -int f2fs_add_link(struct dentry *, struct inode *); +int __f2fs_add_link(struct inode *, const struct qstr *, struct inode *); void f2fs_delete_entry(struct f2fs_dir_entry *, struct page *, struct inode *); int f2fs_make_empty(struct inode *, struct inode *); bool f2fs_empty_dir(struct inode *); +static inline int f2fs_add_link(struct dentry *dentry, struct inode *inode) +{ + return __f2fs_add_link(dentry->d_parent->d_inode, &dentry->d_name, + inode); +} + /* * super.c */ diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c index b571fee677d5..62000422879a 100644 --- a/fs/f2fs/recovery.c +++ b/fs/f2fs/recovery.c @@ -42,7 +42,7 @@ static int recover_dentry(struct page *ipage, struct inode *inode) { struct f2fs_node *raw_node = (struct f2fs_node *)kmap(ipage); struct f2fs_inode *raw_inode = &(raw_node->i); - struct dentry dent, parent; + struct qstr name; struct f2fs_dir_entry *de; struct page *page; struct inode *dir; @@ -57,17 +57,15 @@ static int recover_dentry(struct page *ipage, struct inode *inode) goto out; } - parent.d_inode = dir; - dent.d_parent = &parent; - dent.d_name.len = le32_to_cpu(raw_inode->i_namelen); - dent.d_name.name = raw_inode->i_name; + name.len = le32_to_cpu(raw_inode->i_namelen); + name.name = raw_inode->i_name; - de = f2fs_find_entry(dir, &dent.d_name, &page); + de = f2fs_find_entry(dir, &name, &page); if (de) { kunmap(page); f2fs_put_page(page, 0); } else { - f2fs_add_link(&dent, inode); + __f2fs_add_link(dir, &name, inode); } iput(dir); out: -- cgit From 343d9c283c9847da043fda3e76e3197f27b667dd Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Fri, 8 Feb 2013 13:00:22 -0500 Subject: jbd2: add tracepoints which provide per-handle statistics Handles which stay open a long time are problematic when it comes time to close down a transaction so it can be committed. These tracepoints will help us determine which ones are the problematic ones, and to validate whether changes makes things better or worse. Signed-off-by: "Theodore Ts'o" --- fs/jbd2/transaction.c | 28 ++++++++++++++++++++++++++-- 1 file changed, 26 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index 735609e2d636..b7e2385c6e92 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -30,6 +30,8 @@ #include #include +#include + static void __jbd2_journal_temp_unlink_buffer(struct journal_head *jh); static void __jbd2_journal_unfile_buffer(struct journal_head *jh); @@ -307,6 +309,8 @@ repeat: */ update_t_max_wait(transaction, ts); handle->h_transaction = transaction; + handle->h_requested_credits = nblocks; + handle->h_start_jiffies = jiffies; atomic_inc(&transaction->t_updates); atomic_inc(&transaction->t_handle_count); jbd_debug(4, "Handle %p given %d credits (total %d, free %d)\n", @@ -353,7 +357,8 @@ static handle_t *new_handle(int nblocks) * Return a pointer to a newly allocated handle, or an ERR_PTR() value * on failure. */ -handle_t *jbd2__journal_start(journal_t *journal, int nblocks, gfp_t gfp_mask) +handle_t *jbd2__journal_start(journal_t *journal, int nblocks, gfp_t gfp_mask, + unsigned int type, unsigned int line_no) { handle_t *handle = journal_current_handle(); int err; @@ -379,6 +384,11 @@ handle_t *jbd2__journal_start(journal_t *journal, int nblocks, gfp_t gfp_mask) current->journal_info = NULL; handle = ERR_PTR(err); } + handle->h_type = type; + handle->h_line_no = line_no; + trace_jbd2_handle_start(journal->j_fs_dev->bd_dev, + handle->h_transaction->t_tid, type, + line_no, nblocks); return handle; } EXPORT_SYMBOL(jbd2__journal_start); @@ -386,7 +396,7 @@ EXPORT_SYMBOL(jbd2__journal_start); handle_t *jbd2_journal_start(journal_t *journal, int nblocks) { - return jbd2__journal_start(journal, nblocks, GFP_NOFS); + return jbd2__journal_start(journal, nblocks, GFP_NOFS, 0, 0); } EXPORT_SYMBOL(jbd2_journal_start); @@ -448,7 +458,14 @@ int jbd2_journal_extend(handle_t *handle, int nblocks) goto unlock; } + trace_jbd2_handle_extend(journal->j_fs_dev->bd_dev, + handle->h_transaction->t_tid, + handle->h_type, handle->h_line_no, + handle->h_buffer_credits, + nblocks); + handle->h_buffer_credits += nblocks; + handle->h_requested_credits += nblocks; atomic_add(nblocks, &transaction->t_outstanding_credits); result = 0; @@ -1377,6 +1394,13 @@ int jbd2_journal_stop(handle_t *handle) } jbd_debug(4, "Handle %p going down\n", handle); + trace_jbd2_handle_stats(journal->j_fs_dev->bd_dev, + handle->h_transaction->t_tid, + handle->h_type, handle->h_line_no, + jiffies - handle->h_start_jiffies, + handle->h_sync, handle->h_requested_credits, + (handle->h_requested_credits - + handle->h_buffer_credits)); /* * Implement synchronous transaction batching. If the handle -- cgit From 722887ddc8982ff40e40b650fbca9ae1e56259bc Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Fri, 8 Feb 2013 13:00:31 -0500 Subject: ext4: move the jbd2 wrapper functions out of super.c Move the jbd2 wrapper functions which start and stop handles out of super.c, where they don't really logically belong, and into ext4_jbd2.c. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 2 + fs/ext4/ext4_jbd2.c | 101 +++++++++++++++++++++++++++++++++++++++++++++++++ fs/ext4/super.c | 107 +--------------------------------------------------- 3 files changed, 105 insertions(+), 105 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index d93393eb5f2d..a5ae87c51401 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2149,6 +2149,8 @@ extern void *ext4_kvzalloc(size_t size, gfp_t flags); extern void ext4_kvfree(void *ptr); extern int ext4_alloc_flex_bg_array(struct super_block *sb, ext4_group_t ngroup); +extern const char *ext4_decode_error(struct super_block *sb, int errno, + char nbuf[16]); extern __printf(4, 5) void __ext4_error(struct super_block *, const char *, unsigned int, const char *, ...); diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index b4323ba846b5..6f6114525535 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -6,6 +6,107 @@ #include +/* Just increment the non-pointer handle value */ +static handle_t *ext4_get_nojournal(void) +{ + handle_t *handle = current->journal_info; + unsigned long ref_cnt = (unsigned long)handle; + + BUG_ON(ref_cnt >= EXT4_NOJOURNAL_MAX_REF_COUNT); + + ref_cnt++; + handle = (handle_t *)ref_cnt; + + current->journal_info = handle; + return handle; +} + + +/* Decrement the non-pointer handle value */ +static void ext4_put_nojournal(handle_t *handle) +{ + unsigned long ref_cnt = (unsigned long)handle; + + BUG_ON(ref_cnt == 0); + + ref_cnt--; + handle = (handle_t *)ref_cnt; + + current->journal_info = handle; +} + +/* + * Wrappers for jbd2_journal_start/end. + */ +handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) +{ + journal_t *journal; + + trace_ext4_journal_start(sb, nblocks, _RET_IP_); + if (sb->s_flags & MS_RDONLY) + return ERR_PTR(-EROFS); + + WARN_ON(sb->s_writers.frozen == SB_FREEZE_COMPLETE); + journal = EXT4_SB(sb)->s_journal; + if (!journal) + return ext4_get_nojournal(); + /* + * Special case here: if the journal has aborted behind our + * backs (eg. EIO in the commit thread), then we still need to + * take the FS itself readonly cleanly. + */ + if (is_journal_aborted(journal)) { + ext4_abort(sb, "Detected aborted journal"); + return ERR_PTR(-EROFS); + } + return jbd2_journal_start(journal, nblocks); +} + +int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle) +{ + struct super_block *sb; + int err; + int rc; + + if (!ext4_handle_valid(handle)) { + ext4_put_nojournal(handle); + return 0; + } + sb = handle->h_transaction->t_journal->j_private; + err = handle->h_err; + rc = jbd2_journal_stop(handle); + + if (!err) + err = rc; + if (err) + __ext4_std_error(sb, where, line, err); + return err; +} + +void ext4_journal_abort_handle(const char *caller, unsigned int line, + const char *err_fn, struct buffer_head *bh, + handle_t *handle, int err) +{ + char nbuf[16]; + const char *errstr = ext4_decode_error(NULL, err, nbuf); + + BUG_ON(!ext4_handle_valid(handle)); + + if (bh) + BUFFER_TRACE(bh, "abort"); + + if (!handle->h_err) + handle->h_err = err; + + if (is_handle_aborted(handle)) + return; + + printk(KERN_ERR "EXT4-fs: %s:%d: aborting transaction: %s in %s\n", + caller, line, errstr, err_fn); + + jbd2_journal_abort_handle(handle); +} + int __ext4_journal_get_write_access(const char *where, unsigned int line, handle_t *handle, struct buffer_head *bh) { diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 2e1f94704b1f..cb9d67fbc8f0 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -69,8 +69,6 @@ static void ext4_mark_recovery_complete(struct super_block *sb, static void ext4_clear_journal_err(struct super_block *sb, struct ext4_super_block *es); static int ext4_sync_fs(struct super_block *sb, int wait); -static const char *ext4_decode_error(struct super_block *sb, int errno, - char nbuf[16]); static int ext4_remount(struct super_block *sb, int *flags, char *data); static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf); static int ext4_unfreeze(struct super_block *sb); @@ -296,107 +294,6 @@ void ext4_itable_unused_set(struct super_block *sb, } -/* Just increment the non-pointer handle value */ -static handle_t *ext4_get_nojournal(void) -{ - handle_t *handle = current->journal_info; - unsigned long ref_cnt = (unsigned long)handle; - - BUG_ON(ref_cnt >= EXT4_NOJOURNAL_MAX_REF_COUNT); - - ref_cnt++; - handle = (handle_t *)ref_cnt; - - current->journal_info = handle; - return handle; -} - - -/* Decrement the non-pointer handle value */ -static void ext4_put_nojournal(handle_t *handle) -{ - unsigned long ref_cnt = (unsigned long)handle; - - BUG_ON(ref_cnt == 0); - - ref_cnt--; - handle = (handle_t *)ref_cnt; - - current->journal_info = handle; -} - -/* - * Wrappers for jbd2_journal_start/end. - */ -handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) -{ - journal_t *journal; - - trace_ext4_journal_start(sb, nblocks, _RET_IP_); - if (sb->s_flags & MS_RDONLY) - return ERR_PTR(-EROFS); - - WARN_ON(sb->s_writers.frozen == SB_FREEZE_COMPLETE); - journal = EXT4_SB(sb)->s_journal; - if (!journal) - return ext4_get_nojournal(); - /* - * Special case here: if the journal has aborted behind our - * backs (eg. EIO in the commit thread), then we still need to - * take the FS itself readonly cleanly. - */ - if (is_journal_aborted(journal)) { - ext4_abort(sb, "Detected aborted journal"); - return ERR_PTR(-EROFS); - } - return jbd2_journal_start(journal, nblocks); -} - -int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle) -{ - struct super_block *sb; - int err; - int rc; - - if (!ext4_handle_valid(handle)) { - ext4_put_nojournal(handle); - return 0; - } - sb = handle->h_transaction->t_journal->j_private; - err = handle->h_err; - rc = jbd2_journal_stop(handle); - - if (!err) - err = rc; - if (err) - __ext4_std_error(sb, where, line, err); - return err; -} - -void ext4_journal_abort_handle(const char *caller, unsigned int line, - const char *err_fn, struct buffer_head *bh, - handle_t *handle, int err) -{ - char nbuf[16]; - const char *errstr = ext4_decode_error(NULL, err, nbuf); - - BUG_ON(!ext4_handle_valid(handle)); - - if (bh) - BUFFER_TRACE(bh, "abort"); - - if (!handle->h_err) - handle->h_err = err; - - if (is_handle_aborted(handle)) - return; - - printk(KERN_ERR "EXT4-fs: %s:%d: aborting transaction: %s in %s\n", - caller, line, errstr, err_fn); - - jbd2_journal_abort_handle(handle); -} - static void __save_error_info(struct super_block *sb, const char *func, unsigned int line) { @@ -582,8 +479,8 @@ void ext4_error_file(struct file *file, const char *function, ext4_handle_error(inode->i_sb); } -static const char *ext4_decode_error(struct super_block *sb, int errno, - char nbuf[16]) +const char *ext4_decode_error(struct super_block *sb, int errno, + char nbuf[16]) { char *errstr = NULL; -- cgit From 01a7decf75930925322c5efc87af0b5e58eb8650 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Feb 2013 11:57:27 -0500 Subject: nfsd: keep a checksum of the first 256 bytes of request Now that we're allowing more DRC entries, it becomes a lot easier to hit problems with XID collisions. In order to mitigate those, calculate a checksum of up to the first 256 bytes of each request coming in and store that in the cache entry, along with the total length of the request. This initially used crc32, but Chuck Lever and Jim Rees pointed out that crc32 is probably more heavyweight than we really need for generating these checksums, and recommended looking at using the same routines that are used to generate checksums for IP packets. On an x86_64 KVM guest measurements with ftrace showed ~800ns to use csum_partial vs ~1750ns for crc32. The difference probably isn't terribly significant, but for now we may as well use csum_partial. Signed-off-by: Jeff Layton Stones-thrown-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/cache.h | 5 +++++ fs/nfsd/nfscache.c | 47 ++++++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 49 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/cache.h b/fs/nfsd/cache.h index 9c7232b45103..87fd1410b737 100644 --- a/fs/nfsd/cache.h +++ b/fs/nfsd/cache.h @@ -29,6 +29,8 @@ struct svc_cacherep { u32 c_prot; u32 c_proc; u32 c_vers; + unsigned int c_len; + __wsum c_csum; unsigned long c_timestamp; union { struct kvec u_vec; @@ -73,6 +75,9 @@ enum { /* Cache entries expire after this time period */ #define RC_EXPIRE (120 * HZ) +/* Checksum this amount of the request */ +#define RC_CSUMLEN (256U) + int nfsd_reply_cache_init(void); void nfsd_reply_cache_shutdown(void); int nfsd_cache_lookup(struct svc_rqst *); diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index f7544698e6e6..40db57eb2b06 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -11,6 +11,7 @@ #include #include #include +#include #include "nfsd.h" #include "cache.h" @@ -130,6 +131,7 @@ int nfsd_reply_cache_init(void) INIT_LIST_HEAD(&lru_head); max_drc_entries = nfsd_cache_size_limit(); num_drc_entries = 0; + return 0; out_nomem: printk(KERN_ERR "nfsd: failed to allocate reply cache\n"); @@ -237,13 +239,46 @@ nfsd_reply_cache_shrink(struct shrinker *shrink, struct shrink_control *sc) return num; } +/* + * Walk an xdr_buf and get a CRC for at most the first RC_CSUMLEN bytes + */ +static __wsum +nfsd_cache_csum(struct svc_rqst *rqstp) +{ + int idx; + unsigned int base; + __wsum csum; + struct xdr_buf *buf = &rqstp->rq_arg; + const unsigned char *p = buf->head[0].iov_base; + size_t csum_len = min_t(size_t, buf->head[0].iov_len + buf->page_len, + RC_CSUMLEN); + size_t len = min(buf->head[0].iov_len, csum_len); + + /* rq_arg.head first */ + csum = csum_partial(p, len, 0); + csum_len -= len; + + /* Continue into page array */ + idx = buf->page_base / PAGE_SIZE; + base = buf->page_base & ~PAGE_MASK; + while (csum_len) { + p = page_address(buf->pages[idx]) + base; + len = min(PAGE_SIZE - base, csum_len); + csum = csum_partial(p, len, csum); + csum_len -= len; + base = 0; + ++idx; + } + return csum; +} + /* * Search the request hash for an entry that matches the given rqstp. * Must be called with cache_lock held. Returns the found entry or * NULL on failure. */ static struct svc_cacherep * -nfsd_cache_search(struct svc_rqst *rqstp) +nfsd_cache_search(struct svc_rqst *rqstp, __wsum csum) { struct svc_cacherep *rp; struct hlist_node *hn; @@ -257,6 +292,7 @@ nfsd_cache_search(struct svc_rqst *rqstp) hlist_for_each_entry(rp, hn, rh, c_hash) { if (xid == rp->c_xid && proc == rp->c_proc && proto == rp->c_prot && vers == rp->c_vers && + rqstp->rq_arg.len == rp->c_len && csum == rp->c_csum && rpc_cmp_addr(svc_addr(rqstp), (struct sockaddr *)&rp->c_addr) && rpc_get_port(svc_addr(rqstp)) == rpc_get_port((struct sockaddr *)&rp->c_addr)) return rp; @@ -277,6 +313,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) u32 proto = rqstp->rq_prot, vers = rqstp->rq_vers, proc = rqstp->rq_proc; + __wsum csum; unsigned long age; int type = rqstp->rq_cachetype; int rtn; @@ -287,10 +324,12 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) return RC_DOIT; } + csum = nfsd_cache_csum(rqstp); + spin_lock(&cache_lock); rtn = RC_DOIT; - rp = nfsd_cache_search(rqstp); + rp = nfsd_cache_search(rqstp, csum); if (rp) goto found_entry; @@ -318,7 +357,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) * Must search again just in case someone inserted one * after we dropped the lock above. */ - found = nfsd_cache_search(rqstp); + found = nfsd_cache_search(rqstp, csum); if (found) { nfsd_reply_cache_free_locked(rp); rp = found; @@ -344,6 +383,8 @@ setup_entry: rpc_set_port((struct sockaddr *)&rp->c_addr, rpc_get_port(svc_addr(rqstp))); rp->c_prot = proto; rp->c_vers = vers; + rp->c_len = rqstp->rq_arg.len; + rp->c_csum = csum; hash_refile(rp); lru_put_end(rp); -- cgit From 9924a92a8c217576bd2a2b1bbbb854462f1a00ae Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Fri, 8 Feb 2013 21:59:22 -0500 Subject: ext4: pass context information to jbd2__journal_start() So we can better understand what bits of ext4 are responsible for long-running jbd2 handles, use jbd2__journal_start() so we can pass context information for logging purposes. The recommended way for finding the longer-running handles is: T=/sys/kernel/debug/tracing EVENT=$T/events/jbd2/jbd2_handle_stats echo "interval > 5" > $EVENT/filter echo 1 > $EVENT/enable ./run-my-fs-benchmark cat $T/trace > /tmp/problem-handles This will list handles that were active for longer than 20ms. Having longer-running handles is bad, because a commit started at the wrong time could stall for those 20+ milliseconds, which could delay an fsync() or an O_SYNC operation. Here is an example line from the trace file describing a handle which lived on for 311 jiffies, or over 1.2 seconds: postmark-2917 [000] .... 196.435786: jbd2_handle_stats: dev 254,32 tid 570 type 2 line_no 2541 interval 311 sync 0 requested_blocks 1 dirtied_blocks 0 Signed-off-by: "Theodore Ts'o" --- fs/ext4/acl.c | 5 +++-- fs/ext4/ext4_jbd2.c | 5 +++-- fs/ext4/ext4_jbd2.h | 31 ++++++++++++++++++++++++++++--- fs/ext4/extents.c | 11 ++++++----- fs/ext4/file.c | 2 +- fs/ext4/ialloc.c | 2 +- fs/ext4/indirect.c | 7 ++++--- fs/ext4/inline.c | 10 +++++----- fs/ext4/inode.c | 33 ++++++++++++++++++++------------- fs/ext4/ioctl.c | 4 ++-- fs/ext4/migrate.c | 4 ++-- fs/ext4/move_extent.c | 2 +- fs/ext4/namei.c | 42 ++++++++++++++++++++++++------------------ fs/ext4/resize.c | 10 +++++----- fs/ext4/super.c | 10 +++++----- fs/ext4/xattr.c | 2 +- 16 files changed, 111 insertions(+), 69 deletions(-) (limited to 'fs') diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index e6e0d988439b..406cf8bb12d5 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -324,7 +324,7 @@ ext4_acl_chmod(struct inode *inode) if (error) return error; retry: - handle = ext4_journal_start(inode, + handle = ext4_journal_start(inode, EXT4_HT_XATTR, EXT4_DATA_TRANS_BLOCKS(inode->i_sb)); if (IS_ERR(handle)) { error = PTR_ERR(handle); @@ -422,7 +422,8 @@ ext4_xattr_set_acl(struct dentry *dentry, const char *name, const void *value, acl = NULL; retry: - handle = ext4_journal_start(inode, EXT4_DATA_TRANS_BLOCKS(inode->i_sb)); + handle = ext4_journal_start(inode, EXT4_HT_XATTR, + EXT4_DATA_TRANS_BLOCKS(inode->i_sb)); if (IS_ERR(handle)) { error = PTR_ERR(handle); goto release_and_out; diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index 6f6114525535..7058975e3a55 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -38,7 +38,8 @@ static void ext4_put_nojournal(handle_t *handle) /* * Wrappers for jbd2_journal_start/end. */ -handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) +handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line, + int type, int nblocks) { journal_t *journal; @@ -59,7 +60,7 @@ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) ext4_abort(sb, "Detected aborted journal"); return ERR_PTR(-EROFS); } - return jbd2_journal_start(journal, nblocks); + return jbd2__journal_start(journal, nblocks, GFP_NOFS, type, line); } int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle) diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index 7177f9b21cb2..302814b85945 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -110,6 +110,22 @@ #define EXT4_MAXQUOTAS_INIT_BLOCKS(sb) (MAXQUOTAS*EXT4_QUOTA_INIT_BLOCKS(sb)) #define EXT4_MAXQUOTAS_DEL_BLOCKS(sb) (MAXQUOTAS*EXT4_QUOTA_DEL_BLOCKS(sb)) +/* + * Ext4 handle operation types -- for logging purposes + */ +#define EXT4_HT_MISC 0 +#define EXT4_HT_INODE 1 +#define EXT4_HT_WRITE_PAGE 2 +#define EXT4_HT_MAP_BLOCKS 3 +#define EXT4_HT_DIR 4 +#define EXT4_HT_TRUNCATE 5 +#define EXT4_HT_QUOTA 6 +#define EXT4_HT_RESIZE 7 +#define EXT4_HT_MIGRATE 8 +#define EXT4_HT_MOVE_EXTENTS 9 +#define EXT4_HT_XATTR 10 +#define EXT4_HT_MAX 11 + /** * struct ext4_journal_cb_entry - Base structure for callback information. * @@ -234,7 +250,8 @@ int __ext4_handle_dirty_super(const char *where, unsigned int line, #define ext4_handle_dirty_super(handle, sb) \ __ext4_handle_dirty_super(__func__, __LINE__, (handle), (sb)) -handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks); +handle_t *__ext4_journal_start_sb(struct super_block *sb, unsigned int line, + int type, int nblocks); int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle); #define EXT4_NOJOURNAL_MAX_REF_COUNT ((unsigned long) 4096) @@ -268,9 +285,17 @@ static inline int ext4_handle_has_enough_credits(handle_t *handle, int needed) return 1; } -static inline handle_t *ext4_journal_start(struct inode *inode, int nblocks) +#define ext4_journal_start_sb(sb, type, nblocks) \ + __ext4_journal_start_sb((sb), __LINE__, (type), (nblocks)) + +#define ext4_journal_start(inode, type, nblocks) \ + __ext4_journal_start((inode), __LINE__, (type), (nblocks)) + +static inline handle_t *__ext4_journal_start(struct inode *inode, + unsigned int line, int type, + int nblocks) { - return ext4_journal_start_sb(inode->i_sb, nblocks); + return __ext4_journal_start_sb(inode->i_sb, line, type, nblocks); } #define ext4_journal_stop(handle) \ diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index db55c62e1f01..b6b54d658dc2 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -2656,7 +2656,7 @@ static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start, ext_debug("truncate since %u to %u\n", start, end); /* probably first extent we're gonna free will be last in block */ - handle = ext4_journal_start(inode, depth + 1); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, depth + 1); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -4287,7 +4287,7 @@ void ext4_ext_truncate(struct inode *inode) * probably first extent we're gonna free will be last in block */ err = ext4_writepage_trans_blocks(inode); - handle = ext4_journal_start(inode, err); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, err); if (IS_ERR(handle)) return; @@ -4454,7 +4454,8 @@ retry: while (ret >= 0 && ret < max_blocks) { map.m_lblk = map.m_lblk + ret; map.m_len = max_blocks = max_blocks - ret; - handle = ext4_journal_start(inode, credits); + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, + credits); if (IS_ERR(handle)) { ret = PTR_ERR(handle); break; @@ -4532,7 +4533,7 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset, while (ret >= 0 && ret < max_blocks) { map.m_lblk += ret; map.m_len = (max_blocks -= ret); - handle = ext4_journal_start(inode, credits); + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, credits); if (IS_ERR(handle)) { ret = PTR_ERR(handle); break; @@ -4710,7 +4711,7 @@ int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length) inode_dio_wait(inode); credits = ext4_writepage_trans_blocks(inode); - handle = ext4_journal_start(inode, credits); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); if (IS_ERR(handle)) { err = PTR_ERR(handle); goto out_dio; diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 405565a62277..2cf8ab810687 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -240,7 +240,7 @@ static int ext4_file_open(struct inode * inode, struct file * filp) handle_t *handle; int err; - handle = ext4_journal_start_sb(sb, 1); + handle = ext4_journal_start_sb(sb, EXT4_HT_MISC, 1); if (IS_ERR(handle)) return PTR_ERR(handle); err = ext4_journal_get_write_access(handle, sbi->s_sbh); diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 3f32c8012447..10bd6fecc9ff 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -1137,7 +1137,7 @@ int ext4_init_inode_table(struct super_block *sb, ext4_group_t group, if (gdp->bg_flags & cpu_to_le16(EXT4_BG_INODE_ZEROED)) goto out; - handle = ext4_journal_start_sb(sb, 1); + handle = ext4_journal_start_sb(sb, EXT4_HT_MISC, 1); if (IS_ERR(handle)) { ret = PTR_ERR(handle); goto out; diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 193281098fb6..c541ab8b64dd 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -791,7 +791,7 @@ ssize_t ext4_ind_direct_IO(int rw, struct kiocb *iocb, if (final_size > inode->i_size) { /* Credits for sb + inode write */ - handle = ext4_journal_start(inode, 2); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 2); if (IS_ERR(handle)) { ret = PTR_ERR(handle); goto out; @@ -851,7 +851,7 @@ locked: int err; /* Credits for sb + inode write */ - handle = ext4_journal_start(inode, 2); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 2); if (IS_ERR(handle)) { /* This is really bad luck. We've written the data * but cannot extend i_size. Bail out and pretend @@ -950,7 +950,8 @@ static handle_t *start_transaction(struct inode *inode) { handle_t *result; - result = ext4_journal_start(inode, ext4_blocks_for_truncate(inode)); + result = ext4_journal_start(inode, EXT4_HT_TRUNCATE, + ext4_blocks_for_truncate(inode)); if (!IS_ERR(result)) return result; diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index 93a3408fc89b..bc5f871f0893 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -545,7 +545,7 @@ static int ext4_convert_inline_data_to_extent(struct address_space *mapping, return ret; retry: - handle = ext4_journal_start(inode, needed_blocks); + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks); if (IS_ERR(handle)) { ret = PTR_ERR(handle); handle = NULL; @@ -657,7 +657,7 @@ int ext4_try_to_write_inline_data(struct address_space *mapping, * The possible write could happen in the inode, * so try to reserve the space in inode first. */ - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); if (IS_ERR(handle)) { ret = PTR_ERR(handle); handle = NULL; @@ -853,7 +853,7 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping, if (ret) return ret; - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); if (IS_ERR(handle)) { ret = PTR_ERR(handle); handle = NULL; @@ -1770,7 +1770,7 @@ void ext4_inline_data_truncate(struct inode *inode, int *has_inline) needed_blocks = ext4_writepage_trans_blocks(inode); - handle = ext4_journal_start(inode, needed_blocks); + handle = ext4_journal_start(inode, EXT4_HT_INODE, needed_blocks); if (IS_ERR(handle)) return; @@ -1862,7 +1862,7 @@ int ext4_convert_inline_data(struct inode *inode) if (error) return error; - handle = ext4_journal_start(inode, needed_blocks); + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks); if (IS_ERR(handle)) { error = PTR_ERR(handle); goto out_free; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 07d9defeaf8c..5042c8773ad7 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -234,7 +234,8 @@ void ext4_evict_inode(struct inode *inode) * protection against it */ sb_start_intwrite(inode->i_sb); - handle = ext4_journal_start(inode, ext4_blocks_for_truncate(inode)+3); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, + ext4_blocks_for_truncate(inode)+3); if (IS_ERR(handle)) { ext4_std_error(inode->i_sb, PTR_ERR(handle)); /* @@ -656,7 +657,8 @@ static int _ext4_get_block(struct inode *inode, sector_t iblock, if (map.m_len > DIO_MAX_BLOCKS) map.m_len = DIO_MAX_BLOCKS; dio_credits = ext4_chunk_trans_blocks(inode, map.m_len); - handle = ext4_journal_start(inode, dio_credits); + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, + dio_credits); if (IS_ERR(handle)) { ret = PTR_ERR(handle); return ret; @@ -881,7 +883,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping, } retry: - handle = ext4_journal_start(inode, needed_blocks); + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks); if (IS_ERR(handle)) { ret = PTR_ERR(handle); goto out; @@ -1881,7 +1883,8 @@ static int __ext4_journalled_writepage(struct page *page, * references to buffers so we are safe */ unlock_page(page); - handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode)); + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, + ext4_writepage_trans_blocks(inode)); if (IS_ERR(handle)) { ret = PTR_ERR(handle); goto out; @@ -2312,7 +2315,8 @@ retry: needed_blocks = ext4_da_writepages_trans_blocks(inode); /* start a new transaction*/ - handle = ext4_journal_start(inode, needed_blocks); + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, + needed_blocks); if (IS_ERR(handle)) { ret = PTR_ERR(handle); ext4_msg(inode->i_sb, KERN_CRIT, "%s: jbd2_start: " @@ -2468,7 +2472,7 @@ retry: * to journalling the i_disksize update if writes to the end * of file which has an already mapped buffer. */ - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, 1); if (IS_ERR(handle)) { ret = PTR_ERR(handle); goto out; @@ -4215,8 +4219,9 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) /* (user+group)*(old+new) structure, inode write (sb, * inode block, ? - but truncate inode update has it) */ - handle = ext4_journal_start(inode, (EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb)+ - EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb))+3); + handle = ext4_journal_start(inode, EXT4_HT_QUOTA, + (EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb) + + EXT4_MAXQUOTAS_DEL_BLOCKS(inode->i_sb)) + 3); if (IS_ERR(handle)) { error = PTR_ERR(handle); goto err_out; @@ -4251,7 +4256,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) (attr->ia_size < inode->i_size)) { handle_t *handle; - handle = ext4_journal_start(inode, 3); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 3); if (IS_ERR(handle)) { error = PTR_ERR(handle); goto err_out; @@ -4271,7 +4276,8 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) attr->ia_size); if (error) { /* Do as much error cleanup as possible */ - handle = ext4_journal_start(inode, 3); + handle = ext4_journal_start(inode, + EXT4_HT_INODE, 3); if (IS_ERR(handle)) { ext4_orphan_del(NULL, inode); goto err_out; @@ -4612,7 +4618,7 @@ void ext4_dirty_inode(struct inode *inode, int flags) { handle_t *handle; - handle = ext4_journal_start(inode, 2); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 2); if (IS_ERR(handle)) goto out; @@ -4713,7 +4719,7 @@ int ext4_change_inode_journal_flag(struct inode *inode, int val) /* Finally we can mark the inode as dirty. */ - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -4791,7 +4797,8 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) else get_block = ext4_get_block; retry_alloc: - handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode)); + handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, + ext4_writepage_trans_blocks(inode)); if (IS_ERR(handle)) { ret = VM_FAULT_SIGBUS; goto out; diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 4784ac244fc6..31f4f56a32d6 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -104,7 +104,7 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) } else if (oldflags & EXT4_EOFBLOCKS_FL) ext4_truncate(inode); - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); if (IS_ERR(handle)) { err = PTR_ERR(handle); goto flags_out; @@ -173,7 +173,7 @@ flags_out: } mutex_lock(&inode->i_mutex); - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_INODE, 1); if (IS_ERR(handle)) { err = PTR_ERR(handle); goto unlock_out; diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index db8226d595fa..4e4fcfd342f8 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -456,7 +456,7 @@ int ext4_ext_migrate(struct inode *inode) */ return retval; - handle = ext4_journal_start(inode, + handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, EXT4_DATA_TRANS_BLOCKS(inode->i_sb) + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb) @@ -507,7 +507,7 @@ int ext4_ext_migrate(struct inode *inode) ext4_set_inode_state(inode, EXT4_STATE_EXT_MIGRATE); up_read((&EXT4_I(inode)->i_data_sem)); - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, 1); if (IS_ERR(handle)) { /* * It is impossible to update on-disk structures without diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index e4cdb5188f34..0d6734394eac 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -923,7 +923,7 @@ move_extent_per_page(struct file *o_filp, struct inode *donor_inode, again: *err = 0; jblocks = ext4_writepage_trans_blocks(orig_inode) * 2; - handle = ext4_journal_start(orig_inode, jblocks); + handle = ext4_journal_start(orig_inode, EXT4_HT_MOVE_EXTENTS, jblocks); if (IS_ERR(handle)) { *err = PTR_ERR(handle); return 0; diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 34ed624d2c56..51841032ca03 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2262,9 +2262,10 @@ static int ext4_create(struct inode *dir, struct dentry *dentry, umode_t mode, dquot_initialize(dir); retry: - handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + - EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); + handle = ext4_journal_start(dir, EXT4_HT_DIR, + (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb))); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2298,9 +2299,10 @@ static int ext4_mknod(struct inode *dir, struct dentry *dentry, dquot_initialize(dir); retry: - handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + - EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); + handle = ext4_journal_start(dir, EXT4_HT_DIR, + (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb))); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2414,9 +2416,10 @@ static int ext4_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) dquot_initialize(dir); retry: - handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + - EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); + handle = ext4_journal_start(dir, EXT4_HT_DIR, + (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb))); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2729,7 +2732,8 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) dquot_initialize(dir); dquot_initialize(dentry->d_inode); - handle = ext4_journal_start(dir, EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); + handle = ext4_journal_start(dir, EXT4_HT_DIR, + EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2791,7 +2795,8 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) dquot_initialize(dir); dquot_initialize(dentry->d_inode); - handle = ext4_journal_start(dir, EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); + handle = ext4_journal_start(dir, EXT4_HT_DIR, + EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2870,7 +2875,7 @@ static int ext4_symlink(struct inode *dir, EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb); } retry: - handle = ext4_journal_start(dir, credits); + handle = ext4_journal_start(dir, EXT4_HT_DIR, credits); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2908,7 +2913,7 @@ retry: * Now inode is being linked into dir (EXT4_DATA_TRANS_BLOCKS * + EXT4_INDEX_EXTRA_TRANS_BLOCKS), inode is also modified */ - handle = ext4_journal_start(dir, + handle = ext4_journal_start(dir, EXT4_HT_DIR, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 1); if (IS_ERR(handle)) { @@ -2955,8 +2960,9 @@ static int ext4_link(struct dentry *old_dentry, dquot_initialize(dir); retry: - handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS); + handle = ext4_journal_start(dir, EXT4_HT_DIR, + (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS)); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -3039,9 +3045,9 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, * in separate transaction */ if (new_dentry->d_inode) dquot_initialize(new_dentry->d_inode); - handle = ext4_journal_start(old_dir, 2 * - EXT4_DATA_TRANS_BLOCKS(old_dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 2); + handle = ext4_journal_start(old_dir, EXT4_HT_DIR, + (2 * EXT4_DATA_TRANS_BLOCKS(old_dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 2)); if (IS_ERR(handle)) return PTR_ERR(handle); diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index 8eefb636beb8..c7f4d7584669 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -466,7 +466,7 @@ static int setup_new_flex_group_blocks(struct super_block *sb, meta_bg = EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_META_BG); /* This transaction may be extended/restarted along the way */ - handle = ext4_journal_start_sb(sb, EXT4_MAX_TRANS_DATA); + handle = ext4_journal_start_sb(sb, EXT4_HT_RESIZE, EXT4_MAX_TRANS_DATA); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -1031,7 +1031,7 @@ static void update_backups(struct super_block *sb, int blk_off, char *data, handle_t *handle; int err = 0, err2; - handle = ext4_journal_start_sb(sb, EXT4_MAX_TRANS_DATA); + handle = ext4_journal_start_sb(sb, EXT4_HT_RESIZE, EXT4_MAX_TRANS_DATA); if (IS_ERR(handle)) { group = 1; err = PTR_ERR(handle); @@ -1412,7 +1412,7 @@ static int ext4_flex_group_add(struct super_block *sb, * modify each of the reserved GDT dindirect blocks. */ credit = flex_gd->count * 4 + reserved_gdb; - handle = ext4_journal_start_sb(sb, credit); + handle = ext4_journal_start_sb(sb, EXT4_HT_RESIZE, credit); if (IS_ERR(handle)) { err = PTR_ERR(handle); goto exit; @@ -1624,7 +1624,7 @@ static int ext4_group_extend_no_check(struct super_block *sb, /* We will update the superblock, one block bitmap, and * one group descriptor via ext4_group_add_blocks(). */ - handle = ext4_journal_start_sb(sb, 3); + handle = ext4_journal_start_sb(sb, EXT4_HT_RESIZE, 3); if (IS_ERR(handle)) { err = PTR_ERR(handle); ext4_warning(sb, "error %d on journal start", err); @@ -1788,7 +1788,7 @@ static int ext4_convert_meta_bg(struct super_block *sb, struct inode *inode) credits += 3; /* block bitmap, bg descriptor, resize inode */ } - handle = ext4_journal_start_sb(sb, credits); + handle = ext4_journal_start_sb(sb, EXT4_HT_RESIZE, credits); if (IS_ERR(handle)) return PTR_ERR(handle); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index cb9d67fbc8f0..ef6ac59e4961 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4762,7 +4762,7 @@ static int ext4_write_dquot(struct dquot *dquot) struct inode *inode; inode = dquot_to_inode(dquot); - handle = ext4_journal_start(inode, + handle = ext4_journal_start(inode, EXT4_HT_QUOTA, EXT4_QUOTA_TRANS_BLOCKS(dquot->dq_sb)); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -4778,7 +4778,7 @@ static int ext4_acquire_dquot(struct dquot *dquot) int ret, err; handle_t *handle; - handle = ext4_journal_start(dquot_to_inode(dquot), + handle = ext4_journal_start(dquot_to_inode(dquot), EXT4_HT_QUOTA, EXT4_QUOTA_INIT_BLOCKS(dquot->dq_sb)); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -4794,7 +4794,7 @@ static int ext4_release_dquot(struct dquot *dquot) int ret, err; handle_t *handle; - handle = ext4_journal_start(dquot_to_inode(dquot), + handle = ext4_journal_start(dquot_to_inode(dquot), EXT4_HT_QUOTA, EXT4_QUOTA_DEL_BLOCKS(dquot->dq_sb)); if (IS_ERR(handle)) { /* Release dquot anyway to avoid endless cycle in dqput() */ @@ -4826,7 +4826,7 @@ static int ext4_write_info(struct super_block *sb, int type) handle_t *handle; /* Data block + inode block */ - handle = ext4_journal_start(sb->s_root->d_inode, 2); + handle = ext4_journal_start(sb->s_root->d_inode, EXT4_HT_QUOTA, 2); if (IS_ERR(handle)) return PTR_ERR(handle); ret = dquot_commit_info(sb, type); @@ -4972,7 +4972,7 @@ static int ext4_quota_off(struct super_block *sb, int type) /* Update modification times of quota files when userspace can * start looking at them */ - handle = ext4_journal_start(inode, 1); + handle = ext4_journal_start(inode, EXT4_HT_QUOTA, 1); if (IS_ERR(handle)) goto out; inode->i_mtime = inode->i_ctime = CURRENT_TIME; diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index c68990c392c7..2efc5600b03b 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -1175,7 +1175,7 @@ retry: if (ext4_has_inline_data(inode)) credits += ext4_writepage_trans_blocks(inode) + 1; - handle = ext4_journal_start(inode, credits); + handle = ext4_journal_start(inode, EXT4_HT_XATTR, credits); if (IS_ERR(handle)) { error = PTR_ERR(handle); } else { -- cgit From 47564bfb95bf370d73906fc4ae57c271e8ba96cd Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 09:24:14 -0500 Subject: ext4: grab page before starting transaction handle in write_begin() The grab_cache_page_write_begin() function can potentially sleep for a long time, since it may need to do memory allocation which can block if the system is under significant memory pressure, and because it may be blocked on page writeback. If it does take a long time to grab the page, it's better that we not hold an active jbd2 handle. So grab a handle on the page first, and _then_ start the transaction handle. This commit fixes the following long transaction handle hold time: postmark-2917 [000] .... 196.435786: jbd2_handle_stats: dev 254,32 tid 570 type 2 line_no 2541 interval 311 sync 0 requested_blocks 1 dirtied_blocks 0 Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/inode.c | 111 ++++++++++++++++++++++++++++++++++---------------------- 1 file changed, 68 insertions(+), 43 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 5042c8773ad7..2fa18bb0bf3c 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -875,32 +875,40 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping, ret = ext4_try_to_write_inline_data(mapping, inode, pos, len, flags, pagep); if (ret < 0) - goto out; - if (ret == 1) { - ret = 0; - goto out; - } + return ret; + if (ret == 1) + return 0; } -retry: + /* + * grab_cache_page_write_begin() can take a long time if the + * system is thrashing due to memory pressure, or if the page + * is being written back. So grab it first before we start + * the transaction handle. This also allows us to allocate + * the page (if needed) without using GFP_NOFS. + */ +retry_grab: + page = grab_cache_page_write_begin(mapping, index, flags); + if (!page) + return -ENOMEM; + unlock_page(page); + +retry_journal: handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, needed_blocks); if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - goto out; + page_cache_release(page); + return PTR_ERR(handle); } - /* We cannot recurse into the filesystem as the transaction is already - * started */ - flags |= AOP_FLAG_NOFS; - - page = grab_cache_page_write_begin(mapping, index, flags); - if (!page) { + lock_page(page); + if (page->mapping != mapping) { + /* The page got truncated from under us */ + unlock_page(page); + page_cache_release(page); ext4_journal_stop(handle); - ret = -ENOMEM; - goto out; + goto retry_grab; } - - *pagep = page; + wait_on_page_writeback(page); if (ext4_should_dioread_nolock(inode)) ret = __block_write_begin(page, pos, len, ext4_get_block_write); @@ -915,7 +923,6 @@ retry: if (ret) { unlock_page(page); - page_cache_release(page); /* * __block_write_begin may have instantiated a few blocks * outside i_size. Trim these off again. Don't need @@ -939,11 +946,14 @@ retry: if (inode->i_nlink) ext4_orphan_del(NULL, inode); } - } - if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) - goto retry; -out: + if (ret == -ENOSPC && + ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry_journal; + page_cache_release(page); + return ret; + } + *pagep = page; return ret; } @@ -2458,42 +2468,52 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping, pos, len, flags, pagep, fsdata); if (ret < 0) - goto out; - if (ret == 1) { - ret = 0; - goto out; - } + return ret; + if (ret == 1) + return 0; } -retry: + /* + * grab_cache_page_write_begin() can take a long time if the + * system is thrashing due to memory pressure, or if the page + * is being written back. So grab it first before we start + * the transaction handle. This also allows us to allocate + * the page (if needed) without using GFP_NOFS. + */ +retry_grab: + page = grab_cache_page_write_begin(mapping, index, flags); + if (!page) + return -ENOMEM; + unlock_page(page); + /* * With delayed allocation, we don't log the i_disksize update * if there is delayed block allocation. But we still need * to journalling the i_disksize update if writes to the end * of file which has an already mapped buffer. */ +retry_journal: handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, 1); if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - goto out; + page_cache_release(page); + return PTR_ERR(handle); } - /* We cannot recurse into the filesystem as the transaction is already - * started */ - flags |= AOP_FLAG_NOFS; - page = grab_cache_page_write_begin(mapping, index, flags); - if (!page) { + lock_page(page); + if (page->mapping != mapping) { + /* The page got truncated from under us */ + unlock_page(page); + page_cache_release(page); ext4_journal_stop(handle); - ret = -ENOMEM; - goto out; + goto retry_grab; } - *pagep = page; + /* In case writeback began while the page was unlocked */ + wait_on_page_writeback(page); ret = __block_write_begin(page, pos, len, ext4_da_get_block_prep); if (ret < 0) { unlock_page(page); ext4_journal_stop(handle); - page_cache_release(page); /* * block_write_begin may have instantiated a few blocks * outside i_size. Trim these off again. Don't need @@ -2501,11 +2521,16 @@ retry: */ if (pos + len > inode->i_size) ext4_truncate_failed_write(inode); + + if (ret == -ENOSPC && + ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry_journal; + + page_cache_release(page); + return ret; } - if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) - goto retry; -out: + *pagep = page; return ret; } -- cgit From 931b68649d31b6b52608110f34856f8eb77adb36 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 09:43:39 -0500 Subject: ext4: start handle at the last possible moment in ext4_unlink() Don't start the jbd2 transaction handle until after the directory entry has been found, to minimize the amount of time that a handle is held active. Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/namei.c | 24 ++++++++++++++---------- 1 file changed, 14 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 51841032ca03..4a27069d54ad 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2787,7 +2787,7 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) struct inode *inode; struct buffer_head *bh; struct ext4_dir_entry_2 *de; - handle_t *handle; + handle_t *handle = NULL; trace_ext4_unlink_enter(dir, dentry); /* Initialize quotas before so that eventual writes go @@ -2795,14 +2795,6 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) dquot_initialize(dir); dquot_initialize(dentry->d_inode); - handle = ext4_journal_start(dir, EXT4_HT_DIR, - EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); - if (IS_ERR(handle)) - return PTR_ERR(handle); - - if (IS_DIRSYNC(dir)) - ext4_handle_sync(handle); - retval = -ENOENT; bh = ext4_find_entry(dir, &dentry->d_name, &de, NULL); if (!bh) @@ -2814,6 +2806,17 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) if (le32_to_cpu(de->inode) != inode->i_ino) goto end_unlink; + handle = ext4_journal_start(dir, EXT4_HT_DIR, + EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); + if (IS_ERR(handle)) { + retval = PTR_ERR(handle); + handle = NULL; + goto end_unlink; + } + + if (IS_DIRSYNC(dir)) + ext4_handle_sync(handle); + if (!inode->i_nlink) { ext4_warning(inode->i_sb, "Deleting nonexistent file (%lu), %d", @@ -2834,8 +2837,9 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) retval = 0; end_unlink: - ext4_journal_stop(handle); brelse(bh); + if (handle) + ext4_journal_stop(handle); trace_ext4_unlink_exit(dentry, retval); return retval; } -- cgit From 8dcfaad244cdfa245cc2b4ddf42cea5fd8b80ece Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 09:45:11 -0500 Subject: ext4: start handle at the last possible moment in ext4_rmdir() Don't start the jbd2 transaction handle until after the directory entry has been found, to minimize the amount of time that a handle is held active. Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/namei.c | 24 ++++++++++++++---------- 1 file changed, 14 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 4a27069d54ad..36a4afd12f39 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2725,26 +2725,18 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) struct inode *inode; struct buffer_head *bh; struct ext4_dir_entry_2 *de; - handle_t *handle; + handle_t *handle = NULL; /* Initialize quotas before so that eventual writes go in * separate transaction */ dquot_initialize(dir); dquot_initialize(dentry->d_inode); - handle = ext4_journal_start(dir, EXT4_HT_DIR, - EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); - if (IS_ERR(handle)) - return PTR_ERR(handle); - retval = -ENOENT; bh = ext4_find_entry(dir, &dentry->d_name, &de, NULL); if (!bh) goto end_rmdir; - if (IS_DIRSYNC(dir)) - ext4_handle_sync(handle); - inode = dentry->d_inode; retval = -EIO; @@ -2755,6 +2747,17 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) if (!empty_dir(inode)) goto end_rmdir; + handle = ext4_journal_start(dir, EXT4_HT_DIR, + EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); + if (IS_ERR(handle)) { + retval = PTR_ERR(handle); + handle = NULL; + goto end_rmdir; + } + + if (IS_DIRSYNC(dir)) + ext4_handle_sync(handle); + retval = ext4_delete_entry(handle, dir, de, bh); if (retval) goto end_rmdir; @@ -2776,8 +2779,9 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) ext4_mark_inode_dirty(handle, dir); end_rmdir: - ext4_journal_stop(handle); brelse(bh); + if (handle) + ext4_journal_stop(handle); return retval; } -- cgit From 4b217630d0ec277c961e57f6d2985433b352c2ce Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 12:50:27 -0500 Subject: ext4: fix the number of credits needed for ext4_ext_migrate() The migration ioctl creates a temporary inode. Since this inode is never linked to a directory, we don't need to reserve journal credits required for modifying the directory. Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/migrate.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index 4e4fcfd342f8..480acf4a085f 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -456,11 +456,14 @@ int ext4_ext_migrate(struct inode *inode) */ return retval; + /* + * Worst case we can touch the allocation bitmaps, a bgd + * block, and a block to link in the orphan list. We do need + * need to worry about credits for modifying the quota inode. + */ handle = ext4_journal_start(inode, EXT4_HT_MIGRATE, - EXT4_DATA_TRANS_BLOCKS(inode->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + - EXT4_MAXQUOTAS_INIT_BLOCKS(inode->i_sb) - + 1); + 4 + EXT4_MAXQUOTAS_TRANS_BLOCKS(inode->i_sb)); + if (IS_ERR(handle)) { retval = PTR_ERR(handle); return retval; -- cgit From 64044abf05d0842a7fed30e102fa411a744c7d9f Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 15:06:24 -0500 Subject: ext4: fix the number of credits needed for ext4_unlink() and ext4_rmdir() The ext4_unlink() and ext4_rmdir() don't actually release the blocks associated with the file/directory. This gets done in a separate jbd2 handle called via ext4_evict_inode(). Thus, we don't need to reserve lots of journal credits for the truncate. Note that using too many journal credits is non-optimal because it can leading to the journal transmit getting closed too early, before it is strictly necessary. Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/ext4_jbd2.h | 6 ------ fs/ext4/namei.c | 4 ++-- 2 files changed, 2 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index 302814b85945..c1fc2dca14ae 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -59,12 +59,6 @@ #define EXT4_META_TRANS_BLOCKS(sb) (EXT4_XATTR_TRANS_BLOCKS + \ EXT4_MAXQUOTAS_TRANS_BLOCKS(sb)) -/* Delete operations potentially hit one directory's namespace plus an - * entire inode, plus arbitrary amounts of bitmap/indirection data. Be - * generous. We can grow the delete transaction later if necessary. */ - -#define EXT4_DELETE_TRANS_BLOCKS(sb) (2 * EXT4_DATA_TRANS_BLOCKS(sb) + 64) - /* Define an arbitrary limit for the amount of data we will anticipate * writing to any given transaction. For unbounded transactions such as * write(2) and truncate(2) we can write more than this, but we always diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 36a4afd12f39..5f3d2b569c63 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2748,7 +2748,7 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) goto end_rmdir; handle = ext4_journal_start(dir, EXT4_HT_DIR, - EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); + EXT4_DATA_TRANS_BLOCKS(dir->i_sb)); if (IS_ERR(handle)) { retval = PTR_ERR(handle); handle = NULL; @@ -2811,7 +2811,7 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) goto end_unlink; handle = ext4_journal_start(dir, EXT4_HT_DIR, - EXT4_DELETE_TRANS_BLOCKS(dir->i_sb)); + EXT4_DATA_TRANS_BLOCKS(dir->i_sb)); if (IS_ERR(handle)) { retval = PTR_ERR(handle); handle = NULL; -- cgit From 95eaefbdececae5e781d76d03fe7472a857c8c7a Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 15:23:03 -0500 Subject: ext4: fix the number of credits needed for acl ops with inline data Operations which modify extended attributes may need extra journal credits if inline data is used, since there is a chance that some extended attributes may need to get pushed to an external attribute block. Changes to reflect this was made in xattr.c, but they were missed in fs/ext4/acl.c. To fix this, abstract the calculation of the number of credits needed for xattr operations to an inline function defined in ext4_jbd2.h, and use it in acl.c and xattr.c. Also move the function declarations used in inline.c from xattr.h (where they are non-obviously hidden, and caused problems since ext4_jbd2.h needs to use the function ext4_has_inline_data), and move them to ext4.h. Signed-off-by: "Theodore Ts'o" Reviewed-by: Tao Ma Reviewed-by: Jan Kara --- fs/ext4/acl.c | 4 ++-- fs/ext4/ext4.h | 69 +++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/ext4/ext4_jbd2.h | 14 +++++++++++ fs/ext4/xattr.c | 9 +------ fs/ext4/xattr.h | 68 ---------------------------------------------------- 5 files changed, 86 insertions(+), 78 deletions(-) (limited to 'fs') diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index 406cf8bb12d5..39a54a0e9fe4 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -325,7 +325,7 @@ ext4_acl_chmod(struct inode *inode) return error; retry: handle = ext4_journal_start(inode, EXT4_HT_XATTR, - EXT4_DATA_TRANS_BLOCKS(inode->i_sb)); + ext4_jbd2_credits_xattr(inode)); if (IS_ERR(handle)) { error = PTR_ERR(handle); ext4_std_error(inode->i_sb, error); @@ -423,7 +423,7 @@ ext4_xattr_set_acl(struct dentry *dentry, const char *name, const void *value, retry: handle = ext4_journal_start(inode, EXT4_HT_XATTR, - EXT4_DATA_TRANS_BLOCKS(inode->i_sb)); + ext4_jbd2_credits_xattr(inode)); if (IS_ERR(handle)) { error = PTR_ERR(handle); goto release_and_out; diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index a5ae87c51401..61ecf059f70c 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2456,6 +2456,75 @@ extern const struct file_operations ext4_file_operations; extern loff_t ext4_llseek(struct file *file, loff_t offset, int origin); extern void ext4_unwritten_wait(struct inode *inode); +/* inline.c */ +extern int ext4_has_inline_data(struct inode *inode); +extern int ext4_get_inline_size(struct inode *inode); +extern int ext4_get_max_inline_size(struct inode *inode); +extern int ext4_find_inline_data_nolock(struct inode *inode); +extern void ext4_write_inline_data(struct inode *inode, + struct ext4_iloc *iloc, + void *buffer, loff_t pos, + unsigned int len); +extern int ext4_prepare_inline_data(handle_t *handle, struct inode *inode, + unsigned int len); +extern int ext4_init_inline_data(handle_t *handle, struct inode *inode, + unsigned int len); +extern int ext4_destroy_inline_data(handle_t *handle, struct inode *inode); + +extern int ext4_readpage_inline(struct inode *inode, struct page *page); +extern int ext4_try_to_write_inline_data(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + unsigned flags, + struct page **pagep); +extern int ext4_write_inline_data_end(struct inode *inode, + loff_t pos, unsigned len, + unsigned copied, + struct page *page); +extern struct buffer_head * +ext4_journalled_write_inline_data(struct inode *inode, + unsigned len, + struct page *page); +extern int ext4_da_write_inline_data_begin(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + unsigned flags, + struct page **pagep, + void **fsdata); +extern int ext4_da_write_inline_data_end(struct inode *inode, loff_t pos, + unsigned len, unsigned copied, + struct page *page); +extern int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, + struct inode *inode); +extern int ext4_try_create_inline_dir(handle_t *handle, + struct inode *parent, + struct inode *inode); +extern int ext4_read_inline_dir(struct file *filp, + void *dirent, filldir_t filldir, + int *has_inline_data); +extern struct buffer_head *ext4_find_inline_entry(struct inode *dir, + const struct qstr *d_name, + struct ext4_dir_entry_2 **res_dir, + int *has_inline_data); +extern int ext4_delete_inline_entry(handle_t *handle, + struct inode *dir, + struct ext4_dir_entry_2 *de_del, + struct buffer_head *bh, + int *has_inline_data); +extern int empty_inline_dir(struct inode *dir, int *has_inline_data); +extern struct buffer_head *ext4_get_first_inline_block(struct inode *inode, + struct ext4_dir_entry_2 **parent_de, + int *retval); +extern int ext4_inline_data_fiemap(struct inode *inode, + struct fiemap_extent_info *fieinfo, + int *has_inline); +extern int ext4_try_to_evict_inline_data(handle_t *handle, + struct inode *inode, + int needed); +extern void ext4_inline_data_truncate(struct inode *inode, int *has_inline); + +extern int ext4_convert_inline_data(struct inode *inode); + /* namei.c */ extern const struct inode_operations ext4_dir_inode_operations; extern const struct inode_operations ext4_special_inode_operations; diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index c1fc2dca14ae..4c216b1bf20c 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -104,6 +104,20 @@ #define EXT4_MAXQUOTAS_INIT_BLOCKS(sb) (MAXQUOTAS*EXT4_QUOTA_INIT_BLOCKS(sb)) #define EXT4_MAXQUOTAS_DEL_BLOCKS(sb) (MAXQUOTAS*EXT4_QUOTA_DEL_BLOCKS(sb)) +static inline int ext4_jbd2_credits_xattr(struct inode *inode) +{ + int credits = EXT4_DATA_TRANS_BLOCKS(inode->i_sb); + + /* + * In case of inline data, we may push out the data to a block, + * so we need to reserve credits for this eventuality + */ + if (ext4_has_inline_data(inode)) + credits += ext4_writepage_trans_blocks(inode) + 1; + return credits; +} + + /* * Ext4 handle operation types -- for logging purposes */ diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 2efc5600b03b..cc31da027596 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -1165,16 +1165,9 @@ ext4_xattr_set(struct inode *inode, int name_index, const char *name, { handle_t *handle; int error, retries = 0; - int credits = EXT4_DATA_TRANS_BLOCKS(inode->i_sb); + int credits = ext4_jbd2_credits_xattr(inode); retry: - /* - * In case of inline data, we may push out the data to a block, - * So reserve the journal space first. - */ - if (ext4_has_inline_data(inode)) - credits += ext4_writepage_trans_blocks(inode) + 1; - handle = ext4_journal_start(inode, EXT4_HT_XATTR, credits); if (IS_ERR(handle)) { error = PTR_ERR(handle); diff --git a/fs/ext4/xattr.h b/fs/ext4/xattr.h index 69eda787a96a..aa25deb5c6cd 100644 --- a/fs/ext4/xattr.h +++ b/fs/ext4/xattr.h @@ -125,74 +125,6 @@ extern int ext4_xattr_ibody_inline_set(handle_t *handle, struct inode *inode, struct ext4_xattr_info *i, struct ext4_xattr_ibody_find *is); -extern int ext4_has_inline_data(struct inode *inode); -extern int ext4_get_inline_size(struct inode *inode); -extern int ext4_get_max_inline_size(struct inode *inode); -extern int ext4_find_inline_data_nolock(struct inode *inode); -extern void ext4_write_inline_data(struct inode *inode, - struct ext4_iloc *iloc, - void *buffer, loff_t pos, - unsigned int len); -extern int ext4_prepare_inline_data(handle_t *handle, struct inode *inode, - unsigned int len); -extern int ext4_init_inline_data(handle_t *handle, struct inode *inode, - unsigned int len); -extern int ext4_destroy_inline_data(handle_t *handle, struct inode *inode); - -extern int ext4_readpage_inline(struct inode *inode, struct page *page); -extern int ext4_try_to_write_inline_data(struct address_space *mapping, - struct inode *inode, - loff_t pos, unsigned len, - unsigned flags, - struct page **pagep); -extern int ext4_write_inline_data_end(struct inode *inode, - loff_t pos, unsigned len, - unsigned copied, - struct page *page); -extern struct buffer_head * -ext4_journalled_write_inline_data(struct inode *inode, - unsigned len, - struct page *page); -extern int ext4_da_write_inline_data_begin(struct address_space *mapping, - struct inode *inode, - loff_t pos, unsigned len, - unsigned flags, - struct page **pagep, - void **fsdata); -extern int ext4_da_write_inline_data_end(struct inode *inode, loff_t pos, - unsigned len, unsigned copied, - struct page *page); -extern int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, - struct inode *inode); -extern int ext4_try_create_inline_dir(handle_t *handle, - struct inode *parent, - struct inode *inode); -extern int ext4_read_inline_dir(struct file *filp, - void *dirent, filldir_t filldir, - int *has_inline_data); -extern struct buffer_head *ext4_find_inline_entry(struct inode *dir, - const struct qstr *d_name, - struct ext4_dir_entry_2 **res_dir, - int *has_inline_data); -extern int ext4_delete_inline_entry(handle_t *handle, - struct inode *dir, - struct ext4_dir_entry_2 *de_del, - struct buffer_head *bh, - int *has_inline_data); -extern int empty_inline_dir(struct inode *dir, int *has_inline_data); -extern struct buffer_head *ext4_get_first_inline_block(struct inode *inode, - struct ext4_dir_entry_2 **parent_de, - int *retval); -extern int ext4_inline_data_fiemap(struct inode *inode, - struct fiemap_extent_info *fieinfo, - int *has_inline); -extern int ext4_try_to_evict_inline_data(handle_t *handle, - struct inode *inode, - int needed); -extern void ext4_inline_data_truncate(struct inode *inode, int *has_inline); - -extern int ext4_convert_inline_data(struct inode *inode); - #ifdef CONFIG_EXT4_FS_SECURITY extern int ext4_init_security(handle_t *handle, struct inode *inode, struct inode *dir, const struct qstr *qstr); -- cgit From 1139575a927010390c6b38e4215a6d741b056074 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 16:27:09 -0500 Subject: ext4: start handle at the last possible moment when creating inodes In ext4_{create,mknod,mkdir,symlink}(), don't start the journal handle until the inode has been succesfully allocated. In order to do this, we need to start the handle in the ext4_new_inode(). So create a new variant of this function, ext4_new_inode_start_handle(), so the handle can be created at the last possible minute, before we need to modify the inode allocation bitmap block. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 17 ++++++++-- fs/ext4/ialloc.c | 15 +++++++-- fs/ext4/namei.c | 94 ++++++++++++++++++++++++++------------------------------ 3 files changed, 71 insertions(+), 55 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 61ecf059f70c..fc1c0375c9f2 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2004,9 +2004,20 @@ extern int ext4fs_dirhash(const char *name, int len, struct dx_hash_info *hinfo); /* ialloc.c */ -extern struct inode *ext4_new_inode(handle_t *, struct inode *, umode_t, - const struct qstr *qstr, __u32 goal, - uid_t *owner); +extern struct inode *__ext4_new_inode(handle_t *, struct inode *, umode_t, + const struct qstr *qstr, __u32 goal, + uid_t *owner, int handle_type, + unsigned int line_no, int nblocks); + +#define ext4_new_inode(handle, dir, mode, qstr, goal, owner) \ + __ext4_new_inode((handle), (dir), (mode), (qstr), (goal), (owner), \ + 0, 0, 0) +#define ext4_new_inode_start_handle(dir, mode, qstr, goal, owner, \ + type, nblocks) \ + __ext4_new_inode(NULL, (dir), (mode), (qstr), (goal), (owner), \ + (type), __LINE__, (nblocks)) + + extern void ext4_free_inode(handle_t *, struct inode *); extern struct inode * ext4_orphan_get(struct super_block *, unsigned long); extern unsigned long ext4_count_free_inodes(struct super_block *); diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 10bd6fecc9ff..91d8fe3aced3 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -634,8 +634,10 @@ static int find_group_other(struct super_block *sb, struct inode *parent, * For other inodes, search forward from the parent directory's block * group to find a free inode. */ -struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, umode_t mode, - const struct qstr *qstr, __u32 goal, uid_t *owner) +struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir, + umode_t mode, const struct qstr *qstr, + __u32 goal, uid_t *owner, int handle_type, + unsigned int line_no, int nblocks) { struct super_block *sb; struct buffer_head *inode_bitmap_bh = NULL; @@ -725,6 +727,15 @@ repeat_in_this_group: "inode=%lu", ino + 1); continue; } + if (!handle) { + BUG_ON(nblocks <= 0); + handle = __ext4_journal_start_sb(dir->i_sb, line_no, + handle_type, nblocks); + if (IS_ERR(handle)) { + err = PTR_ERR(handle); + goto fail; + } + } BUFFER_TRACE(inode_bitmap_bh, "get_write_access"); err = ext4_journal_get_write_access(handle, inode_bitmap_bh); if (err) diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 5f3d2b569c63..3e1529c39808 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2257,30 +2257,28 @@ static int ext4_create(struct inode *dir, struct dentry *dentry, umode_t mode, { handle_t *handle; struct inode *inode; - int err, retries = 0; + int err, credits, retries = 0; dquot_initialize(dir); + credits = (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); retry: - handle = ext4_journal_start(dir, EXT4_HT_DIR, - (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + - EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb))); - if (IS_ERR(handle)) - return PTR_ERR(handle); - - if (IS_DIRSYNC(dir)) - ext4_handle_sync(handle); - - inode = ext4_new_inode(handle, dir, mode, &dentry->d_name, 0, NULL); + inode = ext4_new_inode_start_handle(dir, mode, &dentry->d_name, 0, + NULL, EXT4_HT_DIR, credits); + handle = ext4_journal_current_handle(); err = PTR_ERR(inode); if (!IS_ERR(inode)) { inode->i_op = &ext4_file_inode_operations; inode->i_fop = &ext4_file_operations; ext4_set_aops(inode); err = ext4_add_nondir(handle, dentry, inode); + if (!err && IS_DIRSYNC(dir)) + ext4_handle_sync(handle); } - ext4_journal_stop(handle); + if (handle) + ext4_journal_stop(handle); if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries)) goto retry; return err; @@ -2291,32 +2289,30 @@ static int ext4_mknod(struct inode *dir, struct dentry *dentry, { handle_t *handle; struct inode *inode; - int err, retries = 0; + int err, credits, retries = 0; if (!new_valid_dev(rdev)) return -EINVAL; dquot_initialize(dir); + credits = (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); retry: - handle = ext4_journal_start(dir, EXT4_HT_DIR, - (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + - EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb))); - if (IS_ERR(handle)) - return PTR_ERR(handle); - - if (IS_DIRSYNC(dir)) - ext4_handle_sync(handle); - - inode = ext4_new_inode(handle, dir, mode, &dentry->d_name, 0, NULL); + inode = ext4_new_inode_start_handle(dir, mode, &dentry->d_name, 0, + NULL, EXT4_HT_DIR, credits); + handle = ext4_journal_current_handle(); err = PTR_ERR(inode); if (!IS_ERR(inode)) { init_special_inode(inode, inode->i_mode, rdev); inode->i_op = &ext4_special_inode_operations; err = ext4_add_nondir(handle, dentry, inode); + if (!err && IS_DIRSYNC(dir)) + ext4_handle_sync(handle); } - ext4_journal_stop(handle); + if (handle) + ext4_journal_stop(handle); if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries)) goto retry; return err; @@ -2408,26 +2404,21 @@ static int ext4_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) { handle_t *handle; struct inode *inode; - int err, retries = 0; + int err, credits, retries = 0; if (EXT4_DIR_LINK_MAX(dir)) return -EMLINK; dquot_initialize(dir); + credits = (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); retry: - handle = ext4_journal_start(dir, EXT4_HT_DIR, - (EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + - EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb))); - if (IS_ERR(handle)) - return PTR_ERR(handle); - - if (IS_DIRSYNC(dir)) - ext4_handle_sync(handle); - - inode = ext4_new_inode(handle, dir, S_IFDIR | mode, - &dentry->d_name, 0, NULL); + inode = ext4_new_inode_start_handle(dir, S_IFDIR | mode, + &dentry->d_name, + 0, NULL, EXT4_HT_DIR, credits); + handle = ext4_journal_current_handle(); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; @@ -2455,8 +2446,12 @@ out_clear_inode: goto out_clear_inode; unlock_new_inode(inode); d_instantiate(dentry, inode); + if (IS_DIRSYNC(dir)) + ext4_handle_sync(handle); + out_stop: - ext4_journal_stop(handle); + if (handle) + ext4_journal_stop(handle); if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries)) goto retry; return err; @@ -2883,15 +2878,10 @@ static int ext4_symlink(struct inode *dir, EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb); } retry: - handle = ext4_journal_start(dir, EXT4_HT_DIR, credits); - if (IS_ERR(handle)) - return PTR_ERR(handle); - - if (IS_DIRSYNC(dir)) - ext4_handle_sync(handle); - - inode = ext4_new_inode(handle, dir, S_IFLNK|S_IRWXUGO, - &dentry->d_name, 0, NULL); + inode = ext4_new_inode_start_handle(dir, S_IFLNK|S_IRWXUGO, + &dentry->d_name, 0, NULL, + EXT4_HT_DIR, credits); + handle = ext4_journal_current_handle(); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; @@ -2944,8 +2934,12 @@ retry: } EXT4_I(inode)->i_disksize = inode->i_size; err = ext4_add_nondir(handle, dentry, inode); + if (!err && IS_DIRSYNC(dir)) + ext4_handle_sync(handle); + out_stop: - ext4_journal_stop(handle); + if (handle) + ext4_journal_stop(handle); if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries)) goto retry; return err; -- cgit From a0b30c12297eb63e9b994164f9c0937d29b9352d Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 16:28:20 -0500 Subject: ext4: use module parameters instead of debugfs for mballoc_debug There are multiple reasons to move away from debugfs. First of all, we are only using it for a single parameter, and it is much more complicated to set up (some 30 lines of code compared to 3), and one more thing that might fail while loading the ext4 module. Secondly, as a module paramter it can be specified as a boot option if ext4 is built into the kernel, or as a parameter when the module is loaded, and it can also be manipulated dynamically under /sys/module/ext4/parameters/mballoc_debug. So it is more flexible. Ultimately we want to move away from using mb_debug() towards tracepoints, but for now this is still a useful simplification of the code base. Signed-off-by: "Theodore Ts'o" --- fs/ext4/mballoc.c | 47 +++++++++-------------------------------------- fs/ext4/mballoc.h | 4 ++-- 2 files changed, 11 insertions(+), 40 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index e350885aec30..6540ebe058e3 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -23,11 +23,18 @@ #include "ext4_jbd2.h" #include "mballoc.h" -#include #include +#include #include #include +#ifdef CONFIG_EXT4_DEBUG +ushort ext4_mballoc_debug __read_mostly; + +module_param_named(mballoc_debug, ext4_mballoc_debug, ushort, 0644); +MODULE_PARM_DESC(mballoc_debug, "Debugging level for ext4's mballoc"); +#endif + /* * MUSTDO: * - test ext4_ext_search_left() and ext4_ext_search_right() @@ -2660,40 +2667,6 @@ static void ext4_free_data_callback(struct super_block *sb, mb_debug(1, "freed %u blocks in %u structures\n", count, count2); } -#ifdef CONFIG_EXT4_DEBUG -u8 mb_enable_debug __read_mostly; - -static struct dentry *debugfs_dir; -static struct dentry *debugfs_debug; - -static void __init ext4_create_debugfs_entry(void) -{ - debugfs_dir = debugfs_create_dir("ext4", NULL); - if (debugfs_dir) - debugfs_debug = debugfs_create_u8("mballoc-debug", - S_IRUGO | S_IWUSR, - debugfs_dir, - &mb_enable_debug); -} - -static void ext4_remove_debugfs_entry(void) -{ - debugfs_remove(debugfs_debug); - debugfs_remove(debugfs_dir); -} - -#else - -static void __init ext4_create_debugfs_entry(void) -{ -} - -static void ext4_remove_debugfs_entry(void) -{ -} - -#endif - int __init ext4_init_mballoc(void) { ext4_pspace_cachep = KMEM_CACHE(ext4_prealloc_space, @@ -2715,7 +2688,6 @@ int __init ext4_init_mballoc(void) kmem_cache_destroy(ext4_ac_cachep); return -ENOMEM; } - ext4_create_debugfs_entry(); return 0; } @@ -2730,7 +2702,6 @@ void ext4_exit_mballoc(void) kmem_cache_destroy(ext4_ac_cachep); kmem_cache_destroy(ext4_free_data_cachep); ext4_groupinfo_destroy_slabs(); - ext4_remove_debugfs_entry(); } @@ -3876,7 +3847,7 @@ static void ext4_mb_show_ac(struct ext4_allocation_context *ac) struct super_block *sb = ac->ac_sb; ext4_group_t ngroups, i; - if (!mb_enable_debug || + if (!ext4_mballoc_debug || (EXT4_SB(sb)->s_mount_flags & EXT4_MF_FS_ABORTED)) return; diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index 3ccd889ba953..08481ee84cd5 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -37,11 +37,11 @@ /* */ #ifdef CONFIG_EXT4_DEBUG -extern u8 mb_enable_debug; +extern ushort ext4_mballoc_debug; #define mb_debug(n, fmt, a...) \ do { \ - if ((n) <= mb_enable_debug) { \ + if ((n) <= ext4_mballoc_debug) { \ printk(KERN_DEBUG "(%s, %d): %s: ", \ __FILE__, __LINE__, __func__); \ printk(fmt, ## a); \ -- cgit From b6e96d0067d81f6a300bedee661b5ece8164e210 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 9 Feb 2013 16:29:20 -0500 Subject: jbd2: use module parameters instead of debugfs for jbd_debug There are multiple reasons to move away from debugfs. First of all, we are only using it for a single parameter, and it is much more complicated to set up (some 30 lines of code compared to 3), and one more thing that might fail while loading the jbd2 module. Secondly, as a module paramter it can be specified as a boot option if jbd2 is built into the kernel, or as a parameter when the module is loaded, and it can also be manipulated dynamically under /sys/module/jbd2/parameters/jbd2_debug. So it is more flexible. Ultimately we want to move away from using jbd_debug() towards tracepoints, but for now this is still a useful simplification of the code base. Signed-off-by: "Theodore Ts'o" --- fs/jbd2/journal.c | 50 ++++++++------------------------------------------ 1 file changed, 8 insertions(+), 42 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 4ba2e81e35ac..ed10991ab006 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -35,7 +35,6 @@ #include #include #include -#include #include #include #include @@ -51,6 +50,14 @@ #include #include +#ifdef CONFIG_JBD2_DEBUG +ushort jbd2_journal_enable_debug __read_mostly; +EXPORT_SYMBOL(jbd2_journal_enable_debug); + +module_param_named(jbd2_debug, jbd2_journal_enable_debug, ushort, 0644); +MODULE_PARM_DESC(jbd2_debug, "Debugging level for jbd2"); +#endif + EXPORT_SYMBOL(jbd2_journal_extend); EXPORT_SYMBOL(jbd2_journal_stop); EXPORT_SYMBOL(jbd2_journal_lock_updates); @@ -2495,45 +2502,6 @@ restart: spin_unlock(&journal->j_list_lock); } -/* - * debugfs tunables - */ -#ifdef CONFIG_JBD2_DEBUG -u8 jbd2_journal_enable_debug __read_mostly; -EXPORT_SYMBOL(jbd2_journal_enable_debug); - -#define JBD2_DEBUG_NAME "jbd2-debug" - -static struct dentry *jbd2_debugfs_dir; -static struct dentry *jbd2_debug; - -static void __init jbd2_create_debugfs_entry(void) -{ - jbd2_debugfs_dir = debugfs_create_dir("jbd2", NULL); - if (jbd2_debugfs_dir) - jbd2_debug = debugfs_create_u8(JBD2_DEBUG_NAME, - S_IRUGO | S_IWUSR, - jbd2_debugfs_dir, - &jbd2_journal_enable_debug); -} - -static void __exit jbd2_remove_debugfs_entry(void) -{ - debugfs_remove(jbd2_debug); - debugfs_remove(jbd2_debugfs_dir); -} - -#else - -static void __init jbd2_create_debugfs_entry(void) -{ -} - -static void __exit jbd2_remove_debugfs_entry(void) -{ -} - -#endif #ifdef CONFIG_PROC_FS @@ -2619,7 +2587,6 @@ static int __init journal_init(void) ret = journal_init_caches(); if (ret == 0) { - jbd2_create_debugfs_entry(); jbd2_create_jbd_stats_proc_entry(); } else { jbd2_journal_destroy_caches(); @@ -2634,7 +2601,6 @@ static void __exit journal_exit(void) if (n) printk(KERN_EMERG "JBD2: leaked %d journal_heads!\n", n); #endif - jbd2_remove_debugfs_entry(); jbd2_remove_jbd_stats_proc_entry(); jbd2_journal_destroy_caches(); } -- cgit From 7ffdea7ea36cd1efebab65d1b221f05deaab00cd Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 26 Jan 2013 00:11:36 +0000 Subject: locking in fs/9p ->readdir() ... is really excessive. First of all, ->readdir() is serialized by file->f_path.dentry->d_inode->i_mutex; playing with file->f_path.dentry->d_lock is not buying you anything. Moreover, rdir->mutex is pointless for exactly the same reason - you'll never see contention on it. While we are at it, there's no point in having rdir->buf a pointer - you have it point just past the end of rdir, so it might as well be a flex array (and no, it's not a gccism). Absolutely untested patch follows: Signed-off-by: Al Viro Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_dir.c | 92 +++++++++++++++------------------------------------------ 1 file changed, 23 insertions(+), 69 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index ff911e779651..be1e34adc3c6 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -52,10 +52,9 @@ */ struct p9_rdir { - struct mutex mutex; int head; int tail; - uint8_t *buf; + uint8_t buf[]; }; /** @@ -93,33 +92,12 @@ static void p9stat_init(struct p9_wstat *stbuf) * */ -static int v9fs_alloc_rdir_buf(struct file *filp, int buflen) +static struct p9_rdir *v9fs_alloc_rdir_buf(struct file *filp, int buflen) { - struct p9_rdir *rdir; - struct p9_fid *fid; - int err = 0; - - fid = filp->private_data; - if (!fid->rdir) { - rdir = kmalloc(sizeof(struct p9_rdir) + buflen, GFP_KERNEL); - - if (rdir == NULL) { - err = -ENOMEM; - goto exit; - } - spin_lock(&filp->f_dentry->d_lock); - if (!fid->rdir) { - rdir->buf = (uint8_t *)rdir + sizeof(struct p9_rdir); - mutex_init(&rdir->mutex); - rdir->head = rdir->tail = 0; - fid->rdir = (void *) rdir; - rdir = NULL; - } - spin_unlock(&filp->f_dentry->d_lock); - kfree(rdir); - } -exit: - return err; + struct p9_fid *fid = filp->private_data; + if (!fid->rdir) + fid->rdir = kzalloc(sizeof(struct p9_rdir) + buflen, GFP_KERNEL); + return fid->rdir; } /** @@ -145,20 +123,16 @@ static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir) buflen = fid->clnt->msize - P9_IOHDRSZ; - err = v9fs_alloc_rdir_buf(filp, buflen); - if (err) - goto exit; - rdir = (struct p9_rdir *) fid->rdir; + rdir = v9fs_alloc_rdir_buf(filp, buflen); + if (!rdir) + return -ENOMEM; - err = mutex_lock_interruptible(&rdir->mutex); - if (err) - return err; - while (err == 0) { + while (1) { if (rdir->tail == rdir->head) { err = v9fs_file_readn(filp, rdir->buf, NULL, buflen, filp->f_pos); if (err <= 0) - goto unlock_and_exit; + return err; rdir->head = 0; rdir->tail = err; @@ -169,9 +143,8 @@ static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir) rdir->tail - rdir->head, &st); if (err) { p9_debug(P9_DEBUG_VFS, "returned %d\n", err); - err = -EIO; p9stat_free(&st); - goto unlock_and_exit; + return -EIO; } reclen = st.size+2; @@ -180,19 +153,13 @@ static int v9fs_dir_readdir(struct file *filp, void *dirent, filldir_t filldir) p9stat_free(&st); - if (over) { - err = 0; - goto unlock_and_exit; - } + if (over) + return 0; + rdir->head += reclen; filp->f_pos += reclen; } } - -unlock_and_exit: - mutex_unlock(&rdir->mutex); -exit: - return err; } /** @@ -218,21 +185,16 @@ static int v9fs_dir_readdir_dotl(struct file *filp, void *dirent, buflen = fid->clnt->msize - P9_READDIRHDRSZ; - err = v9fs_alloc_rdir_buf(filp, buflen); - if (err) - goto exit; - rdir = (struct p9_rdir *) fid->rdir; + rdir = v9fs_alloc_rdir_buf(filp, buflen); + if (!rdir) + return -ENOMEM; - err = mutex_lock_interruptible(&rdir->mutex); - if (err) - return err; - - while (err == 0) { + while (1) { if (rdir->tail == rdir->head) { err = p9_client_readdir(fid, rdir->buf, buflen, filp->f_pos); if (err <= 0) - goto unlock_and_exit; + return err; rdir->head = 0; rdir->tail = err; @@ -245,8 +207,7 @@ static int v9fs_dir_readdir_dotl(struct file *filp, void *dirent, &curdirent); if (err < 0) { p9_debug(P9_DEBUG_VFS, "returned %d\n", err); - err = -EIO; - goto unlock_and_exit; + return -EIO; } /* d_off in dirent structure tracks the offset into @@ -261,20 +222,13 @@ static int v9fs_dir_readdir_dotl(struct file *filp, void *dirent, curdirent.d_type); oldoffset = curdirent.d_off; - if (over) { - err = 0; - goto unlock_and_exit; - } + if (over) + return 0; filp->f_pos = curdirent.d_off; rdir->head += err; } } - -unlock_and_exit: - mutex_unlock(&rdir->mutex); -exit: - return err; } -- cgit From 03f0e022736d123bc42907a958535f4f10d0c4c3 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Tue, 5 Feb 2013 11:26:28 +0530 Subject: fs/9p: Don't use O_TRUNC flag in TOPEN and TLOPEN request We do the truncate via setattr request, hence don't pass the O_TRUNC flag in open request. Without this patch we end up sending zero sized write request to server when we try to truncate. Some servers (VirtFS) were not handling that properly. Reported-by: M. Mohan Kumar Signed-off-by: Aneesh Kumar K.V Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_file.c | 4 ---- fs/9p/vfs_inode.c | 3 --- fs/9p/vfs_inode_dotl.c | 1 - 3 files changed, 8 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index c2483e97beee..7a3399767570 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -80,10 +80,6 @@ int v9fs_file_open(struct inode *inode, struct file *file) p9_client_clunk(fid); return err; } - if (file->f_flags & O_TRUNC) { - i_size_write(inode, 0); - inode->i_blocks = 0; - } if ((file->f_flags & O_APPEND) && (!v9fs_proto_dotu(v9ses) && !v9fs_proto_dotl(v9ses))) generic_file_llseek(file, 0, SEEK_END); diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 890bed538f9b..57d017ac68e4 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -192,9 +192,6 @@ int v9fs_uflags2omode(int uflags, int extended) break; } - if (uflags & O_TRUNC) - ret |= P9_OTRUNC; - if (extended) { if (uflags & O_EXCL) ret |= P9_OEXCL; diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index 40895546e103..edd41d918e6c 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -186,7 +186,6 @@ static int v9fs_mapped_dotl_flags(int flags) { O_CREAT, P9_DOTL_CREATE }, { O_EXCL, P9_DOTL_EXCL }, { O_NOCTTY, P9_DOTL_NOCTTY }, - { O_TRUNC, P9_DOTL_TRUNC }, { O_APPEND, P9_DOTL_APPEND }, { O_NONBLOCK, P9_DOTL_NONBLOCK }, { O_DSYNC, P9_DOTL_DSYNC }, -- cgit From b6f4bee02f682d1c86ece297871b78ae01afaaf4 Mon Sep 17 00:00:00 2001 From: "M. Mohan Kumar" Date: Tue, 5 Feb 2013 14:25:05 +0530 Subject: fs/9p: Fix atomic_open Return EEXISTS if requested file already exists, without this patch open call will always succeed even if the file exists and user specified O_CREAT|O_EXCL. Following test code can be used to verify this patch. Without this patch executing following test code on 9p mount will result in printing 'test case failed' always. main() { int fd; /* first create the file */ fd = open("./file", O_CREAT|O_WRONLY); if (fd < 0) { perror("open"); return -1; } close(fd); /* Now opening same file with O_CREAT|O_EXCL should fail */ fd = open("./file", O_CREAT|O_EXCL); if (fd < 0 && errno == EEXIST) printf("test case pass\n"); else printf("test case failed\n"); close(fd); return 0; } Signed-off-by: M. Mohan Kumar Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_inode_dotl.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index edd41d918e6c..8d24ad66dfb8 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -267,8 +267,14 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, } /* Only creates */ - if (!(flags & O_CREAT) || dentry->d_inode) - return finish_no_open(file, res); + if (!(flags & O_CREAT)) + return finish_no_open(file, res); + else if (dentry->d_inode) { + if ((flags & (O_CREAT | O_EXCL)) == (O_CREAT | O_EXCL)) + return -EEXIST; + else + return finish_no_open(file, res); + } v9ses = v9fs_inode2v9ses(dir); -- cgit From c137afabe330f64eddcd4dd281258807e27fd430 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 7 Feb 2013 14:41:11 -0500 Subject: NFSv4: Allow the state manager to mark an open_owner as being recovered This patch adds a seqcount_t lock for use by the state manager to signal that an open owner has been recovered. This mechanism will be used by the delegation, open and byte range lock code in order to figure out if they need to replay requests due to collisions with lock recovery. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 3 +++ fs/nfs/nfs4state.c | 10 +++++++++- 2 files changed, 12 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index a3f488b074a2..b12b73472020 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -13,6 +13,8 @@ #define NFS4_MAX_LOOP_ON_RECOVER (10) +#include + struct idmap; enum nfs4_client_state { @@ -90,6 +92,7 @@ struct nfs4_state_owner { unsigned long so_flags; struct list_head so_states; struct nfs_seqid_counter so_seqid; + seqcount_t so_reclaim_seqcount; }; enum { diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index e61f68d5ef21..fff97228cdec 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -518,6 +518,7 @@ nfs4_alloc_state_owner(struct nfs_server *server, nfs4_init_seqid_counter(&sp->so_seqid); atomic_set(&sp->so_count, 1); INIT_LIST_HEAD(&sp->so_lru); + seqcount_init(&sp->so_reclaim_seqcount); return sp; } @@ -1390,8 +1391,9 @@ static int nfs4_reclaim_open_state(struct nfs4_state_owner *sp, const struct nfs * recovering after a network partition or a reboot from a * server that doesn't support a grace period. */ -restart: spin_lock(&sp->so_lock); + write_seqcount_begin(&sp->so_reclaim_seqcount); +restart: list_for_each_entry(state, &sp->so_states, open_states) { if (!test_and_clear_bit(ops->state_flag_bit, &state->flags)) continue; @@ -1412,6 +1414,7 @@ restart: } spin_unlock(&state->state_lock); nfs4_put_open_state(state); + spin_lock(&sp->so_lock); goto restart; } } @@ -1449,12 +1452,17 @@ restart: goto out_err; } nfs4_put_open_state(state); + spin_lock(&sp->so_lock); goto restart; } + write_seqcount_end(&sp->so_reclaim_seqcount); spin_unlock(&sp->so_lock); return 0; out_err: nfs4_put_open_state(state); + spin_lock(&sp->so_lock); + write_seqcount_end(&sp->so_reclaim_seqcount); + spin_unlock(&sp->so_lock); return status; } -- cgit From 9a99af494bd7141d567d00b5ef94b141821e158c Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 4 Feb 2013 20:17:49 -0500 Subject: NFSv4.1: Prevent deadlocks between state recovery and file locking We currently have a deadlock in which the state recovery thread ends up blocking due to one of the locks which it is trying to recover holding the nfs_inode->rwsem. The situation is as follows: the state recovery thread is scheduled in order to recover from a reboot. It immediately drains the session, forcing all ordinary NFSv4.1 calls to nfs41_setup_sequence() to be put to sleep. This includes the file locking process that holds the nfs_inode->rwsem. When the thread gets to nfs4_reclaim_locks(), it tries to grab a write lock on nfs_inode->rwsem, and boom... Fix is to have the lock drop the nfs_inode->rwsem while it is doing RPC calls. We use a sequence lock in order to signal to the locking process whether or not a state recovery thread has run on that inode, in which case it should retry the lock. Reported-by: Andy Adamson Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 62fa4001e47a..92584c1ea725 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4813,8 +4813,10 @@ static int nfs41_lock_expired(struct nfs4_state *state, struct file_lock *reques static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request) { + struct nfs4_state_owner *sp = state->owner; struct nfs_inode *nfsi = NFS_I(state->inode); unsigned char fl_flags = request->fl_flags; + unsigned int seq; int status = -ENOLCK; if ((fl_flags & FL_POSIX) && @@ -4836,9 +4838,16 @@ static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock status = do_vfs_lock(request->fl_file, request); goto out_unlock; } + seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); + up_read(&nfsi->rwsem); status = _nfs4_do_setlk(state, cmd, request, NFS_LOCK_NEW); if (status != 0) + goto out; + down_read(&nfsi->rwsem); + if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) { + status = -NFS4ERR_DELAY; goto out_unlock; + } /* Note: we always want to sleep here! */ request->fl_flags = fl_flags | FL_SLEEP; if (do_vfs_lock(request->fl_file, request) < 0) -- cgit From d25be546a86fcb2ea919fb909a42b79aa40b816c Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 5 Feb 2013 11:43:28 -0500 Subject: NFSv4.1: Don't lose locks when a server reboots during delegation return If the server reboots while we are converting a delegation into OPEN/LOCK stateids as part of a delegation return, the current code will simply exit with an error. This causes us to lose both delegation state and locking state (i.e. locking atomicity). Deal with this by exposing the delegation stateid during delegation return, so that we can recover the delegation, and then resume open/lock recovery. Note that not having to hold the nfs_inode->rwsem across the calls to nfs_delegation_claim_opens() also fixes a deadlock against the NFSv4.1 reboot recovery code. Signed-off-by: Trond Myklebust --- fs/nfs/delegation.c | 147 +++++++++++++++++++++++++++++++++++++--------------- fs/nfs/delegation.h | 1 + fs/nfs/nfs4proc.c | 3 ++ 3 files changed, 108 insertions(+), 43 deletions(-) (limited to 'fs') diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 81c5eec3cf38..2542cdaa1116 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -55,7 +55,8 @@ int nfs4_have_delegation(struct inode *inode, fmode_t flags) flags &= FMODE_READ|FMODE_WRITE; rcu_read_lock(); delegation = rcu_dereference(NFS_I(inode)->delegation); - if (delegation != NULL && (delegation->type & flags) == flags) { + if (delegation != NULL && (delegation->type & flags) == flags && + !test_bit(NFS_DELEGATION_RETURNING, &delegation->flags)) { nfs_mark_delegation_referenced(delegation); ret = 1; } @@ -94,7 +95,9 @@ static int nfs_delegation_claim_opens(struct inode *inode, const nfs4_stateid *s { struct nfs_inode *nfsi = NFS_I(inode); struct nfs_open_context *ctx; + struct nfs4_state_owner *sp; struct nfs4_state *state; + unsigned int seq; int err; again: @@ -109,9 +112,13 @@ again: continue; get_nfs_open_context(ctx); spin_unlock(&inode->i_lock); + sp = state->owner; + seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); err = nfs4_open_delegation_recall(ctx, state, stateid); - if (err >= 0) + if (!err) err = nfs_delegation_claim_locks(ctx, state); + if (!err && read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) + err = -EAGAIN; put_nfs_open_context(ctx); if (err != 0) return err; @@ -181,40 +188,92 @@ static struct inode *nfs_delegation_grab_inode(struct nfs_delegation *delegation return inode; } +static struct nfs_delegation * +nfs_start_delegation_return_locked(struct nfs_inode *nfsi) +{ + struct nfs_delegation *ret = NULL; + struct nfs_delegation *delegation = rcu_dereference(nfsi->delegation); + + if (delegation == NULL) + goto out; + spin_lock(&delegation->lock); + if (!test_and_set_bit(NFS_DELEGATION_RETURNING, &delegation->flags)) + ret = delegation; + spin_unlock(&delegation->lock); +out: + return ret; +} + +static struct nfs_delegation * +nfs_start_delegation_return(struct nfs_inode *nfsi) +{ + struct nfs_delegation *delegation; + + rcu_read_lock(); + delegation = nfs_start_delegation_return_locked(nfsi); + rcu_read_unlock(); + return delegation; +} + +static void +nfs_abort_delegation_return(struct nfs_delegation *delegation, + struct nfs_client *clp) +{ + + spin_lock(&delegation->lock); + clear_bit(NFS_DELEGATION_RETURNING, &delegation->flags); + set_bit(NFS_DELEGATION_RETURN, &delegation->flags); + spin_unlock(&delegation->lock); + set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state); +} + static struct nfs_delegation * nfs_detach_delegation_locked(struct nfs_inode *nfsi, - struct nfs_server *server) + struct nfs_delegation *delegation, + struct nfs_client *clp) { - struct nfs_delegation *delegation = + struct nfs_delegation *deleg_cur = rcu_dereference_protected(nfsi->delegation, - lockdep_is_held(&server->nfs_client->cl_lock)); + lockdep_is_held(&clp->cl_lock)); - if (delegation == NULL) - goto nomatch; + if (deleg_cur == NULL || delegation != deleg_cur) + return NULL; spin_lock(&delegation->lock); + set_bit(NFS_DELEGATION_RETURNING, &delegation->flags); list_del_rcu(&delegation->super_list); delegation->inode = NULL; nfsi->delegation_state = 0; rcu_assign_pointer(nfsi->delegation, NULL); spin_unlock(&delegation->lock); return delegation; -nomatch: - return NULL; } static struct nfs_delegation *nfs_detach_delegation(struct nfs_inode *nfsi, - struct nfs_server *server) + struct nfs_delegation *delegation, + struct nfs_server *server) { struct nfs_client *clp = server->nfs_client; - struct nfs_delegation *delegation; spin_lock(&clp->cl_lock); - delegation = nfs_detach_delegation_locked(nfsi, server); + delegation = nfs_detach_delegation_locked(nfsi, delegation, clp); spin_unlock(&clp->cl_lock); return delegation; } +static struct nfs_delegation * +nfs_inode_detach_delegation(struct inode *inode) +{ + struct nfs_inode *nfsi = NFS_I(inode); + struct nfs_server *server = NFS_SERVER(inode); + struct nfs_delegation *delegation; + + delegation = nfs_start_delegation_return(nfsi); + if (delegation == NULL) + return NULL; + return nfs_detach_delegation(nfsi, delegation, server); +} + /** * nfs_inode_set_delegation - set up a delegation on an inode * @inode: inode to which delegation applies @@ -268,7 +327,10 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct delegation = NULL; goto out; } - freeme = nfs_detach_delegation_locked(nfsi, server); + freeme = nfs_detach_delegation_locked(nfsi, + old_delegation, clp); + if (freeme == NULL) + goto out; } list_add_rcu(&delegation->super_list, &server->delegations); nfsi->delegation_state = delegation->type; @@ -292,19 +354,29 @@ out: /* * Basic procedure for returning a delegation to the server */ -static int __nfs_inode_return_delegation(struct inode *inode, struct nfs_delegation *delegation, int issync) +static int nfs_end_delegation_return(struct inode *inode, struct nfs_delegation *delegation, int issync) { + struct nfs_client *clp = NFS_SERVER(inode)->nfs_client; struct nfs_inode *nfsi = NFS_I(inode); int err; - /* - * Guard against new delegated open/lock/unlock calls and against - * state recovery - */ - down_write(&nfsi->rwsem); - err = nfs_delegation_claim_opens(inode, &delegation->stateid); - up_write(&nfsi->rwsem); - if (err) + if (delegation == NULL) + return 0; + do { + err = nfs_delegation_claim_opens(inode, &delegation->stateid); + if (!issync || err != -EAGAIN) + break; + /* + * Guard against state recovery + */ + err = nfs4_wait_clnt_recover(clp); + } while (err == 0); + + if (err) { + nfs_abort_delegation_return(delegation, clp); + goto out; + } + if (!nfs_detach_delegation(nfsi, delegation, NFS_SERVER(inode))) goto out; err = nfs_do_return_delegation(inode, delegation, issync); @@ -340,13 +412,10 @@ restart: inode = nfs_delegation_grab_inode(delegation); if (inode == NULL) continue; - delegation = nfs_detach_delegation(NFS_I(inode), - server); + delegation = nfs_start_delegation_return_locked(NFS_I(inode)); rcu_read_unlock(); - if (delegation != NULL) - err = __nfs_inode_return_delegation(inode, - delegation, 0); + err = nfs_end_delegation_return(inode, delegation, 0); iput(inode); if (!err) goto restart; @@ -367,15 +436,11 @@ restart: */ void nfs_inode_return_delegation_noreclaim(struct inode *inode) { - struct nfs_server *server = NFS_SERVER(inode); - struct nfs_inode *nfsi = NFS_I(inode); struct nfs_delegation *delegation; - if (rcu_access_pointer(nfsi->delegation) != NULL) { - delegation = nfs_detach_delegation(nfsi, server); - if (delegation != NULL) - nfs_do_return_delegation(inode, delegation, 0); - } + delegation = nfs_inode_detach_delegation(inode); + if (delegation != NULL) + nfs_do_return_delegation(inode, delegation, 0); } /** @@ -390,18 +455,14 @@ void nfs_inode_return_delegation_noreclaim(struct inode *inode) */ int nfs4_inode_return_delegation(struct inode *inode) { - struct nfs_server *server = NFS_SERVER(inode); struct nfs_inode *nfsi = NFS_I(inode); struct nfs_delegation *delegation; int err = 0; nfs_wb_all(inode); - if (rcu_access_pointer(nfsi->delegation) != NULL) { - delegation = nfs_detach_delegation(nfsi, server); - if (delegation != NULL) { - err = __nfs_inode_return_delegation(inode, delegation, 1); - } - } + delegation = nfs_start_delegation_return(nfsi); + if (delegation != NULL) + err = nfs_end_delegation_return(inode, delegation, 1); return err; } @@ -471,7 +532,7 @@ void nfs_remove_bad_delegation(struct inode *inode) { struct nfs_delegation *delegation; - delegation = nfs_detach_delegation(NFS_I(inode), NFS_SERVER(inode)); + delegation = nfs_inode_detach_delegation(inode); if (delegation) { nfs_inode_find_state_and_recover(inode, &delegation->stateid); nfs_free_delegation(delegation); @@ -649,7 +710,7 @@ restart: if (inode == NULL) continue; delegation = nfs_detach_delegation(NFS_I(inode), - server); + delegation, server); rcu_read_unlock(); if (delegation != NULL) diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h index bbc6a4dba0d8..d54d4fca6793 100644 --- a/fs/nfs/delegation.h +++ b/fs/nfs/delegation.h @@ -29,6 +29,7 @@ enum { NFS_DELEGATION_NEED_RECLAIM = 0, NFS_DELEGATION_RETURN, NFS_DELEGATION_REFERENCED, + NFS_DELEGATION_RETURNING, }; int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct nfs_openres *res); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 92584c1ea725..a1e89e270ad8 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -896,6 +896,8 @@ static int can_open_delegated(struct nfs_delegation *delegation, fmode_t fmode) return 0; if (test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags)) return 0; + if (test_bit(NFS_DELEGATION_RETURNING, &delegation->flags)) + return 0; nfs_mark_delegation_referenced(delegation); return 1; } @@ -973,6 +975,7 @@ static int update_open_stateid(struct nfs4_state *state, nfs4_stateid *open_stat spin_lock(&deleg_cur->lock); if (nfsi->delegation != deleg_cur || + test_bit(NFS_DELEGATION_RETURNING, &deleg_cur->flags) || (deleg_cur->type & fmode) != fmode) goto no_delegation_unlock; -- cgit From 37380e4264dbda9753e470a30d4322097aab7152 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 5 Feb 2013 12:35:16 -0500 Subject: NFSv4: Fix up the return values of nfs4_open_delegation_recall Adjust the return values so that they return EAGAIN to the caller in cases where we might want to retry the delegation recall after the state recovery has run. Note that we can't wait and retry in this routine, because the caller may be the state manager thread. If delegation recall fails due to a session or reboot related issue, also ensure that we mark the stateid as delegated so that nfs_delegation_claim_opens can find it again later. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 25 +++++++++++-------------- 1 file changed, 11 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index a1e89e270ad8..f7e05ade5572 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1355,19 +1355,18 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state case -NFS4ERR_BAD_HIGH_SLOT: case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: case -NFS4ERR_DEADSESSION: + set_bit(NFS_DELEGATED_STATE, &state->flags); nfs4_schedule_session_recovery(server->nfs_client->cl_session, err); + err = -EAGAIN; goto out; case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_STALE_STATEID: + set_bit(NFS_DELEGATED_STATE, &state->flags); case -NFS4ERR_EXPIRED: /* Don't recall a delegation if it was lost */ nfs4_schedule_lease_recovery(server->nfs_client); + err = -EAGAIN; goto out; - case -ERESTARTSYS: - /* - * The show must go on: exit, but mark the - * stateid as needing recovery. - */ case -NFS4ERR_DELEG_REVOKED: case -NFS4ERR_ADMIN_REVOKED: case -NFS4ERR_BAD_STATEID: @@ -1378,6 +1377,7 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state err = 0; goto out; } + set_bit(NFS_DELEGATED_STATE, &state->flags); err = nfs4_handle_exception(server, err, &exception); } while (exception.retry); out: @@ -4957,24 +4957,22 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl) case 0: case -ESTALE: goto out; - case -NFS4ERR_EXPIRED: - nfs4_schedule_stateid_recovery(server, state); case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_STALE_STATEID: + set_bit(NFS_DELEGATED_STATE, &state->flags); + case -NFS4ERR_EXPIRED: nfs4_schedule_lease_recovery(server->nfs_client); + err = -EAGAIN; goto out; case -NFS4ERR_BADSESSION: case -NFS4ERR_BADSLOT: case -NFS4ERR_BAD_HIGH_SLOT: case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: case -NFS4ERR_DEADSESSION: + set_bit(NFS_DELEGATED_STATE, &state->flags); nfs4_schedule_session_recovery(server->nfs_client->cl_session, err); + err = -EAGAIN; goto out; - case -ERESTARTSYS: - /* - * The show must go on: exit, but mark the - * stateid as needing recovery. - */ case -NFS4ERR_DELEG_REVOKED: case -NFS4ERR_ADMIN_REVOKED: case -NFS4ERR_BAD_STATEID: @@ -4987,9 +4985,8 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl) /* kill_proc(fl->fl_pid, SIGLOST, 1); */ err = 0; goto out; - case -NFS4ERR_DELAY: - break; } + set_bit(NFS_DELEGATED_STATE, &state->flags); err = nfs4_handle_exception(server, err, &exception); } while (exception.retry); out: -- cgit From 65b62a29f719e937b5be1df472287f4c61e53ac6 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 7 Feb 2013 10:54:07 -0500 Subject: NFSv4: Ensure delegation recall and byte range lock removal don't conflict Add a mutex to the struct nfs4_state_owner to ensure that delegation recall doesn't conflict with byte range lock removal. Note that we nest the new mutex _outside_ the state manager reclaim protection (nfsi->rwsem) in order to avoid deadlocks. Signed-off-by: Trond Myklebust --- fs/nfs/delegation.c | 7 ++++++- fs/nfs/nfs4_fs.h | 1 + fs/nfs/nfs4proc.c | 9 ++++++++- fs/nfs/nfs4state.c | 1 + 4 files changed, 16 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 2542cdaa1116..6390a4b5fee7 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -71,8 +71,10 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_ int status = 0; if (inode->i_flock == NULL) - goto out; + return 0; + if (inode->i_flock == NULL) + goto out; /* Protect inode->i_flock using the file locks lock */ lock_flocks(); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { @@ -113,12 +115,15 @@ again: get_nfs_open_context(ctx); spin_unlock(&inode->i_lock); sp = state->owner; + /* Block nfs4_proc_unlck */ + mutex_lock(&sp->so_delegreturn_mutex); seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); err = nfs4_open_delegation_recall(ctx, state, stateid); if (!err) err = nfs_delegation_claim_locks(ctx, state); if (!err && read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) err = -EAGAIN; + mutex_unlock(&sp->so_delegreturn_mutex); put_nfs_open_context(ctx); if (err != 0) return err; diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index b12b73472020..944c9a5c1039 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -93,6 +93,7 @@ struct nfs4_state_owner { struct list_head so_states; struct nfs_seqid_counter so_seqid; seqcount_t so_reclaim_seqcount; + struct mutex so_delegreturn_mutex; }; enum { diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index f7e05ade5572..d51227371c67 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4485,7 +4485,9 @@ static struct rpc_task *nfs4_do_unlck(struct file_lock *fl, static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request) { - struct nfs_inode *nfsi = NFS_I(state->inode); + struct inode *inode = state->inode; + struct nfs4_state_owner *sp = state->owner; + struct nfs_inode *nfsi = NFS_I(inode); struct nfs_seqid *seqid; struct nfs4_lock_state *lsp; struct rpc_task *task; @@ -4495,12 +4497,17 @@ static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock * status = nfs4_set_lock_state(state, request); /* Unlock _before_ we do the RPC call */ request->fl_flags |= FL_EXISTS; + /* Exclude nfs_delegation_claim_locks() */ + mutex_lock(&sp->so_delegreturn_mutex); + /* Exclude nfs4_reclaim_open_stateid() - note nesting! */ down_read(&nfsi->rwsem); if (do_vfs_lock(request->fl_file, request) == -ENOENT) { up_read(&nfsi->rwsem); + mutex_unlock(&sp->so_delegreturn_mutex); goto out; } up_read(&nfsi->rwsem); + mutex_unlock(&sp->so_delegreturn_mutex); if (status != 0) goto out; /* Is this a delegated lock? */ diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index fff97228cdec..6ace365c6334 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -519,6 +519,7 @@ nfs4_alloc_state_owner(struct nfs_server *server, atomic_set(&sp->so_count, 1); INIT_LIST_HEAD(&sp->so_lru); seqcount_init(&sp->so_reclaim_seqcount); + mutex_init(&sp->so_delegreturn_mutex); return sp; } -- cgit From c21443c2c792cd9b463646d982b0fe48aa6feb0f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 7 Feb 2013 14:26:21 -0500 Subject: NFSv4: Fix a reboot recovery race when opening a file If the server reboots after it has replied to our OPEN, but before we call nfs4_opendata_to_nfs4_state(), then the reboot recovery thread will not see a stateid for this open, and so will fail to recover it. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 50 ++++++++++++++++++++++++++++++++++++++------------ 1 file changed, 38 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index d51227371c67..7cbf7aa6c634 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1848,6 +1848,43 @@ static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata, struct sattr->ia_valid |= ATTR_MTIME; } +static int _nfs4_open_and_get_state(struct nfs4_opendata *opendata, + fmode_t fmode, + int flags, + struct nfs4_state **res) +{ + struct nfs4_state_owner *sp = opendata->owner; + struct nfs_server *server = sp->so_server; + struct nfs4_state *state; + unsigned int seq; + int ret; + + seq = raw_seqcount_begin(&sp->so_reclaim_seqcount); + + ret = _nfs4_proc_open(opendata); + if (ret != 0) + goto out; + + state = nfs4_opendata_to_nfs4_state(opendata); + ret = PTR_ERR(state); + if (IS_ERR(state)) + goto out; + if (server->caps & NFS_CAP_POSIX_LOCK) + set_bit(NFS_STATE_POSIX_LOCKS, &state->flags); + + ret = nfs4_opendata_access(sp->so_cred, opendata, state, fmode, flags); + if (ret != 0) + goto out; + + if (read_seqcount_retry(&sp->so_reclaim_seqcount, seq)) { + nfs4_schedule_stateid_recovery(server, state); + nfs4_wait_clnt_recover(server->nfs_client); + } + *res = state; +out: + return ret; +} + /* * Returns a referenced nfs4_state */ @@ -1892,18 +1929,7 @@ static int _nfs4_do_open(struct inode *dir, if (dentry->d_inode != NULL) opendata->state = nfs4_get_open_state(dentry->d_inode, sp); - status = _nfs4_proc_open(opendata); - if (status != 0) - goto err_opendata_put; - - state = nfs4_opendata_to_nfs4_state(opendata); - status = PTR_ERR(state); - if (IS_ERR(state)) - goto err_opendata_put; - if (server->caps & NFS_CAP_POSIX_LOCK) - set_bit(NFS_STATE_POSIX_LOCKS, &state->flags); - - status = nfs4_opendata_access(cred, opendata, state, fmode, flags); + status = _nfs4_open_and_get_state(opendata, fmode, flags, &state); if (status != 0) goto err_opendata_put; -- cgit From e56a316214d0f1e2446fa7a717309f9414564d9d Mon Sep 17 00:00:00 2001 From: Fengguang Wu Date: Mon, 11 Feb 2013 16:21:42 -0500 Subject: nfsd4: free_stid can be static Reported-by: Fengguang Wu --- fs/nfsd/nfs4state.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 60a2282905e0..c1a6ddf3a84a 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -350,7 +350,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_ol_stateid *stp, struct sv return dp; } -void free_stid(struct nfs4_stid *s, struct kmem_cache *slab) +static void free_stid(struct nfs4_stid *s, struct kmem_cache *slab) { struct idr *stateids = &s->sc_client->cl_stateids; -- cgit From 7d79e75f6420fc13cfe72554b3ea5afad24c8625 Mon Sep 17 00:00:00 2001 From: Changman Lee Date: Wed, 23 Jan 2013 09:40:23 +0900 Subject: f2fs: save device node number into f2fs_inode This patch stores inode->i_rdev into on-disk inode structure. Alun reported that: aspire tmp # mount -t f2fs /dev/sdb mnt aspire tmp # mknod mnt/sda1 b 8 1 aspire tmp # mknod mnt/null c 1 3 aspire tmp # mknod mnt/console c 5 1 aspire tmp # ls -l mnt total 2 crw-r--r-- 1 root root 5, 1 Jan 22 18:44 console crw-r--r-- 1 root root 1, 3 Jan 22 18:44 null brw-r--r-- 1 root root 8, 1 Jan 22 18:44 sda1 aspire tmp # umount mnt aspire tmp # mount -t f2fs /dev/sdb mnt aspire tmp # ls -l mnt total 2 crw-r--r-- 1 root root 0, 0 Jan 22 18:44 console crw-r--r-- 1 root root 0, 0 Jan 22 18:44 null brw-r--r-- 1 root root 0, 0 Jan 22 18:44 sda1 In this report, f2fs lost the major/minor numbers of device files after umount. The reason was revealed that f2fs does not store the inode->i_rdev to the on-disk inode data structure. So, as the other file systems do, f2fs also stores i_rdev into the i_addr fields in on-disk inode structure without any on-disk layout changes. Note that, this bug is limited to device files made by mknod(). Reported-and-Tested-by: Alun Jones Signed-off-by: Changman Lee Signed-off-by: Jaegeuk Kim --- fs/f2fs/inode.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'fs') diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 794241777322..340a21be5a76 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -100,6 +100,10 @@ static int do_read_inode(struct inode *inode) inode->i_ctime.tv_nsec = le32_to_cpu(ri->i_ctime_nsec); inode->i_mtime.tv_nsec = le32_to_cpu(ri->i_mtime_nsec); inode->i_generation = le32_to_cpu(ri->i_generation); + if (ri->i_addr[0]) + inode->i_rdev = old_decode_dev(le32_to_cpu(ri->i_addr[0])); + else + inode->i_rdev = new_decode_dev(le32_to_cpu(ri->i_addr[1])); fi->i_current_depth = le32_to_cpu(ri->i_current_depth); fi->i_xattr_nid = le32_to_cpu(ri->i_xattr_nid); @@ -203,6 +207,20 @@ void update_inode(struct inode *inode, struct page *node_page) ri->i_flags = cpu_to_le32(F2FS_I(inode)->i_flags); ri->i_pino = cpu_to_le32(F2FS_I(inode)->i_pino); ri->i_generation = cpu_to_le32(inode->i_generation); + + if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) { + if (old_valid_dev(inode->i_rdev)) { + ri->i_addr[0] = + cpu_to_le32(old_encode_dev(inode->i_rdev)); + ri->i_addr[1] = 0; + } else { + ri->i_addr[0] = 0; + ri->i_addr[1] = + cpu_to_le32(new_encode_dev(inode->i_rdev)); + ri->i_addr[2] = 0; + } + } + set_cold_node(inode, node_page); set_page_dirty(node_page); } -- cgit From 577e349514452fa3fcd99fd06e587b02d3d1cf28 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Thu, 24 Jan 2013 19:56:11 +0900 Subject: f2fs: prevent checkpoint once any IO failure is detected This patch enhances the checkpoint routine to cope with IO errors. Basically f2fs detects IO errors from end_io_write, and the errors are able to be occurred during one of data, node, and meta page writes. In the previous code, when an IO error is occurred during writes, f2fs sets a flag, CP_ERROR_FLAG, in the raw ckeckpoint buffer which will be written to disk. Afterwards, write_checkpoint() will check the flag and remount f2fs as a read-only (ro) mode. However, even once f2fs is remounted as a ro mode, dirty checkpoint pages are freely able to be written to disk by flusher or kswapd in background. In such a case, after cold reboot, f2fs would restore the checkpoint data having CP_ERROR_FLAG, resulting in disabling write_checkpoint and remounting f2fs as a ro mode again. Therefore, let's prevent any checkpoint page (meta) writes once an IO error is occurred, and remount f2fs as a ro mode right away at that moment. Reported-by: Oliver Winker Signed-off-by: Jaegeuk Kim Reviewed-by: Namjae Jeon --- fs/f2fs/checkpoint.c | 36 +++++++++++++++++++----------------- fs/f2fs/f2fs.h | 3 +-- fs/f2fs/segment.c | 8 ++------ fs/f2fs/super.c | 12 +++++++++--- 4 files changed, 31 insertions(+), 28 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index ff3c8439af87..9c1627165039 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -72,22 +72,22 @@ static int f2fs_write_meta_page(struct page *page, { struct inode *inode = page->mapping->host; struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb); - int err; - wait_on_page_writeback(page); - - err = write_meta_page(sbi, page, wbc); - if (err) { + /* Should not write any meta pages, if any IO error was occurred */ + if (wbc->for_reclaim || + is_set_ckpt_flags(F2FS_CKPT(sbi), CP_ERROR_FLAG)) { + dec_page_count(sbi, F2FS_DIRTY_META); wbc->pages_skipped++; set_page_dirty(page); + return AOP_WRITEPAGE_ACTIVATE; } - dec_page_count(sbi, F2FS_DIRTY_META); + wait_on_page_writeback(page); - /* In this case, we should not unlock this page */ - if (err != AOP_WRITEPAGE_ACTIVATE) - unlock_page(page); - return err; + write_meta_page(sbi, page); + dec_page_count(sbi, F2FS_DIRTY_META); + unlock_page(page); + return 0; } static int f2fs_write_meta_pages(struct address_space *mapping, @@ -138,7 +138,10 @@ long sync_meta_pages(struct f2fs_sb_info *sbi, enum page_type type, BUG_ON(page->mapping != mapping); BUG_ON(!PageDirty(page)); clear_page_dirty_for_io(page); - f2fs_write_meta_page(page, &wbc); + if (f2fs_write_meta_page(page, &wbc)) { + unlock_page(page); + break; + } if (nwritten++ >= nr_to_write) break; } @@ -717,13 +720,12 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, bool is_umount) sbi->alloc_valid_block_count = 0; /* Here, we only have one bio having CP pack */ - if (is_set_ckpt_flags(ckpt, CP_ERROR_FLAG)) - sbi->sb->s_flags |= MS_RDONLY; - else - sync_meta_pages(sbi, META_FLUSH, LONG_MAX); + sync_meta_pages(sbi, META_FLUSH, LONG_MAX); - clear_prefree_segments(sbi); - F2FS_RESET_SB_DIRT(sbi); + if (!is_set_ckpt_flags(ckpt, CP_ERROR_FLAG)) { + clear_prefree_segments(sbi); + F2FS_RESET_SB_DIRT(sbi); + } } /* diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index c8e2d751ef9c..f4f509766465 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -929,8 +929,7 @@ void allocate_new_segments(struct f2fs_sb_info *); struct page *get_sum_page(struct f2fs_sb_info *, unsigned int); struct bio *f2fs_bio_alloc(struct block_device *, int); void f2fs_submit_bio(struct f2fs_sb_info *, enum page_type, bool sync); -int write_meta_page(struct f2fs_sb_info *, struct page *, - struct writeback_control *); +void write_meta_page(struct f2fs_sb_info *, struct page *); void write_node_page(struct f2fs_sb_info *, struct page *, unsigned int, block_t, block_t *); void write_data_page(struct inode *, struct page *, struct dnode_of_data*, diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 4b0099066582..7aa270f3538a 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -600,6 +600,7 @@ static void f2fs_end_io_write(struct bio *bio, int err) if (page->mapping) set_bit(AS_EIO, &page->mapping->flags); set_ckpt_flags(p->sbi->ckpt, CP_ERROR_FLAG); + p->sbi->sb->s_flags |= MS_RDONLY; } end_page_writeback(page); dec_page_count(p->sbi, F2FS_WRITEBACK); @@ -815,15 +816,10 @@ static void do_write_page(struct f2fs_sb_info *sbi, struct page *page, mutex_unlock(&curseg->curseg_mutex); } -int write_meta_page(struct f2fs_sb_info *sbi, struct page *page, - struct writeback_control *wbc) +void write_meta_page(struct f2fs_sb_info *sbi, struct page *page) { - if (wbc->for_reclaim) - return AOP_WRITEPAGE_ACTIVATE; - set_page_writeback(page); submit_write_page(sbi, page, page->index, META); - return 0; } void write_node_page(struct f2fs_sb_info *sbi, struct page *page, diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 37fad04c8669..117ca2a46e39 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -387,10 +387,11 @@ static int sanity_check_raw_super(struct super_block *sb, return 0; } -static int sanity_check_ckpt(struct f2fs_super_block *raw_super, - struct f2fs_checkpoint *ckpt) +static int sanity_check_ckpt(struct f2fs_sb_info *sbi) { unsigned int total, fsmeta; + struct f2fs_super_block *raw_super = F2FS_RAW_SUPER(sbi); + struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi); total = le32_to_cpu(raw_super->segment_count); fsmeta = le32_to_cpu(raw_super->segment_count_ckpt); @@ -401,6 +402,11 @@ static int sanity_check_ckpt(struct f2fs_super_block *raw_super, if (fsmeta >= total) return 1; + + if (is_set_ckpt_flags(ckpt, CP_ERROR_FLAG)) { + f2fs_msg(sbi->sb, KERN_ERR, "A bug case: need to run fsck"); + return 1; + } return 0; } @@ -525,7 +531,7 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent) /* sanity checking of checkpoint */ err = -EINVAL; - if (sanity_check_ckpt(raw_super, sbi->ckpt)) { + if (sanity_check_ckpt(sbi)) { f2fs_msg(sb, KERN_ERR, "Invalid F2FS checkpoint"); goto free_cp; } -- cgit From bd43df021ac37247f2db58ff376fb4032170f754 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Fri, 25 Jan 2013 18:33:41 +0900 Subject: f2fs: cover global locks for reserve_new_block The fill_zero() from fallocate() calls get_new_data_page() in which calls reserve_new_block(). The reserve_new_block() should be covered by *DATA_NEW*, one of global locks. And also, before getting the lock, we should check free sections by calling f2fs_balance_fs(). If we break this rule, f2fs is able to face with out-of-control free space management and fall into infinite loop like the following scenario as well. [f2fs_sync_fs()] [fallocate()] - write_checkpoint() - fill_zero() - block_operations() - get_new_data_page() : grab NODE_NEW - get_dnode_of_data() : get locked dirty node page - sync_node_pages() : try to grab NODE_NEW for data allocation : trylock and skip the dirty node page : call sync_node_pages() repeatedly in order to flush all the dirty node pages! In order to avoid this, we should grab another global lock such as DATA_NEW before calling get_new_data_page() in fill_zero(). Signed-off-by: Jaegeuk Kim --- fs/f2fs/file.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 3191b52aafb0..6cdab2c64fc6 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -387,12 +387,17 @@ const struct inode_operations f2fs_file_inode_operations = { static void fill_zero(struct inode *inode, pgoff_t index, loff_t start, loff_t len) { + struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb); struct page *page; if (!len) return; + f2fs_balance_fs(sbi); + + mutex_lock_op(sbi, DATA_NEW); page = get_new_data_page(inode, index, false); + mutex_unlock_op(sbi, DATA_NEW); if (!IS_ERR(page)) { wait_on_page_writeback(page); -- cgit From aa43507f68e44fbc014d820e47eda5a369b8dc9d Mon Sep 17 00:00:00 2001 From: Alejandro Martinez Ruiz Date: Fri, 25 Jan 2013 19:08:59 +0100 Subject: f2fs: fix disable_ext_identify option spelling There is a typo in the ->show_options function for disable_ext_identify. Fix it to match the spelling from the documentation. Signed-off-by: Alejandro Martinez Ruiz Signed-off-by: Jaegeuk Kim --- fs/f2fs/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 117ca2a46e39..25656b6869a4 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -198,7 +198,7 @@ static int f2fs_show_options(struct seq_file *seq, struct dentry *root) seq_puts(seq, ",noacl"); #endif if (test_opt(sbi, DISABLE_EXT_IDENTIFY)) - seq_puts(seq, ",disable_ext_indentify"); + seq_puts(seq, ",disable_ext_identify"); seq_printf(seq, ",active_logs=%u", sbi->active_logs); -- cgit From a2617dc6863b21a8109c199ab533b3dbfe178f27 Mon Sep 17 00:00:00 2001 From: majianpeng Date: Tue, 29 Jan 2013 16:19:02 +0800 Subject: f2fs: clean up the add_orphan_inode func For the code > prev = list_entry(orphan->list.prev, typeof(*prev), list); if orphan->list.prev == head, it can't get the right prev. And we can use the parameter 'this' to add. Signed-off-by: Jianpeng Ma Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 9c1627165039..d3b34d05211f 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -219,19 +219,11 @@ retry: new->ino = ino; /* add new_oentry into list which is sorted by inode number */ - if (orphan) { - struct orphan_inode_entry *prev; - - /* get previous entry */ - prev = list_entry(orphan->list.prev, typeof(*prev), list); - if (&prev->list != head) - /* insert new orphan inode entry */ - list_add(&new->list, &prev->list); - else - list_add(&new->list, head); - } else { + if (orphan) + list_add(&new->list, this->prev); + else list_add_tail(&new->list, head); - } + sbi->n_orphans++; out: mutex_unlock(&sbi->orphan_inode_mutex); -- cgit From d6212a5f18c8f9f9cc884070a96e11907711217f Mon Sep 17 00:00:00 2001 From: Changman Lee Date: Tue, 29 Jan 2013 18:30:07 +0900 Subject: f2fs: add un/freeze_fs into super_operations This patch supports ioctl FIFREEZE and FITHAW to snapshot filesystem. Before calling f2fs_freeze, all writers would be suspended and sync_fs would be completed. So no f2fs has to do something. Just background gc operation should be skipped due to generate dirty nodes and data until unfreeze. Signed-off-by: Changman Lee Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 5 +++++ fs/f2fs/inode.c | 2 ++ fs/f2fs/super.c | 18 ++++++++++++++++++ 3 files changed, 25 insertions(+) (limited to 'fs') diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index c386910dacc5..9b5d0aad5daf 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -44,6 +44,11 @@ static int gc_thread_func(void *data) if (kthread_should_stop()) break; + if (sbi->sb->s_writers.frozen >= SB_FREEZE_WRITE) { + wait_ms = GC_THREAD_MAX_SLEEP_TIME; + continue; + } + f2fs_balance_fs(sbi); if (!test_opt(sbi, BG_GC)) diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 340a21be5a76..62433c63e2db 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -278,6 +278,7 @@ void f2fs_evict_inode(struct inode *inode) if (inode->i_nlink || is_bad_inode(inode)) goto no_delete; + sb_start_intwrite(inode->i_sb); set_inode_flag(F2FS_I(inode), FI_NO_ALLOC); i_size_write(inode, 0); @@ -285,6 +286,7 @@ void f2fs_evict_inode(struct inode *inode) f2fs_truncate(inode); remove_inode_page(inode); + sb_end_intwrite(inode->i_sb); no_delete: clear_inode(inode); } diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 25656b6869a4..0b18aee2ed25 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -143,6 +143,22 @@ int f2fs_sync_fs(struct super_block *sb, int sync) return 0; } +static int f2fs_freeze(struct super_block *sb) +{ + int err; + + if (sb->s_flags & MS_RDONLY) + return 0; + + err = f2fs_sync_fs(sb, 1); + return err; +} + +static int f2fs_unfreeze(struct super_block *sb) +{ + return 0; +} + static int f2fs_statfs(struct dentry *dentry, struct kstatfs *buf) { struct super_block *sb = dentry->d_sb; @@ -213,6 +229,8 @@ static struct super_operations f2fs_sops = { .evict_inode = f2fs_evict_inode, .put_super = f2fs_put_super, .sync_fs = f2fs_sync_fs, + .freeze_fs = f2fs_freeze, + .unfreeze_fs = f2fs_unfreeze, .statfs = f2fs_statfs, }; -- cgit From 3786dfdf4f00755ce5797c0e87eac8f898acb52c Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Wed, 30 Jan 2013 22:47:02 +0900 Subject: f2fs: avoid redundant call to has_not_enough_free_secs in f2fs_gc After doing a write_checkpoint from garbage collection path if there is still need to do more garbage collection, gc_more label is used to jump and start the process again. And in that process, first step before getting victim is to check if there are not enough free sections, which is already done before doing a jump to gc_more. We can avoid the redundant call to check free sections, by checking the gc_type flag which will remain FG_GC(value 1) under this condition. Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 9b5d0aad5daf..fb03be68cb20 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -672,7 +672,7 @@ gc_more: if (!(sbi->sb->s_flags & MS_ACTIVE)) goto stop; - if (has_not_enough_free_secs(sbi)) + if (gc_type == BG_GC && has_not_enough_free_secs(sbi)) gc_type = FG_GC; if (!__get_victim(sbi, &segno, gc_type, NO_CHECK_TYPE)) -- cgit From a2b52a598a4df55782462acba3bdced1a4610a60 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Wed, 30 Jan 2013 22:47:16 +0900 Subject: f2fs: reorganize code for ra_node_page We can remove unneeded label unlock_out, avoid unnecessary jump and reorganize the returning conditions in this function. Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/node.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 9bda63c9c166..f71dfbbcb2b0 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -874,15 +874,11 @@ void ra_node_page(struct f2fs_sb_info *sbi, nid_t nid) return; if (read_node_page(apage, READA)) - goto unlock_out; + unlock_page(apage); - page_cache_release(apage); - return; - -unlock_out: - unlock_page(apage); release_out: page_cache_release(apage); + return; } struct page *get_node_page(struct f2fs_sb_info *sbi, pgoff_t nid) -- cgit From 324ddc702eb89fb4ec8ed0cb38b4563ba4804f88 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Wed, 30 Jan 2013 22:47:29 +0900 Subject: f2fs: fix typo mistake for data_version description In f2fs_inode_info structure, the description for data_version has a typo mistake. It should be latest instead of lastes. So, correcting that. Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index f4f509766465..975cb44851ab 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -141,7 +141,7 @@ struct f2fs_inode_info { /* Use below internally in f2fs*/ unsigned long flags; /* use to pass per-file flags */ - unsigned long long data_version;/* lastes version of data for fsync */ + unsigned long long data_version;/* latest version of data for fsync */ atomic_t dirty_dents; /* # of dirty dentry pages */ f2fs_hash_t chash; /* hash value of given file name */ unsigned int clevel; /* maximum level of given file name */ -- cgit From 369a708c2a6557fabd409cf17a03db271c54931a Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Thu, 31 Jan 2013 10:15:35 +0900 Subject: f2fs: remove the use of page_cache_release Let's remove the use of page_cache_release() in f2fs, and instead, use f2fs_put_page(page, 0) which is exactly same but for code readability. Signed-off-by: Jaegeuk Kim --- fs/f2fs/node.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index f71dfbbcb2b0..33fa6d506d94 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -104,7 +104,7 @@ static void ra_nat_pages(struct f2fs_sb_info *sbi, int nid) f2fs_put_page(page, 1); continue; } - page_cache_release(page); + f2fs_put_page(page, 0); } } @@ -877,7 +877,7 @@ void ra_node_page(struct f2fs_sb_info *sbi, nid_t nid) unlock_page(apage); release_out: - page_cache_release(apage); + f2fs_put_page(apage, 0); return; } -- cgit From d4686d56ec912d55fd8a9d6d509de50de24e90ab Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Thu, 31 Jan 2013 15:36:04 +0900 Subject: f2fs: avoid balanc_fs during evict_inode 1. Background Previously, if f2fs tries to move data blocks of an *evicting* inode during the cleaning process, it stops the process incompletely and then restarts the whole process, since it needs a locked inode to grab victim data pages in its address space. In order to get a locked inode, iget_locked() by f2fs_iget() is normally used, but, it waits if the inode is on freeing. So, here is a deadlock scenario. 1. f2fs_evict_inode() <- inode "A" 2. f2fs_balance_fs() 3. f2fs_gc() 4. gc_data_segment() 5. f2fs_iget() <- inode "A" too! If step #1 and #5 treat a same inode "A", step #5 would fall into deadlock since the inode "A" is on freeing. In order to resolve this, f2fs_iget_nowait() which skips __wait_on_freeing_inode() was introduced in step #5, and stops f2fs_gc() to complete f2fs_evict_inode(). 1. f2fs_evict_inode() <- inode "A" 2. f2fs_balance_fs() 3. f2fs_gc() 4. gc_data_segment() 5. f2fs_iget_nowait() <- inode "A", then stop f2fs_gc() w/ -ENOENT 2. Problem and Solution In the above scenario, however, f2fs cannot finish f2fs_evict_inode() only if: o there are not enough free sections, and o f2fs_gc() tries to move data blocks of the *evicting* inode repeatedly. So, the final solution is to use f2fs_iget() and remove f2fs_balance_fs() in f2fs_evict_inode(). The f2fs_evict_inode() actually truncates all the data and node blocks, which means that it doesn't produce any dirty node pages accordingly. So, we don't need to do f2fs_balance_fs() in practical. Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 1 - fs/f2fs/file.c | 3 +-- fs/f2fs/gc.c | 2 +- fs/f2fs/inode.c | 33 --------------------------------- fs/f2fs/recovery.c | 2 +- 5 files changed, 3 insertions(+), 38 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 975cb44851ab..5022a7d7f7ca 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -847,7 +847,6 @@ long f2fs_ioctl(struct file *, unsigned int, unsigned long); * inode.c */ void f2fs_set_inode_flags(struct inode *); -struct inode *f2fs_iget_nowait(struct super_block *, unsigned long); struct inode *f2fs_iget(struct super_block *, unsigned long); void update_inode(struct inode *, struct page *); int f2fs_write_inode(struct inode *, struct writeback_control *); diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 6cdab2c64fc6..3efa4739b00c 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -298,8 +298,6 @@ void f2fs_truncate(struct inode *inode) inode->i_mtime = inode->i_ctime = CURRENT_TIME; mark_inode_dirty(inode); } - - f2fs_balance_fs(F2FS_SB(inode->i_sb)); } static int f2fs_getattr(struct vfsmount *mnt, @@ -356,6 +354,7 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr) attr->ia_size != i_size_read(inode)) { truncate_setsize(inode, attr->ia_size); f2fs_truncate(inode); + f2fs_balance_fs(F2FS_SB(inode->i_sb)); } __setattr_copy(inode, attr); diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index fb03be68cb20..375e69e2c6f1 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -579,7 +579,7 @@ next_step: ofs_in_node = le16_to_cpu(entry->ofs_in_node); if (phase == 2) { - inode = f2fs_iget_nowait(sb, dni.ino); + inode = f2fs_iget(sb, dni.ino); if (IS_ERR(inode)) continue; diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 62433c63e2db..ddae412d30c8 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -16,11 +16,6 @@ #include "f2fs.h" #include "node.h" -struct f2fs_iget_args { - u64 ino; - int on_free; -}; - void f2fs_set_inode_flags(struct inode *inode) { unsigned int flags = F2FS_I(inode)->i_flags; @@ -40,34 +35,6 @@ void f2fs_set_inode_flags(struct inode *inode) inode->i_flags |= S_DIRSYNC; } -static int f2fs_iget_test(struct inode *inode, void *data) -{ - struct f2fs_iget_args *args = data; - - if (inode->i_ino != args->ino) - return 0; - if (inode->i_state & (I_FREEING | I_WILL_FREE)) { - args->on_free = 1; - return 0; - } - return 1; -} - -struct inode *f2fs_iget_nowait(struct super_block *sb, unsigned long ino) -{ - struct f2fs_iget_args args = { - .ino = ino, - .on_free = 0 - }; - struct inode *inode = ilookup5(sb, ino, f2fs_iget_test, &args); - - if (inode) - return inode; - if (!args.on_free) - return f2fs_iget(sb, ino); - return ERR_PTR(-ENOENT); -} - static int do_read_inode(struct inode *inode) { struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb); diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c index f42e4060b399..e2a3e1a8eae9 100644 --- a/fs/f2fs/recovery.c +++ b/fs/f2fs/recovery.c @@ -226,7 +226,7 @@ static void check_index_in_prev_nodes(struct f2fs_sb_info *sbi, f2fs_put_page(node_page, 1); /* Deallocate previous index in the node page */ - inode = f2fs_iget_nowait(sbi->sb, ino); + inode = f2fs_iget(sbi->sb, ino); if (IS_ERR(inode)) return; -- cgit From facb0205401a6ce4be410e1f4b9357bbc6caa36a Mon Sep 17 00:00:00 2001 From: Changman Lee Date: Thu, 31 Jan 2013 17:16:56 +0900 Subject: f2fs: stop repeated checking if cp is needed If it is decided that f2fs should do checkpoint, skip next comparison. Signed-off-by: Changman Lee --- fs/f2fs/file.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 3efa4739b00c..33d1736ee5f9 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -157,11 +157,11 @@ int f2fs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) if (!S_ISREG(inode->i_mode) || inode->i_nlink != 1) need_cp = true; - if (is_inode_flag_set(F2FS_I(inode), FI_NEED_CP)) + else if (is_inode_flag_set(F2FS_I(inode), FI_NEED_CP)) need_cp = true; - if (!space_for_roll_forward(sbi)) + else if (!space_for_roll_forward(sbi)) need_cp = true; - if (need_to_sync_dir(sbi, inode)) + else if (need_to_sync_dir(sbi, inode)) need_cp = true; if (need_cp) { -- cgit From f83759e28372e593879f4dd20eb6c5ba6c4f393a Mon Sep 17 00:00:00 2001 From: majianpeng Date: Fri, 1 Feb 2013 15:00:30 +0800 Subject: f2fs: add device name in debugfs In file status, it can't distinguish between different devices. So add device name to do this function. Signed-off-by: Jianpeng Ma Signed-off-by: Jaegeuk Kim --- fs/f2fs/debug.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c index c8c37307b326..025b9e2f935d 100644 --- a/fs/f2fs/debug.c +++ b/fs/f2fs/debug.c @@ -183,10 +183,12 @@ static int stat_show(struct seq_file *s, void *v) mutex_lock(&f2fs_stat_mutex); list_for_each_entry_safe(si, next, &f2fs_stat_list, stat_list) { + char devname[BDEVNAME_SIZE]; update_general_status(si->sbi); - seq_printf(s, "\n=====[ partition info. #%d ]=====\n", i++); + seq_printf(s, "\n=====[ partition info(%s). #%d ]=====\n", + bdevname(si->sbi->sb->s_bdev, devname), i++); seq_printf(s, "[SB: 1] [CP: 2] [SIT: %d] [NAT: %d] ", si->sit_area_segs, si->nat_area_segs); seq_printf(s, "[SSA: %d] [MAIN: %d", -- cgit From 5c9b469295fb6b10d98923eab5e79c4edb80ed20 Mon Sep 17 00:00:00 2001 From: majianpeng Date: Fri, 1 Feb 2013 19:07:57 +0800 Subject: f2fs: use F2FS_BLKSIZE to judge bloksize and page_cache_size In some system PAGE_CACHE_SIZE isn't 4K. So using F2FS_BLKSIZE to judge. By Jaegeuk Kim: o f2fs does not support no other 4KB page cache size. Signed-off-by: Jianpeng Ma Signed-off-by: Jaegeuk Kim --- fs/f2fs/super.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 0b18aee2ed25..1d7fe11fea30 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -384,14 +384,23 @@ static int sanity_check_raw_super(struct super_block *sb, return 1; } + /* Currently, support only 4KB page cache size */ + if (F2FS_BLKSIZE != PAGE_CACHE_SIZE) { + f2fs_msg(sb, KERN_INFO, + "Invalid page_cache_size (%u), supports only 4KB\n", + PAGE_CACHE_SIZE); + return 1; + } + /* Currently, support only 4KB block size */ blocksize = 1 << le32_to_cpu(raw_super->log_blocksize); - if (blocksize != PAGE_CACHE_SIZE) { + if (blocksize != F2FS_BLKSIZE) { f2fs_msg(sb, KERN_INFO, "Invalid blocksize (%u), supports only 4KB\n", blocksize); return 1; } + if (le32_to_cpu(raw_super->log_sectorsize) != F2FS_LOG_SECTOR_SIZE) { f2fs_msg(sb, KERN_INFO, "Invalid log sectorsize"); -- cgit From 14d7e9de0500e9110e5820e2e77d096fc1bab724 Mon Sep 17 00:00:00 2001 From: majianpeng Date: Fri, 1 Feb 2013 19:07:03 +0800 Subject: f2fs: when check superblock failed, try to check another superblock In f2fs, there are two superblocks. So when the first superblock was invalidate, it should try to check another. By Jaegeuk Kim: o Remove a white space for coding style o Clean up for code readability o Fix a typo Signed-off-by: Jianpeng Ma Signed-off-by: Jaegeuk Kim --- fs/f2fs/super.c | 47 +++++++++++++++++++++++++++++++---------------- 1 file changed, 31 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 1d7fe11fea30..ddb665f54d17 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -387,7 +387,7 @@ static int sanity_check_raw_super(struct super_block *sb, /* Currently, support only 4KB page cache size */ if (F2FS_BLKSIZE != PAGE_CACHE_SIZE) { f2fs_msg(sb, KERN_INFO, - "Invalid page_cache_size (%u), supports only 4KB\n", + "Invalid page_cache_size (%lu), supports only 4KB\n", PAGE_CACHE_SIZE); return 1; } @@ -462,6 +462,32 @@ static void init_sb_info(struct f2fs_sb_info *sbi) atomic_set(&sbi->nr_pages[i], 0); } +static int validate_superblock(struct super_block *sb, + struct f2fs_super_block **raw_super, + struct buffer_head **raw_super_buf, sector_t block) +{ + const char *super = (block == 0 ? "first" : "second"); + + /* read f2fs raw super block */ + *raw_super_buf = sb_bread(sb, block); + if (!*raw_super_buf) { + f2fs_msg(sb, KERN_ERR, "unable to read %s superblock", + super); + return 1; + } + + *raw_super = (struct f2fs_super_block *) + ((char *)(*raw_super_buf)->b_data + F2FS_SUPER_OFFSET); + + /* sanity checking of raw super */ + if (!sanity_check_raw_super(sb, *raw_super)) + return 0; + + f2fs_msg(sb, KERN_ERR, "Can't find a valid F2FS filesystem " + "in %s superblock", super); + return 1; +} + static int f2fs_fill_super(struct super_block *sb, void *data, int silent) { struct f2fs_sb_info *sbi; @@ -482,16 +508,11 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent) goto free_sbi; } - /* read f2fs raw super block */ - raw_super_buf = sb_bread(sb, 0); - if (!raw_super_buf) { - err = -EIO; - f2fs_msg(sb, KERN_ERR, "unable to read superblock"); - goto free_sbi; + if (validate_superblock(sb, &raw_super, &raw_super_buf, 0)) { + brelse(raw_super_buf); + if (validate_superblock(sb, &raw_super, &raw_super_buf, 1)) + goto free_sb_buf; } - raw_super = (struct f2fs_super_block *) - ((char *)raw_super_buf->b_data + F2FS_SUPER_OFFSET); - /* init some FS parameters */ sbi->active_logs = NR_CURSEG_TYPE; @@ -507,12 +528,6 @@ static int f2fs_fill_super(struct super_block *sb, void *data, int silent) if (parse_options(sb, sbi, (char *)data)) goto free_sb_buf; - /* sanity checking of raw super */ - if (sanity_check_raw_super(sb, raw_super)) { - f2fs_msg(sb, KERN_ERR, "Can't find a valid F2FS filesystem"); - goto free_sb_buf; - } - sb->s_maxbytes = max_file_size(le32_to_cpu(raw_super->log_blocksize)); sb->s_max_links = F2FS_LINK_MAX; get_random_bytes(&sbi->s_next_generation, sizeof(u32)); -- cgit From 94787d91cba7ba168b028703b50a0224702ace9c Mon Sep 17 00:00:00 2001 From: Changman Lee Date: Fri, 1 Feb 2013 18:05:09 +0900 Subject: f2fs: remove repeated F2FS_SET_SB_DIRT call F2FS_SET_SB_DIRT is called in inc_page_count and it is directly called one more time in the next line. Signed-off-by: Changman Lee Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index d3b34d05211f..2887c196b0a2 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -164,7 +164,6 @@ static int f2fs_set_meta_page_dirty(struct page *page) if (!PageDirty(page)) { __set_page_dirty_nobuffers(page); inc_page_count(sbi, F2FS_DIRTY_META); - F2FS_SET_SB_DIRT(sbi); return 1; } return 0; -- cgit From 48600e44c18e4eb0f7c02ec8633c4c56aef292f0 Mon Sep 17 00:00:00 2001 From: Changman Lee Date: Mon, 4 Feb 2013 10:05:09 +0900 Subject: f2fs: remove unnecessary gc option check and balance_fs 1. If f2fs is mounted with background_gc_off option, checking BG_GC is not redundant. 2. f2fs_balance_fs is checked in f2fs_gc, so this is also redundant. Signed-off-by: Changman Lee Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 375e69e2c6f1..8d293cb685ba 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -49,11 +49,6 @@ static int gc_thread_func(void *data) continue; } - f2fs_balance_fs(sbi); - - if (!test_opt(sbi, BG_GC)) - continue; - /* * [GC triggering condition] * 0. GC is not conducted currently. @@ -96,6 +91,8 @@ int start_gc_thread(struct f2fs_sb_info *sbi) { struct f2fs_gc_kthread *gc_th; + if (!test_opt(sbi, BG_GC)) + return 0; gc_th = kmalloc(sizeof(struct f2fs_gc_kthread), GFP_KERNEL); if (!gc_th) return -ENOMEM; -- cgit From ec7b1f2dd180afb1ce23aa44a7451e59bf2f3eb3 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sat, 2 Feb 2013 23:52:28 +0900 Subject: f2fs: name gc task as per the block device Currently GC task is started for each f2fs formatted/mounted device. But, when we check the task list, using 'ps', there is no distinguishing factor between the tasks. So, name the task as per the block device just like the flusher threads. Also, remove the macro GC_THREAD_NAME and instead use the name: f2fs_gc to avoid name length truncation, as the command length is 16 -> TASK_COMM_LEN 16 and example name like: f2fs_gc_task:8:16 -> this exceeds name length Before Patch for 2 F2FS formatted partitions: root 28061 0.0 0.0 0 0 ? S 10:31 0:00 [f2fs_gc_task] root 28087 0.0 0.0 0 0 ? S 10:32 0:00 [f2fs_gc_task] After Patch: root 16756 0.0 0.0 0 0 ? S 14:57 0:00 [f2fs_gc-8:18] root 16765 0.0 0.0 0 0 ? S 14:57 0:00 [f2fs_gc-8:19] Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 3 ++- fs/f2fs/gc.h | 1 - 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 8d293cb685ba..b1e498503e59 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -90,6 +90,7 @@ static int gc_thread_func(void *data) int start_gc_thread(struct f2fs_sb_info *sbi) { struct f2fs_gc_kthread *gc_th; + dev_t dev = sbi->sb->s_bdev->bd_dev; if (!test_opt(sbi, BG_GC)) return 0; @@ -100,7 +101,7 @@ int start_gc_thread(struct f2fs_sb_info *sbi) sbi->gc_thread = gc_th; init_waitqueue_head(&sbi->gc_thread->gc_wait_queue_head); sbi->gc_thread->f2fs_gc_task = kthread_run(gc_thread_func, sbi, - GC_THREAD_NAME); + "f2fs_gc-%u:%u", MAJOR(dev), MINOR(dev)); if (IS_ERR(gc_th->f2fs_gc_task)) { kfree(gc_th); return -ENOMEM; diff --git a/fs/f2fs/gc.h b/fs/f2fs/gc.h index b026d9354ccd..3abdf83b5c16 100644 --- a/fs/f2fs/gc.h +++ b/fs/f2fs/gc.h @@ -8,7 +8,6 @@ * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ -#define GC_THREAD_NAME "f2fs_gc_task" #define GC_THREAD_MIN_WB_PAGES 1 /* * a threshold to determine * whether IO subsystem is idle -- cgit From 25718423ea6f7118f9c173adff0fac52414571b8 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sat, 2 Feb 2013 23:52:42 +0900 Subject: f2fs: mark gc_thread as NULL when thread creation is failed When gc thread creation is failed, mark gc_thread as NULL to avoid crash while trying to stop invalid thread in stop_gc_thread->kthread_stop. Instead make it return from: if (!gc_th) return; Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index b1e498503e59..16fdec355201 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -104,6 +104,7 @@ int start_gc_thread(struct f2fs_sb_info *sbi) "f2fs_gc-%u:%u", MAJOR(dev), MINOR(dev)); if (IS_ERR(gc_th->f2fs_gc_task)) { kfree(gc_th); + sbi->gc_thread = NULL; return -ENOMEM; } return 0; -- cgit From 5ac206cf4f9aad871aa1f1dd653a3404092db56c Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sat, 2 Feb 2013 23:52:59 +0900 Subject: f2fs: make an accessor to get sections for particular block type Introduce accessor to get the sections based upon the block type (node,dents...) and modify the functions : should_do_checkpoint, has_not_enough_free_secs to use this accessor function to get the node sections and dent sections. Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 8 ++++++++ fs/f2fs/gc.h | 8 ++------ fs/f2fs/segment.h | 9 ++------- 3 files changed, 12 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 5022a7d7f7ca..87840bccaf21 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -573,6 +573,14 @@ static inline int get_pages(struct f2fs_sb_info *sbi, int count_type) return atomic_read(&sbi->nr_pages[count_type]); } +static inline int get_blocktype_secs(struct f2fs_sb_info *sbi, int block_type) +{ + unsigned int pages_per_sec = sbi->segs_per_sec * + (1 << sbi->log_blocks_per_seg); + return ((get_pages(sbi, block_type) + pages_per_sec - 1) + >> sbi->log_blocks_per_seg) / sbi->segs_per_sec; +} + static inline block_t valid_user_blocks(struct f2fs_sb_info *sbi) { block_t ret; diff --git a/fs/f2fs/gc.h b/fs/f2fs/gc.h index 3abdf83b5c16..c407a75a7daa 100644 --- a/fs/f2fs/gc.h +++ b/fs/f2fs/gc.h @@ -106,11 +106,7 @@ static inline int is_idle(struct f2fs_sb_info *sbi) static inline bool should_do_checkpoint(struct f2fs_sb_info *sbi) { - unsigned int pages_per_sec = sbi->segs_per_sec * - (1 << sbi->log_blocks_per_seg); - int node_secs = ((get_pages(sbi, F2FS_DIRTY_NODES) + pages_per_sec - 1) - >> sbi->log_blocks_per_seg) / sbi->segs_per_sec; - int dent_secs = ((get_pages(sbi, F2FS_DIRTY_DENTS) + pages_per_sec - 1) - >> sbi->log_blocks_per_seg) / sbi->segs_per_sec; + int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES); + int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS); return free_sections(sbi) <= (node_secs + 2 * dent_secs + 2); } diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index 66a288a52fd3..a9417b94d6bc 100644 --- a/fs/f2fs/segment.h +++ b/fs/f2fs/segment.h @@ -459,13 +459,8 @@ static inline int get_ssr_segment(struct f2fs_sb_info *sbi, int type) static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi) { - unsigned int pages_per_sec = (1 << sbi->log_blocks_per_seg) * - sbi->segs_per_sec; - int node_secs = ((get_pages(sbi, F2FS_DIRTY_NODES) + pages_per_sec - 1) - >> sbi->log_blocks_per_seg) / sbi->segs_per_sec; - int dent_secs = ((get_pages(sbi, F2FS_DIRTY_DENTS) + pages_per_sec - 1) - >> sbi->log_blocks_per_seg) / sbi->segs_per_sec; - + int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES); + int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS); if (sbi->por_doing) return false; -- cgit From b1f1daf8c72d615b64163e26488d8effeed29b60 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sat, 2 Feb 2013 23:53:15 +0900 Subject: f2fs: optimize the return condition for has_not_enough_free_secs Instead of evaluating the free_sections and then deciding to return true/false from that path. We can directly use the evaluation condition for returning proper value. Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/segment.h | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index a9417b94d6bc..458bf5c726f7 100644 --- a/fs/f2fs/segment.h +++ b/fs/f2fs/segment.h @@ -464,10 +464,8 @@ static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi) if (sbi->por_doing) return false; - if (free_sections(sbi) <= (node_secs + 2 * dent_secs + - reserved_sections(sbi))) - return true; - return false; + return (free_sections(sbi) <= (node_secs + 2 * dent_secs + + reserved_sections(sbi))); } static inline int utilization(struct f2fs_sb_info *sbi) -- cgit From 437275272f9e635673f065300e5d95226a25cb06 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Mon, 4 Feb 2013 15:11:17 +0900 Subject: f2fs: clarify and enhance the f2fs_gc flow This patch makes clearer the ambiguous f2fs_gc flow as follows. 1. Remove intermediate checkpoint condition during f2fs_gc (i.e., should_do_checkpoint() and GC_BLOCKED) 2. Remove unnecessary return values of f2fs_gc because of #1. (i.e., GC_NODE, GC_OK, etc) 3. Simplify write_checkpoint() because of #2. 4. Clarify the main f2fs_gc flow. o monitor how many freed sections during one iteration of do_garbage_collect(). o do GC more without checkpoints if we can't get enough free sections. o do checkpoint once we've got enough free sections through forground GCs. 5. Adopt thread-logging (Slack-Space-Recycle) scheme more aggressively on data log types. See. get_ssr_segement() Signed-off-by: Jaegeuk Kim --- fs/f2fs/checkpoint.c | 10 ++--- fs/f2fs/f2fs.h | 3 +- fs/f2fs/gc.c | 107 ++++++++++++++++++++------------------------------- fs/f2fs/gc.h | 16 -------- fs/f2fs/node.c | 2 +- fs/f2fs/recovery.c | 2 +- fs/f2fs/segment.c | 21 +++++++++- fs/f2fs/segment.h | 12 ++---- fs/f2fs/super.c | 4 +- 9 files changed, 72 insertions(+), 105 deletions(-) (limited to 'fs') diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index 2887c196b0a2..2b6fc131e2ce 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -539,7 +539,7 @@ retry: /* * Freeze all the FS-operations for checkpoint. */ -void block_operations(struct f2fs_sb_info *sbi) +static void block_operations(struct f2fs_sb_info *sbi) { int t; struct writeback_control wbc = { @@ -722,15 +722,13 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, bool is_umount) /* * We guarantee that this checkpoint procedure should not fail. */ -void write_checkpoint(struct f2fs_sb_info *sbi, bool blocked, bool is_umount) +void write_checkpoint(struct f2fs_sb_info *sbi, bool is_umount) { struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi); unsigned long long ckpt_ver; - if (!blocked) { - mutex_lock(&sbi->cp_mutex); - block_operations(sbi); - } + mutex_lock(&sbi->cp_mutex); + block_operations(sbi); f2fs_submit_bio(sbi, DATA, true); f2fs_submit_bio(sbi, NODE, true); diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 87840bccaf21..e7e7a29767d6 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -969,8 +969,7 @@ int get_valid_checkpoint(struct f2fs_sb_info *); void set_dirty_dir_page(struct inode *, struct page *); void remove_dirty_dir_inode(struct inode *); void sync_dirty_dir_inodes(struct f2fs_sb_info *); -void block_operations(struct f2fs_sb_info *); -void write_checkpoint(struct f2fs_sb_info *, bool, bool); +void write_checkpoint(struct f2fs_sb_info *, bool); void init_orphan_info(struct f2fs_sb_info *); int __init create_checkpoint_caches(void); void destroy_checkpoint_caches(void); diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 16fdec355201..52d3a391b922 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -78,7 +78,8 @@ static int gc_thread_func(void *data) sbi->bg_gc++; - if (f2fs_gc(sbi) == GC_NONE) + /* if return value is not zero, no victim was selected */ + if (f2fs_gc(sbi)) wait_ms = GC_THREAD_NOGC_SLEEP_TIME; else if (wait_ms == GC_THREAD_NOGC_SLEEP_TIME) wait_ms = GC_THREAD_MAX_SLEEP_TIME; @@ -360,7 +361,7 @@ static int check_valid_map(struct f2fs_sb_info *sbi, sentry = get_seg_entry(sbi, segno); ret = f2fs_test_bit(offset, sentry->cur_valid_map); mutex_unlock(&sit_i->sentry_lock); - return ret ? GC_OK : GC_NEXT; + return ret; } /* @@ -368,7 +369,7 @@ static int check_valid_map(struct f2fs_sb_info *sbi, * On validity, copy that node with cold status, otherwise (invalid node) * ignore that. */ -static int gc_node_segment(struct f2fs_sb_info *sbi, +static void gc_node_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, unsigned int segno, int gc_type) { bool initial = true; @@ -380,21 +381,12 @@ next_step: for (off = 0; off < sbi->blocks_per_seg; off++, entry++) { nid_t nid = le32_to_cpu(entry->nid); struct page *node_page; - int err; - /* - * It makes sure that free segments are able to write - * all the dirty node pages before CP after this CP. - * So let's check the space of dirty node pages. - */ - if (should_do_checkpoint(sbi)) { - mutex_lock(&sbi->cp_mutex); - block_operations(sbi); - return GC_BLOCKED; - } + /* stop BG_GC if there is not enough free sections. */ + if (gc_type == BG_GC && has_not_enough_free_secs(sbi, 0)) + return; - err = check_valid_map(sbi, segno, off); - if (err == GC_NEXT) + if (check_valid_map(sbi, segno, off) == 0) continue; if (initial) { @@ -424,7 +416,6 @@ next_step: }; sync_node_pages(sbi, 0, &wbc); } - return GC_DONE; } /* @@ -467,13 +458,13 @@ static int check_dnode(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, node_page = get_node_page(sbi, nid); if (IS_ERR(node_page)) - return GC_NEXT; + return 0; get_node_info(sbi, nid, dni); if (sum->version != dni->version) { f2fs_put_page(node_page, 1); - return GC_NEXT; + return 0; } *nofs = ofs_of_node(node_page); @@ -481,8 +472,8 @@ static int check_dnode(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, f2fs_put_page(node_page, 1); if (source_blkaddr != blkaddr) - return GC_NEXT; - return GC_OK; + return 0; + return 1; } static void move_data_page(struct inode *inode, struct page *page, int gc_type) @@ -523,13 +514,13 @@ out: * If the parent node is not valid or the data block address is different, * the victim data block is ignored. */ -static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, +static void gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, struct list_head *ilist, unsigned int segno, int gc_type) { struct super_block *sb = sbi->sb; struct f2fs_summary *entry; block_t start_addr; - int err, off; + int off; int phase = 0; start_addr = START_BLOCK(sbi, segno); @@ -543,20 +534,11 @@ next_step: unsigned int ofs_in_node, nofs; block_t start_bidx; - /* - * It makes sure that free segments are able to write - * all the dirty node pages before CP after this CP. - * So let's check the space of dirty node pages. - */ - if (should_do_checkpoint(sbi)) { - mutex_lock(&sbi->cp_mutex); - block_operations(sbi); - err = GC_BLOCKED; - goto stop; - } + /* stop BG_GC if there is not enough free sections. */ + if (gc_type == BG_GC && has_not_enough_free_secs(sbi, 0)) + return; - err = check_valid_map(sbi, segno, off); - if (err == GC_NEXT) + if (check_valid_map(sbi, segno, off) == 0) continue; if (phase == 0) { @@ -565,8 +547,7 @@ next_step: } /* Get an inode by ino with checking validity */ - err = check_dnode(sbi, entry, &dni, start_addr + off, &nofs); - if (err == GC_NEXT) + if (check_dnode(sbi, entry, &dni, start_addr + off, &nofs) == 0) continue; if (phase == 1) { @@ -606,11 +587,9 @@ next_iput: } if (++phase < 4) goto next_step; - err = GC_DONE; -stop: + if (gc_type == FG_GC) f2fs_submit_bio(sbi, DATA, true); - return err; } static int __get_victim(struct f2fs_sb_info *sbi, unsigned int *victim, @@ -624,17 +603,16 @@ static int __get_victim(struct f2fs_sb_info *sbi, unsigned int *victim, return ret; } -static int do_garbage_collect(struct f2fs_sb_info *sbi, unsigned int segno, +static void do_garbage_collect(struct f2fs_sb_info *sbi, unsigned int segno, struct list_head *ilist, int gc_type) { struct page *sum_page; struct f2fs_summary_block *sum; - int ret = GC_DONE; /* read segment summary of victim */ sum_page = get_sum_page(sbi, segno); if (IS_ERR(sum_page)) - return GC_ERROR; + return; /* * CP needs to lock sum_page. In this time, we don't need @@ -646,17 +624,16 @@ static int do_garbage_collect(struct f2fs_sb_info *sbi, unsigned int segno, switch (GET_SUM_TYPE((&sum->footer))) { case SUM_TYPE_NODE: - ret = gc_node_segment(sbi, sum->entries, segno, gc_type); + gc_node_segment(sbi, sum->entries, segno, gc_type); break; case SUM_TYPE_DATA: - ret = gc_data_segment(sbi, sum->entries, ilist, segno, gc_type); + gc_data_segment(sbi, sum->entries, ilist, segno, gc_type); break; } stat_inc_seg_count(sbi, GET_SUM_TYPE((&sum->footer))); stat_inc_call_count(sbi->stat_info); f2fs_put_page(sum_page, 0); - return ret; } int f2fs_gc(struct f2fs_sb_info *sbi) @@ -664,40 +641,38 @@ int f2fs_gc(struct f2fs_sb_info *sbi) struct list_head ilist; unsigned int segno, i; int gc_type = BG_GC; - int gc_status = GC_NONE; + int nfree = 0; + int ret = -1; INIT_LIST_HEAD(&ilist); gc_more: if (!(sbi->sb->s_flags & MS_ACTIVE)) goto stop; - if (gc_type == BG_GC && has_not_enough_free_secs(sbi)) + if (gc_type == BG_GC && has_not_enough_free_secs(sbi, nfree)) gc_type = FG_GC; if (!__get_victim(sbi, &segno, gc_type, NO_CHECK_TYPE)) goto stop; + ret = 0; - for (i = 0; i < sbi->segs_per_sec; i++) { - /* - * do_garbage_collect will give us three gc_status: - * GC_ERROR, GC_DONE, and GC_BLOCKED. - * If GC is finished uncleanly, we have to return - * the victim to dirty segment list. - */ - gc_status = do_garbage_collect(sbi, segno + i, &ilist, gc_type); - if (gc_status != GC_DONE) - break; - } - if (has_not_enough_free_secs(sbi)) { - write_checkpoint(sbi, (gc_status == GC_BLOCKED), false); - if (has_not_enough_free_secs(sbi)) - goto gc_more; - } + for (i = 0; i < sbi->segs_per_sec; i++) + do_garbage_collect(sbi, segno + i, &ilist, gc_type); + + if (gc_type == FG_GC && + get_valid_blocks(sbi, segno, sbi->segs_per_sec) == 0) + nfree++; + + if (has_not_enough_free_secs(sbi, nfree)) + goto gc_more; + + if (gc_type == FG_GC) + write_checkpoint(sbi, false); stop: mutex_unlock(&sbi->gc_mutex); put_gc_inode(&ilist); - return gc_status; + return ret; } void build_gc_manager(struct f2fs_sb_info *sbi) diff --git a/fs/f2fs/gc.h b/fs/f2fs/gc.h index c407a75a7daa..30b2db003acd 100644 --- a/fs/f2fs/gc.h +++ b/fs/f2fs/gc.h @@ -22,15 +22,6 @@ /* Search max. number of dirty segments to select a victim segment */ #define MAX_VICTIM_SEARCH 20 -enum { - GC_NONE = 0, - GC_ERROR, - GC_OK, - GC_NEXT, - GC_BLOCKED, - GC_DONE, -}; - struct f2fs_gc_kthread { struct task_struct *f2fs_gc_task; wait_queue_head_t gc_wait_queue_head; @@ -103,10 +94,3 @@ static inline int is_idle(struct f2fs_sb_info *sbi) struct request_list *rl = &q->root_rl; return !(rl->count[BLK_RW_SYNC]) && !(rl->count[BLK_RW_ASYNC]); } - -static inline bool should_do_checkpoint(struct f2fs_sb_info *sbi) -{ - int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES); - int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS); - return free_sections(sbi) <= (node_secs + 2 * dent_secs + 2); -} diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 33fa6d506d94..43ce16422b75 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -1135,7 +1135,7 @@ static int f2fs_write_node_pages(struct address_space *mapping, /* First check balancing cached NAT entries */ if (try_to_free_nats(sbi, NAT_ENTRY_PER_BLOCK)) { - write_checkpoint(sbi, false, false); + write_checkpoint(sbi, false); return 0; } diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c index e2a3e1a8eae9..01e1a03b54c8 100644 --- a/fs/f2fs/recovery.c +++ b/fs/f2fs/recovery.c @@ -373,5 +373,5 @@ void recover_fsync_data(struct f2fs_sb_info *sbi) out: destroy_fsync_dnodes(sbi, &inode_list); kmem_cache_destroy(fsync_entry_slab); - write_checkpoint(sbi, false, false); + write_checkpoint(sbi, false); } diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 7aa270f3538a..777f17e496e6 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -29,7 +29,7 @@ void f2fs_balance_fs(struct f2fs_sb_info *sbi) * We should do GC or end up with checkpoint, if there are so many dirty * dir/node pages without enough free segments. */ - if (has_not_enough_free_secs(sbi)) { + if (has_not_enough_free_secs(sbi, 0)) { mutex_lock(&sbi->gc_mutex); f2fs_gc(sbi); } @@ -308,7 +308,7 @@ static unsigned int check_prefree_segments(struct f2fs_sb_info *sbi, * If there is not enough reserved sections, * we should not reuse prefree segments. */ - if (has_not_enough_free_secs(sbi)) + if (has_not_enough_free_secs(sbi, 0)) return NULL_SEGNO; /* @@ -536,6 +536,23 @@ static void change_curseg(struct f2fs_sb_info *sbi, int type, bool reuse) } } +static int get_ssr_segment(struct f2fs_sb_info *sbi, int type) +{ + struct curseg_info *curseg = CURSEG_I(sbi, type); + const struct victim_selection *v_ops = DIRTY_I(sbi)->v_ops; + + if (IS_NODESEG(type) || !has_not_enough_free_secs(sbi, 0)) + return v_ops->get_victim(sbi, + &(curseg)->next_segno, BG_GC, type, SSR); + + /* For data segments, let's do SSR more intensively */ + for (; type >= CURSEG_HOT_DATA; type--) + if (v_ops->get_victim(sbi, &(curseg)->next_segno, + BG_GC, type, SSR)) + return 1; + return 0; +} + /* * flush out current segment and replace it with new segment * This function should be returned with success, otherwise BUG diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index 458bf5c726f7..552dadbb2327 100644 --- a/fs/f2fs/segment.h +++ b/fs/f2fs/segment.h @@ -450,21 +450,15 @@ static inline bool need_SSR(struct f2fs_sb_info *sbi) return (free_sections(sbi) < overprovision_sections(sbi)); } -static inline int get_ssr_segment(struct f2fs_sb_info *sbi, int type) -{ - struct curseg_info *curseg = CURSEG_I(sbi, type); - return DIRTY_I(sbi)->v_ops->get_victim(sbi, - &(curseg)->next_segno, BG_GC, type, SSR); -} - -static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi) +static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi, int freed) { int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES); int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS); + if (sbi->por_doing) return false; - return (free_sections(sbi) <= (node_secs + 2 * dent_secs + + return ((free_sections(sbi) + freed) <= (node_secs + 2 * dent_secs + reserved_sections(sbi))); } diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index ddb665f54d17..8c117649a035 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -112,7 +112,7 @@ static void f2fs_put_super(struct super_block *sb) f2fs_destroy_stats(sbi); stop_gc_thread(sbi); - write_checkpoint(sbi, false, true); + write_checkpoint(sbi, true); iput(sbi->node_inode); iput(sbi->meta_inode); @@ -136,7 +136,7 @@ int f2fs_sync_fs(struct super_block *sb, int sync) return 0; if (sync) - write_checkpoint(sbi, false, false); + write_checkpoint(sbi, false); else f2fs_balance_fs(sbi); -- cgit From b7250d2d845822466356f7f22a650bf807090d7e Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Tue, 5 Feb 2013 13:19:28 +0900 Subject: f2fs: fix calculation of max. gc cost in the SSR case In the SSR case, the max gc cost should be the number of pages in a segment. Otherwise, f2fs is able to fail getting dirty segments frequently for SSR. In get_victim_by_default() previously, while(1) { ... cost = get_gc_cost(); <- cost is between 0 ~ 512. ... if (cost == get_max_cost(sbi, &p)) <- max cost is UINT_MAX due to GC_CB type continue; if (nsearched++ >= MAX_VICTIM_SEARCH) break; } So, if there are a number of fully valid segments in series, f2fs cannot skip those segments by comparing the cost and max cost of each segment. Note that, the cost is the number of valid blocks at the time of the last checkpoint. Signed-off-by: Jaegeuk Kim --- fs/f2fs/gc.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 52d3a391b922..94b8a0c48453 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -146,6 +146,9 @@ static void select_policy(struct f2fs_sb_info *sbi, int gc_type, static unsigned int get_max_cost(struct f2fs_sb_info *sbi, struct victim_sel_policy *p) { + /* SSR allocates in a segment unit */ + if (p->alloc_mode == SSR) + return 1 << sbi->log_blocks_per_seg; if (p->gc_mode == GC_GREEDY) return (1 << sbi->log_blocks_per_seg) * p->ofs_unit; else if (p->gc_mode == GC_CB) -- cgit From e9750824114ff939d9da299e73651add6aa65456 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Mon, 4 Feb 2013 23:41:41 +0900 Subject: f2fs: add compat_ioctl to provide backward compatability adding compat_ioctl to provide support for backward comptability - 32bit binary execution on 64bit kernel. Signed-off-by: Namjae Jeon Signed-off-by: Amit Sahrawat Signed-off-by: Jaegeuk Kim --- fs/f2fs/f2fs.h | 15 +++++++++++++++ fs/f2fs/file.c | 21 +++++++++++++++++++++ 2 files changed, 36 insertions(+) (limited to 'fs') diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index e7e7a29767d6..c3462b69917e 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -103,6 +103,20 @@ static inline int update_sits_in_cursum(struct f2fs_summary_block *rs, int i) return before; } +/* + * ioctl commands + */ +#define F2FS_IOC_GETFLAGS FS_IOC_GETFLAGS +#define F2FS_IOC_SETFLAGS FS_IOC_SETFLAGS + +#if defined(__KERNEL__) && defined(CONFIG_COMPAT) +/* + * ioctl commands in 32 bit emulation + */ +#define F2FS_IOC32_GETFLAGS FS_IOC32_GETFLAGS +#define F2FS_IOC32_SETFLAGS FS_IOC32_SETFLAGS +#endif + /* * For INODE and NODE manager */ @@ -850,6 +864,7 @@ void f2fs_truncate(struct inode *); int f2fs_setattr(struct dentry *, struct iattr *); int truncate_hole(struct inode *, pgoff_t, pgoff_t); long f2fs_ioctl(struct file *, unsigned int, unsigned long); +long f2fs_compat_ioctl(struct file *, unsigned int, unsigned long); /* * inode.c diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 33d1736ee5f9..b7a053d4c6d3 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include @@ -634,6 +635,23 @@ out: } } +#ifdef CONFIG_COMPAT +long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + switch (cmd) { + case F2FS_IOC32_GETFLAGS: + cmd = F2FS_IOC_GETFLAGS; + break; + case F2FS_IOC32_SETFLAGS: + cmd = F2FS_IOC_SETFLAGS; + break; + default: + return -ENOIOCTLCMD; + } + return f2fs_ioctl(file, cmd, (unsigned long) compat_ptr(arg)); +} +#endif + const struct file_operations f2fs_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, @@ -645,6 +663,9 @@ const struct file_operations f2fs_file_operations = { .fsync = f2fs_sync_file, .fallocate = f2fs_fallocate, .unlocked_ioctl = f2fs_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = f2fs_compat_ioctl, +#endif .splice_read = generic_file_splice_read, .splice_write = generic_file_splice_write, }; -- cgit From 7dd690c82029ed34aafdb58ce7463cdead69abb5 Mon Sep 17 00:00:00 2001 From: Jaegeuk Kim Date: Tue, 12 Feb 2013 07:28:55 +0900 Subject: f2fs: avoid build warning This patch removes the following build warning: fs/f2fs/node.c: warning: 'nofs' may be used uninitialized in this function [-Wuninitialized]: => 738:8 Note that this is a false alarm. Signed-off-by: Jaegeuk Kim --- fs/f2fs/node.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 0da252c78af8..e275218904ed 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -660,7 +660,7 @@ int truncate_inode_blocks(struct inode *inode, pgoff_t from) struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb); int err = 0, cont = 1; int level, offset[4], noffset[4]; - unsigned int nofs; + unsigned int nofs = 0; struct f2fs_node *rn; struct dnode_of_data dn; struct page *page; -- cgit From c8da19b9866ea84e9ad1c369393ea95d54ee7845 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 11 Feb 2013 19:01:21 -0500 Subject: NFSv4.1: Fix an ABBA locking issue with session and state serialisation Ensure that if nfs_wait_on_sequence() causes our rpc task to wait for an NFSv4 state serialisation lock, then we also drop the session slot. Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/nfs/nfs4proc.c | 32 ++++++++++++++++++++------------ 1 file changed, 20 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 7cbf7aa6c634..cf252dfd91f8 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1466,7 +1466,7 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata) struct nfs4_state_owner *sp = data->owner; if (nfs_wait_on_sequence(data->o_arg.seqid, task) != 0) - return; + goto out_wait; /* * Check if we still need to send an OPEN call, or if we can use * a delegation instead. @@ -1501,6 +1501,7 @@ unlock_no_action: rcu_read_unlock(); out_no_action: task->tk_action = NULL; +out_wait: nfs4_sequence_done(task, &data->o_res.seq_res); } @@ -2179,7 +2180,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) dprintk("%s: begin!\n", __func__); if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0) - return; + goto out_wait; task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_DOWNGRADE]; calldata->arg.fmode = FMODE_READ|FMODE_WRITE; @@ -2201,16 +2202,14 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) if (!call_close) { /* Note: exit _without_ calling nfs4_close_done */ - task->tk_action = NULL; - nfs4_sequence_done(task, &calldata->res.seq_res); - goto out; + goto out_no_action; } if (calldata->arg.fmode == 0) { task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CLOSE]; if (calldata->roc && pnfs_roc_drain(inode, &calldata->roc_barrier, task)) - goto out; + goto out_wait; } nfs_fattr_init(calldata->res.fattr); @@ -2220,8 +2219,12 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) &calldata->res.seq_res, task) != 0) nfs_release_seqid(calldata->arg.seqid); -out: dprintk("%s: done!\n", __func__); + return; +out_no_action: + task->tk_action = NULL; +out_wait: + nfs4_sequence_done(task, &calldata->res.seq_res); } static const struct rpc_call_ops nfs4_close_ops = { @@ -4452,12 +4455,10 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data) struct nfs4_unlockdata *calldata = data; if (nfs_wait_on_sequence(calldata->arg.seqid, task) != 0) - return; + goto out_wait; if (test_bit(NFS_LOCK_INITIALIZED, &calldata->lsp->ls_flags) == 0) { /* Note: exit _without_ running nfs4_locku_done */ - task->tk_action = NULL; - nfs4_sequence_done(task, &calldata->res.seq_res); - return; + goto out_no_action; } calldata->timestamp = jiffies; if (nfs4_setup_sequence(calldata->server, @@ -4465,6 +4466,11 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data) &calldata->res.seq_res, task) != 0) nfs_release_seqid(calldata->arg.seqid); + return; +out_no_action: + task->tk_action = NULL; +out_wait: + nfs4_sequence_done(task, &calldata->res.seq_res); } static const struct rpc_call_ops nfs4_locku_ops = { @@ -4612,7 +4618,7 @@ static void nfs4_lock_prepare(struct rpc_task *task, void *calldata) dprintk("%s: begin!\n", __func__); if (nfs_wait_on_sequence(data->arg.lock_seqid, task) != 0) - return; + goto out_wait; /* Do we need to do an open_to_lock_owner? */ if (!(data->arg.lock_seqid->sequence->flags & NFS_SEQID_CONFIRMED)) { if (nfs_wait_on_sequence(data->arg.open_seqid, task) != 0) { @@ -4632,6 +4638,8 @@ static void nfs4_lock_prepare(struct rpc_task *task, void *calldata) nfs_release_seqid(data->arg.open_seqid); out_release_lock_seqid: nfs_release_seqid(data->arg.lock_seqid); +out_wait: + nfs4_sequence_done(task, &data->res.seq_res); dprintk("%s: done!, ret = %d\n", __func__, data->rpc_status); } -- cgit From 05cb11c17e892f0e131b6c2ba25d63221aafbd11 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 02:56:19 -0800 Subject: ceph: Translate between uid and gids in cap messages and kuids and kgids - Make the uid and gid arguments of send_cap_msg() used to compose ceph_mds_caps messages of type kuid_t and kgid_t. - Pass inode->i_uid and inode->i_gid in __send_cap to send_cap_msg() through variables of type kuid_t and kgid_t. - Modify struct ceph_cap_snap to store uids and gids in types kuid_t and kgid_t. This allows capturing inode->i_uid and inode->i_gid in ceph_queue_cap_snap() without loss and pssing them to __ceph_flush_snaps() where they are removed from struct ceph_cap_snap and passed to send_cap_msg(). - In handle_cap_grant translate uid and gids in the initial user namespace stored in struct ceph_mds_cap into kuids and kgids before setting inode->i_uid and inode->i_gid. Cc: Sage Weil Signed-off-by: "Eric W. Biederman" --- fs/ceph/caps.c | 14 +++++++------- fs/ceph/super.h | 4 ++-- 2 files changed, 9 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index a1d9bb30c1bf..39eb4665b122 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -930,7 +930,7 @@ static int send_cap_msg(struct ceph_mds_session *session, u64 size, u64 max_size, struct timespec *mtime, struct timespec *atime, u64 time_warp_seq, - uid_t uid, gid_t gid, umode_t mode, + kuid_t uid, kgid_t gid, umode_t mode, u64 xattr_version, struct ceph_buffer *xattrs_buf, u64 follows) @@ -974,8 +974,8 @@ static int send_cap_msg(struct ceph_mds_session *session, ceph_encode_timespec(&fc->atime, atime); fc->time_warp_seq = cpu_to_le32(time_warp_seq); - fc->uid = cpu_to_le32(uid); - fc->gid = cpu_to_le32(gid); + fc->uid = cpu_to_le32(from_kuid(&init_user_ns, uid)); + fc->gid = cpu_to_le32(from_kgid(&init_user_ns, gid)); fc->mode = cpu_to_le32(mode); fc->xattr_version = cpu_to_le64(xattr_version); @@ -1081,8 +1081,8 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, struct timespec mtime, atime; int wake = 0; umode_t mode; - uid_t uid; - gid_t gid; + kuid_t uid; + kgid_t gid; struct ceph_mds_session *session; u64 xattr_version = 0; struct ceph_buffer *xattr_blob = NULL; @@ -2359,8 +2359,8 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant, if ((issued & CEPH_CAP_AUTH_EXCL) == 0) { inode->i_mode = le32_to_cpu(grant->mode); - inode->i_uid = le32_to_cpu(grant->uid); - inode->i_gid = le32_to_cpu(grant->gid); + inode->i_uid = make_kuid(&init_user_ns, le32_to_cpu(grant->uid)); + inode->i_gid = make_kgid(&init_user_ns, le32_to_cpu(grant->gid)); dout("%p mode 0%o uid.gid %d.%d\n", inode, inode->i_mode, inode->i_uid, inode->i_gid); } diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 66ebe720e40d..f053bbd1886f 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -138,8 +138,8 @@ struct ceph_cap_snap { struct ceph_snap_context *context; umode_t mode; - uid_t uid; - gid_t gid; + kuid_t uid; + kgid_t gid; struct ceph_buffer *xattr_blob; u64 xattr_version; -- cgit From ab871b903e9095772c219b512d9eae96c4663a5d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 03:40:12 -0800 Subject: ceph: Translate inode uid and gid attributes to/from kuids and kgids. - In fill_inode() transate uids and gids in the initial user namespace into kuids and kgids stored in inode->i_uid and inode->i_gid. - In ceph_setattr() if they have changed convert inode->i_uid and inode->i_gid into initial user namespace uids and gids for transmission. Cc: Sage Weil Signed-off-by: "Eric W. Biederman" --- fs/ceph/inode.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index 2971eaa65cdc..40e0787022e9 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -612,8 +612,8 @@ static int fill_inode(struct inode *inode, if ((issued & CEPH_CAP_AUTH_EXCL) == 0) { inode->i_mode = le32_to_cpu(info->mode); - inode->i_uid = le32_to_cpu(info->uid); - inode->i_gid = le32_to_cpu(info->gid); + inode->i_uid = make_kuid(&init_user_ns, le32_to_cpu(info->uid)); + inode->i_gid = make_kgid(&init_user_ns, le32_to_cpu(info->gid)); dout("%p mode 0%o uid.gid %d.%d\n", inode, inode->i_mode, inode->i_uid, inode->i_gid); } @@ -1570,8 +1570,9 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) inode->i_uid = attr->ia_uid; dirtied |= CEPH_CAP_AUTH_EXCL; } else if ((issued & CEPH_CAP_AUTH_SHARED) == 0 || - attr->ia_uid != inode->i_uid) { - req->r_args.setattr.uid = cpu_to_le32(attr->ia_uid); + !uid_eq(attr->ia_uid, inode->i_uid)) { + req->r_args.setattr.uid = cpu_to_le32( + from_kuid(&init_user_ns, attr->ia_uid)); mask |= CEPH_SETATTR_UID; release |= CEPH_CAP_AUTH_SHARED; } @@ -1583,8 +1584,9 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) inode->i_gid = attr->ia_gid; dirtied |= CEPH_CAP_AUTH_EXCL; } else if ((issued & CEPH_CAP_AUTH_SHARED) == 0 || - attr->ia_gid != inode->i_gid) { - req->r_args.setattr.gid = cpu_to_le32(attr->ia_gid); + !gid_eq(attr->ia_gid, inode->i_gid)) { + req->r_args.setattr.gid = cpu_to_le32( + from_kgid(&init_user_ns, attr->ia_gid)); mask |= CEPH_SETATTR_GID; release |= CEPH_CAP_AUTH_SHARED; } -- cgit From ff3d0046625c1b37df37beb8477135d44dae2823 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 04:01:53 -0800 Subject: ceph: Convert struct ceph_mds_request to use kuid_t and kgid_t Hold the uid and gid for a pending ceph mds request using the types kuid_t and kgid_t. When a request message is finally created convert the kuid_t and kgid_t values into uids and gids in the initial user namespace. Cc: Sage Weil Signed-off-by: "Eric W. Biederman" --- fs/ceph/mds_client.c | 4 ++-- fs/ceph/mds_client.h | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 9165eb8309eb..7a3dfe0a9a80 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -1658,8 +1658,8 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc, head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch); head->op = cpu_to_le32(req->r_op); - head->caller_uid = cpu_to_le32(req->r_uid); - head->caller_gid = cpu_to_le32(req->r_gid); + head->caller_uid = cpu_to_le32(from_kuid(&init_user_ns, req->r_uid)); + head->caller_gid = cpu_to_le32(from_kgid(&init_user_ns, req->r_gid)); head->args = req->r_args; ceph_encode_filepath(&p, end, ino1, path1); diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index dd26846dd71d..ff4188bf6199 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -184,8 +184,8 @@ struct ceph_mds_request { union ceph_mds_request_args r_args; int r_fmode; /* file mode, if expecting cap */ - uid_t r_uid; - gid_t r_gid; + kuid_t r_uid; + kgid_t r_gid; /* for choosing which mds to send this request to */ int r_direct_mode; -- cgit From bd2bae6a66df9261a39e47291b0a6b00cd0831e0 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 04:05:39 -0800 Subject: ceph: Convert kuids and kgids before printing them. Before printing kuid and kgids values convert them into the initial user namespace. Cc: Sage Weil Signed-off-by: "Eric W. Biederman" --- fs/ceph/caps.c | 3 ++- fs/ceph/inode.c | 9 ++++++--- 2 files changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 39eb4665b122..ae2be696eb5b 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -2362,7 +2362,8 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant, inode->i_uid = make_kuid(&init_user_ns, le32_to_cpu(grant->uid)); inode->i_gid = make_kgid(&init_user_ns, le32_to_cpu(grant->gid)); dout("%p mode 0%o uid.gid %d.%d\n", inode, inode->i_mode, - inode->i_uid, inode->i_gid); + from_kuid(&init_user_ns, inode->i_uid), + from_kgid(&init_user_ns, inode->i_gid)); } if ((issued & CEPH_CAP_LINK_EXCL) == 0) diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index 40e0787022e9..d45895f4a04d 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -615,7 +615,8 @@ static int fill_inode(struct inode *inode, inode->i_uid = make_kuid(&init_user_ns, le32_to_cpu(info->uid)); inode->i_gid = make_kgid(&init_user_ns, le32_to_cpu(info->gid)); dout("%p mode 0%o uid.gid %d.%d\n", inode, inode->i_mode, - inode->i_uid, inode->i_gid); + from_kuid(&init_user_ns, inode->i_uid), + from_kgid(&init_user_ns, inode->i_gid)); } if ((issued & CEPH_CAP_LINK_EXCL) == 0) @@ -1565,7 +1566,8 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) if (ia_valid & ATTR_UID) { dout("setattr %p uid %d -> %d\n", inode, - inode->i_uid, attr->ia_uid); + from_kuid(&init_user_ns, inode->i_uid), + from_kuid(&init_user_ns, attr->ia_uid)); if (issued & CEPH_CAP_AUTH_EXCL) { inode->i_uid = attr->ia_uid; dirtied |= CEPH_CAP_AUTH_EXCL; @@ -1579,7 +1581,8 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) } if (ia_valid & ATTR_GID) { dout("setattr %p gid %d -> %d\n", inode, - inode->i_gid, attr->ia_gid); + from_kgid(&init_user_ns, inode->i_gid), + from_kgid(&init_user_ns, attr->ia_gid)); if (issued & CEPH_CAP_AUTH_EXCL) { inode->i_gid = attr->ia_gid; dirtied |= CEPH_CAP_AUTH_EXCL; -- cgit From f791f7c5e354870eaa5e31c4038c6723683283f1 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 29 Jan 2013 16:09:41 -0800 Subject: 9p: Transmit kuid and kgid values Modify the p9_client_rpc format specifiers of every function that directly transmits a uid or a gid from 'd' to 'u' or 'g' as appropriate. Modify those same functions to take kuid_t and kgid_t parameters instead of uid_t and gid_t parameters. Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: Eric W. Biederman --- fs/9p/v9fs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index d934f04e7736..4e0bd9d62ed6 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -375,7 +375,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->flags &= ~V9FS_ACL_MASK; } - fid = p9_client_attach(v9ses->clnt, NULL, v9ses->uname, ~0, + fid = p9_client_attach(v9ses->clnt, NULL, v9ses->uname, INVALID_UID, v9ses->aname); if (IS_ERR(fid)) { retval = PTR_ERR(fid); -- cgit From 447c50943fd008755122c7a62bac068e73c1cf2c Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 29 Jan 2013 16:18:50 -0800 Subject: 9p: Modify the stat structures to use kuid_t and kgid_t 9p has thre strucrtures that can encode inode stat information. Modify all of those structures to contain kuid_t and kgid_t values. Modify he wire encoders and decoders of those structures to use 'u' and 'g' instead of 'd' in the format string where uids and gids are present. This results in all kuid and kgid conversion to and from on the wire values being performed by the same code in protocol.c where the client is known at the time of the conversion. Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: Eric W. Biederman --- fs/9p/vfs_inode.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 890bed538f9b..1581fe218934 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -228,9 +228,9 @@ v9fs_blank_wstat(struct p9_wstat *wstat) wstat->uid = NULL; wstat->gid = NULL; wstat->muid = NULL; - wstat->n_uid = ~0; - wstat->n_gid = ~0; - wstat->n_muid = ~0; + wstat->n_uid = INVALID_UID; + wstat->n_gid = INVALID_GID; + wstat->n_muid = INVALID_UID; wstat->extension = NULL; } -- cgit From b464255699077c6b33ea58ee01db80f5729511ad Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2013 11:48:53 -0800 Subject: 9p: Modify struct 9p_fid to use a kuid_t not a uid_t Change struct 9p_fid and it's associated functions to use kuid_t's instead of uid_t. Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: "Eric W. Biederman" --- fs/9p/fid.c | 17 +++++++++-------- fs/9p/v9fs.c | 2 +- 2 files changed, 10 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index da8eefbe830d..afd4724b2d92 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -74,19 +74,20 @@ int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) * */ -static struct p9_fid *v9fs_fid_find(struct dentry *dentry, u32 uid, int any) +static struct p9_fid *v9fs_fid_find(struct dentry *dentry, kuid_t uid, int any) { struct v9fs_dentry *dent; struct p9_fid *fid, *ret; p9_debug(P9_DEBUG_VFS, " dentry: %s (%p) uid %d any %d\n", - dentry->d_name.name, dentry, uid, any); + dentry->d_name.name, dentry, from_kuid(&init_user_ns, uid), + any); dent = (struct v9fs_dentry *) dentry->d_fsdata; ret = NULL; if (dent) { spin_lock(&dent->lock); list_for_each_entry(fid, &dent->fidlist, dlist) { - if (any || fid->uid == uid) { + if (any || uid_eq(fid->uid, uid)) { ret = fid; break; } @@ -126,7 +127,7 @@ err_out: } static struct p9_fid *v9fs_fid_lookup_with_uid(struct dentry *dentry, - uid_t uid, int any) + kuid_t uid, int any) { struct dentry *ds; char **wnames, *uname; @@ -233,7 +234,7 @@ err_out: struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) { - uid_t uid; + kuid_t uid; int any, access; struct v9fs_session_info *v9ses; @@ -253,7 +254,7 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) break; default: - uid = ~0; + uid = INVALID_UID; any = 0; break; } @@ -272,7 +273,7 @@ struct p9_fid *v9fs_fid_clone(struct dentry *dentry) return ret; } -static struct p9_fid *v9fs_fid_clone_with_uid(struct dentry *dentry, uid_t uid) +static struct p9_fid *v9fs_fid_clone_with_uid(struct dentry *dentry, kuid_t uid) { struct p9_fid *fid, *ret; @@ -289,7 +290,7 @@ struct p9_fid *v9fs_writeback_fid(struct dentry *dentry) int err; struct p9_fid *fid; - fid = v9fs_fid_clone_with_uid(dentry, 0); + fid = v9fs_fid_clone_with_uid(dentry, GLOBAL_ROOT_UID); if (IS_ERR(fid)) goto error_out; /* diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 4e0bd9d62ed6..d64967cbd73e 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -387,7 +387,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, if ((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_SINGLE) fid->uid = v9ses->uid; else - fid->uid = ~0; + fid->uid = INVALID_UID; #ifdef CONFIG_9P_FSCACHE /* register the session for caching */ -- cgit From 76ed23a5d703d94ede1ef6c12c14a75add691202 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2013 11:57:40 -0800 Subject: 9p: Modify struct v9fs_session_info to use a kuids and kgids Change struct v9fs_session_info and the code that popluates it to use kuids and kgids. When parsing the 9p mount options convert the dfltuid, dflutgid, and the session uid from the current user namespace into kuids and kgids. Modify V9FS_DEFUID and V9FS_DEFGUID to be kuid and kgid values. Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: "Eric W. Biederman" --- fs/9p/v9fs.c | 30 +++++++++++++++++++++++++----- fs/9p/v9fs.h | 10 +++++----- 2 files changed, 30 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index d64967cbd73e..58e6cbce4156 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -161,7 +161,13 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) ret = r; continue; } - v9ses->dfltuid = option; + v9ses->dfltuid = make_kuid(current_user_ns(), option); + if (!uid_valid(v9ses->dfltuid)) { + p9_debug(P9_DEBUG_ERROR, + "uid field, but not a uid?\n"); + ret = -EINVAL; + continue; + } break; case Opt_dfltgid: r = match_int(&args[0], &option); @@ -171,7 +177,13 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) ret = r; continue; } - v9ses->dfltgid = option; + v9ses->dfltgid = make_kgid(current_user_ns(), option); + if (!gid_valid(v9ses->dfltgid)) { + p9_debug(P9_DEBUG_ERROR, + "gid field, but not a gid?\n"); + ret = -EINVAL; + continue; + } break; case Opt_afid: r = match_int(&args[0], &option); @@ -248,8 +260,9 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) else if (strcmp(s, "client") == 0) { v9ses->flags |= V9FS_ACCESS_CLIENT; } else { + uid_t uid; v9ses->flags |= V9FS_ACCESS_SINGLE; - v9ses->uid = simple_strtoul(s, &e, 10); + uid = simple_strtoul(s, &e, 10); if (*e != '\0') { ret = -EINVAL; pr_info("Unknown access argument %s\n", @@ -257,6 +270,13 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) kfree(s); goto free_and_return; } + v9ses->uid = make_kuid(current_user_ns(), uid); + if (!uid_valid(v9ses->uid)) { + ret = -EINVAL; + pr_info("Uknown uid %s\n", s); + kfree(s); + goto free_and_return; + } } kfree(s); @@ -319,7 +339,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, list_add(&v9ses->slist, &v9fs_sessionlist); spin_unlock(&v9fs_sessionlist_lock); - v9ses->uid = ~0; + v9ses->uid = INVALID_UID; v9ses->dfltuid = V9FS_DEFUID; v9ses->dfltgid = V9FS_DEFGID; @@ -364,7 +384,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->flags &= ~V9FS_ACCESS_MASK; v9ses->flags |= V9FS_ACCESS_ANY; - v9ses->uid = ~0; + v9ses->uid = INVALID_UID; } if (!v9fs_proto_dotl(v9ses) || !((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_CLIENT)) { diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index 34c59f14a1c9..a8e127c89627 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -109,9 +109,9 @@ struct v9fs_session_info { char *uname; /* user name to mount as */ char *aname; /* name of remote hierarchy being mounted */ unsigned int maxdata; /* max data for client interface */ - unsigned int dfltuid; /* default uid/muid for legacy support */ - unsigned int dfltgid; /* default gid for legacy support */ - u32 uid; /* if ACCESS_SINGLE, the uid that has access */ + kuid_t dfltuid; /* default uid/muid for legacy support */ + kgid_t dfltgid; /* default gid for legacy support */ + kuid_t uid; /* if ACCESS_SINGLE, the uid that has access */ struct p9_client *clnt; /* 9p client */ struct list_head slist; /* list of sessions registered with v9fs */ struct backing_dev_info bdi; @@ -165,8 +165,8 @@ extern struct inode *v9fs_inode_from_fid_dotl(struct v9fs_session_info *v9ses, #define V9FS_PORT 564 #define V9FS_DEFUSER "nobody" #define V9FS_DEFANAME "" -#define V9FS_DEFUID (-2) -#define V9FS_DEFGID (-2) +#define V9FS_DEFUID KUIDT_INIT(-2) +#define V9FS_DEFGID KGIDT_INIT(-2) static inline struct v9fs_session_info *v9fs_inode2v9ses(struct inode *inode) { -- cgit From d4ef4e3581044890e4a04607a4e8f42d3ebcf704 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2013 12:08:21 -0800 Subject: 9p: Modify v9fs_get_fsgid_for_create to return a kgid Modify v9fs_get_fsgid_for_create to return a kgid and modify all of the variables that hold the result of v9fs_get_fsgid_for_create to be of type kgid_t. Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: "Eric W. Biederman" --- fs/9p/vfs_inode_dotl.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index 40895546e103..e27886573e7d 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -57,7 +57,7 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode, * group of the new file system object. */ -static gid_t v9fs_get_fsgid_for_create(struct inode *dir_inode) +static kgid_t v9fs_get_fsgid_for_create(struct inode *dir_inode) { BUG_ON(dir_inode == NULL); @@ -246,7 +246,7 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, int *opened) { int err = 0; - gid_t gid; + kgid_t gid; umode_t mode; char *name = NULL; struct p9_qid qid; @@ -391,7 +391,7 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, int err; struct v9fs_session_info *v9ses; struct p9_fid *fid = NULL, *dfid = NULL; - gid_t gid; + kgid_t gid; char *name; umode_t mode; struct inode *inode; @@ -692,7 +692,7 @@ v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry, const char *symname) { int err; - gid_t gid; + kgid_t gid; char *name; struct p9_qid qid; struct inode *inode; @@ -832,7 +832,7 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode, dev_t rdev) { int err; - gid_t gid; + kgid_t gid; char *name; umode_t mode; struct v9fs_session_info *v9ses; -- cgit From 66fdb93f882d21612a5287cd1303c9b1391ebf5d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2013 13:04:05 -0800 Subject: afs: Remove unused structure afs_store_status While looking for kuid_t and kgid_t conversions I found this structure that has never been used since it was added to the kernel in 2007. The obvious for this structure to be used is in xdr_encode_AFS_StoreStatus and that function uses a small handful of local variables instead. So remove the unnecessary structure to prevent confusion. Cc: David Howells Signed-off-by: "Eric W. Biederman" --- fs/afs/afs.h | 7 ------- 1 file changed, 7 deletions(-) (limited to 'fs') diff --git a/fs/afs/afs.h b/fs/afs/afs.h index c548aa346f0d..3d8fd35c0dd0 100644 --- a/fs/afs/afs.h +++ b/fs/afs/afs.h @@ -133,13 +133,6 @@ struct afs_file_status { /* * AFS file status change request */ -struct afs_store_status { - u32 mask; /* which bits of the struct are set */ - u32 mtime_client; /* last time client changed data */ - u32 owner; /* owner ID */ - u32 group; /* group ID */ - umode_t mode; /* UNIX mode */ -}; #define AFS_SET_MTIME 0x01 /* set the mtime */ #define AFS_SET_OWNER 0x02 /* set the owner ID */ -- cgit From 3a46741804a4226cd837d1246eed95d4161f7159 Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Tue, 12 Feb 2013 10:56:54 -0700 Subject: eCryptfs: decrypt_pki_encrypted_session_key(): remove kfree() redundant null check smatch analysis: fs/ecryptfs/keystore.c:1206 decrypt_pki_encrypted_session_key() info: redundant null check on msg calling kfree() Cc: Dustin Kirkland Cc: ecryptfs@vger.kernel.org Signed-off-by: Tim Gardner Signed-off-by: Tyler Hicks --- fs/ecryptfs/keystore.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index 6154cde3a052..5aceff202dc0 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -1202,8 +1202,7 @@ decrypt_pki_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, crypt_stat->key_size); } out: - if (msg) - kfree(msg); + kfree(msg); return rc; } -- cgit From 1101d58669a92ed9c9f4c7281404fb1e067a1e28 Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Tue, 12 Feb 2013 11:03:49 -0700 Subject: ecryptfs: ecryptfs_msg_ctx_alloc_to_free(): remove kfree() redundant null check smatch analysis: fs/ecryptfs/messaging.c:101 ecryptfs_msg_ctx_alloc_to_free() info: redundant null check on msg_ctx->msg calling kfree() Cc: Dustin Kirkland Cc: ecryptfs@vger.kernel.org Signed-off-by: Tim Gardner Signed-off-by: Tyler Hicks --- fs/ecryptfs/messaging.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/messaging.c b/fs/ecryptfs/messaging.c index 5fa2471796c2..d5c7297c5816 100644 --- a/fs/ecryptfs/messaging.c +++ b/fs/ecryptfs/messaging.c @@ -97,8 +97,7 @@ static void ecryptfs_msg_ctx_free_to_alloc(struct ecryptfs_msg_ctx *msg_ctx) void ecryptfs_msg_ctx_alloc_to_free(struct ecryptfs_msg_ctx *msg_ctx) { list_move(&(msg_ctx->node), &ecryptfs_msg_ctx_free_list); - if (msg_ctx->msg) - kfree(msg_ctx->msg); + kfree(msg_ctx->msg); msg_ctx->msg = NULL; msg_ctx->state = ECRYPTFS_MSG_CTX_STATE_FREE; } -- cgit From fb0af3f2b1b613e5ea75426d454c7e5b1d1eef49 Mon Sep 17 00:00:00 2001 From: Josh Boyer Date: Tue, 12 Feb 2013 13:07:22 -0800 Subject: pstore: Create a convenient mount point for pstore Using /dev/pstore as a mount point for the pstore filesystem is slightly awkward. We don't normally mount filesystems in /dev/ and the /dev/pstore file isn't created automatically by anything. While this method will still work, we can create a persistent mount point in sysfs. This will put pstore on par with things like cgroups and efivarfs. Signed-off-by: Josh Boyer Acked-by: Kees Cook Signed-off-by: Tony Luck --- fs/pstore/inode.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c index 67de74ca85f4..e4bcb2cf055a 100644 --- a/fs/pstore/inode.c +++ b/fs/pstore/inode.c @@ -418,9 +418,25 @@ static struct file_system_type pstore_fs_type = { .kill_sb = pstore_kill_sb, }; +static struct kobject *pstore_kobj; + static int __init init_pstore_fs(void) { - return register_filesystem(&pstore_fs_type); + int err = 0; + + /* Create a convenient mount point for people to access pstore */ + pstore_kobj = kobject_create_and_add("pstore", fs_kobj); + if (!pstore_kobj) { + err = -ENOMEM; + goto out; + } + + err = register_filesystem(&pstore_fs_type); + if (err < 0) + kobject_put(pstore_kobj); + +out: + return err; } module_init(init_pstore_fs) -- cgit From fd95e81cb1c74c9acd2356821faa9f24c2fec365 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Wed, 13 Feb 2013 12:21:40 +0000 Subject: GFS2: Reinstate withdraw ack system This patch reinstates the ack system which withdraw should be using. It appears to have been accidentally forgotten when the lock module was merged into GFS2, due to two different sysfs files having the same name. Reported-by: David Teigland Signed-off-by: Steven Whitehouse --- fs/gfs2/incore.h | 1 + fs/gfs2/ops_fstype.c | 1 + fs/gfs2/sys.c | 24 +++++++++++++++++++++++- fs/gfs2/util.c | 3 +++ 4 files changed, 28 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 1533cf8b4269..e2601ba38ef5 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -642,6 +642,7 @@ struct gfs2_sbd { wait_queue_head_t sd_glock_wait; atomic_t sd_glock_disposal; struct completion sd_locking_init; + struct completion sd_wdack; struct delayed_work sd_control_work; /* Inode Stuff */ diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index e063f22d9e4c..1b612be4b873 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -81,6 +81,7 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) init_waitqueue_head(&sdp->sd_glock_wait); atomic_set(&sdp->sd_glock_disposal, 0); init_completion(&sdp->sd_locking_init); + init_completion(&sdp->sd_wdack); spin_lock_init(&sdp->sd_statfs_spin); spin_lock_init(&sdp->sd_rindex_spin); diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 462e84142759..4fb9ad80d260 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -330,6 +330,28 @@ static ssize_t block_store(struct gfs2_sbd *sdp, const char *buf, size_t len) return ret; } +static ssize_t wdack_show(struct gfs2_sbd *sdp, char *buf) +{ + int val = completion_done(&sdp->sd_wdack) ? 1 : 0; + + return sprintf(buf, "%d\n", val); +} + +static ssize_t wdack_store(struct gfs2_sbd *sdp, const char *buf, size_t len) +{ + ssize_t ret = len; + int val; + + val = simple_strtol(buf, NULL, 0); + + if ((val == 1) && + !strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm")) + complete(&sdp->sd_wdack); + else + ret = -EINVAL; + return ret; +} + static ssize_t lkfirst_show(struct gfs2_sbd *sdp, char *buf) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; @@ -461,7 +483,7 @@ static struct gfs2_attr gdlm_attr_##_name = __ATTR(_name,_mode,_show,_store) GDLM_ATTR(proto_name, 0444, proto_name_show, NULL); GDLM_ATTR(block, 0644, block_show, block_store); -GDLM_ATTR(withdraw, 0644, withdraw_show, withdraw_store); +GDLM_ATTR(withdraw, 0644, wdack_show, wdack_store); GDLM_ATTR(jid, 0644, jid_show, jid_store); GDLM_ATTR(first, 0644, lkfirst_show, lkfirst_store); GDLM_ATTR(first_done, 0444, first_done_show, NULL); diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c index f00d7c5744f6..6402fb69d71b 100644 --- a/fs/gfs2/util.c +++ b/fs/gfs2/util.c @@ -54,6 +54,9 @@ int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...) kobject_uevent(&sdp->sd_kobj, KOBJ_OFFLINE); + if (!strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm")) + wait_for_completion(&sdp->sd_wdack); + if (lm->lm_unmount) { fs_err(sdp, "telling LM to unmount\n"); lm->lm_unmount(sdp); -- cgit From f74f70f8b10b435f5f20247e70d1d86b53a59685 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 04:23:54 -0800 Subject: afs: Only allow mounting afs in the intial network namespace rxrpc sockets only work in the initial network namespace so it isn't possible to support afs in any other network namespace. Cc: David Howells Signed-off-by: "Eric W. Biederman" --- fs/afs/super.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/afs/super.c b/fs/afs/super.c index 43165009428d..7c31ec399575 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -24,6 +24,8 @@ #include #include #include +#include +#include #include "internal.h" #define AFS_FS_MAGIC 0x6B414653 /* 'kAFS' */ @@ -363,6 +365,10 @@ static struct dentry *afs_mount(struct file_system_type *fs_type, memset(¶ms, 0, sizeof(params)); + ret = -EINVAL; + if (current->nsproxy->net_ns != &init_net) + goto error; + /* parse the options and device name */ if (options) { ret = afs_parse_options(¶ms, options, &dev_name); -- cgit From a0a5386ac6400493cc2eb8b58583e56af0708730 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 7 Feb 2012 16:20:48 -0800 Subject: afs: Support interacting with multiple user namespaces Modify struct afs_file_status to store owner as a kuid_t and group as a kgid_t. In xdr_decode_AFSFetchStatus as owner is now a kuid_t and group is now a kgid_t don't use the EXTRACT macro. Instead perform the work of the extract macro explicitly. Read the value with ntohl and convert it to the appropriate type with make_kuid or make_kgid. Test if the value is different from what is stored in status and update changed. Update the value in status. In xdr_encode_AFS_StoreStatus call from_kuid or from_kgid as we are computing the on the wire encoding. Initialize uids with GLOBAL_ROOT_UID instead of 0. Initialize gids with GLOBAL_ROOT_GID instead of 0. Cc: David Howells Acked-by: Serge Hallyn Signed-off-by: Eric W. Biederman --- fs/afs/afs.h | 4 ++-- fs/afs/fsclient.c | 14 ++++++++++---- fs/afs/inode.c | 6 +++--- 3 files changed, 15 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/afs/afs.h b/fs/afs/afs.h index 3d8fd35c0dd0..3c462ff6db63 100644 --- a/fs/afs/afs.h +++ b/fs/afs/afs.h @@ -119,8 +119,8 @@ struct afs_file_status { u64 size; /* file size */ afs_dataversion_t data_version; /* current data version */ u32 author; /* author ID */ - u32 owner; /* owner ID */ - u32 group; /* group ID */ + kuid_t owner; /* owner ID */ + kgid_t group; /* group ID */ afs_access_t caller_access; /* access rights for authenticated caller */ afs_access_t anon_access; /* access rights for unauthenticated caller */ umode_t mode; /* UNIX mode */ diff --git a/fs/afs/fsclient.c b/fs/afs/fsclient.c index b960ff05ea0b..c2e930ec2888 100644 --- a/fs/afs/fsclient.c +++ b/fs/afs/fsclient.c @@ -42,6 +42,8 @@ static void xdr_decode_AFSFetchStatus(const __be32 **_bp, umode_t mode; u64 data_version, size; u32 changed = 0; /* becomes non-zero if ctime-type changes seen */ + kuid_t owner; + kgid_t group; #define EXTRACT(DST) \ do { \ @@ -56,7 +58,9 @@ static void xdr_decode_AFSFetchStatus(const __be32 **_bp, size = ntohl(*bp++); data_version = ntohl(*bp++); EXTRACT(status->author); - EXTRACT(status->owner); + owner = make_kuid(&init_user_ns, ntohl(*bp++)); + changed |= !uid_eq(owner, status->owner); + status->owner = owner; EXTRACT(status->caller_access); /* call ticket dependent */ EXTRACT(status->anon_access); EXTRACT(status->mode); @@ -65,7 +69,9 @@ static void xdr_decode_AFSFetchStatus(const __be32 **_bp, bp++; /* seg size */ status->mtime_client = ntohl(*bp++); status->mtime_server = ntohl(*bp++); - EXTRACT(status->group); + group = make_kgid(&init_user_ns, ntohl(*bp++)); + changed |= !gid_eq(group, status->group); + status->group = group; bp++; /* sync counter */ data_version |= (u64) ntohl(*bp++) << 32; EXTRACT(status->lock_count); @@ -181,12 +187,12 @@ static void xdr_encode_AFS_StoreStatus(__be32 **_bp, struct iattr *attr) if (attr->ia_valid & ATTR_UID) { mask |= AFS_SET_OWNER; - owner = attr->ia_uid; + owner = from_kuid(&init_user_ns, attr->ia_uid); } if (attr->ia_valid & ATTR_GID) { mask |= AFS_SET_GROUP; - group = attr->ia_gid; + group = from_kgid(&init_user_ns, attr->ia_gid); } if (attr->ia_valid & ATTR_MODE) { diff --git a/fs/afs/inode.c b/fs/afs/inode.c index 95cffd38239f..789bc253b5f6 100644 --- a/fs/afs/inode.c +++ b/fs/afs/inode.c @@ -69,7 +69,7 @@ static int afs_inode_map_status(struct afs_vnode *vnode, struct key *key) set_nlink(inode, vnode->status.nlink); inode->i_uid = vnode->status.owner; - inode->i_gid = 0; + inode->i_gid = GLOBAL_ROOT_GID; inode->i_size = vnode->status.size; inode->i_ctime.tv_sec = vnode->status.mtime_server; inode->i_ctime.tv_nsec = 0; @@ -175,8 +175,8 @@ struct inode *afs_iget_autocell(struct inode *dir, const char *dev_name, inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO; inode->i_op = &afs_autocell_inode_operations; set_nlink(inode, 2); - inode->i_uid = 0; - inode->i_gid = 0; + inode->i_uid = GLOBAL_ROOT_UID; + inode->i_gid = GLOBAL_ROOT_GID; inode->i_ctime.tv_sec = get_seconds(); inode->i_ctime.tv_nsec = 0; inode->i_atime = inode->i_mtime = inode->i_ctime; -- cgit From 9fd973e085f7759f710603422b2e11ad5f2e000d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2013 18:50:54 -0800 Subject: coda: Restrict coda messages to the initial pid namespace Remove the slight chance that pids in coda messages will be interpreted in the wrong pid namespace. - Explicitly send all pids in coda messages in the initial pid namespace. - Only allow mounts from processes in the initial pid namespace. - Only allow processes in the initial pid namespace to open the coda character device to communicate with coda. Cc: Jan Harkes Signed-off-by: "Eric W. Biederman" --- fs/coda/inode.c | 4 ++++ fs/coda/psdev.c | 4 ++++ fs/coda/upcall.c | 4 ++-- 3 files changed, 10 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/coda/inode.c b/fs/coda/inode.c index be2aa4909487..77bbaf4666a7 100644 --- a/fs/coda/inode.c +++ b/fs/coda/inode.c @@ -20,6 +20,7 @@ #include #include #include +#include #include @@ -157,6 +158,9 @@ static int coda_fill_super(struct super_block *sb, void *data, int silent) int error; int idx; + if (task_active_pid_ns(current) != &init_pid_ns) + return -EINVAL; + idx = get_device_index((struct coda_mount_data *) data); /* Ignore errors in data, for backward compatibility */ diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index 761d5b31b18d..dd60f905d4fe 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include #include @@ -266,6 +267,9 @@ static int coda_psdev_open(struct inode * inode, struct file * file) struct venus_comm *vcp; int idx, err; + if (task_active_pid_ns(current) != &init_pid_ns) + return -EINVAL; + idx = iminor(inode); if (idx < 0 || idx >= MAX_CODADEVS) return -ENODEV; diff --git a/fs/coda/upcall.c b/fs/coda/upcall.c index 0c68fd31fbf2..5c6d2cd6ee86 100644 --- a/fs/coda/upcall.c +++ b/fs/coda/upcall.c @@ -50,8 +50,8 @@ static void *alloc_upcall(int opcode, int size) return ERR_PTR(-ENOMEM); inp->ih.opcode = opcode; - inp->ih.pid = current->pid; - inp->ih.pgid = task_pgrp_nr(current); + inp->ih.pid = task_pid_nr_ns(current, &init_pid_ns); + inp->ih.pgid = task_pgrp_nr_ns(current, &init_pid_ns); inp->ih.uid = current_fsuid(); return (void*)inp; -- cgit From d83f5901bc0cd7131a3b8534169ee889efc4c257 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2013 19:21:14 -0800 Subject: coda: Restrict coda messages to the initial user namespace Remove the slight chance that uids and gids in coda messages will be interpreted in the wrong user namespace. - Only allow processes in the initial user namespace to open the coda character device to communicate with coda filesystems. - Explicitly convert the uids in the coda header into the initial user namespace. - In coda_vattr_to_attr make kuids and kgids from the initial user namespace uids and gids in struct coda_vattr that just came from userspace. - In coda_iattr_to_vattr convert kuids and kgids into uids and gids in the intial user namespace and store them in struct coda_vattr for sending to coda userspace programs. Nothing needs to be changed with mounts as coda does not support being mounted in anything other than the initial user namespace. Cc: Jan Harkes Signed-off-by: "Eric W. Biederman" --- fs/coda/coda_linux.c | 8 ++++---- fs/coda/psdev.c | 3 +++ fs/coda/upcall.c | 6 +++--- 3 files changed, 10 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/coda/coda_linux.c b/fs/coda/coda_linux.c index 854ace712685..2849f41e72a2 100644 --- a/fs/coda/coda_linux.c +++ b/fs/coda/coda_linux.c @@ -100,9 +100,9 @@ void coda_vattr_to_iattr(struct inode *inode, struct coda_vattr *attr) if (attr->va_mode != (u_short) -1) inode->i_mode = attr->va_mode | inode_type; if (attr->va_uid != -1) - inode->i_uid = (uid_t) attr->va_uid; + inode->i_uid = make_kuid(&init_user_ns, (uid_t) attr->va_uid); if (attr->va_gid != -1) - inode->i_gid = (gid_t) attr->va_gid; + inode->i_gid = make_kgid(&init_user_ns, (gid_t) attr->va_gid); if (attr->va_nlink != -1) set_nlink(inode, attr->va_nlink); if (attr->va_size != -1) @@ -171,10 +171,10 @@ void coda_iattr_to_vattr(struct iattr *iattr, struct coda_vattr *vattr) vattr->va_mode = iattr->ia_mode; } if ( valid & ATTR_UID ) { - vattr->va_uid = (vuid_t) iattr->ia_uid; + vattr->va_uid = (vuid_t) from_kuid(&init_user_ns, iattr->ia_uid); } if ( valid & ATTR_GID ) { - vattr->va_gid = (vgid_t) iattr->ia_gid; + vattr->va_gid = (vgid_t) from_kgid(&init_user_ns, iattr->ia_gid); } if ( valid & ATTR_SIZE ) { vattr->va_size = iattr->ia_size; diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index dd60f905d4fe..ebc2bae6c289 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -270,6 +270,9 @@ static int coda_psdev_open(struct inode * inode, struct file * file) if (task_active_pid_ns(current) != &init_pid_ns) return -EINVAL; + if (current_user_ns() != &init_user_ns) + return -EINVAL; + idx = iminor(inode); if (idx < 0 || idx >= MAX_CODADEVS) return -ENODEV; diff --git a/fs/coda/upcall.c b/fs/coda/upcall.c index 5c6d2cd6ee86..3a731976dc5e 100644 --- a/fs/coda/upcall.c +++ b/fs/coda/upcall.c @@ -52,7 +52,7 @@ static void *alloc_upcall(int opcode, int size) inp->ih.opcode = opcode; inp->ih.pid = task_pid_nr_ns(current, &init_pid_ns); inp->ih.pgid = task_pgrp_nr_ns(current, &init_pid_ns); - inp->ih.uid = current_fsuid(); + inp->ih.uid = from_kuid(&init_user_ns, current_fsuid()); return (void*)inp; } @@ -157,7 +157,7 @@ int venus_lookup(struct super_block *sb, struct CodaFid *fid, } int venus_close(struct super_block *sb, struct CodaFid *fid, int flags, - vuid_t uid) + kuid_t uid) { union inputArgs *inp; union outputArgs *outp; @@ -166,7 +166,7 @@ int venus_close(struct super_block *sb, struct CodaFid *fid, int flags, insize = SIZE(release); UPARG(CODA_CLOSE); - inp->ih.uid = uid; + inp->ih.uid = from_kuid(&init_user_ns, uid); inp->coda_close.VFid = *fid; inp->coda_close.flags = flags; -- cgit From 17499e332962e495b3ee2e5507e09b853ed6f607 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 30 Jan 2013 19:36:06 -0800 Subject: coda: Cache permisions in struct coda_inode_info in a kuid_t. - Change c_uid in struct coda_indoe_info from a vuid_t to a kuid_t. - Initialize c_uid to GLOBAL_ROOT_UID instead of 0. - Use uid_eq to compare cached kuids. Cc: Jan Harkes Signed-off-by: "Eric W. Biederman" --- fs/coda/cache.c | 4 ++-- fs/coda/coda_fs_i.h | 2 +- fs/coda/inode.c | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/coda/cache.c b/fs/coda/cache.c index 958ae0e0ff8c..1da168c61d35 100644 --- a/fs/coda/cache.c +++ b/fs/coda/cache.c @@ -33,7 +33,7 @@ void coda_cache_enter(struct inode *inode, int mask) spin_lock(&cii->c_lock); cii->c_cached_epoch = atomic_read(&permission_epoch); - if (cii->c_uid != current_fsuid()) { + if (!uid_eq(cii->c_uid, current_fsuid())) { cii->c_uid = current_fsuid(); cii->c_cached_perm = mask; } else @@ -65,7 +65,7 @@ int coda_cache_check(struct inode *inode, int mask) spin_lock(&cii->c_lock); hit = (mask & cii->c_cached_perm) == mask && - cii->c_uid == current_fsuid() && + uid_eq(cii->c_uid, current_fsuid()) && cii->c_cached_epoch == atomic_read(&permission_epoch); spin_unlock(&cii->c_lock); diff --git a/fs/coda/coda_fs_i.h b/fs/coda/coda_fs_i.h index b24fdfd8a3f0..c64075213218 100644 --- a/fs/coda/coda_fs_i.h +++ b/fs/coda/coda_fs_i.h @@ -25,7 +25,7 @@ struct coda_inode_info { u_short c_flags; /* flags (see below) */ unsigned int c_mapcount; /* nr of times this inode is mapped */ unsigned int c_cached_epoch; /* epoch for cached permissions */ - vuid_t c_uid; /* fsuid for cached permissions */ + kuid_t c_uid; /* fsuid for cached permissions */ unsigned int c_cached_perm; /* cached access permissions */ spinlock_t c_lock; struct inode vfs_inode; diff --git a/fs/coda/inode.c b/fs/coda/inode.c index 77bbaf4666a7..cf674e9179a3 100644 --- a/fs/coda/inode.c +++ b/fs/coda/inode.c @@ -49,7 +49,7 @@ static struct inode *coda_alloc_inode(struct super_block *sb) return NULL; memset(&ei->c_fid, 0, sizeof(struct CodaFid)); ei->c_flags = 0; - ei->c_uid = 0; + ei->c_uid = GLOBAL_ROOT_UID; ei->c_cached_perm = 0; spin_lock_init(&ei->c_lock); return &ei->vfs_inode; -- cgit From 9522751cded17c231acd1cf92bc21b3da1b07f38 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 04:49:40 -0800 Subject: ocfs2: Handle kuids and kgids in acl/xattr conversions. Explicitly deal with the different kinds of acls because they need different conversions. Cc: Mark Fasheh Cc: Joel Becker Signed-off-by: "Eric W. Biederman" --- fs/ocfs2/acl.c | 31 +++++++++++++++++++++++++++++-- 1 file changed, 29 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/acl.c b/fs/ocfs2/acl.c index 260b16281fc3..8a404576fb26 100644 --- a/fs/ocfs2/acl.c +++ b/fs/ocfs2/acl.c @@ -65,7 +65,20 @@ static struct posix_acl *ocfs2_acl_from_xattr(const void *value, size_t size) acl->a_entries[n].e_tag = le16_to_cpu(entry->e_tag); acl->a_entries[n].e_perm = le16_to_cpu(entry->e_perm); - acl->a_entries[n].e_id = le32_to_cpu(entry->e_id); + switch(acl->a_entries[n].e_tag) { + case ACL_USER: + acl->a_entries[n].e_uid = + make_kuid(&init_user_ns, + le32_to_cpu(entry->e_id)); + break; + case ACL_GROUP: + acl->a_entries[n].e_gid = + make_kgid(&init_user_ns, + le32_to_cpu(entry->e_id)); + break; + default: + break; + } value += sizeof(struct posix_acl_entry); } @@ -91,7 +104,21 @@ static void *ocfs2_acl_to_xattr(const struct posix_acl *acl, size_t *size) for (n = 0; n < acl->a_count; n++, entry++) { entry->e_tag = cpu_to_le16(acl->a_entries[n].e_tag); entry->e_perm = cpu_to_le16(acl->a_entries[n].e_perm); - entry->e_id = cpu_to_le32(acl->a_entries[n].e_id); + switch(acl->a_entries[n].e_tag) { + case ACL_USER: + entry->e_id = cpu_to_le32( + from_kuid(&init_user_ns, + acl->a_entries[n].e_uid)); + break; + case ACL_GROUP: + entry->e_id = cpu_to_le32( + from_kgid(&init_user_ns, + acl->a_entries[n].e_gid)); + break; + default: + entry->e_id = cpu_to_le32(ACL_UNDEFINED_ID); + break; + } } return ocfs2_acl; } -- cgit From 03ab30f73dbf2f4f719d2c0c2acef81bf0445eb7 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 16:59:23 -0800 Subject: ocfs2: convert between kuids and kgids and DLM locks Convert between uid and gids stored in the on the wire format of dlm locks aka struct ocfs2_meta_lvb and kuids and kgids stored in inode->i_uid and inode->i_gid. Cc: Mark Fasheh Cc: Joel Becker Signed-off-by: "Eric W. Biederman" --- fs/ocfs2/dlmglue.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 4f7795fb5fc0..f99af1cb849c 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -2045,8 +2045,8 @@ static void __ocfs2_stuff_meta_lvb(struct inode *inode) lvb->lvb_version = OCFS2_LVB_VERSION; lvb->lvb_isize = cpu_to_be64(i_size_read(inode)); lvb->lvb_iclusters = cpu_to_be32(oi->ip_clusters); - lvb->lvb_iuid = cpu_to_be32(inode->i_uid); - lvb->lvb_igid = cpu_to_be32(inode->i_gid); + lvb->lvb_iuid = cpu_to_be32(i_uid_read(inode)); + lvb->lvb_igid = cpu_to_be32(i_gid_read(inode)); lvb->lvb_imode = cpu_to_be16(inode->i_mode); lvb->lvb_inlink = cpu_to_be16(inode->i_nlink); lvb->lvb_iatime_packed = @@ -2095,8 +2095,8 @@ static void ocfs2_refresh_inode_from_lvb(struct inode *inode) else inode->i_blocks = ocfs2_inode_sector_count(inode); - inode->i_uid = be32_to_cpu(lvb->lvb_iuid); - inode->i_gid = be32_to_cpu(lvb->lvb_igid); + i_uid_write(inode, be32_to_cpu(lvb->lvb_iuid)); + i_gid_write(inode, be32_to_cpu(lvb->lvb_igid)); inode->i_mode = be16_to_cpu(lvb->lvb_imode); set_nlink(inode, be16_to_cpu(lvb->lvb_inlink)); ocfs2_unpack_timespec(&inode->i_atime, -- cgit From 2c03417627c5edaeeae1e8c5da6e8fd6f8c720d3 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 17:27:11 -0800 Subject: ocfs2: Convert uid and gids between in core and on disk inodes Cc: Mark Fasheh Cc: Joel Becker Signed-off-by: "Eric W. Biederman" --- fs/ocfs2/inode.c | 12 ++++++------ fs/ocfs2/namei.c | 4 ++-- 2 files changed, 8 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index d89e08a81eda..f87f9bd1edff 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -269,8 +269,8 @@ void ocfs2_populate_inode(struct inode *inode, struct ocfs2_dinode *fe, inode->i_generation = le32_to_cpu(fe->i_generation); inode->i_rdev = huge_decode_dev(le64_to_cpu(fe->id1.dev1.i_rdev)); inode->i_mode = le16_to_cpu(fe->i_mode); - inode->i_uid = le32_to_cpu(fe->i_uid); - inode->i_gid = le32_to_cpu(fe->i_gid); + i_uid_write(inode, le32_to_cpu(fe->i_uid)); + i_gid_write(inode, le32_to_cpu(fe->i_gid)); /* Fast symlinks will have i_size but no allocated clusters. */ if (S_ISLNK(inode->i_mode) && !fe->i_clusters) { @@ -1259,8 +1259,8 @@ int ocfs2_mark_inode_dirty(handle_t *handle, fe->i_size = cpu_to_le64(i_size_read(inode)); ocfs2_set_links_count(fe, inode->i_nlink); - fe->i_uid = cpu_to_le32(inode->i_uid); - fe->i_gid = cpu_to_le32(inode->i_gid); + fe->i_uid = cpu_to_le32(i_uid_read(inode)); + fe->i_gid = cpu_to_le32(i_gid_read(inode)); fe->i_mode = cpu_to_le16(inode->i_mode); fe->i_atime = cpu_to_le64(inode->i_atime.tv_sec); fe->i_atime_nsec = cpu_to_le32(inode->i_atime.tv_nsec); @@ -1290,8 +1290,8 @@ void ocfs2_refresh_inode(struct inode *inode, ocfs2_set_inode_flags(inode); i_size_write(inode, le64_to_cpu(fe->i_size)); set_nlink(inode, ocfs2_read_links_count(fe)); - inode->i_uid = le32_to_cpu(fe->i_uid); - inode->i_gid = le32_to_cpu(fe->i_gid); + i_uid_write(inode, le32_to_cpu(fe->i_uid)); + i_gid_write(inode, le32_to_cpu(fe->i_gid)); inode->i_mode = le16_to_cpu(fe->i_mode); if (S_ISLNK(inode->i_mode) && le32_to_cpu(fe->i_clusters) == 0) inode->i_blocks = 0; diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index f1fd0741162b..04ee1b57c243 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -512,8 +512,8 @@ static int __ocfs2_mknod_locked(struct inode *dir, fe->i_suballoc_loc = cpu_to_le64(suballoc_loc); fe->i_suballoc_bit = cpu_to_le16(suballoc_bit); fe->i_suballoc_slot = cpu_to_le16(inode_ac->ac_alloc_slot); - fe->i_uid = cpu_to_le32(inode->i_uid); - fe->i_gid = cpu_to_le32(inode->i_gid); + fe->i_uid = cpu_to_le32(i_uid_read(inode)); + fe->i_gid = cpu_to_le32(i_gid_read(inode)); fe->i_mode = cpu_to_le16(inode->i_mode); if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) fe->id1.dev1.i_rdev = cpu_to_le64(huge_encode_dev(dev)); -- cgit From ba6135609c2b56851e37e1d89ddbdbae4e774a71 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 17:33:53 -0800 Subject: ocfs2: For tracing report the uid and gid values in the initial user namespace Cc: Mark Fasheh Cc: Joel Becker Signed-off-by: "Eric W. Biederman" --- fs/ocfs2/file.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 37d313ede159..8ee93322db2b 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1116,7 +1116,8 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr) (unsigned long long)OCFS2_I(inode)->ip_blkno, dentry->d_name.len, dentry->d_name.name, attr->ia_valid, attr->ia_mode, - attr->ia_uid, attr->ia_gid); + from_kuid(&init_user_ns, attr->ia_uid), + from_kgid(&init_user_ns, attr->ia_gid)); /* ensuring we don't even attempt to truncate a symlink */ if (S_ISLNK(inode->i_mode)) -- cgit From 488c8ef033c6409cd8d23bcd04eed2f56301836d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 17:35:29 -0800 Subject: ocfs2: Compare kuids and kgids using uid_eq and gid_eq Cc: Mark Fasheh Cc: Joel Becker Signed-off-by: "Eric W. Biederman" --- fs/ocfs2/file.c | 8 ++++---- fs/ocfs2/refcounttree.c | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 8ee93322db2b..0a2924a2d9e6 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1175,14 +1175,14 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr) } } - if ((attr->ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || - (attr->ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { + if ((attr->ia_valid & ATTR_UID && !uid_eq(attr->ia_uid, inode->i_uid)) || + (attr->ia_valid & ATTR_GID && !gid_eq(attr->ia_gid, inode->i_gid))) { /* * Gather pointers to quota structures so that allocation / * freeing of quota structures happens here and not inside * dquot_transfer() where we have problems with lock ordering */ - if (attr->ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid + if (attr->ia_valid & ATTR_UID && !uid_eq(attr->ia_uid, inode->i_uid) && OCFS2_HAS_RO_COMPAT_FEATURE(sb, OCFS2_FEATURE_RO_COMPAT_USRQUOTA)) { transfer_to[USRQUOTA] = dqget(sb, make_kqid_uid(attr->ia_uid)); @@ -1191,7 +1191,7 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr) goto bail_unlock; } } - if (attr->ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid + if (attr->ia_valid & ATTR_GID && !gid_eq(attr->ia_gid, inode->i_gid) && OCFS2_HAS_RO_COMPAT_FEATURE(sb, OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)) { transfer_to[GRPQUOTA] = dqget(sb, make_kqid_gid(attr->ia_gid)); diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index 30a055049e16..934a4ac3e7fc 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -4407,7 +4407,7 @@ static int ocfs2_vfs_reflink(struct dentry *old_dentry, struct inode *dir, * rights to do so. */ if (preserve) { - if ((current_fsuid() != inode->i_uid) && !capable(CAP_CHOWN)) + if (!uid_eq(current_fsuid(), inode->i_uid) && !capable(CAP_CHOWN)) return -EPERM; if (!in_group_p(inode->i_gid) && !capable(CAP_CHOWN)) return -EPERM; -- cgit From 393551e9898136513007b1e88a25bd4dcdb0d759 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 02:09:03 -0800 Subject: gfs2: Remove improper checks in gfs2_set_dqblk. In set_dqblk it is an error to look at fdq->d_id or fdq->d_flags. Userspace quota applications do not set these fields when calling quotactl(Q_XSETQLIM,...), and the kernel does not set those fields when quota_setquota calls set_dqblk. gfs2 never looks at fdq->d_id or fdq->d_flags after checking to see if they match the id and type supplied to set_dqblk. No other linux filesystem in set_dqblk looks at either fdq->d_id or fdq->d_flags. Therefore remove these bogus checks from gfs2 and allow normal quota setting applications to work. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/quota.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index ae55e248c3b7..e4f6ccf3da64 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -1543,13 +1543,9 @@ static int gfs2_set_dqblk(struct super_block *sb, struct kqid qid, switch(qid.type) { case USRQUOTA: type = QUOTA_USER; - if (fdq->d_flags != FS_USER_QUOTA) - return -EINVAL; break; case GRPQUOTA: type = QUOTA_GROUP; - if (fdq->d_flags != FS_GROUP_QUOTA) - return -EINVAL; break; default: return -EINVAL; @@ -1557,8 +1553,6 @@ static int gfs2_set_dqblk(struct super_block *sb, struct kqid qid, if (fdq->d_fieldmask & ~GFS2_FIELDMASK) return -EINVAL; - if (fdq->d_id != from_kqid(&init_user_ns, qid)) - return -EINVAL; error = qd_get(sdp, type, from_kqid(&init_user_ns, qid), &qd); if (error) -- cgit From f4108a607f75b073423eed229ee4f95e5fc10631 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 17:49:26 -0800 Subject: gfs2: Split NO_QUOTA_CHANGE inot NO_UID_QUTOA_CHANGE and NO_GID_QUTOA_CHANGE Split NO_QUOTA_CHANGE into NO_UID_QUTOA_CHANGE and NO_GID_QUTOA_CHANGE so the constants may be well typed. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/bmap.c | 2 +- fs/gfs2/dir.c | 2 +- fs/gfs2/inode.c | 10 +++++----- fs/gfs2/quota.c | 4 ++-- fs/gfs2/quota.h | 5 +++-- fs/gfs2/super.c | 2 +- fs/gfs2/xattr.c | 4 ++-- 7 files changed, 15 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index a68e91bcef3d..10c54e3c2e72 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -1098,7 +1098,7 @@ static int trunc_dealloc(struct gfs2_inode *ip, u64 size) if (error) return error; - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) return error; diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 9a35670fdc38..6d3e3e2cabf8 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -1849,7 +1849,7 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, if (!ht) return -ENOMEM; - error = gfs2_quota_hold(dip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_quota_hold(dip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) goto out; diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 2b6f5698ef18..aa1050287f3c 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -548,7 +548,7 @@ static int link_dinode(struct gfs2_inode *dip, const struct qstr *name, if (error) return error; - error = gfs2_quota_lock(dip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_quota_lock(dip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) goto fail; @@ -1589,15 +1589,15 @@ static int setattr_chown(struct inode *inode, struct iattr *attr) ngid = attr->ia_gid; if (!(attr->ia_valid & ATTR_UID) || ouid == nuid) - ouid = nuid = NO_QUOTA_CHANGE; + ouid = nuid = NO_UID_QUOTA_CHANGE; if (!(attr->ia_valid & ATTR_GID) || ogid == ngid) - ogid = ngid = NO_QUOTA_CHANGE; + ogid = ngid = NO_GID_QUOTA_CHANGE; error = gfs2_quota_lock(ip, nuid, ngid); if (error) return error; - if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { + if (ouid != NO_UID_QUOTA_CHANGE || ogid != NO_GID_QUOTA_CHANGE) { error = gfs2_quota_check(ip, nuid, ngid); if (error) goto out_gunlock_q; @@ -1611,7 +1611,7 @@ static int setattr_chown(struct inode *inode, struct iattr *attr) if (error) goto out_end_trans; - if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { + if (ouid != NO_UID_QUOTA_CHANGE || ogid != NO_GID_QUOTA_CHANGE) { u64 blocks = gfs2_get_inode_blocks(&ip->i_inode); gfs2_quota_change(ip, -blocks, ouid, ogid); gfs2_quota_change(ip, blocks, nuid, ngid); diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index e4f6ccf3da64..dbfacaa8b6fb 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -524,7 +524,7 @@ int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid) ip->i_res->rs_qa_qd_num++; qd++; - if (uid != NO_QUOTA_CHANGE && uid != ip->i_inode.i_uid) { + if (uid != NO_UID_QUOTA_CHANGE && uid != ip->i_inode.i_uid) { error = qdsb_get(sdp, QUOTA_USER, uid, qd); if (error) goto out; @@ -532,7 +532,7 @@ int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid) qd++; } - if (gid != NO_QUOTA_CHANGE && gid != ip->i_inode.i_gid) { + if (gid != NO_GID_QUOTA_CHANGE && gid != ip->i_inode.i_gid) { error = qdsb_get(sdp, QUOTA_GROUP, gid, qd); if (error) goto out; diff --git a/fs/gfs2/quota.h b/fs/gfs2/quota.h index f25d98b87904..7f67323a6a72 100644 --- a/fs/gfs2/quota.h +++ b/fs/gfs2/quota.h @@ -14,7 +14,8 @@ struct gfs2_inode; struct gfs2_sbd; struct shrink_control; -#define NO_QUOTA_CHANGE ((u32)-1) +#define NO_UID_QUOTA_CHANGE INVALID_UID +#define NO_GID_QUOTA_CHANGE INVALID_GID extern int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid); extern void gfs2_quota_unhold(struct gfs2_inode *ip); @@ -41,7 +42,7 @@ static inline int gfs2_quota_lock_check(struct gfs2_inode *ip) int ret; if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return 0; - ret = gfs2_quota_lock(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + ret = gfs2_quota_lock(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (ret) return ret; if (sdp->sd_args.ar_quota != GFS2_QUOTA_ON) diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index d6488674d916..7cc5a6054547 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -1429,7 +1429,7 @@ static int gfs2_dinode_dealloc(struct gfs2_inode *ip) if (error) return error; - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) return error; diff --git a/fs/gfs2/xattr.c b/fs/gfs2/xattr.c index 76c144b3c9bb..80c25ae79048 100644 --- a/fs/gfs2/xattr.c +++ b/fs/gfs2/xattr.c @@ -331,7 +331,7 @@ static int ea_remove_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, if (error) return error; - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) goto out_alloc; @@ -1461,7 +1461,7 @@ int gfs2_ea_dealloc(struct gfs2_inode *ip) if (error) return error; - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + error = gfs2_quota_hold(ip, NO_UID_QUOTA_CHANGE, NO_GID_QUOTA_CHANGE); if (error) return error; -- cgit From 558e85289fca3d3397882442d1a695936c4f2662 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 18:15:33 -0800 Subject: gfs2: Report quotas in the caller's user namespace. When a quota is queried return the uid or the gid in the mapped into the caller's user namespace. In addition perform the munged version of the mapping so that instead of -1 a value that does not map is reported as the overflowuid or the overflowgid. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/quota.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index dbfacaa8b6fb..f8279eed56e1 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -1509,7 +1509,7 @@ static int gfs2_get_dqblk(struct super_block *sb, struct kqid qid, qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; fdq->d_version = FS_DQUOT_VERSION; fdq->d_flags = (type == QUOTA_USER) ? FS_USER_QUOTA : FS_GROUP_QUOTA; - fdq->d_id = from_kqid(&init_user_ns, qid); + fdq->d_id = from_kqid_munged(current_user_ns(), qid); fdq->d_blk_hardlimit = be64_to_cpu(qlvb->qb_limit) << sdp->sd_fsb2bb_shift; fdq->d_blk_softlimit = be64_to_cpu(qlvb->qb_warn) << sdp->sd_fsb2bb_shift; fdq->d_bcount = be64_to_cpu(qlvb->qb_value) << sdp->sd_fsb2bb_shift; -- cgit From 2f6c9896f71e6b6c1c565ea76dd9f5e89579c120 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 18:33:38 -0800 Subject: gfs2: Introduce qd2index Both qd_alloc and qd2offset perform the exact same computation to get an index from a gfs2_quota_data. Make life a little simpler and factor out this index computation. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/quota.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index f8279eed56e1..0e7c982377a1 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -120,11 +120,17 @@ out: return (atomic_read(&qd_lru_count) * sysctl_vfs_cache_pressure) / 100; } +static u64 qd2index(struct gfs2_quota_data *qd) +{ + return (2 * (u64)qd->qd_id) + + test_bit(QDF_USER, &qd->qd_flags) ? 0 : 1; +} + static u64 qd2offset(struct gfs2_quota_data *qd) { u64 offset; - offset = 2 * (u64)qd->qd_id + !test_bit(QDF_USER, &qd->qd_flags); + offset = qd2index(qd); offset *= sizeof(struct gfs2_quota); return offset; @@ -147,7 +153,7 @@ static int qd_alloc(struct gfs2_sbd *sdp, int user, u32 id, qd->qd_slot = -1; INIT_LIST_HEAD(&qd->qd_reclaim); - error = gfs2_glock_get(sdp, 2 * (u64)id + !user, + error = gfs2_glock_get(sdp, qd2index(qd), &gfs2_quota_glops, CREATE, &qd->qd_gl); if (error) goto fail; -- cgit From e08d8d7f201dc1e64f5d9d5aa2cd4f37aecaaab4 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 19:25:50 -0800 Subject: gfs2: Modify struct gfs2_quota_change_host to use struct kqid Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/quota.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 0e7c982377a1..02913e95491c 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -71,7 +71,7 @@ struct gfs2_quota_change_host { u64 qc_change; u32 qc_flags; /* GFS2_QCF_... */ - u32 qc_id; + struct kqid qc_id; }; static LIST_HEAD(qd_lru_list); @@ -1200,7 +1200,9 @@ static void gfs2_quota_change_in(struct gfs2_quota_change_host *qc, const void * qc->qc_change = be64_to_cpu(str->qc_change); qc->qc_flags = be32_to_cpu(str->qc_flags); - qc->qc_id = be32_to_cpu(str->qc_id); + qc->qc_id = make_kqid(&init_user_ns, + (qc->qc_flags & GFS2_QCF_USER)?USRQUOTA:GRPQUOTA, + be32_to_cpu(str->qc_id)); } int gfs2_quota_init(struct gfs2_sbd *sdp) @@ -1264,7 +1266,7 @@ int gfs2_quota_init(struct gfs2_sbd *sdp) continue; error = qd_alloc(sdp, (qc.qc_flags & GFS2_QCF_USER), - qc.qc_id, &qd); + from_kqid(&init_user_ns, qc.qc_id), &qd); if (error) { brelse(bh); goto fail; -- cgit From b59c8b6f9d1b1220e5ed72152f42a658bf739d90 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 19:35:56 -0800 Subject: gfs2: Modify qdsb_get to take a struct kqid Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/quota.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 02913e95491c..20762ae2a9c4 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -464,12 +464,13 @@ static void qd_unlock(struct gfs2_quota_data *qd) qd_put(qd); } -static int qdsb_get(struct gfs2_sbd *sdp, int user, u32 id, +static int qdsb_get(struct gfs2_sbd *sdp, struct kqid qid, struct gfs2_quota_data **qdp) { int error; - error = qd_get(sdp, user, id, qdp); + error = qd_get(sdp, qid.type == USRQUOTA ? QUOTA_USER : QUOTA_GROUP, + from_kqid(&init_user_ns, qid), qdp); if (error) return error; @@ -518,20 +519,20 @@ int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid) if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return 0; - error = qdsb_get(sdp, QUOTA_USER, ip->i_inode.i_uid, qd); + error = qdsb_get(sdp, make_kqid_uid(ip->i_inode.i_uid), qd); if (error) goto out; ip->i_res->rs_qa_qd_num++; qd++; - error = qdsb_get(sdp, QUOTA_GROUP, ip->i_inode.i_gid, qd); + error = qdsb_get(sdp, make_kqid_gid(ip->i_inode.i_gid), qd); if (error) goto out; ip->i_res->rs_qa_qd_num++; qd++; if (uid != NO_UID_QUOTA_CHANGE && uid != ip->i_inode.i_uid) { - error = qdsb_get(sdp, QUOTA_USER, uid, qd); + error = qdsb_get(sdp, make_kqid_uid(uid), qd); if (error) goto out; ip->i_res->rs_qa_qd_num++; @@ -539,7 +540,7 @@ int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid) } if (gid != NO_GID_QUOTA_CHANGE && gid != ip->i_inode.i_gid) { - error = qdsb_get(sdp, QUOTA_GROUP, gid, qd); + error = qdsb_get(sdp, make_kqid_gid(gid), qd); if (error) goto out; ip->i_res->rs_qa_qd_num++; -- cgit From ed87dabcc3fc0a5040f95dd3f7206cffebca5c79 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 19:42:40 -0800 Subject: gfs2: Convert gfs2_quota_refresh to take a kqid - In quota_refresh_user_store convert the user supplied uid into a kqid and pass it to gfs2_quota_refresh. - In quota_refresh_group_store convert the user supplied gid into a kqid and pass it to gfs2_quota_refresh. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/quota.c | 5 +++-- fs/gfs2/quota.h | 2 +- fs/gfs2/sys.c | 14 ++++++++++++-- 3 files changed, 16 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 20762ae2a9c4..47315c091a09 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -1177,13 +1177,14 @@ static int gfs2_quota_sync_timeo(struct super_block *sb, int type) return gfs2_quota_sync(sb, type); } -int gfs2_quota_refresh(struct gfs2_sbd *sdp, int user, u32 id) +int gfs2_quota_refresh(struct gfs2_sbd *sdp, struct kqid qid) { struct gfs2_quota_data *qd; struct gfs2_holder q_gh; int error; - error = qd_get(sdp, user, id, &qd); + error = qd_get(sdp, qid.type == USRQUOTA ? QUOTA_USER : QUOTA_GROUP, + from_kqid(&init_user_ns, qid), &qd); if (error) return error; diff --git a/fs/gfs2/quota.h b/fs/gfs2/quota.h index 7f67323a6a72..bef805de8491 100644 --- a/fs/gfs2/quota.h +++ b/fs/gfs2/quota.h @@ -28,7 +28,7 @@ extern void gfs2_quota_change(struct gfs2_inode *ip, s64 change, u32 uid, u32 gid); extern int gfs2_quota_sync(struct super_block *sb, int type); -extern int gfs2_quota_refresh(struct gfs2_sbd *sdp, int user, u32 id); +extern int gfs2_quota_refresh(struct gfs2_sbd *sdp, struct kqid qid); extern int gfs2_quota_init(struct gfs2_sbd *sdp); extern void gfs2_quota_cleanup(struct gfs2_sbd *sdp); diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 8056b7b7238e..e6d8d482422f 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -175,6 +175,7 @@ static ssize_t quota_sync_store(struct gfs2_sbd *sdp, const char *buf, static ssize_t quota_refresh_user_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { + struct kqid qid; int error; u32 id; @@ -183,13 +184,18 @@ static ssize_t quota_refresh_user_store(struct gfs2_sbd *sdp, const char *buf, id = simple_strtoul(buf, NULL, 0); - error = gfs2_quota_refresh(sdp, 1, id); + qid = make_kqid(current_user_ns(), USRQUOTA, id); + if (!qid_valid(qid)) + return -EINVAL; + + error = gfs2_quota_refresh(sdp, qid); return error ? error : len; } static ssize_t quota_refresh_group_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { + struct kqid qid; int error; u32 id; @@ -198,7 +204,11 @@ static ssize_t quota_refresh_group_store(struct gfs2_sbd *sdp, const char *buf, id = simple_strtoul(buf, NULL, 0); - error = gfs2_quota_refresh(sdp, 0, id); + qid = make_kqid(current_user_ns(), GRPQUOTA, id); + if (!qid_valid(qid)) + return -EINVAL; + + error = gfs2_quota_refresh(sdp, qid); return error ? error : len; } -- cgit From 05e0a60d8025e280e56b3fa36ea8facc7c1c65c2 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 19:52:08 -0800 Subject: gfs2: Store qd_id in struct gfs2_quota_data as a struct kqid - Change qd_id in struct gfs2_qutoa_data to struct kqid. - Remove the now unnecessary QDF_USER bit field in qd_flags. - Propopoage this change through the code generally making things simpler along the way. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/incore.h | 3 +-- fs/gfs2/quota.c | 69 ++++++++++++++++++++------------------------------------ 2 files changed, 26 insertions(+), 46 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index c373a24fedd9..5b298bdab90c 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -391,7 +391,6 @@ struct gfs2_revoke_replay { }; enum { - QDF_USER = 0, QDF_CHANGE = 1, QDF_LOCKED = 2, QDF_REFRESH = 3, @@ -403,7 +402,7 @@ struct gfs2_quota_data { atomic_t qd_count; - u32 qd_id; + struct kqid qd_id; unsigned long qd_flags; /* QDF_... */ s64 qd_change; diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 47315c091a09..8cb4d10cb16a 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -122,8 +122,9 @@ out: static u64 qd2index(struct gfs2_quota_data *qd) { - return (2 * (u64)qd->qd_id) + - test_bit(QDF_USER, &qd->qd_flags) ? 0 : 1; + struct kqid qid = qd->qd_id; + return (2 * (u64)from_kqid(&init_user_ns, qid)) + + (qid.type == USRQUOTA) ? 0 : 1; } static u64 qd2offset(struct gfs2_quota_data *qd) @@ -136,7 +137,7 @@ static u64 qd2offset(struct gfs2_quota_data *qd) return offset; } -static int qd_alloc(struct gfs2_sbd *sdp, int user, u32 id, +static int qd_alloc(struct gfs2_sbd *sdp, struct kqid qid, struct gfs2_quota_data **qdp) { struct gfs2_quota_data *qd; @@ -147,9 +148,7 @@ static int qd_alloc(struct gfs2_sbd *sdp, int user, u32 id, return -ENOMEM; atomic_set(&qd->qd_count, 1); - qd->qd_id = id; - if (user) - set_bit(QDF_USER, &qd->qd_flags); + qd->qd_id = qid; qd->qd_slot = -1; INIT_LIST_HEAD(&qd->qd_reclaim); @@ -167,7 +166,7 @@ fail: return error; } -static int qd_get(struct gfs2_sbd *sdp, int user, u32 id, +static int qd_get(struct gfs2_sbd *sdp, struct kqid qid, struct gfs2_quota_data **qdp) { struct gfs2_quota_data *qd = NULL, *new_qd = NULL; @@ -179,8 +178,7 @@ static int qd_get(struct gfs2_sbd *sdp, int user, u32 id, found = 0; spin_lock(&qd_lru_lock); list_for_each_entry(qd, &sdp->sd_quota_list, qd_list) { - if (qd->qd_id == id && - !test_bit(QDF_USER, &qd->qd_flags) == !user) { + if (qid_eq(qd->qd_id, qid)) { if (!atomic_read(&qd->qd_count) && !list_empty(&qd->qd_reclaim)) { /* Remove it from reclaim list */ @@ -214,7 +212,7 @@ static int qd_get(struct gfs2_sbd *sdp, int user, u32 id, return 0; } - error = qd_alloc(sdp, user, id, &new_qd); + error = qd_alloc(sdp, qid, &new_qd); if (error) return error; } @@ -469,8 +467,7 @@ static int qdsb_get(struct gfs2_sbd *sdp, struct kqid qid, { int error; - error = qd_get(sdp, qid.type == USRQUOTA ? QUOTA_USER : QUOTA_GROUP, - from_kqid(&init_user_ns, qid), qdp); + error = qd_get(sdp, qid, qdp); if (error) return error; @@ -574,18 +571,10 @@ static int sort_qd(const void *a, const void *b) const struct gfs2_quota_data *qd_a = *(const struct gfs2_quota_data **)a; const struct gfs2_quota_data *qd_b = *(const struct gfs2_quota_data **)b; - if (!test_bit(QDF_USER, &qd_a->qd_flags) != - !test_bit(QDF_USER, &qd_b->qd_flags)) { - if (test_bit(QDF_USER, &qd_a->qd_flags)) - return -1; - else - return 1; - } - if (qd_a->qd_id < qd_b->qd_id) + if (qid_lt(qd_a->qd_id, qd_b->qd_id)) return -1; - if (qd_a->qd_id > qd_b->qd_id) + if (qid_lt(qd_b->qd_id, qd_a->qd_id)) return 1; - return 0; } @@ -602,9 +591,9 @@ static void do_qc(struct gfs2_quota_data *qd, s64 change) if (!test_bit(QDF_CHANGE, &qd->qd_flags)) { qc->qc_change = 0; qc->qc_flags = 0; - if (test_bit(QDF_USER, &qd->qd_flags)) + if (qd->qd_id.type == USRQUOTA) qc->qc_flags = cpu_to_be32(GFS2_QCF_USER); - qc->qc_id = cpu_to_be32(qd->qd_id); + qc->qc_id = cpu_to_be32(from_kqid(&init_user_ns, qd->qd_id)); } x = be64_to_cpu(qc->qc_change) + change; @@ -1047,8 +1036,8 @@ static int print_message(struct gfs2_quota_data *qd, char *type) printk(KERN_INFO "GFS2: fsid=%s: quota %s for %s %u\n", sdp->sd_fsname, type, - (test_bit(QDF_USER, &qd->qd_flags)) ? "user" : "group", - qd->qd_id); + (qd->qd_id.type == USRQUOTA) ? "user" : "group", + from_kqid(&init_user_ns, qd->qd_id)); return 0; } @@ -1070,8 +1059,8 @@ int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) for (x = 0; x < ip->i_res->rs_qa_qd_num; x++) { qd = ip->i_res->rs_qa_qd[x]; - if (!((qd->qd_id == uid && test_bit(QDF_USER, &qd->qd_flags)) || - (qd->qd_id == gid && !test_bit(QDF_USER, &qd->qd_flags)))) + if (!(qid_eq(qd->qd_id, make_kqid_uid(uid)) || + qid_eq(qd->qd_id, make_kqid_gid(gid)))) continue; value = (s64)be64_to_cpu(qd->qd_qb.qb_value); @@ -1081,10 +1070,7 @@ int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) if (be64_to_cpu(qd->qd_qb.qb_limit) && (s64)be64_to_cpu(qd->qd_qb.qb_limit) < value) { print_message(qd, "exceeded"); - quota_send_warning(make_kqid(&init_user_ns, - test_bit(QDF_USER, &qd->qd_flags) ? - USRQUOTA : GRPQUOTA, - qd->qd_id), + quota_send_warning(qd->qd_id, sdp->sd_vfs->s_dev, QUOTA_NL_BHARDWARN); error = -EDQUOT; @@ -1094,10 +1080,7 @@ int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) time_after_eq(jiffies, qd->qd_last_warn + gfs2_tune_get(sdp, gt_quota_warn_period) * HZ)) { - quota_send_warning(make_kqid(&init_user_ns, - test_bit(QDF_USER, &qd->qd_flags) ? - USRQUOTA : GRPQUOTA, - qd->qd_id), + quota_send_warning(qd->qd_id, sdp->sd_vfs->s_dev, QUOTA_NL_BSOFTWARN); error = print_message(qd, "warning"); qd->qd_last_warn = jiffies; @@ -1121,8 +1104,8 @@ void gfs2_quota_change(struct gfs2_inode *ip, s64 change, for (x = 0; x < ip->i_res->rs_qa_qd_num; x++) { qd = ip->i_res->rs_qa_qd[x]; - if ((qd->qd_id == uid && test_bit(QDF_USER, &qd->qd_flags)) || - (qd->qd_id == gid && !test_bit(QDF_USER, &qd->qd_flags))) { + if (qid_eq(qd->qd_id, make_kqid_uid(uid)) || + qid_eq(qd->qd_id, make_kqid_gid(gid))) { do_qc(qd, change); } } @@ -1183,8 +1166,7 @@ int gfs2_quota_refresh(struct gfs2_sbd *sdp, struct kqid qid) struct gfs2_holder q_gh; int error; - error = qd_get(sdp, qid.type == USRQUOTA ? QUOTA_USER : QUOTA_GROUP, - from_kqid(&init_user_ns, qid), &qd); + error = qd_get(sdp, qid, &qd); if (error) return error; @@ -1267,8 +1249,7 @@ int gfs2_quota_init(struct gfs2_sbd *sdp) if (!qc.qc_change) continue; - error = qd_alloc(sdp, (qc.qc_flags & GFS2_QCF_USER), - from_kqid(&init_user_ns, qc.qc_id), &qd); + error = qd_alloc(sdp, qc.qc_id, &qd); if (error) { brelse(bh); goto fail; @@ -1509,7 +1490,7 @@ static int gfs2_get_dqblk(struct super_block *sb, struct kqid qid, else return -EINVAL; - error = qd_get(sdp, type, from_kqid(&init_user_ns, qid), &qd); + error = qd_get(sdp, qid, &qd); if (error) return error; error = do_glock(qd, FORCE, &q_gh); @@ -1564,7 +1545,7 @@ static int gfs2_set_dqblk(struct super_block *sb, struct kqid qid, if (fdq->d_fieldmask & ~GFS2_FIELDMASK) return -EINVAL; - error = qd_get(sdp, type, from_kqid(&init_user_ns, qid), &qd); + error = qd_get(sdp, qid, &qd); if (error) return error; -- cgit From 236c64e4b79b78059ec3e17362d8f02f6dc06f26 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 20:09:30 -0800 Subject: gfs2: Remove the QUOTA_USER and QUOTA_GROUP defines Remove the QUOTA_USER and QUOTA_GRUP defines. Remove the last vestigal users of QUOTA_USER and QUOTA_GROUP. Now that struct kqid is used throughout the gfs2 quota code the need there is to use QUOTA_USER and QUOTA_GROUP and the defines are just extraneous and confusing. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/quota.c | 25 +++++-------------------- 1 file changed, 5 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 8cb4d10cb16a..0bbb0407fd96 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -65,9 +65,6 @@ #include "inode.h" #include "util.h" -#define QUOTA_USER 1 -#define QUOTA_GROUP 0 - struct gfs2_quota_change_host { u64 qc_change; u32 qc_flags; /* GFS2_QCF_... */ @@ -1476,18 +1473,14 @@ static int gfs2_get_dqblk(struct super_block *sb, struct kqid qid, struct gfs2_quota_data *qd; struct gfs2_holder q_gh; int error; - int type; memset(fdq, 0, sizeof(struct fs_disk_quota)); if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return -ESRCH; /* Crazy XFS error code */ - if (qid.type == USRQUOTA) - type = QUOTA_USER; - else if (qid.type == GRPQUOTA) - type = QUOTA_GROUP; - else + if ((qid.type != USRQUOTA) && + (qid.type != GRPQUOTA)) return -EINVAL; error = qd_get(sdp, qid, &qd); @@ -1499,7 +1492,7 @@ static int gfs2_get_dqblk(struct super_block *sb, struct kqid qid, qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; fdq->d_version = FS_DQUOT_VERSION; - fdq->d_flags = (type == QUOTA_USER) ? FS_USER_QUOTA : FS_GROUP_QUOTA; + fdq->d_flags = (qid.type == USRQUOTA) ? FS_USER_QUOTA : FS_GROUP_QUOTA; fdq->d_id = from_kqid_munged(current_user_ns(), qid); fdq->d_blk_hardlimit = be64_to_cpu(qlvb->qb_limit) << sdp->sd_fsb2bb_shift; fdq->d_blk_softlimit = be64_to_cpu(qlvb->qb_warn) << sdp->sd_fsb2bb_shift; @@ -1526,21 +1519,13 @@ static int gfs2_set_dqblk(struct super_block *sb, struct kqid qid, int alloc_required; loff_t offset; int error; - int type; if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF) return -ESRCH; /* Crazy XFS error code */ - switch(qid.type) { - case USRQUOTA: - type = QUOTA_USER; - break; - case GRPQUOTA: - type = QUOTA_GROUP; - break; - default: + if ((qid.type != USRQUOTA) && + (qid.type != GRPQUOTA)) return -EINVAL; - } if (fdq->d_fieldmask & ~GFS2_FIELDMASK) return -EINVAL; -- cgit From 7c06b5d67225dc99ca81a33db3e055e08da857c3 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 20:27:54 -0800 Subject: gfs2: Use kuid_t and kgid_t types where appropriate. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/inode.c | 3 ++- fs/gfs2/quota.c | 8 ++++---- fs/gfs2/quota.h | 8 ++++---- 3 files changed, 10 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index aa1050287f3c..ce07ce4734f4 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -1580,7 +1580,8 @@ static int setattr_chown(struct inode *inode, struct iattr *attr) { struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); - u32 ouid, ogid, nuid, ngid; + kuid_t ouid, nuid; + kgid_t ogid, ngid; int error; ouid = inode->i_uid; diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 0bbb0407fd96..87f274039c4b 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -492,7 +492,7 @@ static void qdsb_put(struct gfs2_quota_data *qd) qd_put(qd); } -int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid) +int gfs2_quota_hold(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_quota_data **qd; @@ -918,7 +918,7 @@ fail: return error; } -int gfs2_quota_lock(struct gfs2_inode *ip, u32 uid, u32 gid) +int gfs2_quota_lock(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_quota_data *qd; @@ -1039,7 +1039,7 @@ static int print_message(struct gfs2_quota_data *qd, char *type) return 0; } -int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) +int gfs2_quota_check(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_quota_data *qd; @@ -1088,7 +1088,7 @@ int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid) } void gfs2_quota_change(struct gfs2_inode *ip, s64 change, - u32 uid, u32 gid) + kuid_t uid, kgid_t gid) { struct gfs2_quota_data *qd; unsigned int x; diff --git a/fs/gfs2/quota.h b/fs/gfs2/quota.h index bef805de8491..4f5e6e44ed83 100644 --- a/fs/gfs2/quota.h +++ b/fs/gfs2/quota.h @@ -17,15 +17,15 @@ struct shrink_control; #define NO_UID_QUOTA_CHANGE INVALID_UID #define NO_GID_QUOTA_CHANGE INVALID_GID -extern int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid); +extern int gfs2_quota_hold(struct gfs2_inode *ip, kuid_t uid, kgid_t gid); extern void gfs2_quota_unhold(struct gfs2_inode *ip); -extern int gfs2_quota_lock(struct gfs2_inode *ip, u32 uid, u32 gid); +extern int gfs2_quota_lock(struct gfs2_inode *ip, kuid_t uid, kgid_t gid); extern void gfs2_quota_unlock(struct gfs2_inode *ip); -extern int gfs2_quota_check(struct gfs2_inode *ip, u32 uid, u32 gid); +extern int gfs2_quota_check(struct gfs2_inode *ip, kuid_t uid, kgid_t gid); extern void gfs2_quota_change(struct gfs2_inode *ip, s64 change, - u32 uid, u32 gid); + kuid_t uid, kgid_t gid); extern int gfs2_quota_sync(struct super_block *sb, int type); extern int gfs2_quota_refresh(struct gfs2_sbd *sdp, struct kqid qid); -- cgit From 6b24c0d279eacfb499854d09ea7f2b69d1721a29 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 21:56:13 -0800 Subject: gfs2: Use uid_eq and gid_eq where appropriate Where kuid_t values are compared use uid_eq and where kgid_t values are compared use gid_eq. This is unfortunately necessary because of the type safety that keeps someone from accidentally mixing kuids and kgids with other types. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/acl.c | 2 +- fs/gfs2/inode.c | 19 +++++++++++-------- fs/gfs2/quota.c | 6 ++++-- 3 files changed, 16 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/acl.c b/fs/gfs2/acl.c index f850020ad906..f69ac0af5496 100644 --- a/fs/gfs2/acl.c +++ b/fs/gfs2/acl.c @@ -237,7 +237,7 @@ static int gfs2_xattr_system_set(struct dentry *dentry, const char *name, return -EINVAL; if (type == ACL_TYPE_DEFAULT && !S_ISDIR(inode->i_mode)) return value ? -EACCES : 0; - if ((current_fsuid() != inode->i_uid) && !capable(CAP_FOWNER)) + if (!uid_eq(current_fsuid(), inode->i_uid) && !capable(CAP_FOWNER)) return -EPERM; if (S_ISLNK(inode->i_mode)) return -EOPNOTSUPP; diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index ce07ce4734f4..bb7c754f46a2 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -368,10 +368,11 @@ static void munge_mode_uid_gid(const struct gfs2_inode *dip, struct inode *inode) { if (GFS2_SB(&dip->i_inode)->sd_args.ar_suiddir && - (dip->i_inode.i_mode & S_ISUID) && dip->i_inode.i_uid) { + (dip->i_inode.i_mode & S_ISUID) && + !uid_eq(dip->i_inode.i_uid, GLOBAL_ROOT_UID)) { if (S_ISDIR(inode->i_mode)) inode->i_mode |= S_ISUID; - else if (dip->i_inode.i_uid != current_fsuid()) + else if (!uid_eq(dip->i_inode.i_uid, current_fsuid())) inode->i_mode &= ~07111; inode->i_uid = dip->i_inode.i_uid; } else @@ -978,8 +979,8 @@ static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, return -EPERM; if ((dip->i_inode.i_mode & S_ISVTX) && - dip->i_inode.i_uid != current_fsuid() && - ip->i_inode.i_uid != current_fsuid() && !capable(CAP_FOWNER)) + !uid_eq(dip->i_inode.i_uid, current_fsuid()) && + !uid_eq(ip->i_inode.i_uid, current_fsuid()) && !capable(CAP_FOWNER)) return -EPERM; if (IS_APPEND(&dip->i_inode)) @@ -1589,16 +1590,17 @@ static int setattr_chown(struct inode *inode, struct iattr *attr) nuid = attr->ia_uid; ngid = attr->ia_gid; - if (!(attr->ia_valid & ATTR_UID) || ouid == nuid) + if (!(attr->ia_valid & ATTR_UID) || uid_eq(ouid, nuid)) ouid = nuid = NO_UID_QUOTA_CHANGE; - if (!(attr->ia_valid & ATTR_GID) || ogid == ngid) + if (!(attr->ia_valid & ATTR_GID) || gid_eq(ogid, ngid)) ogid = ngid = NO_GID_QUOTA_CHANGE; error = gfs2_quota_lock(ip, nuid, ngid); if (error) return error; - if (ouid != NO_UID_QUOTA_CHANGE || ogid != NO_GID_QUOTA_CHANGE) { + if (!uid_eq(ouid, NO_UID_QUOTA_CHANGE) || + !gid_eq(ogid, NO_GID_QUOTA_CHANGE)) { error = gfs2_quota_check(ip, nuid, ngid); if (error) goto out_gunlock_q; @@ -1612,7 +1614,8 @@ static int setattr_chown(struct inode *inode, struct iattr *attr) if (error) goto out_end_trans; - if (ouid != NO_UID_QUOTA_CHANGE || ogid != NO_GID_QUOTA_CHANGE) { + if (!uid_eq(ouid, NO_UID_QUOTA_CHANGE) || + !gid_eq(ogid, NO_GID_QUOTA_CHANGE)) { u64 blocks = gfs2_get_inode_blocks(&ip->i_inode); gfs2_quota_change(ip, -blocks, ouid, ogid); gfs2_quota_change(ip, blocks, nuid, ngid); diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 87f274039c4b..afd2e5d38e5a 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -525,7 +525,8 @@ int gfs2_quota_hold(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) ip->i_res->rs_qa_qd_num++; qd++; - if (uid != NO_UID_QUOTA_CHANGE && uid != ip->i_inode.i_uid) { + if (!uid_eq(uid, NO_UID_QUOTA_CHANGE) && + !uid_eq(uid, ip->i_inode.i_uid)) { error = qdsb_get(sdp, make_kqid_uid(uid), qd); if (error) goto out; @@ -533,7 +534,8 @@ int gfs2_quota_hold(struct gfs2_inode *ip, kuid_t uid, kgid_t gid) qd++; } - if (gid != NO_GID_QUOTA_CHANGE && gid != ip->i_inode.i_gid) { + if (!gid_eq(gid, NO_GID_QUOTA_CHANGE) && + !gid_eq(gid, ip->i_inode.i_gid)) { error = qdsb_get(sdp, make_kqid_gid(gid), qd); if (error) goto out; -- cgit From d0546426426ca96f0b815581c0a9ed0db28319b6 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 31 Jan 2013 22:08:10 -0800 Subject: gfs2: Convert uids and gids between dinodes and vfs inodes. When reading dinodes from the disk convert uids and gids into kuids and kgids to store in vfs data structures. When writing to dinodes to the disk convert kuids and kgids in the in memory structures into plain uids and gids. For now all on disk data structures are assumed to be stored in the initial user namespace. Cc: Steven Whitehouse Signed-off-by: "Eric W. Biederman" --- fs/gfs2/glops.c | 4 ++-- fs/gfs2/inode.c | 4 ++-- fs/gfs2/super.c | 4 ++-- 3 files changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index 78d4184ffc7d..444b6503ebc4 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -322,8 +322,8 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) break; }; - ip->i_inode.i_uid = be32_to_cpu(str->di_uid); - ip->i_inode.i_gid = be32_to_cpu(str->di_gid); + i_uid_write(&ip->i_inode, be32_to_cpu(str->di_uid)); + i_gid_write(&ip->i_inode, be32_to_cpu(str->di_gid)); gfs2_set_nlink(&ip->i_inode, be32_to_cpu(str->di_nlink)); i_size_write(&ip->i_inode, be64_to_cpu(str->di_size)); gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks)); diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index bb7c754f46a2..b862d114e155 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -456,8 +456,8 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_inode *ip, di->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino); di->di_num.no_addr = cpu_to_be64(ip->i_no_addr); di->di_mode = cpu_to_be32(ip->i_inode.i_mode); - di->di_uid = cpu_to_be32(ip->i_inode.i_uid); - di->di_gid = cpu_to_be32(ip->i_inode.i_gid); + di->di_uid = cpu_to_be32(i_uid_read(&ip->i_inode)); + di->di_gid = cpu_to_be32(i_gid_read(&ip->i_inode)); di->di_nlink = 0; di->di_size = cpu_to_be64(ip->i_inode.i_size); di->di_blocks = cpu_to_be64(1); diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 7cc5a6054547..a03425376500 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -721,8 +721,8 @@ void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) str->di_num.no_addr = cpu_to_be64(ip->i_no_addr); str->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino); str->di_mode = cpu_to_be32(ip->i_inode.i_mode); - str->di_uid = cpu_to_be32(ip->i_inode.i_uid); - str->di_gid = cpu_to_be32(ip->i_inode.i_gid); + str->di_uid = cpu_to_be32(i_uid_read(&ip->i_inode)); + str->di_gid = cpu_to_be32(i_gid_read(&ip->i_inode)); str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink); str->di_size = cpu_to_be64(i_size_read(&ip->i_inode)); str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); -- cgit From 1ac7fd8190b79c822631ed537186fb8b2d9e9b74 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 7 Feb 2012 16:28:28 -0800 Subject: ncpfs: Support interacting with multiple user namespaces ncpfs does not natively support uids and gids so this conversion was simply a matter of updating the the type of the mounteduid, the uid and the gid on the superblock. Fixing the ioctls that read them, updating the mount option parser and the mount option printer. Cc: Petr Vandrovec Acked-by: Serge Hallyn Signed-off-by: Eric W. Biederman --- fs/ncpfs/inode.c | 55 +++++++++++++++++++++++++++++++--------------------- fs/ncpfs/ioctl.c | 25 +++++++++++++----------- fs/ncpfs/ncp_fs_sb.h | 6 +++--- 3 files changed, 50 insertions(+), 36 deletions(-) (limited to 'fs') diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c index 1acdad7fcec7..e2be336d1c22 100644 --- a/fs/ncpfs/inode.c +++ b/fs/ncpfs/inode.c @@ -331,12 +331,15 @@ static int ncp_show_options(struct seq_file *seq, struct dentry *root) struct ncp_server *server = NCP_SBP(root->d_sb); unsigned int tmp; - if (server->m.uid != 0) - seq_printf(seq, ",uid=%u", server->m.uid); - if (server->m.gid != 0) - seq_printf(seq, ",gid=%u", server->m.gid); - if (server->m.mounted_uid != 0) - seq_printf(seq, ",owner=%u", server->m.mounted_uid); + if (!uid_eq(server->m.uid, GLOBAL_ROOT_UID)) + seq_printf(seq, ",uid=%u", + from_kuid_munged(&init_user_ns, server->m.uid)); + if (!gid_eq(server->m.gid, GLOBAL_ROOT_GID)) + seq_printf(seq, ",gid=%u", + from_kgid_munged(&init_user_ns, server->m.gid)); + if (!uid_eq(server->m.mounted_uid, GLOBAL_ROOT_UID)) + seq_printf(seq, ",owner=%u", + from_kuid_munged(&init_user_ns, server->m.mounted_uid)); tmp = server->m.file_mode & S_IALLUGO; if (tmp != NCP_DEFAULT_FILE_MODE) seq_printf(seq, ",mode=0%o", tmp); @@ -381,13 +384,13 @@ static int ncp_parse_options(struct ncp_mount_data_kernel *data, char *options) data->flags = 0; data->int_flags = 0; - data->mounted_uid = 0; + data->mounted_uid = GLOBAL_ROOT_UID; data->wdog_pid = NULL; data->ncp_fd = ~0; data->time_out = NCP_DEFAULT_TIME_OUT; data->retry_count = NCP_DEFAULT_RETRY_COUNT; - data->uid = 0; - data->gid = 0; + data->uid = GLOBAL_ROOT_UID; + data->gid = GLOBAL_ROOT_GID; data->file_mode = NCP_DEFAULT_FILE_MODE; data->dir_mode = NCP_DEFAULT_DIR_MODE; data->info_fd = -1; @@ -399,13 +402,19 @@ static int ncp_parse_options(struct ncp_mount_data_kernel *data, char *options) goto err; switch (optval) { case 'u': - data->uid = optint; + data->uid = make_kuid(current_user_ns(), optint); + if (!uid_valid(data->uid)) + goto err; break; case 'g': - data->gid = optint; + data->gid = make_kgid(current_user_ns(), optint); + if (!gid_valid(data->gid)) + goto err; break; case 'o': - data->mounted_uid = optint; + data->mounted_uid = make_kuid(current_user_ns(), optint); + if (!uid_valid(data->mounted_uid)) + goto err; break; case 'm': data->file_mode = optint; @@ -480,13 +489,13 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) data.flags = md->flags; data.int_flags = NCP_IMOUNT_LOGGEDIN_POSSIBLE; - data.mounted_uid = md->mounted_uid; + data.mounted_uid = make_kuid(current_user_ns(), md->mounted_uid); data.wdog_pid = find_get_pid(md->wdog_pid); data.ncp_fd = md->ncp_fd; data.time_out = md->time_out; data.retry_count = md->retry_count; - data.uid = md->uid; - data.gid = md->gid; + data.uid = make_kuid(current_user_ns(), md->uid); + data.gid = make_kgid(current_user_ns(), md->gid); data.file_mode = md->file_mode; data.dir_mode = md->dir_mode; data.info_fd = -1; @@ -499,13 +508,13 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) struct ncp_mount_data_v4* md = (struct ncp_mount_data_v4*)raw_data; data.flags = md->flags; - data.mounted_uid = md->mounted_uid; + data.mounted_uid = make_kuid(current_user_ns(), md->mounted_uid); data.wdog_pid = find_get_pid(md->wdog_pid); data.ncp_fd = md->ncp_fd; data.time_out = md->time_out; data.retry_count = md->retry_count; - data.uid = md->uid; - data.gid = md->gid; + data.uid = make_kuid(current_user_ns(), md->uid); + data.gid = make_kgid(current_user_ns(), md->gid); data.file_mode = md->file_mode; data.dir_mode = md->dir_mode; data.info_fd = -1; @@ -520,6 +529,10 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) goto out; break; } + error = -EINVAL; + if (!uid_valid(data.mounted_uid) || !uid_valid(data.uid) || + !gid_valid(data.gid)) + goto out; error = -EBADF; ncp_filp = fget(data.ncp_fd); if (!ncp_filp) @@ -886,12 +899,10 @@ int ncp_notify_change(struct dentry *dentry, struct iattr *attr) goto out; result = -EPERM; - if (((attr->ia_valid & ATTR_UID) && - (attr->ia_uid != server->m.uid))) + if ((attr->ia_valid & ATTR_UID) && !uid_eq(attr->ia_uid, server->m.uid)) goto out; - if (((attr->ia_valid & ATTR_GID) && - (attr->ia_gid != server->m.gid))) + if ((attr->ia_valid & ATTR_GID) && !gid_eq(attr->ia_gid, server->m.gid)) goto out; if (((attr->ia_valid & ATTR_MODE) && diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c index 6958adfaff08..d44318d27504 100644 --- a/fs/ncpfs/ioctl.c +++ b/fs/ncpfs/ioctl.c @@ -45,7 +45,7 @@ ncp_get_fs_info(struct ncp_server * server, struct inode *inode, return -EINVAL; } /* TODO: info.addr = server->m.serv_addr; */ - SET_UID(info.mounted_uid, server->m.mounted_uid); + SET_UID(info.mounted_uid, from_kuid_munged(current_user_ns(), server->m.mounted_uid)); info.connection = server->connection; info.buffer_size = server->buffer_size; info.volume_number = NCP_FINFO(inode)->volNumber; @@ -69,7 +69,7 @@ ncp_get_fs_info_v2(struct ncp_server * server, struct inode *inode, DPRINTK("info.version invalid: %d\n", info2.version); return -EINVAL; } - info2.mounted_uid = server->m.mounted_uid; + info2.mounted_uid = from_kuid_munged(current_user_ns(), server->m.mounted_uid); info2.connection = server->connection; info2.buffer_size = server->buffer_size; info2.volume_number = NCP_FINFO(inode)->volNumber; @@ -135,7 +135,7 @@ ncp_get_compat_fs_info_v2(struct ncp_server * server, struct inode *inode, DPRINTK("info.version invalid: %d\n", info2.version); return -EINVAL; } - info2.mounted_uid = server->m.mounted_uid; + info2.mounted_uid = from_kuid_munged(current_user_ns(), server->m.mounted_uid); info2.connection = server->connection; info2.buffer_size = server->buffer_size; info2.volume_number = NCP_FINFO(inode)->volNumber; @@ -348,22 +348,25 @@ static long __ncp_ioctl(struct inode *inode, unsigned int cmd, unsigned long arg { u16 uid; - SET_UID(uid, server->m.mounted_uid); + SET_UID(uid, from_kuid_munged(current_user_ns(), server->m.mounted_uid)); if (put_user(uid, (u16 __user *)argp)) return -EFAULT; return 0; } case NCP_IOC_GETMOUNTUID32: - if (put_user(server->m.mounted_uid, - (u32 __user *)argp)) + { + uid_t uid = from_kuid_munged(current_user_ns(), server->m.mounted_uid); + if (put_user(uid, (u32 __user *)argp)) return -EFAULT; return 0; + } case NCP_IOC_GETMOUNTUID64: - if (put_user(server->m.mounted_uid, - (u64 __user *)argp)) + { + uid_t uid = from_kuid_munged(current_user_ns(), server->m.mounted_uid); + if (put_user(uid, (u64 __user *)argp)) return -EFAULT; return 0; - + } case NCP_IOC_GETROOT: { struct ncp_setroot_ioctl sr; @@ -810,7 +813,7 @@ long ncp_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { struct inode *inode = filp->f_dentry->d_inode; struct ncp_server *server = NCP_SERVER(inode); - uid_t uid = current_uid(); + kuid_t uid = current_uid(); int need_drop_write = 0; long ret; @@ -824,7 +827,7 @@ long ncp_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) } break; } - if (server->m.mounted_uid != uid) { + if (!uid_eq(server->m.mounted_uid, uid)) { switch (cmd) { /* * Only mount owner can issue these ioctls. Information diff --git a/fs/ncpfs/ncp_fs_sb.h b/fs/ncpfs/ncp_fs_sb.h index 54cc0cdb3dcb..c51b2c543539 100644 --- a/fs/ncpfs/ncp_fs_sb.h +++ b/fs/ncpfs/ncp_fs_sb.h @@ -23,15 +23,15 @@ struct ncp_mount_data_kernel { unsigned long flags; /* NCP_MOUNT_* flags */ unsigned int int_flags; /* internal flags */ #define NCP_IMOUNT_LOGGEDIN_POSSIBLE 0x0001 - uid_t mounted_uid; /* Who may umount() this filesystem? */ + kuid_t mounted_uid; /* Who may umount() this filesystem? */ struct pid *wdog_pid; /* Who cares for our watchdog packets? */ unsigned int ncp_fd; /* The socket to the ncp port */ unsigned int time_out; /* How long should I wait after sending a NCP request? */ unsigned int retry_count; /* And how often should I retry? */ unsigned char mounted_vol[NCP_VOLNAME_LEN + 1]; - uid_t uid; - gid_t gid; + kuid_t uid; + kgid_t gid; umode_t file_mode; umode_t dir_mode; int info_fd; -- cgit From ddca4e1730cf1b72044ae76ddf17b29d790b4dbc Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 14:50:52 -0800 Subject: nfs_common: Update the translation between nfsv3 acls linux posix acls - Use kuid_t and kgit in struct nfsacl_encode_desc. - Convert from kuids and kgids when generating on the wire values. - Convert on the wire values to kuids and kgids when read. - Modify cmp_acl_entry to be type safe comparison on posix acls. Only acls with type ACL_USER and ACL_GROUP can appear more than once and as such need to compare more than their tag. - The e_id field is being removed from posix acls so don't initialize it. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfs_common/nfsacl.c | 41 ++++++++++++++++++++++++++--------------- 1 file changed, 26 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/nfs_common/nfsacl.c b/fs/nfs_common/nfsacl.c index 6940439bd609..ed628f71274c 100644 --- a/fs/nfs_common/nfsacl.c +++ b/fs/nfs_common/nfsacl.c @@ -38,8 +38,8 @@ struct nfsacl_encode_desc { unsigned int count; struct posix_acl *acl; int typeflag; - uid_t uid; - gid_t gid; + kuid_t uid; + kgid_t gid; }; struct nfsacl_simple_acl { @@ -60,14 +60,16 @@ xdr_nfsace_encode(struct xdr_array2_desc *desc, void *elem) *p++ = htonl(entry->e_tag | nfsacl_desc->typeflag); switch(entry->e_tag) { case ACL_USER_OBJ: - *p++ = htonl(nfsacl_desc->uid); + *p++ = htonl(from_kuid(&init_user_ns, nfsacl_desc->uid)); break; case ACL_GROUP_OBJ: - *p++ = htonl(nfsacl_desc->gid); + *p++ = htonl(from_kgid(&init_user_ns, nfsacl_desc->gid)); break; case ACL_USER: + *p++ = htonl(from_kuid(&init_user_ns, entry->e_uid)); + break; case ACL_GROUP: - *p++ = htonl(entry->e_id); + *p++ = htonl(from_kgid(&init_user_ns, entry->e_gid)); break; default: /* Solaris depends on that! */ *p++ = 0; @@ -148,6 +150,7 @@ xdr_nfsace_decode(struct xdr_array2_desc *desc, void *elem) (struct nfsacl_decode_desc *) desc; __be32 *p = elem; struct posix_acl_entry *entry; + unsigned int id; if (!nfsacl_desc->acl) { if (desc->array_len > NFS_ACL_MAX_ENTRIES) @@ -160,14 +163,22 @@ xdr_nfsace_decode(struct xdr_array2_desc *desc, void *elem) entry = &nfsacl_desc->acl->a_entries[nfsacl_desc->count++]; entry->e_tag = ntohl(*p++) & ~NFS_ACL_DEFAULT; - entry->e_id = ntohl(*p++); + id = ntohl(*p++); entry->e_perm = ntohl(*p++); switch(entry->e_tag) { - case ACL_USER_OBJ: case ACL_USER: - case ACL_GROUP_OBJ: + entry->e_uid = make_kuid(&init_user_ns, id); + if (!uid_valid(entry->e_uid)) + return -EINVAL; + break; case ACL_GROUP: + entry->e_gid = make_kgid(&init_user_ns, id); + if (!gid_valid(entry->e_gid)) + return -EINVAL; + break; + case ACL_USER_OBJ: + case ACL_GROUP_OBJ: case ACL_OTHER: if (entry->e_perm & ~S_IRWXO) return -EINVAL; @@ -190,9 +201,13 @@ cmp_acl_entry(const void *x, const void *y) if (a->e_tag != b->e_tag) return a->e_tag - b->e_tag; - else if (a->e_id > b->e_id) + else if ((a->e_tag == ACL_USER) && uid_gt(a->e_uid, b->e_uid)) + return 1; + else if ((a->e_tag == ACL_USER) && uid_lt(a->e_uid, b->e_uid)) + return -1; + else if ((a->e_tag == ACL_GROUP) && gid_gt(a->e_gid, b->e_gid)) return 1; - else if (a->e_id < b->e_id) + else if ((a->e_tag == ACL_GROUP) && gid_lt(a->e_gid, b->e_gid)) return -1; else return 0; @@ -213,22 +228,18 @@ posix_acl_from_nfsacl(struct posix_acl *acl) sort(acl->a_entries, acl->a_count, sizeof(struct posix_acl_entry), cmp_acl_entry, NULL); - /* Clear undefined identifier fields and find the ACL_GROUP_OBJ - and ACL_MASK entries. */ + /* Find the ACL_GROUP_OBJ and ACL_MASK entries. */ FOREACH_ACL_ENTRY(pa, acl, pe) { switch(pa->e_tag) { case ACL_USER_OBJ: - pa->e_id = ACL_UNDEFINED_ID; break; case ACL_GROUP_OBJ: - pa->e_id = ACL_UNDEFINED_ID; group_obj = pa; break; case ACL_MASK: mask = pa; /* fall through */ case ACL_OTHER: - pa->e_id = ACL_UNDEFINED_ID; break; } } -- cgit From 4e963d4f3ed1b756ff20f84960fcb4db42509146 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 03:03:16 -0800 Subject: nfs: Pass GLOBAL_ROOT_UID and GLOBAL_ROOT_GID to keyring alloc Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfs/idmap.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index bc3968fa81e5..f77017c4e28b 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -193,7 +193,8 @@ static int nfs_idmap_init_keyring(void) if (!cred) return -ENOMEM; - keyring = keyring_alloc(".id_resolver", 0, 0, cred, + keyring = keyring_alloc(".id_resolver", + GLOBAL_ROOT_UID, GLOBAL_ROOT_GID, cred, (KEY_POS_ALL & ~KEY_POS_SETATTR) | KEY_USR_VIEW | KEY_USR_READ, KEY_ALLOC_NOT_IN_QUOTA, NULL); -- cgit From 9f309c86cf343c59c79d25d9bde5d4a895d2e81f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 03:21:47 -0800 Subject: nfs: Convert idmap to use kuids and kgids Convert nfs_map_name_to_uid to return a kuid_t value. Convert nfs_map_name_to_gid to return a kgid_t value. Convert nfs_map_uid_to_name to take a kuid_t paramater. Convert nfs_map_gid_to_name to take a kgid_t paramater. Tweak nfs_fattr_map_owner_to_name to use a kuid_t intermediate value. Tweak nfs_fattr_map_group_to_name to use a kgid_t intermediate value. Which makes these functions properly handle kuids and kgids, including erroring of the generated kuid or kgid is invalid. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfs/idmap.c | 50 ++++++++++++++++++++++++++++++++++---------------- 1 file changed, 34 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index f77017c4e28b..b9623d19d599 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -97,7 +97,7 @@ static void nfs_fattr_free_group_name(struct nfs_fattr *fattr) static bool nfs_fattr_map_owner_name(struct nfs_server *server, struct nfs_fattr *fattr) { struct nfs4_string *owner = fattr->owner_name; - __u32 uid; + kuid_t uid; if (!(fattr->valid & NFS_ATTR_FATTR_OWNER_NAME)) return false; @@ -111,7 +111,7 @@ static bool nfs_fattr_map_owner_name(struct nfs_server *server, struct nfs_fattr static bool nfs_fattr_map_group_name(struct nfs_server *server, struct nfs_fattr *fattr) { struct nfs4_string *group = fattr->group_name; - __u32 gid; + kgid_t gid; if (!(fattr->valid & NFS_ATTR_FATTR_GROUP_NAME)) return false; @@ -837,43 +837,61 @@ idmap_release_pipe(struct inode *inode) nfs_idmap_abort_pipe_upcall(idmap, -EPIPE); } -int nfs_map_name_to_uid(const struct nfs_server *server, const char *name, size_t namelen, __u32 *uid) +int nfs_map_name_to_uid(const struct nfs_server *server, const char *name, size_t namelen, kuid_t *uid) { struct idmap *idmap = server->nfs_client->cl_idmap; + __u32 id = -1; + int ret = 0; - if (nfs_map_string_to_numeric(name, namelen, uid)) - return 0; - return nfs_idmap_lookup_id(name, namelen, "uid", uid, idmap); + if (!nfs_map_string_to_numeric(name, namelen, &id)) + ret = nfs_idmap_lookup_id(name, namelen, "uid", &id, idmap); + if (ret == 0) { + *uid = make_kuid(&init_user_ns, id); + if (!uid_valid(*uid)) + ret = -ERANGE; + } + return ret; } -int nfs_map_group_to_gid(const struct nfs_server *server, const char *name, size_t namelen, __u32 *gid) +int nfs_map_group_to_gid(const struct nfs_server *server, const char *name, size_t namelen, kgid_t *gid) { struct idmap *idmap = server->nfs_client->cl_idmap; + __u32 id = -1; + int ret = 0; - if (nfs_map_string_to_numeric(name, namelen, gid)) - return 0; - return nfs_idmap_lookup_id(name, namelen, "gid", gid, idmap); + if (!nfs_map_string_to_numeric(name, namelen, &id)) + ret = nfs_idmap_lookup_id(name, namelen, "gid", &id, idmap); + if (ret == 0) { + *gid = make_kgid(&init_user_ns, id); + if (!gid_valid(*gid)) + ret = -ERANGE; + } + return ret; } -int nfs_map_uid_to_name(const struct nfs_server *server, __u32 uid, char *buf, size_t buflen) +int nfs_map_uid_to_name(const struct nfs_server *server, kuid_t uid, char *buf, size_t buflen) { struct idmap *idmap = server->nfs_client->cl_idmap; int ret = -EINVAL; + __u32 id; + id = from_kuid(&init_user_ns, uid); if (!(server->caps & NFS_CAP_UIDGID_NOMAP)) - ret = nfs_idmap_lookup_name(uid, "user", buf, buflen, idmap); + ret = nfs_idmap_lookup_name(id, "user", buf, buflen, idmap); if (ret < 0) - ret = nfs_map_numeric_to_string(uid, buf, buflen); + ret = nfs_map_numeric_to_string(id, buf, buflen); return ret; } -int nfs_map_gid_to_group(const struct nfs_server *server, __u32 gid, char *buf, size_t buflen) +int nfs_map_gid_to_group(const struct nfs_server *server, kgid_t gid, char *buf, size_t buflen) { struct idmap *idmap = server->nfs_client->cl_idmap; int ret = -EINVAL; + __u32 id; + id = from_kgid(&init_user_ns, gid); if (!(server->caps & NFS_CAP_UIDGID_NOMAP)) - ret = nfs_idmap_lookup_name(gid, "group", buf, buflen, idmap); + ret = nfs_idmap_lookup_name(id, "group", buf, buflen, idmap); if (ret < 0) - ret = nfs_map_numeric_to_string(gid, buf, buflen); + ret = nfs_map_numeric_to_string(id, buf, buflen); return ret; } -- cgit From cfa0898d4ff1874b86d21768adb49d2c033fa90b Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 03:45:54 -0800 Subject: nfs: Convert nfs2xdr to use kuids and kgids When reading uids and gids off the wire convert them to kuids and kgids. When putting kuids and kgids onto the wire first convert them to uids and gids the other side will understand. Add an additional failure mode for incoming uid or gids that are invalid. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfs/nfs2xdr.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index 06b9df49f7f7..62db136339ea 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -290,8 +290,13 @@ static int decode_fattr(struct xdr_stream *xdr, struct nfs_fattr *fattr) fattr->mode = be32_to_cpup(p++); fattr->nlink = be32_to_cpup(p++); - fattr->uid = be32_to_cpup(p++); - fattr->gid = be32_to_cpup(p++); + fattr->uid = make_kuid(&init_user_ns, be32_to_cpup(p++)); + if (!uid_valid(fattr->uid)) + goto out_uid; + fattr->gid = make_kgid(&init_user_ns, be32_to_cpup(p++)); + if (!gid_valid(fattr->gid)) + goto out_gid; + fattr->size = be32_to_cpup(p++); fattr->du.nfs2.blocksize = be32_to_cpup(p++); @@ -313,6 +318,12 @@ static int decode_fattr(struct xdr_stream *xdr, struct nfs_fattr *fattr) fattr->change_attr = nfs_timespec_to_change_attr(&fattr->ctime); return 0; +out_uid: + dprintk("NFS: returned invalid uid\n"); + return -EINVAL; +out_gid: + dprintk("NFS: returned invalid gid\n"); + return -EINVAL; out_overflow: print_overflow_msg(__func__, xdr); return -EIO; @@ -351,11 +362,11 @@ static void encode_sattr(struct xdr_stream *xdr, const struct iattr *attr) else *p++ = cpu_to_be32(NFS2_SATTR_NOT_SET); if (attr->ia_valid & ATTR_UID) - *p++ = cpu_to_be32(attr->ia_uid); + *p++ = cpu_to_be32(from_kuid(&init_user_ns, attr->ia_uid)); else *p++ = cpu_to_be32(NFS2_SATTR_NOT_SET); if (attr->ia_valid & ATTR_GID) - *p++ = cpu_to_be32(attr->ia_gid); + *p++ = cpu_to_be32(from_kgid(&init_user_ns, attr->ia_gid)); else *p++ = cpu_to_be32(NFS2_SATTR_NOT_SET); if (attr->ia_valid & ATTR_SIZE) -- cgit From 57a38dae2ac3f3d618bf197314e2805840220d0f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 03:50:00 -0800 Subject: nfs: Convert nfs3xdr to use kuids and kgids When reading uids and gids off the wire convert them to kuids and kgids. When putting kuids and kgids onto the wire first convert them to uids and gids the other side will understand. Add an additional failure mode incoming for uids or gids that are invalid. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfs/nfs3xdr.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index bffc32406fbf..fa6d72131c19 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -592,13 +592,13 @@ static void encode_sattr3(struct xdr_stream *xdr, const struct iattr *attr) if (attr->ia_valid & ATTR_UID) { *p++ = xdr_one; - *p++ = cpu_to_be32(attr->ia_uid); + *p++ = cpu_to_be32(from_kuid(&init_user_ns, attr->ia_uid)); } else *p++ = xdr_zero; if (attr->ia_valid & ATTR_GID) { *p++ = xdr_one; - *p++ = cpu_to_be32(attr->ia_gid); + *p++ = cpu_to_be32(from_kgid(&init_user_ns, attr->ia_gid)); } else *p++ = xdr_zero; @@ -657,8 +657,12 @@ static int decode_fattr3(struct xdr_stream *xdr, struct nfs_fattr *fattr) fattr->mode = (be32_to_cpup(p++) & ~S_IFMT) | fmode; fattr->nlink = be32_to_cpup(p++); - fattr->uid = be32_to_cpup(p++); - fattr->gid = be32_to_cpup(p++); + fattr->uid = make_kuid(&init_user_ns, be32_to_cpup(p++)); + if (!uid_valid(fattr->uid)) + goto out_uid; + fattr->gid = make_kgid(&init_user_ns, be32_to_cpup(p++)); + if (!gid_valid(fattr->gid)) + goto out_gid; p = xdr_decode_size3(p, &fattr->size); p = xdr_decode_size3(p, &fattr->du.nfs3.used); @@ -675,6 +679,12 @@ static int decode_fattr3(struct xdr_stream *xdr, struct nfs_fattr *fattr) fattr->valid |= NFS_ATTR_FATTR_V3; return 0; +out_uid: + dprintk("NFS: returned invalid uid\n"); + return -EINVAL; +out_gid: + dprintk("NFS: returned invalid gid\n"); + return -EINVAL; out_overflow: print_overflow_msg(__func__, xdr); return -EIO; -- cgit From e5782076e72be2a39bf261f7db03e1769ec05198 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 14:22:02 -0800 Subject: nfs: Convert nfs4xdr to use kuids and kgids When reading uids and gids off the wire convert them to kuids and kgids. When putting kuids and kgids onto the wire first convert them to uids and gids the other side will understand. When printing kuids and kgids convert them to values in the initial user namespace then use normal printf formats. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfs/nfs4xdr.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 26b143920433..e3edda554ac7 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -1002,7 +1002,7 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const owner_namelen = nfs_map_uid_to_name(server, iap->ia_uid, owner_name, IDMAP_NAMESZ); if (owner_namelen < 0) { dprintk("nfs: couldn't resolve uid %d to string\n", - iap->ia_uid); + from_kuid(&init_user_ns, iap->ia_uid)); /* XXX */ strcpy(owner_name, "nobody"); owner_namelen = sizeof("nobody") - 1; @@ -1014,7 +1014,7 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const owner_grouplen = nfs_map_gid_to_group(server, iap->ia_gid, owner_group, IDMAP_NAMESZ); if (owner_grouplen < 0) { dprintk("nfs: couldn't resolve gid %d to string\n", - iap->ia_gid); + from_kgid(&init_user_ns, iap->ia_gid)); strcpy(owner_group, "nobody"); owner_grouplen = sizeof("nobody") - 1; /* goto out; */ @@ -3778,14 +3778,14 @@ out_overflow: } static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, - const struct nfs_server *server, uint32_t *uid, + const struct nfs_server *server, kuid_t *uid, struct nfs4_string *owner_name) { uint32_t len; __be32 *p; int ret = 0; - *uid = -2; + *uid = make_kuid(&init_user_ns, -2); if (unlikely(bitmap[1] & (FATTR4_WORD1_OWNER - 1U))) return -EIO; if (likely(bitmap[1] & FATTR4_WORD1_OWNER)) { @@ -3813,7 +3813,7 @@ static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, __func__, len); bitmap[1] &= ~FATTR4_WORD1_OWNER; } - dprintk("%s: uid=%d\n", __func__, (int)*uid); + dprintk("%s: uid=%d\n", __func__, (int)from_kuid(&init_user_ns, *uid)); return ret; out_overflow: print_overflow_msg(__func__, xdr); @@ -3821,14 +3821,14 @@ out_overflow: } static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, - const struct nfs_server *server, uint32_t *gid, + const struct nfs_server *server, kgid_t *gid, struct nfs4_string *group_name) { uint32_t len; __be32 *p; int ret = 0; - *gid = -2; + *gid = make_kgid(&init_user_ns, -2); if (unlikely(bitmap[1] & (FATTR4_WORD1_OWNER_GROUP - 1U))) return -EIO; if (likely(bitmap[1] & FATTR4_WORD1_OWNER_GROUP)) { @@ -3856,7 +3856,7 @@ static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, __func__, len); bitmap[1] &= ~FATTR4_WORD1_OWNER_GROUP; } - dprintk("%s: gid=%d\n", __func__, (int)*gid); + dprintk("%s: gid=%d\n", __func__, (int)from_kgid(&init_user_ns, *gid)); return ret; out_overflow: print_overflow_msg(__func__, xdr); -- cgit From 9ff593c4739c76c31ffdf7f2a1d67489ece728e6 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 14:26:23 -0800 Subject: nfs: kuid and kgid conversions for nfs/inode.c - Use uid_eq and gid_eq when comparing kuids and kgids. - Use make_kuid(&init_user_ns, -2) and make_kgid(&init_user_ns, -2) as the initial uid and gid on nfs inodes, instead of using the typeunsafe value of -2. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfs/inode.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index ebeb94ce1b0b..ce2b0394c4d6 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -332,8 +332,8 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) inode->i_version = 0; inode->i_size = 0; clear_nlink(inode); - inode->i_uid = -2; - inode->i_gid = -2; + inode->i_uid = make_kuid(&init_user_ns, -2); + inode->i_gid = make_kgid(&init_user_ns, -2); inode->i_blocks = 0; memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf)); nfsi->write_io = 0; @@ -1009,9 +1009,9 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat /* Have any file permissions changed? */ if ((fattr->valid & NFS_ATTR_FATTR_MODE) && (inode->i_mode & S_IALLUGO) != (fattr->mode & S_IALLUGO)) invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; - if ((fattr->valid & NFS_ATTR_FATTR_OWNER) && inode->i_uid != fattr->uid) + if ((fattr->valid & NFS_ATTR_FATTR_OWNER) && !uid_eq(inode->i_uid, fattr->uid)) invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; - if ((fattr->valid & NFS_ATTR_FATTR_GROUP) && inode->i_gid != fattr->gid) + if ((fattr->valid & NFS_ATTR_FATTR_GROUP) && !gid_eq(inode->i_gid, fattr->gid)) invalid |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL; /* Has the link count changed? */ @@ -1440,7 +1440,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) | NFS_INO_REVAL_FORCED); if (fattr->valid & NFS_ATTR_FATTR_OWNER) { - if (inode->i_uid != fattr->uid) { + if (!uid_eq(inode->i_uid, fattr->uid)) { invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; inode->i_uid = fattr->uid; } @@ -1451,7 +1451,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) | NFS_INO_REVAL_FORCED); if (fattr->valid & NFS_ATTR_FATTR_GROUP) { - if (inode->i_gid != fattr->gid) { + if (!gid_eq(inode->i_gid, fattr->gid)) { invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; inode->i_gid = fattr->gid; } -- cgit From 6c1810e040d87fcd8fc95aedfd2ef6979d71e517 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 1 Feb 2013 03:38:35 -0800 Subject: nfsd: Remove declaration of nonexistent nfs4_acl_permisison Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/acl.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/acl.h b/fs/nfsd/acl.h index 34e5c40af5ef..8b186a4955cc 100644 --- a/fs/nfsd/acl.h +++ b/fs/nfsd/acl.h @@ -44,8 +44,6 @@ struct nfs4_acl *nfs4_acl_new(int); int nfs4_acl_get_whotype(char *, u32); int nfs4_acl_write_who(int who, char *p); -int nfs4_acl_permission(struct nfs4_acl *acl, uid_t owner, gid_t group, - uid_t who, u32 mask); #define NFS4_ACL_TYPE_DEFAULT 0x01 #define NFS4_ACL_DIR 0x02 -- cgit From b5663898ec3fa7f1a58a9def9592be345bb173c2 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 03:54:35 -0800 Subject: nfsd: idmap use u32 not uid_t as the intermediate type u32 and uid_t have the same size and semantics so this change should have no operational effect. This just removes the WTF factor when looking at variables that hold both uids and gids whos type is uid_t. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/nfs4idmap.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index a1f10c0a6255..7e84dfa23d83 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -65,7 +65,7 @@ MODULE_PARM_DESC(nfs4_disable_idmapping, struct ent { struct cache_head h; int type; /* User / Group */ - uid_t id; + u32 id; char name[IDMAP_NAMESZ]; char authname[IDMAP_NAMESZ]; }; @@ -540,7 +540,7 @@ rqst_authname(struct svc_rqst *rqstp) static __be32 idmap_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen, - uid_t *id) + u32 *id) { struct ent *item, key = { .type = type, @@ -564,7 +564,7 @@ idmap_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen } static int -idmap_id_to_name(struct svc_rqst *rqstp, int type, uid_t id, char *name) +idmap_id_to_name(struct svc_rqst *rqstp, int type, u32 id, char *name) { struct ent *item, key = { .id = id, @@ -587,7 +587,7 @@ idmap_id_to_name(struct svc_rqst *rqstp, int type, uid_t id, char *name) } static bool -numeric_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen, uid_t *id) +numeric_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen, u32 *id) { int ret; char buf[11]; @@ -603,7 +603,7 @@ numeric_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namel } static __be32 -do_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen, uid_t *id) +do_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen, u32 *id) { if (nfs4_disable_idmapping && rqstp->rq_cred.cr_flavor < RPC_AUTH_GSS) if (numeric_name_to_id(rqstp, type, name, namelen, id)) @@ -616,7 +616,7 @@ do_name_to_id(struct svc_rqst *rqstp, int type, const char *name, u32 namelen, u } static int -do_id_to_name(struct svc_rqst *rqstp, int type, uid_t id, char *name) +do_id_to_name(struct svc_rqst *rqstp, int type, u32 id, char *name) { if (nfs4_disable_idmapping && rqstp->rq_cred.cr_flavor < RPC_AUTH_GSS) return sprintf(name, "%u", id); -- cgit From 65e10f6d0ab09ba95c2eb07cac43208692cf670e Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 03:24:21 -0800 Subject: nfsd: Convert idmap to use kuids and kgids Convert nfsd_map_name_to_uid to return a kuid_t value. Convert nfsd_map_name_to_gid to return a kgid_t value. Convert nfsd_map_uid_to_name to take a kuid_t parameter. Convert nfsd_map_gid_to_name to take a kgid_t paramater. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/idmap.h | 8 ++++---- fs/nfsd/nfs4idmap.c | 26 ++++++++++++++++++++------ 2 files changed, 24 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/idmap.h b/fs/nfsd/idmap.h index 9d513efc01ba..bf95f6b817a4 100644 --- a/fs/nfsd/idmap.h +++ b/fs/nfsd/idmap.h @@ -54,9 +54,9 @@ static inline void nfsd_idmap_shutdown(struct net *net) } #endif -__be32 nfsd_map_name_to_uid(struct svc_rqst *, const char *, size_t, __u32 *); -__be32 nfsd_map_name_to_gid(struct svc_rqst *, const char *, size_t, __u32 *); -int nfsd_map_uid_to_name(struct svc_rqst *, __u32, char *); -int nfsd_map_gid_to_name(struct svc_rqst *, __u32, char *); +__be32 nfsd_map_name_to_uid(struct svc_rqst *, const char *, size_t, kuid_t *); +__be32 nfsd_map_name_to_gid(struct svc_rqst *, const char *, size_t, kgid_t *); +int nfsd_map_uid_to_name(struct svc_rqst *, kuid_t, char *); +int nfsd_map_gid_to_name(struct svc_rqst *, kgid_t, char *); #endif /* LINUX_NFSD_IDMAP_H */ diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index 7e84dfa23d83..0ce12346df9c 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -625,26 +625,40 @@ do_id_to_name(struct svc_rqst *rqstp, int type, u32 id, char *name) __be32 nfsd_map_name_to_uid(struct svc_rqst *rqstp, const char *name, size_t namelen, - __u32 *id) + kuid_t *uid) { - return do_name_to_id(rqstp, IDMAP_TYPE_USER, name, namelen, id); + __be32 status; + u32 id = -1; + status = do_name_to_id(rqstp, IDMAP_TYPE_USER, name, namelen, &id); + *uid = make_kuid(&init_user_ns, id); + if (!uid_valid(*uid)) + status = nfserr_badowner; + return status; } __be32 nfsd_map_name_to_gid(struct svc_rqst *rqstp, const char *name, size_t namelen, - __u32 *id) + kgid_t *gid) { - return do_name_to_id(rqstp, IDMAP_TYPE_GROUP, name, namelen, id); + __be32 status; + u32 id = -1; + status = do_name_to_id(rqstp, IDMAP_TYPE_GROUP, name, namelen, &id); + *gid = make_kgid(&init_user_ns, id); + if (!gid_valid(*gid)) + status = nfserr_badowner; + return status; } int -nfsd_map_uid_to_name(struct svc_rqst *rqstp, __u32 id, char *name) +nfsd_map_uid_to_name(struct svc_rqst *rqstp, kuid_t uid, char *name) { + u32 id = from_kuid(&init_user_ns, uid); return do_id_to_name(rqstp, IDMAP_TYPE_USER, id, name); } int -nfsd_map_gid_to_name(struct svc_rqst *rqstp, __u32 id, char *name) +nfsd_map_gid_to_name(struct svc_rqst *rqstp, kgid_t gid, char *name) { + u32 id = from_kgid(&init_user_ns, gid); return do_id_to_name(rqstp, IDMAP_TYPE_GROUP, id, name); } -- cgit From e097258f2eb4a91e7389ae69a3c87df111637a3f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 04:04:22 -0800 Subject: nfsd: Remove nfsd_luid, nfsd_lgid, nfsd_ruid and nfsd_rgid These trivial macros that don't currently do anything are the last vestiages of an old attempt at uid mapping that was removed from the kernel in September of 2002. Remove them to make it clear what the code is currently doing. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/auth.h | 6 ------ fs/nfsd/nfs3xdr.c | 4 ++-- fs/nfsd/nfsxdr.c | 4 ++-- 3 files changed, 4 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/auth.h b/fs/nfsd/auth.h index 78b3c0e93822..53325a12ba62 100644 --- a/fs/nfsd/auth.h +++ b/fs/nfsd/auth.h @@ -1,6 +1,5 @@ /* * nfsd-specific authentication stuff. - * uid/gid mapping not yet implemented. * * Copyright (C) 1995, 1996 Olaf Kirch */ @@ -8,11 +7,6 @@ #ifndef LINUX_NFSD_AUTH_H #define LINUX_NFSD_AUTH_H -#define nfsd_luid(rq, uid) ((u32)(uid)) -#define nfsd_lgid(rq, gid) ((u32)(gid)) -#define nfsd_ruid(rq, uid) ((u32)(uid)) -#define nfsd_rgid(rq, gid) ((u32)(gid)) - /* * Set the current process's fsuid/fsgid etc to those of the NFS * client user diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c index 324c0baf7cda..1884a3fbb584 100644 --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -167,8 +167,8 @@ encode_fattr3(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, *p++ = htonl(nfs3_ftypes[(stat->mode & S_IFMT) >> 12]); *p++ = htonl((u32) stat->mode); *p++ = htonl((u32) stat->nlink); - *p++ = htonl((u32) nfsd_ruid(rqstp, stat->uid)); - *p++ = htonl((u32) nfsd_rgid(rqstp, stat->gid)); + *p++ = htonl((u32) stat->uid); + *p++ = htonl((u32) stat->gid); if (S_ISLNK(stat->mode) && stat->size > NFS3_MAXPATHLEN) { p = xdr_encode_hyper(p, (u64) NFS3_MAXPATHLEN); } else { diff --git a/fs/nfsd/nfsxdr.c b/fs/nfsd/nfsxdr.c index 979b42106979..1e51e7034a89 100644 --- a/fs/nfsd/nfsxdr.c +++ b/fs/nfsd/nfsxdr.c @@ -151,8 +151,8 @@ encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, *p++ = htonl(nfs_ftypes[type >> 12]); *p++ = htonl((u32) stat->mode); *p++ = htonl((u32) stat->nlink); - *p++ = htonl((u32) nfsd_ruid(rqstp, stat->uid)); - *p++ = htonl((u32) nfsd_rgid(rqstp, stat->gid)); + *p++ = htonl((u32) stat->uid); + *p++ = htonl((u32) stat->gid); if (S_ISLNK(type) && stat->size > NFS_MAXPATHLEN) { *p++ = htonl(NFS_MAXPATHLEN); -- cgit From 458878a705c822a6be267977e435b16576bde59b Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 04:16:08 -0800 Subject: nfsd: Convert nfs3xdr to use kuids and kgids When reading uids and gids off the wire convert them to kuids and kgids. When putting kuids and kgids onto the wire first convert them to uids and gids the other side will understand. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/nfs3xdr.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c index 1884a3fbb584..925c944bc0bc 100644 --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -105,12 +105,14 @@ decode_sattr3(__be32 *p, struct iattr *iap) iap->ia_mode = ntohl(*p++); } if (*p++) { - iap->ia_valid |= ATTR_UID; - iap->ia_uid = ntohl(*p++); + iap->ia_uid = make_kuid(&init_user_ns, ntohl(*p++)); + if (uid_valid(iap->ia_uid)) + iap->ia_valid |= ATTR_UID; } if (*p++) { - iap->ia_valid |= ATTR_GID; - iap->ia_gid = ntohl(*p++); + iap->ia_gid = make_kgid(&init_user_ns, ntohl(*p++)); + if (gid_valid(iap->ia_gid)) + iap->ia_valid |= ATTR_GID; } if (*p++) { u64 newsize; @@ -167,8 +169,8 @@ encode_fattr3(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, *p++ = htonl(nfs3_ftypes[(stat->mode & S_IFMT) >> 12]); *p++ = htonl((u32) stat->mode); *p++ = htonl((u32) stat->nlink); - *p++ = htonl((u32) stat->uid); - *p++ = htonl((u32) stat->gid); + *p++ = htonl((u32) from_kuid(&init_user_ns, stat->uid)); + *p++ = htonl((u32) from_kgid(&init_user_ns, stat->gid)); if (S_ISLNK(stat->mode) && stat->size > NFS3_MAXPATHLEN) { p = xdr_encode_hyper(p, (u64) NFS3_MAXPATHLEN); } else { -- cgit From 7c19723e997a3990951c0db0500009fb90c0c5b9 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 04:32:35 -0800 Subject: nfsd: Convert nfsxdr to use kuids and kgids When reading uids and gids off the wire convert them to kuids and kgids. If the conversion results in an invalid result don't set the ATTR_UID or ATTR_GID. When putting kuids and kgids onto the wire first convert them to uids and gids the other side will understand. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/nfsxdr.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsxdr.c b/fs/nfsd/nfsxdr.c index 1e51e7034a89..4201ede0ec91 100644 --- a/fs/nfsd/nfsxdr.c +++ b/fs/nfsd/nfsxdr.c @@ -100,12 +100,14 @@ decode_sattr(__be32 *p, struct iattr *iap) iap->ia_mode = tmp; } if ((tmp = ntohl(*p++)) != (u32)-1) { - iap->ia_valid |= ATTR_UID; - iap->ia_uid = tmp; + iap->ia_uid = make_kuid(&init_user_ns, tmp); + if (uid_valid(iap->ia_uid)) + iap->ia_valid |= ATTR_UID; } if ((tmp = ntohl(*p++)) != (u32)-1) { - iap->ia_valid |= ATTR_GID; - iap->ia_gid = tmp; + iap->ia_gid = make_kgid(&init_user_ns, tmp); + if (gid_valid(iap->ia_gid)) + iap->ia_valid |= ATTR_GID; } if ((tmp = ntohl(*p++)) != (u32)-1) { iap->ia_valid |= ATTR_SIZE; @@ -151,8 +153,8 @@ encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, *p++ = htonl(nfs_ftypes[type >> 12]); *p++ = htonl((u32) stat->mode); *p++ = htonl((u32) stat->nlink); - *p++ = htonl((u32) stat->uid); - *p++ = htonl((u32) stat->gid); + *p++ = htonl((u32) from_kuid(&init_user_ns, stat->uid)); + *p++ = htonl((u32) from_kgid(&init_user_ns, stat->gid)); if (S_ISLNK(type) && stat->size > NFS_MAXPATHLEN) { *p++ = htonl(NFS_MAXPATHLEN); -- cgit From ab8e4aee0a3f73d1b12e6d63b42075f0586ad4fd Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 05:18:08 -0800 Subject: nfsd: Handle kuids and kgids in the nfs4acl to posix_acl conversion In struct nfs4_ace remove the member who and replace it with an anonymous union holding who_uid and who_gid. Allowing typesafe storage uids and gids. Add a helper pace_gt for sorting posix_acl_entries. In struct posix_user_ace_state to replace uid with a union of kuid_t uid and kgid_t gid. Remove all initializations of the deprecated posic_acl_entry e_id field. Which is not present when user namespaces are enabled. Split find_uid into two functions find_uid and find_gid that work in a typesafe manner. In nfs4xdr update nfsd4_encode_fattr to deal with the changes in struct nfs4_ace. Rewrite nfsd4_encode_name to take a kuid_t and a kgid_t instead of a generic id and flag if it is a group or a uid. Replace the group flag with a test for a valid gid. Modify nfsd4_encode_user to take a kuid_t and call the modifed nfsd4_encode_name. Modify nfsd4_encode_group to take a kgid_t and call the modified nfsd4_encode_name. Modify nfsd4_encode_aclname to take an ace instead of taking the fields of an ace broken out. This allows it to detect if the ace is for a user or a group and to pass the appropriate value while still being typesafe. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/nfs4acl.c | 63 ++++++++++++++++++++++++++++++++++++++++--------------- fs/nfsd/nfs4xdr.c | 41 ++++++++++++++++++++++-------------- 2 files changed, 71 insertions(+), 33 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4acl.c b/fs/nfsd/nfs4acl.c index 9c51aff02ae2..8a50b3c18093 100644 --- a/fs/nfsd/nfs4acl.c +++ b/fs/nfsd/nfs4acl.c @@ -264,7 +264,7 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl, ace->flag = eflag; ace->access_mask = deny_mask_from_posix(deny, flags); ace->whotype = NFS4_ACL_WHO_NAMED; - ace->who = pa->e_id; + ace->who_uid = pa->e_uid; ace++; acl->naces++; } @@ -273,7 +273,7 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl, ace->access_mask = mask_from_posix(pa->e_perm & pas.mask, flags); ace->whotype = NFS4_ACL_WHO_NAMED; - ace->who = pa->e_id; + ace->who_uid = pa->e_uid; ace++; acl->naces++; pa++; @@ -300,7 +300,7 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl, ace->access_mask = mask_from_posix(pa->e_perm & pas.mask, flags); ace->whotype = NFS4_ACL_WHO_NAMED; - ace->who = pa->e_id; + ace->who_gid = pa->e_gid; ace++; acl->naces++; pa++; @@ -329,7 +329,7 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl, ace->flag = eflag | NFS4_ACE_IDENTIFIER_GROUP; ace->access_mask = deny_mask_from_posix(deny, flags); ace->whotype = NFS4_ACL_WHO_NAMED; - ace->who = pa->e_id; + ace->who_gid = pa->e_gid; ace++; acl->naces++; } @@ -345,6 +345,18 @@ _posix_to_nfsv4_one(struct posix_acl *pacl, struct nfs4_acl *acl, acl->naces++; } +static bool +pace_gt(struct posix_acl_entry *pace1, struct posix_acl_entry *pace2) +{ + if (pace1->e_tag != pace2->e_tag) + return pace1->e_tag > pace2->e_tag; + if (pace1->e_tag == ACL_USER) + return uid_gt(pace1->e_uid, pace2->e_uid); + if (pace1->e_tag == ACL_GROUP) + return gid_gt(pace1->e_gid, pace2->e_gid); + return false; +} + static void sort_pacl_range(struct posix_acl *pacl, int start, int end) { int sorted = 0, i; @@ -355,8 +367,8 @@ sort_pacl_range(struct posix_acl *pacl, int start, int end) { while (!sorted) { sorted = 1; for (i = start; i < end; i++) { - if (pacl->a_entries[i].e_id - > pacl->a_entries[i+1].e_id) { + if (pace_gt(&pacl->a_entries[i], + &pacl->a_entries[i+1])) { sorted = 0; tmp = pacl->a_entries[i]; pacl->a_entries[i] = pacl->a_entries[i+1]; @@ -398,7 +410,10 @@ struct posix_ace_state { }; struct posix_user_ace_state { - uid_t uid; + union { + kuid_t uid; + kgid_t gid; + }; struct posix_ace_state perms; }; @@ -521,7 +536,6 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags) if (error) goto out_err; low_mode_from_nfs4(state->owner.allow, &pace->e_perm, flags); - pace->e_id = ACL_UNDEFINED_ID; for (i=0; i < state->users->n; i++) { pace++; @@ -531,7 +545,7 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags) goto out_err; low_mode_from_nfs4(state->users->aces[i].perms.allow, &pace->e_perm, flags); - pace->e_id = state->users->aces[i].uid; + pace->e_uid = state->users->aces[i].uid; add_to_mask(state, &state->users->aces[i].perms); } @@ -541,7 +555,6 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags) if (error) goto out_err; low_mode_from_nfs4(state->group.allow, &pace->e_perm, flags); - pace->e_id = ACL_UNDEFINED_ID; add_to_mask(state, &state->group); for (i=0; i < state->groups->n; i++) { @@ -552,14 +565,13 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags) goto out_err; low_mode_from_nfs4(state->groups->aces[i].perms.allow, &pace->e_perm, flags); - pace->e_id = state->groups->aces[i].uid; + pace->e_gid = state->groups->aces[i].gid; add_to_mask(state, &state->groups->aces[i].perms); } pace++; pace->e_tag = ACL_MASK; low_mode_from_nfs4(state->mask.allow, &pace->e_perm, flags); - pace->e_id = ACL_UNDEFINED_ID; pace++; pace->e_tag = ACL_OTHER; @@ -567,7 +579,6 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags) if (error) goto out_err; low_mode_from_nfs4(state->other.allow, &pace->e_perm, flags); - pace->e_id = ACL_UNDEFINED_ID; return pacl; out_err: @@ -587,12 +598,13 @@ static inline void deny_bits(struct posix_ace_state *astate, u32 mask) astate->deny |= mask & ~astate->allow; } -static int find_uid(struct posix_acl_state *state, struct posix_ace_state_array *a, uid_t uid) +static int find_uid(struct posix_acl_state *state, kuid_t uid) { + struct posix_ace_state_array *a = state->users; int i; for (i = 0; i < a->n; i++) - if (a->aces[i].uid == uid) + if (uid_eq(a->aces[i].uid, uid)) return i; /* Not found: */ a->n++; @@ -603,6 +615,23 @@ static int find_uid(struct posix_acl_state *state, struct posix_ace_state_array return i; } +static int find_gid(struct posix_acl_state *state, kgid_t gid) +{ + struct posix_ace_state_array *a = state->groups; + int i; + + for (i = 0; i < a->n; i++) + if (gid_eq(a->aces[i].gid, gid)) + return i; + /* Not found: */ + a->n++; + a->aces[i].gid = gid; + a->aces[i].perms.allow = state->everyone.allow; + a->aces[i].perms.deny = state->everyone.deny; + + return i; +} + static void deny_bits_array(struct posix_ace_state_array *a, u32 mask) { int i; @@ -636,7 +665,7 @@ static void process_one_v4_ace(struct posix_acl_state *state, } break; case ACL_USER: - i = find_uid(state, state->users, ace->who); + i = find_uid(state, ace->who_uid); if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) { allow_bits(&state->users->aces[i].perms, mask); } else { @@ -658,7 +687,7 @@ static void process_one_v4_ace(struct posix_acl_state *state, } break; case ACL_GROUP: - i = find_uid(state, state->groups, ace->who); + i = find_gid(state, ace->who_gid); if (ace->type == NFS4_ACE_ACCESS_ALLOWED_ACE_TYPE) { allow_bits(&state->groups->aces[i].perms, mask); } else { diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 0dc11586682f..3812b06d24b1 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -293,13 +293,13 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, ace->whotype = nfs4_acl_get_whotype(buf, dummy32); status = nfs_ok; if (ace->whotype != NFS4_ACL_WHO_NAMED) - ace->who = 0; + ; else if (ace->flag & NFS4_ACE_IDENTIFIER_GROUP) status = nfsd_map_name_to_gid(argp->rqstp, - buf, dummy32, &ace->who); + buf, dummy32, &ace->who_gid); else status = nfsd_map_name_to_uid(argp->rqstp, - buf, dummy32, &ace->who); + buf, dummy32, &ace->who_uid); if (status) return status; } @@ -1926,7 +1926,7 @@ static u32 nfs4_file_type(umode_t mode) } static __be32 -nfsd4_encode_name(struct svc_rqst *rqstp, int whotype, uid_t id, int group, +nfsd4_encode_name(struct svc_rqst *rqstp, int whotype, kuid_t uid, kgid_t gid, __be32 **p, int *buflen) { int status; @@ -1935,10 +1935,10 @@ nfsd4_encode_name(struct svc_rqst *rqstp, int whotype, uid_t id, int group, return nfserr_resource; if (whotype != NFS4_ACL_WHO_NAMED) status = nfs4_acl_write_who(whotype, (u8 *)(*p + 1)); - else if (group) - status = nfsd_map_gid_to_name(rqstp, id, (u8 *)(*p + 1)); + else if (gid_valid(gid)) + status = nfsd_map_gid_to_name(rqstp, gid, (u8 *)(*p + 1)); else - status = nfsd_map_uid_to_name(rqstp, id, (u8 *)(*p + 1)); + status = nfsd_map_uid_to_name(rqstp, uid, (u8 *)(*p + 1)); if (status < 0) return nfserrno(status); *p = xdr_encode_opaque(*p, NULL, status); @@ -1948,22 +1948,33 @@ nfsd4_encode_name(struct svc_rqst *rqstp, int whotype, uid_t id, int group, } static inline __be32 -nfsd4_encode_user(struct svc_rqst *rqstp, uid_t uid, __be32 **p, int *buflen) +nfsd4_encode_user(struct svc_rqst *rqstp, kuid_t user, __be32 **p, int *buflen) { - return nfsd4_encode_name(rqstp, NFS4_ACL_WHO_NAMED, uid, 0, p, buflen); + return nfsd4_encode_name(rqstp, NFS4_ACL_WHO_NAMED, user, INVALID_GID, + p, buflen); } static inline __be32 -nfsd4_encode_group(struct svc_rqst *rqstp, uid_t gid, __be32 **p, int *buflen) +nfsd4_encode_group(struct svc_rqst *rqstp, kgid_t group, __be32 **p, int *buflen) { - return nfsd4_encode_name(rqstp, NFS4_ACL_WHO_NAMED, gid, 1, p, buflen); + return nfsd4_encode_name(rqstp, NFS4_ACL_WHO_NAMED, INVALID_UID, group, + p, buflen); } static inline __be32 -nfsd4_encode_aclname(struct svc_rqst *rqstp, int whotype, uid_t id, int group, +nfsd4_encode_aclname(struct svc_rqst *rqstp, struct nfs4_ace *ace, __be32 **p, int *buflen) { - return nfsd4_encode_name(rqstp, whotype, id, group, p, buflen); + kuid_t uid = INVALID_UID; + kgid_t gid = INVALID_GID; + + if (ace->whotype == NFS4_ACL_WHO_NAMED) { + if (ace->flag & NFS4_ACE_IDENTIFIER_GROUP) + gid = ace->who_gid; + else + uid = ace->who_uid; + } + return nfsd4_encode_name(rqstp, ace->whotype, uid, gid, p, buflen); } #define WORD0_ABSENT_FS_ATTRS (FATTR4_WORD0_FS_LOCATIONS | FATTR4_WORD0_FSID | \ @@ -2224,9 +2235,7 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, WRITE32(ace->type); WRITE32(ace->flag); WRITE32(ace->access_mask & NFS4_ACE_MASK_ALL); - status = nfsd4_encode_aclname(rqstp, ace->whotype, - ace->who, ace->flag & NFS4_ACE_IDENTIFIER_GROUP, - &p, &buflen); + status = nfsd4_encode_aclname(rqstp, ace, &p, &buflen); if (status == nfserr_resource) goto out_resource; if (status) -- cgit From 03bc6d1cc1759e6b5959cacc02a19ef36e95e741 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 06:24:49 -0800 Subject: nfsd: Modify nfsd4_cb_sec to use kuids and kgids Change uid and gid in struct nfsd4_cb_sec to be of type kuid_t and kgid_t. In nfsd4_decode_cb_sec when reading uids and gids off the wire convert them to kuids and kgids, and if they don't convert to valid kuids or valid kuids ignore RPC_AUTH_UNIX and don't fill in any of the fields. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/nfs4xdr.c | 13 ++++++++++--- fs/nfsd/state.h | 4 ++-- 2 files changed, 12 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 3812b06d24b1..2d1d06bae3a7 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -464,9 +464,16 @@ static __be32 nfsd4_decode_cb_sec(struct nfsd4_compoundargs *argp, struct nfsd4_ READ32(dummy); READ_BUF(dummy * 4); if (cbs->flavor == (u32)(-1)) { - cbs->uid = uid; - cbs->gid = gid; - cbs->flavor = RPC_AUTH_UNIX; + kuid_t kuid = make_kuid(&init_user_ns, uid); + kgid_t kgid = make_kgid(&init_user_ns, gid); + if (uid_valid(kuid) && gid_valid(kgid)) { + cbs->uid = kuid; + cbs->gid = kgid; + cbs->flavor = RPC_AUTH_UNIX; + } else { + dprintk("RPC_AUTH_UNIX with invalid" + "uid or gid ignoring!\n"); + } } break; case RPC_AUTH_GSS: diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h index d1c229feed52..1a8c7391f7ae 100644 --- a/fs/nfsd/state.h +++ b/fs/nfsd/state.h @@ -152,8 +152,8 @@ struct nfsd4_channel_attrs { struct nfsd4_cb_sec { u32 flavor; /* (u32)(-1) used to mean "no valid flavor" */ - u32 uid; - u32 gid; + kuid_t uid; + kgid_t gid; }; struct nfsd4_create_session { -- cgit From 4c1e1b34d5c800ad3ac9a7e2805b0bea70ad2278 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 06:42:53 -0800 Subject: nfsd: Store ex_anon_uid and ex_anon_gid as kuids and kgids Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/auth.c | 2 +- fs/nfsd/export.c | 22 ++++++++++++++-------- 2 files changed, 15 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/auth.c b/fs/nfsd/auth.c index 34a10d78b839..4d6642b38eae 100644 --- a/fs/nfsd/auth.c +++ b/fs/nfsd/auth.c @@ -58,7 +58,7 @@ int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) for (i = 0; i < rqgi->ngroups; i++) { if (gid_eq(GLOBAL_ROOT_GID, GROUP_AT(rqgi, i))) - GROUP_AT(gi, i) = make_kgid(&init_user_ns, exp->ex_anon_gid); + GROUP_AT(gi, i) = exp->ex_anon_gid; else GROUP_AT(gi, i) = GROUP_AT(rqgi, i); } diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index a3946cf13fc8..5681c5906f08 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -544,13 +544,17 @@ static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen) err = get_int(&mesg, &an_int); if (err) goto out3; - exp.ex_anon_uid= an_int; + exp.ex_anon_uid= make_kuid(&init_user_ns, an_int); + if (!uid_valid(exp.ex_anon_uid)) + goto out3; /* anon gid */ err = get_int(&mesg, &an_int); if (err) goto out3; - exp.ex_anon_gid= an_int; + exp.ex_anon_gid= make_kgid(&init_user_ns, an_int); + if (!gid_valid(exp.ex_anon_gid)) + goto out3; /* fsid */ err = get_int(&mesg, &an_int); @@ -613,7 +617,7 @@ out: } static void exp_flags(struct seq_file *m, int flag, int fsid, - uid_t anonu, uid_t anong, struct nfsd4_fs_locations *fslocs); + kuid_t anonu, kgid_t anong, struct nfsd4_fs_locations *fslocs); static void show_secinfo(struct seq_file *m, struct svc_export *exp); static int svc_export_show(struct seq_file *m, @@ -1179,15 +1183,17 @@ static void show_secinfo(struct seq_file *m, struct svc_export *exp) } static void exp_flags(struct seq_file *m, int flag, int fsid, - uid_t anonu, uid_t anong, struct nfsd4_fs_locations *fsloc) + kuid_t anonu, kgid_t anong, struct nfsd4_fs_locations *fsloc) { show_expflags(m, flag, NFSEXP_ALLFLAGS); if (flag & NFSEXP_FSID) seq_printf(m, ",fsid=%d", fsid); - if (anonu != (uid_t)-2 && anonu != (0x10000-2)) - seq_printf(m, ",anonuid=%u", anonu); - if (anong != (gid_t)-2 && anong != (0x10000-2)) - seq_printf(m, ",anongid=%u", anong); + if (!uid_eq(anonu, make_kuid(&init_user_ns, (uid_t)-2)) && + !uid_eq(anonu, make_kuid(&init_user_ns, 0x10000-2))) + seq_printf(m, ",anonuid=%u", from_kuid(&init_user_ns, anonu)); + if (!gid_eq(anong, make_kgid(&init_user_ns, (gid_t)-2)) && + !gid_eq(anong, make_kgid(&init_user_ns, 0x10000-2))) + seq_printf(m, ",anongid=%u", from_kgid(&init_user_ns, anong)); if (fsloc && fsloc->locations_count > 0) { char *loctype = (fsloc->migrated) ? "refer" : "replicas"; int i; -- cgit From 6fab877900030ba3ae11928efb6087589f1e514c Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Feb 2013 06:53:11 -0800 Subject: nfsd: Properly compare and initialize kuids and kgids Use uid_eq(uid, GLOBAL_ROOT_UID) instead of !uid. Use gid_eq(gid, GLOBAL_ROOT_GID) instead of !gid. Use uid_eq(uid, INVALID_UID) instead of uid == -1 Use gid_eq(uid, INVALID_GID) instead of gid == -1 Use uid = GLOBAL_ROOT_UID instead of uid = 0; Use gid = GLOBAL_ROOT_GID instead of gid = 0; Use !uid_eq(uid1, uid2) instead of uid1 != uid2. Use !gid_eq(gid1, gid2) instead of gid1 != gid2. Use uid_eq(uid1, uid2) instead of uid1 == uid2. Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: "Eric W. Biederman" --- fs/nfsd/auth.c | 10 +++++----- fs/nfsd/nfs4recover.c | 4 ++-- fs/nfsd/nfs4state.c | 6 +++--- fs/nfsd/vfs.c | 8 ++++---- 4 files changed, 14 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/auth.c b/fs/nfsd/auth.c index 4d6642b38eae..06cddd572264 100644 --- a/fs/nfsd/auth.c +++ b/fs/nfsd/auth.c @@ -47,9 +47,9 @@ int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) if (!gi) goto oom; } else if (flags & NFSEXP_ROOTSQUASH) { - if (!new->fsuid) + if (uid_eq(new->fsuid, GLOBAL_ROOT_UID)) new->fsuid = exp->ex_anon_uid; - if (!new->fsgid) + if (gid_eq(new->fsgid, GLOBAL_ROOT_GID)) new->fsgid = exp->ex_anon_gid; gi = groups_alloc(rqgi->ngroups); @@ -66,9 +66,9 @@ int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) gi = get_group_info(rqgi); } - if (new->fsuid == (uid_t) -1) + if (uid_eq(new->fsuid, INVALID_UID)) new->fsuid = exp->ex_anon_uid; - if (new->fsgid == (gid_t) -1) + if (gid_eq(new->fsgid, INVALID_GID)) new->fsgid = exp->ex_anon_gid; ret = set_groups(new, gi); @@ -76,7 +76,7 @@ int nfsd_setuser(struct svc_rqst *rqstp, struct svc_export *exp) if (ret < 0) goto error; - if (new->fsuid) + if (!uid_eq(new->fsuid, GLOBAL_ROOT_UID)) new->cap_effective = cap_drop_nfsd_set(new->cap_effective); else new->cap_effective = cap_raise_nfsd_set(new->cap_effective, diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index ba6fdd4a0455..4914af4a817e 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -73,8 +73,8 @@ nfs4_save_creds(const struct cred **original_creds) if (!new) return -ENOMEM; - new->fsuid = 0; - new->fsgid = 0; + new->fsuid = GLOBAL_ROOT_UID; + new->fsgid = GLOBAL_ROOT_GID; *original_creds = override_creds(new); put_cred(new); return 0; diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index ac8ed96c4199..0af6d3c114ed 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1202,7 +1202,7 @@ static bool groups_equal(struct group_info *g1, struct group_info *g2) if (g1->ngroups != g2->ngroups) return false; for (i=0; ingroups; i++) - if (GROUP_AT(g1, i) != GROUP_AT(g2, i)) + if (!gid_eq(GROUP_AT(g1, i), GROUP_AT(g2, i))) return false; return true; } @@ -1227,8 +1227,8 @@ static bool same_creds(struct svc_cred *cr1, struct svc_cred *cr2) { if ((is_gss_cred(cr1) != is_gss_cred(cr2)) - || (cr1->cr_uid != cr2->cr_uid) - || (cr1->cr_gid != cr2->cr_gid) + || (!uid_eq(cr1->cr_uid, cr2->cr_uid)) + || (!gid_eq(cr1->cr_gid, cr2->cr_gid)) || !groups_equal(cr1->cr_group_info, cr2->cr_group_info)) return false; if (cr1->cr_principal == cr2->cr_principal) diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index d586117fa94a..31ff1d642e31 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -401,8 +401,8 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap, /* Revoke setuid/setgid on chown */ if (!S_ISDIR(inode->i_mode) && - (((iap->ia_valid & ATTR_UID) && iap->ia_uid != inode->i_uid) || - ((iap->ia_valid & ATTR_GID) && iap->ia_gid != inode->i_gid))) { + (((iap->ia_valid & ATTR_UID) && !uid_eq(iap->ia_uid, inode->i_uid)) || + ((iap->ia_valid & ATTR_GID) && !gid_eq(iap->ia_gid, inode->i_gid)))) { iap->ia_valid |= ATTR_KILL_PRIV; if (iap->ia_valid & ATTR_MODE) { /* we're setting mode too, just clear the s*id bits */ @@ -1205,7 +1205,7 @@ nfsd_create_setattr(struct svc_rqst *rqstp, struct svc_fh *resfhp, * send along the gid on create when it tries to implement * setgid directories via NFS: */ - if (current_fsuid() != 0) + if (!uid_eq(current_fsuid(), GLOBAL_ROOT_UID)) iap->ia_valid &= ~(ATTR_UID|ATTR_GID); if (iap->ia_valid) return nfsd_setattr(rqstp, resfhp, iap, 0, (time_t)0); @@ -2150,7 +2150,7 @@ nfsd_permission(struct svc_rqst *rqstp, struct svc_export *exp, * with NFSv3. */ if ((acc & NFSD_MAY_OWNER_OVERRIDE) && - inode->i_uid == current_fsuid()) + uid_eq(inode->i_uid, current_fsuid())) return 0; /* This assumes NFSD_MAY_{READ,WRITE,EXEC} == MAY_{READ,WRITE,EXEC} */ -- cgit From 46bbc25f9ff1242cd1c042580aec11ddb5e41369 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 5 Feb 2013 23:55:44 -0800 Subject: cifs: Override unmappable incoming uids and gids The cifs protocol has a 64bit space for uids and gids, while linux only supports a 32bit space today. Instead of silently truncating 64bit cifs ids, replace cifs ids that do not fit in the 32bit linux id space with the default uid and gids for the cifs mount. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/inode.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index ed6208ff85a7..d7ea2a6eaead 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -244,15 +244,19 @@ cifs_unix_basic_to_fattr(struct cifs_fattr *fattr, FILE_UNIX_BASIC_INFO *info, break; } - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) - fattr->cf_uid = cifs_sb->mnt_uid; - else - fattr->cf_uid = le64_to_cpu(info->Uid); - - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) - fattr->cf_gid = cifs_sb->mnt_gid; - else - fattr->cf_gid = le64_to_cpu(info->Gid); + fattr->cf_uid = cifs_sb->mnt_uid; + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID)) { + u64 id = le64_to_cpu(info->Uid); + if (id < ((uid_t)-1)) + fattr->cf_uid = id; + } + + fattr->cf_gid = cifs_sb->mnt_gid; + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID)) { + u64 id = le64_to_cpu(info->Gid); + if (id < ((gid_t)-1)) + fattr->cf_gid = id; + } fattr->cf_nlink = le64_to_cpu(info->Nlinks); } -- cgit From 355958f2893d97375da17696d2566dd4ab7b54a8 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 00:10:23 -0800 Subject: cifs: Use BUILD_BUG_ON to validate uids and gids are the same size The assumption that sizeof(uid_t) is the same as sizeof(gid_t) is completely reasonable but since we can verify the condition at compile time. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsacl.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index 5cbd00e74067..2e811f45b413 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -297,6 +297,7 @@ sid_to_id(struct cifs_sb_info *cifs_sb, struct cifs_sid *psid, * probably a safe assumption but might be better to check based on * sidtype. */ + BUILD_BUG_ON(sizeof(uid_t) != sizeof(gid_t)); if (sidkey->datalen != sizeof(uid_t)) { rc = -EIO; cFYI(1, "%s: Downcall contained malformed key " -- cgit From 8e3028b908a048ac0b3102504b7daad7b5544846 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 00:21:22 -0800 Subject: cifs: Pass GLOBAL_ROOT_UID and GLOBAL_ROOT_GID to keyring_alloc keyring_alloc has been updated to take a kuid_t and kgid_t so pass GLOBAL_ROOT_UID instead of 0 for the uid and GLOBAL_ROOT_GID instead of 0 for the gid. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsacl.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index 2e811f45b413..aaaf5ce142b8 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -347,7 +347,8 @@ init_cifs_idmap(void) if (!cred) return -ENOMEM; - keyring = keyring_alloc(".cifs_idmap", 0, 0, cred, + keyring = keyring_alloc(".cifs_idmap", + GLOBAL_ROOT_UID, GLOBAL_ROOT_GID, cred, (KEY_POS_ALL & ~KEY_POS_SETATTR) | KEY_USR_VIEW | KEY_USR_READ, KEY_ALLOC_NOT_IN_QUOTA, NULL); -- cgit From 8abf2775dd425ec3c767ea7c5a51b45fc8be76c2 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 00:33:17 -0800 Subject: cifs: Use kuids and kgids SID to uid/gid mapping Update id_mode_to_cifs_acl to take a kuid_t and a kgid_t. Replace NO_CHANGE_32 with INVALID_UID and INVALID_GID, and tests for NO_CHANGE_32 with uid_valid and gid_valid. Carefully unpack the value returned from request_key. memcpy the value into the expected type. The convert the uid/gid into a kuid/kgid. And then only if the result is a valid kuid or kgid update fuid/fgid. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsacl.c | 43 +++++++++++++++++++++++++++++-------------- fs/cifs/cifspdu.h | 1 - fs/cifs/cifsproto.h | 2 +- fs/cifs/inode.c | 8 ++++---- 4 files changed, 34 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index aaaf5ce142b8..f1e3f25fe004 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -266,8 +266,8 @@ sid_to_id(struct cifs_sb_info *cifs_sb, struct cifs_sid *psid, struct key *sidkey; char *sidstr; const struct cred *saved_cred; - uid_t fuid = cifs_sb->mnt_uid; - gid_t fgid = cifs_sb->mnt_gid; + kuid_t fuid = cifs_sb->mnt_uid; + kgid_t fgid = cifs_sb->mnt_gid; /* * If we have too many subauthorities, then something is really wrong. @@ -306,10 +306,21 @@ sid_to_id(struct cifs_sb_info *cifs_sb, struct cifs_sid *psid, goto out_key_put; } - if (sidtype == SIDOWNER) - memcpy(&fuid, &sidkey->payload.value, sizeof(uid_t)); - else - memcpy(&fgid, &sidkey->payload.value, sizeof(gid_t)); + if (sidtype == SIDOWNER) { + kuid_t uid; + uid_t id; + memcpy(&id, &sidkey->payload.value, sizeof(uid_t)); + uid = make_kuid(&init_user_ns, id); + if (uid_valid(uid)) + fuid = uid; + } else { + kgid_t gid; + gid_t id; + memcpy(&id, &sidkey->payload.value, sizeof(gid_t)); + gid = make_kgid(&init_user_ns, id); + if (gid_valid(gid)) + fgid = gid; + } out_key_put: key_put(sidkey); @@ -776,7 +787,7 @@ static int parse_sec_desc(struct cifs_sb_info *cifs_sb, /* Convert permission bits from mode to equivalent CIFS ACL */ static int build_sec_desc(struct cifs_ntsd *pntsd, struct cifs_ntsd *pnntsd, - __u32 secdesclen, __u64 nmode, uid_t uid, gid_t gid, int *aclflag) + __u32 secdesclen, __u64 nmode, kuid_t uid, kgid_t gid, int *aclflag) { int rc = 0; __u32 dacloffset; @@ -808,17 +819,19 @@ static int build_sec_desc(struct cifs_ntsd *pntsd, struct cifs_ntsd *pnntsd, *aclflag = CIFS_ACL_DACL; } else { memcpy(pnntsd, pntsd, secdesclen); - if (uid != NO_CHANGE_32) { /* chown */ + if (uid_valid(uid)) { /* chown */ + uid_t id; owner_sid_ptr = (struct cifs_sid *)((char *)pnntsd + le32_to_cpu(pnntsd->osidoffset)); nowner_sid_ptr = kmalloc(sizeof(struct cifs_sid), GFP_KERNEL); if (!nowner_sid_ptr) return -ENOMEM; - rc = id_to_sid(uid, SIDOWNER, nowner_sid_ptr); + id = from_kuid(&init_user_ns, uid); + rc = id_to_sid(id, SIDOWNER, nowner_sid_ptr); if (rc) { cFYI(1, "%s: Mapping error %d for owner id %d", - __func__, rc, uid); + __func__, rc, id); kfree(nowner_sid_ptr); return rc; } @@ -826,17 +839,19 @@ static int build_sec_desc(struct cifs_ntsd *pntsd, struct cifs_ntsd *pnntsd, kfree(nowner_sid_ptr); *aclflag = CIFS_ACL_OWNER; } - if (gid != NO_CHANGE_32) { /* chgrp */ + if (gid_valid(gid)) { /* chgrp */ + gid_t id; group_sid_ptr = (struct cifs_sid *)((char *)pnntsd + le32_to_cpu(pnntsd->gsidoffset)); ngroup_sid_ptr = kmalloc(sizeof(struct cifs_sid), GFP_KERNEL); if (!ngroup_sid_ptr) return -ENOMEM; - rc = id_to_sid(gid, SIDGROUP, ngroup_sid_ptr); + id = from_kgid(&init_user_ns, gid); + rc = id_to_sid(id, SIDGROUP, ngroup_sid_ptr); if (rc) { cFYI(1, "%s: Mapping error %d for group id %d", - __func__, rc, gid); + __func__, rc, id); kfree(ngroup_sid_ptr); return rc; } @@ -1004,7 +1019,7 @@ cifs_acl_to_fattr(struct cifs_sb_info *cifs_sb, struct cifs_fattr *fattr, /* Convert mode bits to an ACL so we can update the ACL on the server */ int id_mode_to_cifs_acl(struct inode *inode, const char *path, __u64 nmode, - uid_t uid, gid_t gid) + kuid_t uid, kgid_t gid) { int rc = 0; int aclflag = CIFS_ACL_DACL; /* default flag to set */ diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index b9d59a948a2c..e996ff6b26d1 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -277,7 +277,6 @@ #define CIFS_NO_HANDLE 0xFFFF #define NO_CHANGE_64 0xFFFFFFFFFFFFFFFFULL -#define NO_CHANGE_32 0xFFFFFFFFUL /* IPC$ in ASCII */ #define CIFS_IPC_RESOURCE "\x49\x50\x43\x24" diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 1988c1baa224..0060ec2124bf 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -161,7 +161,7 @@ extern int cifs_acl_to_fattr(struct cifs_sb_info *cifs_sb, struct cifs_fattr *fattr, struct inode *inode, const char *path, const __u16 *pfid); extern int id_mode_to_cifs_acl(struct inode *inode, const char *path, __u64, - uid_t, gid_t); + kuid_t, kgid_t); extern struct cifs_ntsd *get_cifs_acl(struct cifs_sb_info *, struct inode *, const char *, u32 *); extern int set_cifs_acl(struct cifs_ntsd *, __u32, struct inode *, diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index d7ea2a6eaead..d4cf7509c106 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -2090,8 +2090,8 @@ static int cifs_setattr_nounix(struct dentry *direntry, struct iattr *attrs) { unsigned int xid; - uid_t uid = NO_CHANGE_32; - gid_t gid = NO_CHANGE_32; + kuid_t uid = INVALID_UID; + kgid_t gid = INVALID_GID; struct inode *inode = direntry->d_inode; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsInodeInfo *cifsInode = CIFS_I(inode); @@ -2150,7 +2150,7 @@ cifs_setattr_nounix(struct dentry *direntry, struct iattr *attrs) #ifdef CONFIG_CIFS_ACL if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) { - if (uid != NO_CHANGE_32 || gid != NO_CHANGE_32) { + if (uid_valid(uid) || gid_valid(gid)) { rc = id_mode_to_cifs_acl(inode, full_path, NO_CHANGE_64, uid, gid); if (rc) { @@ -2174,7 +2174,7 @@ cifs_setattr_nounix(struct dentry *direntry, struct iattr *attrs) #ifdef CONFIG_CIFS_ACL if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) { rc = id_mode_to_cifs_acl(inode, full_path, mode, - NO_CHANGE_32, NO_CHANGE_32); + INVALID_UID, INVALID_GID); if (rc) { cFYI(1, "%s: Setting ACL failed with error: %d", __func__, rc); -- cgit From dbfb98af18194cff87d4c1dea8d43faf14eae2e7 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 00:44:13 -0800 Subject: cifs: Convert from a kuid before printing current_fsuid Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsproto.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 0060ec2124bf..feb9b6e2ab5a 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -46,7 +46,8 @@ extern void _free_xid(unsigned int); ({ \ unsigned int __xid = _get_xid(); \ cFYI(1, "CIFS VFS: in %s as Xid: %u with uid: %d", \ - __func__, __xid, current_fsuid()); \ + __func__, __xid, \ + from_kuid(&init_user_ns, current_fsuid())); \ __xid; \ }) -- cgit From 49418b2c28c901294f8b36ff14c766c9458c3623 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 00:57:56 -0800 Subject: cifs: Modify struct cifs_unix_set_info_args to hold a kuid_t and a kgid_t Use INVALID_UID and INVALID_GID instead of NO_CHANGE_64 to indicate the value should not be changed. In cifs_fill_unix_set_info convert from kuids and kgids into uids and gids that will fit in FILE_UNIX_BASIC_INFO. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsproto.h | 4 ++-- fs/cifs/cifssmb.c | 10 ++++++++-- fs/cifs/dir.c | 18 +++++++++--------- fs/cifs/file.c | 4 ++-- fs/cifs/inode.c | 14 +++++++------- 5 files changed, 28 insertions(+), 22 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index feb9b6e2ab5a..f450f0683ddd 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -305,8 +305,8 @@ struct cifs_unix_set_info_args { __u64 atime; __u64 mtime; __u64 mode; - __u64 uid; - __u64 gid; + kuid_t uid; + kgid_t gid; dev_t device; }; diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 76d0d2998850..00e12f2d626b 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -5819,8 +5819,14 @@ static void cifs_fill_unix_set_info(FILE_UNIX_BASIC_INFO *data_offset, const struct cifs_unix_set_info_args *args) { + u64 uid = NO_CHANGE_64, gid = NO_CHANGE_64; u64 mode = args->mode; + if (uid_valid(args->uid)) + uid = from_kuid(&init_user_ns, args->uid); + if (gid_valid(args->gid)) + gid = from_kgid(&init_user_ns, args->gid); + /* * Samba server ignores set of file size to zero due to bugs in some * older clients, but we should be precise - we use SetFileSize to @@ -5833,8 +5839,8 @@ cifs_fill_unix_set_info(FILE_UNIX_BASIC_INFO *data_offset, data_offset->LastStatusChange = cpu_to_le64(args->ctime); data_offset->LastAccessTime = cpu_to_le64(args->atime); data_offset->LastModificationTime = cpu_to_le64(args->mtime); - data_offset->Uid = cpu_to_le64(args->uid); - data_offset->Gid = cpu_to_le64(args->gid); + data_offset->Uid = cpu_to_le64(uid); + data_offset->Gid = cpu_to_le64(gid); /* better to leave device as zero when it is */ data_offset->DevMajor = cpu_to_le64(MAJOR(args->device)); data_offset->DevMinor = cpu_to_le64(MINOR(args->device)); diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 8719bbe0dcc3..1cd016217448 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -342,14 +342,14 @@ cifs_do_create(struct inode *inode, struct dentry *direntry, unsigned int xid, *created |= FILE_CREATED; if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) { - args.uid = (__u64) current_fsuid(); + args.uid = current_fsuid(); if (inode->i_mode & S_ISGID) - args.gid = (__u64) inode->i_gid; + args.gid = inode->i_gid; else - args.gid = (__u64) current_fsgid(); + args.gid = current_fsgid(); } else { - args.uid = NO_CHANGE_64; - args.gid = NO_CHANGE_64; + args.uid = INVALID_UID; /* no change */ + args.gid = INVALID_GID; /* no change */ } CIFSSMBUnixSetFileInfo(xid, tcon, &args, fid->netfid, current->tgid); @@ -588,11 +588,11 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, umode_t mode, .device = device_number, }; if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) { - args.uid = (__u64) current_fsuid(); - args.gid = (__u64) current_fsgid(); + args.uid = current_fsuid(); + args.gid = current_fsgid(); } else { - args.uid = NO_CHANGE_64; - args.gid = NO_CHANGE_64; + args.uid = INVALID_UID; /* no change */ + args.gid = INVALID_GID; /* no change */ } rc = CIFSSMBUnixSetPathInfo(xid, pTcon, full_path, &args, cifs_sb->local_nls, diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 0a6677ba212b..b9baf5f66349 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -487,8 +487,8 @@ int cifs_open(struct inode *inode, struct file *file) */ struct cifs_unix_set_info_args args = { .mode = inode->i_mode, - .uid = NO_CHANGE_64, - .gid = NO_CHANGE_64, + .uid = INVALID_UID, /* no change */ + .gid = INVALID_GID, /* no change */ .ctime = NO_CHANGE_64, .atime = NO_CHANGE_64, .mtime = NO_CHANGE_64, diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index d4cf7509c106..737643940540 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1249,14 +1249,14 @@ cifs_mkdir_qinfo(struct inode *parent, struct dentry *dentry, umode_t mode, .device = 0, }; if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) { - args.uid = (__u64)current_fsuid(); + args.uid = current_fsuid(); if (parent->i_mode & S_ISGID) - args.gid = (__u64)parent->i_gid; + args.gid = parent->i_gid; else - args.gid = (__u64)current_fsgid(); + args.gid = current_fsgid(); } else { - args.uid = NO_CHANGE_64; - args.gid = NO_CHANGE_64; + args.uid = INVALID_UID; /* no change */ + args.gid = INVALID_GID; /* no change */ } CIFSSMBUnixSetPathInfo(xid, tcon, full_path, &args, cifs_sb->local_nls, @@ -2017,12 +2017,12 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs) if (attrs->ia_valid & ATTR_UID) args->uid = attrs->ia_uid; else - args->uid = NO_CHANGE_64; + args->uid = INVALID_UID; /* no change */ if (attrs->ia_valid & ATTR_GID) args->gid = attrs->ia_gid; else - args->gid = NO_CHANGE_64; + args->gid = INVALID_GID; /* no change */ if (attrs->ia_valid & ATTR_ATIME) args->atime = cifs_UnixTimeToNT(attrs->ia_atime); -- cgit From 6d4a083205c27bc2e2a1c03dff4acf2fe457d1c4 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 01:48:56 -0800 Subject: cifs: Convert struct tcon_link to use a kuid. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsglob.h | 2 +- fs/cifs/connect.c | 12 ++++++------ 2 files changed, 7 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index aea1eec64911..7993a7a07ae2 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -837,7 +837,7 @@ struct cifs_tcon { */ struct tcon_link { struct rb_node tl_rbnode; - uid_t tl_uid; + kuid_t tl_uid; unsigned long tl_flags; #define TCON_LINK_MASTER 0 #define TCON_LINK_PENDING 1 diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 17c3643e5950..541169459223 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -3919,7 +3919,7 @@ cifs_set_vol_auth(struct smb_vol *vol, struct cifs_ses *ses) } static struct cifs_tcon * -cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid) +cifs_construct_tcon(struct cifs_sb_info *cifs_sb, kuid_t fsuid) { int rc; struct cifs_tcon *master_tcon = cifs_sb_master_tcon(cifs_sb); @@ -3989,7 +3989,7 @@ cifs_sb_tcon_pending_wait(void *unused) /* find and return a tlink with given uid */ static struct tcon_link * -tlink_rb_search(struct rb_root *root, uid_t uid) +tlink_rb_search(struct rb_root *root, kuid_t uid) { struct rb_node *node = root->rb_node; struct tcon_link *tlink; @@ -3997,9 +3997,9 @@ tlink_rb_search(struct rb_root *root, uid_t uid) while (node) { tlink = rb_entry(node, struct tcon_link, tl_rbnode); - if (tlink->tl_uid > uid) + if (uid_gt(tlink->tl_uid, uid)) node = node->rb_left; - else if (tlink->tl_uid < uid) + else if (uid_lt(tlink->tl_uid, uid)) node = node->rb_right; else return tlink; @@ -4018,7 +4018,7 @@ tlink_rb_insert(struct rb_root *root, struct tcon_link *new_tlink) tlink = rb_entry(*new, struct tcon_link, tl_rbnode); parent = *new; - if (tlink->tl_uid > new_tlink->tl_uid) + if (uid_gt(tlink->tl_uid, new_tlink->tl_uid)) new = &((*new)->rb_left); else new = &((*new)->rb_right); @@ -4048,7 +4048,7 @@ struct tcon_link * cifs_sb_tlink(struct cifs_sb_info *cifs_sb) { int ret; - uid_t fsuid = current_fsuid(); + kuid_t fsuid = current_fsuid(); struct tcon_link *tlink, *newtlink; if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER)) -- cgit From 4a2c8cf56953a6ebe3c8671433607b7f96f200d5 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 01:53:25 -0800 Subject: cifs: Convert struct cifs_fattr to use kuid and kgids In cifs_unix_to_basic_fattr only update the cifs_fattr with an id if it is valid after conversion. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsglob.h | 4 ++-- fs/cifs/inode.c | 14 ++++++++++---- 2 files changed, 12 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 7993a7a07ae2..98312fda0d35 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -1245,8 +1245,8 @@ struct cifs_fattr { u64 cf_eof; u64 cf_bytes; u64 cf_createtime; - uid_t cf_uid; - gid_t cf_gid; + kuid_t cf_uid; + kgid_t cf_gid; umode_t cf_mode; dev_t cf_rdev; unsigned int cf_nlink; diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 737643940540..9638233964fc 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -247,15 +247,21 @@ cifs_unix_basic_to_fattr(struct cifs_fattr *fattr, FILE_UNIX_BASIC_INFO *info, fattr->cf_uid = cifs_sb->mnt_uid; if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID)) { u64 id = le64_to_cpu(info->Uid); - if (id < ((uid_t)-1)) - fattr->cf_uid = id; + if (id < ((uid_t)-1)) { + kuid_t uid = make_kuid(&init_user_ns, id); + if (uid_valid(uid)) + fattr->cf_uid = uid; + } } fattr->cf_gid = cifs_sb->mnt_gid; if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID)) { u64 id = le64_to_cpu(info->Gid); - if (id < ((gid_t)-1)) - fattr->cf_gid = id; + if (id < ((gid_t)-1)) { + kgid_t gid = make_kgid(&init_user_ns, id); + if (gid_valid(gid)) + fattr->cf_gid = gid; + } } fattr->cf_nlink = le64_to_cpu(info->Nlinks); -- cgit From fef59fd728366aa9bf125b8859aff84fa0bd1a36 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 02:23:02 -0800 Subject: cifs: Convert struct cifsFileInfo to use a kuid Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsglob.h | 2 +- fs/cifs/file.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 98312fda0d35..04aa74e5b348 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -930,7 +930,7 @@ struct cifsFileInfo { struct list_head tlist; /* pointer to next fid owned by tcon */ struct list_head flist; /* next fid (file instance) for this inode */ struct cifs_fid_locks *llist; /* brlocks held by this fid */ - unsigned int uid; /* allows finding which FileInfo structure */ + kuid_t uid; /* allows finding which FileInfo structure */ __u32 pid; /* process id who opened file */ struct cifs_fid fid; /* file id from remote */ /* BB add lock scope info here if needed */ ; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index b9baf5f66349..c23fbd81fe1a 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -1649,7 +1649,7 @@ struct cifsFileInfo *find_readable_file(struct cifsInodeInfo *cifs_inode, are always at the end of the list but since the first entry might have a close pending, we go through the whole list */ list_for_each_entry(open_file, &cifs_inode->openFileList, flist) { - if (fsuid_only && open_file->uid != current_fsuid()) + if (fsuid_only && !uid_eq(open_file->uid, current_fsuid())) continue; if (OPEN_FMODE(open_file->f_flags) & FMODE_READ) { if (!open_file->invalidHandle) { @@ -1702,7 +1702,7 @@ refind_writable: list_for_each_entry(open_file, &cifs_inode->openFileList, flist) { if (!any_available && open_file->pid != current->tgid) continue; - if (fsuid_only && open_file->uid != current_fsuid()) + if (fsuid_only && !uid_eq(open_file->uid, current_fsuid())) continue; if (OPEN_FMODE(open_file->f_flags) & FMODE_WRITE) { if (!open_file->invalidHandle) { -- cgit From 3da46565043a3d7b9fd5c924e7de2d3e65e9d2a9 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 01:37:39 -0800 Subject: cifs: Modify struct smb_vol to use kuids and kgids Add two helper functions get_option_uid and get_option_gid to handle the work of parsing uid and gids paramaters from the command line and making kuids and kgids out of them. Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsglob.h | 10 +++++----- fs/cifs/connect.c | 50 ++++++++++++++++++++++++++++++++++++++++---------- 2 files changed, 45 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 04aa74e5b348..e308e8bd0772 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -399,11 +399,11 @@ struct smb_vol { char *iocharset; /* local code page for mapping to and from Unicode */ char source_rfc1001_name[RFC1001_NAME_LEN_WITH_NULL]; /* clnt nb name */ char target_rfc1001_name[RFC1001_NAME_LEN_WITH_NULL]; /* srvr nb name */ - uid_t cred_uid; - uid_t linux_uid; - gid_t linux_gid; - uid_t backupuid; - gid_t backupgid; + kuid_t cred_uid; + kuid_t linux_uid; + kgid_t linux_gid; + kuid_t backupuid; + kgid_t backupgid; umode_t file_mode; umode_t dir_mode; unsigned secFlg; diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 541169459223..9bd13a75602d 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -987,6 +987,41 @@ static int get_option_ul(substring_t args[], unsigned long *option) return rc; } +static int get_option_uid(substring_t args[], kuid_t *result) +{ + unsigned long value; + kuid_t uid; + int rc; + + rc = get_option_ul(args, &value); + if (rc) + return rc; + + uid = make_kuid(current_user_ns(), value); + if (!uid_valid(uid)) + return -EINVAL; + + *result = uid; + return 0; +} + +static int get_option_gid(substring_t args[], kgid_t *result) +{ + unsigned long value; + kgid_t gid; + int rc; + + rc = get_option_ul(args, &value); + if (rc) + return rc; + + gid = make_kgid(current_user_ns(), value); + if (!gid_valid(gid)) + return -EINVAL; + + *result = gid; + return 0; +} static int cifs_parse_security_flavors(char *value, struct smb_vol *vol) @@ -1424,47 +1459,42 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, /* Numeric Values */ case Opt_backupuid: - if (get_option_ul(args, &option)) { + if (get_option_uid(args, &vol->backupuid)) { cERROR(1, "%s: Invalid backupuid value", __func__); goto cifs_parse_mount_err; } - vol->backupuid = option; vol->backupuid_specified = true; break; case Opt_backupgid: - if (get_option_ul(args, &option)) { + if (get_option_gid(args, &vol->backupgid)) { cERROR(1, "%s: Invalid backupgid value", __func__); goto cifs_parse_mount_err; } - vol->backupgid = option; vol->backupgid_specified = true; break; case Opt_uid: - if (get_option_ul(args, &option)) { + if (get_option_uid(args, &vol->linux_uid)) { cERROR(1, "%s: Invalid uid value", __func__); goto cifs_parse_mount_err; } - vol->linux_uid = option; uid_specified = true; break; case Opt_cruid: - if (get_option_ul(args, &option)) { + if (get_option_uid(args, &vol->cred_uid)) { cERROR(1, "%s: Invalid cruid value", __func__); goto cifs_parse_mount_err; } - vol->cred_uid = option; break; case Opt_gid: - if (get_option_ul(args, &option)) { + if (get_option_gid(args, &vol->linux_gid)) { cERROR(1, "%s: Invalid gid value", __func__); goto cifs_parse_mount_err; } - vol->linux_gid = option; gid_specified = true; break; case Opt_file_mode: -- cgit From 1f68233c52e9f2bb53130a0063bc1e6864f6d204 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 01:20:20 -0800 Subject: cifs: Convert struct cifs_sb_info to use kuids and kgids Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifs_fs_sb.h | 8 ++++---- fs/cifs/cifsfs.c | 14 ++++++++++---- fs/cifs/connect.c | 2 +- fs/cifs/misc.c | 2 +- 4 files changed, 16 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifs_fs_sb.h b/fs/cifs/cifs_fs_sb.h index c865bfdfe819..37e4a72a7d1c 100644 --- a/fs/cifs/cifs_fs_sb.h +++ b/fs/cifs/cifs_fs_sb.h @@ -55,10 +55,10 @@ struct cifs_sb_info { unsigned int wsize; unsigned long actimeo; /* attribute cache timeout (jiffies) */ atomic_t active; - uid_t mnt_uid; - gid_t mnt_gid; - uid_t mnt_backupuid; - gid_t mnt_backupgid; + kuid_t mnt_uid; + kgid_t mnt_gid; + kuid_t mnt_backupuid; + kgid_t mnt_backupgid; umode_t mnt_file_mode; umode_t mnt_dir_mode; unsigned int mnt_cifs_flags; diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index f653835d067b..17590731786d 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -376,13 +376,15 @@ cifs_show_options(struct seq_file *s, struct dentry *root) (int)(srcaddr->sa_family)); } - seq_printf(s, ",uid=%u", cifs_sb->mnt_uid); + seq_printf(s, ",uid=%u", + from_kuid_munged(&init_user_ns, cifs_sb->mnt_uid)); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) seq_printf(s, ",forceuid"); else seq_printf(s, ",noforceuid"); - seq_printf(s, ",gid=%u", cifs_sb->mnt_gid); + seq_printf(s, ",gid=%u", + from_kgid_munged(&init_user_ns, cifs_sb->mnt_gid)); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) seq_printf(s, ",forcegid"); else @@ -437,9 +439,13 @@ cifs_show_options(struct seq_file *s, struct dentry *root) if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_PERM) seq_printf(s, ",noperm"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_BACKUPUID) - seq_printf(s, ",backupuid=%u", cifs_sb->mnt_backupuid); + seq_printf(s, ",backupuid=%u", + from_kuid_munged(&init_user_ns, + cifs_sb->mnt_backupuid)); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_BACKUPGID) - seq_printf(s, ",backupgid=%u", cifs_sb->mnt_backupgid); + seq_printf(s, ",backupgid=%u", + from_kgid_munged(&init_user_ns, + cifs_sb->mnt_backupgid)); seq_printf(s, ",rsize=%u", cifs_sb->rsize); seq_printf(s, ",wsize=%u", cifs_sb->wsize); diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 9bd13a75602d..01279b8f5d1f 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -2743,7 +2743,7 @@ compare_mount_options(struct super_block *sb, struct cifs_mnt_data *mnt_data) if (new->rsize && new->rsize < old->rsize) return 0; - if (old->mnt_uid != new->mnt_uid || old->mnt_gid != new->mnt_gid) + if (!uid_eq(old->mnt_uid, new->mnt_uid) || !gid_eq(old->mnt_gid, new->mnt_gid)) return 0; if (old->mnt_file_mode != new->mnt_file_mode || diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 3a00c0d0cead..1b15bf839f37 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -569,7 +569,7 @@ bool backup_cred(struct cifs_sb_info *cifs_sb) { if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_BACKUPUID) { - if (cifs_sb->mnt_backupuid == current_fsuid()) + if (uid_eq(cifs_sb->mnt_backupuid, current_fsuid())) return true; } if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_BACKUPGID) { -- cgit From 64ed39dd1ef284c0338799a6167b77a6d6e01982 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 6 Feb 2013 02:30:39 -0800 Subject: cifs: Convert struct cifs_ses to use a kuid_t and a kgid_t Cc: Steve French Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifs_spnego.c | 6 ++++-- fs/cifs/cifsglob.h | 4 ++-- fs/cifs/connect.c | 2 +- 3 files changed, 7 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c index 086f381d6489..10e774761299 100644 --- a/fs/cifs/cifs_spnego.c +++ b/fs/cifs/cifs_spnego.c @@ -149,10 +149,12 @@ cifs_get_spnego_key(struct cifs_ses *sesInfo) goto out; dp = description + strlen(description); - sprintf(dp, ";uid=0x%x", sesInfo->linux_uid); + sprintf(dp, ";uid=0x%x", + from_kuid_munged(&init_user_ns, sesInfo->linux_uid)); dp = description + strlen(description); - sprintf(dp, ";creduid=0x%x", sesInfo->cred_uid); + sprintf(dp, ";creduid=0x%x", + from_kuid_munged(&init_user_ns, sesInfo->cred_uid)); if (sesInfo->user_name) { dp = description + strlen(description); diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index e308e8bd0772..aaef57beba0e 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -702,8 +702,8 @@ struct cifs_ses { char *serverNOS; /* name of network operating system of server */ char *serverDomain; /* security realm of server */ __u64 Suid; /* remote smb uid */ - uid_t linux_uid; /* overriding owner of files on the mount */ - uid_t cred_uid; /* owner of credentials */ + kuid_t linux_uid; /* overriding owner of files on the mount */ + kuid_t cred_uid; /* owner of credentials */ unsigned int capabilities; char serverName[SERVER_NAME_LEN_WITH_NULL * 2]; /* BB make bigger for TCP names - will ipv6 and sctp addresses fit? */ diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 01279b8f5d1f..067b0e169fe4 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -2271,7 +2271,7 @@ static int match_session(struct cifs_ses *ses, struct smb_vol *vol) { switch (ses->server->secType) { case Kerberos: - if (vol->cred_uid != ses->cred_uid) + if (!uid_eq(vol->cred_uid, ses->cred_uid)) return 0; break; default: -- cgit From 39b648d9ec7d4ab0b4362872c6284a12c582afa6 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Thu, 31 Jan 2013 11:53:05 -0800 Subject: ceph: remove 'ceph.layout' virtual xattr This has been deprecated since v3.3, 114fc474. Kill it. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 7 ------- 1 file changed, 7 deletions(-) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 2c2ae5be9902..c2048b1a5395 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -123,13 +123,6 @@ static size_t ceph_vxattrcb_file_layout(struct ceph_inode_info *ci, char *val, static struct ceph_vxattr ceph_file_vxattrs[] = { XATTR_NAME_CEPH(file, layout), - /* The following extended attribute name is deprecated */ - { - .name = XATTR_CEPH_PREFIX "layout", - .name_size = sizeof (XATTR_CEPH_PREFIX "layout"), - .getxattr_cb = ceph_vxattrcb_file_layout, - .readonly = true, - }, { 0 } /* Required table terminator */ }; static size_t ceph_file_vxattrs_name_size; /* total size of all names */ -- cgit From 8860147a01c4243f64f7d602dbf8342ca616ed45 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Thu, 31 Jan 2013 11:53:27 -0800 Subject: ceph: support hidden vxattrs Add ability to flag virtual xattrs as hidden, such that you can getxattr them but they do not appear in listxattr. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index c2048b1a5395..43063d0dee8f 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -29,7 +29,7 @@ struct ceph_vxattr { size_t name_size; /* strlen(name) + 1 (for '\0') */ size_t (*getxattr_cb)(struct ceph_inode_info *ci, char *val, size_t size); - bool readonly; + bool readonly, hidden; }; /* directories */ @@ -85,13 +85,14 @@ static size_t ceph_vxattrcb_dir_rctime(struct ceph_inode_info *ci, char *val, #define CEPH_XATTR_NAME(_type, _name) XATTR_CEPH_PREFIX #_type "." #_name -#define XATTR_NAME_CEPH(_type, _name) \ - { \ - .name = CEPH_XATTR_NAME(_type, _name), \ - .name_size = sizeof (CEPH_XATTR_NAME(_type, _name)), \ - .getxattr_cb = ceph_vxattrcb_ ## _type ## _ ## _name, \ - .readonly = true, \ - } +#define XATTR_NAME_CEPH(_type, _name) \ + { \ + .name = CEPH_XATTR_NAME(_type, _name), \ + .name_size = sizeof (CEPH_XATTR_NAME(_type, _name)), \ + .getxattr_cb = ceph_vxattrcb_ ## _type ## _ ## _name, \ + .readonly = true, \ + .hidden = false, \ + } static struct ceph_vxattr ceph_dir_vxattrs[] = { XATTR_NAME_CEPH(dir, entries), @@ -157,7 +158,8 @@ static size_t __init vxattrs_name_size(struct ceph_vxattr *vxattrs) size_t size = 0; for (vxattr = vxattrs; vxattr->name; vxattr++) - size += vxattr->name_size; + if (!vxattr->hidden) + size += vxattr->name_size; return size; } -- cgit From 3adf654ddbc355c23d75c6684128d4b067a7b792 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Thu, 31 Jan 2013 11:53:41 -0800 Subject: ceph: pass unhandled ceph.* setxattrs through to MDS If we do not specifically understand a setxattr on a ceph.* virtual xattr, send it through to the MDS. This allows us to implement new functionality via the MDS without direct support on the client side. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 43063d0dee8f..edc47de77fed 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -777,6 +777,10 @@ int ceph_setxattr(struct dentry *dentry, const char *name, if (vxattr && vxattr->readonly) return -EOPNOTSUPP; + /* pass any unhandled ceph.* xattrs through to the MDS */ + if (!strncmp(name, XATTR_CEPH_PREFIX, XATTR_CEPH_PREFIX_LEN)) + goto do_sync_unlocked; + /* preallocate memory for xattr name, value, index node */ err = -ENOMEM; newname = kmemdup(name, name_len + 1, GFP_NOFS); @@ -833,6 +837,7 @@ retry: do_sync: spin_unlock(&ci->i_ceph_lock); +do_sync_unlocked: err = ceph_sync_setxattr(dentry, name, value, size, flags); out: kfree(newname); -- cgit From d421acb1ad7dfa31b7463b67f1593714b0b727c3 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sun, 20 Jan 2013 21:55:30 -0800 Subject: ceph: pass ceph.* removexattrs through to MDS If we do not explicitly recognized a vxattr (e.g., as readonly), pass the request through to the MDS and deal with it there. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index edc47de77fed..234270f00c2a 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -892,6 +892,10 @@ int ceph_removexattr(struct dentry *dentry, const char *name) if (vxattr && vxattr->readonly) return -EOPNOTSUPP; + /* pass any unhandled ceph.* xattrs through to the MDS */ + if (!strncmp(name, XATTR_CEPH_PREFIX, XATTR_CEPH_PREFIX_LEN)) + goto do_sync_unlocked; + err = -ENOMEM; spin_lock(&ci->i_ceph_lock); retry: @@ -931,6 +935,7 @@ retry: return err; do_sync: spin_unlock(&ci->i_ceph_lock); +do_sync_unlocked: err = ceph_send_removexattr(dentry, name); out: return err; -- cgit From f36e4472969a78ae65e514b553e9a0feacb40a28 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sun, 20 Jan 2013 21:59:29 -0800 Subject: ceph: add exists_cb to vxattr struct Allow for a callback to dynamically determine if a vxattr exists for the given inode. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 234270f00c2a..06344da4e968 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -30,6 +30,7 @@ struct ceph_vxattr { size_t (*getxattr_cb)(struct ceph_inode_info *ci, char *val, size_t size); bool readonly, hidden; + bool (*exists_cb)(struct ceph_inode_info *ci); }; /* directories */ @@ -92,6 +93,7 @@ static size_t ceph_vxattrcb_dir_rctime(struct ceph_inode_info *ci, char *val, .getxattr_cb = ceph_vxattrcb_ ## _type ## _ ## _name, \ .readonly = true, \ .hidden = false, \ + .exists_cb = NULL, \ } static struct ceph_vxattr ceph_dir_vxattrs[] = { -- cgit From 0bee82fb4b8d49541fe474ed460d2b917f329568 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sun, 20 Jan 2013 22:00:58 -0800 Subject: ceph: fix getxattr vxattr handling Change the vxattr handling for getxattr so that vxattrs are checked prior to any xattr content, and never after. Enforce vxattr existence via the exists_cb callback. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 06344da4e968..87b85f3403d4 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -569,13 +569,17 @@ ssize_t ceph_getxattr(struct dentry *dentry, const char *name, void *value, if (!ceph_is_valid_xattr(name)) return -ENODATA; - /* let's see if a virtual xattr was requested */ - vxattr = ceph_match_vxattr(inode, name); - spin_lock(&ci->i_ceph_lock); dout("getxattr %p ver=%lld index_ver=%lld\n", inode, ci->i_xattrs.version, ci->i_xattrs.index_version); + /* let's see if a virtual xattr was requested */ + vxattr = ceph_match_vxattr(inode, name); + if (vxattr && !(vxattr->exists_cb && !vxattr->exists_cb(ci))) { + err = vxattr->getxattr_cb(ci, value, size); + goto out; + } + if (__ceph_caps_issued_mask(ci, CEPH_CAP_XATTR_SHARED, 1) && (ci->i_xattrs.index_version >= ci->i_xattrs.version)) { goto get_xattr; @@ -589,11 +593,6 @@ ssize_t ceph_getxattr(struct dentry *dentry, const char *name, void *value, spin_lock(&ci->i_ceph_lock); - if (vxattr && vxattr->readonly) { - err = vxattr->getxattr_cb(ci, value, size); - goto out; - } - err = __build_xattrs(inode); if (err < 0) goto out; @@ -601,11 +600,8 @@ ssize_t ceph_getxattr(struct dentry *dentry, const char *name, void *value, get_xattr: err = -ENODATA; /* == ENOATTR */ xattr = __get_xattr(ci, name); - if (!xattr) { - if (vxattr) - err = vxattr->getxattr_cb(ci, value, size); + if (!xattr) goto out; - } err = -ERANGE; if (size && size < xattr->val_len) -- cgit From b65917dd2700b7d12e25e2e0bbfd58eb3c932158 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sun, 20 Jan 2013 22:02:39 -0800 Subject: ceph: fix listxattr handling for vxattrs Only include vxattrs in the result if they are not hidden and exist (as determined by the exists_cb callback). Note that the buffer size we return when 0 is passed in always includes vxattrs that *might* exist, forming an upper bound. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 87b85f3403d4..ec09ea5c4f07 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -657,23 +657,30 @@ list_xattr: vir_namelen = ceph_vxattrs_name_size(vxattrs); /* adding 1 byte per each variable due to the null termination */ - namelen = vir_namelen + ci->i_xattrs.names_size + ci->i_xattrs.count; + namelen = ci->i_xattrs.names_size + ci->i_xattrs.count; err = -ERANGE; - if (size && namelen > size) + if (size && vir_namelen + namelen > size) goto out; - err = namelen; + err = namelen + vir_namelen; if (size == 0) goto out; names = __copy_xattr_names(ci, names); /* virtual xattr names, too */ - if (vxattrs) + err = namelen; + if (vxattrs) { for (i = 0; vxattrs[i].name; i++) { - len = sprintf(names, "%s", vxattrs[i].name); - names += len + 1; + if (!vxattrs[i].hidden && + !(vxattrs[i].exists_cb && + !vxattrs[i].exists_cb(ci))) { + len = sprintf(names, "%s", vxattrs[i].name); + names += len + 1; + err += len + 1; + } } + } out: spin_unlock(&ci->i_ceph_lock); -- cgit From 32ab0bd78d7d9235efb38ad5cba6a3a6b39a1da6 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sat, 19 Jan 2013 16:46:32 -0800 Subject: ceph: change ceph.file.layout.* implementation, content Implement a new method to generate the ceph.file.layout vxattr using the new framework. Use 'stripe_unit' instead of 'chunk_size'. Include pool name, either as a string or as an integer. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 67 +++++++++++++++++++++++++++++++++++++++++++++------------ 1 file changed, 53 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index ec09ea5c4f07..532c95a6c9fa 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -33,6 +33,50 @@ struct ceph_vxattr { bool (*exists_cb)(struct ceph_inode_info *ci); }; +/* layouts */ + +static bool ceph_vxattrcb_layout_exists(struct ceph_inode_info *ci) +{ + size_t s; + char *p = (char *)&ci->i_layout; + + for (s = 0; s < sizeof(ci->i_layout); s++, p++) + if (*p) + return true; + return false; +} + +static size_t ceph_vxattrcb_layout(struct ceph_inode_info *ci, char *val, + size_t size) +{ + int ret; + struct ceph_fs_client *fsc = ceph_sb_to_client(ci->vfs_inode.i_sb); + struct ceph_osd_client *osdc = &fsc->client->osdc; + s64 pool = ceph_file_layout_pg_pool(ci->i_layout); + const char *pool_name; + + dout("ceph_vxattrcb_layout %p\n", &ci->vfs_inode); + down_read(&osdc->map_sem); + pool_name = ceph_pg_pool_name_by_id(osdc->osdmap, pool); + if (pool_name) + ret = snprintf(val, size, + "stripe_unit=%lld stripe_count=%lld object_size=%lld pool=%s", + (unsigned long long)ceph_file_layout_su(ci->i_layout), + (unsigned long long)ceph_file_layout_stripe_count(ci->i_layout), + (unsigned long long)ceph_file_layout_object_size(ci->i_layout), + pool_name); + else + ret = snprintf(val, size, + "stripe_unit=%lld stripe_count=%lld object_size=%lld pool=%lld", + (unsigned long long)ceph_file_layout_su(ci->i_layout), + (unsigned long long)ceph_file_layout_stripe_count(ci->i_layout), + (unsigned long long)ceph_file_layout_object_size(ci->i_layout), + (unsigned long long)pool); + + up_read(&osdc->map_sem); + return ret; +} + /* directories */ static size_t ceph_vxattrcb_dir_entries(struct ceph_inode_info *ci, char *val, @@ -84,6 +128,7 @@ static size_t ceph_vxattrcb_dir_rctime(struct ceph_inode_info *ci, char *val, (long)ci->i_rctime.tv_nsec); } + #define CEPH_XATTR_NAME(_type, _name) XATTR_CEPH_PREFIX #_type "." #_name #define XATTR_NAME_CEPH(_type, _name) \ @@ -111,21 +156,15 @@ static size_t ceph_dir_vxattrs_name_size; /* total size of all names */ /* files */ -static size_t ceph_vxattrcb_file_layout(struct ceph_inode_info *ci, char *val, - size_t size) -{ - int ret; - - ret = snprintf(val, size, - "chunk_bytes=%lld\nstripe_count=%lld\nobject_size=%lld\n", - (unsigned long long)ceph_file_layout_su(ci->i_layout), - (unsigned long long)ceph_file_layout_stripe_count(ci->i_layout), - (unsigned long long)ceph_file_layout_object_size(ci->i_layout)); - return ret; -} - static struct ceph_vxattr ceph_file_vxattrs[] = { - XATTR_NAME_CEPH(file, layout), + { + .name = "ceph.file.layout", + .name_size = sizeof("ceph.file.layout"), + .getxattr_cb = ceph_vxattrcb_layout, + .readonly = false, + .hidden = false, + .exists_cb = ceph_vxattrcb_layout_exists, + }, { 0 } /* Required table terminator */ }; static size_t ceph_file_vxattrs_name_size; /* total size of all names */ -- cgit From 1f08f2b056ea6c2e12f4e95e88949a882a996208 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sun, 20 Jan 2013 22:07:12 -0800 Subject: ceph: add ceph.dir.layout vxattr This virtual xattr will only appear when there is a dir layout policy set on the directory. It can be set via setxattr and removed via removexattr (implemented by the MDS). Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 532c95a6c9fa..f3c4fe7202c7 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -142,6 +142,14 @@ static size_t ceph_vxattrcb_dir_rctime(struct ceph_inode_info *ci, char *val, } static struct ceph_vxattr ceph_dir_vxattrs[] = { + { + .name = "ceph.dir.layout", + .name_size = sizeof("ceph.dir.layout"), + .getxattr_cb = ceph_vxattrcb_layout, + .readonly = false, + .hidden = false, + .exists_cb = ceph_vxattrcb_layout_exists, + }, XATTR_NAME_CEPH(dir, entries), XATTR_NAME_CEPH(dir, files), XATTR_NAME_CEPH(dir, subdirs), -- cgit From 695b711933689ea51af782760f4b1e2c6a42a631 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sun, 20 Jan 2013 22:08:33 -0800 Subject: ceph: implement hidden per-field ceph.*.layout.* vxattrs Allow individual fields of the layout to be fetched via getxattr. The ceph.dir.layout.* vxattr with "disappear" if the exists_cb indicates there no dir layout set. Signed-off-by: Sage Weil Reviewed-by: Sam Lang --- fs/ceph/xattr.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 59 insertions(+) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index f3c4fe7202c7..2135817e708d 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -77,6 +77,46 @@ static size_t ceph_vxattrcb_layout(struct ceph_inode_info *ci, char *val, return ret; } +static size_t ceph_vxattrcb_layout_stripe_unit(struct ceph_inode_info *ci, + char *val, size_t size) +{ + return snprintf(val, size, "%lld", + (unsigned long long)ceph_file_layout_su(ci->i_layout)); +} + +static size_t ceph_vxattrcb_layout_stripe_count(struct ceph_inode_info *ci, + char *val, size_t size) +{ + return snprintf(val, size, "%lld", + (unsigned long long)ceph_file_layout_stripe_count(ci->i_layout)); +} + +static size_t ceph_vxattrcb_layout_object_size(struct ceph_inode_info *ci, + char *val, size_t size) +{ + return snprintf(val, size, "%lld", + (unsigned long long)ceph_file_layout_object_size(ci->i_layout)); +} + +static size_t ceph_vxattrcb_layout_pool(struct ceph_inode_info *ci, + char *val, size_t size) +{ + int ret; + struct ceph_fs_client *fsc = ceph_sb_to_client(ci->vfs_inode.i_sb); + struct ceph_osd_client *osdc = &fsc->client->osdc; + s64 pool = ceph_file_layout_pg_pool(ci->i_layout); + const char *pool_name; + + down_read(&osdc->map_sem); + pool_name = ceph_pg_pool_name_by_id(osdc->osdmap, pool); + if (pool_name) + ret = snprintf(val, size, "%s", pool_name); + else + ret = snprintf(val, size, "%lld", (unsigned long long)pool); + up_read(&osdc->map_sem); + return ret; +} + /* directories */ static size_t ceph_vxattrcb_dir_entries(struct ceph_inode_info *ci, char *val, @@ -130,6 +170,8 @@ static size_t ceph_vxattrcb_dir_rctime(struct ceph_inode_info *ci, char *val, #define CEPH_XATTR_NAME(_type, _name) XATTR_CEPH_PREFIX #_type "." #_name +#define CEPH_XATTR_NAME2(_type, _name, _name2) \ + XATTR_CEPH_PREFIX #_type "." #_name "." #_name2 #define XATTR_NAME_CEPH(_type, _name) \ { \ @@ -140,6 +182,15 @@ static size_t ceph_vxattrcb_dir_rctime(struct ceph_inode_info *ci, char *val, .hidden = false, \ .exists_cb = NULL, \ } +#define XATTR_LAYOUT_FIELD(_type, _name, _field) \ + { \ + .name = CEPH_XATTR_NAME2(_type, _name, _field), \ + .name_size = sizeof (CEPH_XATTR_NAME2(_type, _name, _field)), \ + .getxattr_cb = ceph_vxattrcb_ ## _name ## _ ## _field, \ + .readonly = false, \ + .hidden = true, \ + .exists_cb = ceph_vxattrcb_layout_exists, \ + } static struct ceph_vxattr ceph_dir_vxattrs[] = { { @@ -150,6 +201,10 @@ static struct ceph_vxattr ceph_dir_vxattrs[] = { .hidden = false, .exists_cb = ceph_vxattrcb_layout_exists, }, + XATTR_LAYOUT_FIELD(dir, layout, stripe_unit), + XATTR_LAYOUT_FIELD(dir, layout, stripe_count), + XATTR_LAYOUT_FIELD(dir, layout, object_size), + XATTR_LAYOUT_FIELD(dir, layout, pool), XATTR_NAME_CEPH(dir, entries), XATTR_NAME_CEPH(dir, files), XATTR_NAME_CEPH(dir, subdirs), @@ -173,6 +228,10 @@ static struct ceph_vxattr ceph_file_vxattrs[] = { .hidden = false, .exists_cb = ceph_vxattrcb_layout_exists, }, + XATTR_LAYOUT_FIELD(file, layout, stripe_unit), + XATTR_LAYOUT_FIELD(file, layout, stripe_count), + XATTR_LAYOUT_FIELD(file, layout, object_size), + XATTR_LAYOUT_FIELD(file, layout, pool), { 0 } /* Required table terminator */ }; static size_t ceph_file_vxattrs_name_size; /* total size of all names */ -- cgit From fd9a8d7160937f94aad36ac80d7255b4988740ac Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 12 Feb 2013 09:48:42 -0500 Subject: NFSv4.1: Fix bulk recall and destroy of layouts The current code in pnfs_destroy_all_layouts() assumes that removing the layout from the server->layouts list is sufficient to make it invisible to other processes. This ignores the fact that most users access the layout through the nfs_inode->layout... There is further breakage due to lack of reference counting of the layouts, meaning that the whole thing Oopses at the drop of a hat. The code in initiate_bulk_draining() is almost correct, and can be used as a model for pnfs_destroy_all_layouts(), so move that code to pnfs.c, and refactor the code to allow us to choose between a single filesystem bulk recall, and a recall of all layouts. Also note that initiate_bulk_draining() currently calls iput() while holding locks. Fix that too. Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/nfs/callback_proc.c | 61 +++----------------- fs/nfs/pnfs.c | 150 ++++++++++++++++++++++++++++++++++++++++++------- fs/nfs/pnfs.h | 7 ++- 3 files changed, 144 insertions(+), 74 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index 264d1aa935f2..2960512792c2 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -183,60 +183,15 @@ static u32 initiate_file_draining(struct nfs_client *clp, static u32 initiate_bulk_draining(struct nfs_client *clp, struct cb_layoutrecallargs *args) { - struct nfs_server *server; - struct pnfs_layout_hdr *lo; - struct inode *ino; - u32 rv = NFS4ERR_NOMATCHING_LAYOUT; - struct pnfs_layout_hdr *tmp; - LIST_HEAD(recall_list); - LIST_HEAD(free_me_list); - struct pnfs_layout_range range = { - .iomode = IOMODE_ANY, - .offset = 0, - .length = NFS4_MAX_UINT64, - }; - - spin_lock(&clp->cl_lock); - rcu_read_lock(); - list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) { - if ((args->cbl_recall_type == RETURN_FSID) && - memcmp(&server->fsid, &args->cbl_fsid, - sizeof(struct nfs_fsid))) - continue; + int stat; - list_for_each_entry(lo, &server->layouts, plh_layouts) { - ino = igrab(lo->plh_inode); - if (!ino) - continue; - spin_lock(&ino->i_lock); - /* Is this layout in the process of being freed? */ - if (NFS_I(ino)->layout != lo) { - spin_unlock(&ino->i_lock); - iput(ino); - continue; - } - pnfs_get_layout_hdr(lo); - spin_unlock(&ino->i_lock); - list_add(&lo->plh_bulk_recall, &recall_list); - } - } - rcu_read_unlock(); - spin_unlock(&clp->cl_lock); - - list_for_each_entry_safe(lo, tmp, - &recall_list, plh_bulk_recall) { - ino = lo->plh_inode; - spin_lock(&ino->i_lock); - set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags); - if (pnfs_mark_matching_lsegs_invalid(lo, &free_me_list, &range)) - rv = NFS4ERR_DELAY; - list_del_init(&lo->plh_bulk_recall); - spin_unlock(&ino->i_lock); - pnfs_free_lseg_list(&free_me_list); - pnfs_put_layout_hdr(lo); - iput(ino); - } - return rv; + if (args->cbl_recall_type == RETURN_FSID) + stat = pnfs_destroy_layouts_byfsid(clp, &args->cbl_fsid, true); + else + stat = pnfs_destroy_layouts_byclid(clp, true); + if (stat != 0) + return NFS4ERR_DELAY; + return NFS4ERR_NOMATCHING_LAYOUT; } static u32 do_callback_layoutrecall(struct nfs_client *clp, diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index d00260b08103..6be70f622b62 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -505,37 +505,147 @@ pnfs_destroy_layout(struct nfs_inode *nfsi) } EXPORT_SYMBOL_GPL(pnfs_destroy_layout); -/* - * Called by the state manger to remove all layouts established under an - * expired lease. - */ -void -pnfs_destroy_all_layouts(struct nfs_client *clp) +static bool +pnfs_layout_add_bulk_destroy_list(struct inode *inode, + struct list_head *layout_list) { - struct nfs_server *server; struct pnfs_layout_hdr *lo; - LIST_HEAD(tmp_list); + bool ret = false; - nfs4_deviceid_mark_client_invalid(clp); - nfs4_deviceid_purge_client(clp); + spin_lock(&inode->i_lock); + lo = NFS_I(inode)->layout; + if (lo != NULL && list_empty(&lo->plh_bulk_destroy)) { + pnfs_get_layout_hdr(lo); + list_add(&lo->plh_bulk_destroy, layout_list); + ret = true; + } + spin_unlock(&inode->i_lock); + return ret; +} + +/* Caller must hold rcu_read_lock and clp->cl_lock */ +static int +pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp, + struct nfs_server *server, + struct list_head *layout_list) +{ + struct pnfs_layout_hdr *lo, *next; + struct inode *inode; + + list_for_each_entry_safe(lo, next, &server->layouts, plh_layouts) { + inode = igrab(lo->plh_inode); + if (inode == NULL) + continue; + list_del_init(&lo->plh_layouts); + if (pnfs_layout_add_bulk_destroy_list(inode, layout_list)) + continue; + rcu_read_unlock(); + spin_unlock(&clp->cl_lock); + iput(inode); + spin_lock(&clp->cl_lock); + rcu_read_lock(); + return -EAGAIN; + } + return 0; +} + +static int +pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list, + bool is_bulk_recall) +{ + struct pnfs_layout_hdr *lo; + struct inode *inode; + struct pnfs_layout_range range = { + .iomode = IOMODE_ANY, + .offset = 0, + .length = NFS4_MAX_UINT64, + }; + LIST_HEAD(lseg_list); + int ret = 0; + + while (!list_empty(layout_list)) { + lo = list_entry(layout_list->next, struct pnfs_layout_hdr, + plh_bulk_destroy); + dprintk("%s freeing layout for inode %lu\n", __func__, + lo->plh_inode->i_ino); + inode = lo->plh_inode; + spin_lock(&inode->i_lock); + list_del_init(&lo->plh_bulk_destroy); + lo->plh_block_lgets++; /* permanently block new LAYOUTGETs */ + if (is_bulk_recall) + set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags); + if (pnfs_mark_matching_lsegs_invalid(lo, &lseg_list, &range)) + ret = -EAGAIN; + spin_unlock(&inode->i_lock); + pnfs_free_lseg_list(&lseg_list); + pnfs_put_layout_hdr(lo); + iput(inode); + } + return ret; +} + +int +pnfs_destroy_layouts_byfsid(struct nfs_client *clp, + struct nfs_fsid *fsid, + bool is_recall) +{ + struct nfs_server *server; + LIST_HEAD(layout_list); spin_lock(&clp->cl_lock); rcu_read_lock(); +restart: list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) { - if (!list_empty(&server->layouts)) - list_splice_init(&server->layouts, &tmp_list); + if (memcmp(&server->fsid, fsid, sizeof(*fsid)) != 0) + continue; + if (pnfs_layout_bulk_destroy_byserver_locked(clp, + server, + &layout_list) != 0) + goto restart; } rcu_read_unlock(); spin_unlock(&clp->cl_lock); - while (!list_empty(&tmp_list)) { - lo = list_entry(tmp_list.next, struct pnfs_layout_hdr, - plh_layouts); - dprintk("%s freeing layout for inode %lu\n", __func__, - lo->plh_inode->i_ino); - list_del_init(&lo->plh_layouts); - pnfs_destroy_layout(NFS_I(lo->plh_inode)); + if (list_empty(&layout_list)) + return 0; + return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall); +} + +int +pnfs_destroy_layouts_byclid(struct nfs_client *clp, + bool is_recall) +{ + struct nfs_server *server; + LIST_HEAD(layout_list); + + spin_lock(&clp->cl_lock); + rcu_read_lock(); +restart: + list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) { + if (pnfs_layout_bulk_destroy_byserver_locked(clp, + server, + &layout_list) != 0) + goto restart; } + rcu_read_unlock(); + spin_unlock(&clp->cl_lock); + + if (list_empty(&layout_list)) + return 0; + return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall); +} + +/* + * Called by the state manger to remove all layouts established under an + * expired lease. + */ +void +pnfs_destroy_all_layouts(struct nfs_client *clp) +{ + nfs4_deviceid_mark_client_invalid(clp); + nfs4_deviceid_purge_client(clp); + + pnfs_destroy_layouts_byclid(clp, false); } /* @@ -888,7 +998,7 @@ alloc_init_layout_hdr(struct inode *ino, atomic_set(&lo->plh_refcount, 1); INIT_LIST_HEAD(&lo->plh_layouts); INIT_LIST_HEAD(&lo->plh_segs); - INIT_LIST_HEAD(&lo->plh_bulk_recall); + INIT_LIST_HEAD(&lo->plh_bulk_destroy); lo->plh_inode = ino; lo->plh_lc_cred = get_rpccred(ctx->state->owner->so_cred); return lo; diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index dbf7bba52da0..97cb358bb882 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -132,7 +132,7 @@ struct pnfs_layoutdriver_type { struct pnfs_layout_hdr { atomic_t plh_refcount; struct list_head plh_layouts; /* other client layouts */ - struct list_head plh_bulk_recall; /* clnt list of bulk recalls */ + struct list_head plh_bulk_destroy; struct list_head plh_segs; /* layout segments list */ nfs4_stateid plh_stateid; atomic_t plh_outstanding; /* number of RPCs out */ @@ -196,6 +196,11 @@ struct pnfs_layout_segment *pnfs_layout_process(struct nfs4_layoutget *lgp); void pnfs_free_lseg_list(struct list_head *tmp_list); void pnfs_destroy_layout(struct nfs_inode *); void pnfs_destroy_all_layouts(struct nfs_client *); +int pnfs_destroy_layouts_byfsid(struct nfs_client *clp, + struct nfs_fsid *fsid, + bool is_recall); +int pnfs_destroy_layouts_byclid(struct nfs_client *clp, + bool is_recall); void pnfs_put_layout_hdr(struct pnfs_layout_hdr *lo); void pnfs_set_layout_stateid(struct pnfs_layout_hdr *lo, const nfs4_stateid *new, -- cgit From 8de5c325b4ee7d5b23b95edd28b81c9a2a9f427f Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Thu, 14 Feb 2013 15:11:41 -0500 Subject: ext4: use KERN_WARNING for warning messages Some messages printed related to a WARN_ON(1) were printed using KERN_NOTICE. Use KERN_WARNING or ext4_warning() instead so that context related to the WARN_ON() is printed at the same printk warning level (and log files, etc.) Signed-off-by: "Theodore Ts'o" --- fs/ext4/ialloc.c | 12 ++++++------ fs/ext4/inode.c | 6 +++--- 2 files changed, 9 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 91d8fe3aced3..32fd2b9075dd 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -1028,17 +1028,17 @@ iget_failed: inode = NULL; bad_orphan: ext4_warning(sb, "bad orphan inode %lu! e2fsck was run?", ino); - printk(KERN_NOTICE "ext4_test_bit(bit=%d, block=%llu) = %d\n", + printk(KERN_WARNING "ext4_test_bit(bit=%d, block=%llu) = %d\n", bit, (unsigned long long)bitmap_bh->b_blocknr, ext4_test_bit(bit, bitmap_bh->b_data)); - printk(KERN_NOTICE "inode=%p\n", inode); + printk(KERN_WARNING "inode=%p\n", inode); if (inode) { - printk(KERN_NOTICE "is_bad_inode(inode)=%d\n", + printk(KERN_WARNING "is_bad_inode(inode)=%d\n", is_bad_inode(inode)); - printk(KERN_NOTICE "NEXT_ORPHAN(inode)=%u\n", + printk(KERN_WARNING "NEXT_ORPHAN(inode)=%u\n", NEXT_ORPHAN(inode)); - printk(KERN_NOTICE "max_ino=%lu\n", max_ino); - printk(KERN_NOTICE "i_nlink=%u\n", inode->i_nlink); + printk(KERN_WARNING "max_ino=%lu\n", max_ino); + printk(KERN_WARNING "i_nlink=%u\n", inode->i_nlink); /* Avoid freeing blocks if we got a bad deleted inode */ if (inode->i_nlink == 0) inode->i_blocks = 0; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 2fa18bb0bf3c..b85d5dae726b 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -343,7 +343,7 @@ void ext4_da_update_reserve_space(struct inode *inode, spin_lock(&ei->i_block_reservation_lock); trace_ext4_da_update_reserve_space(inode, used, quota_claim); if (unlikely(used > ei->i_reserved_data_blocks)) { - ext4_msg(inode->i_sb, KERN_NOTICE, "%s: ino %lu, used %d " + ext4_warning(inode->i_sb, "%s: ino %lu, used %d " "with only %d reserved data blocks", __func__, inode->i_ino, used, ei->i_reserved_data_blocks); @@ -352,7 +352,7 @@ void ext4_da_update_reserve_space(struct inode *inode, } if (unlikely(ei->i_allocated_meta_blocks > ei->i_reserved_meta_blocks)) { - ext4_msg(inode->i_sb, KERN_NOTICE, "%s: ino %lu, allocated %d " + ext4_warning(inode->i_sb, "%s: ino %lu, allocated %d " "with only %d reserved metadata blocks\n", __func__, inode->i_ino, ei->i_allocated_meta_blocks, ei->i_reserved_meta_blocks); @@ -1263,7 +1263,7 @@ static void ext4_da_release_space(struct inode *inode, int to_free) * function is called from invalidate page, it's * harmless to return without any action. */ - ext4_msg(inode->i_sb, KERN_NOTICE, "ext4_da_release_space: " + ext4_warning(inode->i_sb, "ext4_da_release_space: " "ino %lu, to_free %d with only %d reserved " "data blocks", inode->i_ino, to_free, ei->i_reserved_data_blocks); -- cgit From 01a523eb51cb505a4bc1eaffeeccd2527d6ab619 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Thu, 14 Feb 2013 15:51:58 -0500 Subject: ext4: add debugging context for warning in ext4_da_update_reserve_space() Print some additional debugging context to hopefully help to debug a warning which is getting triggered by xfstests #74. Also remove extraneous newlines from when printk's were converted to ext4_warning() and ext4_msg(). Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index b85d5dae726b..c4e9177f60c6 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -352,10 +352,12 @@ void ext4_da_update_reserve_space(struct inode *inode, } if (unlikely(ei->i_allocated_meta_blocks > ei->i_reserved_meta_blocks)) { - ext4_warning(inode->i_sb, "%s: ino %lu, allocated %d " - "with only %d reserved metadata blocks\n", __func__, - inode->i_ino, ei->i_allocated_meta_blocks, - ei->i_reserved_meta_blocks); + ext4_warning(inode->i_sb, "ino %lu, allocated %d " + "with only %d reserved metadata blocks " + "(releasing %d blocks with reserved %d data blocks)", + inode->i_ino, ei->i_allocated_meta_blocks, + ei->i_reserved_meta_blocks, used, + ei->i_reserved_data_blocks); WARN_ON(1); ei->i_allocated_meta_blocks = ei->i_reserved_meta_blocks; } @@ -1609,7 +1611,7 @@ static void mpage_da_map_and_submit(struct mpage_da_data *mpd) (unsigned long long) next, mpd->b_size >> mpd->inode->i_blkbits, err); ext4_msg(sb, KERN_CRIT, - "This should not happen!! Data will be lost\n"); + "This should not happen!! Data will be lost"); if (err == -ENOSPC) ext4_print_free_blocks(mpd->inode); } -- cgit From a1e16c26660b301cc8423185924cf1b0b16ea92b Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 11 Feb 2013 16:05:01 +1100 Subject: xfs: limit speculative prealloc size on sparse files Speculative preallocation based on the current file size works well for contiguous files, but is sub-optimal for sparse files where the EOF preallocation can fill holes and result in large amounts of zeros being written when it is not necessary. The algorithm is modified to prevent EOF speculative preallocation from triggering larger allocations on IO patterns of truncate--to-zero-seek-write-seek-write-.... which results in non-sparse files for large files. This, unfortunately, is the way cp now behaves when copying sparse files and so needs to be fixed. What this code does is that it looks at the existing extent adjacent to the current EOF and if it determines that it is a hole we disable speculative preallocation altogether. To avoid the next write from doing a large prealloc, it takes the size of subsequent preallocations from the current size of the existing EOF extent. IOWs, if you leave a hole in the file, it resets preallocation behaviour to the same as if it was a zero size file. Example new behaviour: $ xfs_io -f -c "pwrite 0 31m" \ -c "pwrite 33m 1m" \ -c "pwrite 128m 1m" \ -c "fiemap -v" /mnt/scratch/blah wrote 32505856/32505856 bytes at offset 0 31 MiB, 7936 ops; 0.0000 sec (1.608 GiB/sec and 421432.7439 ops/sec) wrote 1048576/1048576 bytes at offset 34603008 1 MiB, 256 ops; 0.0000 sec (1.462 GiB/sec and 383233.5329 ops/sec) wrote 1048576/1048576 bytes at offset 134217728 1 MiB, 256 ops; 0.0000 sec (1.719 GiB/sec and 450704.2254 ops/sec) /mnt/scratch/blah: EXT: FILE-OFFSET BLOCK-RANGE TOTAL FLAGS 0: [0..65535]: 96..65631 65536 0x0 1: [65536..67583]: hole 2048 2: [67584..69631]: 67680..69727 2048 0x0 3: [69632..262143]: hole 192512 4: [262144..264191]: 262240..264287 2048 0x1 Signed-off-by: Dave Chinner Reviewed-by: Mark Tinguely Reviewed-by: Brian Foster Signed-off-by: Ben Myers --- fs/xfs/xfs_iomap.c | 77 +++++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 67 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 364818eef40e..912d83d8860a 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -310,6 +310,62 @@ xfs_iomap_eof_want_preallocate( return 0; } +/* + * Determine the initial size of the preallocation. We are beyond the current + * EOF here, but we need to take into account whether this is a sparse write or + * an extending write when determining the preallocation size. Hence we need to + * look up the extent that ends at the current write offset and use the result + * to determine the preallocation size. + * + * If the extent is a hole, then preallocation is essentially disabled. + * Otherwise we take the size of the preceeding data extent as the basis for the + * preallocation size. If the size of the extent is greater than half the + * maximum extent length, then use the current offset as the basis. This ensures + * that for large files the preallocation size always extends to MAXEXTLEN + * rather than falling short due to things like stripe unit/width alignment of + * real extents. + */ +STATIC int +xfs_iomap_eof_prealloc_initial_size( + struct xfs_mount *mp, + struct xfs_inode *ip, + xfs_off_t offset, + xfs_bmbt_irec_t *imap, + int nimaps) +{ + xfs_fileoff_t start_fsb; + int imaps = 1; + int error; + + ASSERT(nimaps >= imaps); + + /* if we are using a specific prealloc size, return now */ + if (mp->m_flags & XFS_MOUNT_DFLT_IOSIZE) + return 0; + + /* + * As we write multiple pages, the offset will always align to the + * start of a page and hence point to a hole at EOF. i.e. if the size is + * 4096 bytes, we only have one block at FSB 0, but XFS_B_TO_FSB(4096) + * will return FSB 1. Hence if there are blocks in the file, we want to + * point to the block prior to the EOF block and not the hole that maps + * directly at @offset. + */ + start_fsb = XFS_B_TO_FSB(mp, offset); + if (start_fsb) + start_fsb--; + error = xfs_bmapi_read(ip, start_fsb, 1, imap, &imaps, XFS_BMAPI_ENTIRE); + if (error) + return 0; + + ASSERT(imaps == 1); + if (imap[0].br_startblock == HOLESTARTBLOCK) + return 0; + if (imap[0].br_blockcount <= (MAXEXTLEN >> 1)) + return imap[0].br_blockcount; + return XFS_B_TO_FSB(mp, offset); +} + /* * If we don't have a user specified preallocation size, dynamically increase * the preallocation size as the size of the file grows. Cap the maximum size @@ -319,20 +375,19 @@ xfs_iomap_eof_want_preallocate( STATIC xfs_fsblock_t xfs_iomap_prealloc_size( struct xfs_mount *mp, - struct xfs_inode *ip) + struct xfs_inode *ip, + xfs_off_t offset, + struct xfs_bmbt_irec *imap, + int nimaps) { xfs_fsblock_t alloc_blocks = 0; - if (!(mp->m_flags & XFS_MOUNT_DFLT_IOSIZE)) { + alloc_blocks = xfs_iomap_eof_prealloc_initial_size(mp, ip, offset, + imap, nimaps); + if (alloc_blocks > 0) { int shift = 0; int64_t freesp; - /* - * rounddown_pow_of_two() returns an undefined result - * if we pass in alloc_blocks = 0. Hence the "+ 1" to - * ensure we always pass in a non-zero value. - */ - alloc_blocks = XFS_B_TO_FSB(mp, XFS_ISIZE(ip)) + 1; alloc_blocks = XFS_FILEOFF_MIN(MAXEXTLEN, rounddown_pow_of_two(alloc_blocks)); @@ -399,7 +454,6 @@ xfs_iomap_write_delay( extsz = xfs_get_extsz_hint(ip); offset_fsb = XFS_B_TO_FSBT(mp, offset); - error = xfs_iomap_eof_want_preallocate(mp, ip, offset, count, imap, XFS_WRITE_IMAPS, &prealloc); if (error) @@ -407,7 +461,10 @@ xfs_iomap_write_delay( retry: if (prealloc) { - xfs_fsblock_t alloc_blocks = xfs_iomap_prealloc_size(mp, ip); + xfs_fsblock_t alloc_blocks; + + alloc_blocks = xfs_iomap_prealloc_size(mp, ip, offset, imap, + XFS_WRITE_IMAPS); aligned_offset = XFS_WRITEIO_ALIGN(mp, (offset + count - 1)); ioalign = XFS_B_TO_FSBT(mp, aligned_offset); -- cgit From 5337fe9b108d602c483fe9d62ffef9227acf3a74 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Mon, 11 Feb 2013 10:08:21 -0500 Subject: xfs: recheck buffer pinned status after push trylock failure The buffer pinned check and trylock sequence in xfs_buf_item_push() can race with an active transaction on marking the buffer pinned. This can result in the buffer becoming pinned and stale after the initial check and the trylock failure, but before the check in xfs_buf_trylock() that issues a log force. If the log force is issued from this context, a spinlock recursion occurs on xa_lock. Prepare xfs_buf_item_push() to handle the race by detecting a pinned buffer after the trylock failure so xfsaild issues a log force from a safe context. This, along with various previous fixes, renders the log force in xfs_buf_trylock() redundant. Signed-off-by: Brian Foster Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_buf_item.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index 9c4c050e4270..cf263476d6b4 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -469,8 +469,18 @@ xfs_buf_item_push( if (xfs_buf_ispinned(bp)) return XFS_ITEM_PINNED; - if (!xfs_buf_trylock(bp)) + if (!xfs_buf_trylock(bp)) { + /* + * If we have just raced with a buffer being pinned and it has + * been marked stale, we could end up stalling until someone else + * issues a log force to unpin the stale buffer. Check for the + * race condition here so xfsaild recognizes the buffer is pinned + * and queues a log force to move it along. + */ + if (xfs_buf_ispinned(bp)) + return XFS_ITEM_PINNED; return XFS_ITEM_LOCKED; + } ASSERT(!(bip->bli_flags & XFS_BLI_STALE)); -- cgit From fa5566e4ffb918131a054413eb42075a77a41413 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Mon, 11 Feb 2013 10:08:22 -0500 Subject: xfs: remove log force from xfs_buf_trylock() The trylock log force invoked via xfs_buf_item_push() can attempt to acquire xa_lock, thus leading to a recursion bug when called with xa_lock held. This log force was originally added to xfs_buf_trylock() to address xfsaild stalls due to pinned and stale buffers. Since the addition of this behavior, the log item pushing code had been reworked to detect and track pinned items to inform xfsaild to issue a log force itself when necessary. As such, the log force on trylock failure is redundant and safe to remove. Signed-off-by: Brian Foster Reviewed-by: Dave Chinner Signed-off-by: Ben Myers --- fs/xfs/xfs_buf.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index fbbb9eb92e32..4e8f0df82d02 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -951,8 +951,6 @@ xfs_buf_trylock( locked = down_trylock(&bp->b_sema) == 0; if (locked) XB_SET_OWNER(bp); - else if (atomic_read(&bp->b_pin_count) && (bp->b_flags & XBF_STALE)) - xfs_log_force(bp->b_target->bt_mount, 0); trace_xfs_buf_trylock(bp, _RET_IP_); return locked; -- cgit From 1e82379b018ceed0f0912327c60d73107dacbcb3 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 11 Feb 2013 15:58:13 +1100 Subject: xfs: xfs_bmap_add_attrfork_local is too generic When we are converting local data to an extent format as a result of adding an attribute, the type of data contained in the local fork determines the behaviour that needs to occur. xfs_bmap_add_attrfork_local() already handles the directory data case specially by using S_ISDIR() and calling out to xfs_dir2_sf_to_block(), but with verifiers we now need to handle each different type of metadata specially and different metadata formats require different verifiers (and eventually block header initialisation). There is only a single place that we add and attribute fork to the inode, but that is in the attribute code and it knows nothing about the specific contents of the data fork. It is only the case of local data that is the issue here, so adding code to hadnle this case in the attribute specific code is wrong. Hence we are really stuck trying to detect the data fork contents in xfs_bmap_add_attrfork_local() and performing the correct callout there. Luckily the current cases can be determined by S_IS* macros, and we can push the work off to data specific callouts, but each of those callouts does a lot of work in common with xfs_bmap_local_to_extents(). The only reason that this fails for symlinks right now is is that xfs_bmap_local_to_extents() assumes the data fork contains extent data, and so attaches a a bmap extent data verifier to the buffer and simply copies the data fork information straight into it. To fix this, allow us to pass a "formatting" callback into xfs_bmap_local_to_extents() which is responsible for setting the buffer type, initialising it and copying the data fork contents over to the new buffer. This allows callers to specify how they want to format the new buffer (which is necessary for the upcoming CRC enabled metadata blocks) and hence make xfs_bmap_local_to_extents() useful for any type of data fork content. Signed-off-by: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers --- fs/xfs/xfs_bmap.c | 114 ++++++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 93 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 491f35e2dd26..b44af9211bd9 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -147,7 +147,10 @@ xfs_bmap_local_to_extents( xfs_fsblock_t *firstblock, /* first block allocated in xaction */ xfs_extlen_t total, /* total blocks needed by transaction */ int *logflagsp, /* inode logging flags */ - int whichfork); /* data or attr fork */ + int whichfork, /* data or attr fork */ + void (*init_fn)(struct xfs_buf *bp, + struct xfs_inode *ip, + struct xfs_ifork *ifp)); /* * Search the extents list for the inode, for the extent containing bno. @@ -357,7 +360,42 @@ xfs_bmap_add_attrfork_extents( } /* - * Called from xfs_bmap_add_attrfork to handle local format files. + * Block initialisation functions for local to extent format conversion. + * As these get more complex, they will be moved to the relevant files, + * but for now they are too simple to worry about. + */ +STATIC void +xfs_bmap_local_to_extents_init_fn( + struct xfs_buf *bp, + struct xfs_inode *ip, + struct xfs_ifork *ifp) +{ + bp->b_ops = &xfs_bmbt_buf_ops; + memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes); +} + +STATIC void +xfs_symlink_local_to_remote( + struct xfs_buf *bp, + struct xfs_inode *ip, + struct xfs_ifork *ifp) +{ + /* remote symlink blocks are not verifiable until CRCs come along */ + bp->b_ops = NULL; + memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes); +} + +/* + * Called from xfs_bmap_add_attrfork to handle local format files. Each + * different data fork content type needs a different callout to do the + * conversion. Some are basic and only require special block initialisation + * callouts for the data formating, others (directories) are so specialised they + * handle everything themselves. + * + * XXX (dgc): investigate whether directory conversion can use the generic + * formatting callout. It should be possible - it's just a very complex + * formatter. it would also require passing the transaction through to the init + * function. */ STATIC int /* error */ xfs_bmap_add_attrfork_local( @@ -368,25 +406,29 @@ xfs_bmap_add_attrfork_local( int *flags) /* inode logging flags */ { xfs_da_args_t dargs; /* args for dir/attr code */ - int error; /* error return value */ - xfs_mount_t *mp; /* mount structure pointer */ if (ip->i_df.if_bytes <= XFS_IFORK_DSIZE(ip)) return 0; + if (S_ISDIR(ip->i_d.di_mode)) { - mp = ip->i_mount; memset(&dargs, 0, sizeof(dargs)); dargs.dp = ip; dargs.firstblock = firstblock; dargs.flist = flist; - dargs.total = mp->m_dirblkfsbs; + dargs.total = ip->i_mount->m_dirblkfsbs; dargs.whichfork = XFS_DATA_FORK; dargs.trans = tp; - error = xfs_dir2_sf_to_block(&dargs); - } else - error = xfs_bmap_local_to_extents(tp, ip, firstblock, 1, flags, - XFS_DATA_FORK); - return error; + return xfs_dir2_sf_to_block(&dargs); + } + + if (S_ISLNK(ip->i_d.di_mode)) + return xfs_bmap_local_to_extents(tp, ip, firstblock, 1, + flags, XFS_DATA_FORK, + xfs_symlink_local_to_remote); + + return xfs_bmap_local_to_extents(tp, ip, firstblock, 1, flags, + XFS_DATA_FORK, + xfs_bmap_local_to_extents_init_fn); } /* @@ -3219,7 +3261,10 @@ xfs_bmap_local_to_extents( xfs_fsblock_t *firstblock, /* first block allocated in xaction */ xfs_extlen_t total, /* total blocks needed by transaction */ int *logflagsp, /* inode logging flags */ - int whichfork) /* data or attr fork */ + int whichfork, + void (*init_fn)(struct xfs_buf *bp, + struct xfs_inode *ip, + struct xfs_ifork *ifp)) { int error; /* error return value */ int flags; /* logging flags returned */ @@ -3239,12 +3284,12 @@ xfs_bmap_local_to_extents( xfs_buf_t *bp; /* buffer for extent block */ xfs_bmbt_rec_host_t *ep;/* extent record pointer */ + ASSERT((ifp->if_flags & + (XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) == XFS_IFINLINE); memset(&args, 0, sizeof(args)); args.tp = tp; args.mp = ip->i_mount; args.firstblock = *firstblock; - ASSERT((ifp->if_flags & - (XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) == XFS_IFINLINE); /* * Allocate a block. We know we need only one, since the * file currently fits in an inode. @@ -3258,17 +3303,20 @@ xfs_bmap_local_to_extents( } args.total = total; args.minlen = args.maxlen = args.prod = 1; - if ((error = xfs_alloc_vextent(&args))) + error = xfs_alloc_vextent(&args); + if (error) goto done; - /* - * Can't fail, the space was reserved. - */ + + /* Can't fail, the space was reserved. */ ASSERT(args.fsbno != NULLFSBLOCK); ASSERT(args.len == 1); *firstblock = args.fsbno; bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0); - bp->b_ops = &xfs_bmbt_buf_ops; - memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes); + + /* initialise the block and copy the data */ + init_fn(bp, ip, ifp); + + /* account for the change in fork size and log everything */ xfs_trans_log_buf(tp, bp, 0, ifp->if_bytes - 1); xfs_bmap_forkoff_reset(args.mp, ip, whichfork); xfs_idata_realloc(ip, -ifp->if_bytes, whichfork); @@ -4915,8 +4963,32 @@ xfs_bmapi_write( XFS_STATS_INC(xs_blk_mapw); if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) { + /* + * XXX (dgc): This assumes we are only called for inodes that + * contain content neutral data in local format. Anything that + * contains caller-specific data in local format that needs + * transformation to move to a block format needs to do the + * conversion to extent format itself. + * + * Directory data forks and attribute forks handle this + * themselves, but with the addition of metadata verifiers every + * data fork in local format now contains caller specific data + * and as such conversion through this function is likely to be + * broken. + * + * The only likely user of this branch is for remote symlinks, + * but we cannot overwrite the data fork contents of the symlink + * (EEXIST occurs higher up the stack) and so it will never go + * from local format to extent format here. Hence I don't think + * this branch is ever executed intentionally and we should + * consider removing it and asserting that xfs_bmapi_write() + * cannot be called directly on local format forks. i.e. callers + * are completely responsible for local to extent format + * conversion, not xfs_bmapi_write(). + */ error = xfs_bmap_local_to_extents(tp, ip, firstblock, total, - &bma.logflags, whichfork); + &bma.logflags, whichfork, + xfs_bmap_local_to_extents_init_fn); if (error) goto error0; } -- cgit From 2a745b14bc99d52c29d0c886a110321f651cf183 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Wed, 13 Feb 2013 04:20:01 -0700 Subject: Btrfs: fix crash in log replay with qgroups enabled When replaying a log tree with qgroups enabled, tree_mod_log_rewind does a sanity-check of the number of items against the maximum possible number. It calculates that number with the nodesize of fs_root. Unfortunately fs_root is not yet set at this stage. So instead use the nodesize from tree_root, which is already initialized. Signed-off-by: Arne Jansen Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index eea5da7a2b9a..6eff0fa9ecaa 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1222,7 +1222,7 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb, __tree_mod_log_rewind(eb_rewin, time_seq, tm); WARN_ON(btrfs_header_nritems(eb_rewin) > - BTRFS_NODEPTRS_PER_BLOCK(fs_info->fs_root)); + BTRFS_NODEPTRS_PER_BLOCK(fs_info->tree_root)); return eb_rewin; } -- cgit From dc6982ff4db1f47da73b1967ef5302d6721e5b95 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Thu, 14 Feb 2013 23:59:26 -0500 Subject: ext4: refactor code to read directory blocks into ext4_read_dirblock() The code to read in directory blocks and verify their metadata checksums was replicated in ten different places across fs/ext4/namei.c, and the code was buggy in subtle ways in a number of those replicated sites. In some cases, ext4_error() was called with a training newline. In others, in particularly in empty_dir(), it was possible to call ext4_dirent_csum_verify() on an index block, which would trigger false warnings requesting the system adminsitrator to run e2fsck. By refactoring the code, we make the code more readable, as well as shrinking the compiled object file by over 700 bytes and 50 lines of code. Signed-off-by: "Theodore Ts'o" --- fs/ext4/namei.c | 291 +++++++++++++++++++++++--------------------------------- 1 file changed, 119 insertions(+), 172 deletions(-) (limited to 'fs') diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 3e1529c39808..0e28c749e273 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -83,6 +83,86 @@ static struct buffer_head *ext4_append(handle_t *handle, return bh; } +static int ext4_dx_csum_verify(struct inode *inode, + struct ext4_dir_entry *dirent); + +typedef enum { + EITHER, INDEX, DIRENT +} dirblock_type_t; + +#define ext4_read_dirblock(inode, block, type) \ + __ext4_read_dirblock((inode), (block), (type), __LINE__) + +static struct buffer_head *__ext4_read_dirblock(struct inode *inode, + ext4_lblk_t block, + dirblock_type_t type, + unsigned int line) +{ + struct buffer_head *bh; + struct ext4_dir_entry *dirent; + int err = 0, is_dx_block = 0; + + bh = ext4_bread(NULL, inode, block, 0, &err); + if (!bh) { + if (err == 0) { + ext4_error_inode(inode, __func__, line, block, + "Directory hole found"); + return ERR_PTR(-EIO); + } + __ext4_warning(inode->i_sb, __func__, line, + "error reading directory block " + "(ino %lu, block %lu)", inode->i_ino, + (unsigned long) block); + return ERR_PTR(err); + } + dirent = (struct ext4_dir_entry *) bh->b_data; + /* Determine whether or not we have an index block */ + if (is_dx(inode)) { + if (block == 0) + is_dx_block = 1; + else if (ext4_rec_len_from_disk(dirent->rec_len, + inode->i_sb->s_blocksize) == + inode->i_sb->s_blocksize) + is_dx_block = 1; + } + if (!is_dx_block && type == INDEX) { + ext4_error_inode(inode, __func__, line, block, + "directory leaf block found instead of index block"); + return ERR_PTR(-EIO); + } + if (!EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb, + EXT4_FEATURE_RO_COMPAT_METADATA_CSUM) || + buffer_verified(bh)) + return bh; + + /* + * An empty leaf block can get mistaken for a index block; for + * this reason, we can only check the index checksum when the + * caller is sure it should be an index block. + */ + if (is_dx_block && type == INDEX) { + if (ext4_dx_csum_verify(inode, dirent)) + set_buffer_verified(bh); + else { + ext4_error_inode(inode, __func__, line, block, + "Directory index failed checksum"); + brelse(bh); + return ERR_PTR(-EIO); + } + } + if (!is_dx_block) { + if (ext4_dirent_csum_verify(inode, dirent)) + set_buffer_verified(bh); + else { + ext4_error_inode(inode, __func__, line, block, + "Directory block failed checksum"); + brelse(bh); + return ERR_PTR(-EIO); + } + } + return bh; +} + #ifndef assert #define assert(test) J_ASSERT(test) #endif @@ -604,9 +684,9 @@ dx_probe(const struct qstr *d_name, struct inode *dir, u32 hash; frame->bh = NULL; - if (!(bh = ext4_bread(NULL, dir, 0, 0, err))) { - if (*err == 0) - *err = ERR_BAD_DX_DIR; + bh = ext4_read_dirblock(dir, 0, INDEX); + if (IS_ERR(bh)) { + *err = PTR_ERR(bh); goto fail; } root = (struct dx_root *) bh->b_data; @@ -643,15 +723,6 @@ dx_probe(const struct qstr *d_name, struct inode *dir, goto fail; } - if (!buffer_verified(bh) && - !ext4_dx_csum_verify(dir, (struct ext4_dir_entry *)bh->b_data)) { - ext4_warning(dir->i_sb, "Root failed checksum"); - brelse(bh); - *err = ERR_BAD_DX_DIR; - goto fail; - } - set_buffer_verified(bh); - entries = (struct dx_entry *) (((char *)&root->info) + root->info.info_length); @@ -709,23 +780,13 @@ dx_probe(const struct qstr *d_name, struct inode *dir, frame->entries = entries; frame->at = at; if (!indirect--) return frame; - if (!(bh = ext4_bread(NULL, dir, dx_get_block(at), 0, err))) { - if (!(*err)) - *err = ERR_BAD_DX_DIR; + bh = ext4_read_dirblock(dir, dx_get_block(at), INDEX); + if (IS_ERR(bh)) { + *err = PTR_ERR(bh); goto fail2; } entries = ((struct dx_node *) bh->b_data)->entries; - if (!buffer_verified(bh) && - !ext4_dx_csum_verify(dir, - (struct ext4_dir_entry *)bh->b_data)) { - ext4_warning(dir->i_sb, "Node failed checksum"); - brelse(bh); - *err = ERR_BAD_DX_DIR; - goto fail2; - } - set_buffer_verified(bh); - if (dx_get_limit(entries) != dx_node_limit (dir)) { ext4_warning(dir->i_sb, "dx entry: limit != node limit"); @@ -783,7 +844,7 @@ static int ext4_htree_next_block(struct inode *dir, __u32 hash, { struct dx_frame *p; struct buffer_head *bh; - int err, num_frames = 0; + int num_frames = 0; __u32 bhash; p = frame; @@ -822,26 +883,9 @@ static int ext4_htree_next_block(struct inode *dir, __u32 hash, * block so no check is necessary */ while (num_frames--) { - if (!(bh = ext4_bread(NULL, dir, dx_get_block(p->at), - 0, &err))) { - if (!err) { - ext4_error(dir->i_sb, - "Directory hole detected on inode %lu\n", - dir->i_ino); - return -EIO; - } - return err; /* Failure */ - } - - if (!buffer_verified(bh) && - !ext4_dx_csum_verify(dir, - (struct ext4_dir_entry *)bh->b_data)) { - ext4_warning(dir->i_sb, "Node failed checksum"); - brelse(bh); - return -EIO; - } - set_buffer_verified(bh); - + bh = ext4_read_dirblock(dir, dx_get_block(p->at), INDEX); + if (IS_ERR(bh)) + return PTR_ERR(bh); p++; brelse(p->bh); p->bh = bh; @@ -867,23 +911,9 @@ static int htree_dirblock_to_tree(struct file *dir_file, dxtrace(printk(KERN_INFO "In htree dirblock_to_tree: block %lu\n", (unsigned long)block)); - if (!(bh = ext4_bread(NULL, dir, block, 0, &err))) { - if (!err) { - err = -EIO; - ext4_error(dir->i_sb, - "Directory hole detected on inode %lu\n", - dir->i_ino); - } - return err; - } - - if (!buffer_verified(bh) && - !ext4_dirent_csum_verify(dir, - (struct ext4_dir_entry *)bh->b_data)) { - brelse(bh); - return -EIO; - } - set_buffer_verified(bh); + bh = ext4_read_dirblock(dir, block, DIRENT); + if (IS_ERR(bh)) + return PTR_ERR(bh); de = (struct ext4_dir_entry_2 *) bh->b_data; top = (struct ext4_dir_entry_2 *) ((char *) de + @@ -1337,26 +1367,11 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q return NULL; do { block = dx_get_block(frame->at); - if (!(bh = ext4_bread(NULL, dir, block, 0, err))) { - if (!(*err)) { - *err = -EIO; - ext4_error(dir->i_sb, - "Directory hole detected on inode %lu\n", - dir->i_ino); - } + bh = ext4_read_dirblock(dir, block, DIRENT); + if (IS_ERR(bh)) { + *err = PTR_ERR(bh); goto errout; } - - if (!buffer_verified(bh) && - !ext4_dirent_csum_verify(dir, - (struct ext4_dir_entry *)bh->b_data)) { - EXT4_ERROR_INODE(dir, "checksumming directory " - "block %lu", (unsigned long)block); - brelse(bh); - *err = -EIO; - goto errout; - } - set_buffer_verified(bh); retval = search_dirblock(bh, dir, d_name, block << EXT4_BLOCK_SIZE_BITS(sb), res_dir); @@ -1920,22 +1935,10 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, } blocks = dir->i_size >> sb->s_blocksize_bits; for (block = 0; block < blocks; block++) { - if (!(bh = ext4_bread(handle, dir, block, 0, &retval))) { - if (!retval) { - retval = -EIO; - ext4_error(inode->i_sb, - "Directory hole detected on inode %lu\n", - inode->i_ino); - } - return retval; - } - if (!buffer_verified(bh) && - !ext4_dirent_csum_verify(dir, - (struct ext4_dir_entry *)bh->b_data)) { - brelse(bh); - return -EIO; - } - set_buffer_verified(bh); + bh = ext4_read_dirblock(dir, block, DIRENT); + if (IS_ERR(bh)) + return PTR_ERR(bh); + retval = add_dirent_to_buf(handle, dentry, inode, NULL, bh); if (retval != -ENOSPC) { brelse(bh); @@ -1986,22 +1989,13 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, return err; entries = frame->entries; at = frame->at; - - if (!(bh = ext4_bread(handle, dir, dx_get_block(frame->at), 0, &err))) { - if (!err) { - err = -EIO; - ext4_error(dir->i_sb, - "Directory hole detected on inode %lu\n", - dir->i_ino); - } + bh = ext4_read_dirblock(dir, dx_get_block(frame->at), DIRENT); + if (IS_ERR(bh)) { + err = PTR_ERR(bh); + bh = NULL; goto cleanup; } - if (!buffer_verified(bh) && - !ext4_dirent_csum_verify(dir, (struct ext4_dir_entry *)bh->b_data)) - goto journal_error; - set_buffer_verified(bh); - BUFFER_TRACE(bh, "get_write_access"); err = ext4_journal_get_write_access(handle, bh); if (err) @@ -2352,6 +2346,7 @@ static int ext4_init_new_dir(handle_t *handle, struct inode *dir, struct buffer_head *dir_block = NULL; struct ext4_dir_entry_2 *de; struct ext4_dir_entry_tail *t; + ext4_lblk_t block = 0; unsigned int blocksize = dir->i_sb->s_blocksize; int csum_size = 0; int err; @@ -2368,16 +2363,9 @@ static int ext4_init_new_dir(handle_t *handle, struct inode *dir, goto out; } - inode->i_size = EXT4_I(inode)->i_disksize = blocksize; - if (!(dir_block = ext4_bread(handle, inode, 0, 1, &err))) { - if (!err) { - err = -EIO; - ext4_error(inode->i_sb, - "Directory hole detected on inode %lu\n", - inode->i_ino); - } + inode->i_size = 0; + if (!(dir_block = ext4_append(handle, inode, &block, &err))) goto out; - } BUFFER_TRACE(dir_block, "get_write_access"); err = ext4_journal_get_write_access(handle, dir_block); if (err) @@ -2477,26 +2465,14 @@ static int empty_dir(struct inode *inode) } sb = inode->i_sb; - if (inode->i_size < EXT4_DIR_REC_LEN(1) + EXT4_DIR_REC_LEN(2) || - !(bh = ext4_bread(NULL, inode, 0, 0, &err))) { - if (err) - EXT4_ERROR_INODE(inode, - "error %d reading directory lblock 0", err); - else - ext4_warning(inode->i_sb, - "bad directory (dir #%lu) - no data block", - inode->i_ino); + if (inode->i_size < EXT4_DIR_REC_LEN(1) + EXT4_DIR_REC_LEN(2)) { + EXT4_ERROR_INODE(inode, "invalid size"); return 1; } - if (!buffer_verified(bh) && - !ext4_dirent_csum_verify(inode, - (struct ext4_dir_entry *)bh->b_data)) { - EXT4_ERROR_INODE(inode, "checksum error reading directory " - "lblock 0"); - brelse(bh); - return -EIO; - } - set_buffer_verified(bh); + bh = ext4_read_dirblock(inode, 0, EITHER); + if (IS_ERR(bh)) + return 1; + de = (struct ext4_dir_entry_2 *) bh->b_data; de1 = ext4_next_entry(de, sb->s_blocksize); if (le32_to_cpu(de->inode) != inode->i_ino || @@ -2519,29 +2495,9 @@ static int empty_dir(struct inode *inode) err = 0; brelse(bh); lblock = offset >> EXT4_BLOCK_SIZE_BITS(sb); - bh = ext4_bread(NULL, inode, lblock, 0, &err); - if (!bh) { - if (err) - EXT4_ERROR_INODE(inode, - "error %d reading directory " - "lblock %u", err, lblock); - else - ext4_warning(inode->i_sb, - "bad directory (dir #%lu) - no data block", - inode->i_ino); - - offset += sb->s_blocksize; - continue; - } - if (!buffer_verified(bh) && - !ext4_dirent_csum_verify(inode, - (struct ext4_dir_entry *)bh->b_data)) { - EXT4_ERROR_INODE(inode, "checksum error " - "reading directory lblock 0"); - brelse(bh); - return -EIO; - } - set_buffer_verified(bh); + bh = ext4_read_dirblock(inode, lblock, EITHER); + if (IS_ERR(bh)) + return 1; de = (struct ext4_dir_entry_2 *) bh->b_data; } if (ext4_check_dir_entry(inode, NULL, de, bh, @@ -3004,13 +2960,9 @@ static struct buffer_head *ext4_get_first_dir_block(handle_t *handle, struct buffer_head *bh; if (!ext4_has_inline_data(inode)) { - if (!(bh = ext4_bread(handle, inode, 0, 0, retval))) { - if (!*retval) { - *retval = -EIO; - ext4_error(inode->i_sb, - "Directory hole detected on inode %lu\n", - inode->i_ino); - } + bh = ext4_read_dirblock(inode, 0, EITHER); + if (IS_ERR(bh)) { + *retval = PTR_ERR(bh); return NULL; } *parent_de = ext4_next_entry( @@ -3089,11 +3041,6 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, &inlined); if (!dir_bh) goto end_rename; - if (!inlined && !buffer_verified(dir_bh) && - !ext4_dirent_csum_verify(old_inode, - (struct ext4_dir_entry *)dir_bh->b_data)) - goto end_rename; - set_buffer_verified(dir_bh); if (le32_to_cpu(parent_de->inode) != old_dir->i_ino) goto end_rename; retval = -EMLINK; -- cgit From 0f70b40613ee14b0cadafeb461034cff81b4419a Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Fri, 15 Feb 2013 03:35:57 -0500 Subject: ext4: use ERR_PTR() abstraction for ext4_append() Use ERR_PTR()/IS_ERR() abstraction instead of passing in a separate pointer to an integer for the error code, as a code cleanup. Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 2 ++ fs/ext4/namei.c | 65 +++++++++++++++++++++++++++------------------------------ 2 files changed, 33 insertions(+), 34 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index c4e9177f60c6..f4466c3650dc 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -707,6 +707,8 @@ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode, /* ensure we send some value back into *errp */ *errp = 0; + if (create && err == 0) + err = -ENOSPC; /* should never happen */ if (err < 0) *errp = err; if (err <= 0) diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 0e28c749e273..f58c053d13e0 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -51,34 +51,28 @@ static struct buffer_head *ext4_append(handle_t *handle, struct inode *inode, - ext4_lblk_t *block, int *err) + ext4_lblk_t *block) { struct buffer_head *bh; + int err = 0; if (unlikely(EXT4_SB(inode->i_sb)->s_max_dir_size_kb && ((inode->i_size >> 10) >= - EXT4_SB(inode->i_sb)->s_max_dir_size_kb))) { - *err = -ENOSPC; - return NULL; - } + EXT4_SB(inode->i_sb)->s_max_dir_size_kb))) + return ERR_PTR(-ENOSPC); *block = inode->i_size >> inode->i_sb->s_blocksize_bits; - bh = ext4_bread(handle, inode, *block, 1, err); - if (bh) { - inode->i_size += inode->i_sb->s_blocksize; - EXT4_I(inode)->i_disksize = inode->i_size; - *err = ext4_journal_get_write_access(handle, bh); - if (*err) { - brelse(bh); - bh = NULL; - } - } - if (!bh && !(*err)) { - *err = -EIO; - ext4_error(inode->i_sb, - "Directory hole detected on inode %lu\n", - inode->i_ino); + bh = ext4_bread(handle, inode, *block, 1, &err); + if (!bh) + return ERR_PTR(err); + inode->i_size += inode->i_sb->s_blocksize; + EXT4_I(inode)->i_disksize = inode->i_size; + err = ext4_journal_get_write_access(handle, bh); + if (err) { + brelse(bh); + ext4_std_error(inode->i_sb, err); + return ERR_PTR(err); } return bh; } @@ -1555,11 +1549,12 @@ static struct ext4_dir_entry_2 *do_split(handle_t *handle, struct inode *dir, EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) csum_size = sizeof(struct ext4_dir_entry_tail); - bh2 = ext4_append (handle, dir, &newblock, &err); - if (!(bh2)) { + bh2 = ext4_append(handle, dir, &newblock); + if (IS_ERR(bh2)) { brelse(*bh); *bh = NULL; - goto errout; + *error = PTR_ERR(bh2); + return NULL; } BUFFER_TRACE(*bh, "get_write_access"); @@ -1640,7 +1635,6 @@ journal_error: brelse(bh2); *bh = NULL; ext4_std_error(dir->i_sb, err); -errout: *error = err; return NULL; } @@ -1815,10 +1809,10 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, len = ((char *) root) + (blocksize - csum_size) - (char *) de; /* Allocate new block for the 0th block's dirents */ - bh2 = ext4_append(handle, dir, &block, &retval); - if (!(bh2)) { + bh2 = ext4_append(handle, dir, &block); + if (IS_ERR(bh2)) { brelse(bh); - return retval; + return PTR_ERR(bh2); } ext4_set_inode_flag(dir, EXT4_INODE_INDEX); data1 = bh2->b_data; @@ -1950,9 +1944,9 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, return make_indexed_dir(handle, dentry, inode, bh); brelse(bh); } - bh = ext4_append(handle, dir, &block, &retval); - if (!bh) - return retval; + bh = ext4_append(handle, dir, &block); + if (IS_ERR(bh)) + return PTR_ERR(bh); de = (struct ext4_dir_entry_2 *) bh->b_data; de->inode = 0; de->rec_len = ext4_rec_len_to_disk(blocksize - csum_size, blocksize); @@ -2023,9 +2017,11 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, err = -ENOSPC; goto cleanup; } - bh2 = ext4_append (handle, dir, &newblock, &err); - if (!(bh2)) + bh2 = ext4_append(handle, dir, &newblock); + if (IS_ERR(bh2)) { + err = PTR_ERR(bh2); goto cleanup; + } node2 = (struct dx_node *)(bh2->b_data); entries2 = node2->entries; memset(&node2->fake, 0, sizeof(struct fake_dirent)); @@ -2364,8 +2360,9 @@ static int ext4_init_new_dir(handle_t *handle, struct inode *dir, } inode->i_size = 0; - if (!(dir_block = ext4_append(handle, inode, &block, &err))) - goto out; + dir_block = ext4_append(handle, inode, &block); + if (IS_ERR(dir_block)) + return PTR_ERR(dir_block); BUFFER_TRACE(dir_block, "get_write_access"); err = ext4_journal_get_write_access(handle, dir_block); if (err) -- cgit From 483479c26a65e5f0cc95e9324f313bc95c7dc6fd Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Mon, 4 Feb 2013 14:02:35 +0300 Subject: NFS: use SUNRPC cache creation and destruction helper for DNS cache This cache was the first containerized and doesn't use net-aware cache creation and destruction helpers. This is a cleanup patch which just makes code looks clearer and reduce amount of lines of code. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfs/dns_resolve.c | 60 ++++++++++++++++++++++------------------------------ 1 file changed, 25 insertions(+), 35 deletions(-) (limited to 'fs') diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index 62c8c4127219..d5ce5f4f0f5f 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -353,48 +353,39 @@ ssize_t nfs_dns_resolve_name(struct net *net, char *name, } EXPORT_SYMBOL_GPL(nfs_dns_resolve_name); +static struct cache_detail nfs_dns_resolve_template = { + .owner = THIS_MODULE, + .hash_size = NFS_DNS_HASHTBL_SIZE, + .name = "dns_resolve", + .cache_put = nfs_dns_ent_put, + .cache_upcall = nfs_dns_upcall, + .cache_parse = nfs_dns_parse, + .cache_show = nfs_dns_show, + .match = nfs_dns_match, + .init = nfs_dns_ent_init, + .update = nfs_dns_ent_update, + .alloc = nfs_dns_ent_alloc, +}; + + int nfs_dns_resolver_cache_init(struct net *net) { - int err = -ENOMEM; + int err; struct nfs_net *nn = net_generic(net, nfs_net_id); - struct cache_detail *cd; - struct cache_head **tbl; - cd = kzalloc(sizeof(struct cache_detail), GFP_KERNEL); - if (cd == NULL) - goto err_cd; - - tbl = kzalloc(NFS_DNS_HASHTBL_SIZE * sizeof(struct cache_head *), - GFP_KERNEL); - if (tbl == NULL) - goto err_tbl; - - cd->owner = THIS_MODULE, - cd->hash_size = NFS_DNS_HASHTBL_SIZE, - cd->hash_table = tbl, - cd->name = "dns_resolve", - cd->cache_put = nfs_dns_ent_put, - cd->cache_upcall = nfs_dns_upcall, - cd->cache_parse = nfs_dns_parse, - cd->cache_show = nfs_dns_show, - cd->match = nfs_dns_match, - cd->init = nfs_dns_ent_init, - cd->update = nfs_dns_ent_update, - cd->alloc = nfs_dns_ent_alloc, - - nfs_cache_init(cd); - err = nfs_cache_register_net(net, cd); + nn->nfs_dns_resolve = cache_create_net(&nfs_dns_resolve_template, net); + if (IS_ERR(nn->nfs_dns_resolve)) + return PTR_ERR(nn->nfs_dns_resolve); + + nfs_cache_init(nn->nfs_dns_resolve); + err = nfs_cache_register_net(net, nn->nfs_dns_resolve); if (err) goto err_reg; - nn->nfs_dns_resolve = cd; return 0; err_reg: - nfs_cache_destroy(cd); - kfree(cd->hash_table); -err_tbl: - kfree(cd); -err_cd: + nfs_cache_destroy(nn->nfs_dns_resolve); + cache_destroy_net(nn->nfs_dns_resolve, net); return err; } @@ -405,8 +396,7 @@ void nfs_dns_resolver_cache_destroy(struct net *net) nfs_cache_unregister_net(net, cd); nfs_cache_destroy(cd); - kfree(cd->hash_table); - kfree(cd); + cache_destroy_net(nn->nfs_dns_resolve, net); } static int rpc_pipefs_event(struct notifier_block *nb, unsigned long event, -- cgit From 462b8f6bf1d3f5feb7a346394036dbc1df3a8ed5 Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Mon, 4 Feb 2013 14:02:40 +0300 Subject: NFS: simplify and clean cache library This is a cleanup patch. Such helpers like nfs_cache_init() and nfs_cache_destroy() are redundant, because they are just a wrappers around sunrpc_init_cache_detail() and sunrpc_destroy_cache_detail() respectively. So let's remove them completely and move corresponding logic to nfs_cache_register_net() and nfs_cache_unregister_net() respectively (since they are called together anyway). Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfs/cache_lib.c | 12 +++--------- fs/nfs/cache_lib.h | 2 -- fs/nfs/dns_resolve.c | 6 +----- 3 files changed, 4 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfs/cache_lib.c b/fs/nfs/cache_lib.c index 862a2f16db64..5f7b053720ee 100644 --- a/fs/nfs/cache_lib.c +++ b/fs/nfs/cache_lib.c @@ -128,10 +128,13 @@ int nfs_cache_register_net(struct net *net, struct cache_detail *cd) struct super_block *pipefs_sb; int ret = 0; + sunrpc_init_cache_detail(cd); pipefs_sb = rpc_get_sb_net(net); if (pipefs_sb) { ret = nfs_cache_register_sb(pipefs_sb, cd); rpc_put_sb_net(net); + if (ret) + sunrpc_destroy_cache_detail(cd); } return ret; } @@ -151,14 +154,5 @@ void nfs_cache_unregister_net(struct net *net, struct cache_detail *cd) nfs_cache_unregister_sb(pipefs_sb, cd); rpc_put_sb_net(net); } -} - -void nfs_cache_init(struct cache_detail *cd) -{ - sunrpc_init_cache_detail(cd); -} - -void nfs_cache_destroy(struct cache_detail *cd) -{ sunrpc_destroy_cache_detail(cd); } diff --git a/fs/nfs/cache_lib.h b/fs/nfs/cache_lib.h index 317db95e37f8..4116d2c3f52f 100644 --- a/fs/nfs/cache_lib.h +++ b/fs/nfs/cache_lib.h @@ -23,8 +23,6 @@ extern struct nfs_cache_defer_req *nfs_cache_defer_req_alloc(void); extern void nfs_cache_defer_req_put(struct nfs_cache_defer_req *dreq); extern int nfs_cache_wait_for_upcall(struct nfs_cache_defer_req *dreq); -extern void nfs_cache_init(struct cache_detail *cd); -extern void nfs_cache_destroy(struct cache_detail *cd); extern int nfs_cache_register_net(struct net *net, struct cache_detail *cd); extern void nfs_cache_unregister_net(struct net *net, struct cache_detail *cd); extern int nfs_cache_register_sb(struct super_block *sb, diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index d5ce5f4f0f5f..9cbc98a4159d 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -377,14 +377,12 @@ int nfs_dns_resolver_cache_init(struct net *net) if (IS_ERR(nn->nfs_dns_resolve)) return PTR_ERR(nn->nfs_dns_resolve); - nfs_cache_init(nn->nfs_dns_resolve); err = nfs_cache_register_net(net, nn->nfs_dns_resolve); if (err) goto err_reg; return 0; err_reg: - nfs_cache_destroy(nn->nfs_dns_resolve); cache_destroy_net(nn->nfs_dns_resolve, net); return err; } @@ -392,10 +390,8 @@ err_reg: void nfs_dns_resolver_cache_destroy(struct net *net) { struct nfs_net *nn = net_generic(net, nfs_net_id); - struct cache_detail *cd = nn->nfs_dns_resolve; - nfs_cache_unregister_net(net, cd); - nfs_cache_destroy(cd); + nfs_cache_unregister_net(net, nn->nfs_dns_resolve); cache_destroy_net(nn->nfs_dns_resolve, net); } -- cgit From 73fb847a44224d5708550e4be7baba9da75e00af Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Mon, 4 Feb 2013 14:02:45 +0300 Subject: SUNRPC: introduce cache_detail->cache_request callback This callback will allow to simplify upcalls in further patches in this series. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfs/dns_resolve.c | 3 ++- fs/nfsd/export.c | 6 ++++-- fs/nfsd/nfs4idmap.c | 6 ++++-- 3 files changed, 10 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index 9cbc98a4159d..55bc5d11457c 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -144,7 +144,7 @@ static int nfs_dns_upcall(struct cache_detail *cd, ret = nfs_cache_upcall(cd, key->hostname); if (ret) - ret = sunrpc_cache_pipe_upcall(cd, ch, nfs_dns_request); + ret = sunrpc_cache_pipe_upcall(cd, ch, cd->cache_request); return ret; } @@ -359,6 +359,7 @@ static struct cache_detail nfs_dns_resolve_template = { .name = "dns_resolve", .cache_put = nfs_dns_ent_put, .cache_upcall = nfs_dns_upcall, + .cache_request = nfs_dns_request, .cache_parse = nfs_dns_parse, .cache_show = nfs_dns_show, .match = nfs_dns_match, diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index 8e9df45b3d3c..0e16c7fa6800 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -69,7 +69,7 @@ static void expkey_request(struct cache_detail *cd, static int expkey_upcall(struct cache_detail *cd, struct cache_head *h) { - return sunrpc_cache_pipe_upcall(cd, h, expkey_request); + return sunrpc_cache_pipe_upcall(cd, h, cd->cache_request); } static struct svc_expkey *svc_expkey_update(struct cache_detail *cd, struct svc_expkey *new, @@ -246,6 +246,7 @@ static struct cache_detail svc_expkey_cache_template = { .name = "nfsd.fh", .cache_put = expkey_put, .cache_upcall = expkey_upcall, + .cache_request = expkey_request, .cache_parse = expkey_parse, .cache_show = expkey_show, .match = expkey_match, @@ -340,7 +341,7 @@ static void svc_export_request(struct cache_detail *cd, static int svc_export_upcall(struct cache_detail *cd, struct cache_head *h) { - return sunrpc_cache_pipe_upcall(cd, h, svc_export_request); + return sunrpc_cache_pipe_upcall(cd, h, cd->cache_request); } static struct svc_export *svc_export_update(struct svc_export *new, @@ -714,6 +715,7 @@ static struct cache_detail svc_export_cache_template = { .name = "nfsd.export", .cache_put = svc_export_put, .cache_upcall = svc_export_upcall, + .cache_request = svc_export_request, .cache_parse = svc_export_parse, .cache_show = svc_export_show, .match = svc_export_match, diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index a1f10c0a6255..9033dfde1812 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -142,7 +142,7 @@ idtoname_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, static int idtoname_upcall(struct cache_detail *cd, struct cache_head *ch) { - return sunrpc_cache_pipe_upcall(cd, ch, idtoname_request); + return sunrpc_cache_pipe_upcall(cd, ch, cd->cache_request); } static int @@ -193,6 +193,7 @@ static struct cache_detail idtoname_cache_template = { .name = "nfs4.idtoname", .cache_put = ent_put, .cache_upcall = idtoname_upcall, + .cache_request = idtoname_request, .cache_parse = idtoname_parse, .cache_show = idtoname_show, .warn_no_listener = warn_no_idmapd, @@ -323,7 +324,7 @@ nametoid_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, static int nametoid_upcall(struct cache_detail *cd, struct cache_head *ch) { - return sunrpc_cache_pipe_upcall(cd, ch, nametoid_request); + return sunrpc_cache_pipe_upcall(cd, ch, cd->cache_request); } static int @@ -366,6 +367,7 @@ static struct cache_detail nametoid_cache_template = { .name = "nfs4.nametoid", .cache_put = ent_put, .cache_upcall = nametoid_upcall, + .cache_request = nametoid_request, .cache_parse = nametoid_parse, .cache_show = nametoid_show, .warn_no_listener = warn_no_idmapd, -- cgit From 2d4383383b0b04ca380b67aa2d7397d0b399dcbf Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Mon, 4 Feb 2013 14:02:50 +0300 Subject: SUNRPC: rework cache upcall logic For most of SUNRPC caches (except NFS DNS cache) cache_detail->cache_upcall is redundant since all that it's implementations are doing is calling sunrpc_cache_pipe_upcall() with proper function address argument. Cache request function address is now stored on cache_detail structure and thus all the code can be simplified. Now, for those cache details, which doesn't have cache_upcall callback (the only one, which still has is nfs_dns_resolve_template) sunrpc_cache_pipe_upcall will be called instead. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfsd/export.c | 12 ------------ fs/nfsd/nfs4idmap.c | 14 -------------- 2 files changed, 26 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index 0e16c7fa6800..15ebf91982b0 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -67,11 +67,6 @@ static void expkey_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } -static int expkey_upcall(struct cache_detail *cd, struct cache_head *h) -{ - return sunrpc_cache_pipe_upcall(cd, h, cd->cache_request); -} - static struct svc_expkey *svc_expkey_update(struct cache_detail *cd, struct svc_expkey *new, struct svc_expkey *old); static struct svc_expkey *svc_expkey_lookup(struct cache_detail *cd, struct svc_expkey *); @@ -245,7 +240,6 @@ static struct cache_detail svc_expkey_cache_template = { .hash_size = EXPKEY_HASHMAX, .name = "nfsd.fh", .cache_put = expkey_put, - .cache_upcall = expkey_upcall, .cache_request = expkey_request, .cache_parse = expkey_parse, .cache_show = expkey_show, @@ -339,11 +333,6 @@ static void svc_export_request(struct cache_detail *cd, (*bpp)[-1] = '\n'; } -static int svc_export_upcall(struct cache_detail *cd, struct cache_head *h) -{ - return sunrpc_cache_pipe_upcall(cd, h, cd->cache_request); -} - static struct svc_export *svc_export_update(struct svc_export *new, struct svc_export *old); static struct svc_export *svc_export_lookup(struct svc_export *); @@ -714,7 +703,6 @@ static struct cache_detail svc_export_cache_template = { .hash_size = EXPORT_HASHMAX, .name = "nfsd.export", .cache_put = svc_export_put, - .cache_upcall = svc_export_upcall, .cache_request = svc_export_request, .cache_parse = svc_export_parse, .cache_show = svc_export_show, diff --git a/fs/nfsd/nfs4idmap.c b/fs/nfsd/nfs4idmap.c index 9033dfde1812..d9402ea9d751 100644 --- a/fs/nfsd/nfs4idmap.c +++ b/fs/nfsd/nfs4idmap.c @@ -139,12 +139,6 @@ idtoname_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, (*bpp)[-1] = '\n'; } -static int -idtoname_upcall(struct cache_detail *cd, struct cache_head *ch) -{ - return sunrpc_cache_pipe_upcall(cd, ch, cd->cache_request); -} - static int idtoname_match(struct cache_head *ca, struct cache_head *cb) { @@ -192,7 +186,6 @@ static struct cache_detail idtoname_cache_template = { .hash_size = ENT_HASHMAX, .name = "nfs4.idtoname", .cache_put = ent_put, - .cache_upcall = idtoname_upcall, .cache_request = idtoname_request, .cache_parse = idtoname_parse, .cache_show = idtoname_show, @@ -321,12 +314,6 @@ nametoid_request(struct cache_detail *cd, struct cache_head *ch, char **bpp, (*bpp)[-1] = '\n'; } -static int -nametoid_upcall(struct cache_detail *cd, struct cache_head *ch) -{ - return sunrpc_cache_pipe_upcall(cd, ch, cd->cache_request); -} - static int nametoid_match(struct cache_head *ca, struct cache_head *cb) { @@ -366,7 +353,6 @@ static struct cache_detail nametoid_cache_template = { .hash_size = ENT_HASHMAX, .name = "nfs4.nametoid", .cache_put = ent_put, - .cache_upcall = nametoid_upcall, .cache_request = nametoid_request, .cache_parse = nametoid_parse, .cache_show = nametoid_show, -- cgit From 21cd1254d3402a72927ed744e8ac1a7cf532f1ea Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Mon, 4 Feb 2013 14:02:55 +0300 Subject: SUNRPC: remove "cache_request" argument in sunrpc_cache_pipe_upcall() function Passing this pointer is redundant since it's stored on cache_detail structure, which is also passed to sunrpc_cache_pipe_upcall () function. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfs/dns_resolve.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index 55bc5d11457c..945527092295 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -144,7 +144,7 @@ static int nfs_dns_upcall(struct cache_detail *cd, ret = nfs_cache_upcall(cd, key->hostname); if (ret) - ret = sunrpc_cache_pipe_upcall(cd, ch, cd->cache_request); + ret = sunrpc_cache_pipe_upcall(cd, ch); return ret; } -- cgit From 1ac8362977b9ec75779170ac3074c7b36ab19b82 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 14 Feb 2013 16:45:13 -0500 Subject: nfsd: fix comments on nfsd_cache_lookup Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 40db57eb2b06..2f9c2d26a2b9 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -302,8 +302,10 @@ nfsd_cache_search(struct svc_rqst *rqstp, __wsum csum) /* * Try to find an entry matching the current call in the cache. When none - * is found, we grab the oldest unlocked entry off the LRU list. - * Note that no operation within the loop may sleep. + * is found, we try to grab the oldest expired entry off the LRU list. If + * a suitable one isn't there, then drop the cache_lock and allocate a + * new one, then search again in case one got inserted while this thread + * didn't hold the lock. */ int nfsd_cache_lookup(struct svc_rqst *rqstp) @@ -344,6 +346,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) } } + /* Drop the lock and allocate a new entry */ spin_unlock(&cache_lock); rp = nfsd_reply_cache_alloc(); if (!rp) { -- cgit From 11f779421a39b86da8a523d97e5fd3477878d44f Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Fri, 1 Feb 2013 15:56:12 +0300 Subject: nfsd: containerize NFSd filesystem This patch makes NFSD file system superblock to be created per net. This makes possible to get proper network namespace from superblock instead of using hard-coded "init_net". Note: NFSd fs super-block holds network namespace. This garantees, that network namespace won't disappear from underneath of it. This, obviously, means, that in case of kill of a container's "init" (which is not a mount namespace, but network namespace creator) netowrk namespace won't be destroyed. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 46 +++++++++++++++++++++++++++++++++------------- fs/nfsd/nfssvc.c | 5 ++--- 2 files changed, 35 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 29c3f0d25469..f6d448e6eb28 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -220,6 +220,7 @@ static ssize_t write_unlock_ip(struct file *file, char *buf, size_t size) struct sockaddr *sap = (struct sockaddr *)&address; size_t salen = sizeof(address); char *fo_path; + struct net *net = file->f_dentry->d_sb->s_fs_info; /* sanity check */ if (size == 0) @@ -232,7 +233,7 @@ static ssize_t write_unlock_ip(struct file *file, char *buf, size_t size) if (qword_get(&buf, fo_path, size) < 0) return -EINVAL; - if (rpc_pton(&init_net, fo_path, size, sap, salen) == 0) + if (rpc_pton(net, fo_path, size, sap, salen) == 0) return -EINVAL; return nlmsvc_unlock_all_by_ip(sap); @@ -317,6 +318,7 @@ static ssize_t write_filehandle(struct file *file, char *buf, size_t size) int len; struct auth_domain *dom; struct knfsd_fh fh; + struct net *net = file->f_dentry->d_sb->s_fs_info; if (size == 0) return -EINVAL; @@ -352,7 +354,7 @@ static ssize_t write_filehandle(struct file *file, char *buf, size_t size) if (!dom) return -ENOMEM; - len = exp_rootfh(&init_net, dom, path, &fh, maxsize); + len = exp_rootfh(net, dom, path, &fh, maxsize); auth_domain_put(dom); if (len) return len; @@ -396,7 +398,7 @@ static ssize_t write_threads(struct file *file, char *buf, size_t size) { char *mesg = buf; int rv; - struct net *net = &init_net; + struct net *net = file->f_dentry->d_sb->s_fs_info; if (size > 0) { int newthreads; @@ -447,7 +449,7 @@ static ssize_t write_pool_threads(struct file *file, char *buf, size_t size) int len; int npools; int *nthreads; - struct net *net = &init_net; + struct net *net = file->f_dentry->d_sb->s_fs_info; mutex_lock(&nfsd_mutex); npools = nfsd_nrpools(net); @@ -510,7 +512,7 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) unsigned minor; ssize_t tlen = 0; char *sep; - struct net *net = &init_net; + struct net *net = file->f_dentry->d_sb->s_fs_info; struct nfsd_net *nn = net_generic(net, nfsd_net_id); if (size>0) { @@ -792,7 +794,7 @@ static ssize_t __write_ports(struct file *file, char *buf, size_t size, static ssize_t write_ports(struct file *file, char *buf, size_t size) { ssize_t rv; - struct net *net = &init_net; + struct net *net = file->f_dentry->d_sb->s_fs_info; mutex_lock(&nfsd_mutex); rv = __write_ports(file, buf, size, net); @@ -827,7 +829,7 @@ int nfsd_max_blksize; static ssize_t write_maxblksize(struct file *file, char *buf, size_t size) { char *mesg = buf; - struct net *net = &init_net; + struct net *net = file->f_dentry->d_sb->s_fs_info; struct nfsd_net *nn = net_generic(net, nfsd_net_id); if (size > 0) { @@ -923,7 +925,8 @@ static ssize_t nfsd4_write_time(struct file *file, char *buf, size_t size, */ static ssize_t write_leasetime(struct file *file, char *buf, size_t size) { - struct nfsd_net *nn = net_generic(&init_net, nfsd_net_id); + struct net *net = file->f_dentry->d_sb->s_fs_info; + struct nfsd_net *nn = net_generic(net, nfsd_net_id); return nfsd4_write_time(file, buf, size, &nn->nfsd4_lease, nn); } @@ -939,7 +942,8 @@ static ssize_t write_leasetime(struct file *file, char *buf, size_t size) */ static ssize_t write_gracetime(struct file *file, char *buf, size_t size) { - struct nfsd_net *nn = net_generic(&init_net, nfsd_net_id); + struct net *net = file->f_dentry->d_sb->s_fs_info; + struct nfsd_net *nn = net_generic(net, nfsd_net_id); return nfsd4_write_time(file, buf, size, &nn->nfsd4_grace, nn); } @@ -995,7 +999,8 @@ static ssize_t __write_recoverydir(struct file *file, char *buf, size_t size, static ssize_t write_recoverydir(struct file *file, char *buf, size_t size) { ssize_t rv; - struct nfsd_net *nn = net_generic(&init_net, nfsd_net_id); + struct net *net = file->f_dentry->d_sb->s_fs_info; + struct nfsd_net *nn = net_generic(net, nfsd_net_id); mutex_lock(&nfsd_mutex); rv = __write_recoverydir(file, buf, size, nn); @@ -1037,20 +1042,35 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent) #endif /* last one */ {""} }; - return simple_fill_super(sb, 0x6e667364, nfsd_files); + struct net *net = data; + int ret; + + ret = simple_fill_super(sb, 0x6e667364, nfsd_files); + if (ret) + return ret; + sb->s_fs_info = get_net(net); + return 0; } static struct dentry *nfsd_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) { - return mount_single(fs_type, flags, data, nfsd_fill_super); + return mount_ns(fs_type, flags, current->nsproxy->net_ns, nfsd_fill_super); +} + +static void nfsd_umount(struct super_block *sb) +{ + struct net *net = sb->s_fs_info; + + kill_litter_super(sb); + put_net(net); } static struct file_system_type nfsd_fs_type = { .owner = THIS_MODULE, .name = "nfsd", .mount = nfsd_mount, - .kill_sb = kill_litter_super, + .kill_sb = nfsd_umount, }; #ifdef CONFIG_PROC_FS diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 40cb1cbaba4e..6cee5db72047 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -702,8 +702,7 @@ nfsd_dispatch(struct svc_rqst *rqstp, __be32 *statp) int nfsd_pool_stats_open(struct inode *inode, struct file *file) { int ret; - struct net *net = &init_net; - struct nfsd_net *nn = net_generic(net, nfsd_net_id); + struct nfsd_net *nn = net_generic(inode->i_sb->s_fs_info, nfsd_net_id); mutex_lock(&nfsd_mutex); if (nn->nfsd_serv == NULL) { @@ -720,7 +719,7 @@ int nfsd_pool_stats_open(struct inode *inode, struct file *file) int nfsd_pool_stats_release(struct inode *inode, struct file *file) { int ret = seq_release(inode, file); - struct net *net = &init_net; + struct net *net = inode->i_sb->s_fs_info; mutex_lock(&nfsd_mutex); /* this function really, really should have been called svc_put() */ -- cgit From 96d851c4d28de8cc83fe2bd5c6bc2eb8f253a6c5 Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Fri, 1 Feb 2013 15:56:17 +0300 Subject: nfsd: use proper net while reading "exports" file Functuon "exports_open" is used for both "/proc/fs/nfs/exports" and "/proc/fs/nfsd/exports" files. Now NFSd filesystem is containerised, so proper net can be taken from superblock for "/proc/fs/nfsd/exports" reader. But for "/proc/fs/nfsd/exports" only current->nsproxy->net_ns can be used. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 31 +++++++++++++++++++++++++------ 1 file changed, 25 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index f6d448e6eb28..8ead2c25ce65 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -125,11 +125,11 @@ static const struct file_operations transaction_ops = { .llseek = default_llseek, }; -static int exports_open(struct inode *inode, struct file *file) +static int exports_net_open(struct net *net, struct file *file) { int err; struct seq_file *seq; - struct nfsd_net *nn = net_generic(&init_net, nfsd_net_id); + struct nfsd_net *nn = net_generic(net, nfsd_net_id); err = seq_open(file, &nfs_exports_op); if (err) @@ -140,8 +140,26 @@ static int exports_open(struct inode *inode, struct file *file) return 0; } -static const struct file_operations exports_operations = { - .open = exports_open, +static int exports_proc_open(struct inode *inode, struct file *file) +{ + return exports_net_open(current->nsproxy->net_ns, file); +} + +static const struct file_operations exports_proc_operations = { + .open = exports_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, + .owner = THIS_MODULE, +}; + +static int exports_nfsd_open(struct inode *inode, struct file *file) +{ + return exports_net_open(inode->i_sb->s_fs_info, file); +} + +static const struct file_operations exports_nfsd_operations = { + .open = exports_nfsd_open, .read = seq_read, .llseek = seq_lseek, .release = seq_release, @@ -1018,7 +1036,7 @@ static ssize_t write_recoverydir(struct file *file, char *buf, size_t size) static int nfsd_fill_super(struct super_block * sb, void * data, int silent) { static struct tree_descr nfsd_files[] = { - [NFSD_List] = {"exports", &exports_operations, S_IRUGO}, + [NFSD_List] = {"exports", &exports_nfsd_operations, S_IRUGO}, [NFSD_Export_features] = {"export_features", &export_features_operations, S_IRUGO}, [NFSD_FO_UnlockIP] = {"unlock_ip", @@ -1081,7 +1099,8 @@ static int create_proc_exports_entry(void) entry = proc_mkdir("fs/nfs", NULL); if (!entry) return -ENOMEM; - entry = proc_create("exports", 0, entry, &exports_operations); + entry = proc_create("exports", 0, entry, + &exports_proc_operations); if (!entry) return -ENOMEM; return 0; -- cgit From 71a50306934f416e74ba27cbfb88855c22251525 Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Fri, 1 Feb 2013 15:56:22 +0300 Subject: nfsd: disable usermode helper client tracker in container This tracker uses khelper kthread to execute binaries. Execution itself is done from kthread context - i.e. global root is used. This is not suitable for containers with own root. So, disable this tracker for a while. Note: one of possible solutions can be pass "init" callback to khelper, which will swap root to desired one. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4recover.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index ba6fdd4a0455..e0ae1cf18a82 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -1185,6 +1185,12 @@ bin_to_hex_dup(const unsigned char *src, int srclen) static int nfsd4_umh_cltrack_init(struct net __attribute__((unused)) *net) { + /* XXX: The usermode helper s not working in container yet. */ + if (net != &init_net) { + WARN(1, KERN_ERR "NFSD: attempt to initialize umh client " + "tracking in a container!\n"); + return -EINVAL; + } return nfsd4_umh_cltrack_upcall("init", NULL, NULL); } -- cgit From deb4534f4f3be7aea7d9d24c3b0d58f370cbf9ef Mon Sep 17 00:00:00 2001 From: Stanislav Kinsbursky Date: Fri, 1 Feb 2013 15:56:27 +0300 Subject: nfsd: enable NFSv4 state in containers Currently, NFSd is ready to operate in network namespace based containers. So let's drop check for "init_net" and make it able to fly. Signed-off-by: Stanislav Kinsbursky Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 10 ---------- 1 file changed, 10 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index c1a6ddf3a84a..f194f869be4c 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -4939,16 +4939,6 @@ nfs4_state_start_net(struct net *net) struct nfsd_net *nn = net_generic(net, nfsd_net_id); int ret; - /* - * FIXME: For now, we hang most of the pernet global stuff off of - * init_net until nfsd is fully containerized. Eventually, we'll - * need to pass a net pointer into this function, take a reference - * to that instead and then do most of the rest of this on a per-net - * basis. - */ - if (net != &init_net) - return -EINVAL; - ret = nfs4_state_create_net(net); if (ret) return ret; -- cgit From f25cc71e634edcf8a15bc60a48f2b5f3ec9fbb1d Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Wed, 13 Feb 2013 08:40:16 -0700 Subject: lockd: nlmclnt_reclaim(): avoid stack overflow Even though nlmclnt_reclaim() is only one call into the stack frame, 928 bytes on the stack seems like a lot. Recode to dynamically allocate the request structure once from within the reclaimer task, then pass this pointer into nlmclnt_reclaim() for reuse on subsequent calls. smatch analysis: fs/lockd/clntproc.c:620 nlmclnt_reclaim() warn: 'reqst' puts 928 bytes on stack Also remove redundant assignment of 0 after memset. Cc: Trond Myklebust Signed-off-by: Tim Gardner Reviewed-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/lockd/clntlock.c | 12 +++++++++++- fs/lockd/clntproc.c | 6 ++---- 2 files changed, 13 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index 4885b53e56bd..6cd673d34fb9 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -220,10 +220,19 @@ reclaimer(void *ptr) { struct nlm_host *host = (struct nlm_host *) ptr; struct nlm_wait *block; + struct nlm_rqst *req; struct file_lock *fl, *next; u32 nsmstate; struct net *net = host->net; + req = kmalloc(sizeof(*req), GFP_KERNEL); + if (!req) { + printk(KERN_ERR "lockd: reclaimer unable to alloc memory." + " Locks for %s won't be reclaimed!\n", + host->h_name); + return 0; + } + allow_signal(SIGKILL); down_write(&host->h_rwsem); @@ -253,7 +262,7 @@ restart: */ if (signalled()) continue; - if (nlmclnt_reclaim(host, fl) != 0) + if (nlmclnt_reclaim(host, fl, req) != 0) continue; list_add_tail(&fl->fl_u.nfs_fl.list, &host->h_granted); if (host->h_nsmstate != nsmstate) { @@ -279,5 +288,6 @@ restart: /* Release host handle after use */ nlmclnt_release_host(host); lockd_down(net); + kfree(req); return 0; } diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 54f9e6ce0430..b43114c4332a 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -615,17 +615,15 @@ out_unlock: * RECLAIM: Try to reclaim a lock */ int -nlmclnt_reclaim(struct nlm_host *host, struct file_lock *fl) +nlmclnt_reclaim(struct nlm_host *host, struct file_lock *fl, + struct nlm_rqst *req) { - struct nlm_rqst reqst, *req; int status; - req = &reqst; memset(req, 0, sizeof(*req)); locks_init_lock(&req->a_args.lock.fl); locks_init_lock(&req->a_res.lock.fl); req->a_host = host; - req->a_flags = 0; /* Set up the argument struct */ nlmclnt_setlockargs(req, fl); -- cgit From 6f60cbd3ae442cb35861bb522f388db123d42ec1 Mon Sep 17 00:00:00 2001 From: David Sterba Date: Fri, 15 Feb 2013 11:31:02 -0700 Subject: btrfs: access superblock via pagecache in scan_one_device btrfs_scan_one_device is calling set_blocksize() which can race with a concurrent process making dirty page cache pages. It can end up dropping dirty page cache pages on the floor, which isn't very nice when someone is just running btrfs dev scan to find filesystems on the box. Now that udev is registering btrfs devices as it discovers them, we can actually end up racing with our own mkfs program too. When this happens, we drop some of the important blocks written by mkfs. This commit changes scan_one_device to read the super out of the page cache instead of trying to use bread. This way we don't have to care about the blocksize of the device. This also drops the invalidate_bdev() call. It wasn't very polite to invalidate during the scan either. mkfs is putting the super into the page cache, there's no reason to invalidate at this point. Signed-off-by: David Sterba Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 70 +++++++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 64 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 5cbb7f4b1672..5349e17d8863 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -792,26 +792,77 @@ int btrfs_open_devices(struct btrfs_fs_devices *fs_devices, return ret; } +/* + * Look for a btrfs signature on a device. This may be called out of the mount path + * and we are not allowed to call set_blocksize during the scan. The superblock + * is read via pagecache + */ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, struct btrfs_fs_devices **fs_devices_ret) { struct btrfs_super_block *disk_super; struct block_device *bdev; - struct buffer_head *bh; - int ret; + struct page *page; + void *p; + int ret = -EINVAL; u64 devid; u64 transid; u64 total_devices; + u64 bytenr; + pgoff_t index; + /* + * we would like to check all the supers, but that would make + * a btrfs mount succeed after a mkfs from a different FS. + * So, we need to add a special mount option to scan for + * later supers, using BTRFS_SUPER_MIRROR_MAX instead + */ + bytenr = btrfs_sb_offset(0); flags |= FMODE_EXCL; mutex_lock(&uuid_mutex); - ret = btrfs_get_bdev_and_sb(path, flags, holder, 0, &bdev, &bh); - if (ret) + + bdev = blkdev_get_by_path(path, flags, holder); + + if (IS_ERR(bdev)) { + ret = PTR_ERR(bdev); + printk(KERN_INFO "btrfs: open %s failed\n", path); goto error; - disk_super = (struct btrfs_super_block *)bh->b_data; + } + + /* make sure our super fits in the device */ + if (bytenr + PAGE_CACHE_SIZE >= i_size_read(bdev->bd_inode)) + goto error_bdev_put; + + /* make sure our super fits in the page */ + if (sizeof(*disk_super) > PAGE_CACHE_SIZE) + goto error_bdev_put; + + /* make sure our super doesn't straddle pages on disk */ + index = bytenr >> PAGE_CACHE_SHIFT; + if ((bytenr + sizeof(*disk_super) - 1) >> PAGE_CACHE_SHIFT != index) + goto error_bdev_put; + + /* pull in the page with our super */ + page = read_cache_page_gfp(bdev->bd_inode->i_mapping, + index, GFP_NOFS); + + if (IS_ERR_OR_NULL(page)) + goto error_bdev_put; + + p = kmap(page); + + /* align our pointer to the offset of the super block */ + disk_super = p + (bytenr & ~PAGE_CACHE_MASK); + + if (btrfs_super_bytenr(disk_super) != bytenr || + strncmp((char *)(&disk_super->magic), BTRFS_MAGIC, + sizeof(disk_super->magic))) + goto error_unmap; + devid = btrfs_stack_device_id(&disk_super->dev_item); transid = btrfs_super_generation(disk_super); total_devices = btrfs_super_num_devices(disk_super); + if (disk_super->label[0]) { if (disk_super->label[BTRFS_LABEL_SIZE - 1]) disk_super->label[BTRFS_LABEL_SIZE - 1] = '\0'; @@ -819,12 +870,19 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, } else { printk(KERN_INFO "device fsid %pU ", disk_super->fsid); } + printk(KERN_CONT "devid %llu transid %llu %s\n", (unsigned long long)devid, (unsigned long long)transid, path); + ret = device_list_add(path, disk_super, devid, fs_devices_ret); if (!ret && fs_devices_ret) (*fs_devices_ret)->total_devices = total_devices; - brelse(bh); + +error_unmap: + kunmap(page); + page_cache_release(page); + +error_bdev_put: blkdev_put(bdev, flags); error: mutex_unlock(&uuid_mutex); -- cgit From 56edc86b5a72bdbc86358e57fb09165136baf0b8 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 15 Feb 2013 13:36:34 -0500 Subject: nfsd: fix compiler warning about ambiguous types in nfsd_cache_csum kbuild test robot says: tree: git://linux-nfs.org/~bfields/linux.git for-3.9 head: deb4534f4f3be7aea7d9d24c3b0d58f370cbf9ef commit: 01a7decf75930925322c5efc87af0b5e58eb8650 [32/44] nfsd: keep a checksum of the first 256 bytes of request config: i386-randconfig-x088 (attached as .config) All warnings: fs/nfsd/nfscache.c: In function 'nfsd_cache_csum': >> fs/nfsd/nfscache.c:266:9: warning: comparison of distinct pointer types lacks a cast [enabled by default] vim +266 fs/nfsd/nfscache.c 250 __wsum csum; 251 struct xdr_buf *buf = &rqstp->rq_arg; 252 const unsigned char *p = buf->head[0].iov_base; 253 size_t csum_len = min_t(size_t, buf->head[0].iov_len + buf->page_len, 254 RC_CSUMLEN); 255 size_t len = min(buf->head[0].iov_len, csum_len); 256 257 /* rq_arg.head first */ 258 csum = csum_partial(p, len, 0); 259 csum_len -= len; 260 261 /* Continue into page array */ 262 idx = buf->page_base / PAGE_SIZE; 263 base = buf->page_base & ~PAGE_MASK; 264 while (csum_len) { 265 p = page_address(buf->pages[idx]) + base; > 266 len = min(PAGE_SIZE - base, csum_len); 267 csum = csum_partial(p, len, csum); 268 csum_len -= len; 269 base = 0; 270 ++idx; 271 } 272 return csum; 273 } 274 Signed-off-by: Jeff Layton Reported-by: kbuild test robot Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 2f9c2d26a2b9..ca43664422f6 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -263,7 +263,7 @@ nfsd_cache_csum(struct svc_rqst *rqstp) base = buf->page_base & ~PAGE_MASK; while (csum_len) { p = page_address(buf->pages[idx]) + base; - len = min(PAGE_SIZE - base, csum_len); + len = min_t(size_t, PAGE_SIZE - base, csum_len); csum = csum_partial(p, len, csum); csum_len -= len; base = 0; -- cgit From 085b7a45c63d3da5be155faab9249a5cab224561 Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Fri, 15 Feb 2013 16:03:46 -0500 Subject: NFSv4.1: Don't decode skipped layoutgets layoutget's prepare hook can call rpc_exit with status = NFS4_OK (0). Because of this, nfs4_proc_layoutget can't depend on a 0 status to mean that the RPC was successfully sent, received and parsed. To fix this, use the result's len member to see if parsing took place. This fixes the following OOPS -- calling xdr_init_decode() with a buffer length 0 doesn't set the stream's 'p' member and ends up using uninitialized memory in filelayout_decode_layout. BUG: unable to handle kernel paging request at 0000000000008050 IP: [] memcpy+0x18/0x120 PGD 0 Oops: 0000 [#1] SMP last sysfs file: /sys/devices/pci0000:00/0000:00:11.0/0000:02:01.0/irq CPU 1 Modules linked in: nfs_layout_nfsv41_files nfs lockd fscache auth_rpcgss nfs_acl autofs4 sunrpc ipt_REJECT nf_conntrack_ipv4 nf_defrag_ipv4 iptable_filter ip_tables ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables ipv6 dm_mirror dm_region_hash dm_log dm_mod ppdev parport_pc parport snd_ens1371 snd_rawmidi snd_ac97_codec ac97_bus snd_seq snd_seq_device snd_pcm snd_timer snd soundcore snd_page_alloc e1000 microcode vmware_balloon i2c_piix4 i2c_core sg shpchp ext4 mbcache jbd2 sr_mod cdrom sd_mod crc_t10dif pata_acpi ata_generic ata_piix mptspi mptscsih mptbase scsi_transport_spi [last unloaded: speedstep_lib] Pid: 1665, comm: flush-0:22 Not tainted 2.6.32-356-test-2 #2 VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform RIP: 0010:[] [] memcpy+0x18/0x120 RSP: 0018:ffff88003dfab588 EFLAGS: 00010206 RAX: ffff88003dc42000 RBX: ffff88003dfab610 RCX: 0000000000000009 RDX: 000000003f807ff0 RSI: 0000000000008050 RDI: ffff88003dc42000 RBP: ffff88003dfab5b0 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000080 R12: 0000000000000024 R13: ffff88003dc42000 R14: ffff88003f808030 R15: ffff88003dfab6a0 FS: 0000000000000000(0000) GS:ffff880003420000(0000) knlGS:0000000000000000 CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b CR2: 0000000000008050 CR3: 000000003bc92000 CR4: 00000000001407e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process flush-0:22 (pid: 1665, threadinfo ffff88003dfaa000, task ffff880037f77540) Stack: ffffffffa0398ac1 ffff8800397c5940 ffff88003dfab610 ffff88003dfab6a0 ffff88003dfab5d0 ffff88003dfab680 ffffffffa01c150b ffffea0000d82e70 000000508116713b 0000000000000000 0000000000000000 0000000000000000 Call Trace: [] ? xdr_inline_decode+0xb1/0x120 [sunrpc] [] filelayout_decode_layout+0xeb/0x350 [nfs_layout_nfsv41_files] [] filelayout_alloc_lseg+0x8c/0x3c0 [nfs_layout_nfsv41_files] [] ? __wait_on_bit+0x7e/0x90 Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/nfs/nfs4proc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index cf252dfd91f8..eae83bf96c6d 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -6184,7 +6184,8 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags) status = nfs4_wait_for_completion_rpc_task(task); if (status == 0) status = task->tk_status; - if (status == 0) + /* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */ + if (status == 0 && lgp->res.layoutp->len) lseg = pnfs_layout_process(lgp); rpc_put_task(task); dprintk("<-- %s status=%d\n", __func__, status); -- cgit From 96aa1549afa6c79ae4a4f099de861efd218c38d8 Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Tue, 12 Feb 2013 13:03:42 -0700 Subject: nfs: remove kfree() redundant null checks smatch analysis: fs/nfs/getroot.c:130 nfs_get_root() info: redundant null check on name calling kfree() fs/nfs/unlink.c:272 nfs_async_unlink() info: redundant null check on devname_garbage calling kfree() Cc: Trond Myklebust Cc: linux-nfs@vger.kernel.org Signed-off-by: Tim Gardner Signed-off-by: Trond Myklebust --- fs/nfs/getroot.c | 3 +-- fs/nfs/unlink.c | 3 +-- 2 files changed, 2 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c index 033803c36644..44efaa8c5f78 100644 --- a/fs/nfs/getroot.c +++ b/fs/nfs/getroot.c @@ -126,8 +126,7 @@ struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh, } spin_unlock(&ret->d_lock); out: - if (name) - kfree(name); + kfree(name); nfs_free_fattr(fsinfo.fattr); return ret; } diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index 13cea637eff8..d26a32f5b53b 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -268,8 +268,7 @@ nfs_async_unlink(struct inode *dir, struct dentry *dentry) * point dentry is definitely not a root, so we won't need * that anymore. */ - if (devname_garbage) - kfree(devname_garbage); + kfree(devname_garbage); return 0; out_unlock: spin_unlock(&dentry->d_lock); -- cgit From 5a12cca697aca5dfba42a7d4c3356acc0445a2b0 Mon Sep 17 00:00:00 2001 From: fanchaoting Date: Mon, 4 Feb 2013 21:15:02 +0800 Subject: umount oops when remove blocklayoutdriver first now pnfs client uses block layout, maybe we can remove blocklayoutdriver first. if we umount later, it can cause oops in unset_pnfs_layoutdriver. because nfss->pnfs_curr_ld->clear_layoutdriver is invalid. reproduce it: modprobe blocklayoutdriver mount -t nfs4 -o minorversion=1 pnfsip:/ /mnt/ rmmod blocklayoutdriver umount /mnt then you can see following CPU 0 Pid: 17023, comm: umount.nfs4 Tainted: GF O 3.7.0-rc6-pnfs #1 VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform RIP: 0010:[] [] unset_pnfs_layoutdriver+0x1d/0x70 [nfsv4] RSP: 0018:ffff8800022d9e48 EFLAGS: 00010286 RAX: ffffffffa04a1b00 RBX: ffff88000b013800 RCX: 0000000000000001 RDX: ffffffff81ae8ee0 RSI: ffff880001ee94b8 RDI: ffff88000b013800 RBP: ffff8800022d9e58 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffff880001ee9400 R13: ffff8800105978c0 R14: 00007fff25846c08 R15: 0000000001bba550 FS: 00007f45ae7f0700(0000) GS:ffff880012c00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: ffffffffa04a1b38 CR3: 0000000002c0c000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process umount.nfs4 (pid: 17023, threadinfo ffff8800022d8000, task ffff880006e48aa0) Stack: ffff8800105978c0 ffff88000b013800 ffff8800022d9e78 ffffffffa04cd0ce ffff8800022d9e78 ffff88000b013800 ffff8800022d9ea8 ffffffffa04755a7 ffff8800022d9ea8 ffff880002f96400 ffff88000b013800 ffff880002f96400 Call Trace: [] nfs4_destroy_server+0x1e/0x30 [nfsv4] [] nfs_free_server+0xb7/0x150 [nfs] [] nfs_kill_super+0x35/0x40 [nfs] [] deactivate_locked_super+0x45/0x70 [] deactivate_super+0x4a/0x70 [] mntput_no_expire+0xd2/0x130 [] sys_umount+0x72/0xe0 [] system_call_fastpath+0x16/0x1b Code: 06 e1 b8 ea ff ff ff eb 9e 0f 1f 44 00 00 55 48 89 e5 53 48 83 ec 08 66 66 66 66 90 48 8b 87 80 03 00 00 48 89 fb 48 85 c0 74 29 <48> 8b 40 38 48 85 c0 74 02 ff d0 48 8b 03 3e ff 48 04 0f 94 c2 RIP [] unset_pnfs_layoutdriver+0x1d/0x70 [nfsv4] RSP CR2: ffffffffa04a1b38 ---[ end trace 29f75aaedda058bf ]--- Signed-off-by: fanchaoting Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/nfs/blocklayout/blocklayout.c | 1 + fs/nfs/objlayout/objio_osd.c | 1 + 2 files changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/nfs/blocklayout/blocklayout.c b/fs/nfs/blocklayout/blocklayout.c index 4fa788c93f46..434b93ec0970 100644 --- a/fs/nfs/blocklayout/blocklayout.c +++ b/fs/nfs/blocklayout/blocklayout.c @@ -1273,6 +1273,7 @@ static const struct nfs_pageio_ops bl_pg_write_ops = { static struct pnfs_layoutdriver_type blocklayout_type = { .id = LAYOUT_BLOCK_VOLUME, .name = "LAYOUT_BLOCK_VOLUME", + .owner = THIS_MODULE, .read_pagelist = bl_read_pagelist, .write_pagelist = bl_write_pagelist, .alloc_layout_hdr = bl_alloc_layout_hdr, diff --git a/fs/nfs/objlayout/objio_osd.c b/fs/nfs/objlayout/objio_osd.c index c6f990656f89..88f9611a945c 100644 --- a/fs/nfs/objlayout/objio_osd.c +++ b/fs/nfs/objlayout/objio_osd.c @@ -647,6 +647,7 @@ static struct pnfs_layoutdriver_type objlayout_type = { .flags = PNFS_LAYOUTRET_ON_SETATTR | PNFS_LAYOUTRET_ON_ERROR, + .owner = THIS_MODULE, .alloc_layout_hdr = objlayout_alloc_layout_hdr, .free_layout_hdr = objlayout_free_layout_hdr, -- cgit From 06b0c886214a223dde7b21cbfc3008fd20a8ce16 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:26:51 -0500 Subject: ext4: refine extent status tree This commit refines the extent status tree code. 1) A prefix 'es_' is added to to the extent status tree structure members. 2) Refactored es_remove_extent() so that __es_remove_extent() can be used by es_insert_extent() to remove the old extent entry(-ies) before inserting a new one. 3) Rename extent_status_end() to ext4_es_end() 4) ext4_es_can_be_merged() is define to check whether two extents can be merged or not. 5) Update and clarified comments. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/extents.c | 21 ++-- fs/ext4/extents_status.c | 322 ++++++++++++++++++++++++++--------------------- fs/ext4/extents_status.h | 8 +- fs/ext4/file.c | 12 +- 4 files changed, 201 insertions(+), 162 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index b6b54d658dc2..37f94a751ad7 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3528,13 +3528,14 @@ static int ext4_find_delalloc_range(struct inode *inode, { struct extent_status es; - es.start = lblk_start; - ext4_es_find_extent(inode, &es); - if (es.len == 0) + es.es_lblk = lblk_start; + (void)ext4_es_find_extent(inode, &es); + if (es.es_len == 0) return 0; /* there is no delay extent in this tree */ - else if (es.start <= lblk_start && lblk_start < es.start + es.len) + else if (es.es_lblk <= lblk_start && + lblk_start < es.es_lblk + es.es_len) return 1; - else if (lblk_start <= es.start && es.start <= lblk_end) + else if (lblk_start <= es.es_lblk && es.es_lblk <= lblk_end) return 1; else return 0; @@ -4569,7 +4570,7 @@ static int ext4_find_delayed_extent(struct inode *inode, struct extent_status es; ext4_lblk_t next_del; - es.start = newex->ec_block; + es.es_lblk = newex->ec_block; next_del = ext4_es_find_extent(inode, &es); if (newex->ec_start == 0) { @@ -4577,18 +4578,18 @@ static int ext4_find_delayed_extent(struct inode *inode, * No extent in extent-tree contains block @newex->ec_start, * then the block may stay in 1)a hole or 2)delayed-extent. */ - if (es.len == 0) + if (es.es_len == 0) /* A hole found. */ return 0; - if (es.start > newex->ec_block) { + if (es.es_lblk > newex->ec_block) { /* A hole found. */ - newex->ec_len = min(es.start - newex->ec_block, + newex->ec_len = min(es.es_lblk - newex->ec_block, newex->ec_len); return 0; } - newex->ec_len = es.start + es.len - newex->ec_block; + newex->ec_len = es.es_lblk + es.es_len - newex->ec_block; } return next_del; diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 564d981a2fcc..c9921cf4f817 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -23,40 +23,53 @@ * (e.g. Reservation space warning), and provide extent-level locking. * Delay extent tree is the first step to achieve this goal. It is * original built by Yongqiang Yang. At that time it is called delay - * extent tree, whose goal is only track delay extent in memory to + * extent tree, whose goal is only track delayed extents in memory to * simplify the implementation of fiemap and bigalloc, and introduce * lseek SEEK_DATA/SEEK_HOLE support. That is why it is still called - * delay extent tree at the following comment. But for better - * understand what it does, it has been rename to extent status tree. + * delay extent tree at the first commit. But for better understand + * what it does, it has been rename to extent status tree. * - * Currently the first step has been done. All delay extents are - * tracked in the tree. It maintains the delay extent when a delay - * allocation is issued, and the delay extent is written out or + * Step1: + * Currently the first step has been done. All delayed extents are + * tracked in the tree. It maintains the delayed extent when a delayed + * allocation is issued, and the delayed extent is written out or * invalidated. Therefore the implementation of fiemap and bigalloc * are simplified, and SEEK_DATA/SEEK_HOLE are introduced. * * The following comment describes the implemenmtation of extent * status tree and future works. + * + * Step2: + * In this step all extent status are tracked by extent status tree. + * Thus, we can first try to lookup a block mapping in this tree before + * finding it in extent tree. Hence, single extent cache can be removed + * because extent status tree can do a better job. Extents in status + * tree are loaded on-demand. Therefore, the extent status tree may not + * contain all of the extents in a file. Meanwhile we define a shrinker + * to reclaim memory from extent status tree because fragmented extent + * tree will make status tree cost too much memory. written/unwritten/- + * hole extents in the tree will be reclaimed by this shrinker when we + * are under high memory pressure. Delayed extents will not be + * reclimed because fiemap, bigalloc, and seek_data/hole need it. */ /* - * extents status tree implementation for ext4. + * Extent status tree implementation for ext4. * * * ========================================================================== - * Extents status encompass delayed extents and extent locks + * Extent status tree tracks all extent status. * - * 1. Why delayed extent implementation ? + * 1. Why we need to implement extent status tree? * - * Without delayed extent, ext4 identifies a delayed extent by looking + * Without extent status tree, ext4 identifies a delayed extent by looking * up page cache, this has several deficiencies - complicated, buggy, * and inefficient code. * - * FIEMAP, SEEK_HOLE/DATA, bigalloc, punch hole and writeout all need - * to know if a block or a range of blocks are belonged to a delayed - * extent. + * FIEMAP, SEEK_HOLE/DATA, bigalloc, and writeout all need to know if a + * block or a range of blocks are belonged to a delayed extent. * - * Let us have a look at how they do without delayed extents implementation. + * Let us have a look at how they do without extent status tree. * -- FIEMAP * FIEMAP looks up page cache to identify delayed allocations from holes. * @@ -68,47 +81,48 @@ * already under delayed allocation or not to determine whether * quota reserving is needed for the cluster. * - * -- punch hole - * punch hole looks up page cache to identify a delayed extent. - * * -- writeout * Writeout looks up whole page cache to see if a buffer is * mapped, If there are not very many delayed buffers, then it is * time comsuming. * - * With delayed extents implementation, FIEMAP, SEEK_HOLE/DATA, + * With extent status tree implementation, FIEMAP, SEEK_HOLE/DATA, * bigalloc and writeout can figure out if a block or a range of * blocks is under delayed allocation(belonged to a delayed extent) or - * not by searching the delayed extent tree. + * not by searching the extent tree. * * * ========================================================================== - * 2. ext4 delayed extents impelmentation + * 2. Ext4 extent status tree impelmentation + * + * -- extent + * A extent is a range of blocks which are contiguous logically and + * physically. Unlike extent in extent tree, this extent in ext4 is + * a in-memory struct, there is no corresponding on-disk data. There + * is no limit on length of extent, so an extent can contain as many + * blocks as they are contiguous logically and physically. * - * -- delayed extent - * A delayed extent is a range of blocks which are contiguous - * logically and under delayed allocation. Unlike extent in - * ext4, delayed extent in ext4 is a in-memory struct, there is - * no corresponding on-disk data. There is no limit on length of - * delayed extent, so a delayed extent can contain as many blocks - * as they are contiguous logically. + * -- extent status tree + * Every inode has an extent status tree and all allocation blocks + * are added to the tree with different status. The extent in the + * tree are ordered by logical block no. * - * -- delayed extent tree - * Every inode has a delayed extent tree and all under delayed - * allocation blocks are added to the tree as delayed extents. - * Delayed extents in the tree are ordered by logical block no. + * -- operations on a extent status tree + * There are three important operations on a delayed extent tree: find + * next extent, adding a extent(a range of blocks) and removing a extent. * - * -- operations on a delayed extent tree - * There are three operations on a delayed extent tree: find next - * delayed extent, adding a space(a range of blocks) and removing - * a space. + * -- race on a extent status tree + * Extent status tree is protected by inode->i_es_lock. * - * -- race on a delayed extent tree - * Delayed extent tree is protected inode->i_es_lock. + * -- memory consumption + * Fragmented extent tree will make extent status tree cost too much + * memory. Hence, we will reclaim written/unwritten/hole extents from + * the tree under a heavy memory pressure. * * * ========================================================================== - * 3. performance analysis + * 3. Performance analysis + * * -- overhead * 1. There is a cache extent for write access, so if writes are * not very random, adding space operaions are in O(1) time. @@ -120,15 +134,19 @@ * * ========================================================================== * 4. TODO list - * -- Track all extent status * - * -- Improve get block process + * -- Refactor delayed space reservation * * -- Extent-level locking */ static struct kmem_cache *ext4_es_cachep; +static int __es_insert_extent(struct ext4_es_tree *tree, + struct extent_status *newes); +static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, + ext4_lblk_t end); + int __init ext4_init_es(void) { ext4_es_cachep = KMEM_CACHE(extent_status, SLAB_RECLAIM_ACCOUNT); @@ -161,7 +179,7 @@ static void ext4_es_print_tree(struct inode *inode) while (node) { struct extent_status *es; es = rb_entry(node, struct extent_status, rb_node); - printk(KERN_DEBUG " [%u/%u)", es->start, es->len); + printk(KERN_DEBUG " [%u/%u)", es->es_lblk, es->es_len); node = rb_next(node); } printk(KERN_DEBUG "\n"); @@ -170,10 +188,10 @@ static void ext4_es_print_tree(struct inode *inode) #define ext4_es_print_tree(inode) #endif -static inline ext4_lblk_t extent_status_end(struct extent_status *es) +static inline ext4_lblk_t ext4_es_end(struct extent_status *es) { - BUG_ON(es->start + es->len < es->start); - return es->start + es->len - 1; + BUG_ON(es->es_lblk + es->es_len < es->es_lblk); + return es->es_lblk + es->es_len - 1; } /* @@ -181,25 +199,25 @@ static inline ext4_lblk_t extent_status_end(struct extent_status *es) * it can't be found, try to find next extent. */ static struct extent_status *__es_tree_search(struct rb_root *root, - ext4_lblk_t offset) + ext4_lblk_t lblk) { struct rb_node *node = root->rb_node; struct extent_status *es = NULL; while (node) { es = rb_entry(node, struct extent_status, rb_node); - if (offset < es->start) + if (lblk < es->es_lblk) node = node->rb_left; - else if (offset > extent_status_end(es)) + else if (lblk > ext4_es_end(es)) node = node->rb_right; else return es; } - if (es && offset < es->start) + if (es && lblk < es->es_lblk) return es; - if (es && offset > extent_status_end(es)) { + if (es && lblk > ext4_es_end(es)) { node = rb_next(&es->rb_node); return node ? rb_entry(node, struct extent_status, rb_node) : NULL; @@ -209,8 +227,8 @@ static struct extent_status *__es_tree_search(struct rb_root *root, } /* - * ext4_es_find_extent: find the 1st delayed extent covering @es->start - * if it exists, otherwise, the next extent after @es->start. + * ext4_es_find_extent: find the 1st delayed extent covering @es->lblk + * if it exists, otherwise, the next extent after @es->lblk. * * @inode: the inode which owns delayed extents * @es: delayed extent that we found @@ -226,7 +244,7 @@ ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es) struct rb_node *node; ext4_lblk_t ret = EXT_MAX_BLOCKS; - trace_ext4_es_find_extent_enter(inode, es->start); + trace_ext4_es_find_extent_enter(inode, es->es_lblk); read_lock(&EXT4_I(inode)->i_es_lock); tree = &EXT4_I(inode)->i_es_tree; @@ -234,25 +252,25 @@ ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es) /* find delay extent in cache firstly */ if (tree->cache_es) { es1 = tree->cache_es; - if (in_range(es->start, es1->start, es1->len)) { + if (in_range(es->es_lblk, es1->es_lblk, es1->es_len)) { es_debug("%u cached by [%u/%u)\n", - es->start, es1->start, es1->len); + es->es_lblk, es1->es_lblk, es1->es_len); goto out; } } - es->len = 0; - es1 = __es_tree_search(&tree->root, es->start); + es->es_len = 0; + es1 = __es_tree_search(&tree->root, es->es_lblk); out: if (es1) { tree->cache_es = es1; - es->start = es1->start; - es->len = es1->len; + es->es_lblk = es1->es_lblk; + es->es_len = es1->es_len; node = rb_next(&es1->rb_node); if (node) { es1 = rb_entry(node, struct extent_status, rb_node); - ret = es1->start; + ret = es1->es_lblk; } } @@ -263,14 +281,14 @@ out: } static struct extent_status * -ext4_es_alloc_extent(ext4_lblk_t start, ext4_lblk_t len) +ext4_es_alloc_extent(ext4_lblk_t lblk, ext4_lblk_t len) { struct extent_status *es; es = kmem_cache_alloc(ext4_es_cachep, GFP_ATOMIC); if (es == NULL) return NULL; - es->start = start; - es->len = len; + es->es_lblk = lblk; + es->es_len = len; return es; } @@ -279,6 +297,20 @@ static void ext4_es_free_extent(struct extent_status *es) kmem_cache_free(ext4_es_cachep, es); } +/* + * Check whether or not two extents can be merged + * Condition: + * - logical block number is contiguous + */ +static int ext4_es_can_be_merged(struct extent_status *es1, + struct extent_status *es2) +{ + if (es1->es_lblk + es1->es_len != es2->es_lblk) + return 0; + + return 1; +} + static struct extent_status * ext4_es_try_to_merge_left(struct ext4_es_tree *tree, struct extent_status *es) { @@ -290,8 +322,8 @@ ext4_es_try_to_merge_left(struct ext4_es_tree *tree, struct extent_status *es) return es; es1 = rb_entry(node, struct extent_status, rb_node); - if (es->start == extent_status_end(es1) + 1) { - es1->len += es->len; + if (ext4_es_can_be_merged(es1, es)) { + es1->es_len += es->es_len; rb_erase(&es->rb_node, &tree->root); ext4_es_free_extent(es); es = es1; @@ -311,8 +343,8 @@ ext4_es_try_to_merge_right(struct ext4_es_tree *tree, struct extent_status *es) return es; es1 = rb_entry(node, struct extent_status, rb_node); - if (es1->start == extent_status_end(es) + 1) { - es->len += es1->len; + if (ext4_es_can_be_merged(es, es1)) { + es->es_len += es1->es_len; rb_erase(node, &tree->root); ext4_es_free_extent(es1); } @@ -320,60 +352,43 @@ ext4_es_try_to_merge_right(struct ext4_es_tree *tree, struct extent_status *es) return es; } -static int __es_insert_extent(struct ext4_es_tree *tree, ext4_lblk_t offset, - ext4_lblk_t len) +static int __es_insert_extent(struct ext4_es_tree *tree, + struct extent_status *newes) { struct rb_node **p = &tree->root.rb_node; struct rb_node *parent = NULL; struct extent_status *es; - ext4_lblk_t end = offset + len - 1; - - BUG_ON(end < offset); - es = tree->cache_es; - if (es && offset == (extent_status_end(es) + 1)) { - es_debug("cached by [%u/%u)\n", es->start, es->len); - es->len += len; - es = ext4_es_try_to_merge_right(tree, es); - goto out; - } else if (es && es->start == end + 1) { - es_debug("cached by [%u/%u)\n", es->start, es->len); - es->start = offset; - es->len += len; - es = ext4_es_try_to_merge_left(tree, es); - goto out; - } else if (es && es->start <= offset && - end <= extent_status_end(es)) { - es_debug("cached by [%u/%u)\n", es->start, es->len); - goto out; - } while (*p) { parent = *p; es = rb_entry(parent, struct extent_status, rb_node); - if (offset < es->start) { - if (es->start == end + 1) { - es->start = offset; - es->len += len; + if (newes->es_lblk < es->es_lblk) { + if (ext4_es_can_be_merged(newes, es)) { + /* + * Here we can modify es_lblk directly + * because it isn't overlapped. + */ + es->es_lblk = newes->es_lblk; + es->es_len += newes->es_len; es = ext4_es_try_to_merge_left(tree, es); goto out; } p = &(*p)->rb_left; - } else if (offset > extent_status_end(es)) { - if (offset == extent_status_end(es) + 1) { - es->len += len; + } else if (newes->es_lblk > ext4_es_end(es)) { + if (ext4_es_can_be_merged(es, newes)) { + es->es_len += newes->es_len; es = ext4_es_try_to_merge_right(tree, es); goto out; } p = &(*p)->rb_right; } else { - if (extent_status_end(es) <= end) - es->len = offset - es->start + len; - goto out; + BUG_ON(1); + return -EINVAL; } } - es = ext4_es_alloc_extent(offset, len); + es = ext4_es_alloc_extent(newes->es_lblk, newes->es_len); if (!es) return -ENOMEM; rb_link_node(&es->rb_node, parent, p); @@ -385,27 +400,38 @@ out: } /* - * ext4_es_insert_extent() adds a space to a delayed extent tree. - * Caller holds inode->i_es_lock. + * ext4_es_insert_extent() adds a space to a extent status tree. * * ext4_es_insert_extent is called by ext4_da_write_begin and * ext4_es_remove_extent. * * Return 0 on success, error code on failure. */ -int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t offset, +int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len) { struct ext4_es_tree *tree; + struct extent_status newes; + ext4_lblk_t end = lblk + len - 1; int err = 0; - trace_ext4_es_insert_extent(inode, offset, len); + trace_ext4_es_insert_extent(inode, lblk, len); es_debug("add [%u/%u) to extent status tree of inode %lu\n", - offset, len, inode->i_ino); + lblk, len, inode->i_ino); + + BUG_ON(end < lblk); + + newes.es_lblk = lblk; + newes.es_len = len; write_lock(&EXT4_I(inode)->i_es_lock); tree = &EXT4_I(inode)->i_es_tree; - err = __es_insert_extent(tree, offset, len); + err = __es_remove_extent(tree, lblk, end); + if (err != 0) + goto error; + err = __es_insert_extent(tree, &newes); + +error: write_unlock(&EXT4_I(inode)->i_es_lock); ext4_es_print_tree(inode); @@ -413,57 +439,45 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t offset, return err; } -/* - * ext4_es_remove_extent() removes a space from a delayed extent tree. - * Caller holds inode->i_es_lock. - * - * Return 0 on success, error code on failure. - */ -int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t offset, - ext4_lblk_t len) +static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, + ext4_lblk_t end) { struct rb_node *node; - struct ext4_es_tree *tree; struct extent_status *es; struct extent_status orig_es; - ext4_lblk_t len1, len2, end; + ext4_lblk_t len1, len2; int err = 0; - trace_ext4_es_remove_extent(inode, offset, len); - es_debug("remove [%u/%u) from extent status tree of inode %lu\n", - offset, len, inode->i_ino); - - end = offset + len - 1; - BUG_ON(end < offset); - write_lock(&EXT4_I(inode)->i_es_lock); - tree = &EXT4_I(inode)->i_es_tree; - es = __es_tree_search(&tree->root, offset); + es = __es_tree_search(&tree->root, lblk); if (!es) goto out; - if (es->start > end) + if (es->es_lblk > end) goto out; /* Simply invalidate cache_es. */ tree->cache_es = NULL; - orig_es.start = es->start; - orig_es.len = es->len; - len1 = offset > es->start ? offset - es->start : 0; - len2 = extent_status_end(es) > end ? - extent_status_end(es) - end : 0; + orig_es.es_lblk = es->es_lblk; + orig_es.es_len = es->es_len; + len1 = lblk > es->es_lblk ? lblk - es->es_lblk : 0; + len2 = ext4_es_end(es) > end ? ext4_es_end(es) - end : 0; if (len1 > 0) - es->len = len1; + es->es_len = len1; if (len2 > 0) { if (len1 > 0) { - err = __es_insert_extent(tree, end + 1, len2); + struct extent_status newes; + + newes.es_lblk = end + 1; + newes.es_len = len2; + err = __es_insert_extent(tree, &newes); if (err) { - es->start = orig_es.start; - es->len = orig_es.len; + es->es_lblk = orig_es.es_lblk; + es->es_len = orig_es.es_len; goto out; } } else { - es->start = end + 1; - es->len = len2; + es->es_lblk = end + 1; + es->es_len = len2; } goto out; } @@ -476,7 +490,7 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t offset, es = NULL; } - while (es && extent_status_end(es) <= end) { + while (es && ext4_es_end(es) <= end) { node = rb_next(&es->rb_node); rb_erase(&es->rb_node, &tree->root); ext4_es_free_extent(es); @@ -487,13 +501,39 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t offset, es = rb_entry(node, struct extent_status, rb_node); } - if (es && es->start < end + 1) { - len1 = extent_status_end(es) - end; - es->start = end + 1; - es->len = len1; + if (es && es->es_lblk < end + 1) { + len1 = ext4_es_end(es) - end; + es->es_lblk = end + 1; + es->es_len = len1; } out: + return err; +} + +/* + * ext4_es_remove_extent() removes a space from a extent status tree. + * + * Return 0 on success, error code on failure. + */ +int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, + ext4_lblk_t len) +{ + struct ext4_es_tree *tree; + ext4_lblk_t end; + int err = 0; + + trace_ext4_es_remove_extent(inode, lblk, len); + es_debug("remove [%u/%u) from extent status tree of inode %lu\n", + lblk, len, inode->i_ino); + + end = lblk + len - 1; + BUG_ON(end < lblk); + + tree = &EXT4_I(inode)->i_es_tree; + + write_lock(&EXT4_I(inode)->i_es_lock); + err = __es_remove_extent(tree, lblk, end); write_unlock(&EXT4_I(inode)->i_es_lock); ext4_es_print_tree(inode); return err; diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index 077f82db092a..81e9339f23f1 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -22,8 +22,8 @@ struct extent_status { struct rb_node rb_node; - ext4_lblk_t start; /* first block extent covers */ - ext4_lblk_t len; /* length of extent in block */ + ext4_lblk_t es_lblk; /* first logical block extent covers */ + ext4_lblk_t es_len; /* length of extent in block */ }; struct ext4_es_tree { @@ -35,9 +35,9 @@ extern int __init ext4_init_es(void); extern void ext4_exit_es(void); extern void ext4_es_init_tree(struct ext4_es_tree *tree); -extern int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t start, +extern int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len); -extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t start, +extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len); extern ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es); diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 2cf8ab810687..2df9354b105e 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -464,10 +464,9 @@ static loff_t ext4_seek_data(struct file *file, loff_t offset, loff_t maxsize) * If there is a delay extent at this offset, * it will be as a data. */ - es.start = last; + es.es_lblk = last; (void)ext4_es_find_extent(inode, &es); - if (last >= es.start && - last < es.start + es.len) { + if (es.es_len != 0 && in_range(last, es.es_lblk, es.es_len)) { if (last != start) dataoff = last << blkbits; break; @@ -549,11 +548,10 @@ static loff_t ext4_seek_hole(struct file *file, loff_t offset, loff_t maxsize) * If there is a delay extent at this offset, * we will skip this extent. */ - es.start = last; + es.es_lblk = last; (void)ext4_es_find_extent(inode, &es); - if (last >= es.start && - last < es.start + es.len) { - last = es.start + es.len; + if (es.es_len != 0 && in_range(last, es.es_lblk, es.es_len)) { + last = es.es_lblk + es.es_len; holeoff = last << blkbits; continue; } -- cgit From fdc0212e86ca15c5cfed77088af7cc5eb79ccbc7 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:26:51 -0500 Subject: ext4: add physical block and status member into extent status tree This commit adds two members in extent_status structure to let it record physical block and extent status. Here es_pblk is used to record both of them because physical block only has 48 bits. So extent status could be stashed into it so that we can save some memory. Now written, unwritten, delayed and hole are defined as status. Due to new member is added into extent status tree, all interfaces need to be adjusted. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/extents_status.c | 67 +++++++++++++++++++++++++++++++++++++++--------- fs/ext4/extents_status.h | 64 ++++++++++++++++++++++++++++++++++++++++++++- fs/ext4/inode.c | 3 ++- 3 files changed, 120 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index c9921cf4f817..1f5fd44993e9 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -179,7 +179,9 @@ static void ext4_es_print_tree(struct inode *inode) while (node) { struct extent_status *es; es = rb_entry(node, struct extent_status, rb_node); - printk(KERN_DEBUG " [%u/%u)", es->es_lblk, es->es_len); + printk(KERN_DEBUG " [%u/%u) %llu %llx", + es->es_lblk, es->es_len, + ext4_es_pblock(es), ext4_es_status(es)); node = rb_next(node); } printk(KERN_DEBUG "\n"); @@ -234,7 +236,7 @@ static struct extent_status *__es_tree_search(struct rb_root *root, * @es: delayed extent that we found * * Returns the first block of the next extent after es, otherwise - * EXT_MAX_BLOCKS if no delay extent is found. + * EXT_MAX_BLOCKS if no extent is found. * Delayed extent is returned via @es. */ ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es) @@ -249,17 +251,18 @@ ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es) read_lock(&EXT4_I(inode)->i_es_lock); tree = &EXT4_I(inode)->i_es_tree; - /* find delay extent in cache firstly */ + /* find extent in cache firstly */ + es->es_len = es->es_pblk = 0; if (tree->cache_es) { es1 = tree->cache_es; if (in_range(es->es_lblk, es1->es_lblk, es1->es_len)) { - es_debug("%u cached by [%u/%u)\n", - es->es_lblk, es1->es_lblk, es1->es_len); + es_debug("%u cached by [%u/%u) %llu %llx\n", + es->es_lblk, es1->es_lblk, es1->es_len, + ext4_es_pblock(es1), ext4_es_status(es1)); goto out; } } - es->es_len = 0; es1 = __es_tree_search(&tree->root, es->es_lblk); out: @@ -267,6 +270,7 @@ out: tree->cache_es = es1; es->es_lblk = es1->es_lblk; es->es_len = es1->es_len; + es->es_pblk = es1->es_pblk; node = rb_next(&es1->rb_node); if (node) { es1 = rb_entry(node, struct extent_status, rb_node); @@ -281,7 +285,7 @@ out: } static struct extent_status * -ext4_es_alloc_extent(ext4_lblk_t lblk, ext4_lblk_t len) +ext4_es_alloc_extent(ext4_lblk_t lblk, ext4_lblk_t len, ext4_fsblk_t pblk) { struct extent_status *es; es = kmem_cache_alloc(ext4_es_cachep, GFP_ATOMIC); @@ -289,6 +293,7 @@ ext4_es_alloc_extent(ext4_lblk_t lblk, ext4_lblk_t len) return NULL; es->es_lblk = lblk; es->es_len = len; + es->es_pblk = pblk; return es; } @@ -301,6 +306,8 @@ static void ext4_es_free_extent(struct extent_status *es) * Check whether or not two extents can be merged * Condition: * - logical block number is contiguous + * - physical block number is contiguous + * - status is equal */ static int ext4_es_can_be_merged(struct extent_status *es1, struct extent_status *es2) @@ -308,6 +315,13 @@ static int ext4_es_can_be_merged(struct extent_status *es1, if (es1->es_lblk + es1->es_len != es2->es_lblk) return 0; + if (ext4_es_status(es1) != ext4_es_status(es2)) + return 0; + + if ((ext4_es_is_written(es1) || ext4_es_is_unwritten(es1)) && + (ext4_es_pblock(es1) + es1->es_len != ext4_es_pblock(es2))) + return 0; + return 1; } @@ -371,6 +385,10 @@ static int __es_insert_extent(struct ext4_es_tree *tree, */ es->es_lblk = newes->es_lblk; es->es_len += newes->es_len; + if (ext4_es_is_written(es) || + ext4_es_is_unwritten(es)) + ext4_es_store_pblock(es, + newes->es_pblk); es = ext4_es_try_to_merge_left(tree, es); goto out; } @@ -388,7 +406,8 @@ static int __es_insert_extent(struct ext4_es_tree *tree, } } - es = ext4_es_alloc_extent(newes->es_lblk, newes->es_len); + es = ext4_es_alloc_extent(newes->es_lblk, newes->es_len, + newes->es_pblk); if (!es) return -ENOMEM; rb_link_node(&es->rb_node, parent, p); @@ -408,21 +427,24 @@ out: * Return 0 on success, error code on failure. */ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, - ext4_lblk_t len) + ext4_lblk_t len, ext4_fsblk_t pblk, + unsigned long long status) { struct ext4_es_tree *tree; struct extent_status newes; ext4_lblk_t end = lblk + len - 1; int err = 0; - trace_ext4_es_insert_extent(inode, lblk, len); - es_debug("add [%u/%u) to extent status tree of inode %lu\n", - lblk, len, inode->i_ino); + es_debug("add [%u/%u) %llu %llx to extent status tree of inode %lu\n", + lblk, len, pblk, status, inode->i_ino); BUG_ON(end < lblk); newes.es_lblk = lblk; newes.es_len = len; + ext4_es_store_pblock(&newes, pblk); + ext4_es_store_status(&newes, status); + trace_ext4_es_insert_extent(inode, &newes); write_lock(&EXT4_I(inode)->i_es_lock); tree = &EXT4_I(inode)->i_es_tree; @@ -446,6 +468,7 @@ static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, struct extent_status *es; struct extent_status orig_es; ext4_lblk_t len1, len2; + ext4_fsblk_t block; int err = 0; es = __es_tree_search(&tree->root, lblk); @@ -459,6 +482,8 @@ static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, orig_es.es_lblk = es->es_lblk; orig_es.es_len = es->es_len; + orig_es.es_pblk = es->es_pblk; + len1 = lblk > es->es_lblk ? lblk - es->es_lblk : 0; len2 = ext4_es_end(es) > end ? ext4_es_end(es) - end : 0; if (len1 > 0) @@ -469,6 +494,13 @@ static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, newes.es_lblk = end + 1; newes.es_len = len2; + if (ext4_es_is_written(&orig_es) || + ext4_es_is_unwritten(&orig_es)) { + block = ext4_es_pblock(&orig_es) + + orig_es.es_len - len2; + ext4_es_store_pblock(&newes, block); + } + ext4_es_store_status(&newes, ext4_es_status(&orig_es)); err = __es_insert_extent(tree, &newes); if (err) { es->es_lblk = orig_es.es_lblk; @@ -478,6 +510,11 @@ static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, } else { es->es_lblk = end + 1; es->es_len = len2; + if (ext4_es_is_written(es) || + ext4_es_is_unwritten(es)) { + block = orig_es.es_pblk + orig_es.es_len - len2; + ext4_es_store_pblock(es, block); + } } goto out; } @@ -502,9 +539,15 @@ static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, } if (es && es->es_lblk < end + 1) { + ext4_lblk_t orig_len = es->es_len; + len1 = ext4_es_end(es) - end; es->es_lblk = end + 1; es->es_len = len1; + if (ext4_es_is_written(es) || ext4_es_is_unwritten(es)) { + block = es->es_pblk + orig_len - len1; + ext4_es_store_pblock(es, block); + } } out: diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index 81e9339f23f1..3cad83303adb 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -20,10 +20,21 @@ #define es_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) #endif +#define EXTENT_STATUS_WRITTEN 0x80000000 /* written extent */ +#define EXTENT_STATUS_UNWRITTEN 0x40000000 /* unwritten extent */ +#define EXTENT_STATUS_DELAYED 0x20000000 /* delayed extent */ +#define EXTENT_STATUS_HOLE 0x10000000 /* hole */ + +#define EXTENT_STATUS_FLAGS (EXTENT_STATUS_WRITTEN | \ + EXTENT_STATUS_UNWRITTEN | \ + EXTENT_STATUS_DELAYED | \ + EXTENT_STATUS_HOLE) + struct extent_status { struct rb_node rb_node; ext4_lblk_t es_lblk; /* first logical block extent covers */ ext4_lblk_t es_len; /* length of extent in block */ + ext4_fsblk_t es_pblk; /* first physical block */ }; struct ext4_es_tree { @@ -36,10 +47,61 @@ extern void ext4_exit_es(void); extern void ext4_es_init_tree(struct ext4_es_tree *tree); extern int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, - ext4_lblk_t len); + ext4_lblk_t len, ext4_fsblk_t pblk, + unsigned long long status); extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len); extern ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es); +static inline int ext4_es_is_written(struct extent_status *es) +{ + return (es->es_pblk & EXTENT_STATUS_WRITTEN); +} + +static inline int ext4_es_is_unwritten(struct extent_status *es) +{ + return (es->es_pblk & EXTENT_STATUS_UNWRITTEN); +} + +static inline int ext4_es_is_delayed(struct extent_status *es) +{ + return (es->es_pblk & EXTENT_STATUS_DELAYED); +} + +static inline int ext4_es_is_hole(struct extent_status *es) +{ + return (es->es_pblk & EXTENT_STATUS_HOLE); +} + +static inline ext4_fsblk_t ext4_es_status(struct extent_status *es) +{ + return (es->es_pblk & EXTENT_STATUS_FLAGS); +} + +static inline ext4_fsblk_t ext4_es_pblock(struct extent_status *es) +{ + return (es->es_pblk & ~EXTENT_STATUS_FLAGS); +} + +static inline void ext4_es_store_pblock(struct extent_status *es, + ext4_fsblk_t pb) +{ + ext4_fsblk_t block; + + block = (pb & ~EXTENT_STATUS_FLAGS) | + (es->es_pblk & EXTENT_STATUS_FLAGS); + es->es_pblk = block; +} + +static inline void ext4_es_store_status(struct extent_status *es, + unsigned long long status) +{ + ext4_fsblk_t block; + + block = (status & EXTENT_STATUS_FLAGS) | + (es->es_pblk & ~EXTENT_STATUS_FLAGS); + es->es_pblk = block; +} + #endif /* _EXT4_EXTENTS_STATUS_H */ diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index f4466c3650dc..e0e1cb0863f4 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1784,7 +1784,8 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, goto out_unlock; } - retval = ext4_es_insert_extent(inode, map->m_lblk, map->m_len); + retval = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, + ~0, EXTENT_STATUS_DELAYED); if (retval) goto out_unlock; -- cgit From be401363ac5ec652c706263a59b0bd0acc3612e8 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:27:26 -0500 Subject: ext4: rename and improbe ext4_es_find_extent() This commit renames ext4_es_find_extent with ext4_es_find_delayed_extent and improve this function. First, we split input and output parameter. Second, this function never return the first block of the next delayed extent after 'es'. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Jan kara --- fs/ext4/extents.c | 15 ++++++++++----- fs/ext4/extents_status.c | 40 ++++++++++++++++++++-------------------- fs/ext4/extents_status.h | 4 ++-- fs/ext4/file.c | 6 ++---- 4 files changed, 34 insertions(+), 31 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 37f94a751ad7..895c19595ecf 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3528,8 +3528,7 @@ static int ext4_find_delalloc_range(struct inode *inode, { struct extent_status es; - es.es_lblk = lblk_start; - (void)ext4_es_find_extent(inode, &es); + ext4_es_find_delayed_extent(inode, lblk_start, &es); if (es.es_len == 0) return 0; /* there is no delay extent in this tree */ else if (es.es_lblk <= lblk_start && @@ -4568,10 +4567,9 @@ static int ext4_find_delayed_extent(struct inode *inode, struct ext4_ext_cache *newex) { struct extent_status es; - ext4_lblk_t next_del; + ext4_lblk_t block, next_del; - es.es_lblk = newex->ec_block; - next_del = ext4_es_find_extent(inode, &es); + ext4_es_find_delayed_extent(inode, newex->ec_block, &es); if (newex->ec_start == 0) { /* @@ -4592,6 +4590,13 @@ static int ext4_find_delayed_extent(struct inode *inode, newex->ec_len = es.es_lblk + es.es_len - newex->ec_block; } + block = newex->ec_block + newex->ec_len; + ext4_es_find_delayed_extent(inode, block, &es); + if (es.es_len == 0) + next_del = EXT_MAX_BLOCKS; + else + next_del = es.es_lblk; + return next_del; } /* fiemap flags we can handle specified here */ diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 1f5fd44993e9..76f4351ea821 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -229,59 +229,59 @@ static struct extent_status *__es_tree_search(struct rb_root *root, } /* - * ext4_es_find_extent: find the 1st delayed extent covering @es->lblk + * ext4_es_find_delayed_extent: find the 1st delayed extent covering @es->lblk * if it exists, otherwise, the next extent after @es->lblk. * * @inode: the inode which owns delayed extents + * @lblk: the offset where we start to search * @es: delayed extent that we found - * - * Returns the first block of the next extent after es, otherwise - * EXT_MAX_BLOCKS if no extent is found. - * Delayed extent is returned via @es. */ -ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es) +void ext4_es_find_delayed_extent(struct inode *inode, ext4_lblk_t lblk, + struct extent_status *es) { struct ext4_es_tree *tree = NULL; struct extent_status *es1 = NULL; struct rb_node *node; - ext4_lblk_t ret = EXT_MAX_BLOCKS; - trace_ext4_es_find_extent_enter(inode, es->es_lblk); + BUG_ON(es == NULL); + trace_ext4_es_find_delayed_extent_enter(inode, lblk); read_lock(&EXT4_I(inode)->i_es_lock); tree = &EXT4_I(inode)->i_es_tree; /* find extent in cache firstly */ - es->es_len = es->es_pblk = 0; + es->es_lblk = es->es_len = es->es_pblk = 0; if (tree->cache_es) { es1 = tree->cache_es; - if (in_range(es->es_lblk, es1->es_lblk, es1->es_len)) { + if (in_range(lblk, es1->es_lblk, es1->es_len)) { es_debug("%u cached by [%u/%u) %llu %llx\n", - es->es_lblk, es1->es_lblk, es1->es_len, + lblk, es1->es_lblk, es1->es_len, ext4_es_pblock(es1), ext4_es_status(es1)); goto out; } } - es1 = __es_tree_search(&tree->root, es->es_lblk); + es1 = __es_tree_search(&tree->root, lblk); out: - if (es1) { + if (es1 && !ext4_es_is_delayed(es1)) { + while ((node = rb_next(&es1->rb_node)) != NULL) { + es1 = rb_entry(node, struct extent_status, rb_node); + if (ext4_es_is_delayed(es1)) + break; + } + } + + if (es1 && ext4_es_is_delayed(es1)) { tree->cache_es = es1; es->es_lblk = es1->es_lblk; es->es_len = es1->es_len; es->es_pblk = es1->es_pblk; - node = rb_next(&es1->rb_node); - if (node) { - es1 = rb_entry(node, struct extent_status, rb_node); - ret = es1->es_lblk; - } } read_unlock(&EXT4_I(inode)->i_es_lock); - trace_ext4_es_find_extent_exit(inode, es, ret); - return ret; + trace_ext4_es_find_delayed_extent_exit(inode, es); } static struct extent_status * diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index 3cad83303adb..3f69d097c6e7 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -51,8 +51,8 @@ extern int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, unsigned long long status); extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len); -extern ext4_lblk_t ext4_es_find_extent(struct inode *inode, - struct extent_status *es); +extern void ext4_es_find_delayed_extent(struct inode *inode, ext4_lblk_t lblk, + struct extent_status *es); static inline int ext4_es_is_written(struct extent_status *es) { diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 2df9354b105e..7e85a10a6f4f 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -464,8 +464,7 @@ static loff_t ext4_seek_data(struct file *file, loff_t offset, loff_t maxsize) * If there is a delay extent at this offset, * it will be as a data. */ - es.es_lblk = last; - (void)ext4_es_find_extent(inode, &es); + ext4_es_find_delayed_extent(inode, last, &es); if (es.es_len != 0 && in_range(last, es.es_lblk, es.es_len)) { if (last != start) dataoff = last << blkbits; @@ -548,8 +547,7 @@ static loff_t ext4_seek_hole(struct file *file, loff_t offset, loff_t maxsize) * If there is a delay extent at this offset, * we will skip this extent. */ - es.es_lblk = last; - (void)ext4_es_find_extent(inode, &es); + ext4_es_find_delayed_extent(inode, last, &es); if (es.es_len != 0 && in_range(last, es.es_lblk, es.es_len)) { last = es.es_lblk + es.es_len; holeoff = last << blkbits; -- cgit From a25a4e1a5d5dc0f97dddbca44e695c532d8228c1 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:28:04 -0500 Subject: ext4: let ext4_ext_map_blocks return EXT4_MAP_UNWRITTEN flag This commit lets ext4_ext_map_blocks return EXT4_MAP_UNWRITTEN flag because in later commit ext4_map_blocks needs to use this flag to determine the extent status. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/extents.c | 6 +++++- fs/ext4/inode.c | 12 +++--------- 2 files changed, 8 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 895c19595ecf..cae8ae3c1746 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3659,6 +3659,7 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, ext4_set_io_unwritten_flag(inode, io); else ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN); + map->m_flags |= EXT4_MAP_UNWRITTEN; if (ext4_should_dioread_nolock(inode)) map->m_flags |= EXT4_MAP_UNINIT; goto out; @@ -3680,8 +3681,10 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, * repeat fallocate creation request * we already have an unwritten extent */ - if (flags & EXT4_GET_BLOCKS_UNINIT_EXT) + if (flags & EXT4_GET_BLOCKS_UNINIT_EXT) { + map->m_flags |= EXT4_MAP_UNWRITTEN; goto map_out; + } /* buffered READ or buffered write_begin() lookup */ if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { @@ -4111,6 +4114,7 @@ got_allocated_blocks: /* Mark uninitialized */ if (flags & EXT4_GET_BLOCKS_UNINIT_EXT){ ext4_ext_mark_uninitialized(&newex); + map->m_flags |= EXT4_MAP_UNWRITTEN; /* * io_end structure was created for every IO write to an * uninitialized extent. To avoid unnecessary conversion, diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index e0e1cb0863f4..084b8212cf95 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -559,16 +559,10 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, return retval; /* - * When we call get_blocks without the create flag, the - * BH_Unwritten flag could have gotten set if the blocks - * requested were part of a uninitialized extent. We need to - * clear this flag now that we are committed to convert all or - * part of the uninitialized extent to be an initialized - * extent. This is because we need to avoid the combination - * of BH_Unwritten and BH_Mapped flags being simultaneously - * set on the buffer_head. + * Here we clear m_flags because after allocating an new extent, + * it will be set again. */ - map->m_flags &= ~EXT4_MAP_UNWRITTEN; + map->m_flags &= ~EXT4_MAP_FLAGS; /* * New blocks allocate and/or writing to uninitialized extent -- cgit From f7fec032aa782d3fd7e51fbdf08aa3a296c01500 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:28:47 -0500 Subject: ext4: track all extent status in extent status tree By recording the phycisal block and status, extent status tree is able to track the status of every extents. When we call _map_blocks functions to lookup an extent or create a new written/unwritten/delayed extent, this extent will be inserted into extent status tree. We don't load all extents from disk in alloc_inode() because it costs too much memory, and if a file is opened and closed frequently it will takes too much time to load all extent information. So currently when we create/lookup an extent, this extent will be inserted into extent status tree. Hence, the extent status tree may not comprehensively contain all of the extents found in the file. Here a condition we need to take care is that an extent might contains unwritten and delayed status simultaneously because an extent is delayed allocated and could be allocated by fallocate. At this time we need to keep delayed status because later we need to update delayed reservation space using it. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Jan kara --- fs/ext4/ext4.h | 3 +++ fs/ext4/extents.c | 20 +++++++++++---- fs/ext4/inode.c | 76 +++++++++++++++++++++++++++++++++++++------------------ 3 files changed, 69 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index fc1c0375c9f2..5c31d6ac9500 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2602,6 +2602,9 @@ extern struct ext4_ext_path *ext4_ext_find_extent(struct inode *, ext4_lblk_t, struct ext4_ext_path *); extern void ext4_ext_drop_refs(struct ext4_ext_path *); extern int ext4_ext_check_inode(struct inode *inode); +extern int ext4_find_delalloc_range(struct inode *inode, + ext4_lblk_t lblk_start, + ext4_lblk_t lblk_end); extern int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk); extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index cae8ae3c1746..be0b1b3eed97 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -2076,8 +2076,18 @@ static int ext4_fill_fiemap_extents(struct inode *inode, break; } - /* This is possible iff next == next_del == EXT_MAX_BLOCKS */ - if (next == next_del) { + /* + * This is possible iff next == next_del == EXT_MAX_BLOCKS. + * we need to check next == EXT_MAX_BLOCKS because it is + * possible that an extent is with unwritten and delayed + * status due to when an extent is delayed allocated and + * is allocated by fallocate status tree will track both of + * them in a extent. + * + * So we could return a unwritten and delayed extent, and + * its block is equal to 'next'. + */ + if (next == next_del && next == EXT_MAX_BLOCKS) { flags |= FIEMAP_EXTENT_LAST; if (unlikely(next_del != EXT_MAX_BLOCKS || next != EXT_MAX_BLOCKS)) { @@ -3522,9 +3532,9 @@ out: * * Return 1 if there is a delalloc block in the range, otherwise 0. */ -static int ext4_find_delalloc_range(struct inode *inode, - ext4_lblk_t lblk_start, - ext4_lblk_t lblk_end) +int ext4_find_delalloc_range(struct inode *inode, + ext4_lblk_t lblk_start, + ext4_lblk_t lblk_end) { struct extent_status es; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 084b8212cf95..576b586b61aa 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -526,20 +526,26 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, retval = ext4_ind_map_blocks(handle, inode, map, flags & EXT4_GET_BLOCKS_KEEP_SIZE); } + if (retval > 0) { + int ret; + unsigned long long status; + + status = map->m_flags & EXT4_MAP_UNWRITTEN ? + EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; + if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) && + ext4_find_delalloc_range(inode, map->m_lblk, + map->m_lblk + map->m_len - 1)) + status |= EXTENT_STATUS_DELAYED; + ret = ext4_es_insert_extent(inode, map->m_lblk, + map->m_len, map->m_pblk, status); + if (ret < 0) + retval = ret; + } if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) up_read((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { - int ret; - if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) { - /* delayed alloc may be allocated by fallocate and - * coverted to initialized by directIO. - * we need to handle delayed extent here. - */ - down_write((&EXT4_I(inode)->i_data_sem)); - goto delayed_mapped; - } - ret = check_block_validity(inode, map); + int ret = check_block_validity(inode, map); if (ret != 0) return ret; } @@ -608,18 +614,23 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE)) ext4_da_update_reserve_space(inode, retval, 1); } - if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) { + if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) ext4_clear_inode_state(inode, EXT4_STATE_DELALLOC_RESERVED); - if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { - int ret; -delayed_mapped: - /* delayed allocation blocks has been allocated */ - ret = ext4_es_remove_extent(inode, map->m_lblk, - map->m_len); - if (ret < 0) - retval = ret; - } + if (retval > 0) { + int ret; + unsigned long long status; + + status = map->m_flags & EXT4_MAP_UNWRITTEN ? + EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; + if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) && + ext4_find_delalloc_range(inode, map->m_lblk, + map->m_lblk + map->m_len - 1)) + status |= EXTENT_STATUS_DELAYED; + ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, + map->m_pblk, status); + if (ret < 0) + retval = ret; } up_write((&EXT4_I(inode)->i_data_sem)); @@ -1765,6 +1776,7 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, retval = ext4_ind_map_blocks(NULL, inode, map, 0); if (retval == 0) { + int ret; /* * XXX: __block_prepare_write() unmaps passed block, * is it OK? @@ -1772,16 +1784,20 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, /* If the block was allocated from previously allocated cluster, * then we dont need to reserve it again. */ if (!(map->m_flags & EXT4_MAP_FROM_CLUSTER)) { - retval = ext4_da_reserve_space(inode, iblock); - if (retval) + ret = ext4_da_reserve_space(inode, iblock); + if (ret) { /* not enough space to reserve */ + retval = ret; goto out_unlock; + } } - retval = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, - ~0, EXTENT_STATUS_DELAYED); - if (retval) + ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, + ~0, EXTENT_STATUS_DELAYED); + if (ret) { + retval = ret; goto out_unlock; + } /* Clear EXT4_MAP_FROM_CLUSTER flag since its purpose is served * and it should not appear on the bh->b_state. @@ -1791,6 +1807,16 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, map_bh(bh, inode->i_sb, invalid_block); set_buffer_new(bh); set_buffer_delay(bh); + } else if (retval > 0) { + int ret; + unsigned long long status; + + status = map->m_flags & EXT4_MAP_UNWRITTEN ? + EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; + ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, + map->m_pblk, status); + if (ret != 0) + retval = ret; } out_unlock: -- cgit From d100eef2440fea13e4f09e88b1c8bcbca64beb9f Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:29:59 -0500 Subject: ext4: lookup block mapping in extent status tree After tracking all extent status, we already have a extent cache in memory. Every time we want to lookup a block mapping, we can first try to lookup it in extent status tree to avoid a potential disk I/O. A new function called ext4_es_lookup_extent is defined to finish this work. When we try to lookup a block mapping, we always call ext4_map_blocks and/or ext4_da_map_blocks. So in these functions we first try to lookup a block mapping in extent status tree. A new flag EXT4_GET_BLOCKS_NO_PUT_HOLE is used in ext4_da_map_blocks in order not to put a hole into extent status tree because this hole will be converted to delayed extent in the tree immediately. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Jan kara --- fs/ext4/ext4.h | 2 ++ fs/ext4/extents.c | 9 ++++++- fs/ext4/extents_status.c | 60 +++++++++++++++++++++++++++++++++++++++++++ fs/ext4/extents_status.h | 2 ++ fs/ext4/inode.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++-- 5 files changed, 136 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 5c31d6ac9500..329e7fba47d6 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -579,6 +579,8 @@ enum { #define EXT4_GET_BLOCKS_KEEP_SIZE 0x0080 /* Do not take i_data_sem locking in ext4_map_blocks */ #define EXT4_GET_BLOCKS_NO_LOCK 0x0100 + /* Do not put hole in extent cache */ +#define EXT4_GET_BLOCKS_NO_PUT_HOLE 0x0200 /* * Flags used by ext4_free_blocks diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index be0b1b3eed97..b9d7a2363736 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -2167,6 +2167,9 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path, block, le32_to_cpu(ex->ee_block), ext4_ext_get_actual_len(ex)); + if (!ext4_find_delalloc_range(inode, lblock, lblock + len - 1)) + ext4_es_insert_extent(inode, lblock, len, ~0, + EXTENT_STATUS_HOLE); } else if (block >= le32_to_cpu(ex->ee_block) + ext4_ext_get_actual_len(ex)) { ext4_lblk_t next; @@ -2180,6 +2183,9 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path, block); BUG_ON(next == lblock); len = next - lblock; + if (!ext4_find_delalloc_range(inode, lblock, lblock + len - 1)) + ext4_es_insert_extent(inode, lblock, len, ~0, + EXTENT_STATUS_HOLE); } else { lblock = len = 0; BUG(); @@ -4018,7 +4024,8 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, * put just found gap into cache to speed up * subsequent requests */ - ext4_ext_put_gap_in_cache(inode, path, map->m_lblk); + if ((flags & EXT4_GET_BLOCKS_NO_PUT_HOLE) == 0) + ext4_ext_put_gap_in_cache(inode, path, map->m_lblk); goto out2; } diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 76f4351ea821..eeb893122d8d 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -461,6 +461,66 @@ error: return err; } +/* + * ext4_es_lookup_extent() looks up an extent in extent status tree. + * + * ext4_es_lookup_extent is called by ext4_map_blocks/ext4_da_map_blocks. + * + * Return: 1 on found, 0 on not + */ +int ext4_es_lookup_extent(struct inode *inode, ext4_lblk_t lblk, + struct extent_status *es) +{ + struct ext4_es_tree *tree; + struct extent_status *es1 = NULL; + struct rb_node *node; + int found = 0; + + trace_ext4_es_lookup_extent_enter(inode, lblk); + es_debug("lookup extent in block %u\n", lblk); + + tree = &EXT4_I(inode)->i_es_tree; + read_lock(&EXT4_I(inode)->i_es_lock); + + /* find extent in cache firstly */ + es->es_lblk = es->es_len = es->es_pblk = 0; + if (tree->cache_es) { + es1 = tree->cache_es; + if (in_range(lblk, es1->es_lblk, es1->es_len)) { + es_debug("%u cached by [%u/%u)\n", + lblk, es1->es_lblk, es1->es_len); + found = 1; + goto out; + } + } + + node = tree->root.rb_node; + while (node) { + es1 = rb_entry(node, struct extent_status, rb_node); + if (lblk < es1->es_lblk) + node = node->rb_left; + else if (lblk > ext4_es_end(es1)) + node = node->rb_right; + else { + found = 1; + break; + } + } + +out: + if (found) { + BUG_ON(!es1); + es->es_lblk = es1->es_lblk; + es->es_len = es1->es_len; + es->es_pblk = es1->es_pblk; + } + + read_unlock(&EXT4_I(inode)->i_es_lock); + + trace_ext4_es_lookup_extent_exit(inode, es, found); + return found; +} + static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, ext4_lblk_t end) { diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index 3f69d097c6e7..8ffc90c784fa 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -53,6 +53,8 @@ extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len); extern void ext4_es_find_delayed_extent(struct inode *inode, ext4_lblk_t lblk, struct extent_status *es); +extern int ext4_es_lookup_extent(struct inode *inode, ext4_lblk_t lblk, + struct extent_status *es); static inline int ext4_es_is_written(struct extent_status *es) { diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 576b586b61aa..95a0c62c5683 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -507,12 +507,33 @@ static pgoff_t ext4_num_dirty_pages(struct inode *inode, pgoff_t idx, int ext4_map_blocks(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, int flags) { + struct extent_status es; int retval; map->m_flags = 0; ext_debug("ext4_map_blocks(): inode %lu, flag %d, max_blocks %u," "logical block %lu\n", inode->i_ino, flags, map->m_len, (unsigned long) map->m_lblk); + + /* Lookup extent status tree firstly */ + if (ext4_es_lookup_extent(inode, map->m_lblk, &es)) { + if (ext4_es_is_written(&es) || ext4_es_is_unwritten(&es)) { + map->m_pblk = ext4_es_pblock(&es) + + map->m_lblk - es.es_lblk; + map->m_flags |= ext4_es_is_written(&es) ? + EXT4_MAP_MAPPED : EXT4_MAP_UNWRITTEN; + retval = es.es_len - (map->m_lblk - es.es_lblk); + if (retval > map->m_len) + retval = map->m_len; + map->m_len = retval; + } else if (ext4_es_is_delayed(&es) || ext4_es_is_hole(&es)) { + retval = 0; + } else { + BUG_ON(1); + } + goto found; + } + /* * Try to see if we can get the block without requesting a new * file system block. @@ -544,6 +565,7 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) up_read((&EXT4_I(inode)->i_data_sem)); +found: if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { int ret = check_block_validity(inode, map); if (ret != 0) @@ -1743,6 +1765,7 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, struct ext4_map_blocks *map, struct buffer_head *bh) { + struct extent_status es; int retval; sector_t invalid_block = ~((sector_t) 0xffff); @@ -1753,6 +1776,42 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, ext_debug("ext4_da_map_blocks(): inode %lu, max_blocks %u," "logical block %lu\n", inode->i_ino, map->m_len, (unsigned long) map->m_lblk); + + /* Lookup extent status tree firstly */ + if (ext4_es_lookup_extent(inode, iblock, &es)) { + + if (ext4_es_is_hole(&es)) { + retval = 0; + down_read((&EXT4_I(inode)->i_data_sem)); + goto add_delayed; + } + + /* + * Delayed extent could be allocated by fallocate. + * So we need to check it. + */ + if (ext4_es_is_delayed(&es) && !ext4_es_is_unwritten(&es)) { + map_bh(bh, inode->i_sb, invalid_block); + set_buffer_new(bh); + set_buffer_delay(bh); + return 0; + } + + map->m_pblk = ext4_es_pblock(&es) + iblock - es.es_lblk; + retval = es.es_len - (iblock - es.es_lblk); + if (retval > map->m_len) + retval = map->m_len; + map->m_len = retval; + if (ext4_es_is_written(&es)) + map->m_flags |= EXT4_MAP_MAPPED; + else if (ext4_es_is_unwritten(&es)) + map->m_flags |= EXT4_MAP_UNWRITTEN; + else + BUG_ON(1); + + return retval; + } + /* * Try to see if we can get the block without requesting a new * file system block. @@ -1771,10 +1830,13 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, map->m_flags |= EXT4_MAP_FROM_CLUSTER; retval = 0; } else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) - retval = ext4_ext_map_blocks(NULL, inode, map, 0); + retval = ext4_ext_map_blocks(NULL, inode, map, + EXT4_GET_BLOCKS_NO_PUT_HOLE); else - retval = ext4_ind_map_blocks(NULL, inode, map, 0); + retval = ext4_ind_map_blocks(NULL, inode, map, + EXT4_GET_BLOCKS_NO_PUT_HOLE); +add_delayed: if (retval == 0) { int ret; /* -- cgit From 69eb33dc24dc44d1128aa5e82d0976c42b440c1a Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:31:07 -0500 Subject: ext4: remove single extent cache Single extent cache could be removed because we have extent status tree as a extent cache, and it would be better. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Jan kara --- fs/ext4/ext4.h | 12 ---- fs/ext4/ext4_extents.h | 6 -- fs/ext4/extents.c | 179 +++++++++++-------------------------------------- fs/ext4/move_extent.c | 3 - fs/ext4/super.c | 1 - 5 files changed, 38 insertions(+), 163 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 329e7fba47d6..0c565c941f7a 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -809,17 +809,6 @@ do { \ #endif /* defined(__KERNEL__) || defined(__linux__) */ -/* - * storage for cached extent - * If ec_len == 0, then the cache is invalid. - * If ec_start == 0, then the cache represents a gap (null mapping) - */ -struct ext4_ext_cache { - ext4_fsblk_t ec_start; - ext4_lblk_t ec_block; - __u32 ec_len; /* must be 32bit to return holes */ -}; - #include "extents_status.h" /* @@ -886,7 +875,6 @@ struct ext4_inode_info { struct inode vfs_inode; struct jbd2_inode *jinode; - struct ext4_ext_cache i_cached_extent; /* * File creation time. Its function is same as that of * struct timespec i_{a,c,m}time in the generic inode. diff --git a/fs/ext4/ext4_extents.h b/fs/ext4/ext4_extents.h index 487fda12bc00..8643ff5bbeb7 100644 --- a/fs/ext4/ext4_extents.h +++ b/fs/ext4/ext4_extents.h @@ -193,12 +193,6 @@ static inline unsigned short ext_depth(struct inode *inode) return le16_to_cpu(ext_inode_hdr(inode)->eh_depth); } -static inline void -ext4_ext_invalidate_cache(struct inode *inode) -{ - EXT4_I(inode)->i_cached_extent.ec_len = 0; -} - static inline void ext4_ext_mark_uninitialized(struct ext4_extent *ext) { /* We can not have an uninitialized extent of zero length! */ diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index b9d7a2363736..372b2cbee07e 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -112,7 +112,7 @@ static int ext4_split_extent_at(handle_t *handle, int flags); static int ext4_find_delayed_extent(struct inode *inode, - struct ext4_ext_cache *newex); + struct extent_status *newes); static int ext4_ext_truncate_extend_restart(handle_t *handle, struct inode *inode, @@ -714,7 +714,6 @@ int ext4_ext_tree_init(handle_t *handle, struct inode *inode) eh->eh_magic = EXT4_EXT_MAGIC; eh->eh_max = cpu_to_le16(ext4_ext_space_root(inode, 0)); ext4_mark_inode_dirty(handle, inode); - ext4_ext_invalidate_cache(inode); return 0; } @@ -1963,7 +1962,6 @@ cleanup: ext4_ext_drop_refs(npath); kfree(npath); } - ext4_ext_invalidate_cache(inode); return err; } @@ -1972,8 +1970,8 @@ static int ext4_fill_fiemap_extents(struct inode *inode, struct fiemap_extent_info *fieinfo) { struct ext4_ext_path *path = NULL; - struct ext4_ext_cache newex; struct ext4_extent *ex; + struct extent_status es; ext4_lblk_t next, next_del, start = 0, end = 0; ext4_lblk_t last = block + num; int exists, depth = 0, err = 0; @@ -2047,31 +2045,31 @@ static int ext4_fill_fiemap_extents(struct inode *inode, BUG_ON(end <= start); if (!exists) { - newex.ec_block = start; - newex.ec_len = end - start; - newex.ec_start = 0; + es.es_lblk = start; + es.es_len = end - start; + es.es_pblk = 0; } else { - newex.ec_block = le32_to_cpu(ex->ee_block); - newex.ec_len = ext4_ext_get_actual_len(ex); - newex.ec_start = ext4_ext_pblock(ex); + es.es_lblk = le32_to_cpu(ex->ee_block); + es.es_len = ext4_ext_get_actual_len(ex); + es.es_pblk = ext4_ext_pblock(ex); if (ext4_ext_is_uninitialized(ex)) flags |= FIEMAP_EXTENT_UNWRITTEN; } /* - * Find delayed extent and update newex accordingly. We call - * it even in !exists case to find out whether newex is the + * Find delayed extent and update es accordingly. We call + * it even in !exists case to find out whether es is the * last existing extent or not. */ - next_del = ext4_find_delayed_extent(inode, &newex); + next_del = ext4_find_delayed_extent(inode, &es); if (!exists && next_del) { exists = 1; flags |= FIEMAP_EXTENT_DELALLOC; } up_read(&EXT4_I(inode)->i_data_sem); - if (unlikely(newex.ec_len == 0)) { - EXT4_ERROR_INODE(inode, "newex.ec_len == 0"); + if (unlikely(es.es_len == 0)) { + EXT4_ERROR_INODE(inode, "es.es_len == 0"); err = -EIO; break; } @@ -2102,9 +2100,9 @@ static int ext4_fill_fiemap_extents(struct inode *inode, if (exists) { err = fiemap_fill_next_extent(fieinfo, - (__u64)newex.ec_block << blksize_bits, - (__u64)newex.ec_start << blksize_bits, - (__u64)newex.ec_len << blksize_bits, + (__u64)es.es_lblk << blksize_bits, + (__u64)es.es_pblk << blksize_bits, + (__u64)es.es_len << blksize_bits, flags); if (err < 0) break; @@ -2114,7 +2112,7 @@ static int ext4_fill_fiemap_extents(struct inode *inode, } } - block = newex.ec_block + newex.ec_len; + block = es.es_lblk + es.es_len; } if (path) { @@ -2125,21 +2123,6 @@ static int ext4_fill_fiemap_extents(struct inode *inode, return err; } -static void -ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block, - __u32 len, ext4_fsblk_t start) -{ - struct ext4_ext_cache *cex; - BUG_ON(len == 0); - spin_lock(&EXT4_I(inode)->i_block_reservation_lock); - trace_ext4_ext_put_in_cache(inode, block, len, start); - cex = &EXT4_I(inode)->i_cached_extent; - cex->ec_block = block; - cex->ec_len = len; - cex->ec_start = start; - spin_unlock(&EXT4_I(inode)->i_block_reservation_lock); -} - /* * ext4_ext_put_gap_in_cache: * calculate boundaries of the gap that the requested block fits into @@ -2156,9 +2139,10 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path, ex = path[depth].p_ext; if (ex == NULL) { - /* there is no extent yet, so gap is [0;-] */ - lblock = 0; - len = EXT_MAX_BLOCKS; + /* + * there is no extent yet, so gap is [0;-] and we + * don't cache it + */ ext_debug("cache gap(whole file):"); } else if (block < le32_to_cpu(ex->ee_block)) { lblock = block; @@ -2192,52 +2176,6 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path, } ext_debug(" -> %u:%lu\n", lblock, len); - ext4_ext_put_in_cache(inode, lblock, len, 0); -} - -/* - * ext4_ext_in_cache() - * Checks to see if the given block is in the cache. - * If it is, the cached extent is stored in the given - * cache extent pointer. - * - * @inode: The files inode - * @block: The block to look for in the cache - * @ex: Pointer where the cached extent will be stored - * if it contains block - * - * Return 0 if cache is invalid; 1 if the cache is valid - */ -static int -ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block, - struct ext4_extent *ex) -{ - struct ext4_ext_cache *cex; - int ret = 0; - - /* - * We borrow i_block_reservation_lock to protect i_cached_extent - */ - spin_lock(&EXT4_I(inode)->i_block_reservation_lock); - cex = &EXT4_I(inode)->i_cached_extent; - - /* has cache valid data? */ - if (cex->ec_len == 0) - goto errout; - - if (in_range(block, cex->ec_block, cex->ec_len)) { - ex->ee_block = cpu_to_le32(cex->ec_block); - ext4_ext_store_pblock(ex, cex->ec_start); - ex->ee_len = cpu_to_le16(cex->ec_len); - ext_debug("%u cached by %u:%u:%llu\n", - block, - cex->ec_block, cex->ec_len, cex->ec_start); - ret = 1; - } -errout: - trace_ext4_ext_in_cache(inode, block, ret); - spin_unlock(&EXT4_I(inode)->i_block_reservation_lock); - return ret; } /* @@ -2677,8 +2615,6 @@ static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start, return PTR_ERR(handle); again: - ext4_ext_invalidate_cache(inode); - trace_ext4_ext_remove_space(inode, start, depth); /* @@ -3920,35 +3856,6 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, map->m_lblk, map->m_len, inode->i_ino); trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); - /* check in cache */ - if (ext4_ext_in_cache(inode, map->m_lblk, &newex)) { - if (!newex.ee_start_lo && !newex.ee_start_hi) { - if ((sbi->s_cluster_ratio > 1) && - ext4_find_delalloc_cluster(inode, map->m_lblk)) - map->m_flags |= EXT4_MAP_FROM_CLUSTER; - - if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { - /* - * block isn't allocated yet and - * user doesn't want to allocate it - */ - goto out2; - } - /* we should allocate requested block */ - } else { - /* block is already allocated */ - if (sbi->s_cluster_ratio > 1) - map->m_flags |= EXT4_MAP_FROM_CLUSTER; - newblock = map->m_lblk - - le32_to_cpu(newex.ee_block) - + ext4_ext_pblock(&newex); - /* number of remaining blocks in the extent */ - allocated = ext4_ext_get_actual_len(&newex) - - (map->m_lblk - le32_to_cpu(newex.ee_block)); - goto out; - } - } - /* find extent for this block */ path = ext4_ext_find_extent(inode, map->m_lblk, NULL); if (IS_ERR(path)) { @@ -3995,15 +3902,9 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk, ee_block, ee_len, newblock); - /* - * Do not put uninitialized extent - * in the cache - */ - if (!ext4_ext_is_uninitialized(ex)) { - ext4_ext_put_in_cache(inode, ee_block, - ee_len, ee_start); + if (!ext4_ext_is_uninitialized(ex)) goto out; - } + allocated = ext4_ext_handle_uninitialized_extents( handle, inode, map, path, flags, allocated, newblock); @@ -4265,10 +4166,9 @@ got_allocated_blocks: * Cache the extent and update transaction to commit on fdatasync only * when it is _not_ an uninitialized extent. */ - if ((flags & EXT4_GET_BLOCKS_UNINIT_EXT) == 0) { - ext4_ext_put_in_cache(inode, map->m_lblk, allocated, newblock); + if ((flags & EXT4_GET_BLOCKS_UNINIT_EXT) == 0) ext4_update_inode_fsync_trans(handle, inode, 1); - } else + else ext4_update_inode_fsync_trans(handle, inode, 0); out: if (allocated > map->m_len) @@ -4327,7 +4227,6 @@ void ext4_ext_truncate(struct inode *inode) goto out_stop; down_write(&EXT4_I(inode)->i_data_sem); - ext4_ext_invalidate_cache(inode); ext4_discard_preallocations(inode); @@ -4576,42 +4475,42 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset, } /* - * If newex is not existing extent (newex->ec_start equals zero) find - * delayed extent at start of newex and update newex accordingly and + * If newes is not existing extent (newes->ec_pblk equals zero) find + * delayed extent at start of newes and update newes accordingly and * return start of the next delayed extent. * - * If newex is existing extent (newex->ec_start is not equal zero) + * If newes is existing extent (newes->ec_pblk is not equal zero) * return start of next delayed extent or EXT_MAX_BLOCKS if no delayed - * extent found. Leave newex unmodified. + * extent found. Leave newes unmodified. */ static int ext4_find_delayed_extent(struct inode *inode, - struct ext4_ext_cache *newex) + struct extent_status *newes) { struct extent_status es; ext4_lblk_t block, next_del; - ext4_es_find_delayed_extent(inode, newex->ec_block, &es); + ext4_es_find_delayed_extent(inode, newes->es_lblk, &es); - if (newex->ec_start == 0) { + if (newes->es_pblk == 0) { /* - * No extent in extent-tree contains block @newex->ec_start, + * No extent in extent-tree contains block @newes->es_pblk, * then the block may stay in 1)a hole or 2)delayed-extent. */ if (es.es_len == 0) /* A hole found. */ return 0; - if (es.es_lblk > newex->ec_block) { + if (es.es_lblk > newes->es_lblk) { /* A hole found. */ - newex->ec_len = min(es.es_lblk - newex->ec_block, - newex->ec_len); + newes->es_len = min(es.es_lblk - newes->es_lblk, + newes->es_len); return 0; } - newex->ec_len = es.es_lblk + es.es_len - newex->ec_block; + newes->es_len = es.es_lblk + es.es_len - newes->es_lblk; } - block = newex->ec_block + newex->ec_len; + block = newes->es_lblk + newes->es_len; ext4_es_find_delayed_extent(inode, block, &es); if (es.es_len == 0) next_del = EXT_MAX_BLOCKS; @@ -4815,14 +4714,12 @@ int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length) goto out; down_write(&EXT4_I(inode)->i_data_sem); - ext4_ext_invalidate_cache(inode); ext4_discard_preallocations(inode); err = ext4_es_remove_extent(inode, first_block, stop_block - first_block); err = ext4_ext_remove_space(inode, first_block, stop_block - 1); - ext4_ext_invalidate_cache(inode); ext4_discard_preallocations(inode); if (IS_SYNC(inode)) diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index 0d6734394eac..d78c33eed7e5 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -764,9 +764,6 @@ out: kfree(donor_path); } - ext4_ext_invalidate_cache(orig_inode); - ext4_ext_invalidate_cache(donor_inode); - return replaced_count; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index ef6ac59e4961..d80bfe5ac11c 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -836,7 +836,6 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) return NULL; ei->vfs_inode.i_version = 1; - memset(&ei->i_cached_extent, 0, sizeof(struct ext4_ext_cache)); INIT_LIST_HEAD(&ei->i_prealloc_list); spin_lock_init(&ei->i_prealloc_lock); ext4_es_init_tree(&ei->i_es_tree); -- cgit From bdedbb7b8d5b960e1ff0d116f5d4935febe73183 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:32:02 -0500 Subject: ext4: adjust some functions for reclaiming extents from extent status tree This commit changes some interfaces in extent status tree because we need to use inode to count the cached objects in a extent status tree. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Jan kara --- fs/ext4/extents_status.c | 50 +++++++++++++++++++++++------------------------- 1 file changed, 24 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index eeb893122d8d..cce152c3c8dc 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -142,9 +142,8 @@ static struct kmem_cache *ext4_es_cachep; -static int __es_insert_extent(struct ext4_es_tree *tree, - struct extent_status *newes); -static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, +static int __es_insert_extent(struct inode *inode, struct extent_status *newes); +static int __es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t end); int __init ext4_init_es(void) @@ -285,7 +284,8 @@ out: } static struct extent_status * -ext4_es_alloc_extent(ext4_lblk_t lblk, ext4_lblk_t len, ext4_fsblk_t pblk) +ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len, + ext4_fsblk_t pblk) { struct extent_status *es; es = kmem_cache_alloc(ext4_es_cachep, GFP_ATOMIC); @@ -297,7 +297,7 @@ ext4_es_alloc_extent(ext4_lblk_t lblk, ext4_lblk_t len, ext4_fsblk_t pblk) return es; } -static void ext4_es_free_extent(struct extent_status *es) +static void ext4_es_free_extent(struct inode *inode, struct extent_status *es) { kmem_cache_free(ext4_es_cachep, es); } @@ -326,8 +326,9 @@ static int ext4_es_can_be_merged(struct extent_status *es1, } static struct extent_status * -ext4_es_try_to_merge_left(struct ext4_es_tree *tree, struct extent_status *es) +ext4_es_try_to_merge_left(struct inode *inode, struct extent_status *es) { + struct ext4_es_tree *tree = &EXT4_I(inode)->i_es_tree; struct extent_status *es1; struct rb_node *node; @@ -339,7 +340,7 @@ ext4_es_try_to_merge_left(struct ext4_es_tree *tree, struct extent_status *es) if (ext4_es_can_be_merged(es1, es)) { es1->es_len += es->es_len; rb_erase(&es->rb_node, &tree->root); - ext4_es_free_extent(es); + ext4_es_free_extent(inode, es); es = es1; } @@ -347,8 +348,9 @@ ext4_es_try_to_merge_left(struct ext4_es_tree *tree, struct extent_status *es) } static struct extent_status * -ext4_es_try_to_merge_right(struct ext4_es_tree *tree, struct extent_status *es) +ext4_es_try_to_merge_right(struct inode *inode, struct extent_status *es) { + struct ext4_es_tree *tree = &EXT4_I(inode)->i_es_tree; struct extent_status *es1; struct rb_node *node; @@ -360,15 +362,15 @@ ext4_es_try_to_merge_right(struct ext4_es_tree *tree, struct extent_status *es) if (ext4_es_can_be_merged(es, es1)) { es->es_len += es1->es_len; rb_erase(node, &tree->root); - ext4_es_free_extent(es1); + ext4_es_free_extent(inode, es1); } return es; } -static int __es_insert_extent(struct ext4_es_tree *tree, - struct extent_status *newes) +static int __es_insert_extent(struct inode *inode, struct extent_status *newes) { + struct ext4_es_tree *tree = &EXT4_I(inode)->i_es_tree; struct rb_node **p = &tree->root.rb_node; struct rb_node *parent = NULL; struct extent_status *es; @@ -389,14 +391,14 @@ static int __es_insert_extent(struct ext4_es_tree *tree, ext4_es_is_unwritten(es)) ext4_es_store_pblock(es, newes->es_pblk); - es = ext4_es_try_to_merge_left(tree, es); + es = ext4_es_try_to_merge_left(inode, es); goto out; } p = &(*p)->rb_left; } else if (newes->es_lblk > ext4_es_end(es)) { if (ext4_es_can_be_merged(es, newes)) { es->es_len += newes->es_len; - es = ext4_es_try_to_merge_right(tree, es); + es = ext4_es_try_to_merge_right(inode, es); goto out; } p = &(*p)->rb_right; @@ -406,7 +408,7 @@ static int __es_insert_extent(struct ext4_es_tree *tree, } } - es = ext4_es_alloc_extent(newes->es_lblk, newes->es_len, + es = ext4_es_alloc_extent(inode, newes->es_lblk, newes->es_len, newes->es_pblk); if (!es) return -ENOMEM; @@ -430,7 +432,6 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len, ext4_fsblk_t pblk, unsigned long long status) { - struct ext4_es_tree *tree; struct extent_status newes; ext4_lblk_t end = lblk + len - 1; int err = 0; @@ -447,11 +448,10 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, trace_ext4_es_insert_extent(inode, &newes); write_lock(&EXT4_I(inode)->i_es_lock); - tree = &EXT4_I(inode)->i_es_tree; - err = __es_remove_extent(tree, lblk, end); + err = __es_remove_extent(inode, lblk, end); if (err != 0) goto error; - err = __es_insert_extent(tree, &newes); + err = __es_insert_extent(inode, &newes); error: write_unlock(&EXT4_I(inode)->i_es_lock); @@ -521,9 +521,10 @@ out: return found; } -static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, - ext4_lblk_t end) +static int __es_remove_extent(struct inode *inode, ext4_lblk_t lblk, + ext4_lblk_t end) { + struct ext4_es_tree *tree = &EXT4_I(inode)->i_es_tree; struct rb_node *node; struct extent_status *es; struct extent_status orig_es; @@ -561,7 +562,7 @@ static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, ext4_es_store_pblock(&newes, block); } ext4_es_store_status(&newes, ext4_es_status(&orig_es)); - err = __es_insert_extent(tree, &newes); + err = __es_insert_extent(inode, &newes); if (err) { es->es_lblk = orig_es.es_lblk; es->es_len = orig_es.es_len; @@ -590,7 +591,7 @@ static int __es_remove_extent(struct ext4_es_tree *tree, ext4_lblk_t lblk, while (es && ext4_es_end(es) <= end) { node = rb_next(&es->rb_node); rb_erase(&es->rb_node, &tree->root); - ext4_es_free_extent(es); + ext4_es_free_extent(inode, es); if (!node) { es = NULL; break; @@ -622,7 +623,6 @@ out: int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len) { - struct ext4_es_tree *tree; ext4_lblk_t end; int err = 0; @@ -633,10 +633,8 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, end = lblk + len - 1; BUG_ON(end < lblk); - tree = &EXT4_I(inode)->i_es_tree; - write_lock(&EXT4_I(inode)->i_es_lock); - err = __es_remove_extent(tree, lblk, end); + err = __es_remove_extent(inode, lblk, end); write_unlock(&EXT4_I(inode)->i_es_lock); ext4_es_print_tree(inode); return err; -- cgit From 74cd15cd02708c7188581f279f33a98b2ae8d322 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Mon, 18 Feb 2013 00:32:55 -0500 Subject: ext4: reclaim extents from extent status tree Although extent status is loaded on-demand, we also need to reclaim extent from the tree when we are under a heavy memory pressure because in some cases fragmented extent tree causes status tree costs too much memory. Here we maintain a lru list in super_block. When the extent status of an inode is accessed and changed, this inode will be move to the tail of the list. The inode will be dropped from this list when it is cleared. In the inode, a counter is added to count the number of cached objects in extent status tree. Here only written/unwritten/hole extent is counted because delayed extent doesn't be reclaimed due to fiemap, bigalloc and seek_data/hole need it. The counter will be increased as a new extent is allocated, and it will be decreased as a extent is freed. In this commit we use normal shrinker framework to reclaim memory from the status tree. ext4_es_reclaim_extents_count() traverses the lru list to count the number of reclaimable extents. ext4_es_shrink() tries to reclaim written/unwritten/hole extents from extent status tree. The inode that has been shrunk is moved to the tail of lru list. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Jan kara --- fs/ext4/ext4.h | 7 +++ fs/ext4/extents_status.c | 156 +++++++++++++++++++++++++++++++++++++++++++++++ fs/ext4/extents_status.h | 5 ++ fs/ext4/super.c | 7 +++ 4 files changed, 175 insertions(+) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 0c565c941f7a..6e16c1867959 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -888,6 +888,8 @@ struct ext4_inode_info { /* extents status tree */ struct ext4_es_tree i_es_tree; rwlock_t i_es_lock; + struct list_head i_es_lru; + unsigned int i_es_lru_nr; /* protected by i_es_lock */ /* ialloc */ ext4_group_t i_last_alloc_group; @@ -1303,6 +1305,11 @@ struct ext4_sb_info { /* Precomputed FS UUID checksum for seeding other checksums */ __u32 s_csum_seed; + + /* Reclaim extents from extent status tree */ + struct shrinker s_es_shrinker; + struct list_head s_es_lru; + spinlock_t s_es_lru_lock ____cacheline_aligned_in_smp; }; static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb) diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index cce152c3c8dc..9f1380e05474 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -145,6 +145,9 @@ static struct kmem_cache *ext4_es_cachep; static int __es_insert_extent(struct inode *inode, struct extent_status *newes); static int __es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t end); +static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei, + int nr_to_scan); +static int ext4_es_reclaim_extents_count(struct super_block *sb); int __init ext4_init_es(void) { @@ -280,6 +283,7 @@ out: read_unlock(&EXT4_I(inode)->i_es_lock); + ext4_es_lru_add(inode); trace_ext4_es_find_delayed_extent_exit(inode, es); } @@ -294,11 +298,24 @@ ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len, es->es_lblk = lblk; es->es_len = len; es->es_pblk = pblk; + + /* + * We don't count delayed extent because we never try to reclaim them + */ + if (!ext4_es_is_delayed(es)) + EXT4_I(inode)->i_es_lru_nr++; + return es; } static void ext4_es_free_extent(struct inode *inode, struct extent_status *es) { + /* Decrease the lru counter when this es is not delayed */ + if (!ext4_es_is_delayed(es)) { + BUG_ON(EXT4_I(inode)->i_es_lru_nr == 0); + EXT4_I(inode)->i_es_lru_nr--; + } + kmem_cache_free(ext4_es_cachep, es); } @@ -456,6 +473,7 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, error: write_unlock(&EXT4_I(inode)->i_es_lock); + ext4_es_lru_add(inode); ext4_es_print_tree(inode); return err; @@ -517,6 +535,7 @@ out: read_unlock(&EXT4_I(inode)->i_es_lock); + ext4_es_lru_add(inode); trace_ext4_es_lookup_extent_exit(inode, es, found); return found; } @@ -639,3 +658,140 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_es_print_tree(inode); return err; } + +static int ext4_es_shrink(struct shrinker *shrink, struct shrink_control *sc) +{ + struct ext4_sb_info *sbi = container_of(shrink, + struct ext4_sb_info, s_es_shrinker); + struct ext4_inode_info *ei; + struct list_head *cur, *tmp, scanned; + int nr_to_scan = sc->nr_to_scan; + int ret, nr_shrunk = 0; + + trace_ext4_es_shrink_enter(sbi->s_sb, nr_to_scan); + + if (!nr_to_scan) + return ext4_es_reclaim_extents_count(sbi->s_sb); + + INIT_LIST_HEAD(&scanned); + + spin_lock(&sbi->s_es_lru_lock); + list_for_each_safe(cur, tmp, &sbi->s_es_lru) { + list_move_tail(cur, &scanned); + + ei = list_entry(cur, struct ext4_inode_info, i_es_lru); + + read_lock(&ei->i_es_lock); + if (ei->i_es_lru_nr == 0) { + read_unlock(&ei->i_es_lock); + continue; + } + read_unlock(&ei->i_es_lock); + + write_lock(&ei->i_es_lock); + ret = __es_try_to_reclaim_extents(ei, nr_to_scan); + write_unlock(&ei->i_es_lock); + + nr_shrunk += ret; + nr_to_scan -= ret; + if (nr_to_scan == 0) + break; + } + list_splice_tail(&scanned, &sbi->s_es_lru); + spin_unlock(&sbi->s_es_lru_lock); + trace_ext4_es_shrink_exit(sbi->s_sb, nr_shrunk); + + return ext4_es_reclaim_extents_count(sbi->s_sb); +} + +void ext4_es_register_shrinker(struct super_block *sb) +{ + struct ext4_sb_info *sbi; + + sbi = EXT4_SB(sb); + INIT_LIST_HEAD(&sbi->s_es_lru); + spin_lock_init(&sbi->s_es_lru_lock); + sbi->s_es_shrinker.shrink = ext4_es_shrink; + sbi->s_es_shrinker.seeks = DEFAULT_SEEKS; + register_shrinker(&sbi->s_es_shrinker); +} + +void ext4_es_unregister_shrinker(struct super_block *sb) +{ + unregister_shrinker(&EXT4_SB(sb)->s_es_shrinker); +} + +void ext4_es_lru_add(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + + spin_lock(&sbi->s_es_lru_lock); + if (list_empty(&ei->i_es_lru)) + list_add_tail(&ei->i_es_lru, &sbi->s_es_lru); + else + list_move_tail(&ei->i_es_lru, &sbi->s_es_lru); + spin_unlock(&sbi->s_es_lru_lock); +} + +void ext4_es_lru_del(struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + + spin_lock(&sbi->s_es_lru_lock); + if (!list_empty(&ei->i_es_lru)) + list_del_init(&ei->i_es_lru); + spin_unlock(&sbi->s_es_lru_lock); +} + +static int ext4_es_reclaim_extents_count(struct super_block *sb) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_inode_info *ei; + struct list_head *cur; + int nr_cached = 0; + + spin_lock(&sbi->s_es_lru_lock); + list_for_each(cur, &sbi->s_es_lru) { + ei = list_entry(cur, struct ext4_inode_info, i_es_lru); + read_lock(&ei->i_es_lock); + nr_cached += ei->i_es_lru_nr; + read_unlock(&ei->i_es_lock); + } + spin_unlock(&sbi->s_es_lru_lock); + trace_ext4_es_reclaim_extents_count(sb, nr_cached); + return nr_cached; +} + +static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei, + int nr_to_scan) +{ + struct inode *inode = &ei->vfs_inode; + struct ext4_es_tree *tree = &ei->i_es_tree; + struct rb_node *node; + struct extent_status *es; + int nr_shrunk = 0; + + if (ei->i_es_lru_nr == 0) + return 0; + + node = rb_first(&tree->root); + while (node != NULL) { + es = rb_entry(node, struct extent_status, rb_node); + node = rb_next(&es->rb_node); + /* + * We can't reclaim delayed extent from status tree because + * fiemap, bigallic, and seek_data/hole need to use it. + */ + if (!ext4_es_is_delayed(es)) { + rb_erase(&es->rb_node, &tree->root); + ext4_es_free_extent(inode, es); + nr_shrunk++; + if (--nr_to_scan == 0) + break; + } + } + tree->cache_es = NULL; + return nr_shrunk; +} diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index 8ffc90c784fa..cf83e77b16cb 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -106,4 +106,9 @@ static inline void ext4_es_store_status(struct extent_status *es, es->es_pblk = block; } +extern void ext4_es_register_shrinker(struct super_block *sb); +extern void ext4_es_unregister_shrinker(struct super_block *sb); +extern void ext4_es_lru_add(struct inode *inode); +extern void ext4_es_lru_del(struct inode *inode); + #endif /* _EXT4_EXTENTS_STATUS_H */ diff --git a/fs/ext4/super.c b/fs/ext4/super.c index d80bfe5ac11c..373d46cd5d3f 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -755,6 +755,7 @@ static void ext4_put_super(struct super_block *sb) ext4_abort(sb, "Couldn't clean up the journal"); } + ext4_es_unregister_shrinker(sb); del_timer(&sbi->s_err_report); ext4_release_system_zone(sb); ext4_mb_release(sb); @@ -840,6 +841,8 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) spin_lock_init(&ei->i_prealloc_lock); ext4_es_init_tree(&ei->i_es_tree); rwlock_init(&ei->i_es_lock); + INIT_LIST_HEAD(&ei->i_es_lru); + ei->i_es_lru_nr = 0; ei->i_reserved_data_blocks = 0; ei->i_reserved_meta_blocks = 0; ei->i_allocated_meta_blocks = 0; @@ -928,6 +931,7 @@ void ext4_clear_inode(struct inode *inode) dquot_drop(inode); ext4_discard_preallocations(inode); ext4_es_remove_extent(inode, 0, EXT_MAX_BLOCKS); + ext4_es_lru_del(inode); if (EXT4_I(inode)->jinode) { jbd2_journal_release_jbd_inode(EXT4_JOURNAL(inode), EXT4_I(inode)->jinode); @@ -3693,6 +3697,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sbi->s_max_writeback_mb_bump = 128; sbi->s_extent_max_zeroout_kb = 32; + /* Register extent status tree shrinker */ + ext4_es_register_shrinker(sb); + /* * set up enough so that it can read an inode */ -- cgit From 1231b3a1eb5740192aeebf5344dd6d6da000febf Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Mon, 18 Feb 2013 12:12:07 -0500 Subject: ext4: fix xattr block allocation/release with bigalloc Currently when new xattr block is created or released we we would call dquot_free_block() or dquot_alloc_block() respectively, among the else decrementing or incrementing the number of blocks assigned to the inode by one block. This however does not work for bigalloc file system because we always allocate/free the whole cluster so we have to count with that in dquot_free_block() and dquot_alloc_block() as well. Use the clusters-to-blocks conversion EXT4_C2B() when passing number of blocks to the dquot_alloc/free functions to fix the problem. The problem has been revealed by xfstests #117 (and possibly others). Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" Reviewed-by: Eric Sandeen Cc: stable@vger.kernel.org --- fs/ext4/xattr.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index cc31da027596..3a120b277240 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -549,7 +549,7 @@ ext4_xattr_release_block(handle_t *handle, struct inode *inode, error = ext4_handle_dirty_xattr_block(handle, inode, bh); if (IS_SYNC(inode)) ext4_handle_sync(handle); - dquot_free_block(inode, 1); + dquot_free_block(inode, EXT4_C2B(EXT4_SB(inode->i_sb), 1)); ea_bdebug(bh, "refcount now=%d; releasing", le32_to_cpu(BHDR(bh)->h_refcount)); } @@ -832,7 +832,8 @@ inserted: else { /* The old block is released after updating the inode. */ - error = dquot_alloc_block(inode, 1); + error = dquot_alloc_block(inode, + EXT4_C2B(EXT4_SB(sb), 1)); if (error) goto cleanup; error = ext4_journal_get_write_access(handle, @@ -929,7 +930,7 @@ cleanup: return error; cleanup_dquot: - dquot_free_block(inode, 1); + dquot_free_block(inode, EXT4_C2B(EXT4_SB(sb), 1)); goto cleanup; bad_block: -- cgit From 87f979d390f9ecfa3d0038a9f9a002a62f8a1895 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Fri, 15 Feb 2013 11:42:29 -0600 Subject: ceph: kill ceph_osdc_writepages() "nofail" parameter There is only one caller of ceph_osdc_writepages(), and it always passes the value true as its "nofail" argument. Get rid of that argument and replace its use in ceph_osdc_writepages() with the constant value true. This and a number of cleanup patches that follow resolve: http://tracker.ceph.com/issues/4126 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/addr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 064d1a68d2c1..c7e401c96fc9 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -493,7 +493,7 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) page_off, len, ci->i_truncate_seq, ci->i_truncate_size, &inode->i_mtime, - &page, 1, 0, 0, true); + &page, 1, 0, 0); if (err < 0) { dout("writepage setting page/mapping error %d %p\n", err, page); SetPageError(page); -- cgit From fbf8685fb155e12a9f4d4b966c7b3442ed557687 Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Fri, 15 Feb 2013 11:42:29 -0600 Subject: ceph: kill ceph_osdc_writepages() "dosync" parameter There is only one caller of ceph_osdc_writepages(), and it always passes 0 as its "dosync" argument. Get rid of that argument and replace its use in ceph_osdc_writepages() with 0. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/addr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index c7e401c96fc9..bef552819312 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -493,7 +493,7 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) page_off, len, ci->i_truncate_seq, ci->i_truncate_size, &inode->i_mtime, - &page, 1, 0, 0); + &page, 1, 0); if (err < 0) { dout("writepage setting page/mapping error %d %p\n", err, page); SetPageError(page); -- cgit From 2480882611e3ab844563dd3d0a822227604ab8fe Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Fri, 15 Feb 2013 11:42:29 -0600 Subject: ceph: kill ceph_osdc_writepages() "flags" parameter There is only one caller of ceph_osdc_writepages(), and it always passes 0 as its "flags" argument. Get rid of that argument and replace its use in ceph_osdc_writepages() with 0. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/addr.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index bef552819312..8d3240d6f289 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -492,8 +492,7 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) &ci->i_layout, snapc, page_off, len, ci->i_truncate_seq, ci->i_truncate_size, - &inode->i_mtime, - &page, 1, 0); + &inode->i_mtime, &page, 1); if (err < 0) { dout("writepage setting page/mapping error %d %p\n", err, page); SetPageError(page); -- cgit From a3bea47e8bdd51d921e5b2045720d60140612c7c Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Fri, 15 Feb 2013 11:42:29 -0600 Subject: ceph: kill ceph_osdc_new_request() "num_reply" parameter The "num_reply" parameter to ceph_osdc_new_request() is never used inside that function, so get rid of it. Note that ceph_sync_write() passes 2 for that argument, while all other callers pass 1. It doesn't matter, but perhaps someone should verify this doesn't indicate a problem. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/addr.c | 4 ++-- fs/ceph/file.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 8d3240d6f289..fc613715af46 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -315,7 +315,7 @@ static int start_read(struct inode *inode, struct list_head *page_list, int max) CEPH_OSD_OP_READ, CEPH_OSD_FLAG_READ, NULL, 0, ci->i_truncate_seq, ci->i_truncate_size, - NULL, false, 1, 0); + NULL, false, 0); if (IS_ERR(req)) return PTR_ERR(req); @@ -837,7 +837,7 @@ get_more_pages: snapc, do_sync, ci->i_truncate_seq, ci->i_truncate_size, - &inode->i_mtime, true, 1, 0); + &inode->i_mtime, true, 0); if (IS_ERR(req)) { rc = PTR_ERR(req); diff --git a/fs/ceph/file.c b/fs/ceph/file.c index a1e5b81e8118..9c4325e654ca 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -541,7 +541,7 @@ more: ci->i_snap_realm->cached_context, do_sync, ci->i_truncate_seq, ci->i_truncate_size, - &mtime, false, 2, page_align); + &mtime, false, page_align); if (IS_ERR(req)) return PTR_ERR(req); -- cgit From 0eb40bf65e2fcc1c23ed7c9a10cd0890ee59e68f Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Fri, 15 Feb 2013 11:42:30 -0600 Subject: libceph: update ceph_mds_state_name() and ceph_mds_op_name() Update ceph_mds_state_name() and ceph_mds_op_name() to include the newly-added definitions in "ceph_fs.h", and to match its counterpart in the user space code. Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/strings.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/ceph/strings.c b/fs/ceph/strings.c index cd5097d7c804..89fa4a940a0f 100644 --- a/fs/ceph/strings.c +++ b/fs/ceph/strings.c @@ -15,6 +15,7 @@ const char *ceph_mds_state_name(int s) case CEPH_MDS_STATE_BOOT: return "up:boot"; case CEPH_MDS_STATE_STANDBY: return "up:standby"; case CEPH_MDS_STATE_STANDBY_REPLAY: return "up:standby-replay"; + case CEPH_MDS_STATE_REPLAYONCE: return "up:oneshot-replay"; case CEPH_MDS_STATE_CREATING: return "up:creating"; case CEPH_MDS_STATE_STARTING: return "up:starting"; /* up and in */ @@ -50,10 +51,13 @@ const char *ceph_mds_op_name(int op) case CEPH_MDS_OP_LOOKUP: return "lookup"; case CEPH_MDS_OP_LOOKUPHASH: return "lookuphash"; case CEPH_MDS_OP_LOOKUPPARENT: return "lookupparent"; + case CEPH_MDS_OP_LOOKUPINO: return "lookupino"; case CEPH_MDS_OP_GETATTR: return "getattr"; case CEPH_MDS_OP_SETXATTR: return "setxattr"; case CEPH_MDS_OP_SETATTR: return "setattr"; case CEPH_MDS_OP_RMXATTR: return "rmxattr"; + case CEPH_MDS_OP_SETLAYOUT: return "setlayou"; + case CEPH_MDS_OP_SETDIRLAYOUT: return "setdirlayout"; case CEPH_MDS_OP_READDIR: return "readdir"; case CEPH_MDS_OP_MKNOD: return "mknod"; case CEPH_MDS_OP_LINK: return "link"; -- cgit From b4278c961aca320839964e23cfc7906ff61af0c2 Mon Sep 17 00:00:00 2001 From: Gao feng Date: Mon, 18 Feb 2013 01:34:55 +0000 Subject: net: proc: remove proc_net_fops_create proc_net_fops_create has been replaced by proc_create, we can remove it now. Signed-off-by: Gao feng Signed-off-by: David S. Miller --- fs/proc/proc_net.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'fs') diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c index fe72cd073dea..30f7c678424b 100644 --- a/fs/proc/proc_net.c +++ b/fs/proc/proc_net.c @@ -177,14 +177,6 @@ const struct file_operations proc_net_operations = { .readdir = proc_tgid_net_readdir, }; - -struct proc_dir_entry *proc_net_fops_create(struct net *net, - const char *name, umode_t mode, const struct file_operations *fops) -{ - return proc_create(name, mode, net->proc_net, fops); -} -EXPORT_SYMBOL_GPL(proc_net_fops_create); - void proc_net_remove(struct net *net, const char *name) { remove_proc_entry(name, net->proc_net); -- cgit From c2399059a389ba686fa7f45d8913a708914752c4 Mon Sep 17 00:00:00 2001 From: Gao feng Date: Mon, 18 Feb 2013 01:34:57 +0000 Subject: net: proc: remove proc_net_remove proc_net_remove has been replaced by remove_proc_entry. we can remove it now. Signed-off-by: Gao feng Signed-off-by: David S. Miller --- fs/proc/proc_net.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c index 30f7c678424b..3131a03d7d37 100644 --- a/fs/proc/proc_net.c +++ b/fs/proc/proc_net.c @@ -177,12 +177,6 @@ const struct file_operations proc_net_operations = { .readdir = proc_tgid_net_readdir, }; -void proc_net_remove(struct net *net, const char *name) -{ - remove_proc_entry(name, net->proc_net); -} -EXPORT_SYMBOL_GPL(proc_net_remove); - static __net_init int proc_net_ns_init(struct net *net) { struct proc_dir_entry *netd, *net_statd; -- cgit From eece09ec213e93333010bf4c6bb9175b32229c54 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sun, 17 Jul 2011 21:25:03 +0200 Subject: locking: Various static lock initializer fixes The static lock initializers want to be fed the proper name of the lock and not some random string. In mainline random strings are obfuscating the readability of debug output, but for RT they prevent the spinlock substitution. Fix it up. Signed-off-by: Thomas Gleixner --- fs/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/file.c b/fs/file.c index 2b3570b7caeb..3906d9577a18 100644 --- a/fs/file.c +++ b/fs/file.c @@ -516,7 +516,7 @@ struct files_struct init_files = { .close_on_exec = init_files.close_on_exec_init, .open_fds = init_files.open_fds_init, }, - .file_lock = __SPIN_LOCK_UNLOCKED(init_task.file_lock), + .file_lock = __SPIN_LOCK_UNLOCKED(init_files.file_lock), }; /* -- cgit From 666b3d803a511fbc9bc5e5ea8ce66010cf03ea13 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 19 Feb 2013 12:04:42 -0500 Subject: NLM: Ensure that we resend all pending blocking locks after a reclaim Currently, nlmclnt_lock will break out of the for(;;) loop when the reclaimer wakes up the blocking lock thread by setting nlm_lck_denied_grace_period. This causes the lock request to fail with an ENOLCK error. The intention was always to ensure that we resend the lock request after the grace period has expired. Reported-by: Wangyuan Zhang Signed-off-by: Trond Myklebust Cc: stable@vger.kernel.org --- fs/lockd/clntproc.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 54f9e6ce0430..52e5120bb159 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -550,6 +550,9 @@ again: status = nlmclnt_block(block, req, NLMCLNT_POLL_TIMEOUT); if (status < 0) break; + /* Resend the blocking lock request after a server reboot */ + if (resp->status == nlm_lck_denied_grace_period) + continue; if (resp->status != nlm_lck_blocked) break; } -- cgit From 9e0eb85d5861d512759caf1301670b36d4c221ed Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Wed, 6 Feb 2013 13:11:38 -0600 Subject: ceph: remove a few bogus declarations There are three ceph page vector functions declared in "fs/ceph/super.h" that don't belong there. They're probably left over from some long-ago code reorganization. They're properly declared in "include/linux/ceph/libceph.h" so just delete the ones in "super.h". This and the next few commits resolve: http://tracker.ceph.com/issues/4053 Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/super.h | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 66ebe720e40d..9861cce10a49 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -798,13 +798,7 @@ extern int ceph_mmap(struct file *file, struct vm_area_struct *vma); /* file.c */ extern const struct file_operations ceph_file_fops; extern const struct address_space_operations ceph_aops; -extern int ceph_copy_to_page_vector(struct page **pages, - const char *data, - loff_t off, size_t len); -extern int ceph_copy_from_page_vector(struct page **pages, - char *data, - loff_t off, size_t len); -extern struct page **ceph_alloc_page_vector(int num_pages, gfp_t flags); + extern int ceph_open(struct inode *inode, struct file *file); extern int ceph_atomic_open(struct inode *dir, struct dentry *dentry, struct file *file, unsigned flags, umode_t mode, -- cgit From 78a6184a3ff9041280ee56273c01e5679a831b39 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 21 Nov 2012 02:21:28 +0000 Subject: Btrfs: use slabs for delayed reference allocation The delayed reference allocation is in the fast path of the IO, so use slabs to improve the speed of the allocation. And besides that, it can do check for leaked objects when the module is removed. Signed-off-by: Miao Xie --- fs/btrfs/delayed-ref.c | 74 ++++++++++++++++++++++++++++++++++++++++++-------- fs/btrfs/delayed-ref.h | 37 ++++++++++++++++++++++++- fs/btrfs/disk-io.c | 2 +- fs/btrfs/extent-tree.c | 14 +++++----- fs/btrfs/super.c | 9 +++++- 5 files changed, 115 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index ae9411773397..455894f1ca3b 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -23,6 +23,10 @@ #include "delayed-ref.h" #include "transaction.h" +struct kmem_cache *btrfs_delayed_ref_head_cachep; +struct kmem_cache *btrfs_delayed_tree_ref_cachep; +struct kmem_cache *btrfs_delayed_data_ref_cachep; +struct kmem_cache *btrfs_delayed_extent_op_cachep; /* * delayed back reference update tracking. For subvolume trees * we queue up extent allocations and backref maintenance for @@ -511,7 +515,7 @@ update_existing_head_ref(struct btrfs_delayed_ref_node *existing, ref->extent_op->flags_to_set; existing_ref->extent_op->update_flags = 1; } - kfree(ref->extent_op); + btrfs_free_delayed_extent_op(ref->extent_op); } } /* @@ -592,7 +596,7 @@ static noinline void add_delayed_ref_head(struct btrfs_fs_info *fs_info, * we've updated the existing ref, free the newly * allocated ref */ - kfree(head_ref); + kmem_cache_free(btrfs_delayed_ref_head_cachep, head_ref); } else { delayed_refs->num_heads++; delayed_refs->num_heads_ready++; @@ -653,7 +657,7 @@ static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info, * we've updated the existing ref, free the newly * allocated ref */ - kfree(full_ref); + kmem_cache_free(btrfs_delayed_tree_ref_cachep, full_ref); } else { delayed_refs->num_entries++; trans->delayed_ref_updates++; @@ -714,7 +718,7 @@ static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info, * we've updated the existing ref, free the newly * allocated ref */ - kfree(full_ref); + kmem_cache_free(btrfs_delayed_data_ref_cachep, full_ref); } else { delayed_refs->num_entries++; trans->delayed_ref_updates++; @@ -738,13 +742,13 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_root *delayed_refs; BUG_ON(extent_op && extent_op->is_data); - ref = kmalloc(sizeof(*ref), GFP_NOFS); + ref = kmem_cache_alloc(btrfs_delayed_tree_ref_cachep, GFP_NOFS); if (!ref) return -ENOMEM; - head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS); + head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS); if (!head_ref) { - kfree(ref); + kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref); return -ENOMEM; } @@ -786,13 +790,13 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_root *delayed_refs; BUG_ON(extent_op && !extent_op->is_data); - ref = kmalloc(sizeof(*ref), GFP_NOFS); + ref = kmem_cache_alloc(btrfs_delayed_data_ref_cachep, GFP_NOFS); if (!ref) return -ENOMEM; - head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS); + head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS); if (!head_ref) { - kfree(ref); + kmem_cache_free(btrfs_delayed_data_ref_cachep, ref); return -ENOMEM; } @@ -826,7 +830,7 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_head *head_ref; struct btrfs_delayed_ref_root *delayed_refs; - head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS); + head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS); if (!head_ref) return -ENOMEM; @@ -860,3 +864,51 @@ btrfs_find_delayed_ref_head(struct btrfs_trans_handle *trans, u64 bytenr) return btrfs_delayed_node_to_head(ref); return NULL; } + +void btrfs_delayed_ref_exit(void) +{ + if (btrfs_delayed_ref_head_cachep) + kmem_cache_destroy(btrfs_delayed_ref_head_cachep); + if (btrfs_delayed_tree_ref_cachep) + kmem_cache_destroy(btrfs_delayed_tree_ref_cachep); + if (btrfs_delayed_data_ref_cachep) + kmem_cache_destroy(btrfs_delayed_data_ref_cachep); + if (btrfs_delayed_extent_op_cachep) + kmem_cache_destroy(btrfs_delayed_extent_op_cachep); +} + +int btrfs_delayed_ref_init(void) +{ + btrfs_delayed_ref_head_cachep = kmem_cache_create( + "btrfs_delayed_ref_head", + sizeof(struct btrfs_delayed_ref_head), 0, + SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL); + if (!btrfs_delayed_ref_head_cachep) + goto fail; + + btrfs_delayed_tree_ref_cachep = kmem_cache_create( + "btrfs_delayed_tree_ref", + sizeof(struct btrfs_delayed_tree_ref), 0, + SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL); + if (!btrfs_delayed_tree_ref_cachep) + goto fail; + + btrfs_delayed_data_ref_cachep = kmem_cache_create( + "btrfs_delayed_data_ref", + sizeof(struct btrfs_delayed_data_ref), 0, + SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL); + if (!btrfs_delayed_data_ref_cachep) + goto fail; + + btrfs_delayed_extent_op_cachep = kmem_cache_create( + "btrfs_delayed_extent_op", + sizeof(struct btrfs_delayed_extent_op), 0, + SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL); + if (!btrfs_delayed_extent_op_cachep) + goto fail; + + return 0; +fail: + btrfs_delayed_ref_exit(); + return -ENOMEM; +} diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index c9d703693df0..fe50392cdf76 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -141,12 +141,47 @@ struct btrfs_delayed_ref_root { u64 run_delayed_start; }; +extern struct kmem_cache *btrfs_delayed_ref_head_cachep; +extern struct kmem_cache *btrfs_delayed_tree_ref_cachep; +extern struct kmem_cache *btrfs_delayed_data_ref_cachep; +extern struct kmem_cache *btrfs_delayed_extent_op_cachep; + +int btrfs_delayed_ref_init(void); +void btrfs_delayed_ref_exit(void); + +static inline struct btrfs_delayed_extent_op * +btrfs_alloc_delayed_extent_op(void) +{ + return kmem_cache_alloc(btrfs_delayed_extent_op_cachep, GFP_NOFS); +} + +static inline void +btrfs_free_delayed_extent_op(struct btrfs_delayed_extent_op *op) +{ + if (op) + kmem_cache_free(btrfs_delayed_extent_op_cachep, op); +} + static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref) { WARN_ON(atomic_read(&ref->refs) == 0); if (atomic_dec_and_test(&ref->refs)) { WARN_ON(ref->in_tree); - kfree(ref); + switch (ref->type) { + case BTRFS_TREE_BLOCK_REF_KEY: + case BTRFS_SHARED_BLOCK_REF_KEY: + kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref); + break; + case BTRFS_EXTENT_DATA_REF_KEY: + case BTRFS_SHARED_DATA_REF_KEY: + kmem_cache_free(btrfs_delayed_data_ref_cachep, ref); + break; + case 0: + kmem_cache_free(btrfs_delayed_ref_head_cachep, ref); + break; + default: + BUG(); + } } } diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 65f03670a952..4438aac4947f 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3614,7 +3614,7 @@ int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, continue; } - kfree(head->extent_op); + btrfs_free_delayed_extent_op(head->extent_op); delayed_refs->num_heads--; if (list_empty(&head->cluster)) delayed_refs->num_heads_ready--; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 61da9d0bb805..9bd87f0d73d0 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2285,7 +2285,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, ref = &locked_ref->node; if (extent_op && must_insert_reserved) { - kfree(extent_op); + btrfs_free_delayed_extent_op(extent_op); extent_op = NULL; } @@ -2294,7 +2294,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, ret = run_delayed_extent_op(trans, root, ref, extent_op); - kfree(extent_op); + btrfs_free_delayed_extent_op(extent_op); if (ret) { list_del_init(&locked_ref->cluster); @@ -2338,7 +2338,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, must_insert_reserved); btrfs_put_delayed_ref(ref); - kfree(extent_op); + btrfs_free_delayed_extent_op(extent_op); count++; if (ret) { @@ -2586,7 +2586,7 @@ int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans, struct btrfs_delayed_extent_op *extent_op; int ret; - extent_op = kmalloc(sizeof(*extent_op), GFP_NOFS); + extent_op = btrfs_alloc_delayed_extent_op(); if (!extent_op) return -ENOMEM; @@ -2598,7 +2598,7 @@ int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans, ret = btrfs_add_delayed_extent_op(root->fs_info, trans, bytenr, num_bytes, extent_op); if (ret) - kfree(extent_op); + btrfs_free_delayed_extent_op(extent_op); return ret; } @@ -5330,7 +5330,7 @@ static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans, if (head->extent_op) { if (!head->must_insert_reserved) goto out; - kfree(head->extent_op); + btrfs_free_delayed_extent_op(head->extent_op); head->extent_op = NULL; } @@ -6400,7 +6400,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, if (root_objectid != BTRFS_TREE_LOG_OBJECTID) { struct btrfs_delayed_extent_op *extent_op; - extent_op = kmalloc(sizeof(*extent_op), GFP_NOFS); + extent_op = btrfs_alloc_delayed_extent_op(); BUG_ON(!extent_op); /* -ENOMEM */ if (key) memcpy(&extent_op->key, key, sizeof(extent_op->key)); diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index d8982e9601d3..67b373bf3ff9 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1684,10 +1684,14 @@ static int __init init_btrfs_fs(void) if (err) goto free_delayed_inode; - err = btrfs_interface_init(); + err = btrfs_delayed_ref_init(); if (err) goto free_auto_defrag; + err = btrfs_interface_init(); + if (err) + goto free_delayed_ref; + err = register_filesystem(&btrfs_fs_type); if (err) goto unregister_ioctl; @@ -1699,6 +1703,8 @@ static int __init init_btrfs_fs(void) unregister_ioctl: btrfs_interface_exit(); +free_delayed_ref: + btrfs_delayed_ref_exit(); free_auto_defrag: btrfs_auto_defrag_exit(); free_delayed_inode: @@ -1720,6 +1726,7 @@ free_compress: static void __exit exit_btrfs_fs(void) { btrfs_destroy_cachep(); + btrfs_delayed_ref_exit(); btrfs_auto_defrag_exit(); btrfs_delayed_inode_exit(); ordered_data_exit(); -- cgit From 7b5a1c5310a50abc96c9ca07039688027d0a4282 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 15 Nov 2012 08:14:11 +0000 Subject: Btrfs: cleanup unnecessary clear when freeing a transaction or a trans handle We clear the transaction object and the trans handle when they are about to be freed, it is unnecessary, cleanup it. Signed-off-by: Miao Xie --- fs/btrfs/transaction.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index fc03aa60b684..a1455f1e4676 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -40,7 +40,6 @@ void put_transaction(struct btrfs_transaction *transaction) if (atomic_dec_and_test(&transaction->use_count)) { BUG_ON(!list_empty(&transaction->list)); WARN_ON(transaction->delayed_refs.root.rb_node); - memset(transaction, 0, sizeof(*transaction)); kmem_cache_free(btrfs_transaction_cachep, transaction); } } @@ -650,7 +649,6 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, } assert_qgroups_uptodate(trans); - memset(trans, 0, sizeof(*trans)); kmem_cache_free(btrfs_trans_handle_cachep, trans); return err; } -- cgit From 7892b5afe4a1a00af25107e27357db30434ab876 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 15 Nov 2012 08:14:47 +0000 Subject: Btrfs: use common work instead of delayed work Since we do not want to delay the async transaction commit, we should use common work, not delayed work. Signed-off-by: Miao Xie --- fs/btrfs/transaction.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index a1455f1e4676..86bb105b3982 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1305,13 +1305,13 @@ static void wait_current_trans_commit_start_and_unblock(struct btrfs_root *root, struct btrfs_async_commit { struct btrfs_trans_handle *newtrans; struct btrfs_root *root; - struct delayed_work work; + struct work_struct work; }; static void do_async_commit(struct work_struct *work) { struct btrfs_async_commit *ac = - container_of(work, struct btrfs_async_commit, work.work); + container_of(work, struct btrfs_async_commit, work); /* * We've got freeze protection passed with the transaction. @@ -1339,7 +1339,7 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans, if (!ac) return -ENOMEM; - INIT_DELAYED_WORK(&ac->work, do_async_commit); + INIT_WORK(&ac->work, do_async_commit); ac->root = root; ac->newtrans = btrfs_join_transaction(root); if (IS_ERR(ac->newtrans)) { @@ -1363,7 +1363,7 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans, &root->fs_info->sb->s_writers.lock_map[SB_FREEZE_FS-1], 1, _THIS_IP_); - schedule_delayed_work(&ac->work, 0); + schedule_work(&ac->work); /* wait for transaction to start and unblock */ if (wait_for_unblock) -- cgit From 4ea41ce07d8aed9e64c546abf341aea44e782e64 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 19 Dec 2012 06:59:03 +0000 Subject: Btrfs: cleanup similar code in delayed inode The delayed item commit code in several functions is similar, so cleanup it. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/delayed-inode.c | 83 +++++++++++++++++++++--------------------------- 1 file changed, 37 insertions(+), 46 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index 34836036f01b..4e204bb8ba13 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -1110,6 +1110,25 @@ static int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans, return 0; } +static inline int +__btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, + struct btrfs_path *path, + struct btrfs_delayed_node *node) +{ + int ret; + + ret = btrfs_insert_delayed_items(trans, path, node->root, node); + if (ret) + return ret; + + ret = btrfs_delete_delayed_items(trans, path, node->root, node); + if (ret) + return ret; + + ret = btrfs_update_delayed_inode(trans, node->root, path, node); + return ret; +} + /* * Called when committing the transaction. * Returns 0 on success. @@ -1119,7 +1138,6 @@ static int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans, static int __btrfs_run_delayed_items(struct btrfs_trans_handle *trans, struct btrfs_root *root, int nr) { - struct btrfs_root *curr_root = root; struct btrfs_delayed_root *delayed_root; struct btrfs_delayed_node *curr_node, *prev_node; struct btrfs_path *path; @@ -1142,15 +1160,8 @@ static int __btrfs_run_delayed_items(struct btrfs_trans_handle *trans, curr_node = btrfs_first_delayed_node(delayed_root); while (curr_node && (!count || (count && nr--))) { - curr_root = curr_node->root; - ret = btrfs_insert_delayed_items(trans, path, curr_root, - curr_node); - if (!ret) - ret = btrfs_delete_delayed_items(trans, path, - curr_root, curr_node); - if (!ret) - ret = btrfs_update_delayed_inode(trans, curr_root, - path, curr_node); + ret = __btrfs_commit_inode_delayed_items(trans, path, + curr_node); if (ret) { btrfs_release_delayed_node(curr_node); curr_node = NULL; @@ -1183,36 +1194,12 @@ int btrfs_run_delayed_items_nr(struct btrfs_trans_handle *trans, return __btrfs_run_delayed_items(trans, root, nr); } -static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, - struct btrfs_delayed_node *node) -{ - struct btrfs_path *path; - struct btrfs_block_rsv *block_rsv; - int ret; - - path = btrfs_alloc_path(); - if (!path) - return -ENOMEM; - path->leave_spinning = 1; - - block_rsv = trans->block_rsv; - trans->block_rsv = &node->root->fs_info->delayed_block_rsv; - - ret = btrfs_insert_delayed_items(trans, path, node->root, node); - if (!ret) - ret = btrfs_delete_delayed_items(trans, path, node->root, node); - if (!ret) - ret = btrfs_update_delayed_inode(trans, node->root, path, node); - btrfs_free_path(path); - - trans->block_rsv = block_rsv; - return ret; -} - int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, struct inode *inode) { struct btrfs_delayed_node *delayed_node = btrfs_get_delayed_node(inode); + struct btrfs_path *path; + struct btrfs_block_rsv *block_rsv; int ret; if (!delayed_node) @@ -1226,8 +1213,20 @@ int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, } mutex_unlock(&delayed_node->mutex); - ret = __btrfs_commit_inode_delayed_items(trans, delayed_node); + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + path->leave_spinning = 1; + + block_rsv = trans->block_rsv; + trans->block_rsv = &delayed_node->root->fs_info->delayed_block_rsv; + + ret = __btrfs_commit_inode_delayed_items(trans, path, delayed_node); + btrfs_release_delayed_node(delayed_node); + btrfs_free_path(path); + trans->block_rsv = block_rsv; + return ret; } @@ -1258,7 +1257,6 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work) struct btrfs_root *root; struct btrfs_block_rsv *block_rsv; int need_requeue = 0; - int ret; async_node = container_of(work, struct btrfs_async_delayed_node, work); @@ -1277,14 +1275,7 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work) block_rsv = trans->block_rsv; trans->block_rsv = &root->fs_info->delayed_block_rsv; - ret = btrfs_insert_delayed_items(trans, path, root, delayed_node); - if (!ret) - ret = btrfs_delete_delayed_items(trans, path, root, - delayed_node); - - if (!ret) - btrfs_update_delayed_inode(trans, root, path, delayed_node); - + __btrfs_commit_inode_delayed_items(trans, path, delayed_node); /* * Maybe new delayed items have been inserted, so we need requeue * the work. Besides that, we must dequeue the empty delayed nodes -- cgit From 0e8c36a9fd8169a8b96c2ddc8446894bcd07b6b1 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 19 Dec 2012 06:59:51 +0000 Subject: Btrfs: fix lots of orphan inodes when the space is not enough We're running into having 50-100 orphans left over with xfstests 83 because of ENOSPC when trying to start the transaction for the inode update. But in fact, it makes no sense in updating the inode for the new size while we're deleting the stupid thing. This patch fixes this problem. Reported-by: Josef Bacik Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/delayed-inode.c | 90 +++++++++++++++++++++++++++++++++++++++++------- fs/btrfs/delayed-inode.h | 1 + fs/btrfs/inode.c | 11 +++--- 3 files changed, 85 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index 4e204bb8ba13..c7e75061b894 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -1065,32 +1065,25 @@ static void btrfs_release_delayed_inode(struct btrfs_delayed_node *delayed_node) } } -static int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_path *path, - struct btrfs_delayed_node *node) +static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct btrfs_delayed_node *node) { struct btrfs_key key; struct btrfs_inode_item *inode_item; struct extent_buffer *leaf; int ret; - mutex_lock(&node->mutex); - if (!node->inode_dirty) { - mutex_unlock(&node->mutex); - return 0; - } - key.objectid = node->inode_id; btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY); key.offset = 0; + ret = btrfs_lookup_inode(trans, root, path, &key, 1); if (ret > 0) { btrfs_release_path(path); - mutex_unlock(&node->mutex); return -ENOENT; } else if (ret < 0) { - mutex_unlock(&node->mutex); return ret; } @@ -1105,11 +1098,28 @@ static int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans, btrfs_delayed_inode_release_metadata(root, node); btrfs_release_delayed_inode(node); - mutex_unlock(&node->mutex); return 0; } +static inline int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct btrfs_delayed_node *node) +{ + int ret; + + mutex_lock(&node->mutex); + if (!node->inode_dirty) { + mutex_unlock(&node->mutex); + return 0; + } + + ret = __btrfs_update_delayed_inode(trans, root, path, node); + mutex_unlock(&node->mutex); + return ret; +} + static inline int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, struct btrfs_path *path, @@ -1230,6 +1240,60 @@ int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, return ret; } +int btrfs_commit_inode_delayed_inode(struct inode *inode) +{ + struct btrfs_trans_handle *trans; + struct btrfs_delayed_node *delayed_node = btrfs_get_delayed_node(inode); + struct btrfs_path *path; + struct btrfs_block_rsv *block_rsv; + int ret; + + if (!delayed_node) + return 0; + + mutex_lock(&delayed_node->mutex); + if (!delayed_node->inode_dirty) { + mutex_unlock(&delayed_node->mutex); + btrfs_release_delayed_node(delayed_node); + return 0; + } + mutex_unlock(&delayed_node->mutex); + + trans = btrfs_join_transaction(delayed_node->root); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto out; + } + + path = btrfs_alloc_path(); + if (!path) { + ret = -ENOMEM; + goto trans_out; + } + path->leave_spinning = 1; + + block_rsv = trans->block_rsv; + trans->block_rsv = &delayed_node->root->fs_info->delayed_block_rsv; + + mutex_lock(&delayed_node->mutex); + if (delayed_node->inode_dirty) + ret = __btrfs_update_delayed_inode(trans, delayed_node->root, + path, delayed_node); + else + ret = 0; + mutex_unlock(&delayed_node->mutex); + + btrfs_free_path(path); + trans->block_rsv = block_rsv; +trans_out: + btrfs_end_transaction(trans, delayed_node->root); + btrfs_btree_balance_dirty(delayed_node->root); +out: + btrfs_release_delayed_node(delayed_node); + + return ret; +} + void btrfs_remove_delayed_node(struct inode *inode) { struct btrfs_delayed_node *delayed_node; diff --git a/fs/btrfs/delayed-inode.h b/fs/btrfs/delayed-inode.h index 4f808e1baeed..78b6ad0fc669 100644 --- a/fs/btrfs/delayed-inode.h +++ b/fs/btrfs/delayed-inode.h @@ -117,6 +117,7 @@ int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, /* Used for evicting the inode. */ void btrfs_remove_delayed_node(struct inode *inode); void btrfs_kill_delayed_inode_items(struct inode *inode); +int btrfs_commit_inode_delayed_inode(struct inode *inode); int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index ca7ace7b7b52..1ea0988b82e1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3904,6 +3904,12 @@ void btrfs_evict_inode(struct inode *inode) goto no_delete; } + ret = btrfs_commit_inode_delayed_inode(inode); + if (ret) { + btrfs_orphan_del(NULL, inode); + goto no_delete; + } + rsv = btrfs_alloc_block_rsv(root, BTRFS_BLOCK_RSV_TEMP); if (!rsv) { btrfs_orphan_del(NULL, inode); @@ -3941,7 +3947,7 @@ void btrfs_evict_inode(struct inode *inode) goto no_delete; } - trans = btrfs_start_transaction_lflush(root, 1); + trans = btrfs_join_transaction(root); if (IS_ERR(trans)) { btrfs_orphan_del(NULL, inode); btrfs_free_block_rsv(root, rsv); @@ -3955,9 +3961,6 @@ void btrfs_evict_inode(struct inode *inode) break; trans->block_rsv = &root->fs_info->trans_block_rsv; - ret = btrfs_update_inode(trans, root, inode); - BUG_ON(ret); - btrfs_end_transaction(trans, root); trans = NULL; btrfs_btree_balance_dirty(root); -- cgit From 093486c453a55230ccdad4b48863b872fe68c46e Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 19 Dec 2012 08:10:10 +0000 Subject: Btrfs: make delayed ref lock logic more readable Locking and unlocking delayed ref mutex are in the different functions, and the name of lock functions is not uniform, so the readability is not so good, this patch optimizes the lock logic and makes it more readable. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/delayed-ref.c | 8 ++++++++ fs/btrfs/delayed-ref.h | 6 ++++++ fs/btrfs/extent-tree.c | 42 ++++++++++++++++++++++++------------------ 3 files changed, 38 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index 455894f1ca3b..b7a0641ead77 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -426,6 +426,14 @@ again: return 1; } +void btrfs_release_ref_cluster(struct list_head *cluster) +{ + struct list_head *pos, *q; + + list_for_each_safe(pos, q, cluster) + list_del_init(pos); +} + /* * helper function to update an extent delayed ref in the * rbtree. existing and update must both have the same diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index fe50392cdf76..7939149f8f27 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -211,8 +211,14 @@ struct btrfs_delayed_ref_head * btrfs_find_delayed_ref_head(struct btrfs_trans_handle *trans, u64 bytenr); int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head *head); +static inline void btrfs_delayed_ref_unlock(struct btrfs_delayed_ref_head *head) +{ + mutex_unlock(&head->mutex); +} + int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans, struct list_head *cluster, u64 search_start); +void btrfs_release_ref_cluster(struct list_head *cluster); int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_root *delayed_refs, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 9bd87f0d73d0..b4cb8186035f 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2143,7 +2143,6 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans, node->num_bytes); } } - mutex_unlock(&head->mutex); return ret; } @@ -2258,7 +2257,7 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, * process of being added. Don't run this ref yet. */ list_del_init(&locked_ref->cluster); - mutex_unlock(&locked_ref->mutex); + btrfs_delayed_ref_unlock(locked_ref); locked_ref = NULL; delayed_refs->num_heads_ready++; spin_unlock(&delayed_refs->lock); @@ -2297,25 +2296,22 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, btrfs_free_delayed_extent_op(extent_op); if (ret) { - list_del_init(&locked_ref->cluster); - mutex_unlock(&locked_ref->mutex); - - printk(KERN_DEBUG "btrfs: run_delayed_extent_op returned %d\n", ret); + printk(KERN_DEBUG + "btrfs: run_delayed_extent_op " + "returned %d\n", ret); spin_lock(&delayed_refs->lock); + btrfs_delayed_ref_unlock(locked_ref); return ret; } goto next; } - - list_del_init(&locked_ref->cluster); - locked_ref = NULL; } ref->in_tree = 0; rb_erase(&ref->rb_node, &delayed_refs->root); delayed_refs->num_entries--; - if (locked_ref) { + if (!btrfs_delayed_ref_is_head(ref)) { /* * when we play the delayed ref, also correct the * ref_mod on head @@ -2337,20 +2333,29 @@ static noinline int run_clustered_refs(struct btrfs_trans_handle *trans, ret = run_one_delayed_ref(trans, root, ref, extent_op, must_insert_reserved); - btrfs_put_delayed_ref(ref); btrfs_free_delayed_extent_op(extent_op); - count++; - if (ret) { - if (locked_ref) { - list_del_init(&locked_ref->cluster); - mutex_unlock(&locked_ref->mutex); - } - printk(KERN_DEBUG "btrfs: run_one_delayed_ref returned %d\n", ret); + btrfs_delayed_ref_unlock(locked_ref); + btrfs_put_delayed_ref(ref); + printk(KERN_DEBUG + "btrfs: run_one_delayed_ref returned %d\n", ret); spin_lock(&delayed_refs->lock); return ret; } + /* + * If this node is a head, that means all the refs in this head + * have been dealt with, and we will pick the next head to deal + * with, so we must unlock the head and drop it from the cluster + * list before we release it. + */ + if (btrfs_delayed_ref_is_head(ref)) { + list_del_init(&locked_ref->cluster); + btrfs_delayed_ref_unlock(locked_ref); + locked_ref = NULL; + } + btrfs_put_delayed_ref(ref); + count++; next: cond_resched(); spin_lock(&delayed_refs->lock); @@ -2500,6 +2505,7 @@ again: ret = run_clustered_refs(trans, root, &cluster); if (ret < 0) { + btrfs_release_ref_cluster(&cluster); spin_unlock(&delayed_refs->lock); btrfs_abort_transaction(trans, root, ret); return ret; -- cgit From da633a42170165cbf20a2d3886c7480ccc832ec3 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 20 Dec 2012 11:19:09 +0000 Subject: Btrfs: flush all dirty inodes if writeback can not start We may try to flush some dirty pages when there is no enough space to reserve. But it is possible that this operation fails, in order to get enough space to reserve successfully, we will sync all the delalloc file. This operation is safe, we needn't worry about the case that the filesystem goes from r/w to r/o. because the filesystem should guarantee all the dirty pages have been written into the disk after it becomes readonly, so the sync operation will do nothing if the filesystem is already readonly. Though it may waste lots of time, as a corner case, we needn't care. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 40 +++++++++++++++++++++++++++++++--------- 1 file changed, 31 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index b4cb8186035f..d5e60d25ca51 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3695,12 +3695,15 @@ static int can_overcommit(struct btrfs_root *root, return 0; } -static int writeback_inodes_sb_nr_if_idle_safe(struct super_block *sb, - unsigned long nr_pages, - enum wb_reason reason) +static inline int writeback_inodes_sb_nr_if_idle_safe(struct super_block *sb, + unsigned long nr_pages, + enum wb_reason reason) { - if (!writeback_in_progress(sb->s_bdi) && - down_read_trylock(&sb->s_umount)) { + /* the flusher is dealing with the dirty inodes now. */ + if (writeback_in_progress(sb->s_bdi)) + return 1; + + if (down_read_trylock(&sb->s_umount)) { writeback_inodes_sb_nr(sb, nr_pages, reason); up_read(&sb->s_umount); return 1; @@ -3709,6 +3712,28 @@ static int writeback_inodes_sb_nr_if_idle_safe(struct super_block *sb, return 0; } +void btrfs_writeback_inodes_sb_nr(struct btrfs_root *root, + unsigned long nr_pages) +{ + struct super_block *sb = root->fs_info->sb; + int started; + + /* If we can not start writeback, just sync all the delalloc file. */ + started = writeback_inodes_sb_nr_if_idle_safe(sb, nr_pages, + WB_REASON_FS_FREE_SPACE); + if (!started) { + /* + * We needn't worry the filesystem going from r/w to r/o though + * we don't acquire ->s_umount mutex, because the filesystem + * should guarantee the delalloc inodes list be empty after + * the filesystem is readonly(all dirty pages are written to + * the disk). + */ + btrfs_start_delalloc_inodes(root, 0); + btrfs_wait_ordered_extents(root, 0); + } +} + /* * shrink metadata reservation for delalloc */ @@ -3741,10 +3766,7 @@ static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig, while (delalloc_bytes && loops < 3) { max_reclaim = min(delalloc_bytes, to_reclaim); nr_pages = max_reclaim >> PAGE_CACHE_SHIFT; - writeback_inodes_sb_nr_if_idle_safe(root->fs_info->sb, - nr_pages, - WB_REASON_FS_FREE_SPACE); - + btrfs_writeback_inodes_sb_nr(root, nr_pages); /* * We need to wait for the async pages to actually start before * we do anything. -- cgit From 4eee4fa4f8ab8c2b2623c22be4c3cb91d525aa57 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Fri, 21 Dec 2012 09:17:45 +0000 Subject: Btrfs: use wrapper page_offset Use wrapper page_offset to get byte-offset into filesystem object for page. Signed-off-by: Liu Bo Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 2 +- fs/btrfs/extent_io.c | 24 +++++++++++------------- fs/btrfs/inode.c | 5 ++--- fs/btrfs/relocation.c | 2 +- 4 files changed, 15 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 4438aac4947f..a762f9137610 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -420,7 +420,7 @@ static int btree_read_extent_buffer_pages(struct btrfs_root *root, static int csum_dirty_buffer(struct btrfs_root *root, struct page *page) { struct extent_io_tree *tree; - u64 start = (u64)page->index << PAGE_CACHE_SHIFT; + u64 start = page_offset(page); u64 found_start; struct extent_buffer *eb; diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 1b319df29eee..bda36fef301e 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -1834,7 +1834,7 @@ int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end, */ static void check_page_uptodate(struct extent_io_tree *tree, struct page *page) { - u64 start = (u64)page->index << PAGE_CACHE_SHIFT; + u64 start = page_offset(page); u64 end = start + PAGE_CACHE_SIZE - 1; if (test_range_bit(tree, start, end, EXTENT_UPTODATE, 1, NULL)) SetPageUptodate(page); @@ -1846,7 +1846,7 @@ static void check_page_uptodate(struct extent_io_tree *tree, struct page *page) */ static void check_page_locked(struct extent_io_tree *tree, struct page *page) { - u64 start = (u64)page->index << PAGE_CACHE_SHIFT; + u64 start = page_offset(page); u64 end = start + PAGE_CACHE_SIZE - 1; if (!test_range_bit(tree, start, end, EXTENT_LOCKED, 0, NULL)) unlock_page(page); @@ -1960,7 +1960,7 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 start, return -EIO; } bio->bi_bdev = dev->bdev; - bio_add_page(bio, page, length, start-page_offset(page)); + bio_add_page(bio, page, length, start - page_offset(page)); btrfsic_submit_bio(WRITE_SYNC, bio); wait_for_completion(&compl); @@ -2293,8 +2293,7 @@ static void end_bio_extent_writepage(struct bio *bio, int err) struct page *page = bvec->bv_page; tree = &BTRFS_I(page->mapping->host)->io_tree; - start = ((u64)page->index << PAGE_CACHE_SHIFT) + - bvec->bv_offset; + start = page_offset(page) + bvec->bv_offset; end = start + bvec->bv_len - 1; if (bvec->bv_offset == 0 && bvec->bv_len == PAGE_CACHE_SIZE) @@ -2353,8 +2352,7 @@ static void end_bio_extent_readpage(struct bio *bio, int err) (long int)bio->bi_bdev); tree = &BTRFS_I(page->mapping->host)->io_tree; - start = ((u64)page->index << PAGE_CACHE_SHIFT) + - bvec->bv_offset; + start = page_offset(page) + bvec->bv_offset; end = start + bvec->bv_len - 1; if (bvec->bv_offset == 0 && bvec->bv_len == PAGE_CACHE_SIZE) @@ -2471,7 +2469,7 @@ static int __must_check submit_one_bio(int rw, struct bio *bio, struct extent_io_tree *tree = bio->bi_private; u64 start; - start = ((u64)page->index << PAGE_CACHE_SHIFT) + bvec->bv_offset; + start = page_offset(page) + bvec->bv_offset; bio->bi_private = NULL; @@ -2595,7 +2593,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree, unsigned long *bio_flags) { struct inode *inode = page->mapping->host; - u64 start = (u64)page->index << PAGE_CACHE_SHIFT; + u64 start = page_offset(page); u64 page_end = start + PAGE_CACHE_SIZE - 1; u64 end; u64 cur = start; @@ -2806,7 +2804,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, struct inode *inode = page->mapping->host; struct extent_page_data *epd = data; struct extent_io_tree *tree = epd->tree; - u64 start = (u64)page->index << PAGE_CACHE_SHIFT; + u64 start = page_offset(page); u64 delalloc_start; u64 page_end = start + PAGE_CACHE_SIZE - 1; u64 end; @@ -3674,7 +3672,7 @@ int extent_invalidatepage(struct extent_io_tree *tree, struct page *page, unsigned long offset) { struct extent_state *cached_state = NULL; - u64 start = ((u64)page->index << PAGE_CACHE_SHIFT); + u64 start = page_offset(page); u64 end = start + PAGE_CACHE_SIZE - 1; size_t blocksize = page->mapping->host->i_sb->s_blocksize; @@ -3700,7 +3698,7 @@ int try_release_extent_state(struct extent_map_tree *map, struct extent_io_tree *tree, struct page *page, gfp_t mask) { - u64 start = (u64)page->index << PAGE_CACHE_SHIFT; + u64 start = page_offset(page); u64 end = start + PAGE_CACHE_SIZE - 1; int ret = 1; @@ -3739,7 +3737,7 @@ int try_release_extent_mapping(struct extent_map_tree *map, gfp_t mask) { struct extent_map *em; - u64 start = (u64)page->index << PAGE_CACHE_SHIFT; + u64 start = page_offset(page); u64 end = start + PAGE_CACHE_SIZE - 1; if ((mask & __GFP_WAIT) && diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 1ea0988b82e1..35d152444932 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2062,7 +2062,7 @@ static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end, static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end, struct extent_state *state, int mirror) { - size_t offset = start - ((u64)page->index << PAGE_CACHE_SHIFT); + size_t offset = start - page_offset(page); struct inode *inode = page->mapping->host; struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; char *kaddr; @@ -6752,8 +6752,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset) return; } lock_extent_bits(tree, page_start, page_end, 0, &cached_state); - ordered = btrfs_lookup_ordered_extent(inode, - page_offset(page)); + ordered = btrfs_lookup_ordered_extent(inode, page_offset(page)); if (ordered) { /* * IO on this page will never be started, so we need diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 300e09ac3659..ba5a3210da9a 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -3017,7 +3017,7 @@ static int relocate_file_extent_cluster(struct inode *inode, } } - page_start = (u64)page->index << PAGE_CACHE_SHIFT; + page_start = page_offset(page); page_end = page_start + PAGE_CACHE_SIZE - 1; lock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end); -- cgit From dfd79829b709af3c2ac55951353a874ae89f41c3 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Fri, 21 Dec 2012 09:21:30 +0000 Subject: Btrfs: fix trivial error in btrfs_ioctl_resize() This patch fixes the following problem: - improper return value - unnecessary read-only check Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index a31cd931d36e..744aabf11ce5 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1329,9 +1329,6 @@ static noinline int btrfs_ioctl_resize(struct file *file, int ret = 0; int mod = 0; - if (root->fs_info->sb->s_flags & MS_RDONLY) - return -EROFS; - if (!capable(CAP_SYS_ADMIN)) return -EPERM; @@ -1363,6 +1360,10 @@ static noinline int btrfs_ioctl_resize(struct file *file, *devstr = '\0'; devstr = vol_args->name; devid = simple_strtoull(devstr, &end, 10); + if (!devid) { + ret = -EINVAL; + goto out_free; + } printk(KERN_INFO "btrfs: resizing devid %llu\n", (unsigned long long)devid); } @@ -1371,7 +1372,7 @@ static noinline int btrfs_ioctl_resize(struct file *file, if (!device) { printk(KERN_INFO "btrfs: resizer unable to find device %llu\n", (unsigned long long)devid); - ret = -EINVAL; + ret = -ENODEV; goto out_free; } @@ -1379,7 +1380,7 @@ static noinline int btrfs_ioctl_resize(struct file *file, printk(KERN_INFO "btrfs: resizer unable to apply on " "readonly device %llu\n", (unsigned long long)devid); - ret = -EINVAL; + ret = -EPERM; goto out_free; } @@ -1401,7 +1402,7 @@ static noinline int btrfs_ioctl_resize(struct file *file, } if (device->is_tgtdev_for_dev_replace) { - ret = -EINVAL; + ret = -EPERM; goto out_free; } -- cgit From 2ab28f322f9896782da904f5942f3873432addc8 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 12 Oct 2012 15:27:49 -0400 Subject: Btrfs: wait on ordered extents at the last possible moment Since we don't actually copy the extent information from the source tree in the fast case we don't need to wait for ordered io to be completed in order to fsync, we just need to wait for the io to be completed. So when we're logging our file just attach all of the ordered extents to the log, and then when the log syncs just wait for IO_DONE on the ordered extents and then write the super. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 3 ++ fs/btrfs/disk-io.c | 4 ++ fs/btrfs/file.c | 30 +++++++++-- fs/btrfs/inode.c | 8 +++ fs/btrfs/ordered-data.c | 68 +++++++++++++++++++++++++ fs/btrfs/ordered-data.h | 11 ++++ fs/btrfs/tree-log.c | 132 ++++++++++++++++++++++++++++++++++++++++++++++-- 7 files changed, 247 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 547b7b05727f..411c8d97074e 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1623,6 +1623,9 @@ struct btrfs_root { struct list_head root_list; + spinlock_t log_extents_lock[2]; + struct list_head logged_list[2]; + spinlock_t orphan_lock; atomic_t orphan_inodes; struct btrfs_block_rsv *orphan_block_rsv; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index a762f9137610..1db8a9938829 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1178,9 +1178,13 @@ static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, INIT_LIST_HEAD(&root->dirty_list); INIT_LIST_HEAD(&root->root_list); + INIT_LIST_HEAD(&root->logged_list[0]); + INIT_LIST_HEAD(&root->logged_list[1]); spin_lock_init(&root->orphan_lock); spin_lock_init(&root->inode_lock); spin_lock_init(&root->accounting_lock); + spin_lock_init(&root->log_extents_lock[0]); + spin_lock_init(&root->log_extents_lock[1]); mutex_init(&root->objectid_mutex); mutex_init(&root->log_mutex); init_waitqueue_head(&root->log_writer_wait); diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index b06d289f998f..083abca56055 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1655,16 +1655,21 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) struct btrfs_root *root = BTRFS_I(inode)->root; int ret = 0; struct btrfs_trans_handle *trans; + bool full_sync = 0; trace_btrfs_sync_file(file, datasync); /* * We write the dirty pages in the range and wait until they complete * out of the ->i_mutex. If so, we can flush the dirty pages by - * multi-task, and make the performance up. + * multi-task, and make the performance up. See + * btrfs_wait_ordered_range for an explanation of the ASYNC check. */ atomic_inc(&BTRFS_I(inode)->sync_writers); - ret = filemap_write_and_wait_range(inode->i_mapping, start, end); + ret = filemap_fdatawrite_range(inode->i_mapping, start, end); + if (!ret && test_bit(BTRFS_INODE_HAS_ASYNC_EXTENT, + &BTRFS_I(inode)->runtime_flags)) + ret = filemap_fdatawrite_range(inode->i_mapping, start, end); atomic_dec(&BTRFS_I(inode)->sync_writers); if (ret) return ret; @@ -1676,7 +1681,10 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) * range being left. */ atomic_inc(&root->log_batch); - btrfs_wait_ordered_range(inode, start, end - start + 1); + full_sync = test_bit(BTRFS_INODE_NEEDS_FULL_SYNC, + &BTRFS_I(inode)->runtime_flags); + if (full_sync) + btrfs_wait_ordered_range(inode, start, end - start + 1); atomic_inc(&root->log_batch); /* @@ -1743,13 +1751,25 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) if (ret != BTRFS_NO_LOG_SYNC) { if (ret > 0) { + /* + * If we didn't already wait for ordered extents we need + * to do that now. + */ + if (!full_sync) + btrfs_wait_ordered_range(inode, start, + end - start + 1); ret = btrfs_commit_transaction(trans, root); } else { ret = btrfs_sync_log(trans, root); - if (ret == 0) + if (ret == 0) { ret = btrfs_end_transaction(trans, root); - else + } else { + if (!full_sync) + btrfs_wait_ordered_range(inode, start, + end - + start + 1); ret = btrfs_commit_transaction(trans, root); + } } } else { ret = btrfs_end_transaction(trans, root); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 35d152444932..31a871ec48f2 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -700,6 +700,8 @@ retry: em->start = async_extent->start; em->len = async_extent->ram_size; em->orig_start = em->start; + em->mod_start = em->start; + em->mod_len = em->len; em->block_start = ins.objectid; em->block_len = ins.offset; @@ -892,6 +894,8 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans, em->orig_start = em->start; ram_size = ins.offset; em->len = ins.offset; + em->mod_start = em->start; + em->mod_len = em->len; em->block_start = ins.objectid; em->block_len = ins.offset; @@ -1338,6 +1342,8 @@ out_check: em->block_start = disk_bytenr; em->orig_block_len = disk_num_bytes; em->bdev = root->fs_info->fs_devices->latest_bdev; + em->mod_start = em->start; + em->mod_len = em->len; set_bit(EXTENT_FLAG_PINNED, &em->flags); set_bit(EXTENT_FLAG_FILLING, &em->flags); em->generation = -1; @@ -5966,6 +5972,8 @@ static struct extent_map *create_pinned_em(struct inode *inode, u64 start, em->start = start; em->orig_start = orig_start; + em->mod_start = start; + em->mod_len = len; em->len = len; em->block_len = block_len; em->block_start = block_start; diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index e5ed56729607..f14b17432117 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -196,6 +196,9 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, entry->file_offset = file_offset; entry->start = start; entry->len = len; + if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM) && + !(type == BTRFS_ORDERED_NOCOW)) + entry->csum_bytes_left = disk_len; entry->disk_len = disk_len; entry->bytes_left = len; entry->inode = igrab(inode); @@ -213,6 +216,7 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, INIT_LIST_HEAD(&entry->root_extent_list); INIT_LIST_HEAD(&entry->work_list); init_completion(&entry->completion); + INIT_LIST_HEAD(&entry->log_list); trace_btrfs_ordered_extent_add(inode, entry); @@ -270,6 +274,10 @@ void btrfs_add_ordered_sum(struct inode *inode, tree = &BTRFS_I(inode)->ordered_tree; spin_lock_irq(&tree->lock); list_add_tail(&sum->list, &entry->list); + WARN_ON(entry->csum_bytes_left < sum->len); + entry->csum_bytes_left -= sum->len; + if (entry->csum_bytes_left == 0) + wake_up(&entry->wait); spin_unlock_irq(&tree->lock); } @@ -405,6 +413,66 @@ out: return ret == 0; } +/* Needs to either be called under a log transaction or the log_mutex */ +void btrfs_get_logged_extents(struct btrfs_root *log, struct inode *inode) +{ + struct btrfs_ordered_inode_tree *tree; + struct btrfs_ordered_extent *ordered; + struct rb_node *n; + int index = log->log_transid % 2; + + tree = &BTRFS_I(inode)->ordered_tree; + spin_lock_irq(&tree->lock); + for (n = rb_first(&tree->tree); n; n = rb_next(n)) { + ordered = rb_entry(n, struct btrfs_ordered_extent, rb_node); + spin_lock(&log->log_extents_lock[index]); + if (list_empty(&ordered->log_list)) { + list_add_tail(&ordered->log_list, &log->logged_list[index]); + atomic_inc(&ordered->refs); + } + spin_unlock(&log->log_extents_lock[index]); + } + spin_unlock_irq(&tree->lock); +} + +void btrfs_wait_logged_extents(struct btrfs_root *log, u64 transid) +{ + struct btrfs_ordered_extent *ordered; + int index = transid % 2; + + spin_lock_irq(&log->log_extents_lock[index]); + while (!list_empty(&log->logged_list[index])) { + ordered = list_first_entry(&log->logged_list[index], + struct btrfs_ordered_extent, + log_list); + list_del_init(&ordered->log_list); + spin_unlock_irq(&log->log_extents_lock[index]); + wait_event(ordered->wait, test_bit(BTRFS_ORDERED_IO_DONE, + &ordered->flags)); + btrfs_put_ordered_extent(ordered); + spin_lock_irq(&log->log_extents_lock[index]); + } + spin_unlock_irq(&log->log_extents_lock[index]); +} + +void btrfs_free_logged_extents(struct btrfs_root *log, u64 transid) +{ + struct btrfs_ordered_extent *ordered; + int index = transid % 2; + + spin_lock_irq(&log->log_extents_lock[index]); + while (!list_empty(&log->logged_list[index])) { + ordered = list_first_entry(&log->logged_list[index], + struct btrfs_ordered_extent, + log_list); + list_del_init(&ordered->log_list); + spin_unlock_irq(&log->log_extents_lock[index]); + btrfs_put_ordered_extent(ordered); + spin_lock_irq(&log->log_extents_lock[index]); + } + spin_unlock_irq(&log->log_extents_lock[index]); +} + /* * used to drop a reference on an ordered extent. This will free * the extent if the last reference is dropped diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h index efc7c2930c17..d523dbd2314d 100644 --- a/fs/btrfs/ordered-data.h +++ b/fs/btrfs/ordered-data.h @@ -79,6 +79,8 @@ struct btrfs_ordered_sum { #define BTRFS_ORDERED_UPDATED_ISIZE 7 /* indicates wether this ordered extent * has done its due diligence in updating * the isize. */ +#define BTRFS_ORDERED_LOGGED_CSUM 8 /* We've logged the csums on this ordered + ordered extent */ struct btrfs_ordered_extent { /* logical offset in the file */ @@ -96,6 +98,9 @@ struct btrfs_ordered_extent { /* number of bytes that still need writing */ u64 bytes_left; + /* number of bytes that still need csumming */ + u64 csum_bytes_left; + /* * the end of the ordered extent which is behind it but * didn't update disk_i_size. Please see the comment of @@ -118,6 +123,9 @@ struct btrfs_ordered_extent { /* list of checksums for insertion when the extent io is done */ struct list_head list; + /* If we need to wait on this to be done */ + struct list_head log_list; + /* used to wait for the BTRFS_ORDERED_COMPLETE bit */ wait_queue_head_t wait; @@ -194,6 +202,9 @@ void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct inode *inode); void btrfs_wait_ordered_extents(struct btrfs_root *root, int delay_iput); +void btrfs_get_logged_extents(struct btrfs_root *log, struct inode *inode); +void btrfs_wait_logged_extents(struct btrfs_root *log, u64 transid); +void btrfs_free_logged_extents(struct btrfs_root *log, u64 transid); int __init ordered_data_init(void); void ordered_data_exit(void); #endif diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 9027bb1e7466..7de720d22b74 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -2281,6 +2281,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, unsigned long log_transid = 0; mutex_lock(&root->log_mutex); + log_transid = root->log_transid; index1 = root->log_transid % 2; if (atomic_read(&root->log_commit[index1])) { wait_log_commit(trans, root, root->log_transid); @@ -2308,11 +2309,11 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, /* bail out if we need to do a full commit */ if (root->fs_info->last_trans_log_full_commit == trans->transid) { ret = -EAGAIN; + btrfs_free_logged_extents(log, log_transid); mutex_unlock(&root->log_mutex); goto out; } - log_transid = root->log_transid; if (log_transid % 2 == 0) mark = EXTENT_DIRTY; else @@ -2324,6 +2325,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, ret = btrfs_write_marked_extents(log, &log->dirty_log_pages, mark); if (ret) { btrfs_abort_transaction(trans, root, ret); + btrfs_free_logged_extents(log, log_transid); mutex_unlock(&root->log_mutex); goto out; } @@ -2363,6 +2365,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, } root->fs_info->last_trans_log_full_commit = trans->transid; btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); + btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); ret = -EAGAIN; goto out; @@ -2373,6 +2376,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); wait_log_commit(trans, log_root_tree, log_root_tree->log_transid); + btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); ret = 0; goto out; @@ -2392,6 +2396,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, */ if (root->fs_info->last_trans_log_full_commit == trans->transid) { btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); + btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); ret = -EAGAIN; goto out_wake_log_root; @@ -2402,10 +2407,12 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, EXTENT_DIRTY | EXTENT_NEW); if (ret) { btrfs_abort_transaction(trans, root, ret); + btrfs_free_logged_extents(log, log_transid); mutex_unlock(&log_root_tree->log_mutex); goto out_wake_log_root; } btrfs_wait_marked_extents(log, &log->dirty_log_pages, mark); + btrfs_wait_logged_extents(log, log_transid); btrfs_set_super_log_root(root->fs_info->super_for_commit, log_root_tree->node->start); @@ -2475,6 +2482,14 @@ static void free_log_tree(struct btrfs_trans_handle *trans, EXTENT_DIRTY | EXTENT_NEW, GFP_NOFS); } + /* + * We may have short-circuited the log tree with the full commit logic + * and left ordered extents on our list, so clear these out to keep us + * from leaking inodes and memory. + */ + btrfs_free_logged_extents(log, 0); + btrfs_free_logged_extents(log, 1); + free_extent_buffer(log->node); kfree(log); } @@ -3271,14 +3286,18 @@ static int log_one_extent(struct btrfs_trans_handle *trans, struct btrfs_root *log = root->log_root; struct btrfs_file_extent_item *fi; struct extent_buffer *leaf; + struct btrfs_ordered_extent *ordered; struct list_head ordered_sums; struct btrfs_map_token token; struct btrfs_key key; - u64 csum_offset = em->mod_start - em->start; - u64 csum_len = em->mod_len; + u64 mod_start = em->mod_start; + u64 mod_len = em->mod_len; + u64 csum_offset; + u64 csum_len; u64 extent_offset = em->start - em->orig_start; u64 block_len; int ret; + int index = log->log_transid % 2; bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM; INIT_LIST_HEAD(&ordered_sums); @@ -3362,6 +3381,92 @@ static int log_one_extent(struct btrfs_trans_handle *trans, csum_len = block_len; } + /* + * First check and see if our csums are on our outstanding ordered + * extents. + */ +again: + spin_lock_irq(&log->log_extents_lock[index]); + list_for_each_entry(ordered, &log->logged_list[index], log_list) { + struct btrfs_ordered_sum *sum; + + if (!mod_len) + break; + + if (ordered->inode != inode) + continue; + + if (ordered->file_offset + ordered->len <= mod_start || + mod_start + mod_len <= ordered->file_offset) + continue; + + /* + * We are going to copy all the csums on this ordered extent, so + * go ahead and adjust mod_start and mod_len in case this + * ordered extent has already been logged. + */ + if (ordered->file_offset > mod_start) { + if (ordered->file_offset + ordered->len >= + mod_start + mod_len) + mod_len = ordered->file_offset - mod_start; + /* + * If we have this case + * + * |--------- logged extent ---------| + * |----- ordered extent ----| + * + * Just don't mess with mod_start and mod_len, we'll + * just end up logging more csums than we need and it + * will be ok. + */ + } else { + if (ordered->file_offset + ordered->len < + mod_start + mod_len) { + mod_len = (mod_start + mod_len) - + (ordered->file_offset + ordered->len); + mod_start = ordered->file_offset + + ordered->len; + } else { + mod_len = 0; + } + } + + /* + * To keep us from looping for the above case of an ordered + * extent that falls inside of the logged extent. + */ + if (test_and_set_bit(BTRFS_ORDERED_LOGGED_CSUM, + &ordered->flags)) + continue; + atomic_inc(&ordered->refs); + spin_unlock_irq(&log->log_extents_lock[index]); + /* + * we've dropped the lock, we must either break or + * start over after this. + */ + + wait_event(ordered->wait, ordered->csum_bytes_left == 0); + + list_for_each_entry(sum, &ordered->list, list) { + ret = btrfs_csum_file_blocks(trans, log, sum); + if (ret) { + btrfs_put_ordered_extent(ordered); + goto unlocked; + } + } + btrfs_put_ordered_extent(ordered); + goto again; + + } + spin_unlock_irq(&log->log_extents_lock[index]); +unlocked: + + if (!mod_len || ret) + return ret; + + csum_offset = mod_start - em->start; + csum_len = mod_len; + /* block start is already adjusted for the file extent offset. */ ret = btrfs_lookup_csums_range(log->fs_info->csum_root, em->block_start + csum_offset, @@ -3393,6 +3498,7 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans, struct extent_map_tree *tree = &BTRFS_I(inode)->extent_tree; u64 test_gen; int ret = 0; + int num = 0; INIT_LIST_HEAD(&extents); @@ -3401,16 +3507,31 @@ static int btrfs_log_changed_extents(struct btrfs_trans_handle *trans, list_for_each_entry_safe(em, n, &tree->modified_extents, list) { list_del_init(&em->list); + + /* + * Just an arbitrary number, this can be really CPU intensive + * once we start getting a lot of extents, and really once we + * have a bunch of extents we just want to commit since it will + * be faster. + */ + if (++num > 32768) { + list_del_init(&tree->modified_extents); + ret = -EFBIG; + goto process; + } + if (em->generation <= test_gen) continue; /* Need a ref to keep it from getting evicted from cache */ atomic_inc(&em->refs); set_bit(EXTENT_FLAG_LOGGING, &em->flags); list_add_tail(&em->list, &extents); + num++; } list_sort(NULL, &extents, extent_cmp); +process: while (!list_empty(&extents)) { em = list_entry(extents.next, struct extent_map, list); @@ -3513,6 +3634,8 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, mutex_lock(&BTRFS_I(inode)->log_mutex); + btrfs_get_logged_extents(log, inode); + /* * a brute force approach to making sure we get the most uptodate * copies of everything. @@ -3656,6 +3779,8 @@ log_extents: BTRFS_I(inode)->logged_trans = trans->transid; BTRFS_I(inode)->last_log_commit = BTRFS_I(inode)->last_sub_trans; out_unlock: + if (err) + btrfs_free_logged_extents(log, log->log_transid); mutex_unlock(&BTRFS_I(inode)->log_mutex); btrfs_free_path(path); @@ -3822,7 +3947,6 @@ int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, end_trans: dput(old_parent); if (ret < 0) { - WARN_ON(ret != -ENOSPC); root->fs_info->last_trans_log_full_commit = trans->transid; ret = 1; } -- cgit From c6b305a89b1903d63652691ad5eb9f05aa0326b8 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 18 Dec 2012 09:16:16 -0500 Subject: Btrfs: don't re-enter when allocating a chunk If we start running low on metadata space we will try to allocate a chunk, which could then try to allocate a chunk to add the device entry. The thing is we allocate a chunk before we try really hard to make the allocation, so we should be able to find space for the device entry. Add a flag to the trans handle so we know we're currently allocating a chunk so we can just bail out if we try to allocate another chunk. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 7 +++++++ fs/btrfs/transaction.c | 1 + fs/btrfs/transaction.h | 1 + 3 files changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index d5e60d25ca51..c642fc2cfc2b 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3570,6 +3570,10 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, int wait_for_alloc = 0; int ret = 0; + /* Don't re-enter if we're already allocating a chunk */ + if (trans->allocating_chunk) + return -ENOSPC; + space_info = __find_space_info(extent_root->fs_info, flags); if (!space_info) { ret = update_space_info(extent_root->fs_info, flags, @@ -3612,6 +3616,8 @@ again: goto again; } + trans->allocating_chunk = true; + /* * If we have mixed data/metadata chunks we want to make sure we keep * allocating mixed chunks instead of individual chunks. @@ -3638,6 +3644,7 @@ again: check_system_chunk(trans, extent_root, flags); ret = btrfs_alloc_chunk(trans, extent_root, flags); + trans->allocating_chunk = false; if (ret < 0 && ret != -ENOSPC) goto out; diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 86bb105b3982..24fde97cba81 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -385,6 +385,7 @@ again: h->qgroup_reserved = qgroup_reserved; h->delayed_ref_elem.seq = 0; h->type = type; + h->allocating_chunk = false; INIT_LIST_HEAD(&h->qgroup_ref_list); INIT_LIST_HEAD(&h->new_bgs); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 0e8aa1e6c287..69700f7b20ac 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -68,6 +68,7 @@ struct btrfs_trans_handle { struct btrfs_block_rsv *orig_rsv; short aborted; short adding_csums; + bool allocating_chunk; enum btrfs_trans_type type; /* * this root is only needed to validate that the root passed to -- cgit From 17b85495cf87ac9b2d9e0878f62fa9f3ae630a6e Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Thu, 27 Dec 2012 09:01:17 +0000 Subject: Btrfs: remove deprecated comments commit d53ba47484ed6245e640ee4bfe9d21e9bfc15765 (Btrfs: use commit root when loading free space cache) has remove the deadlock check, and the related comments can be removed as well. Signed-off-by: Liu Bo Reviewed-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index c642fc2cfc2b..a4a062a0a79c 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -527,12 +527,6 @@ static int cache_block_group(struct btrfs_block_group_cache *cache, cache->cached = BTRFS_CACHE_FAST; spin_unlock(&cache->lock); - /* - * We can't do the read from on-disk cache during a commit since we need - * to have the normal tree locking. Also if we are currently trying to - * allocate blocks for the tree root we can't do the fast caching since - * we likely hold important locks. - */ if (fs_info->mount_opt & BTRFS_MOUNT_SPACE_CACHE) { ret = load_free_space_cache(fs_info, cache); -- cgit From f6373bf3dc6b78cd2c8cfba0215104003ba4a1d0 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Thu, 27 Dec 2012 09:01:18 +0000 Subject: Btrfs: kill unused arguments of cache_block_group Argument 'trans' and 'root' are not used any more. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a4a062a0a79c..a9dc12d5ac1b 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -468,8 +468,6 @@ out: } static int cache_block_group(struct btrfs_block_group_cache *cache, - struct btrfs_trans_handle *trans, - struct btrfs_root *root, int load_cache_only) { DEFINE_WAIT(wait); @@ -4802,7 +4800,7 @@ static int update_block_group(struct btrfs_trans_handle *trans, * space back to the block group, otherwise we will leak space. */ if (!alloc && cache->cached == BTRFS_CACHE_NO) - cache_block_group(cache, trans, NULL, 1); + cache_block_group(cache, 1); byte_in_group = bytenr - cache->key.objectid; WARN_ON(byte_in_group > cache->key.offset); @@ -4917,7 +4915,7 @@ int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans, * to one because the slow code to read in the free extents does check * the pinned extents. */ - cache_block_group(cache, trans, root, 1); + cache_block_group(cache, 1); pin_down_extent(root, cache, bytenr, num_bytes, 0); @@ -5707,8 +5705,7 @@ have_block_group: cached = block_group_cache_done(block_group); if (unlikely(!cached)) { found_uncached_bg = true; - ret = cache_block_group(block_group, trans, - orig_root, 0); + ret = cache_block_group(block_group, 0); BUG_ON(ret < 0); ret = 0; } @@ -6244,7 +6241,7 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, u64 num_bytes = ins->offset; block_group = btrfs_lookup_block_group(root->fs_info, ins->objectid); - cache_block_group(block_group, trans, NULL, 0); + cache_block_group(block_group, 0); caching_ctl = get_caching_control(block_group); if (!caching_ctl) { @@ -8187,7 +8184,7 @@ int btrfs_trim_fs(struct btrfs_root *root, struct fstrim_range *range) if (end - start >= range->minlen) { if (!block_group_cache_done(cache)) { - ret = cache_block_group(cache, NULL, root, 0); + ret = cache_block_group(cache, 0); if (!ret) wait_block_group_cache_done(cache); } -- cgit From c53d613e5293ef248e42195bb0595502b646a5a6 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Thu, 27 Dec 2012 09:01:19 +0000 Subject: Btrfs: kill unused argument of update_block_group Argument 'trans' is not used any more. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a9dc12d5ac1b..a162c7cb8fb5 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -72,8 +72,7 @@ enum { RESERVE_ALLOC_NO_ACCOUNT = 2, }; -static int update_block_group(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +static int update_block_group(struct btrfs_root *root, u64 bytenr, u64 num_bytes, int alloc); static int __btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -4762,8 +4761,7 @@ void btrfs_delalloc_release_space(struct inode *inode, u64 num_bytes) btrfs_free_reserved_data_space(inode, num_bytes); } -static int update_block_group(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +static int update_block_group(struct btrfs_root *root, u64 bytenr, u64 num_bytes, int alloc) { struct btrfs_block_group_cache *cache = NULL; @@ -5312,7 +5310,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, } } - ret = update_block_group(trans, root, bytenr, num_bytes, 0); + ret = update_block_group(root, bytenr, num_bytes, 0); if (ret) { btrfs_abort_transaction(trans, extent_root, ret); goto out; @@ -6134,7 +6132,7 @@ static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(path->nodes[0]); btrfs_free_path(path); - ret = update_block_group(trans, root, ins->objectid, ins->offset, 1); + ret = update_block_group(root, ins->objectid, ins->offset, 1); if (ret) { /* -ENOENT, logic error */ printk(KERN_ERR "btrfs update block group failed for %llu " "%llu\n", (unsigned long long)ins->objectid, @@ -6198,7 +6196,7 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(leaf); btrfs_free_path(path); - ret = update_block_group(trans, root, ins->objectid, ins->offset, 1); + ret = update_block_group(root, ins->objectid, ins->offset, 1); if (ret) { /* -ENOENT, logic error */ printk(KERN_ERR "btrfs update block group failed for %llu " "%llu\n", (unsigned long long)ins->objectid, -- cgit From dcfac4156fa102c1bab0e4e31df37e47278292f6 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Thu, 27 Dec 2012 09:01:20 +0000 Subject: Btrfs: kill unused argument of btrfs_pin_extent_for_log_replay Argument 'trans' is not used any more. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 3 +-- fs/btrfs/extent-tree.c | 3 +-- fs/btrfs/tree-log.c | 3 +-- 3 files changed, 3 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 411c8d97074e..22f012d41fd0 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2939,8 +2939,7 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans, u64 num_bytes, u64 *refs, u64 *flags); int btrfs_pin_extent(struct btrfs_root *root, u64 bytenr, u64 num, int reserved); -int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +int btrfs_pin_extent_for_log_replay(struct btrfs_root *root, u64 bytenr, u64 num_bytes); int btrfs_cross_ref_exist(struct btrfs_trans_handle *trans, struct btrfs_root *root, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a162c7cb8fb5..825f23b13b58 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4898,8 +4898,7 @@ int btrfs_pin_extent(struct btrfs_root *root, /* * this function must be called within transaction */ -int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans, - struct btrfs_root *root, +int btrfs_pin_extent_for_log_replay(struct btrfs_root *root, u64 bytenr, u64 num_bytes) { struct btrfs_block_group_cache *cache; diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 7de720d22b74..deb7f66356bd 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -278,8 +278,7 @@ static int process_one_buffer(struct btrfs_root *log, struct walk_control *wc, u64 gen) { if (wc->pin) - btrfs_pin_extent_for_log_replay(wc->trans, - log->fs_info->extent_root, + btrfs_pin_extent_for_log_replay(log->fs_info->extent_root, eb->start, eb->len); if (btrfs_buffer_uptodate(eb, gen, 0)) { -- cgit From 51fab693477c39ba82c153d3fb30a4cc6de49874 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Thu, 27 Dec 2012 09:01:21 +0000 Subject: Btrfs: use token to avoid times mapping extent buffer The API in tree log code has done sort of changes, and it proves that we can benifit from using token, so do the same thing here. function_graph tracer's timer shows that it costs nearly half time of before(39.788us -> 22.391us). Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 63 +++++++++++++++++++++++++++++++------------------------- 1 file changed, 35 insertions(+), 28 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 31a871ec48f2..280e4d487636 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2715,34 +2715,41 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, struct btrfs_inode_item *item, struct inode *inode) { - btrfs_set_inode_uid(leaf, item, i_uid_read(inode)); - btrfs_set_inode_gid(leaf, item, i_gid_read(inode)); - btrfs_set_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size); - btrfs_set_inode_mode(leaf, item, inode->i_mode); - btrfs_set_inode_nlink(leaf, item, inode->i_nlink); - - btrfs_set_timespec_sec(leaf, btrfs_inode_atime(item), - inode->i_atime.tv_sec); - btrfs_set_timespec_nsec(leaf, btrfs_inode_atime(item), - inode->i_atime.tv_nsec); - - btrfs_set_timespec_sec(leaf, btrfs_inode_mtime(item), - inode->i_mtime.tv_sec); - btrfs_set_timespec_nsec(leaf, btrfs_inode_mtime(item), - inode->i_mtime.tv_nsec); - - btrfs_set_timespec_sec(leaf, btrfs_inode_ctime(item), - inode->i_ctime.tv_sec); - btrfs_set_timespec_nsec(leaf, btrfs_inode_ctime(item), - inode->i_ctime.tv_nsec); - - btrfs_set_inode_nbytes(leaf, item, inode_get_bytes(inode)); - btrfs_set_inode_generation(leaf, item, BTRFS_I(inode)->generation); - btrfs_set_inode_sequence(leaf, item, inode->i_version); - btrfs_set_inode_transid(leaf, item, trans->transid); - btrfs_set_inode_rdev(leaf, item, inode->i_rdev); - btrfs_set_inode_flags(leaf, item, BTRFS_I(inode)->flags); - btrfs_set_inode_block_group(leaf, item, 0); + struct btrfs_map_token token; + + btrfs_init_map_token(&token); + + btrfs_set_token_inode_uid(leaf, item, i_uid_read(inode), &token); + btrfs_set_token_inode_gid(leaf, item, i_gid_read(inode), &token); + btrfs_set_token_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size, + &token); + btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token); + btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token); + + btrfs_set_token_timespec_sec(leaf, btrfs_inode_atime(item), + inode->i_atime.tv_sec, &token); + btrfs_set_token_timespec_nsec(leaf, btrfs_inode_atime(item), + inode->i_atime.tv_nsec, &token); + + btrfs_set_token_timespec_sec(leaf, btrfs_inode_mtime(item), + inode->i_mtime.tv_sec, &token); + btrfs_set_token_timespec_nsec(leaf, btrfs_inode_mtime(item), + inode->i_mtime.tv_nsec, &token); + + btrfs_set_token_timespec_sec(leaf, btrfs_inode_ctime(item), + inode->i_ctime.tv_sec, &token); + btrfs_set_token_timespec_nsec(leaf, btrfs_inode_ctime(item), + inode->i_ctime.tv_nsec, &token); + + btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode), + &token); + btrfs_set_token_inode_generation(leaf, item, BTRFS_I(inode)->generation, + &token); + btrfs_set_token_inode_sequence(leaf, item, inode->i_version, &token); + btrfs_set_token_inode_transid(leaf, item, trans->transid, &token); + btrfs_set_token_inode_rdev(leaf, item, inode->i_rdev, &token); + btrfs_set_token_inode_flags(leaf, item, BTRFS_I(inode)->flags, &token); + btrfs_set_token_inode_block_group(leaf, item, 0, &token); } /* -- cgit From 39f9d028c9906cc7b625df84442205f6bcc84477 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Thu, 27 Dec 2012 09:01:22 +0000 Subject: Btrfs: save us a read_lock This does not change the logic of code, but can save us a read_lock. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/locking.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c index 2a1762c66041..e95df435d897 100644 --- a/fs/btrfs/locking.c +++ b/fs/btrfs/locking.c @@ -113,11 +113,10 @@ again: read_unlock(&eb->lock); return; } - read_unlock(&eb->lock); - wait_event(eb->write_lock_wq, atomic_read(&eb->blocking_writers) == 0); - read_lock(&eb->lock); if (atomic_read(&eb->blocking_writers)) { read_unlock(&eb->lock); + wait_event(eb->write_lock_wq, + atomic_read(&eb->blocking_writers) == 0); goto again; } atomic_inc(&eb->read_locks); -- cgit From a1897fddd28daf6b23d05a30dc2a18836f77f8e3 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Thu, 27 Dec 2012 09:01:23 +0000 Subject: Btrfs: record first logical byte in memory This'd save us a rbtree search which may become expensive in large filesystem. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 1 + fs/btrfs/disk-io.c | 1 + fs/btrfs/extent-tree.c | 19 ++++++++++++++++++- 3 files changed, 20 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 22f012d41fd0..9ee099f3f834 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1250,6 +1250,7 @@ struct btrfs_fs_info { /* block group cache stuff */ spinlock_t block_group_cache_lock; + u64 first_logical_byte; struct rb_root block_group_cache_tree; /* keep track of unallocated space */ diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 1db8a9938829..04f98e3ffd90 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2130,6 +2130,7 @@ int open_ctree(struct super_block *sb, spin_lock_init(&fs_info->block_group_cache_lock); fs_info->block_group_cache_tree = RB_ROOT; + fs_info->first_logical_byte = (u64)-1; extent_io_tree_init(&fs_info->freed_extents[0], fs_info->btree_inode->i_mapping); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 825f23b13b58..82400b2b2517 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -161,6 +161,10 @@ static int btrfs_add_block_group_cache(struct btrfs_fs_info *info, rb_link_node(&block_group->cache_node, parent, p); rb_insert_color(&block_group->cache_node, &info->block_group_cache_tree); + + if (info->first_logical_byte > block_group->key.objectid) + info->first_logical_byte = block_group->key.objectid; + spin_unlock(&info->block_group_cache_lock); return 0; @@ -202,8 +206,11 @@ block_group_cache_tree_search(struct btrfs_fs_info *info, u64 bytenr, break; } } - if (ret) + if (ret) { btrfs_get_block_group(ret); + if (bytenr == 0 && info->first_logical_byte > ret->key.objectid) + info->first_logical_byte = ret->key.objectid; + } spin_unlock(&info->block_group_cache_lock); return ret; @@ -4848,6 +4855,13 @@ static u64 first_logical_byte(struct btrfs_root *root, u64 search_start) struct btrfs_block_group_cache *cache; u64 bytenr; + spin_lock(&root->fs_info->block_group_cache_lock); + bytenr = root->fs_info->first_logical_byte; + spin_unlock(&root->fs_info->block_group_cache_lock); + + if (bytenr < (u64)-1) + return bytenr; + cache = btrfs_lookup_first_block_group(root->fs_info, search_start); if (!cache) return 0; @@ -8059,6 +8073,9 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, spin_lock(&root->fs_info->block_group_cache_lock); rb_erase(&block_group->cache_node, &root->fs_info->block_group_cache_tree); + + if (root->fs_info->first_logical_byte == block_group->key.objectid) + root->fs_info->first_logical_byte = (u64)-1; spin_unlock(&root->fs_info->block_group_cache_lock); down_write(&block_group->space_info->groups_sem); -- cgit From e6ec716f0ddbe51741ef261d0804f0c28038dda4 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 17 Jan 2013 05:38:51 +0000 Subject: Btrfs: make raid attr array more readable The current code of raid attr arry is hard to understand and it is easy to introduce some problem if we modify the array. So I changed it and made it more readable. Cc: Liu Bo Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 10 +++++++++- fs/btrfs/extent-tree.c | 22 +++++++++------------- fs/btrfs/volumes.c | 47 +++++++++++++++++++++++++++++++++++++++++------ 3 files changed, 59 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 9ee099f3f834..541ce9a9949e 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -953,7 +953,15 @@ struct btrfs_dev_replace_item { #define BTRFS_BLOCK_GROUP_DUP (1ULL << 5) #define BTRFS_BLOCK_GROUP_RAID10 (1ULL << 6) #define BTRFS_BLOCK_GROUP_RESERVED BTRFS_AVAIL_ALLOC_BIT_SINGLE -#define BTRFS_NR_RAID_TYPES 5 + +enum btrfs_raid_types { + BTRFS_RAID_RAID10, + BTRFS_RAID_RAID1, + BTRFS_RAID_DUP, + BTRFS_RAID_RAID0, + BTRFS_RAID_SINGLE, + BTRFS_NR_RAID_TYPES +}; #define BTRFS_BLOCK_GROUP_TYPE_MASK (BTRFS_BLOCK_GROUP_DATA | \ BTRFS_BLOCK_GROUP_SYSTEM | \ diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 82400b2b2517..174c4d5c692c 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5545,20 +5545,16 @@ wait_block_group_cache_done(struct btrfs_block_group_cache *cache) int __get_raid_index(u64 flags) { - int index; - if (flags & BTRFS_BLOCK_GROUP_RAID10) - index = 0; + return BTRFS_RAID_RAID10; else if (flags & BTRFS_BLOCK_GROUP_RAID1) - index = 1; + return BTRFS_RAID_RAID1; else if (flags & BTRFS_BLOCK_GROUP_DUP) - index = 2; + return BTRFS_RAID_DUP; else if (flags & BTRFS_BLOCK_GROUP_RAID0) - index = 3; + return BTRFS_RAID_RAID0; else - index = 4; - - return index; + return BTRFS_RAID_SINGLE; } static int get_block_group_index(struct btrfs_block_group_cache *cache) @@ -7518,16 +7514,16 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr) index = get_block_group_index(block_group); } - if (index == 0) { + if (index == BTRFS_RAID_RAID10) { dev_min = 4; /* Divide by 2 */ min_free >>= 1; - } else if (index == 1) { + } else if (index == BTRFS_RAID_RAID1) { dev_min = 2; - } else if (index == 2) { + } else if (index == BTRFS_RAID_DUP) { /* Multiply by 2 */ min_free <<= 1; - } else if (index == 3) { + } else if (index == BTRFS_RAID_RAID0) { dev_min = fs_devices->rw_devices; do_div(min_free, dev_min); } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 5349e17d8863..1ce581d58d87 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3562,13 +3562,48 @@ static int btrfs_cmp_device_info(const void *a, const void *b) } struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = { - { 2, 1, 0, 4, 2, 2 /* raid10 */ }, - { 1, 1, 2, 2, 2, 2 /* raid1 */ }, - { 1, 2, 1, 1, 1, 2 /* dup */ }, - { 1, 1, 0, 2, 1, 1 /* raid0 */ }, - { 1, 1, 1, 1, 1, 1 /* single */ }, + [BTRFS_RAID_RAID10] = { + .sub_stripes = 2, + .dev_stripes = 1, + .devs_max = 0, /* 0 == as many as possible */ + .devs_min = 4, + .devs_increment = 2, + .ncopies = 2, + }, + [BTRFS_RAID_RAID1] = { + .sub_stripes = 1, + .dev_stripes = 1, + .devs_max = 2, + .devs_min = 2, + .devs_increment = 2, + .ncopies = 2, + }, + [BTRFS_RAID_DUP] = { + .sub_stripes = 1, + .dev_stripes = 2, + .devs_max = 1, + .devs_min = 1, + .devs_increment = 1, + .ncopies = 2, + }, + [BTRFS_RAID_RAID0] = { + .sub_stripes = 1, + .dev_stripes = 1, + .devs_max = 0, + .devs_min = 2, + .devs_increment = 1, + .ncopies = 1, + }, + [BTRFS_RAID_SINGLE] = { + .sub_stripes = 1, + .dev_stripes = 1, + .devs_max = 1, + .devs_min = 1, + .devs_increment = 1, + .ncopies = 1, + }, }; - + static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, struct btrfs_root *extent_root, struct map_lookup **map_ret, -- cgit From 3edb2a68cb23cd6ca84022421eeae2604722cdc4 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 22 Jan 2013 10:49:33 +0000 Subject: Btrfs: check the return value of btrfs_start_delalloc_inodes() We forget to check the return value of btrfs_start_delalloc_inodes(), fix it. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/dev-replace.c | 6 +++++- fs/btrfs/transaction.c | 4 +++- 2 files changed, 8 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c index 66dbc8dbddf7..7ba7b3900cb8 100644 --- a/fs/btrfs/dev-replace.c +++ b/fs/btrfs/dev-replace.c @@ -465,7 +465,11 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info, * flush all outstanding I/O and inode extent mappings before the * copy operation is declared as being finished */ - btrfs_start_delalloc_inodes(root, 0); + ret = btrfs_start_delalloc_inodes(root, 0); + if (ret) { + mutex_unlock(&dev_replace->lock_finishing_cancel_unmount); + return ret; + } btrfs_wait_ordered_extents(root, 0); trans = btrfs_start_transaction(root, 0); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 24fde97cba81..42dac27207ab 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1427,7 +1427,9 @@ static int btrfs_flush_all_pending_stuffs(struct btrfs_trans_handle *trans, } if (flush_on_commit || snap_pending) { - btrfs_start_delalloc_inodes(root, 1); + ret = btrfs_start_delalloc_inodes(root, 1); + if (ret) + return ret; btrfs_wait_ordered_extents(root, 1); } -- cgit From eebc60840636e7351371fc17bcd057384bf0c16a Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 22 Jan 2013 10:50:01 +0000 Subject: Btrfs: check the return value of btrfs_run_ordered_operations() We forget to check the return value of btrfs_run_ordered_operations() when flushing all the pending stuffs, fix it. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 42dac27207ab..34610dc6d140 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1451,9 +1451,9 @@ static int btrfs_flush_all_pending_stuffs(struct btrfs_trans_handle *trans, * it here and no for sure that nothing new will be added * to the list */ - btrfs_run_ordered_operations(root, 1); + ret = btrfs_run_ordered_operations(root, 1); - return 0; + return ret; } /* -- cgit From 63607cc86ab808e077a895be2d7895dcbf6f4cf3 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 22 Jan 2013 10:50:35 +0000 Subject: Btrfs: traverse and flush the delalloc inodes once btrfs_start_delalloc_inodes() needn't traverse and flush the delalloc inodes repeatedly. It is because we can regard the data that the users write after we start delalloc inodes flush as the one which is after the delalloc inodes flush is done, and we can flush it next time. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 280e4d487636..d1e3a2d15f91 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7614,7 +7614,7 @@ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput) INIT_LIST_HEAD(&works); INIT_LIST_HEAD(&splice); -again: + spin_lock(&root->fs_info->delalloc_lock); list_splice_init(&root->fs_info->delalloc_inodes, &splice); while (!list_empty(&splice)) { @@ -7650,13 +7650,6 @@ again: btrfs_wait_and_free_delalloc_work(work); } - spin_lock(&root->fs_info->delalloc_lock); - if (!list_empty(&root->fs_info->delalloc_inodes)) { - spin_unlock(&root->fs_info->delalloc_lock); - goto again; - } - spin_unlock(&root->fs_info->delalloc_lock); - /* the filemap_flush will queue IO into the worker threads, but * we have to make sure the IO is actually started and that * ordered extents get created before we return -- cgit From 5b947f1ba959d4b19068f687ecd39900d1eea1aa Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 22 Jan 2013 10:52:04 +0000 Subject: Btrfs: don't traverse the ordered operation list repeatedly btrfs_run_ordered_operations() needn't traverse the ordered operation list repeatedly, it is because the transaction commiter will invoke it again when there is no other writer in this transaction, it can ensure that no one can add new objects into the ordered operation list. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ordered-data.c | 19 +++++-------------- 1 file changed, 5 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index f14b17432117..9489fa96e3ed 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -626,14 +626,10 @@ int btrfs_run_ordered_operations(struct btrfs_root *root, int wait) mutex_lock(&root->fs_info->ordered_operations_mutex); spin_lock(&root->fs_info->ordered_extent_lock); -again: list_splice_init(&root->fs_info->ordered_operations, &splice); - while (!list_empty(&splice)) { - btrfs_inode = list_entry(splice.next, struct btrfs_inode, ordered_operations); - inode = &btrfs_inode->vfs_inode; list_del_init(&btrfs_inode->ordered_operations); @@ -642,22 +638,20 @@ again: * the inode may be getting freed (in sys_unlink path). */ inode = igrab(inode); - - if (!wait && inode) { - list_add_tail(&BTRFS_I(inode)->ordered_operations, - &root->fs_info->ordered_operations); - } - if (!inode) continue; + + if (!wait) + list_add_tail(&BTRFS_I(inode)->ordered_operations, + &root->fs_info->ordered_operations); spin_unlock(&root->fs_info->ordered_extent_lock); work = btrfs_alloc_delalloc_work(inode, wait, 1); if (!work) { + spin_lock(&root->fs_info->ordered_extent_lock); if (list_empty(&BTRFS_I(inode)->ordered_operations)) list_add_tail(&btrfs_inode->ordered_operations, &splice); - spin_lock(&root->fs_info->ordered_extent_lock); list_splice_tail(&splice, &root->fs_info->ordered_operations); spin_unlock(&root->fs_info->ordered_extent_lock); @@ -671,9 +665,6 @@ again: cond_resched(); spin_lock(&root->fs_info->ordered_extent_lock); } - if (wait && !list_empty(&root->fs_info->ordered_operations)) - goto again; - spin_unlock(&root->fs_info->ordered_extent_lock); out: list_for_each_entry_safe(work, next, &works, list) { -- cgit From fe5fafbebd7e648aa6cb0a419dc812188b4d4ac8 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 24 Jan 2013 13:49:14 -0500 Subject: Revert "Btrfs: fix permissions of empty files not affected by umask" This reverts commit 2794ed013b3551cbae887ea1b93c52aaacb7370d. Wasn't supposed to get used in btrfs_mknod, it was supposed to be in btrfs_create, which was done in commit 9185aa587b7425f8f4520da2e66792f5f3c2b815. Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index d1e3a2d15f91..60ec7589900c 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5036,12 +5036,6 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry, goto out_unlock; } - err = btrfs_update_inode(trans, root, inode); - if (err) { - drop_inode = 1; - goto out_unlock; - } - /* * If the active LSM wants to access the inode during * d_instantiate it needs these. Smack checks to see -- cgit From 82b22ac8f651aebbf0149cbbbd5808f07753b09c Mon Sep 17 00:00:00 2001 From: Kusanagi Kouichi Date: Mon, 28 Jan 2013 11:33:31 +0000 Subject: Btrfs: Check CAP_DAC_READ_SEARCH for BTRFS_IOC_INO_PATHS CAP_DAC_READ_SEARCH overrides read and search permission check on file and directory. It seems fit for BTRFS_IOC_INO_PATHS. Signed-off-by: Kusanagi Kouichi Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 744aabf11ce5..1b554b47e814 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3290,7 +3290,7 @@ static long btrfs_ioctl_ino_to_path(struct btrfs_root *root, void __user *arg) struct inode_fs_paths *ipath = NULL; struct btrfs_path *path; - if (!capable(CAP_SYS_ADMIN)) + if (!capable(CAP_DAC_READ_SEARCH)) return -EPERM; path = btrfs_alloc_path(); -- cgit From 55e301fd57a6239ec14b91a1cf2e70b3dd135194 Mon Sep 17 00:00:00 2001 From: Filipe Brandenburger Date: Tue, 29 Jan 2013 06:04:50 +0000 Subject: Btrfs: move fs/btrfs/ioctl.h to include/uapi/linux/btrfs.h The header file will then be installed under /usr/include/linux so that userspace applications can refer to Btrfs ioctls by name and use the same structs used internally in the kernel. Signed-off-by: Filipe Brandenburger Signed-off-by: Josef Bacik --- fs/btrfs/backref.h | 2 +- fs/btrfs/ctree.h | 2 +- fs/btrfs/file.c | 2 +- fs/btrfs/inode.c | 2 +- fs/btrfs/ioctl.c | 2 +- fs/btrfs/ioctl.h | 502 ----------------------------------------------------- fs/btrfs/qgroup.c | 2 +- fs/btrfs/super.c | 2 +- fs/btrfs/volumes.h | 2 +- 9 files changed, 8 insertions(+), 510 deletions(-) delete mode 100644 fs/btrfs/ioctl.h (limited to 'fs') diff --git a/fs/btrfs/backref.h b/fs/btrfs/backref.h index d61feca79455..310a7f6d09b1 100644 --- a/fs/btrfs/backref.h +++ b/fs/btrfs/backref.h @@ -19,7 +19,7 @@ #ifndef __BTRFS_BACKREF__ #define __BTRFS_BACKREF__ -#include "ioctl.h" +#include #include "ulist.h" #include "extent_io.h" diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 541ce9a9949e..69321013683c 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -31,10 +31,10 @@ #include #include #include +#include #include "extent_io.h" #include "extent_map.h" #include "async-thread.h" -#include "ioctl.h" struct btrfs_trans_handle; struct btrfs_transaction; diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 083abca56055..13c78ea3ebce 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -30,11 +30,11 @@ #include #include #include +#include #include "ctree.h" #include "disk-io.h" #include "transaction.h" #include "btrfs_inode.h" -#include "ioctl.h" #include "print-tree.h" #include "tree-log.h" #include "locking.h" diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 60ec7589900c..fc8aa8bf80a1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -39,12 +39,12 @@ #include #include #include +#include #include "compat.h" #include "ctree.h" #include "disk-io.h" #include "transaction.h" #include "btrfs_inode.h" -#include "ioctl.h" #include "print-tree.h" #include "ordered-data.h" #include "xattr.h" diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 1b554b47e814..96ecefc1724f 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -42,12 +42,12 @@ #include #include #include +#include #include "compat.h" #include "ctree.h" #include "disk-io.h" #include "transaction.h" #include "btrfs_inode.h" -#include "ioctl.h" #include "print-tree.h" #include "volumes.h" #include "locking.h" diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h deleted file mode 100644 index dabca9cc8c2e..000000000000 --- a/fs/btrfs/ioctl.h +++ /dev/null @@ -1,502 +0,0 @@ -/* - * Copyright (C) 2007 Oracle. All rights reserved. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public - * License v2 as published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * General Public License for more details. - * - * You should have received a copy of the GNU General Public - * License along with this program; if not, write to the - * Free Software Foundation, Inc., 59 Temple Place - Suite 330, - * Boston, MA 021110-1307, USA. - */ - -#ifndef __IOCTL_ -#define __IOCTL_ -#include - -#define BTRFS_IOCTL_MAGIC 0x94 -#define BTRFS_VOL_NAME_MAX 255 - -/* this should be 4k */ -#define BTRFS_PATH_NAME_MAX 4087 -struct btrfs_ioctl_vol_args { - __s64 fd; - char name[BTRFS_PATH_NAME_MAX + 1]; -}; - -#define BTRFS_DEVICE_PATH_NAME_MAX 1024 - -#define BTRFS_SUBVOL_CREATE_ASYNC (1ULL << 0) -#define BTRFS_SUBVOL_RDONLY (1ULL << 1) -#define BTRFS_SUBVOL_QGROUP_INHERIT (1ULL << 2) -#define BTRFS_FSID_SIZE 16 -#define BTRFS_UUID_SIZE 16 - -#define BTRFS_QGROUP_INHERIT_SET_LIMITS (1ULL << 0) - -struct btrfs_qgroup_limit { - __u64 flags; - __u64 max_rfer; - __u64 max_excl; - __u64 rsv_rfer; - __u64 rsv_excl; -}; - -struct btrfs_qgroup_inherit { - __u64 flags; - __u64 num_qgroups; - __u64 num_ref_copies; - __u64 num_excl_copies; - struct btrfs_qgroup_limit lim; - __u64 qgroups[0]; -}; - -struct btrfs_ioctl_qgroup_limit_args { - __u64 qgroupid; - struct btrfs_qgroup_limit lim; -}; - -#define BTRFS_SUBVOL_NAME_MAX 4039 -struct btrfs_ioctl_vol_args_v2 { - __s64 fd; - __u64 transid; - __u64 flags; - union { - struct { - __u64 size; - struct btrfs_qgroup_inherit __user *qgroup_inherit; - }; - __u64 unused[4]; - }; - char name[BTRFS_SUBVOL_NAME_MAX + 1]; -}; - -/* - * structure to report errors and progress to userspace, either as a - * result of a finished scrub, a canceled scrub or a progress inquiry - */ -struct btrfs_scrub_progress { - __u64 data_extents_scrubbed; /* # of data extents scrubbed */ - __u64 tree_extents_scrubbed; /* # of tree extents scrubbed */ - __u64 data_bytes_scrubbed; /* # of data bytes scrubbed */ - __u64 tree_bytes_scrubbed; /* # of tree bytes scrubbed */ - __u64 read_errors; /* # of read errors encountered (EIO) */ - __u64 csum_errors; /* # of failed csum checks */ - __u64 verify_errors; /* # of occurences, where the metadata - * of a tree block did not match the - * expected values, like generation or - * logical */ - __u64 no_csum; /* # of 4k data block for which no csum - * is present, probably the result of - * data written with nodatasum */ - __u64 csum_discards; /* # of csum for which no data was found - * in the extent tree. */ - __u64 super_errors; /* # of bad super blocks encountered */ - __u64 malloc_errors; /* # of internal kmalloc errors. These - * will likely cause an incomplete - * scrub */ - __u64 uncorrectable_errors; /* # of errors where either no intact - * copy was found or the writeback - * failed */ - __u64 corrected_errors; /* # of errors corrected */ - __u64 last_physical; /* last physical address scrubbed. In - * case a scrub was aborted, this can - * be used to restart the scrub */ - __u64 unverified_errors; /* # of occurences where a read for a - * full (64k) bio failed, but the re- - * check succeeded for each 4k piece. - * Intermittent error. */ -}; - -#define BTRFS_SCRUB_READONLY 1 -struct btrfs_ioctl_scrub_args { - __u64 devid; /* in */ - __u64 start; /* in */ - __u64 end; /* in */ - __u64 flags; /* in */ - struct btrfs_scrub_progress progress; /* out */ - /* pad to 1k */ - __u64 unused[(1024-32-sizeof(struct btrfs_scrub_progress))/8]; -}; - -#define BTRFS_IOCTL_DEV_REPLACE_CONT_READING_FROM_SRCDEV_MODE_ALWAYS 0 -#define BTRFS_IOCTL_DEV_REPLACE_CONT_READING_FROM_SRCDEV_MODE_AVOID 1 -struct btrfs_ioctl_dev_replace_start_params { - __u64 srcdevid; /* in, if 0, use srcdev_name instead */ - __u64 cont_reading_from_srcdev_mode; /* in, see #define - * above */ - __u8 srcdev_name[BTRFS_DEVICE_PATH_NAME_MAX + 1]; /* in */ - __u8 tgtdev_name[BTRFS_DEVICE_PATH_NAME_MAX + 1]; /* in */ -}; - -#define BTRFS_IOCTL_DEV_REPLACE_STATE_NEVER_STARTED 0 -#define BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED 1 -#define BTRFS_IOCTL_DEV_REPLACE_STATE_FINISHED 2 -#define BTRFS_IOCTL_DEV_REPLACE_STATE_CANCELED 3 -#define BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED 4 -struct btrfs_ioctl_dev_replace_status_params { - __u64 replace_state; /* out, see #define above */ - __u64 progress_1000; /* out, 0 <= x <= 1000 */ - __u64 time_started; /* out, seconds since 1-Jan-1970 */ - __u64 time_stopped; /* out, seconds since 1-Jan-1970 */ - __u64 num_write_errors; /* out */ - __u64 num_uncorrectable_read_errors; /* out */ -}; - -#define BTRFS_IOCTL_DEV_REPLACE_CMD_START 0 -#define BTRFS_IOCTL_DEV_REPLACE_CMD_STATUS 1 -#define BTRFS_IOCTL_DEV_REPLACE_CMD_CANCEL 2 -#define BTRFS_IOCTL_DEV_REPLACE_RESULT_NO_ERROR 0 -#define BTRFS_IOCTL_DEV_REPLACE_RESULT_NOT_STARTED 1 -#define BTRFS_IOCTL_DEV_REPLACE_RESULT_ALREADY_STARTED 2 -struct btrfs_ioctl_dev_replace_args { - __u64 cmd; /* in */ - __u64 result; /* out */ - - union { - struct btrfs_ioctl_dev_replace_start_params start; - struct btrfs_ioctl_dev_replace_status_params status; - }; /* in/out */ - - __u64 spare[64]; -}; - -struct btrfs_ioctl_dev_info_args { - __u64 devid; /* in/out */ - __u8 uuid[BTRFS_UUID_SIZE]; /* in/out */ - __u64 bytes_used; /* out */ - __u64 total_bytes; /* out */ - __u64 unused[379]; /* pad to 4k */ - __u8 path[BTRFS_DEVICE_PATH_NAME_MAX]; /* out */ -}; - -struct btrfs_ioctl_fs_info_args { - __u64 max_id; /* out */ - __u64 num_devices; /* out */ - __u8 fsid[BTRFS_FSID_SIZE]; /* out */ - __u64 reserved[124]; /* pad to 1k */ -}; - -/* balance control ioctl modes */ -#define BTRFS_BALANCE_CTL_PAUSE 1 -#define BTRFS_BALANCE_CTL_CANCEL 2 - -/* - * this is packed, because it should be exactly the same as its disk - * byte order counterpart (struct btrfs_disk_balance_args) - */ -struct btrfs_balance_args { - __u64 profiles; - __u64 usage; - __u64 devid; - __u64 pstart; - __u64 pend; - __u64 vstart; - __u64 vend; - - __u64 target; - - __u64 flags; - - __u64 unused[8]; -} __attribute__ ((__packed__)); - -/* report balance progress to userspace */ -struct btrfs_balance_progress { - __u64 expected; /* estimated # of chunks that will be - * relocated to fulfill the request */ - __u64 considered; /* # of chunks we have considered so far */ - __u64 completed; /* # of chunks relocated so far */ -}; - -#define BTRFS_BALANCE_STATE_RUNNING (1ULL << 0) -#define BTRFS_BALANCE_STATE_PAUSE_REQ (1ULL << 1) -#define BTRFS_BALANCE_STATE_CANCEL_REQ (1ULL << 2) - -struct btrfs_ioctl_balance_args { - __u64 flags; /* in/out */ - __u64 state; /* out */ - - struct btrfs_balance_args data; /* in/out */ - struct btrfs_balance_args meta; /* in/out */ - struct btrfs_balance_args sys; /* in/out */ - - struct btrfs_balance_progress stat; /* out */ - - __u64 unused[72]; /* pad to 1k */ -}; - -#define BTRFS_INO_LOOKUP_PATH_MAX 4080 -struct btrfs_ioctl_ino_lookup_args { - __u64 treeid; - __u64 objectid; - char name[BTRFS_INO_LOOKUP_PATH_MAX]; -}; - -struct btrfs_ioctl_search_key { - /* which root are we searching. 0 is the tree of tree roots */ - __u64 tree_id; - - /* keys returned will be >= min and <= max */ - __u64 min_objectid; - __u64 max_objectid; - - /* keys returned will be >= min and <= max */ - __u64 min_offset; - __u64 max_offset; - - /* max and min transids to search for */ - __u64 min_transid; - __u64 max_transid; - - /* keys returned will be >= min and <= max */ - __u32 min_type; - __u32 max_type; - - /* - * how many items did userland ask for, and how many are we - * returning - */ - __u32 nr_items; - - /* align to 64 bits */ - __u32 unused; - - /* some extra for later */ - __u64 unused1; - __u64 unused2; - __u64 unused3; - __u64 unused4; -}; - -struct btrfs_ioctl_search_header { - __u64 transid; - __u64 objectid; - __u64 offset; - __u32 type; - __u32 len; -}; - -#define BTRFS_SEARCH_ARGS_BUFSIZE (4096 - sizeof(struct btrfs_ioctl_search_key)) -/* - * the buf is an array of search headers where - * each header is followed by the actual item - * the type field is expanded to 32 bits for alignment - */ -struct btrfs_ioctl_search_args { - struct btrfs_ioctl_search_key key; - char buf[BTRFS_SEARCH_ARGS_BUFSIZE]; -}; - -struct btrfs_ioctl_clone_range_args { - __s64 src_fd; - __u64 src_offset, src_length; - __u64 dest_offset; -}; - -/* flags for the defrag range ioctl */ -#define BTRFS_DEFRAG_RANGE_COMPRESS 1 -#define BTRFS_DEFRAG_RANGE_START_IO 2 - -struct btrfs_ioctl_space_info { - __u64 flags; - __u64 total_bytes; - __u64 used_bytes; -}; - -struct btrfs_ioctl_space_args { - __u64 space_slots; - __u64 total_spaces; - struct btrfs_ioctl_space_info spaces[0]; -}; - -struct btrfs_data_container { - __u32 bytes_left; /* out -- bytes not needed to deliver output */ - __u32 bytes_missing; /* out -- additional bytes needed for result */ - __u32 elem_cnt; /* out */ - __u32 elem_missed; /* out */ - __u64 val[0]; /* out */ -}; - -struct btrfs_ioctl_ino_path_args { - __u64 inum; /* in */ - __u64 size; /* in */ - __u64 reserved[4]; - /* struct btrfs_data_container *fspath; out */ - __u64 fspath; /* out */ -}; - -struct btrfs_ioctl_logical_ino_args { - __u64 logical; /* in */ - __u64 size; /* in */ - __u64 reserved[4]; - /* struct btrfs_data_container *inodes; out */ - __u64 inodes; -}; - -enum btrfs_dev_stat_values { - /* disk I/O failure stats */ - BTRFS_DEV_STAT_WRITE_ERRS, /* EIO or EREMOTEIO from lower layers */ - BTRFS_DEV_STAT_READ_ERRS, /* EIO or EREMOTEIO from lower layers */ - BTRFS_DEV_STAT_FLUSH_ERRS, /* EIO or EREMOTEIO from lower layers */ - - /* stats for indirect indications for I/O failures */ - BTRFS_DEV_STAT_CORRUPTION_ERRS, /* checksum error, bytenr error or - * contents is illegal: this is an - * indication that the block was damaged - * during read or write, or written to - * wrong location or read from wrong - * location */ - BTRFS_DEV_STAT_GENERATION_ERRS, /* an indication that blocks have not - * been written */ - - BTRFS_DEV_STAT_VALUES_MAX -}; - -/* Reset statistics after reading; needs SYS_ADMIN capability */ -#define BTRFS_DEV_STATS_RESET (1ULL << 0) - -struct btrfs_ioctl_get_dev_stats { - __u64 devid; /* in */ - __u64 nr_items; /* in/out */ - __u64 flags; /* in/out */ - - /* out values: */ - __u64 values[BTRFS_DEV_STAT_VALUES_MAX]; - - __u64 unused[128 - 2 - BTRFS_DEV_STAT_VALUES_MAX]; /* pad to 1k */ -}; - -#define BTRFS_QUOTA_CTL_ENABLE 1 -#define BTRFS_QUOTA_CTL_DISABLE 2 -#define BTRFS_QUOTA_CTL_RESCAN 3 -struct btrfs_ioctl_quota_ctl_args { - __u64 cmd; - __u64 status; -}; - -struct btrfs_ioctl_qgroup_assign_args { - __u64 assign; - __u64 src; - __u64 dst; -}; - -struct btrfs_ioctl_qgroup_create_args { - __u64 create; - __u64 qgroupid; -}; -struct btrfs_ioctl_timespec { - __u64 sec; - __u32 nsec; -}; - -struct btrfs_ioctl_received_subvol_args { - char uuid[BTRFS_UUID_SIZE]; /* in */ - __u64 stransid; /* in */ - __u64 rtransid; /* out */ - struct btrfs_ioctl_timespec stime; /* in */ - struct btrfs_ioctl_timespec rtime; /* out */ - __u64 flags; /* in */ - __u64 reserved[16]; /* in */ -}; - -struct btrfs_ioctl_send_args { - __s64 send_fd; /* in */ - __u64 clone_sources_count; /* in */ - __u64 __user *clone_sources; /* in */ - __u64 parent_root; /* in */ - __u64 flags; /* in */ - __u64 reserved[4]; /* in */ -}; - -#define BTRFS_IOC_SNAP_CREATE _IOW(BTRFS_IOCTL_MAGIC, 1, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_DEFRAG _IOW(BTRFS_IOCTL_MAGIC, 2, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_RESIZE _IOW(BTRFS_IOCTL_MAGIC, 3, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_SCAN_DEV _IOW(BTRFS_IOCTL_MAGIC, 4, \ - struct btrfs_ioctl_vol_args) -/* trans start and trans end are dangerous, and only for - * use by applications that know how to avoid the - * resulting deadlocks - */ -#define BTRFS_IOC_TRANS_START _IO(BTRFS_IOCTL_MAGIC, 6) -#define BTRFS_IOC_TRANS_END _IO(BTRFS_IOCTL_MAGIC, 7) -#define BTRFS_IOC_SYNC _IO(BTRFS_IOCTL_MAGIC, 8) - -#define BTRFS_IOC_CLONE _IOW(BTRFS_IOCTL_MAGIC, 9, int) -#define BTRFS_IOC_ADD_DEV _IOW(BTRFS_IOCTL_MAGIC, 10, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_RM_DEV _IOW(BTRFS_IOCTL_MAGIC, 11, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_BALANCE _IOW(BTRFS_IOCTL_MAGIC, 12, \ - struct btrfs_ioctl_vol_args) - -#define BTRFS_IOC_CLONE_RANGE _IOW(BTRFS_IOCTL_MAGIC, 13, \ - struct btrfs_ioctl_clone_range_args) - -#define BTRFS_IOC_SUBVOL_CREATE _IOW(BTRFS_IOCTL_MAGIC, 14, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_SNAP_DESTROY _IOW(BTRFS_IOCTL_MAGIC, 15, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_DEFRAG_RANGE _IOW(BTRFS_IOCTL_MAGIC, 16, \ - struct btrfs_ioctl_defrag_range_args) -#define BTRFS_IOC_TREE_SEARCH _IOWR(BTRFS_IOCTL_MAGIC, 17, \ - struct btrfs_ioctl_search_args) -#define BTRFS_IOC_INO_LOOKUP _IOWR(BTRFS_IOCTL_MAGIC, 18, \ - struct btrfs_ioctl_ino_lookup_args) -#define BTRFS_IOC_DEFAULT_SUBVOL _IOW(BTRFS_IOCTL_MAGIC, 19, u64) -#define BTRFS_IOC_SPACE_INFO _IOWR(BTRFS_IOCTL_MAGIC, 20, \ - struct btrfs_ioctl_space_args) -#define BTRFS_IOC_START_SYNC _IOR(BTRFS_IOCTL_MAGIC, 24, __u64) -#define BTRFS_IOC_WAIT_SYNC _IOW(BTRFS_IOCTL_MAGIC, 22, __u64) -#define BTRFS_IOC_SNAP_CREATE_V2 _IOW(BTRFS_IOCTL_MAGIC, 23, \ - struct btrfs_ioctl_vol_args_v2) -#define BTRFS_IOC_SUBVOL_CREATE_V2 _IOW(BTRFS_IOCTL_MAGIC, 24, \ - struct btrfs_ioctl_vol_args_v2) -#define BTRFS_IOC_SUBVOL_GETFLAGS _IOR(BTRFS_IOCTL_MAGIC, 25, __u64) -#define BTRFS_IOC_SUBVOL_SETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 26, __u64) -#define BTRFS_IOC_SCRUB _IOWR(BTRFS_IOCTL_MAGIC, 27, \ - struct btrfs_ioctl_scrub_args) -#define BTRFS_IOC_SCRUB_CANCEL _IO(BTRFS_IOCTL_MAGIC, 28) -#define BTRFS_IOC_SCRUB_PROGRESS _IOWR(BTRFS_IOCTL_MAGIC, 29, \ - struct btrfs_ioctl_scrub_args) -#define BTRFS_IOC_DEV_INFO _IOWR(BTRFS_IOCTL_MAGIC, 30, \ - struct btrfs_ioctl_dev_info_args) -#define BTRFS_IOC_FS_INFO _IOR(BTRFS_IOCTL_MAGIC, 31, \ - struct btrfs_ioctl_fs_info_args) -#define BTRFS_IOC_BALANCE_V2 _IOWR(BTRFS_IOCTL_MAGIC, 32, \ - struct btrfs_ioctl_balance_args) -#define BTRFS_IOC_BALANCE_CTL _IOW(BTRFS_IOCTL_MAGIC, 33, int) -#define BTRFS_IOC_BALANCE_PROGRESS _IOR(BTRFS_IOCTL_MAGIC, 34, \ - struct btrfs_ioctl_balance_args) -#define BTRFS_IOC_INO_PATHS _IOWR(BTRFS_IOCTL_MAGIC, 35, \ - struct btrfs_ioctl_ino_path_args) -#define BTRFS_IOC_LOGICAL_INO _IOWR(BTRFS_IOCTL_MAGIC, 36, \ - struct btrfs_ioctl_ino_path_args) -#define BTRFS_IOC_SET_RECEIVED_SUBVOL _IOWR(BTRFS_IOCTL_MAGIC, 37, \ - struct btrfs_ioctl_received_subvol_args) -#define BTRFS_IOC_SEND _IOW(BTRFS_IOCTL_MAGIC, 38, struct btrfs_ioctl_send_args) -#define BTRFS_IOC_DEVICES_READY _IOR(BTRFS_IOCTL_MAGIC, 39, \ - struct btrfs_ioctl_vol_args) -#define BTRFS_IOC_QUOTA_CTL _IOWR(BTRFS_IOCTL_MAGIC, 40, \ - struct btrfs_ioctl_quota_ctl_args) -#define BTRFS_IOC_QGROUP_ASSIGN _IOW(BTRFS_IOCTL_MAGIC, 41, \ - struct btrfs_ioctl_qgroup_assign_args) -#define BTRFS_IOC_QGROUP_CREATE _IOW(BTRFS_IOCTL_MAGIC, 42, \ - struct btrfs_ioctl_qgroup_create_args) -#define BTRFS_IOC_QGROUP_LIMIT _IOR(BTRFS_IOCTL_MAGIC, 43, \ - struct btrfs_ioctl_qgroup_limit_args) -#define BTRFS_IOC_GET_DEV_STATS _IOWR(BTRFS_IOCTL_MAGIC, 52, \ - struct btrfs_ioctl_get_dev_stats) -#define BTRFS_IOC_DEV_REPLACE _IOWR(BTRFS_IOCTL_MAGIC, 53, \ - struct btrfs_ioctl_dev_replace_args) - -#endif diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index a5c856234323..a0d6368249fa 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -23,13 +23,13 @@ #include #include #include +#include #include "ctree.h" #include "transaction.h" #include "disk-io.h" #include "locking.h" #include "ulist.h" -#include "ioctl.h" #include "backref.h" /* TODO XXX FIXME diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 67b373bf3ff9..6846ededfe95 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -41,13 +41,13 @@ #include #include #include +#include #include "compat.h" #include "delayed-inode.h" #include "ctree.h" #include "disk-io.h" #include "transaction.h" #include "btrfs_inode.h" -#include "ioctl.h" #include "print-tree.h" #include "xattr.h" #include "volumes.h" diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index d3c3939ac751..12bb84166a5f 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -21,8 +21,8 @@ #include #include +#include #include "async-thread.h" -#include "ioctl.h" #define BTRFS_STRIPE_LEN (64 * 1024) -- cgit From 8c6a3ee6dbd564d05019d396ecb5dc5b35cbc273 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 29 Jan 2013 10:05:05 +0000 Subject: Btrfs: add a comment for fs_info->max_inline Though ->max_inline is a 64bit variant, and may be accessed by multi-task, but it is just suggestive number, so we needn't add anything to protect fs_info->max_inline, just add a comment to explain wny we don't use a lock to protect it. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 69321013683c..169baa817d96 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1297,6 +1297,12 @@ struct btrfs_fs_info { u64 last_trans_log_full_commit; unsigned long mount_opt; unsigned long compress_type:4; + /* + * It is a suggestive number, the read side is safe even it gets a + * wrong number because we will write out the data into a regular + * extent. The write side(mount/remount) is under ->s_umount lock, + * so it is also safe. + */ u64 max_inline; u64 alloc_start; struct btrfs_transaction *running_transaction; -- cgit From c018daecead7a46a575e2a1397fea850b83396c8 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 29 Jan 2013 10:07:33 +0000 Subject: Btrfs: protect fs_info->alloc_start fs_info->alloc_start is a 64bits variant, can be accessed by multi-task, but it is not protected strictly, it can be changed while we are accessing it. On 32bit machine, we will get wrong value because we access it by two instructions.(In fact, it is also possible that the same problem happens on the 64bit machine, because the compiler may split the 64bit operation into two 32bit operation.) For example: Assuming -> alloc_start is 0x0000 0000 0001 0000 at the beginning, then we remount and set ->alloc_start to 0x0000 0100 0000 0000. Task0 Task1 load high 32 bits set high 32 bits set low 32 bits load low 32 bits Task1 will get 0. This patch fixes this problem by using two locks to protect it fs_info->chunk_mutex sb->s_umount On the read side, we just need get one of these two locks, and on the write side, we must lock all of them. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 10 ++++++++++ fs/btrfs/super.c | 4 ++++ 2 files changed, 14 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 169baa817d96..d8e539fe5544 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1304,6 +1304,16 @@ struct btrfs_fs_info { * so it is also safe. */ u64 max_inline; + /* + * Protected by ->chunk_mutex and sb->s_umount. + * + * The reason that we use two lock to protect it is because only + * remount and mount operations can change it and these two operations + * are under sb->s_umount, but the read side (chunk allocation) can not + * acquire sb->s_umount or the deadlock would happen. So we use two + * locks to protect it. On the write side, we must acquire two locks, + * and on the read side, we just need acquire one of them. + */ u64 alloc_start; struct btrfs_transaction *running_transaction; wait_queue_head_t transaction_throttle; diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 6846ededfe95..eda330df45a4 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -519,7 +519,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) case Opt_alloc_start: num = match_strdup(&args[0]); if (num) { + mutex_lock(&info->chunk_mutex); info->alloc_start = memparse(num, NULL); + mutex_unlock(&info->chunk_mutex); kfree(num); printk(KERN_INFO "btrfs: allocations start at %llu\n", @@ -1289,7 +1291,9 @@ restore: fs_info->mount_opt = old_opts; fs_info->compress_type = old_compress_type; fs_info->max_inline = old_max_inline; + mutex_lock(&fs_info->chunk_mutex); fs_info->alloc_start = old_alloc_start; + mutex_unlock(&fs_info->chunk_mutex); btrfs_resize_thread_pool(fs_info, old_thread_pool_size, fs_info->thread_pool_size); fs_info->metadata_ratio = old_metadata_ratio; -- cgit From e2d845211eda9cf296e8edf6724b3d541f4fbfd5 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 29 Jan 2013 10:09:20 +0000 Subject: Btrfs: use percpu counter for dirty metadata count ->dirty_metadata_bytes is accessed very frequently, so use percpu counter instead of the u64 variant to reduce the contention of the lock. This patch also fixed the problem that we access it without lock protection in __btrfs_btree_balance_dirty(), which may cause we skip the dirty pages flush. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 9 ++++---- fs/btrfs/disk-io.c | 64 ++++++++++++++++++++++++++++------------------------ fs/btrfs/extent_io.c | 9 +++----- 3 files changed, 42 insertions(+), 40 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index d8e539fe5544..4c476281b66b 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -191,6 +191,8 @@ static int btrfs_csum_sizes[] = { 4, 0 }; /* ioprio of readahead is set to idle */ #define BTRFS_IOPRIO_READA (IOPRIO_PRIO_VALUE(IOPRIO_CLASS_IDLE, 0)) +#define BTRFS_DIRTY_METADATA_THRESH (32 * 1024 * 1024) + /* * The key defines the order in the tree, and so it also defines (optimal) * block layout. @@ -1448,10 +1450,9 @@ struct btrfs_fs_info { u64 total_pinned; - /* protected by the delalloc lock, used to keep from writing - * metadata until there is a nice batch - */ - u64 dirty_metadata_bytes; + /* used to keep from writing metadata until there is a nice batch */ + struct percpu_counter dirty_metadata_bytes; + s32 dirty_metadata_batch; struct list_head dirty_cowonly_roots; struct btrfs_fs_devices *fs_devices; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 04f98e3ffd90..34ace168eebc 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -946,18 +946,20 @@ static int btree_writepages(struct address_space *mapping, struct writeback_control *wbc) { struct extent_io_tree *tree; + struct btrfs_fs_info *fs_info; + int ret; + tree = &BTRFS_I(mapping->host)->io_tree; if (wbc->sync_mode == WB_SYNC_NONE) { - struct btrfs_root *root = BTRFS_I(mapping->host)->root; - u64 num_dirty; - unsigned long thresh = 32 * 1024 * 1024; if (wbc->for_kupdate) return 0; + fs_info = BTRFS_I(mapping->host)->root->fs_info; /* this is a bit racy, but that's ok */ - num_dirty = root->fs_info->dirty_metadata_bytes; - if (num_dirty < thresh) + ret = percpu_counter_compare(&fs_info->dirty_metadata_bytes, + BTRFS_DIRTY_METADATA_THRESH); + if (ret < 0) return 0; } return btree_write_cache_pages(mapping, wbc); @@ -1125,24 +1127,16 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr, void clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf) { + struct btrfs_fs_info *fs_info = root->fs_info; + if (btrfs_header_generation(buf) == - root->fs_info->running_transaction->transid) { + fs_info->running_transaction->transid) { btrfs_assert_tree_locked(buf); if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &buf->bflags)) { - spin_lock(&root->fs_info->delalloc_lock); - if (root->fs_info->dirty_metadata_bytes >= buf->len) - root->fs_info->dirty_metadata_bytes -= buf->len; - else { - spin_unlock(&root->fs_info->delalloc_lock); - btrfs_panic(root->fs_info, -EOVERFLOW, - "Can't clear %lu bytes from " - " dirty_mdatadata_bytes (%llu)", - buf->len, - root->fs_info->dirty_metadata_bytes); - } - spin_unlock(&root->fs_info->delalloc_lock); - + __percpu_counter_add(&fs_info->dirty_metadata_bytes, + -buf->len, + fs_info->dirty_metadata_batch); /* ugh, clear_extent_buffer_dirty needs to lock the page */ btrfs_set_lock_blocking(buf); clear_extent_buffer_dirty(buf); @@ -2008,10 +2002,18 @@ int open_ctree(struct super_block *sb, goto fail_srcu; } + ret = percpu_counter_init(&fs_info->dirty_metadata_bytes, 0); + if (ret) { + err = ret; + goto fail_bdi; + } + fs_info->dirty_metadata_batch = PAGE_CACHE_SIZE * + (1 + ilog2(nr_cpu_ids)); + fs_info->btree_inode = new_inode(sb); if (!fs_info->btree_inode) { err = -ENOMEM; - goto fail_bdi; + goto fail_dirty_metadata_bytes; } mapping_set_gfp_mask(fs_info->btree_inode->i_mapping, GFP_NOFS); @@ -2266,6 +2268,7 @@ int open_ctree(struct super_block *sb, leafsize = btrfs_super_leafsize(disk_super); sectorsize = btrfs_super_sectorsize(disk_super); stripesize = btrfs_super_stripesize(disk_super); + fs_info->dirty_metadata_batch = leafsize * (1 + ilog2(nr_cpu_ids)); /* * mixed block groups end up with duplicate but slightly offset @@ -2728,6 +2731,8 @@ fail_iput: invalidate_inode_pages2(fs_info->btree_inode->i_mapping); iput(fs_info->btree_inode); +fail_dirty_metadata_bytes: + percpu_counter_destroy(&fs_info->dirty_metadata_bytes); fail_bdi: bdi_destroy(&fs_info->bdi); fail_srcu: @@ -3406,6 +3411,7 @@ int close_ctree(struct btrfs_root *root) btrfs_close_devices(fs_info->fs_devices); btrfs_mapping_tree_free(&fs_info->mapping_tree); + percpu_counter_destroy(&fs_info->dirty_metadata_bytes); bdi_destroy(&fs_info->bdi); cleanup_srcu_struct(&fs_info->subvol_srcu); @@ -3448,11 +3454,10 @@ void btrfs_mark_buffer_dirty(struct extent_buffer *buf) (unsigned long long)transid, (unsigned long long)root->fs_info->generation); was_dirty = set_extent_buffer_dirty(buf); - if (!was_dirty) { - spin_lock(&root->fs_info->delalloc_lock); - root->fs_info->dirty_metadata_bytes += buf->len; - spin_unlock(&root->fs_info->delalloc_lock); - } + if (!was_dirty) + __percpu_counter_add(&root->fs_info->dirty_metadata_bytes, + buf->len, + root->fs_info->dirty_metadata_batch); } static void __btrfs_btree_balance_dirty(struct btrfs_root *root, @@ -3462,8 +3467,7 @@ static void __btrfs_btree_balance_dirty(struct btrfs_root *root, * looks as though older kernels can get into trouble with * this code, they end up stuck in balance_dirty_pages forever */ - u64 num_dirty; - unsigned long thresh = 32 * 1024 * 1024; + int ret; if (current->flags & PF_MEMALLOC) return; @@ -3471,9 +3475,9 @@ static void __btrfs_btree_balance_dirty(struct btrfs_root *root, if (flush_delayed) btrfs_balance_delayed_items(root); - num_dirty = root->fs_info->dirty_metadata_bytes; - - if (num_dirty > thresh) { + ret = percpu_counter_compare(&root->fs_info->dirty_metadata_bytes, + BTRFS_DIRTY_METADATA_THRESH); + if (ret > 0) { balance_dirty_pages_ratelimited_nr( root->fs_info->btree_inode->i_mapping, 1); } diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index bda36fef301e..b9019c847c58 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3122,12 +3122,9 @@ static int lock_extent_buffer_for_io(struct extent_buffer *eb, set_bit(EXTENT_BUFFER_WRITEBACK, &eb->bflags); spin_unlock(&eb->refs_lock); btrfs_set_header_flag(eb, BTRFS_HEADER_FLAG_WRITTEN); - spin_lock(&fs_info->delalloc_lock); - if (fs_info->dirty_metadata_bytes >= eb->len) - fs_info->dirty_metadata_bytes -= eb->len; - else - WARN_ON(1); - spin_unlock(&fs_info->delalloc_lock); + __percpu_counter_add(&fs_info->dirty_metadata_bytes, + -eb->len, + fs_info->dirty_metadata_batch); ret = 1; } else { spin_unlock(&eb->refs_lock); -- cgit From 963d678b0f7649300e3a67f2513ca9d830c6e303 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 29 Jan 2013 10:10:51 +0000 Subject: Btrfs: use percpu counter for fs_info->delalloc_bytes fs_info->delalloc_bytes is accessed very frequently, so use percpu counter instead of the u64 variant for it to reduce the lock contention. This patch also fixed the problem that we access the variant without the lock protection.At worst, we would not flush the delalloc inodes, and just return ENOSPC error when we still have some free space in the fs. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 7 ++++--- fs/btrfs/disk-io.c | 18 ++++++++++++++---- fs/btrfs/extent-tree.c | 6 ++++-- fs/btrfs/inode.c | 6 ++++-- 4 files changed, 26 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 4c476281b66b..50def99f5379 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1392,6 +1392,7 @@ struct btrfs_fs_info { */ struct list_head ordered_extents; + spinlock_t delalloc_lock; /* * all of the inodes that have delalloc bytes. It is possible for * this list to be empty even when there is still dirty data=ordered @@ -1452,7 +1453,10 @@ struct btrfs_fs_info { /* used to keep from writing metadata until there is a nice batch */ struct percpu_counter dirty_metadata_bytes; + struct percpu_counter delalloc_bytes; s32 dirty_metadata_batch; + s32 delalloc_batch; + struct list_head dirty_cowonly_roots; struct btrfs_fs_devices *fs_devices; @@ -1468,9 +1472,6 @@ struct btrfs_fs_info { struct reloc_control *reloc_ctl; - spinlock_t delalloc_lock; - u64 delalloc_bytes; - /* data_alloc_cluster is only used in ssd mode */ struct btrfs_free_cluster data_alloc_cluster; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 34ace168eebc..2c9498aefe86 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2010,10 +2010,16 @@ int open_ctree(struct super_block *sb, fs_info->dirty_metadata_batch = PAGE_CACHE_SIZE * (1 + ilog2(nr_cpu_ids)); + ret = percpu_counter_init(&fs_info->delalloc_bytes, 0); + if (ret) { + err = ret; + goto fail_dirty_metadata_bytes; + } + fs_info->btree_inode = new_inode(sb); if (!fs_info->btree_inode) { err = -ENOMEM; - goto fail_dirty_metadata_bytes; + goto fail_delalloc_bytes; } mapping_set_gfp_mask(fs_info->btree_inode->i_mapping, GFP_NOFS); @@ -2269,6 +2275,7 @@ int open_ctree(struct super_block *sb, sectorsize = btrfs_super_sectorsize(disk_super); stripesize = btrfs_super_stripesize(disk_super); fs_info->dirty_metadata_batch = leafsize * (1 + ilog2(nr_cpu_ids)); + fs_info->delalloc_batch = sectorsize * 512 * (1 + ilog2(nr_cpu_ids)); /* * mixed block groups end up with duplicate but slightly offset @@ -2731,6 +2738,8 @@ fail_iput: invalidate_inode_pages2(fs_info->btree_inode->i_mapping); iput(fs_info->btree_inode); +fail_delalloc_bytes: + percpu_counter_destroy(&fs_info->delalloc_bytes); fail_dirty_metadata_bytes: percpu_counter_destroy(&fs_info->dirty_metadata_bytes); fail_bdi: @@ -3362,9 +3371,9 @@ int close_ctree(struct btrfs_root *root) btrfs_free_qgroup_config(root->fs_info); - if (fs_info->delalloc_bytes) { - printk(KERN_INFO "btrfs: at unmount delalloc count %llu\n", - (unsigned long long)fs_info->delalloc_bytes); + if (percpu_counter_sum(&fs_info->delalloc_bytes)) { + printk(KERN_INFO "btrfs: at unmount delalloc count %lld\n", + percpu_counter_sum(&fs_info->delalloc_bytes)); } free_extent_buffer(fs_info->extent_root->node); @@ -3412,6 +3421,7 @@ int close_ctree(struct btrfs_root *root) btrfs_mapping_tree_free(&fs_info->mapping_tree); percpu_counter_destroy(&fs_info->dirty_metadata_bytes); + percpu_counter_destroy(&fs_info->delalloc_bytes); bdi_destroy(&fs_info->bdi); cleanup_srcu_struct(&fs_info->subvol_srcu); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 174c4d5c692c..115d1646bf50 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3760,7 +3760,8 @@ static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig, space_info = block_rsv->space_info; smp_mb(); - delalloc_bytes = root->fs_info->delalloc_bytes; + delalloc_bytes = percpu_counter_sum_positive( + &root->fs_info->delalloc_bytes); if (delalloc_bytes == 0) { if (trans) return; @@ -3799,7 +3800,8 @@ static void shrink_delalloc(struct btrfs_root *root, u64 to_reclaim, u64 orig, break; } smp_mb(); - delalloc_bytes = root->fs_info->delalloc_bytes; + delalloc_bytes = percpu_counter_sum_positive( + &root->fs_info->delalloc_bytes); } } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index fc8aa8bf80a1..24c0b7805fe1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1516,7 +1516,8 @@ static void btrfs_set_bit_hook(struct inode *inode, spin_lock(&root->fs_info->delalloc_lock); BTRFS_I(inode)->delalloc_bytes += len; - root->fs_info->delalloc_bytes += len; + __percpu_counter_add(&root->fs_info->delalloc_bytes, len, + root->fs_info->delalloc_batch); if (do_list && list_empty(&BTRFS_I(inode)->delalloc_inodes)) { list_add_tail(&BTRFS_I(inode)->delalloc_inodes, &root->fs_info->delalloc_inodes); @@ -1557,7 +1558,8 @@ static void btrfs_clear_bit_hook(struct inode *inode, btrfs_free_reserved_data_space(inode, len); spin_lock(&root->fs_info->delalloc_lock); - root->fs_info->delalloc_bytes -= len; + __percpu_counter_add(&root->fs_info->delalloc_bytes, -len, + root->fs_info->delalloc_batch); BTRFS_I(inode)->delalloc_bytes -= len; if (do_list && BTRFS_I(inode)->delalloc_bytes == 0 && -- cgit From df0af1a57f72c74d53a9377c60ff20095afab97d Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 29 Jan 2013 10:11:59 +0000 Subject: Btrfs: use the inode own lock to protect its delalloc_bytes We need not use a global lock to protect the delalloc_bytes of the inode, just use its own lock. In this way, we can reduce the lock contention and ->delalloc_lock will just protect delalloc inode list. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/btrfs_inode.h | 1 + fs/btrfs/disk-io.c | 2 ++ fs/btrfs/inode.c | 47 ++++++++++++++++++++++++++++++++++------------- 3 files changed, 37 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index 2a8c242bc4f5..c935a774a9f8 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -40,6 +40,7 @@ #define BTRFS_INODE_HAS_ASYNC_EXTENT 6 #define BTRFS_INODE_NEEDS_FULL_SYNC 7 #define BTRFS_INODE_COPY_EVERYTHING 8 +#define BTRFS_INODE_IN_DELALLOC_LIST 9 /* in memory btrfs inode */ struct btrfs_inode { diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 2c9498aefe86..11f6dbcb1191 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3690,6 +3690,8 @@ static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root) delalloc_inodes); list_del_init(&btrfs_inode->delalloc_inodes); + clear_bit(BTRFS_INODE_IN_DELALLOC_LIST, + &btrfs_inode->runtime_flags); btrfs_invalidate_inodes(btrfs_inode->root); } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 24c0b7805fe1..82c7c66f8523 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1514,15 +1514,22 @@ static void btrfs_set_bit_hook(struct inode *inode, spin_unlock(&BTRFS_I(inode)->lock); } - spin_lock(&root->fs_info->delalloc_lock); - BTRFS_I(inode)->delalloc_bytes += len; __percpu_counter_add(&root->fs_info->delalloc_bytes, len, root->fs_info->delalloc_batch); - if (do_list && list_empty(&BTRFS_I(inode)->delalloc_inodes)) { - list_add_tail(&BTRFS_I(inode)->delalloc_inodes, - &root->fs_info->delalloc_inodes); + spin_lock(&BTRFS_I(inode)->lock); + BTRFS_I(inode)->delalloc_bytes += len; + if (do_list && !test_bit(BTRFS_INODE_IN_DELALLOC_LIST, + &BTRFS_I(inode)->runtime_flags)) { + spin_lock(&root->fs_info->delalloc_lock); + if (list_empty(&BTRFS_I(inode)->delalloc_inodes)) { + list_add_tail(&BTRFS_I(inode)->delalloc_inodes, + &root->fs_info->delalloc_inodes); + set_bit(BTRFS_INODE_IN_DELALLOC_LIST, + &BTRFS_I(inode)->runtime_flags); + } + spin_unlock(&root->fs_info->delalloc_lock); } - spin_unlock(&root->fs_info->delalloc_lock); + spin_unlock(&BTRFS_I(inode)->lock); } } @@ -1557,16 +1564,22 @@ static void btrfs_clear_bit_hook(struct inode *inode, && do_list) btrfs_free_reserved_data_space(inode, len); - spin_lock(&root->fs_info->delalloc_lock); __percpu_counter_add(&root->fs_info->delalloc_bytes, -len, root->fs_info->delalloc_batch); + spin_lock(&BTRFS_I(inode)->lock); BTRFS_I(inode)->delalloc_bytes -= len; - if (do_list && BTRFS_I(inode)->delalloc_bytes == 0 && - !list_empty(&BTRFS_I(inode)->delalloc_inodes)) { - list_del_init(&BTRFS_I(inode)->delalloc_inodes); + test_bit(BTRFS_INODE_IN_DELALLOC_LIST, + &BTRFS_I(inode)->runtime_flags)) { + spin_lock(&root->fs_info->delalloc_lock); + if (!list_empty(&BTRFS_I(inode)->delalloc_inodes)) { + list_del_init(&BTRFS_I(inode)->delalloc_inodes); + clear_bit(BTRFS_INODE_IN_DELALLOC_LIST, + &BTRFS_I(inode)->runtime_flags); + } + spin_unlock(&root->fs_info->delalloc_lock); } - spin_unlock(&root->fs_info->delalloc_lock); + spin_unlock(&BTRFS_I(inode)->lock); } } @@ -7326,14 +7339,19 @@ fail: static int btrfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) { + u64 delalloc_bytes; struct inode *inode = dentry->d_inode; u32 blocksize = inode->i_sb->s_blocksize; generic_fillattr(inode, stat); stat->dev = BTRFS_I(inode)->root->anon_dev; stat->blksize = PAGE_CACHE_SIZE; + + spin_lock(&BTRFS_I(inode)->lock); + delalloc_bytes = BTRFS_I(inode)->delalloc_bytes; + spin_unlock(&BTRFS_I(inode)->lock); stat->blocks = (ALIGN(inode_get_bytes(inode), blocksize) + - ALIGN(BTRFS_I(inode)->delalloc_bytes, blocksize)) >> 9; + ALIGN(delalloc_bytes, blocksize)) >> 9; return 0; } @@ -7620,8 +7638,11 @@ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput) list_del_init(&binode->delalloc_inodes); inode = igrab(&binode->vfs_inode); - if (!inode) + if (!inode) { + clear_bit(BTRFS_INODE_IN_DELALLOC_LIST, + &binode->runtime_flags); continue; + } list_add_tail(&binode->delalloc_inodes, &root->fs_info->delalloc_inodes); -- cgit From de98ced9e743656d108de41841797def0f5cb951 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 29 Jan 2013 10:13:12 +0000 Subject: Btrfs: use seqlock to protect fs_info->avail_{data, metadata, system}_alloc_bits There is no lock to protect fs_info->avail_{data, metadata, system}_alloc_bits, it may introduce some problem, such as the wrong profile information, so we add a seqlock to protect them. Signed-off-by: Zhao Lei Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 2 ++ fs/btrfs/disk-io.c | 1 + fs/btrfs/extent-tree.c | 22 ++++++++++++++------ fs/btrfs/volumes.c | 56 +++++++++++++++++++++++++++----------------------- 4 files changed, 49 insertions(+), 32 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 50def99f5379..5f5c30aaef36 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1483,6 +1483,8 @@ struct btrfs_fs_info { struct rb_root defrag_inodes; atomic_t defrag_running; + /* Used to protect avail_{data, metadata, system}_alloc_bits */ + seqlock_t profiles_lock; /* * these three are in extended format (availability of single * chunks is denoted by BTRFS_AVAIL_ALLOC_BIT_SINGLE bit, other diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 11f6dbcb1191..00b6742fdde7 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2040,6 +2040,7 @@ int open_ctree(struct super_block *sb, spin_lock_init(&fs_info->tree_mod_seq_lock); rwlock_init(&fs_info->tree_mod_log_lock); mutex_init(&fs_info->reloc_mutex); + seqlock_init(&fs_info->profiles_lock); init_completion(&fs_info->kobj_unregister); INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 115d1646bf50..faff98f720de 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3227,12 +3227,14 @@ static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags) u64 extra_flags = chunk_to_extended(flags) & BTRFS_EXTENDED_PROFILE_MASK; + write_seqlock(&fs_info->profiles_lock); if (flags & BTRFS_BLOCK_GROUP_DATA) fs_info->avail_data_alloc_bits |= extra_flags; if (flags & BTRFS_BLOCK_GROUP_METADATA) fs_info->avail_metadata_alloc_bits |= extra_flags; if (flags & BTRFS_BLOCK_GROUP_SYSTEM) fs_info->avail_system_alloc_bits |= extra_flags; + write_sequnlock(&fs_info->profiles_lock); } /* @@ -3324,12 +3326,18 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags) static u64 get_alloc_profile(struct btrfs_root *root, u64 flags) { - if (flags & BTRFS_BLOCK_GROUP_DATA) - flags |= root->fs_info->avail_data_alloc_bits; - else if (flags & BTRFS_BLOCK_GROUP_SYSTEM) - flags |= root->fs_info->avail_system_alloc_bits; - else if (flags & BTRFS_BLOCK_GROUP_METADATA) - flags |= root->fs_info->avail_metadata_alloc_bits; + unsigned seq; + + do { + seq = read_seqbegin(&root->fs_info->profiles_lock); + + if (flags & BTRFS_BLOCK_GROUP_DATA) + flags |= root->fs_info->avail_data_alloc_bits; + else if (flags & BTRFS_BLOCK_GROUP_SYSTEM) + flags |= root->fs_info->avail_system_alloc_bits; + else if (flags & BTRFS_BLOCK_GROUP_METADATA) + flags |= root->fs_info->avail_metadata_alloc_bits; + } while (read_seqretry(&root->fs_info->profiles_lock, seq)); return btrfs_reduce_alloc_profile(root, flags); } @@ -7967,12 +7975,14 @@ static void clear_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags) u64 extra_flags = chunk_to_extended(flags) & BTRFS_EXTENDED_PROFILE_MASK; + write_seqlock(&fs_info->profiles_lock); if (flags & BTRFS_BLOCK_GROUP_DATA) fs_info->avail_data_alloc_bits &= ~extra_flags; if (flags & BTRFS_BLOCK_GROUP_METADATA) fs_info->avail_metadata_alloc_bits &= ~extra_flags; if (flags & BTRFS_BLOCK_GROUP_SYSTEM) fs_info->avail_system_alloc_bits &= ~extra_flags; + write_sequnlock(&fs_info->profiles_lock); } int btrfs_remove_block_group(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 1ce581d58d87..8c9ea4cd66bb 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1430,14 +1430,19 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) u64 devid; u64 num_devices; u8 *dev_uuid; + unsigned seq; int ret = 0; bool clear_super = false; mutex_lock(&uuid_mutex); - all_avail = root->fs_info->avail_data_alloc_bits | - root->fs_info->avail_system_alloc_bits | - root->fs_info->avail_metadata_alloc_bits; + do { + seq = read_seqbegin(&root->fs_info->profiles_lock); + + all_avail = root->fs_info->avail_data_alloc_bits | + root->fs_info->avail_system_alloc_bits | + root->fs_info->avail_metadata_alloc_bits; + } while (read_seqretry(&root->fs_info->profiles_lock, seq)); num_devices = root->fs_info->fs_devices->num_devices; btrfs_dev_replace_lock(&root->fs_info->dev_replace); @@ -3043,6 +3048,7 @@ int btrfs_balance(struct btrfs_balance_control *bctl, int mixed = 0; int ret; u64 num_devices; + unsigned seq; if (btrfs_fs_closing(fs_info) || atomic_read(&fs_info->balance_pause_req) || @@ -3126,22 +3132,26 @@ int btrfs_balance(struct btrfs_balance_control *bctl, /* allow to reduce meta or sys integrity only if force set */ allowed = BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10; - if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) && - (fs_info->avail_system_alloc_bits & allowed) && - !(bctl->sys.target & allowed)) || - ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) && - (fs_info->avail_metadata_alloc_bits & allowed) && - !(bctl->meta.target & allowed))) { - if (bctl->flags & BTRFS_BALANCE_FORCE) { - printk(KERN_INFO "btrfs: force reducing metadata " - "integrity\n"); - } else { - printk(KERN_ERR "btrfs: balance will reduce metadata " - "integrity, use force if you want this\n"); - ret = -EINVAL; - goto out; + do { + seq = read_seqbegin(&fs_info->profiles_lock); + + if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) && + (fs_info->avail_system_alloc_bits & allowed) && + !(bctl->sys.target & allowed)) || + ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) && + (fs_info->avail_metadata_alloc_bits & allowed) && + !(bctl->meta.target & allowed))) { + if (bctl->flags & BTRFS_BALANCE_FORCE) { + printk(KERN_INFO "btrfs: force reducing metadata " + "integrity\n"); + } else { + printk(KERN_ERR "btrfs: balance will reduce metadata " + "integrity, use force if you want this\n"); + ret = -EINVAL; + goto out; + } } - } + } while (read_seqretry(&fs_info->profiles_lock, seq)); if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) { int num_tolerated_disk_barrier_failures; @@ -3980,10 +3990,7 @@ static noinline int init_first_rw_device(struct btrfs_trans_handle *trans, if (ret) return ret; - alloc_profile = BTRFS_BLOCK_GROUP_METADATA | - fs_info->avail_metadata_alloc_bits; - alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile); - + alloc_profile = btrfs_get_alloc_profile(extent_root, 0); ret = __btrfs_alloc_chunk(trans, extent_root, &map, &chunk_size, &stripe_size, chunk_offset, alloc_profile); if (ret) @@ -3991,10 +3998,7 @@ static noinline int init_first_rw_device(struct btrfs_trans_handle *trans, sys_chunk_offset = chunk_offset + chunk_size; - alloc_profile = BTRFS_BLOCK_GROUP_SYSTEM | - fs_info->avail_system_alloc_bits; - alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile); - + alloc_profile = btrfs_get_alloc_profile(fs_info->chunk_root, 0); ret = __btrfs_alloc_chunk(trans, extent_root, &sys_map, &sys_chunk_size, &sys_stripe_size, sys_chunk_offset, alloc_profile); -- cgit From 87533c475187c1420794a2e164bc67a7974f1327 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Tue, 29 Jan 2013 10:14:48 +0000 Subject: Btrfs: use bit operation for ->fs_state There is no lock to protect fs_info->fs_state, it will introduce some problems, such as the value may be covered by the other task when several tasks modify it. For example: Task0 - CPU0 Task1 - CPU1 mov %fs_state rax or $0x1 rax mov %fs_state rax or $0x2 rax mov rax %fs_state mov rax %fs_state The expected value is 3, but in fact, it is 2. Though this problem doesn't happen now (because there is only one flag currently), the code is error prone, if we add other flags, the above problem will happen to a certainty. Now we use bit operation for it to fix the above problem. In this way, we can make the code more robust and be easy to add new flags. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 4 +++- fs/btrfs/disk-io.c | 5 +++-- fs/btrfs/file.c | 2 +- fs/btrfs/scrub.c | 2 +- fs/btrfs/super.c | 4 ++-- fs/btrfs/transaction.c | 9 ++++----- 6 files changed, 14 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 5f5c30aaef36..78fba44ad3d9 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -338,7 +338,9 @@ static inline unsigned long btrfs_chunk_item_size(int num_stripes) /* * File system states */ +#define BTRFS_FS_STATE_ERROR 0 +/* Super block flags */ /* Errors detected */ #define BTRFS_SUPER_FLAG_ERROR (1ULL << 2) @@ -1549,7 +1551,7 @@ struct btrfs_fs_info { u64 qgroup_seq; /* filesystem state */ - u64 fs_state; + unsigned long fs_state; struct btrfs_delayed_root *delayed_root; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 00b6742fdde7..8e58a1f90546 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2201,7 +2201,8 @@ int open_ctree(struct super_block *sb, goto fail_alloc; /* check FS state, whether FS is broken. */ - fs_info->fs_state |= btrfs_super_flags(disk_super); + if (btrfs_super_flags(disk_super) & BTRFS_SUPER_FLAG_ERROR) + set_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state); ret = btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY); if (ret) { @@ -3359,7 +3360,7 @@ int close_ctree(struct btrfs_root *root) printk(KERN_ERR "btrfs: commit super ret %d\n", ret); } - if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) + if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) btrfs_error_commit_super(root); btrfs_put_block_group_cache(fs_info); diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 13c78ea3ebce..75d0fe134be3 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1545,7 +1545,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, * although we have opened a file as writable, we have * to stop this write operation to ensure FS consistency. */ - if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { + if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) { mutex_unlock(&inode->i_mutex); err = -EROFS; goto out; diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index 67783e03d121..c78b2a3fc335 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -2708,7 +2708,7 @@ static noinline_for_stack int scrub_supers(struct scrub_ctx *sctx, int ret; struct btrfs_root *root = sctx->dev_root; - if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) + if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) return -EIO; gen = root->fs_info->last_trans_committed; diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index eda330df45a4..4152f9ea34f5 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -98,7 +98,7 @@ static void __save_error_info(struct btrfs_fs_info *fs_info) * today we only save the error info into ram. Long term we'll * also send it down to the disk */ - fs_info->fs_state = BTRFS_SUPER_FLAG_ERROR; + set_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state); } static void save_error_info(struct btrfs_fs_info *fs_info) @@ -114,7 +114,7 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info) if (sb->s_flags & MS_RDONLY) return; - if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { + if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) { sb->s_flags |= MS_RDONLY; printk(KERN_INFO "btrfs is forced readonly\n"); /* diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 34610dc6d140..baf6d74fd0f2 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -61,7 +61,7 @@ static noinline int join_transaction(struct btrfs_root *root, int type) spin_lock(&fs_info->trans_lock); loop: /* The file system has been taken offline. No new transactions. */ - if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { + if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) { spin_unlock(&fs_info->trans_lock); return -EROFS; } @@ -113,7 +113,7 @@ loop: kmem_cache_free(btrfs_transaction_cachep, cur_trans); cur_trans = fs_info->running_transaction; goto loop; - } else if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { + } else if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) { spin_unlock(&fs_info->trans_lock); kmem_cache_free(btrfs_transaction_cachep, cur_trans); return -EROFS; @@ -301,7 +301,7 @@ start_transaction(struct btrfs_root *root, u64 num_items, int type, int ret; u64 qgroup_reserved = 0; - if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) + if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) return ERR_PTR(-EROFS); if (current->journal_info) { @@ -645,9 +645,8 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, btrfs_run_delayed_iputs(root); if (trans->aborted || - root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { + test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) err = -EIO; - } assert_qgroups_uptodate(trans); kmem_cache_free(btrfs_trans_handle_cachep, trans); -- cgit From 0448748849ef7c593be40e2c1404f7974bd3aac6 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 29 Jan 2013 15:03:37 -0500 Subject: Btrfs: fix chunk allocation error handling If we error out allocating a dev extent we will have already created the block group and such which will cause problems since the allocator may have tried to allocate out of the block group that no longer exists. This will cause BUG_ON()'s in the bio submission path. This also makes a failure to allocate a dev extent a non-abort error, we will just clean up the dev extents we did allocate and exit. Now if we fail to delete the dev extents we will abort since we can't have half of the dev extents hanging around, but this will make us much less likely to abort. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/volumes.c | 32 ++++++++++++++++++++++---------- 1 file changed, 22 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 8c9ea4cd66bb..13efbcf03122 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3825,12 +3825,6 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, if (ret) goto error; - ret = btrfs_make_block_group(trans, extent_root, 0, type, - BTRFS_FIRST_CHUNK_TREE_OBJECTID, - start, num_bytes); - if (ret) - goto error; - for (i = 0; i < map->num_stripes; ++i) { struct btrfs_device *device; u64 dev_offset; @@ -3842,15 +3836,33 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, info->chunk_root->root_key.objectid, BTRFS_FIRST_CHUNK_TREE_OBJECTID, start, dev_offset, stripe_size); - if (ret) { - btrfs_abort_transaction(trans, extent_root, ret); - goto error; - } + if (ret) + goto error_dev_extent; + } + + ret = btrfs_make_block_group(trans, extent_root, 0, type, + BTRFS_FIRST_CHUNK_TREE_OBJECTID, + start, num_bytes); + if (ret) { + i = map->num_stripes - 1; + goto error_dev_extent; } kfree(devices_info); return 0; +error_dev_extent: + for (; i >= 0; i--) { + struct btrfs_device *device; + int err; + + device = map->stripes[i].dev; + err = btrfs_free_dev_extent(trans, device, start); + if (err) { + btrfs_abort_transaction(trans, extent_root, err); + break; + } + } error: kfree(map); kfree(devices_info); -- cgit From 0f5d42b287f32417e54485d79f2318cf2970b37d Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 31 Jan 2013 10:23:04 -0500 Subject: Btrfs: remove extent mapping if we fail to add chunk I got a double free error when unmounting a file system that failed to add a chunk during its operation. This is because we will kfree the mapping that we created but leave the extent_map in the em_tree for chunks. So to fix this just remove the extent_map when we error out so we don't run into this problem. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/volumes.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 13efbcf03122..c7843349c795 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3821,9 +3821,10 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, write_lock(&em_tree->lock); ret = add_extent_mapping(em_tree, em); write_unlock(&em_tree->lock); - free_extent_map(em); - if (ret) + if (ret) { + free_extent_map(em); goto error; + } for (i = 0; i < map->num_stripes; ++i) { struct btrfs_device *device; @@ -3848,6 +3849,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, goto error_dev_extent; } + free_extent_map(em); kfree(devices_info); return 0; @@ -3863,6 +3865,14 @@ error_dev_extent: break; } } + write_lock(&em_tree->lock); + remove_extent_mapping(em_tree, em); + write_unlock(&em_tree->lock); + + /* One for our allocation */ + free_extent_map(em); + /* One for the tree reference */ + free_extent_map(em); error: kfree(map); kfree(devices_info); -- cgit From 96f1bb57771f71bf1d55d5031a1cf47908494330 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 30 Jan 2013 17:02:51 -0500 Subject: Btrfs: do not overcommit if we don't have enough space for global rsv Because of how little we allocate chunks now we can get really tight on metadata space before we will allocate a new chunk. This resulted in being unable to add device extents when allocating a new metadata chunk as we did not have enough space. This is because we were allowed to overcommit too much metadata without actually making sure we had enough space to make allocations. The idea behind overcommit is that we are allowed to say "sure you can have that reservation" when most of the free space is occupied by reservations, not actual allocations. But in this case where a majority of the total space is in use by actual allocations we can screw ourselves by not being able to make real allocations when it matters. So make sure we have enough real space for our global reserve, and if not then don't allow overcommitting. Thanks, Reported-and-tested-by: Jim Schutt Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 21 +++++++++++++++++++-- 1 file changed, 19 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index faff98f720de..e035731b3608 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3672,13 +3672,30 @@ static int can_overcommit(struct btrfs_root *root, struct btrfs_space_info *space_info, u64 bytes, enum btrfs_reserve_flush_enum flush) { + struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv; u64 profile = btrfs_get_alloc_profile(root, 0); + u64 rsv_size = 0; u64 avail; u64 used; used = space_info->bytes_used + space_info->bytes_reserved + - space_info->bytes_pinned + space_info->bytes_readonly + - space_info->bytes_may_use; + space_info->bytes_pinned + space_info->bytes_readonly; + + spin_lock(&global_rsv->lock); + rsv_size = global_rsv->size; + spin_unlock(&global_rsv->lock); + + /* + * We only want to allow over committing if we have lots of actual space + * free, but if we don't have enough space to handle the global reserve + * space then we could end up having a real enospc problem when trying + * to allocate a chunk or some other such important allocation. + */ + rsv_size <<= 1; + if (used + rsv_size >= space_info->total_bytes) + return 0; + + used += space_info->bytes_may_use; spin_lock(&root->fs_info->free_chunk_lock); avail = root->fs_info->free_chunk_space; -- cgit From b4c6f7b75ce9fa3bb78af314e1841af5716bb643 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:52 +0000 Subject: btrfs: remove unused fd in btrfs_ioctl_send() All we do is set it to NULL and test it :) Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/send.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index 321b7fb4e441..614da0d44d56 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -4536,7 +4536,6 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_) struct btrfs_fs_info *fs_info; struct btrfs_ioctl_send_args *arg = NULL; struct btrfs_key key; - struct file *filp = NULL; struct send_ctx *sctx = NULL; u32 i; u64 *clone_sources_tmp = NULL; @@ -4673,8 +4672,6 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_) goto out; out: - if (filp) - fput(filp); kfree(arg); vfree(clone_sources_tmp); -- cgit From d1d3cd27a32b819a0b0e2b9b6b5f9e80890f9821 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:53 +0000 Subject: btrfs: list_entry can't return NULL No need to test the result, we can't get a null pointer from list_entry() Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 8e58a1f90546..bd48bf21118e 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3846,8 +3846,6 @@ int btrfs_cleanup_transaction(struct btrfs_root *root) while (!list_empty(&list)) { t = list_entry(list.next, struct btrfs_transaction, list); - if (!t) - break; btrfs_destroy_ordered_operations(root); -- cgit From 5a01604783c23f28bd4212e839e79277193f974f Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:54 +0000 Subject: btrfs: remove unused fs_info from btrfs_decode_error() Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/super.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 4152f9ea34f5..fd6a07c30ee9 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -63,8 +63,7 @@ static const struct super_operations btrfs_super_ops; static struct file_system_type btrfs_fs_type; -static const char *btrfs_decode_error(struct btrfs_fs_info *fs_info, int errno, - char nbuf[16]) +static const char *btrfs_decode_error(int errno, char nbuf[16]) { char *errstr = NULL; @@ -152,7 +151,7 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, if (errno == -EROFS && (sb->s_flags & MS_RDONLY)) return; - errstr = btrfs_decode_error(fs_info, errno, nbuf); + errstr = btrfs_decode_error(errno, nbuf); if (fmt) { struct va_format vaf = { .fmt = fmt, @@ -261,7 +260,7 @@ void __btrfs_abort_transaction(struct btrfs_trans_handle *trans, char nbuf[16]; const char *errstr; - errstr = btrfs_decode_error(root->fs_info, errno, nbuf); + errstr = btrfs_decode_error(errno, nbuf); btrfs_printk(root->fs_info, "%s:%d: Aborting unused transaction(%s).\n", function, line, errstr); @@ -289,7 +288,7 @@ void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function, va_start(args, fmt); vaf.va = &args; - errstr = btrfs_decode_error(fs_info, errno, nbuf); + errstr = btrfs_decode_error(errno, nbuf); if (fs_info->mount_opt & BTRFS_MOUNT_PANIC_ON_FATAL_ERROR) panic(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (%s)\n", s_id, function, line, &vaf, errstr); -- cgit From aa43a17c21cf35329b1a495dd6876798dd8b016b Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:55 +0000 Subject: btrfs: handle null fs_info in btrfs_panic() At least backref_tree_panic() can apparently pass in a null fs_info, so handle that in __btrfs_panic to get the message out on the console. The btrfs_panic macro also uses fs_info, but that's largely pointless; it's testing to see if BTRFS_MOUNT_PANIC_ON_FATAL_ERROR is not set. But if it *were* set, __btrfs_panic() would have, well, paniced and we wouldn't be here, testing it! So just BUG() at this point. And since we only use fs_info once now, just use it directly. Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 9 ++++++--- fs/btrfs/super.c | 2 +- 2 files changed, 7 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 78fba44ad3d9..1e3a994e1899 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3653,11 +3653,14 @@ __printf(5, 6) void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function, unsigned int line, int errno, const char *fmt, ...); +/* + * If BTRFS_MOUNT_PANIC_ON_FATAL_ERROR is in mount_opt, __btrfs_panic + * will panic(). Otherwise we BUG() here. + */ #define btrfs_panic(fs_info, errno, fmt, args...) \ do { \ - struct btrfs_fs_info *_i = (fs_info); \ - __btrfs_panic(_i, __func__, __LINE__, errno, fmt, ##args); \ - BUG_ON(!(_i->mount_opt & BTRFS_MOUNT_PANIC_ON_FATAL_ERROR)); \ + __btrfs_panic(fs_info, __func__, __LINE__, errno, fmt, ##args); \ + BUG(); \ } while (0) /* acl.c */ diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index fd6a07c30ee9..06537d217419 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -289,7 +289,7 @@ void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function, vaf.va = &args; errstr = btrfs_decode_error(errno, nbuf); - if (fs_info->mount_opt & BTRFS_MOUNT_PANIC_ON_FATAL_ERROR) + if (fs_info && (fs_info->mount_opt & BTRFS_MOUNT_PANIC_ON_FATAL_ERROR)) panic(KERN_CRIT "BTRFS panic (device %s) in %s:%d: %pV (%s)\n", s_id, function, line, &vaf, errstr); -- cgit From 1c697d4acc2e10b2a65b22abba2687a2897ecd74 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:56 +0000 Subject: btrfs: annotate intentional switch case fallthroughs This keeps static checkers happy. Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/ctree.c | 1 + fs/btrfs/super.c | 1 + 2 files changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 6eff0fa9ecaa..35444013f0cf 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1138,6 +1138,7 @@ __tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq, switch (tm->op) { case MOD_LOG_KEY_REMOVE_WHILE_FREEING: BUG_ON(tm->slot < n); + /* Fallthrough */ case MOD_LOG_KEY_REMOVE_WHILE_MOVING: case MOD_LOG_KEY_REMOVE: btrfs_set_node_key(eb, &tm->key, tm->slot); diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 06537d217419..976f6aac102e 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -437,6 +437,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) case Opt_compress_force: case Opt_compress_force_type: compress_force = true; + /* Fallthrough */ case Opt_compress: case Opt_compress_type: if (token == Opt_compress || -- cgit From 0e6360274ff3bd4fb8cf97c5c3483cf5a42a49f4 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:57 +0000 Subject: btrfs: add missing break in btrfs_print_leaf() I don't think that BTRFS_DEV_EXTENT_KEY is supposed to fall through to BTRFS_DEV_STATS_KEY ... Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/print-tree.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/print-tree.c b/fs/btrfs/print-tree.c index 50d95fd190a5..920957ecb27e 100644 --- a/fs/btrfs/print-tree.c +++ b/fs/btrfs/print-tree.c @@ -294,6 +294,7 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l) btrfs_dev_extent_chunk_offset(l, dev_extent), (unsigned long long) btrfs_dev_extent_length(l, dev_extent)); + break; case BTRFS_DEV_STATS_KEY: printk(KERN_INFO "\t\tdevice stats\n"); break; -- cgit From 37252a66f34940421942725ca75d12e6008682dc Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:58 +0000 Subject: btrfs: fix varargs in __btrfs_std_error __btrfs_std_error didn't always properly call va_end, and might call va_start even if fmt was NULL. Move all the varargs handling into the block where we have fmt. Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/super.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 976f6aac102e..e0ea1278377a 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -141,8 +141,6 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, struct super_block *sb = fs_info->sb; char nbuf[16]; const char *errstr; - va_list args; - va_start(args, fmt); /* * Special case: if the error is EROFS, and we're already @@ -153,13 +151,16 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, errstr = btrfs_decode_error(errno, nbuf); if (fmt) { - struct va_format vaf = { - .fmt = fmt, - .va = &args, - }; + struct va_format vaf; + va_list args; + + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: %s (%pV)\n", sb->s_id, function, line, errstr, &vaf); + va_end(args); } else { printk(KERN_CRIT "BTRFS error (device %s) in %s:%d: %s\n", sb->s_id, function, line, errstr); @@ -170,7 +171,6 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, save_error_info(fs_info); btrfs_handle_error(fs_info); } - va_end(args); } static const char * const logtypes[] = { -- cgit From d4c0a7da214d7915200d3e64c058ef36771186f2 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:54:59 +0000 Subject: btrfs: remove unused "item" in btrfs_insert_delayed_item() "item" was set but never used in this function. Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/delayed-inode.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index c7e75061b894..0b278b117cbe 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -875,7 +875,6 @@ static int btrfs_insert_delayed_item(struct btrfs_trans_handle *trans, struct btrfs_delayed_item *delayed_item) { struct extent_buffer *leaf; - struct btrfs_item *item; char *ptr; int ret; @@ -886,7 +885,6 @@ static int btrfs_insert_delayed_item(struct btrfs_trans_handle *trans, leaf = path->nodes[0]; - item = btrfs_item_nr(leaf, path->slots[0]); ptr = btrfs_item_ptr(leaf, path->slots[0], char); write_extent_buffer(leaf, delayed_item->data, (unsigned long)ptr, -- cgit From 1971e917c8c99ff190264305803fbafcbbac1422 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:55:00 +0000 Subject: btrfs: remove unnecessary DEFINE_WAIT() declarations No point in DEFINE_WAIT(wait) if it's not used! Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index e035731b3608..3158817cd5a9 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5541,7 +5541,6 @@ wait_block_group_cache_progress(struct btrfs_block_group_cache *cache, u64 num_bytes) { struct btrfs_caching_control *caching_ctl; - DEFINE_WAIT(wait); caching_ctl = get_caching_control(cache); if (!caching_ctl) @@ -5558,7 +5557,6 @@ static noinline int wait_block_group_cache_done(struct btrfs_block_group_cache *cache) { struct btrfs_caching_control *caching_ctl; - DEFINE_WAIT(wait); caching_ctl = get_caching_control(cache); if (!caching_ctl) -- cgit From 063d006fa06fbf73fab370921120380333a33e85 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 00:55:01 +0000 Subject: btrfs: ensure we don't overrun devices_info[] in __btrfs_alloc_chunk WARN_ON isn't enough, we need to stop the loop if for any reason we would overrun the devices_info array. I tried to track down the connection between the length of the alloc_devices list and the rw_devices counter but it wasn't immediately obvious, so be defensive about it. Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/volumes.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index c7843349c795..305b6a63ab1f 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3734,12 +3734,16 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, if (max_avail < BTRFS_STRIPE_LEN * dev_stripes) continue; + if (ndevs == fs_devices->rw_devices) { + WARN(1, "%s: found more than %llu devices\n", + __func__, fs_devices->rw_devices); + break; + } devices_info[ndevs].dev_offset = dev_offset; devices_info[ndevs].max_avail = max_avail; devices_info[ndevs].total_avail = total_avail; devices_info[ndevs].dev = device; ++ndevs; - WARN_ON(ndevs > fs_devices->rw_devices); } /* -- cgit From eb12db690c7eb0f6593ba5792f5861409e88bc03 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 30 Jan 2013 16:03:59 -0500 Subject: Btrfs: fix freeing delayed ref head while still holding its mutex I hit this error when reproducing a bug that would end in a transaction abort. We take the delayed ref head's mutex to keep anybody from processing it while we're destroying it, but we fail to drop the mutex before we carry on and free the damned thing. Fix this by doing the remove logic for the head ourselves and unlock the mutex, that way we can avoid use after free's or hung tasks waiting on that mutex to come back so they know the delayed ref completed. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index bd48bf21118e..8140cb01951f 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3615,11 +3615,11 @@ int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, } while ((node = rb_first(&delayed_refs->root)) != NULL) { - ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node); + struct btrfs_delayed_ref_head *head = NULL; + ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node); atomic_set(&ref->refs, 1); if (btrfs_delayed_ref_is_head(ref)) { - struct btrfs_delayed_ref_head *head; head = btrfs_delayed_node_to_head(ref); if (!mutex_trylock(&head->mutex)) { @@ -3641,10 +3641,12 @@ int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, delayed_refs->num_heads_ready--; list_del_init(&head->cluster); } + ref->in_tree = 0; rb_erase(&ref->rb_node, &delayed_refs->root); delayed_refs->num_entries--; - + if (head) + mutex_unlock(&head->mutex); spin_unlock(&delayed_refs->lock); btrfs_put_delayed_ref(ref); -- cgit From 779880ef35b60ac82eed1dcaec6db5b34a170df8 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 31 Jan 2013 14:30:08 -0500 Subject: Btrfs: fix how we discard outstanding ordered extents on abort When we abort we've been just free'ing up all the ordered extents and hoping for the best. This results in lots of warnings from various places, warnings from btrfs_destroy_inode() because it's ENOSPC accounting isn't fixed. It will also screw up lots of pages who have been set private but never get cleared because the ordered extents are never allowed to be submitted. This patch fixes those warnings. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 33 +++++++-------------------------- 1 file changed, 7 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 8140cb01951f..259d7891d10c 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3565,35 +3565,16 @@ static void btrfs_destroy_ordered_operations(struct btrfs_root *root) static void btrfs_destroy_ordered_extents(struct btrfs_root *root) { - struct list_head splice; struct btrfs_ordered_extent *ordered; - struct inode *inode; - - INIT_LIST_HEAD(&splice); spin_lock(&root->fs_info->ordered_extent_lock); - - list_splice_init(&root->fs_info->ordered_extents, &splice); - while (!list_empty(&splice)) { - ordered = list_entry(splice.next, struct btrfs_ordered_extent, - root_extent_list); - - list_del_init(&ordered->root_extent_list); - atomic_inc(&ordered->refs); - - /* the inode may be getting freed (in sys_unlink path). */ - inode = igrab(ordered->inode); - - spin_unlock(&root->fs_info->ordered_extent_lock); - if (inode) - iput(inode); - - atomic_set(&ordered->refs, 1); - btrfs_put_ordered_extent(ordered); - - spin_lock(&root->fs_info->ordered_extent_lock); - } - + /* + * This will just short circuit the ordered completion stuff which will + * make sure the ordered extent gets properly cleaned up. + */ + list_for_each_entry(ordered, &root->fs_info->ordered_extents, + root_extent_list) + set_bit(BTRFS_ORDERED_IOERR, &ordered->flags); spin_unlock(&root->fs_info->ordered_extent_lock); } -- cgit From 0bec9ef525e33233d7739b71be83bb78746f6e94 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 31 Jan 2013 14:58:00 -0500 Subject: Btrfs: unreserve space if our ordered extent fails to work When a transaction aborts or there's an EIO on an ordered extent or any error really we will not free up the space we reserved for this ordered extent. This results in warnings from the block group cache cleanup in the case of a transaction abort, or leaking space in the case of EIO on an ordered extent. Fix this up by free'ing the reserved space if we have an error at all trying to complete an ordered extent. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 82c7c66f8523..2b48d2d51d50 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2022,11 +2022,23 @@ out: if (trans) btrfs_end_transaction(trans, root); - if (ret) + if (ret) { clear_extent_uptodate(io_tree, ordered_extent->file_offset, ordered_extent->file_offset + ordered_extent->len - 1, NULL, GFP_NOFS); + /* + * If the ordered extent had an IOERR or something else went + * wrong we need to return the space for this ordered extent + * back to the allocator. + */ + if (!test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags) && + !test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) + btrfs_free_reserved_extent(root, ordered_extent->start, + ordered_extent->disk_len); + } + + /* * This needs to be done to make sure anybody waiting knows we are done * updating everything for this ordered extent. -- cgit From 925396ecf251432d6d0f703a6cfd0cb9e651d936 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 1 Feb 2013 15:57:47 -0500 Subject: Btrfs: account for orphan inodes properly during cleanup Dave sent me a panic where we were doing the orphan cleanup and panic'ed trying to release our reservation from the orphan block rsv. The reason for this is because our orphan block rsv had been free'd out from underneath us because the transaction commit found that there were no orphan inodes according to its count and decided to free it. This is incorrect so make sure we inc the orphan inodes count so the accounting is all done properly. This would also cause the warning in the orphan commit code normally if you had any orphans to cleanup as they would only decrement the orphan count so you'd get a negative orphan count which could cause problems during runtime. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 2b48d2d51d50..2dc59734e071 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2502,6 +2502,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) */ set_bit(BTRFS_INODE_HAS_ORPHAN_ITEM, &BTRFS_I(inode)->runtime_flags); + atomic_inc(&root->orphan_inodes); /* if we have links, this was a truncate, lets do that */ if (inode->i_nlink) { -- cgit From 70afa3998c9baed4186df38988246de1abdab56d Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Feb 2013 13:53:19 -0500 Subject: Btrfs: rework the overcommit logic to be based on the total size People have been complaining about random ENOSPC errors that will clear up after a umount or just a given amount of time. Chris was able to reproduce this with stress.sh and lots of processes and so was I. Basically the overcommit stuff would really let us get out of hand, in my tests I saw up to 30 gigs of outstanding reservations with only 2 gigs total of metadata space. This usually worked out fine but with so much outstanding reservation the flushing stuff short circuits to make sure we don't hang forever flushing when we really need ENOSPC. Plus we allocate chunks in order to alleviate the pressure, but this doesn't actually help us since we only use the non-allocated area in our over commit logic. So instead of basing overcommit on the amount of non-allocated space, instead just do it based on how much total space we have, and then limit it to the non-allocated space in case we are short on space to spill over into. This allows us to have the same performance as well as no longer giving random ENOSPC. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 3158817cd5a9..81aa7cf3ae86 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3677,6 +3677,7 @@ static int can_overcommit(struct btrfs_root *root, u64 rsv_size = 0; u64 avail; u64 used; + u64 to_add; used = space_info->bytes_used + space_info->bytes_reserved + space_info->bytes_pinned + space_info->bytes_readonly; @@ -3710,17 +3711,25 @@ static int can_overcommit(struct btrfs_root *root, BTRFS_BLOCK_GROUP_RAID10)) avail >>= 1; + to_add = space_info->total_bytes; + /* * If we aren't flushing all things, let us overcommit up to * 1/2th of the space. If we can flush, don't let us overcommit * too much, let it overcommit up to 1/8 of the space. */ if (flush == BTRFS_RESERVE_FLUSH_ALL) - avail >>= 3; + to_add >>= 3; else - avail >>= 1; + to_add >>= 1; + + /* + * Limit the overcommit to the amount of free space we could possibly + * allocate for chunks. + */ + to_add = min(avail, to_add); - if (used + bytes < space_info->total_bytes + avail) + if (used + bytes < space_info->total_bytes + to_add) return 1; return 0; } -- cgit From 3e04e7f10b68999e0d8321516ea19d9d5b044dee Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Feb 2013 16:49:15 -0500 Subject: Btrfs: handle errors in compression submission path I noticed we would deadlock if we aborted a transaction while doing compressed io. This is because we don't unlock our pages if something goes horribly wrong. To fix this we need to make sure that we call extent_clear_unlock_delalloc in order to unlock all the pages. If we have to cow in the async submission thread we need to make sure to unlock our locked_page as the cow error path will not unlock the locked page as it depends on the caller to unlock that page. With this patch we no longer deadlock on the page lock when we have an aborted transaction. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 38 ++++++++++++++++++++++++++++---------- 1 file changed, 28 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 2dc59734e071..348b7bb920ef 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -608,7 +608,7 @@ static noinline int submit_compressed_extents(struct inode *inode, if (list_empty(&async_cow->extents)) return 0; - +again: while (!list_empty(&async_cow->extents)) { async_extent = list_entry(async_cow->extents.next, struct async_extent, list); @@ -648,6 +648,8 @@ retry: async_extent->ram_size - 1, btrfs_get_extent, WB_SYNC_ALL); + else if (ret) + unlock_page(async_cow->locked_page); kfree(async_extent); cond_resched(); continue; @@ -672,6 +674,7 @@ retry: if (ret) { int i; + for (i = 0; i < async_extent->nr_pages; i++) { WARN_ON(async_extent->pages[i]->mapping); page_cache_release(async_extent->pages[i]); @@ -679,12 +682,10 @@ retry: kfree(async_extent->pages); async_extent->nr_pages = 0; async_extent->pages = NULL; - unlock_extent(io_tree, async_extent->start, - async_extent->start + - async_extent->ram_size - 1); + if (ret == -ENOSPC) goto retry; - goto out_free; /* JDM: Requeue? */ + goto out_free; } /* @@ -696,7 +697,8 @@ retry: async_extent->ram_size - 1, 0); em = alloc_extent_map(); - BUG_ON(!em); /* -ENOMEM */ + if (!em) + goto out_free_reserve; em->start = async_extent->start; em->len = async_extent->ram_size; em->orig_start = em->start; @@ -728,6 +730,9 @@ retry: async_extent->ram_size - 1, 0); } + if (ret) + goto out_free_reserve; + ret = btrfs_add_ordered_extent_compress(inode, async_extent->start, ins.objectid, @@ -735,7 +740,8 @@ retry: ins.offset, BTRFS_ORDERED_COMPRESSED, async_extent->compress_type); - BUG_ON(ret); /* -ENOMEM */ + if (ret) + goto out_free_reserve; /* * clear dirty, set writeback and unlock the pages. @@ -756,18 +762,30 @@ retry: ins.objectid, ins.offset, async_extent->pages, async_extent->nr_pages); - - BUG_ON(ret); /* -ENOMEM */ alloc_hint = ins.objectid + ins.offset; kfree(async_extent); + if (ret) + goto out; cond_resched(); } ret = 0; out: return ret; +out_free_reserve: + btrfs_free_reserved_extent(root, ins.objectid, ins.offset); out_free: + extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree, + async_extent->start, + async_extent->start + + async_extent->ram_size - 1, + NULL, EXTENT_CLEAR_UNLOCK_PAGE | + EXTENT_CLEAR_UNLOCK | + EXTENT_CLEAR_DELALLOC | + EXTENT_CLEAR_DIRTY | + EXTENT_SET_WRITEBACK | + EXTENT_END_WRITEBACK); kfree(async_extent); - goto out; + goto again; } static u64 get_extent_allocation_hint(struct inode *inode, u64 start, -- cgit From e4a2bcaca9643e7430207810653222fc5187f2be Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Feb 2013 16:55:41 -0500 Subject: Btrfs: if we aren't committing just end the transaction if we error out I hit a deadlock where transaction commit was waiting on num_writers to be 0. This happened because somebody came into btrfs_commit_transaction and noticed we had aborted and it went to cleanup_transaction. This shouldn't happen because cleanup_transaction is really to fixup a bad commit, it doesn't do the normal trans handle cleanup things. So if we have an error just do the normal btrfs_end_transaction dance and return. Once we are in the actual commit path we can use cleanup_transaction and be good to go. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index baf6d74fd0f2..5144ad19ef47 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1476,21 +1476,25 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, ret = btrfs_run_ordered_operations(root, 0); if (ret) { btrfs_abort_transaction(trans, root, ret); - goto cleanup_transaction; + btrfs_end_transaction(trans, root); + return ret; } /* Stop the commit early if ->aborted is set */ if (unlikely(ACCESS_ONCE(cur_trans->aborted))) { ret = cur_trans->aborted; - goto cleanup_transaction; + btrfs_end_transaction(trans, root); + return ret; } /* make a pass through all the delayed refs we have so far * any runnings procs may add more while we are here */ ret = btrfs_run_delayed_refs(trans, root, 0); - if (ret) - goto cleanup_transaction; + if (ret) { + btrfs_end_transaction(trans, root); + return ret; + } btrfs_trans_release_metadata(trans, root); trans->block_rsv = NULL; @@ -1507,8 +1511,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, btrfs_create_pending_block_groups(trans, root); ret = btrfs_run_delayed_refs(trans, root, 0); - if (ret) - goto cleanup_transaction; + if (ret) { + btrfs_end_transaction(trans, root); + return ret; + } spin_lock(&cur_trans->commit_lock); if (cur_trans->in_commit) { -- cgit From de78b51a2852bddccd6535e9e12de65f92787a1e Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 31 Jan 2013 18:21:12 +0000 Subject: btrfs: remove cache only arguments from defrag path The entry point at the defrag ioctl always sets "cache only" to 0; the codepaths haven't run for a long time as far as I can tell. Chris says they're dead code, so remove them. Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik --- fs/btrfs/ctree.c | 65 ++++++++------------------------------------------ fs/btrfs/ctree.h | 8 +++---- fs/btrfs/ioctl.c | 8 +++---- fs/btrfs/transaction.c | 8 +++---- fs/btrfs/transaction.h | 2 +- fs/btrfs/tree-defrag.c | 19 ++++++--------- fs/btrfs/tree-log.c | 4 ++-- 7 files changed, 32 insertions(+), 82 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 35444013f0cf..ecd25a1b4e51 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1442,7 +1442,7 @@ int btrfs_comp_cpu_keys(struct btrfs_key *k1, struct btrfs_key *k2) */ int btrfs_realloc_node(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *parent, - int start_slot, int cache_only, u64 *last_ret, + int start_slot, u64 *last_ret, struct btrfs_key *progress) { struct extent_buffer *cur; @@ -1462,8 +1462,6 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans, struct btrfs_disk_key disk_key; parent_level = btrfs_header_level(parent); - if (cache_only && parent_level != 1) - return 0; WARN_ON(trans->transaction != root->fs_info->running_transaction); WARN_ON(trans->transid != root->fs_info->generation); @@ -1509,10 +1507,6 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans, else uptodate = 0; if (!cur || !uptodate) { - if (cache_only) { - free_extent_buffer(cur); - continue; - } if (!cur) { cur = read_tree_block(root, blocknr, blocksize, gen); @@ -4826,8 +4820,8 @@ int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path) /* * A helper function to walk down the tree starting at min_key, and looking - * for nodes or leaves that are either in cache or have a minimum - * transaction id. This is used by the btree defrag code, and tree logging + * for nodes or leaves that are have a minimum transaction id. + * This is used by the btree defrag code, and tree logging * * This does not cow, but it does stuff the starting key it finds back * into min_key, so you can call btrfs_search_slot with cow=1 on the @@ -4848,7 +4842,7 @@ int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path) */ int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key, struct btrfs_key *max_key, - struct btrfs_path *path, int cache_only, + struct btrfs_path *path, u64 min_trans) { struct extent_buffer *cur; @@ -4888,15 +4882,12 @@ again: if (sret && slot > 0) slot--; /* - * check this node pointer against the cache_only and - * min_trans parameters. If it isn't in cache or is too - * old, skip to the next one. + * check this node pointer against the min_trans parameters. + * If it is too old, old, skip to the next one. */ while (slot < nritems) { u64 blockptr; u64 gen; - struct extent_buffer *tmp; - struct btrfs_disk_key disk_key; blockptr = btrfs_node_blockptr(cur, slot); gen = btrfs_node_ptr_generation(cur, slot); @@ -4904,27 +4895,7 @@ again: slot++; continue; } - if (!cache_only) - break; - - if (max_key) { - btrfs_node_key(cur, &disk_key, slot); - if (comp_keys(&disk_key, max_key) >= 0) { - ret = 1; - goto out; - } - } - - tmp = btrfs_find_tree_block(root, blockptr, - btrfs_level_size(root, level - 1)); - - if (tmp && btrfs_buffer_uptodate(tmp, gen, 1) > 0) { - free_extent_buffer(tmp); - break; - } - if (tmp) - free_extent_buffer(tmp); - slot++; + break; } find_next_key: /* @@ -4935,7 +4906,7 @@ find_next_key: path->slots[level] = slot; btrfs_set_path_blocking(path); sret = btrfs_find_next_key(root, path, min_key, level, - cache_only, min_trans); + min_trans); if (sret == 0) { btrfs_release_path(path); goto again; @@ -5400,8 +5371,7 @@ out: /* * this is similar to btrfs_next_leaf, but does not try to preserve * and fixup the path. It looks for and returns the next key in the - * tree based on the current path and the cache_only and min_trans - * parameters. + * tree based on the current path and the min_trans parameters. * * 0 is returned if another key is found, < 0 if there are any errors * and 1 is returned if there are no higher keys in the tree @@ -5410,8 +5380,7 @@ out: * calling this function. */ int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path, - struct btrfs_key *key, int level, - int cache_only, u64 min_trans) + struct btrfs_key *key, int level, u64 min_trans) { int slot; struct extent_buffer *c; @@ -5462,22 +5431,8 @@ next: if (level == 0) btrfs_item_key_to_cpu(c, key, slot); else { - u64 blockptr = btrfs_node_blockptr(c, slot); u64 gen = btrfs_node_ptr_generation(c, slot); - if (cache_only) { - struct extent_buffer *cur; - cur = btrfs_find_tree_block(root, blockptr, - btrfs_level_size(root, level - 1)); - if (!cur || - btrfs_buffer_uptodate(cur, gen, 1) <= 0) { - slot++; - if (cur) - free_extent_buffer(cur); - goto next; - } - free_extent_buffer(cur); - } if (gen < min_trans) { slot++; goto next; diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 1e3a994e1899..7e2cffd2a5d8 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3125,10 +3125,10 @@ struct extent_buffer *btrfs_root_node(struct btrfs_root *root); struct extent_buffer *btrfs_lock_root_node(struct btrfs_root *root); int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path, struct btrfs_key *key, int lowest_level, - int cache_only, u64 min_trans); + u64 min_trans); int btrfs_search_forward(struct btrfs_root *root, struct btrfs_key *min_key, struct btrfs_key *max_key, - struct btrfs_path *path, int cache_only, + struct btrfs_path *path, u64 min_trans); enum btrfs_compare_tree_result { BTRFS_COMPARE_TREE_NEW, @@ -3181,7 +3181,7 @@ int btrfs_search_slot_for_read(struct btrfs_root *root, int find_higher, int return_any); int btrfs_realloc_node(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *parent, - int start_slot, int cache_only, u64 *last_ret, + int start_slot, u64 *last_ret, struct btrfs_key *progress); void btrfs_release_path(struct btrfs_path *p); struct btrfs_path *btrfs_alloc_path(void); @@ -3576,7 +3576,7 @@ int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode, /* tree-defrag.c */ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, - struct btrfs_root *root, int cache_only); + struct btrfs_root *root); /* sysfs.c */ int btrfs_init_sysfs(void); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 96ecefc1724f..daea831f3d36 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -818,7 +818,7 @@ static int find_new_extents(struct btrfs_root *root, while(1) { ret = btrfs_search_forward(root, &min_key, &max_key, - path, 0, newer_than); + path, newer_than); if (ret != 0) goto none; if (min_key.objectid != ino) @@ -1864,7 +1864,7 @@ static noinline int search_ioctl(struct inode *inode, path->keep_locks = 1; while(1) { - ret = btrfs_search_forward(root, &key, &max_key, path, 0, + ret = btrfs_search_forward(root, &key, &max_key, path, sk->min_transid); if (ret != 0) { if (ret > 0) @@ -2212,10 +2212,10 @@ static int btrfs_ioctl_defrag(struct file *file, void __user *argp) ret = -EPERM; goto out; } - ret = btrfs_defrag_root(root, 0); + ret = btrfs_defrag_root(root); if (ret) goto out; - ret = btrfs_defrag_root(root->fs_info->extent_root, 0); + ret = btrfs_defrag_root(root->fs_info->extent_root); break; case S_IFREG: if (!(file->f_mode & FMODE_WRITE)) { diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 5144ad19ef47..60481a53e004 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -959,10 +959,10 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans, } /* - * defrag a given btree. If cacheonly == 1, this won't read from the disk, - * otherwise every leaf in the btree is read and defragged. + * defrag a given btree. + * Every leaf in the btree is read and defragged. */ -int btrfs_defrag_root(struct btrfs_root *root, int cacheonly) +int btrfs_defrag_root(struct btrfs_root *root) { struct btrfs_fs_info *info = root->fs_info; struct btrfs_trans_handle *trans; @@ -976,7 +976,7 @@ int btrfs_defrag_root(struct btrfs_root *root, int cacheonly) if (IS_ERR(trans)) return PTR_ERR(trans); - ret = btrfs_defrag_leaves(trans, root, cacheonly); + ret = btrfs_defrag_leaves(trans, root); btrfs_end_transaction(trans, root); btrfs_btree_balance_dirty(info->tree_root); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 69700f7b20ac..46628210e5d8 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -117,7 +117,7 @@ int btrfs_write_and_wait_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root); int btrfs_add_dead_root(struct btrfs_root *root); -int btrfs_defrag_root(struct btrfs_root *root, int cacheonly); +int btrfs_defrag_root(struct btrfs_root *root); int btrfs_clean_old_snapshots(struct btrfs_root *root); int btrfs_commit_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root); diff --git a/fs/btrfs/tree-defrag.c b/fs/btrfs/tree-defrag.c index 3b580ee8ab1d..94e05c1f118a 100644 --- a/fs/btrfs/tree-defrag.c +++ b/fs/btrfs/tree-defrag.c @@ -23,13 +23,14 @@ #include "transaction.h" #include "locking.h" -/* defrag all the leaves in a given btree. If cache_only == 1, don't read - * things from disk, otherwise read all the leaves and try to get key order to +/* + * Defrag all the leaves in a given btree. + * Read all the leaves and try to get key order to * better reflect disk order */ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, - struct btrfs_root *root, int cache_only) + struct btrfs_root *root) { struct btrfs_path *path = NULL; struct btrfs_key key; @@ -41,9 +42,6 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, u64 last_ret = 0; u64 min_trans = 0; - if (cache_only) - goto out; - if (root->fs_info->extent_root == root) { /* * there's recursion here right now in the tree locking, @@ -86,11 +84,8 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, } path->keep_locks = 1; - if (cache_only) - min_trans = root->defrag_trans_start; - ret = btrfs_search_forward(root, &key, NULL, path, - cache_only, min_trans); + ret = btrfs_search_forward(root, &key, NULL, path, min_trans); if (ret < 0) goto out; if (ret > 0) { @@ -109,11 +104,11 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, goto out; } path->slots[1] = btrfs_header_nritems(path->nodes[1]); - next_key_ret = btrfs_find_next_key(root, path, &key, 1, cache_only, + next_key_ret = btrfs_find_next_key(root, path, &key, 1, min_trans); ret = btrfs_realloc_node(trans, root, path->nodes[1], 0, - cache_only, &last_ret, + &last_ret, &root->defrag_progress); if (ret) { WARN_ON(ret == -EAGAIN); diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index deb7f66356bd..1a79087c4575 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -2738,7 +2738,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, path->keep_locks = 1; ret = btrfs_search_forward(root, &min_key, &max_key, - path, 0, trans->transid); + path, trans->transid); /* * we didn't find anything from this transaction, see if there @@ -3680,7 +3680,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, while (1) { ins_nr = 0; ret = btrfs_search_forward(root, &min_key, &max_key, - path, 0, trans->transid); + path, trans->transid); if (ret != 0) break; again: -- cgit From 2f697dc6a648d3a16f512fe7a53281d55cce1570 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 4 Feb 2013 13:12:18 +0000 Subject: Btrfs: extend the checksum item as much as possible For write, we also reserve some space for COW blocks during updating the checksum tree, and we calculate the number of blocks by checking if the number of bytes outstanding that are going to need csums needs one more block for csum. When we add these checksum into the checksum tree, we use ordered sums list. Every ordered sum contains csums for each sector, and we'll first try to look up an existing csum item, a) if we don't yet have a proper csum item, then we need to insert one, b) or if we find one but the csum item is not big enough, then we need to extend it. The point is we'll unlock the whole path and then insert or extend. So others can hack in and update the tree. Each insert or extend needs update the tree with COW on, and we may need to insert/extend for many times. That means what we've reserved for updating checksum tree is NOT enough indeed. The case is even more serious with having several write threads at the same time, it can end up eating our reserved space quickly and starting eating globle reserve pool instead. I don't yet come up with a way to calculate the worse case for updating csum, but extending the checksum item as much as possible can be helpful in my test. The idea behind is that it can reduce the times we insert/extend so that it saves us precious reserved space. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/file-item.c | 67 ++++++++++++++++++++++++++++++++++++---------------- 1 file changed, 46 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index 94aa53b38721..ec160202be3e 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -684,6 +684,24 @@ out: return ret; } +static u64 btrfs_sector_sum_left(struct btrfs_ordered_sum *sums, + struct btrfs_sector_sum *sector_sum, + u64 total_bytes, u64 sectorsize) +{ + u64 tmp = sectorsize; + u64 next_sector = sector_sum->bytenr; + struct btrfs_sector_sum *next = sector_sum + 1; + + while ((tmp + total_bytes) < sums->len) { + if (next_sector + sectorsize != next->bytenr) + break; + tmp += sectorsize; + next_sector = next->bytenr; + next++; + } + return tmp; +} + int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_ordered_sum *sums) @@ -789,20 +807,32 @@ again: goto insert; } - if (csum_offset >= btrfs_item_size_nr(leaf, path->slots[0]) / + if (csum_offset == btrfs_item_size_nr(leaf, path->slots[0]) / csum_size) { - u32 diff = (csum_offset + 1) * csum_size; + int extend_nr; + u64 tmp; + u32 diff; + u32 free_space; - /* - * is the item big enough already? we dropped our lock - * before and need to recheck - */ - if (diff < btrfs_item_size_nr(leaf, path->slots[0])) - goto csum; + if (btrfs_leaf_free_space(root, leaf) < + sizeof(struct btrfs_item) + csum_size * 2) + goto insert; + + free_space = btrfs_leaf_free_space(root, leaf) - + sizeof(struct btrfs_item) - csum_size; + tmp = btrfs_sector_sum_left(sums, sector_sum, total_bytes, + root->sectorsize); + tmp >>= root->fs_info->sb->s_blocksize_bits; + WARN_ON(tmp < 1); + + extend_nr = max_t(int, 1, (int)tmp); + diff = (csum_offset + extend_nr) * csum_size; + diff = min(diff, MAX_CSUM_ITEMS(root, csum_size) * csum_size); diff = diff - btrfs_item_size_nr(leaf, path->slots[0]); - if (diff != csum_size) - goto insert; + diff = min(free_space, diff); + diff /= csum_size; + diff *= csum_size; btrfs_extend_item(trans, root, path, diff); goto csum; @@ -812,19 +842,14 @@ insert: btrfs_release_path(path); csum_offset = 0; if (found_next) { - u64 tmp = total_bytes + root->sectorsize; - u64 next_sector = sector_sum->bytenr; - struct btrfs_sector_sum *next = sector_sum + 1; + u64 tmp; - while (tmp < sums->len) { - if (next_sector + root->sectorsize != next->bytenr) - break; - tmp += root->sectorsize; - next_sector = next->bytenr; - next++; - } - tmp = min(tmp, next_offset - file_key.offset); + tmp = btrfs_sector_sum_left(sums, sector_sum, total_bytes, + root->sectorsize); tmp >>= root->fs_info->sb->s_blocksize_bits; + tmp = min(tmp, (next_offset - file_key.offset) >> + root->fs_info->sb->s_blocksize_bits); + tmp = max((u64)1, tmp); tmp = min(tmp, (u64)MAX_CSUM_ITEMS(root, csum_size)); ins_size = csum_size * tmp; -- cgit From cb95e7bf7ba481c3d35b238b1cd671b63f54238a Mon Sep 17 00:00:00 2001 From: Mark Fasheh Date: Mon, 4 Feb 2013 20:54:57 +0000 Subject: btrfs: add "no file data" flag to btrfs send ioctl This patch adds the flag, BTRFS_SEND_FLAG_NO_FILE_DATA to the btrfs send ioctl code. When this flag is set, the btrfs send code will never write file data into the stream (thus also avoiding expensive reads of that data in the first place). BTRFS_SEND_C_UPDATE_EXTENT commands will be sent (instead of BTRFS_SEND_C_WRITE) with an offset, length pair indicating the extent in question. This patch does not affect the operation of BTRFS_SEND_C_CLONE commands - they will continue to be sent when a search finds an appropriate extent to clone from. Signed-off-by: Mark Fasheh Signed-off-by: Josef Bacik --- fs/btrfs/send.c | 50 ++++++++++++++++++++++++++++++++++++++++++++++---- fs/btrfs/send.h | 1 + 2 files changed, 47 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index 614da0d44d56..68da757615ae 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -85,6 +85,7 @@ struct send_ctx { u32 send_max_size; u64 total_send_size; u64 cmd_send_size[BTRFS_SEND_C_MAX + 1]; + u64 flags; /* 'flags' member of btrfs_ioctl_send_args is u64 */ struct vfsmount *mnt; @@ -3709,6 +3710,39 @@ out: return ret; } +/* + * Send an update extent command to user space. + */ +static int send_update_extent(struct send_ctx *sctx, + u64 offset, u32 len) +{ + int ret = 0; + struct fs_path *p; + + p = fs_path_alloc(sctx); + if (!p) + return -ENOMEM; + + ret = begin_cmd(sctx, BTRFS_SEND_C_UPDATE_EXTENT); + if (ret < 0) + goto out; + + ret = get_cur_path(sctx, sctx->cur_ino, sctx->cur_inode_gen, p); + if (ret < 0) + goto out; + + TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); + TLV_PUT_U64(sctx, BTRFS_SEND_A_FILE_OFFSET, offset); + TLV_PUT_U64(sctx, BTRFS_SEND_A_SIZE, len); + + ret = send_cmd(sctx); + +tlv_put_failure: +out: + fs_path_free(sctx, p); + return ret; +} + static int send_write_or_clone(struct send_ctx *sctx, struct btrfs_path *path, struct btrfs_key *key, @@ -3744,7 +3778,11 @@ static int send_write_or_clone(struct send_ctx *sctx, goto out; } - if (!clone_root) { + if (clone_root) { + ret = send_clone(sctx, offset, len, clone_root); + } else if (sctx->flags & BTRFS_SEND_FLAG_NO_FILE_DATA) { + ret = send_update_extent(sctx, offset, len); + } else { while (pos < len) { l = len - pos; if (l > BTRFS_SEND_READ_SIZE) @@ -3757,10 +3795,7 @@ static int send_write_or_clone(struct send_ctx *sctx, pos += ret; } ret = 0; - } else { - ret = send_clone(sctx, offset, len, clone_root); } - out: return ret; } @@ -4560,6 +4595,11 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_) goto out; } + if (arg->flags & ~BTRFS_SEND_FLAG_NO_FILE_DATA) { + ret = -EINVAL; + goto out; + } + sctx = kzalloc(sizeof(struct send_ctx), GFP_NOFS); if (!sctx) { ret = -ENOMEM; @@ -4571,6 +4611,8 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_) INIT_RADIX_TREE(&sctx->name_cache, GFP_NOFS); INIT_LIST_HEAD(&sctx->name_cache_list); + sctx->flags = arg->flags; + sctx->send_filp = fget(arg->send_fd); if (IS_ERR(sctx->send_filp)) { ret = PTR_ERR(sctx->send_filp); diff --git a/fs/btrfs/send.h b/fs/btrfs/send.h index 1bf4f32fd4ef..8bb18f7ccaa6 100644 --- a/fs/btrfs/send.h +++ b/fs/btrfs/send.h @@ -86,6 +86,7 @@ enum btrfs_send_cmd { BTRFS_SEND_C_UTIMES, BTRFS_SEND_C_END, + BTRFS_SEND_C_UPDATE_EXTENT, __BTRFS_SEND_C_MAX, }; #define BTRFS_SEND_C_MAX (__BTRFS_SEND_C_MAX - 1) -- cgit From 2b8195bb5717729e4e94ab4ad73a543feaafb0a2 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 7 Feb 2013 06:01:35 +0000 Subject: Btrfs: fix the race between bio and btrfs_stop_workers open_ctree() need read the metadata to initialize the global information of btrfs. But it may fail after it submit some bio, and then it will jump to the error path. Unfortunately, it doesn't check if there are some bios in flight, and just stop all the worker threads. As a result, when the submitted bios end, they can not find any worker thread which can deal with subsequent work, then oops happen. kernel BUG at fs/btrfs/async-thread.c:605! Fix this problem by invoking invalidate_inode_pages2() before we stop the worker threads. This function will wait until the bio end because it need lock the pages which are going to be invalidated, and if a page is under disk read IO, it must be locked. invalidate_inode_pages2() need wait until end bio handler to unlocked it. Reported-and-Tested-by: Tsutomu Itoh Signed-off-by: Eric Sandeen Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 259d7891d10c..9fcce54ecde4 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2711,13 +2711,13 @@ fail_cleaner: * kthreads */ filemap_write_and_wait(fs_info->btree_inode->i_mapping); - invalidate_inode_pages2(fs_info->btree_inode->i_mapping); fail_block_groups: btrfs_free_block_groups(fs_info); fail_tree_roots: free_root_pointers(fs_info, 1); + invalidate_inode_pages2(fs_info->btree_inode->i_mapping); fail_sb_buffer: btrfs_stop_workers(&fs_info->generic_worker); @@ -2738,7 +2738,6 @@ fail_alloc: fail_iput: btrfs_mapping_tree_free(&fs_info->mapping_tree); - invalidate_inode_pages2(fs_info->btree_inode->i_mapping); iput(fs_info->btree_inode); fail_delalloc_bytes: percpu_counter_destroy(&fs_info->delalloc_bytes); -- cgit From 8696c53304f16fde9368b9d5c89a5acb4a815d4c Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 7 Feb 2013 06:02:44 +0000 Subject: Btrfs: fix memory leak of pending_snapshot->inherit The argument "inherit" of btrfs_ioctl_snap_create_transid() was assigned to NULL during we created the snapshots, so we didn't free it though we called kfree() in the caller. But since we are sure the snapshot creation is done after the function - btrfs_ioctl_snap_create_transid() - completes, it is safe that we don't assign the pointer "inherit" to NULL, and just free it in the caller of btrfs_ioctl_snap_create_transid(). In this way, the code can become more readable. Reported-by: Alex Lyakas Cc: Arne Jansen Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index daea831f3d36..6d6314406e27 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -367,7 +367,7 @@ static noinline int create_subvol(struct btrfs_root *root, struct dentry *dentry, char *name, int namelen, u64 *async_transid, - struct btrfs_qgroup_inherit **inherit) + struct btrfs_qgroup_inherit *inherit) { struct btrfs_trans_handle *trans; struct btrfs_key key; @@ -401,8 +401,7 @@ static noinline int create_subvol(struct btrfs_root *root, if (IS_ERR(trans)) return PTR_ERR(trans); - ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid, - inherit ? *inherit : NULL); + ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid, inherit); if (ret) goto fail; @@ -533,7 +532,7 @@ fail: static int create_snapshot(struct btrfs_root *root, struct dentry *dentry, char *name, int namelen, u64 *async_transid, - bool readonly, struct btrfs_qgroup_inherit **inherit) + bool readonly, struct btrfs_qgroup_inherit *inherit) { struct inode *inode; struct btrfs_pending_snapshot *pending_snapshot; @@ -552,10 +551,7 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry, pending_snapshot->dentry = dentry; pending_snapshot->root = root; pending_snapshot->readonly = readonly; - if (inherit) { - pending_snapshot->inherit = *inherit; - *inherit = NULL; /* take responsibility to free it */ - } + pending_snapshot->inherit = inherit; trans = btrfs_start_transaction(root->fs_info->extent_root, 6); if (IS_ERR(trans)) { @@ -695,7 +691,7 @@ static noinline int btrfs_mksubvol(struct path *parent, char *name, int namelen, struct btrfs_root *snap_src, u64 *async_transid, bool readonly, - struct btrfs_qgroup_inherit **inherit) + struct btrfs_qgroup_inherit *inherit) { struct inode *dir = parent->dentry->d_inode; struct dentry *dentry; @@ -1458,7 +1454,7 @@ out: static noinline int btrfs_ioctl_snap_create_transid(struct file *file, char *name, unsigned long fd, int subvol, u64 *transid, bool readonly, - struct btrfs_qgroup_inherit **inherit) + struct btrfs_qgroup_inherit *inherit) { int namelen; int ret = 0; @@ -1567,7 +1563,7 @@ static noinline int btrfs_ioctl_snap_create_v2(struct file *file, ret = btrfs_ioctl_snap_create_transid(file, vol_args->name, vol_args->fd, subvol, ptr, - readonly, &inherit); + readonly, inherit); if (ret == 0 && ptr && copy_to_user(arg + -- cgit From 5d80366e9b5e56b3ffc1923b4995e83bbbf605e3 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 7 Feb 2013 16:06:02 -0500 Subject: Btrfs: steal from global reserve if we are cleaning up orphans Sometimes xfstest 83 will fail to remount the scratch device because we've gotten ourselves so full that we cannot cleanup the orphan items. In this case check to see if we're doing the orphan cleanup and if we are allow us to steal our reservation from the global block rsv. With this patch I've not been able to reproduce the failed mount problem. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 5 +++++ fs/btrfs/extent-tree.c | 11 +++++++++++ fs/btrfs/inode.c | 5 ----- 3 files changed, 16 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 7e2cffd2a5d8..f1cc247f3178 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1237,6 +1237,11 @@ struct seq_list { u64 seq; }; +enum btrfs_orphan_cleanup_state { + ORPHAN_CLEANUP_STARTED = 1, + ORPHAN_CLEANUP_DONE = 2, +}; + /* fs_info */ struct reloc_control; struct btrfs_device; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 81aa7cf3ae86..1818dd90c27e 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -102,6 +102,8 @@ static void dump_space_info(struct btrfs_space_info *info, u64 bytes, int dump_block_groups); static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, u64 num_bytes, int reserve); +static int block_rsv_use_bytes(struct btrfs_block_rsv *block_rsv, + u64 num_bytes); static noinline int block_group_cache_done(struct btrfs_block_group_cache *cache) @@ -4099,6 +4101,15 @@ again: goto again; out: + if (ret == -ENOSPC && + unlikely(root->orphan_cleanup_state == ORPHAN_CLEANUP_STARTED)) { + struct btrfs_block_rsv *global_rsv = + &root->fs_info->global_block_rsv; + + if (block_rsv != global_rsv && + !block_rsv_use_bytes(global_rsv, orig_bytes)) + ret = 0; + } if (flushing) { spin_lock(&space_info->lock); space_info->flush = 0; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 348b7bb920ef..cf26778085e0 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2218,11 +2218,6 @@ void btrfs_run_delayed_iputs(struct btrfs_root *root) } } -enum btrfs_orphan_cleanup_state { - ORPHAN_CLEANUP_STARTED = 1, - ORPHAN_CLEANUP_DONE = 2, -}; - /* * This is called in transaction commit time. If there are no orphan * files in the subvolume, it removes orphan item and frees block_rsv -- cgit From 4a7d0f6854c4a4ad1dba00a3b128a32d39b9a742 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 7 Feb 2013 16:27:28 -0500 Subject: Btrfs: cleanup orphan reservation if truncate fails I noticed we were getting lots of warnings with xfstest 83 because we have reservations outstanding. This is because we moved the orphan add outside of the truncate, but we don't actually cleanup our reservation if something fails. This fixes the problem and I no longer see warnings. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index cf26778085e0..16925807a9ed 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2538,6 +2538,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) goto out; ret = btrfs_truncate(inode); + if (ret) + btrfs_orphan_del(NULL, inode); } else { nr_unlink++; } -- cgit From 0934856d4697e63c14056375e26e3bd6e8ebd34b Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 7 Feb 2013 10:12:07 +0000 Subject: Btrfs: fix deadlock due to unsubmitted The deadlock problem happened when running fsstress(a test program in LTP). Steps to reproduce: # mkfs.btrfs -b 100M # mount # /fsstress -p 3 -n 10000000 -d The reason is: btrfs_direct_IO() |->do_direct_IO() |->get_page() |->get_blocks() | |->btrfs_delalloc_resereve_space() | |->btrfs_add_ordered_extent() ------- Add a new ordered extent |->dio_send_cur_page(page0) -------------- We didn't submit bio here |->get_page() |->get_blocks() |->btrfs_delalloc_resereve_space() |->flush_space() |->btrfs_start_ordered_extent() |->wait_event() ---------- Wait the completion of the ordered extent that is mentioned above But because we didn't submit the bio that is mentioned above, the ordered extent can not complete, we would wait for its completion forever. There are two methods which can fix this deadlock problem: 1. submit the bio before we invoke get_blocks() 2. reserve the space before we do dio Though the 1st is the simplest way, we need modify the code of VFS, and it is likely to break contiguous requests, and introduce performance regression for the other filesystems. So we have to choose the 2nd way. Signed-off-by: Miao Xie Cc: Josef Bacik Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 3 +- fs/btrfs/inode.c | 81 +++++++++++++++++++++++++------------------------- 2 files changed, 43 insertions(+), 41 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 1818dd90c27e..51e1151a0a02 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4748,7 +4748,8 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) spin_lock(&BTRFS_I(inode)->lock); dropped = drop_outstanding_extent(inode); - to_free = calc_csum_metadata_size(inode, num_bytes, 0); + if (num_bytes) + to_free = calc_csum_metadata_size(inode, num_bytes, 0); spin_unlock(&BTRFS_I(inode)->lock); if (dropped > 0) to_free += btrfs_calc_trans_metadata_size(root, dropped); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 16925807a9ed..d11f38d8696c 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6059,16 +6059,15 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, u64 len = bh_result->b_size; struct btrfs_trans_handle *trans; int unlock_bits = EXTENT_LOCKED; - int ret; + int ret = 0; if (create) { - ret = btrfs_delalloc_reserve_space(inode, len); - if (ret) - return ret; + spin_lock(&BTRFS_I(inode)->lock); + BTRFS_I(inode)->outstanding_extents++; + spin_unlock(&BTRFS_I(inode)->lock); unlock_bits |= EXTENT_DELALLOC | EXTENT_DIRTY; - } else { + } else len = min_t(u64, len, root->sectorsize); - } lockstart = start; lockend = start + len - 1; @@ -6080,14 +6079,6 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, if (lock_extent_direct(inode, lockstart, lockend, &cached_state, create)) return -ENOTBLK; - if (create) { - ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, - lockend, EXTENT_DELALLOC, NULL, - &cached_state, GFP_NOFS); - if (ret) - goto unlock_err; - } - em = btrfs_get_extent(inode, NULL, 0, start, len, 0); if (IS_ERR(em)) { ret = PTR_ERR(em); @@ -6119,7 +6110,6 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, if (!create && (em->block_start == EXTENT_MAP_HOLE || test_bit(EXTENT_FLAG_PREALLOC, &em->flags))) { free_extent_map(em); - ret = 0; goto unlock_err; } @@ -6217,6 +6207,11 @@ unlock: */ if (start + len > i_size_read(inode)) i_size_write(inode, start + len); + + ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, + lockstart + len - 1, EXTENT_DELALLOC, NULL, + &cached_state, GFP_NOFS); + BUG_ON(ret); } /* @@ -6225,24 +6220,9 @@ unlock: * aren't using if there is any left over space. */ if (lockstart < lockend) { - if (create && len < lockend - lockstart) { - clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, - lockstart + len - 1, - unlock_bits | EXTENT_DEFRAG, 1, 0, - &cached_state, GFP_NOFS); - /* - * Beside unlock, we also need to cleanup reserved space - * for the left range by attaching EXTENT_DO_ACCOUNTING. - */ - clear_extent_bit(&BTRFS_I(inode)->io_tree, - lockstart + len, lockend, - unlock_bits | EXTENT_DO_ACCOUNTING | - EXTENT_DEFRAG, 1, 0, NULL, GFP_NOFS); - } else { - clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, - lockend, unlock_bits, 1, 0, - &cached_state, GFP_NOFS); - } + clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, + lockend, unlock_bits, 1, 0, + &cached_state, GFP_NOFS); } else { free_extent_state(cached_state); } @@ -6252,9 +6232,6 @@ unlock: return 0; unlock_err: - if (create) - unlock_bits |= EXTENT_DO_ACCOUNTING; - clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend, unlock_bits, 1, 0, &cached_state, GFP_NOFS); return ret; @@ -6692,15 +6669,39 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; + size_t count = 0; + ssize_t ret; if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iov, offset, nr_segs)) return 0; - return __blockdev_direct_IO(rw, iocb, inode, - BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, - iov, offset, nr_segs, btrfs_get_blocks_direct, NULL, - btrfs_submit_direct, 0); + if (rw & WRITE) { + count = iov_length(iov, nr_segs); + ret = btrfs_delalloc_reserve_space(inode, count); + if (ret) + return ret; + } + + ret = __blockdev_direct_IO(rw, iocb, inode, + BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, + iov, offset, nr_segs, btrfs_get_blocks_direct, NULL, + btrfs_submit_direct, 0); + + if (rw & WRITE) { + if (ret < 0 && ret != -EIOCBQUEUED) + btrfs_delalloc_release_space(inode, count); + else if (ret > 0 && (size_t)ret < count) { + spin_lock(&BTRFS_I(inode)->lock); + BTRFS_I(inode)->outstanding_extents++; + spin_unlock(&BTRFS_I(inode)->lock); + btrfs_delalloc_release_space(inode, + count - (size_t)ret); + } + btrfs_delalloc_release_metadata(inode, 0); + } + + return ret; } #define BTRFS_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) -- cgit From 2e60a51e62185cce48758e596ae7cb2da673b58f Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Fri, 8 Feb 2013 07:01:08 +0000 Subject: Btrfs: serialize unlocked dio reads with truncate Currently, we can do unlocked dio reads, but the following race is possible: dio_read_task truncate_task ->btrfs_setattr() ->btrfs_direct_IO ->__blockdev_direct_IO ->btrfs_get_block ->btrfs_truncate() #alloc truncated blocks #to other inode ->submit_io() #INFORMATION LEAK In order to avoid this problem, we must serialize unlocked dio reads with truncate. There are two approaches: - use extent lock to protect the extent that we truncate - use inode_dio_wait() to make sure the truncating task will wait for the read DIO. If we use the 1st one, we will meet the endless truncation problem due to the nonlocked read DIO after we implement the nonlocked write DIO. It is because we still need invoke inode_dio_wait() avoid the race between write DIO and truncation. By that time, we have to introduce btrfs_inode_{block, resume}_nolock_dio() again. That is we have to implement this patch again, so I choose the 2nd way to fix the problem. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/btrfs_inode.h | 19 +++++++++++++++++++ fs/btrfs/inode.c | 23 +++++++++++++++++++++-- 2 files changed, 40 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index c935a774a9f8..d9b97d4960e6 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -41,6 +41,7 @@ #define BTRFS_INODE_NEEDS_FULL_SYNC 7 #define BTRFS_INODE_COPY_EVERYTHING 8 #define BTRFS_INODE_IN_DELALLOC_LIST 9 +#define BTRFS_INODE_READDIO_NEED_LOCK 10 /* in memory btrfs inode */ struct btrfs_inode { @@ -217,4 +218,22 @@ static inline int btrfs_inode_in_log(struct inode *inode, u64 generation) return 0; } +/* + * Disable DIO read nolock optimization, so new dio readers will be forced + * to grab i_mutex. It is used to avoid the endless truncate due to + * nonlocked dio read. + */ +static inline void btrfs_inode_block_unlocked_dio(struct inode *inode) +{ + set_bit(BTRFS_INODE_READDIO_NEED_LOCK, &BTRFS_I(inode)->runtime_flags); + smp_mb(); +} + +static inline void btrfs_inode_resume_unlocked_dio(struct inode *inode) +{ + smp_mb__before_clear_bit(); + clear_bit(BTRFS_INODE_READDIO_NEED_LOCK, + &BTRFS_I(inode)->runtime_flags); +} + #endif diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index d11f38d8696c..c6ee8f1063ff 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3888,6 +3888,12 @@ static int btrfs_setsize(struct inode *inode, struct iattr *attr) /* we don't support swapfiles, so vmtruncate shouldn't fail */ truncate_setsize(inode, newsize); + + /* Disable nonlocked read DIO to avoid the end less truncate */ + btrfs_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); + btrfs_inode_resume_unlocked_dio(inode); + ret = btrfs_truncate(inode); if (ret && inode->i_nlink) btrfs_orphan_del(NULL, inode); @@ -6670,6 +6676,8 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; size_t count = 0; + int flags = 0; + bool wakeup = false; ssize_t ret; if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iov, @@ -6681,13 +6689,22 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, ret = btrfs_delalloc_reserve_space(inode, count); if (ret) return ret; + } else { + atomic_inc(&inode->i_dio_count); + smp_mb__after_atomic_inc(); + if (unlikely(test_bit(BTRFS_INODE_READDIO_NEED_LOCK, + &BTRFS_I(inode)->runtime_flags))) { + inode_dio_done(inode); + flags = DIO_LOCKING | DIO_SKIP_HOLES; + } else { + wakeup = true; + } } ret = __blockdev_direct_IO(rw, iocb, inode, BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, iov, offset, nr_segs, btrfs_get_blocks_direct, NULL, - btrfs_submit_direct, 0); - + btrfs_submit_direct, flags); if (rw & WRITE) { if (ret < 0 && ret != -EIOCBQUEUED) btrfs_delalloc_release_space(inode, count); @@ -6700,6 +6717,8 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, } btrfs_delalloc_release_metadata(inode, 0); } + if (wakeup) + inode_dio_done(inode); return ret; } -- cgit From 38851cc19adbfa1def2b47106d8050a80e0a3673 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Fri, 8 Feb 2013 07:04:11 +0000 Subject: Btrfs: implement unlocked dio write This idea is from ext4. By this patch, we can make the dio write parallel, and improve the performance. But because we can not update isize without i_mutex, the unlocked dio write just can be done in front of the EOF. We needn't worry about the race between dio write and truncate, because the truncate need wait untill all the dio write end. And we also needn't worry about the race between dio write and punch hole, because we have extent lock to protect our operation. I ran fio to test the performance of this feature. == Hardware == CPU: Intel(R) Core(TM)2 Duo CPU E7500 @ 2.93GHz Mem: 2GB SSD: Intel X25-M 120GB (Test Partition: 60GB) == config file == [global] ioengine=psync direct=1 bs=4k size=32G runtime=60 directory=/mnt/btrfs/ filename=testfile group_reporting thread [file1] numjobs=1 # 2 4 rw=randwrite == result (KBps) == write 1 2 4 lock 24936 24738 24726 nolock 24962 30866 32101 == result (iops) == write 1 2 4 lock 6234 6184 6181 nolock 6240 7716 8025 Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 35 +++++++++++++++++++++++------------ 1 file changed, 23 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index c6ee8f1063ff..fce61991213c 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6677,28 +6677,36 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, struct inode *inode = file->f_mapping->host; size_t count = 0; int flags = 0; - bool wakeup = false; + bool wakeup = true; + bool relock = false; ssize_t ret; if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iov, offset, nr_segs)) return 0; + atomic_inc(&inode->i_dio_count); + smp_mb__after_atomic_inc(); + if (rw & WRITE) { count = iov_length(iov, nr_segs); + /* + * If the write DIO is beyond the EOF, we need update + * the isize, but it is protected by i_mutex. So we can + * not unlock the i_mutex at this case. + */ + if (offset + count <= inode->i_size) { + mutex_unlock(&inode->i_mutex); + relock = true; + } ret = btrfs_delalloc_reserve_space(inode, count); if (ret) - return ret; - } else { - atomic_inc(&inode->i_dio_count); - smp_mb__after_atomic_inc(); - if (unlikely(test_bit(BTRFS_INODE_READDIO_NEED_LOCK, - &BTRFS_I(inode)->runtime_flags))) { - inode_dio_done(inode); - flags = DIO_LOCKING | DIO_SKIP_HOLES; - } else { - wakeup = true; - } + goto out; + } else if (unlikely(test_bit(BTRFS_INODE_READDIO_NEED_LOCK, + &BTRFS_I(inode)->runtime_flags))) { + inode_dio_done(inode); + flags = DIO_LOCKING | DIO_SKIP_HOLES; + wakeup = false; } ret = __blockdev_direct_IO(rw, iocb, inode, @@ -6717,8 +6725,11 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, } btrfs_delalloc_release_metadata(inode, 0); } +out: if (wakeup) inode_dio_done(inode); + if (relock) + mutex_lock(&inode->i_mutex); return ret; } -- cgit From b069e0c3450ae388d7a9c94ded6d938a465de262 Mon Sep 17 00:00:00 2001 From: David Sterba Date: Fri, 8 Feb 2013 21:28:17 +0000 Subject: btrfs: put some enospc messages under enospc_debug The warning in use_block_rsv is not useful for users and may fill the logs unnecessarily. Signed-off-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 26 +++++++++++++++----------- 1 file changed, 15 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 51e1151a0a02..88831fab85d6 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -6408,12 +6408,14 @@ use_block_rsv(struct btrfs_trans_handle *trans, if (!ret) return block_rsv; if (ret && !block_rsv->failfast) { - static DEFINE_RATELIMIT_STATE(_rs, - DEFAULT_RATELIMIT_INTERVAL, - /*DEFAULT_RATELIMIT_BURST*/ 2); - if (__ratelimit(&_rs)) - WARN(1, KERN_DEBUG "btrfs: block rsv returned %d\n", - ret); + if (btrfs_test_opt(root, ENOSPC_DEBUG)) { + static DEFINE_RATELIMIT_STATE(_rs, + DEFAULT_RATELIMIT_INTERVAL * 10, + /*DEFAULT_RATELIMIT_BURST*/ 1); + if (__ratelimit(&_rs)) + WARN(1, KERN_DEBUG + "btrfs: block rsv returned %d\n", ret); + } ret = reserve_metadata_bytes(root, block_rsv, blocksize, BTRFS_RESERVE_NO_FLUSH); if (!ret) { @@ -7730,11 +7732,13 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info) space_info = list_entry(info->space_info.next, struct btrfs_space_info, list); - if (space_info->bytes_pinned > 0 || - space_info->bytes_reserved > 0 || - space_info->bytes_may_use > 0) { - WARN_ON(1); - dump_space_info(space_info, 0, 0); + if (btrfs_test_opt(info->tree_root, ENOSPC_DEBUG)) { + if (space_info->bytes_pinned > 0 || + space_info->bytes_reserved > 0 || + space_info->bytes_may_use > 0) { + WARN_ON(1); + dump_space_info(space_info, 0, 0); + } } list_del(&space_info->list); kfree(space_info); -- cgit From 210549ebe9047ae5a8cc47487203d3ee16a7749b Mon Sep 17 00:00:00 2001 From: David Sterba Date: Sat, 9 Feb 2013 23:38:06 +0000 Subject: btrfs: add cancellation points to defrag The defrag operation can take very long, we want to have a way how to cancel it. The code checks for a pending signal at safe points in the defrag loops and returns EAGAIN. This means a user can press ^C after running 'btrfs fi defrag', woks for both defrag modes, files and root. Returning from the command was instant in my light tests, but may take longer depending on the aging factor of the filesystem. Signed-off-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 7 +++++++ fs/btrfs/ioctl.c | 6 ++++++ fs/btrfs/transaction.c | 6 ++++++ 3 files changed, 19 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index f1cc247f3178..14f01dc70ea6 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3786,4 +3786,11 @@ static inline int is_fstree(u64 rootid) return 1; return 0; } + +static inline int btrfs_defrag_cancelled(struct btrfs_fs_info *fs_info) +{ + return signal_pending(current); +} + + #endif diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 6d6314406e27..d02ec577f70f 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1202,6 +1202,12 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, if (!(inode->i_sb->s_flags & MS_ACTIVE)) break; + if (btrfs_defrag_cancelled(root->fs_info)) { + printk(KERN_DEBUG "btrfs: defrag_file cancelled\n"); + ret = -EAGAIN; + break; + } + if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT, extent_thresh, &last_len, &skip, &defrag_end, range->flags & diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 60481a53e004..d574d830a1c4 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -984,6 +984,12 @@ int btrfs_defrag_root(struct btrfs_root *root) if (btrfs_fs_closing(root->fs_info) || ret != -EAGAIN) break; + + if (btrfs_defrag_cancelled(root->fs_info)) { + printk(KERN_DEBUG "btrfs: defrag_root cancelled\n"); + ret = -EAGAIN; + break; + } } root->defrag_running = 0; return ret; -- cgit From c8f2f24bd56a9f0bc7372e6d6aded6fc122b9120 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 11 Feb 2013 11:33:00 -0500 Subject: Btrfs: remove unused extent io tree ops V2 Nobody uses these io tree ops anymore so just remove them and clean up the code a bit. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/extent_io.c | 35 +++++++++++------------------------ fs/btrfs/extent_io.h | 3 --- 2 files changed, 11 insertions(+), 27 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index b9019c847c58..5c00d6aeae75 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -2646,6 +2646,8 @@ static int __extent_read_full_page(struct extent_io_tree *tree, } } while (cur <= end) { + unsigned long pnr = (last_byte >> PAGE_CACHE_SHIFT) + 1; + if (cur >= last_byte) { char *userpage; struct extent_state *cached = NULL; @@ -2733,26 +2735,17 @@ static int __extent_read_full_page(struct extent_io_tree *tree, continue; } - ret = 0; - if (tree->ops && tree->ops->readpage_io_hook) { - ret = tree->ops->readpage_io_hook(page, cur, - cur + iosize - 1); - } - if (!ret) { - unsigned long pnr = (last_byte >> PAGE_CACHE_SHIFT) + 1; - pnr -= page->index; - ret = submit_extent_page(READ, tree, page, + pnr -= page->index; + ret = submit_extent_page(READ, tree, page, sector, disk_io_size, pg_offset, bdev, bio, pnr, end_bio_extent_readpage, mirror_num, *bio_flags, this_bio_flag); - if (!ret) { - nr++; - *bio_flags = this_bio_flag; - } - } - if (ret) { + if (!ret) { + nr++; + *bio_flags = this_bio_flag; + } else { SetPageError(page); unlock_extent(tree, cur, cur + iosize - 1); } @@ -3441,15 +3434,9 @@ retry: * swizzled back from swapper_space to tmpfs file * mapping */ - if (tree->ops && - tree->ops->write_cache_pages_lock_hook) { - tree->ops->write_cache_pages_lock_hook(page, - data, flush_fn); - } else { - if (!trylock_page(page)) { - flush_fn(data); - lock_page(page); - } + if (!trylock_page(page)) { + flush_fn(data); + lock_page(page); } if (unlikely(page->mapping != mapping)) { diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index 2eacfabd3263..ff182322d112 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -75,7 +75,6 @@ struct extent_io_ops { int (*merge_bio_hook)(struct page *page, unsigned long offset, size_t size, struct bio *bio, unsigned long bio_flags); - int (*readpage_io_hook)(struct page *page, u64 start, u64 end); int (*readpage_io_failed_hook)(struct page *page, int failed_mirror); int (*readpage_end_io_hook)(struct page *page, u64 start, u64 end, struct extent_state *state, int mirror); @@ -90,8 +89,6 @@ struct extent_io_ops { struct extent_state *other); void (*split_extent_hook)(struct inode *inode, struct extent_state *orig, u64 split); - int (*write_cache_pages_lock_hook)(struct page *page, void *data, - void (*flush_fn)(void *)); }; struct extent_io_tree { -- cgit From bf023ecfcaf114c410759f375073e78b0871c120 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Tue, 12 Feb 2013 16:27:46 +0000 Subject: Btrfs: eliminate a use-after-free in btrfs_balance() Commit 5af3e8cc introduced a use-after-free at volumes.c:3139: bctl is freed above in __cancel_balance() in all cases except for balance pause. Fix this by moving the offending check a couple statements above, the meaning of the check is preserved. Reported-by: Chris Mason Signed-off-by: Ilya Dryomov Signed-off-by: Josef Bacik --- fs/btrfs/volumes.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 305b6a63ab1f..d778e9666597 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3195,6 +3195,11 @@ int btrfs_balance(struct btrfs_balance_control *bctl, mutex_lock(&fs_info->balance_mutex); atomic_dec(&fs_info->balance_running); + if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) { + fs_info->num_tolerated_disk_barrier_failures = + btrfs_calc_num_tolerated_disk_barrier_failures(fs_info); + } + if (bargs) { memset(bargs, 0, sizeof(*bargs)); update_ioctl_balance_args(fs_info, 0, bargs); @@ -3205,11 +3210,6 @@ int btrfs_balance(struct btrfs_balance_control *bctl, __cancel_balance(fs_info); } - if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) { - fs_info->num_tolerated_disk_barrier_failures = - btrfs_calc_num_tolerated_disk_barrier_failures(fs_info); - } - wake_up(&fs_info->balance_wait_q); return ret; -- cgit From 3e39cea61c4ecb5774c08f3e30507fdd58c1cca9 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Tue, 12 Feb 2013 16:28:59 +0000 Subject: Btrfs: allow for selecting only completely empty chunks Enhance balance usage filter by making it possible to balance out only completely empty chunks. Today, usage filter properly acts on values from 1 to 99 inclusive, usage=100 selects all chunks, and usage=0 selects no chunks. This commit changes the usage=0 case: the new meaning is to restripe only completely empty chunks and nothing else. Suggested-by: David Sterba Signed-off-by: Ilya Dryomov Signed-off-by: Josef Bacik --- fs/btrfs/volumes.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index d778e9666597..567076542d1d 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2679,7 +2679,7 @@ static int chunk_usage_filter(struct btrfs_fs_info *fs_info, u64 chunk_offset, chunk_used = btrfs_block_group_used(&cache->item); if (bargs->usage == 0) - user_thresh = 0; + user_thresh = 1; else if (bargs->usage > 100) user_thresh = cache->key.offset; else -- cgit From dde5740fdd6175fc95aecf4ccc7856fbbad9b44e Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 12 Feb 2013 14:07:51 -0500 Subject: Btrfs: relax the block group size limit for bitmaps Dave pointed out that xfstests 273 will tell you that it failed to load the space cache for a block group when it remounts. This is because we run out of space writing out the block group cache. This is ok and is working as it should, but let's try to be a bit nicer. This happens because the block group was 100mb, but bitmap entries cover 128mb, so we were only getting extent entries for this block group, which ended up being too many to fit in the free space cache. So relax the bitmap size requirements to block groups that are at least half the size a bitmap will cover or larger, that way we can still keep the amount of space used in the free space cache low enough to be able to write it out. With this patch I no longer fail to write out the free space cache. Thanks, Reported-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/free-space-cache.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 0be7a8742a43..c8090f18c217 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -1356,6 +1356,8 @@ static void recalculate_thresholds(struct btrfs_free_space_ctl *ctl) u64 bytes_per_bg = BITS_PER_BITMAP * ctl->unit; int max_bitmaps = div64_u64(size + bytes_per_bg - 1, bytes_per_bg); + max_bitmaps = max(max_bitmaps, 1); + BUG_ON(ctl->total_bitmaps > max_bitmaps); /* @@ -1636,10 +1638,14 @@ static bool use_bitmap(struct btrfs_free_space_ctl *ctl, } /* - * some block groups are so tiny they can't be enveloped by a bitmap, so - * don't even bother to create a bitmap for this + * The original block groups from mkfs can be really small, like 8 + * megabytes, so don't bother with a bitmap for those entries. However + * some block groups can be smaller than what a bitmap would cover but + * are still large enough that they could overflow the 32k memory limit, + * so allow those block groups to still be allowed to have a bitmap + * entry. */ - if (BITS_PER_BITMAP * ctl->unit > block_group->key.offset) + if (((BITS_PER_BITMAP * ctl->unit) >> 1) > block_group->key.offset) return false; return true; -- cgit From 569e0f358c0c37f6733702d4a5d2c412860f7169 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 13 Feb 2013 11:09:14 -0500 Subject: Btrfs: place ordered operations on a per transaction list Miao made the ordered operations stuff run async, which introduced a deadlock where we could get somebody (sync) racing in and committing the transaction while a commit was already happening. The new committer would try and flush ordered operations which would hang waiting for the commit to finish because it is done asynchronously and no longer inherits the callers trans handle. To fix this we need to make the ordered operations list a per transaction list. We can get new inodes added to the ordered operation list by truncating them and then having another process writing to them, so this makes it so that anybody trying to add an ordered operation _must_ start a transaction in order to add itself to the list, which will keep new inodes from getting added to the ordered operations list after we start committing. This should fix the deadlock and also keeps us from doing a lot more work than we need to during commit. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 7 ------- fs/btrfs/disk-io.c | 11 ++++++----- fs/btrfs/file.c | 15 ++++++++++++++- fs/btrfs/ordered-data.c | 13 ++++++++----- fs/btrfs/ordered-data.h | 3 ++- fs/btrfs/transaction.c | 5 +++-- fs/btrfs/transaction.h | 1 + 7 files changed, 34 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 14f01dc70ea6..961ff2986341 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1407,13 +1407,6 @@ struct btrfs_fs_info { */ struct list_head delalloc_inodes; - /* - * special rename and truncate targets that must be on disk before - * we're allowed to commit. This is basically the ext3 style - * data=ordered list. - */ - struct list_head ordered_operations; - /* * there is a pool of worker threads for checksumming during writes * and a pool for checksumming after reads. This is because readers diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 9fcce54ecde4..e511d9f78c19 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -56,7 +56,8 @@ static void end_workqueue_fn(struct btrfs_work *work); static void free_fs_root(struct btrfs_root *root); static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info, int read_only); -static void btrfs_destroy_ordered_operations(struct btrfs_root *root); +static void btrfs_destroy_ordered_operations(struct btrfs_transaction *t, + struct btrfs_root *root); static void btrfs_destroy_ordered_extents(struct btrfs_root *root); static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, struct btrfs_root *root); @@ -2029,7 +2030,6 @@ int open_ctree(struct super_block *sb, INIT_LIST_HEAD(&fs_info->dead_roots); INIT_LIST_HEAD(&fs_info->delayed_iputs); INIT_LIST_HEAD(&fs_info->delalloc_inodes); - INIT_LIST_HEAD(&fs_info->ordered_operations); INIT_LIST_HEAD(&fs_info->caching_block_groups); spin_lock_init(&fs_info->delalloc_lock); spin_lock_init(&fs_info->trans_lock); @@ -3538,7 +3538,8 @@ void btrfs_error_commit_super(struct btrfs_root *root) btrfs_cleanup_transaction(root); } -static void btrfs_destroy_ordered_operations(struct btrfs_root *root) +static void btrfs_destroy_ordered_operations(struct btrfs_transaction *t, + struct btrfs_root *root) { struct btrfs_inode *btrfs_inode; struct list_head splice; @@ -3548,7 +3549,7 @@ static void btrfs_destroy_ordered_operations(struct btrfs_root *root) mutex_lock(&root->fs_info->ordered_operations_mutex); spin_lock(&root->fs_info->ordered_extent_lock); - list_splice_init(&root->fs_info->ordered_operations, &splice); + list_splice_init(&t->ordered_operations, &splice); while (!list_empty(&splice)) { btrfs_inode = list_entry(splice.next, struct btrfs_inode, ordered_operations); @@ -3829,7 +3830,7 @@ int btrfs_cleanup_transaction(struct btrfs_root *root) while (!list_empty(&list)) { t = list_entry(list.next, struct btrfs_transaction, list); - btrfs_destroy_ordered_operations(root); + btrfs_destroy_ordered_operations(t, root); btrfs_destroy_ordered_extents(root); diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 75d0fe134be3..b12ba52c4505 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1628,7 +1628,20 @@ int btrfs_release_file(struct inode *inode, struct file *filp) */ if (test_and_clear_bit(BTRFS_INODE_ORDERED_DATA_CLOSE, &BTRFS_I(inode)->runtime_flags)) { - btrfs_add_ordered_operation(NULL, BTRFS_I(inode)->root, inode); + struct btrfs_trans_handle *trans; + struct btrfs_root *root = BTRFS_I(inode)->root; + + /* + * We need to block on a committing transaction to keep us from + * throwing a ordered operation on to the list and causing + * something like sync to deadlock trying to flush out this + * inode. + */ + trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) + return PTR_ERR(trans); + btrfs_add_ordered_operation(trans, BTRFS_I(inode)->root, inode); + btrfs_end_transaction(trans, root); if (inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT) filemap_flush(inode->i_mapping); } diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index 9489fa96e3ed..dc08d77b717e 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -612,10 +612,12 @@ void btrfs_wait_ordered_extents(struct btrfs_root *root, int delay_iput) * extra check to make sure the ordered operation list really is empty * before we return */ -int btrfs_run_ordered_operations(struct btrfs_root *root, int wait) +int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans, + struct btrfs_root *root, int wait) { struct btrfs_inode *btrfs_inode; struct inode *inode; + struct btrfs_transaction *cur_trans = trans->transaction; struct list_head splice; struct list_head works; struct btrfs_delalloc_work *work, *next; @@ -626,7 +628,7 @@ int btrfs_run_ordered_operations(struct btrfs_root *root, int wait) mutex_lock(&root->fs_info->ordered_operations_mutex); spin_lock(&root->fs_info->ordered_extent_lock); - list_splice_init(&root->fs_info->ordered_operations, &splice); + list_splice_init(&cur_trans->ordered_operations, &splice); while (!list_empty(&splice)) { btrfs_inode = list_entry(splice.next, struct btrfs_inode, ordered_operations); @@ -643,7 +645,7 @@ int btrfs_run_ordered_operations(struct btrfs_root *root, int wait) if (!wait) list_add_tail(&BTRFS_I(inode)->ordered_operations, - &root->fs_info->ordered_operations); + &cur_trans->ordered_operations); spin_unlock(&root->fs_info->ordered_extent_lock); work = btrfs_alloc_delalloc_work(inode, wait, 1); @@ -653,7 +655,7 @@ int btrfs_run_ordered_operations(struct btrfs_root *root, int wait) list_add_tail(&btrfs_inode->ordered_operations, &splice); list_splice_tail(&splice, - &root->fs_info->ordered_operations); + &cur_trans->ordered_operations); spin_unlock(&root->fs_info->ordered_extent_lock); ret = -ENOMEM; goto out; @@ -1033,6 +1035,7 @@ out: void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct inode *inode) { + struct btrfs_transaction *cur_trans = trans->transaction; u64 last_mod; last_mod = max(BTRFS_I(inode)->generation, BTRFS_I(inode)->last_trans); @@ -1047,7 +1050,7 @@ void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans, spin_lock(&root->fs_info->ordered_extent_lock); if (list_empty(&BTRFS_I(inode)->ordered_operations)) { list_add_tail(&BTRFS_I(inode)->ordered_operations, - &root->fs_info->ordered_operations); + &cur_trans->ordered_operations); } spin_unlock(&root->fs_info->ordered_extent_lock); } diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h index d523dbd2314d..267ac99095f6 100644 --- a/fs/btrfs/ordered-data.h +++ b/fs/btrfs/ordered-data.h @@ -197,7 +197,8 @@ struct btrfs_ordered_extent *btrfs_lookup_ordered_range(struct inode *inode, int btrfs_ordered_update_i_size(struct inode *inode, u64 offset, struct btrfs_ordered_extent *ordered); int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr, u32 *sum); -int btrfs_run_ordered_operations(struct btrfs_root *root, int wait); +int btrfs_run_ordered_operations(struct btrfs_trans_handle *trans, + struct btrfs_root *root, int wait); void btrfs_add_ordered_operation(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct inode *inode); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index d574d830a1c4..0c87d18d1881 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -157,6 +157,7 @@ loop: spin_lock_init(&cur_trans->delayed_refs.lock); INIT_LIST_HEAD(&cur_trans->pending_snapshots); + INIT_LIST_HEAD(&cur_trans->ordered_operations); list_add_tail(&cur_trans->list, &fs_info->trans_list); extent_io_tree_init(&cur_trans->dirty_pages, fs_info->btree_inode->i_mapping); @@ -1456,7 +1457,7 @@ static int btrfs_flush_all_pending_stuffs(struct btrfs_trans_handle *trans, * it here and no for sure that nothing new will be added * to the list */ - ret = btrfs_run_ordered_operations(root, 1); + ret = btrfs_run_ordered_operations(trans, root, 1); return ret; } @@ -1479,7 +1480,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, int should_grow = 0; unsigned long now = get_seconds(); - ret = btrfs_run_ordered_operations(root, 0); + ret = btrfs_run_ordered_operations(trans, root, 0); if (ret) { btrfs_abort_transaction(trans, root, ret); btrfs_end_transaction(trans, root); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 46628210e5d8..3f772fd0191a 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -43,6 +43,7 @@ struct btrfs_transaction { wait_queue_head_t writer_wait; wait_queue_head_t commit_wait; struct list_head pending_snapshots; + struct list_head ordered_operations; struct btrfs_delayed_ref_root delayed_refs; int aborted; }; -- cgit From 867ab667e74377160c4a683375ee5b8bf8801724 Mon Sep 17 00:00:00 2001 From: "jeff.liu" Date: Sat, 5 Jan 2013 02:48:01 +0000 Subject: Btrfs: Add a new ioctl to get the label of a mounted file system Add a new ioctl(2) BTRFS_IOC_GET_FSLABLE, so that we can get the label upon a mounted filesystem. Signed-off-by: Jie Liu Signed-off-by: Anand Jain Cc: Miao Xie Cc: Goffredo Baroncelli Cc: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index d02ec577f70f..fcc15a6804a9 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3917,6 +3917,25 @@ out: return ret; } +static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg) +{ + struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + const char *label = root->fs_info->super_copy->label; + size_t len = strnlen(label, BTRFS_LABEL_SIZE); + int ret; + + if (len == BTRFS_LABEL_SIZE) { + pr_warn("btrfs: label is too long, return the first %zu bytes\n", + --len); + } + + mutex_lock(&root->fs_info->volume_mutex); + ret = copy_to_user(arg, label, len); + mutex_unlock(&root->fs_info->volume_mutex); + + return ret ? -EFAULT : 0; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -4017,6 +4036,8 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_qgroup_limit(file, argp); case BTRFS_IOC_DEV_REPLACE: return btrfs_ioctl_dev_replace(root, argp); + case BTRFS_IOC_GET_FSLABEL: + return btrfs_ioctl_get_fslabel(file, argp); } return -ENOTTY; -- cgit From a8bfd4abea3da0e28f215e2a2b8c2f1ca27ebe80 Mon Sep 17 00:00:00 2001 From: "jeff.liu" Date: Sat, 5 Jan 2013 02:48:08 +0000 Subject: Btrfs: set/change the label of a mounted file system With this new ioctl(2) BTRFS_IOC_SET_FSLABEL, we can set/change the label of a mounted file system. Signed-off-by: Jie Liu Signed-off-by: Anand Jain Reviewed-by: Miao Xie Reviewed-by: Goffredo Baroncelli Reviewed-by: David Sterba Reviewed-by: Goffredo Baroncelli Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 42 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 42 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index fcc15a6804a9..0f68729f261e 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3936,6 +3936,46 @@ static int btrfs_ioctl_get_fslabel(struct file *file, void __user *arg) return ret ? -EFAULT : 0; } +static int btrfs_ioctl_set_fslabel(struct file *file, void __user *arg) +{ + struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_super_block *super_block = root->fs_info->super_copy; + struct btrfs_trans_handle *trans; + char label[BTRFS_LABEL_SIZE]; + int ret; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (copy_from_user(label, arg, sizeof(label))) + return -EFAULT; + + if (strnlen(label, BTRFS_LABEL_SIZE) == BTRFS_LABEL_SIZE) { + pr_err("btrfs: unable to set label with more than %d bytes\n", + BTRFS_LABEL_SIZE - 1); + return -EINVAL; + } + + ret = mnt_want_write_file(file); + if (ret) + return ret; + + mutex_lock(&root->fs_info->volume_mutex); + trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto out_unlock; + } + + strcpy(super_block->label, label); + ret = btrfs_end_transaction(trans, root); + +out_unlock: + mutex_unlock(&root->fs_info->volume_mutex); + mnt_drop_write_file(file); + return ret; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -4038,6 +4078,8 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_dev_replace(root, argp); case BTRFS_IOC_GET_FSLABEL: return btrfs_ioctl_get_fslabel(file, argp); + case BTRFS_IOC_SET_FSLABEL: + return btrfs_ioctl_set_fslabel(file, argp); } return -ENOTTY; -- cgit From cdb4c5748cb3ac533889a6b0b95aa10651e68785 Mon Sep 17 00:00:00 2001 From: Zach Brown Date: Wed, 20 Feb 2013 00:55:13 +0000 Subject: btrfs: define BTRFS_MAGIC as a u64 value super.magic is an le64 but it's treated as an unterminated string when compared against BTRFS_MAGIC which is defined as a string. Instead define BTRFS_MAGIC as a normal hex value and use endian helpers to compare it to the super's magic. I tested this by mounting an fs made before the change and made sure that it didn't introduce sparse errors. This matches a similar cleanup that is pending in btrfs-progs. David Sterba pointed out that we should fix the kernel side as well :). Signed-off-by: Zach Brown Signed-off-by: Josef Bacik --- fs/btrfs/check-integrity.c | 3 +-- fs/btrfs/ctree.h | 2 +- fs/btrfs/disk-io.c | 6 ++---- fs/btrfs/volumes.c | 3 +-- 4 files changed, 5 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/check-integrity.c b/fs/btrfs/check-integrity.c index 11d47bfb62b4..18af6f48781a 100644 --- a/fs/btrfs/check-integrity.c +++ b/fs/btrfs/check-integrity.c @@ -813,8 +813,7 @@ static int btrfsic_process_superblock_dev_mirror( (bh->b_data + (dev_bytenr & 4095)); if (btrfs_super_bytenr(super_tmp) != dev_bytenr || - strncmp((char *)(&(super_tmp->magic)), BTRFS_MAGIC, - sizeof(super_tmp->magic)) || + super_tmp->magic != cpu_to_le64(BTRFS_MAGIC) || memcmp(device->uuid, super_tmp->dev_item.uuid, BTRFS_UUID_SIZE) || btrfs_super_nodesize(super_tmp) != state->metablock_size || btrfs_super_leafsize(super_tmp) != state->metablock_size || diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 961ff2986341..1679051f4d39 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -46,7 +46,7 @@ extern struct kmem_cache *btrfs_path_cachep; extern struct kmem_cache *btrfs_free_space_cachep; struct btrfs_ordered_sum; -#define BTRFS_MAGIC "_BHRfS_M" +#define BTRFS_MAGIC 0x4D5F53665248425FULL /* ascii _BHRfS_M, no null */ #define BTRFS_MAX_MIRRORS 3 diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index e511d9f78c19..39ff34a62a24 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2407,8 +2407,7 @@ int open_ctree(struct super_block *sb, sb->s_blocksize = sectorsize; sb->s_blocksize_bits = blksize_bits(sectorsize); - if (strncmp((char *)(&disk_super->magic), BTRFS_MAGIC, - sizeof(disk_super->magic))) { + if (disk_super->magic != cpu_to_le64(BTRFS_MAGIC)) { printk(KERN_INFO "btrfs: valid FS not found on %s\n", sb->s_id); goto fail_sb_buffer; } @@ -2815,8 +2814,7 @@ struct buffer_head *btrfs_read_dev_super(struct block_device *bdev) super = (struct btrfs_super_block *)bh->b_data; if (btrfs_super_bytenr(super) != bytenr || - strncmp((char *)(&super->magic), BTRFS_MAGIC, - sizeof(super->magic))) { + super->magic != cpu_to_le64(BTRFS_MAGIC)) { brelse(bh); continue; } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 567076542d1d..72b1cf1b2b5e 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -855,8 +855,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, disk_super = p + (bytenr & ~PAGE_CACHE_MASK); if (btrfs_super_bytenr(disk_super) != bytenr || - strncmp((char *)(&disk_super->magic), BTRFS_MAGIC, - sizeof(disk_super->magic))) + disk_super->magic != cpu_to_le64(BTRFS_MAGIC)) goto error_unmap; devid = btrfs_stack_device_id(&disk_super->dev_item); -- cgit From 4b82490649f5c8ecbf888752c325ea68831c497e Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 20 Feb 2013 09:13:32 +0000 Subject: Btrfs: fix the qgroup reserved space is released prematurely In start_transactio(), we will try to join the transaction again after the current transaction is committed, so we should not release the reserved space of the qgroup. Fix it. Cc: Arne Jansen Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 0c87d18d1881..425d5b57d377 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -383,7 +383,7 @@ again: h->block_rsv = NULL; h->orig_rsv = NULL; h->aborted = 0; - h->qgroup_reserved = qgroup_reserved; + h->qgroup_reserved = 0; h->delayed_ref_elem.seq = 0; h->type = type; h->allocating_chunk = false; @@ -402,6 +402,7 @@ again: h->block_rsv = &root->fs_info->trans_block_rsv; h->bytes_reserved = num_bytes; } + h->qgroup_reserved = qgroup_reserved; got_it: btrfs_record_root_in_trans(h, root); -- cgit From 178260b2c14969f29ba39a78df74ed485abc6203 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 20 Feb 2013 09:16:24 +0000 Subject: Btrfs: fix the deadlock between the transaction start/attach and commit Now btrfs_commit_transaction() does this ret = btrfs_run_ordered_operations(root, 0) which async flushes all inodes on the ordered operations list, it introduced a deadlock that transaction-start task, transaction-commit task and the flush workers waited for each other. (See the following URL to get the detail http://marc.info/?l=linux-btrfs&m=136070705732646&w=2) As we know, if ->in_commit is set, it means someone is committing the current transaction, we should not try to join it if we are not JOIN or JOIN_NOLOCK, wait is the best choice for it. In this way, we can avoid the above problem. In this way, there is another benefit: there is no new transaction handle to block the transaction which is on the way of commit, once we set ->in_commit. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 425d5b57d377..5767ea1c0150 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -50,6 +50,14 @@ static noinline void switch_commit_root(struct btrfs_root *root) root->commit_root = btrfs_root_node(root); } +static inline int can_join_transaction(struct btrfs_transaction *trans, + int type) +{ + return !(trans->in_commit && + type != TRANS_JOIN && + type != TRANS_JOIN_NOLOCK); +} + /* * either allocate a new transaction or hop into the existing one */ @@ -85,6 +93,10 @@ loop: spin_unlock(&fs_info->trans_lock); return cur_trans->aborted; } + if (!can_join_transaction(cur_trans, type)) { + spin_unlock(&fs_info->trans_lock); + return -EBUSY; + } atomic_inc(&cur_trans->use_count); atomic_inc(&cur_trans->num_writers); cur_trans->num_joined++; @@ -360,8 +372,11 @@ again: do { ret = join_transaction(root, type); - if (ret == -EBUSY) + if (ret == -EBUSY) { wait_current_trans(root); + if (unlikely(type == TRANS_ATTACH)) + ret = -ENOENT; + } } while (ret == -EBUSY); if (ret < 0) { -- cgit From d4edf39bd5db443151efc993dac67ec9d6b5b8c1 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 20 Feb 2013 09:17:06 +0000 Subject: Btrfs: fix uncompleted transaction In some cases, we need commit the current transaction, but don't want to start a new one if there is no running transaction, so we introduce the function - btrfs_attach_transaction(), which can catch the current transaction, and return -ENOENT if there is no running transaction. But no running transaction doesn't mean the current transction completely, because we removed the running transaction before it completes. In some cases, it doesn't matter. But in some special cases, such as freeze fs, we hope the transaction is fully on disk, it will introduce some bugs, for example, we may feeze the fs and dump the data in the disk, if the transction doesn't complete, we would dump inconsistent data. So we need fix the above problem for those cases. We fixes this problem by introducing a function: btrfs_attach_transaction_barrier() if we hope all the transaction is fully on the disk, even they are not running, we can use this function. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 2 +- fs/btrfs/super.c | 4 ++-- fs/btrfs/transaction.c | 32 ++++++++++++++++++++++++++++++++ fs/btrfs/transaction.h | 2 ++ 4 files changed, 37 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 0f68729f261e..e0893036da08 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3114,7 +3114,7 @@ static noinline long btrfs_ioctl_start_sync(struct btrfs_root *root, u64 transid; int ret; - trans = btrfs_attach_transaction(root); + trans = btrfs_attach_transaction_barrier(root); if (IS_ERR(trans)) { if (PTR_ERR(trans) != -ENOENT) return PTR_ERR(trans); diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index e0ea1278377a..db1ba9a2ed64 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -878,7 +878,7 @@ int btrfs_sync_fs(struct super_block *sb, int wait) btrfs_wait_ordered_extents(root, 0); - trans = btrfs_attach_transaction(root); + trans = btrfs_attach_transaction_barrier(root); if (IS_ERR(trans)) { /* no transaction, don't bother */ if (PTR_ERR(trans) == -ENOENT) @@ -1563,7 +1563,7 @@ static int btrfs_freeze(struct super_block *sb) struct btrfs_trans_handle *trans; struct btrfs_root *root = btrfs_sb(sb)->tree_root; - trans = btrfs_attach_transaction(root); + trans = btrfs_attach_transaction_barrier(root); if (IS_ERR(trans)) { /* no transaction, don't bother */ if (PTR_ERR(trans) == -ENOENT) diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 5767ea1c0150..c1ce664c0c39 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -469,11 +469,43 @@ struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *root return start_transaction(root, 0, TRANS_USERSPACE, 0); } +/* + * btrfs_attach_transaction() - catch the running transaction + * + * It is used when we want to commit the current the transaction, but + * don't want to start a new one. + * + * Note: If this function return -ENOENT, it just means there is no + * running transaction. But it is possible that the inactive transaction + * is still in the memory, not fully on disk. If you hope there is no + * inactive transaction in the fs when -ENOENT is returned, you should + * invoke + * btrfs_attach_transaction_barrier() + */ struct btrfs_trans_handle *btrfs_attach_transaction(struct btrfs_root *root) { return start_transaction(root, 0, TRANS_ATTACH, 0); } +/* + * btrfs_attach_transaction() - catch the running transaction + * + * It is similar to the above function, the differentia is this one + * will wait for all the inactive transactions until they fully + * complete. + */ +struct btrfs_trans_handle * +btrfs_attach_transaction_barrier(struct btrfs_root *root) +{ + struct btrfs_trans_handle *trans; + + trans = start_transaction(root, 0, TRANS_ATTACH, 0); + if (IS_ERR(trans) && PTR_ERR(trans) == -ENOENT) + btrfs_wait_for_commit(root, 0); + + return trans; +} + /* wait for a transaction commit to be fully complete */ static noinline void wait_for_commit(struct btrfs_root *root, struct btrfs_transaction *commit) diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 3f772fd0191a..5afd7b1dceac 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -112,6 +112,8 @@ struct btrfs_trans_handle *btrfs_start_transaction_lflush( struct btrfs_trans_handle *btrfs_join_transaction(struct btrfs_root *root); struct btrfs_trans_handle *btrfs_join_transaction_nolock(struct btrfs_root *root); struct btrfs_trans_handle *btrfs_attach_transaction(struct btrfs_root *root); +struct btrfs_trans_handle *btrfs_attach_transaction_barrier( + struct btrfs_root *root); struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *root); int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid); int btrfs_write_and_wait_transaction(struct btrfs_trans_handle *trans, -- cgit From fa6ac8765c48a06dfed914e8c8c3a903f9d313a0 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Wed, 20 Feb 2013 14:10:23 +0000 Subject: Btrfs: fix cleaner thread not working with inode cache option Right now inode cache inode is treated as the same as space cache inode, ie. keep inode in memory till putting super. But this leads to an awkward situation. If we're going to delete a snapshot/subvolume, btrfs will not actually delete it and return free space, but will add it to dead roots list until the last inode on this snap/subvol being destroyed. Then we'll fetch deleted roots and cleanup them via cleaner thread. So here is the problem, if we enable inode cache option, each snap/subvol has a cached inode which is used to store inode allcation information. And this cache inode will be kept in memory, as the above said. So with inode cache, snap/subvol can only be added into dead roots list during freeing roots stage in umount, so that we can ONLY get space back after another remount(we cleanup dead roots on mount). But the real thing is we'll no more use the snap/subvol if we mark it deleted, so we can safely iput its cache inode when we delete snap/subvol. Another thing is that we need to change the rules of droping inode, we don't keep snap/subvol's cache inode in memory till end so that we can add snap/subvol into dead roots list in time. Reported-by: Mitch Harder Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 3 ++- fs/btrfs/ioctl.c | 6 ++++++ 2 files changed, 8 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index fce61991213c..b009fb52bd8a 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7315,8 +7315,9 @@ int btrfs_drop_inode(struct inode *inode) { struct btrfs_root *root = BTRFS_I(inode)->root; + /* the snap/subvol tree is on deleting */ if (btrfs_root_refs(&root->root_item) == 0 && - !btrfs_is_free_space_inode(inode)) + root != root->fs_info->tree_root) return 1; else return generic_drop_inode(inode); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index e0893036da08..2bbbed506f5b 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2174,6 +2174,12 @@ out_unlock: shrink_dcache_sb(root->fs_info->sb); btrfs_invalidate_inodes(dest); d_delete(dentry); + + /* the last ref */ + if (dest->cache_inode) { + iput(dest->cache_inode); + dest->cache_inode = NULL; + } } out_dput: dput(dentry); -- cgit From 683cebda90787730d5260eac0da7c2bff9fb40ef Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Wed, 20 Feb 2013 14:16:14 +0000 Subject: Btrfs: fix missing check before disabling quota The original code forget to check whether quota has been disabled firstly, and it will return 'EINVAL' and return error to users if quota has been disabled,it will be unfriendly and confusing for users to see that. So just return directly if quota has been disabled. Signed-off-by: Wang Shilong Cc: Arne Jansen Signed-off-by: Josef Bacik --- fs/btrfs/qgroup.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index a0d6368249fa..88ab785bbd73 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -847,6 +847,10 @@ int btrfs_quota_disable(struct btrfs_trans_handle *trans, int ret = 0; spin_lock(&fs_info->qgroup_lock); + if (!fs_info->quota_root) { + spin_unlock(&fs_info->qgroup_lock); + return 0; + } fs_info->quota_enabled = 0; fs_info->pending_quota_state = 0; quota_root = fs_info->quota_root; -- cgit From 272d26d0ad8c0e326689f2fa3cdc6a5fcc8e74e0 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 20 Feb 2013 14:16:39 +0000 Subject: Btrfs: fix missing release of qgroup reservation in commit_transaction() We forget to free qgroup reservation in commit_transaction(),fix it. Signed-off-by: Miao Xie Signed-off-by: Wang Shilong Cc: Arne Jansen Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index c1ce664c0c39..955204ca0447 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1553,6 +1553,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, btrfs_trans_release_metadata(trans, root); trans->block_rsv = NULL; + if (trans->qgroup_reserved) { + btrfs_qgroup_free(root, trans->qgroup_reserved); + trans->qgroup_reserved = 0; + } cur_trans = trans->transaction; @@ -1833,6 +1837,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, cleanup_transaction: btrfs_trans_release_metadata(trans, root); trans->block_rsv = NULL; + if (trans->qgroup_reserved) { + btrfs_qgroup_free(root, trans->qgroup_reserved); + trans->qgroup_reserved = 0; + } btrfs_printk(root->fs_info, "Skipping commit of aborted transaction.\n"); // WARN_ON(1); if (current->journal_info == trans) -- cgit From 1cba0cdf5e4dbcd9e5fa5b54d7a028e55e2ca057 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Wed, 20 Feb 2013 14:06:20 -0500 Subject: btrfs: Init io_lock after cloning btrfs device struct __btrfs_close_devices() clones btrfs device structs with memcpy(). Some of the fields in the clone are reinitialized, but it's missing to init io_lock. In mainline this goes unnoticed, but on RT it leaves the plist pointing to the original about to be freed lock struct. Initialize io_lock after cloning, so no references to the original struct are left. Reported-and-tested-by: Mike Galbraith Cc: stable@vger.kernel.org Signed-off-by: Thomas Gleixner Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 7992dc4ea4cc..5d6010ba8b7e 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -650,6 +650,7 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices) new_device->writeable = 0; new_device->in_fs_metadata = 0; new_device->can_discard = 0; + spin_lock_init(&new_device->io_lock); list_replace_rcu(&device->dev_list, &new_device->dev_list); call_rcu(&device->rcu, free_device); -- cgit From 24542bf7ea5e4fdfdb5157ff544c093fa4dcb536 Mon Sep 17 00:00:00 2001 From: Zach Brown Date: Fri, 16 Nov 2012 00:04:43 +0000 Subject: btrfs: limit fallocate extent reservation to 256MB Very large fallocate requests are cpu bound and result in extents with a repeating pattern of ever decreasing size: $ time fallocate -l 1T file real 0m13.039s ( an excerpt of the extents from btrfs-debug-tree: ) prealloc data disk byte 1536292564992 nr 397312 prealloc data disk byte 1536292962304 nr 196608 prealloc data disk byte 1536293158912 nr 98304 prealloc data disk byte 1536293257216 nr 49152 prealloc data disk byte 1536293306368 nr 24576 prealloc data disk byte 1536293330944 nr 12288 prealloc data disk byte 1536293343232 nr 8192 prealloc data disk byte 1536293351424 nr 4096 prealloc data disk byte 1536293355520 nr 4096 prealloc data disk byte 1536293359616 nr 4096 The excessive cpu use comes from __btrfs_prealloc_file_range() trying to allocate the entire remaining size after each extent is allocated. btrfs_reserve_extent() repeatedly cuts this requested size in half until it gets down to the size that the allocators can return. We limit the problem for now by capping each reservation at 256 meg. The small extents come from a masking bug when decreasing the requested reservation size. The high 32bits are cleared and the remaining low bits might happen to reserve a small size. Fix this by using round_down() which properly casts the mask. After these fixes huge fallocate requests are fast and result in nice large extents: $ time fallocate -l 1T file real 0m0.082s prealloc data disk byte 1112425889792 nr 268435456 prealloc data disk byte 1112694325248 nr 268435456 prealloc data disk byte 1112962760704 nr 268435456 Reported-by: Eric Sandeen Signed-off-by: Zach Brown Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 2 +- fs/btrfs/inode.c | 5 +++-- 2 files changed, 4 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index b3ecca447ddf..d2b3a5e9a621 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -6143,7 +6143,7 @@ again: if (ret == -ENOSPC) { if (!final_tried) { num_bytes = num_bytes >> 1; - num_bytes = num_bytes & ~(root->sectorsize - 1); + num_bytes = round_down(num_bytes, root->sectorsize); num_bytes = max(num_bytes, min_alloc_size); if (num_bytes == min_alloc_size) final_tried = true; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 4e6a11c2cfdd..3bc62b181ef8 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7894,8 +7894,9 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode, } } - ret = btrfs_reserve_extent(trans, root, num_bytes, min_size, - 0, *alloc_hint, &ins, 1); + ret = btrfs_reserve_extent(trans, root, + min(num_bytes, 256ULL * 1024 * 1024), + min_size, 0, *alloc_hint, &ins, 1); if (ret) { if (own_trans) btrfs_end_transaction(trans, root); -- cgit From 86db25785a6e7cacc1ee868fe437e8a2957eae94 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Wed, 20 Feb 2013 16:23:40 -0500 Subject: Btrfs: fix max chunk size on raid5/6 We try to limit the size of a chunk to 10GB, which keeps the unit of work reasonable during balance and resize operations. The limit checks were taking into account the number of copies of the data we had but what they really should be doing is comparing against the logical size of the chunk we're creating. This moves the code around a little to use the count of data stripes from raid5/6. Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 5d6010ba8b7e..538c5cfa005f 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3837,10 +3837,6 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, */ data_stripes = num_stripes / ncopies; - if (stripe_size * ndevs > max_chunk_size * ncopies) { - stripe_size = max_chunk_size * ncopies; - do_div(stripe_size, ndevs); - } if (type & BTRFS_BLOCK_GROUP_RAID5) { raid_stripe_len = find_raid56_stripe_len(ndevs - 1, btrfs_super_stripesize(info->super_copy)); @@ -3851,6 +3847,27 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, btrfs_super_stripesize(info->super_copy)); data_stripes = num_stripes - 2; } + + /* + * Use the number of data stripes to figure out how big this chunk + * is really going to be in terms of logical address space, + * and compare that answer with the max chunk size + */ + if (stripe_size * data_stripes > max_chunk_size) { + u64 mask = (1ULL << 24) - 1; + stripe_size = max_chunk_size; + do_div(stripe_size, data_stripes); + + /* bump the answer up to a 16MB boundary */ + stripe_size = (stripe_size + mask) & ~mask; + + /* but don't go higher than the limits we found + * while searching for free extents + */ + if (stripe_size > devices_info[ndevs-1].max_avail) + stripe_size = devices_info[ndevs-1].max_avail; + } + do_div(stripe_size, dev_stripes); /* align to BTRFS_STRIPE_LEN */ -- cgit From 38c227d87c49ad5d173cb5d4374d49acec6a495d Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Tue, 29 Jan 2013 03:18:40 +0000 Subject: Btrfs: snapshot-aware defrag This comes from one of btrfs's project ideas, As we defragment files, we break any sharing from other snapshots. The balancing code will preserve the sharing, and defrag needs to grow this as well. Now we're able to fill the blank with this patch, in which we make full use of backref walking stuff. Here is the basic idea, o set the writeback ranges started by defragment with flag EXTENT_DEFRAG o at endio, after we finish updating fs tree, we use backref walking to find all parents of the ranges and re-link them with the new COWed file layout by adding corresponding backrefs. Signed-off-by: Li Zefan Signed-off-by: Liu Bo Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 654 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 654 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 3bc62b181ef8..4d0aec0cf5d8 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -55,6 +55,7 @@ #include "locking.h" #include "free-space-cache.h" #include "inode-map.h" +#include "backref.h" struct btrfs_iget_args { u64 ino; @@ -1932,6 +1933,640 @@ out: return ret; } +/* snapshot-aware defrag */ +struct sa_defrag_extent_backref { + struct rb_node node; + struct old_sa_defrag_extent *old; + u64 root_id; + u64 inum; + u64 file_pos; + u64 extent_offset; + u64 num_bytes; + u64 generation; +}; + +struct old_sa_defrag_extent { + struct list_head list; + struct new_sa_defrag_extent *new; + + u64 extent_offset; + u64 bytenr; + u64 offset; + u64 len; + int count; +}; + +struct new_sa_defrag_extent { + struct rb_root root; + struct list_head head; + struct btrfs_path *path; + struct inode *inode; + u64 file_pos; + u64 len; + u64 bytenr; + u64 disk_len; + u8 compress_type; +}; + +static int backref_comp(struct sa_defrag_extent_backref *b1, + struct sa_defrag_extent_backref *b2) +{ + if (b1->root_id < b2->root_id) + return -1; + else if (b1->root_id > b2->root_id) + return 1; + + if (b1->inum < b2->inum) + return -1; + else if (b1->inum > b2->inum) + return 1; + + if (b1->file_pos < b2->file_pos) + return -1; + else if (b1->file_pos > b2->file_pos) + return 1; + + /* + * [------------------------------] ===> (a range of space) + * |<--->| |<---->| =============> (fs/file tree A) + * |<---------------------------->| ===> (fs/file tree B) + * + * A range of space can refer to two file extents in one tree while + * refer to only one file extent in another tree. + * + * So we may process a disk offset more than one time(two extents in A) + * and locate at the same extent(one extent in B), then insert two same + * backrefs(both refer to the extent in B). + */ + return 0; +} + +static void backref_insert(struct rb_root *root, + struct sa_defrag_extent_backref *backref) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *parent = NULL; + struct sa_defrag_extent_backref *entry; + int ret; + + while (*p) { + parent = *p; + entry = rb_entry(parent, struct sa_defrag_extent_backref, node); + + ret = backref_comp(backref, entry); + if (ret < 0) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&backref->node, parent, p); + rb_insert_color(&backref->node, root); +} + +/* + * Note the backref might has changed, and in this case we just return 0. + */ +static noinline int record_one_backref(u64 inum, u64 offset, u64 root_id, + void *ctx) +{ + struct btrfs_file_extent_item *extent; + struct btrfs_fs_info *fs_info; + struct old_sa_defrag_extent *old = ctx; + struct new_sa_defrag_extent *new = old->new; + struct btrfs_path *path = new->path; + struct btrfs_key key; + struct btrfs_root *root; + struct sa_defrag_extent_backref *backref; + struct extent_buffer *leaf; + struct inode *inode = new->inode; + int slot; + int ret; + u64 extent_offset; + u64 num_bytes; + + if (BTRFS_I(inode)->root->root_key.objectid == root_id && + inum == btrfs_ino(inode)) + return 0; + + key.objectid = root_id; + key.type = BTRFS_ROOT_ITEM_KEY; + key.offset = (u64)-1; + + fs_info = BTRFS_I(inode)->root->fs_info; + root = btrfs_read_fs_root_no_name(fs_info, &key); + if (IS_ERR(root)) { + if (PTR_ERR(root) == -ENOENT) + return 0; + WARN_ON(1); + pr_debug("inum=%llu, offset=%llu, root_id=%llu\n", + inum, offset, root_id); + return PTR_ERR(root); + } + + key.objectid = inum; + key.type = BTRFS_EXTENT_DATA_KEY; + if (offset > (u64)-1 << 32) + key.offset = 0; + else + key.offset = offset; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) { + WARN_ON(1); + return ret; + } + + while (1) { + cond_resched(); + + leaf = path->nodes[0]; + slot = path->slots[0]; + + if (slot >= btrfs_header_nritems(leaf)) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) { + goto out; + } else if (ret > 0) { + ret = 0; + goto out; + } + continue; + } + + path->slots[0]++; + + btrfs_item_key_to_cpu(leaf, &key, slot); + + if (key.objectid > inum) + goto out; + + if (key.objectid < inum || key.type != BTRFS_EXTENT_DATA_KEY) + continue; + + extent = btrfs_item_ptr(leaf, slot, + struct btrfs_file_extent_item); + + if (btrfs_file_extent_disk_bytenr(leaf, extent) != old->bytenr) + continue; + + extent_offset = btrfs_file_extent_offset(leaf, extent); + if (key.offset - extent_offset != offset) + continue; + + num_bytes = btrfs_file_extent_num_bytes(leaf, extent); + if (extent_offset >= old->extent_offset + old->offset + + old->len || extent_offset + num_bytes <= + old->extent_offset + old->offset) + continue; + + break; + } + + backref = kmalloc(sizeof(*backref), GFP_NOFS); + if (!backref) { + ret = -ENOENT; + goto out; + } + + backref->root_id = root_id; + backref->inum = inum; + backref->file_pos = offset + extent_offset; + backref->num_bytes = num_bytes; + backref->extent_offset = extent_offset; + backref->generation = btrfs_file_extent_generation(leaf, extent); + backref->old = old; + backref_insert(&new->root, backref); + old->count++; +out: + btrfs_release_path(path); + WARN_ON(ret); + return ret; +} + +static noinline bool record_extent_backrefs(struct btrfs_path *path, + struct new_sa_defrag_extent *new) +{ + struct btrfs_fs_info *fs_info = BTRFS_I(new->inode)->root->fs_info; + struct old_sa_defrag_extent *old, *tmp; + int ret; + + new->path = path; + + list_for_each_entry_safe(old, tmp, &new->head, list) { + ret = iterate_inodes_from_logical(old->bytenr, fs_info, + path, record_one_backref, + old); + BUG_ON(ret < 0 && ret != -ENOENT); + + /* no backref to be processed for this extent */ + if (!old->count) { + list_del(&old->list); + kfree(old); + } + } + + if (list_empty(&new->head)) + return false; + + return true; +} + +static int relink_is_mergable(struct extent_buffer *leaf, + struct btrfs_file_extent_item *fi, + u64 disk_bytenr) +{ + if (btrfs_file_extent_disk_bytenr(leaf, fi) != disk_bytenr) + return 0; + + if (btrfs_file_extent_type(leaf, fi) != BTRFS_FILE_EXTENT_REG) + return 0; + + if (btrfs_file_extent_compression(leaf, fi) || + btrfs_file_extent_encryption(leaf, fi) || + btrfs_file_extent_other_encoding(leaf, fi)) + return 0; + + return 1; +} + +/* + * Note the backref might has changed, and in this case we just return 0. + */ +static noinline int relink_extent_backref(struct btrfs_path *path, + struct sa_defrag_extent_backref *prev, + struct sa_defrag_extent_backref *backref) +{ + struct btrfs_file_extent_item *extent; + struct btrfs_file_extent_item *item; + struct btrfs_ordered_extent *ordered; + struct btrfs_trans_handle *trans; + struct btrfs_fs_info *fs_info; + struct btrfs_root *root; + struct btrfs_key key; + struct extent_buffer *leaf; + struct old_sa_defrag_extent *old = backref->old; + struct new_sa_defrag_extent *new = old->new; + struct inode *src_inode = new->inode; + struct inode *inode; + struct extent_state *cached = NULL; + int ret = 0; + u64 start; + u64 len; + u64 lock_start; + u64 lock_end; + bool merge = false; + int index; + + if (prev && prev->root_id == backref->root_id && + prev->inum == backref->inum && + prev->file_pos + prev->num_bytes == backref->file_pos) + merge = true; + + /* step 1: get root */ + key.objectid = backref->root_id; + key.type = BTRFS_ROOT_ITEM_KEY; + key.offset = (u64)-1; + + fs_info = BTRFS_I(src_inode)->root->fs_info; + index = srcu_read_lock(&fs_info->subvol_srcu); + + root = btrfs_read_fs_root_no_name(fs_info, &key); + if (IS_ERR(root)) { + srcu_read_unlock(&fs_info->subvol_srcu, index); + if (PTR_ERR(root) == -ENOENT) + return 0; + return PTR_ERR(root); + } + if (btrfs_root_refs(&root->root_item) == 0) { + srcu_read_unlock(&fs_info->subvol_srcu, index); + /* parse ENOENT to 0 */ + return 0; + } + + /* step 2: get inode */ + key.objectid = backref->inum; + key.type = BTRFS_INODE_ITEM_KEY; + key.offset = 0; + + inode = btrfs_iget(fs_info->sb, &key, root, NULL); + if (IS_ERR(inode)) { + srcu_read_unlock(&fs_info->subvol_srcu, index); + return 0; + } + + srcu_read_unlock(&fs_info->subvol_srcu, index); + + /* step 3: relink backref */ + lock_start = backref->file_pos; + lock_end = backref->file_pos + backref->num_bytes - 1; + lock_extent_bits(&BTRFS_I(inode)->io_tree, lock_start, lock_end, + 0, &cached); + + ordered = btrfs_lookup_first_ordered_extent(inode, lock_end); + if (ordered) { + btrfs_put_ordered_extent(ordered); + goto out_unlock; + } + + trans = btrfs_join_transaction(root); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto out_unlock; + } + + key.objectid = backref->inum; + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = backref->file_pos; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) { + goto out_free_path; + } else if (ret > 0) { + ret = 0; + goto out_free_path; + } + + extent = btrfs_item_ptr(path->nodes[0], path->slots[0], + struct btrfs_file_extent_item); + + if (btrfs_file_extent_generation(path->nodes[0], extent) != + backref->generation) + goto out_free_path; + + btrfs_release_path(path); + + start = backref->file_pos; + if (backref->extent_offset < old->extent_offset + old->offset) + start += old->extent_offset + old->offset - + backref->extent_offset; + + len = min(backref->extent_offset + backref->num_bytes, + old->extent_offset + old->offset + old->len); + len -= max(backref->extent_offset, old->extent_offset + old->offset); + + ret = btrfs_drop_extents(trans, root, inode, start, + start + len, 1); + if (ret) + goto out_free_path; +again: + key.objectid = btrfs_ino(inode); + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = start; + + if (merge) { + struct btrfs_file_extent_item *fi; + u64 extent_len; + struct btrfs_key found_key; + + ret = btrfs_search_slot(trans, root, &key, path, 1, 1); + if (ret < 0) + goto out_free_path; + + path->slots[0]--; + leaf = path->nodes[0]; + btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); + + fi = btrfs_item_ptr(leaf, path->slots[0], + struct btrfs_file_extent_item); + extent_len = btrfs_file_extent_num_bytes(leaf, fi); + + if (relink_is_mergable(leaf, fi, new->bytenr) && + extent_len + found_key.offset == start) { + btrfs_set_file_extent_num_bytes(leaf, fi, + extent_len + len); + btrfs_mark_buffer_dirty(leaf); + inode_add_bytes(inode, len); + + ret = 1; + goto out_free_path; + } else { + merge = false; + btrfs_release_path(path); + goto again; + } + } + + ret = btrfs_insert_empty_item(trans, root, path, &key, + sizeof(*extent)); + if (ret) { + btrfs_abort_transaction(trans, root, ret); + goto out_free_path; + } + + leaf = path->nodes[0]; + item = btrfs_item_ptr(leaf, path->slots[0], + struct btrfs_file_extent_item); + btrfs_set_file_extent_disk_bytenr(leaf, item, new->bytenr); + btrfs_set_file_extent_disk_num_bytes(leaf, item, new->disk_len); + btrfs_set_file_extent_offset(leaf, item, start - new->file_pos); + btrfs_set_file_extent_num_bytes(leaf, item, len); + btrfs_set_file_extent_ram_bytes(leaf, item, new->len); + btrfs_set_file_extent_generation(leaf, item, trans->transid); + btrfs_set_file_extent_type(leaf, item, BTRFS_FILE_EXTENT_REG); + btrfs_set_file_extent_compression(leaf, item, new->compress_type); + btrfs_set_file_extent_encryption(leaf, item, 0); + btrfs_set_file_extent_other_encoding(leaf, item, 0); + + btrfs_mark_buffer_dirty(leaf); + inode_add_bytes(inode, len); + + ret = btrfs_inc_extent_ref(trans, root, new->bytenr, + new->disk_len, 0, + backref->root_id, backref->inum, + new->file_pos, 0); /* start - extent_offset */ + if (ret) { + btrfs_abort_transaction(trans, root, ret); + goto out_free_path; + } + + ret = 1; +out_free_path: + btrfs_release_path(path); + btrfs_end_transaction(trans, root); +out_unlock: + unlock_extent_cached(&BTRFS_I(inode)->io_tree, lock_start, lock_end, + &cached, GFP_NOFS); + iput(inode); + return ret; +} + +static void relink_file_extents(struct new_sa_defrag_extent *new) +{ + struct btrfs_path *path; + struct old_sa_defrag_extent *old, *tmp; + struct sa_defrag_extent_backref *backref; + struct sa_defrag_extent_backref *prev = NULL; + struct inode *inode; + struct btrfs_root *root; + struct rb_node *node; + int ret; + + inode = new->inode; + root = BTRFS_I(inode)->root; + + path = btrfs_alloc_path(); + if (!path) + return; + + if (!record_extent_backrefs(path, new)) { + btrfs_free_path(path); + goto out; + } + btrfs_release_path(path); + + while (1) { + node = rb_first(&new->root); + if (!node) + break; + rb_erase(node, &new->root); + + backref = rb_entry(node, struct sa_defrag_extent_backref, node); + + ret = relink_extent_backref(path, prev, backref); + WARN_ON(ret < 0); + + kfree(prev); + + if (ret == 1) + prev = backref; + else + prev = NULL; + cond_resched(); + } + kfree(prev); + + btrfs_free_path(path); + + list_for_each_entry_safe(old, tmp, &new->head, list) { + list_del(&old->list); + kfree(old); + } +out: + atomic_dec(&root->fs_info->defrag_running); + wake_up(&root->fs_info->transaction_wait); + + kfree(new); +} + +static struct new_sa_defrag_extent * +record_old_file_extents(struct inode *inode, + struct btrfs_ordered_extent *ordered) +{ + struct btrfs_root *root = BTRFS_I(inode)->root; + struct btrfs_path *path; + struct btrfs_key key; + struct old_sa_defrag_extent *old, *tmp; + struct new_sa_defrag_extent *new; + int ret; + + new = kmalloc(sizeof(*new), GFP_NOFS); + if (!new) + return NULL; + + new->inode = inode; + new->file_pos = ordered->file_offset; + new->len = ordered->len; + new->bytenr = ordered->start; + new->disk_len = ordered->disk_len; + new->compress_type = ordered->compress_type; + new->root = RB_ROOT; + INIT_LIST_HEAD(&new->head); + + path = btrfs_alloc_path(); + if (!path) + goto out_kfree; + + key.objectid = btrfs_ino(inode); + key.type = BTRFS_EXTENT_DATA_KEY; + key.offset = new->file_pos; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + goto out_free_path; + if (ret > 0 && path->slots[0] > 0) + path->slots[0]--; + + /* find out all the old extents for the file range */ + while (1) { + struct btrfs_file_extent_item *extent; + struct extent_buffer *l; + int slot; + u64 num_bytes; + u64 offset; + u64 end; + u64 disk_bytenr; + u64 extent_offset; + + l = path->nodes[0]; + slot = path->slots[0]; + + if (slot >= btrfs_header_nritems(l)) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) + goto out_free_list; + else if (ret > 0) + break; + continue; + } + + btrfs_item_key_to_cpu(l, &key, slot); + + if (key.objectid != btrfs_ino(inode)) + break; + if (key.type != BTRFS_EXTENT_DATA_KEY) + break; + if (key.offset >= new->file_pos + new->len) + break; + + extent = btrfs_item_ptr(l, slot, struct btrfs_file_extent_item); + + num_bytes = btrfs_file_extent_num_bytes(l, extent); + if (key.offset + num_bytes < new->file_pos) + goto next; + + disk_bytenr = btrfs_file_extent_disk_bytenr(l, extent); + if (!disk_bytenr) + goto next; + + extent_offset = btrfs_file_extent_offset(l, extent); + + old = kmalloc(sizeof(*old), GFP_NOFS); + if (!old) + goto out_free_list; + + offset = max(new->file_pos, key.offset); + end = min(new->file_pos + new->len, key.offset + num_bytes); + + old->bytenr = disk_bytenr; + old->extent_offset = extent_offset; + old->offset = offset - key.offset; + old->len = end - offset; + old->new = new; + old->count = 0; + list_add_tail(&old->list, &new->head); +next: + path->slots[0]++; + cond_resched(); + } + + btrfs_free_path(path); + atomic_inc(&root->fs_info->defrag_running); + + return new; + +out_free_list: + list_for_each_entry_safe(old, tmp, &new->head, list) { + list_del(&old->list); + kfree(old); + } +out_free_path: + btrfs_free_path(path); +out_kfree: + kfree(new); + return NULL; +} + /* * helper function for btrfs_finish_ordered_io, this * just reads in some of the csum leaves to prime them into ram @@ -1949,6 +2584,7 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) struct btrfs_trans_handle *trans = NULL; struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; struct extent_state *cached_state = NULL; + struct new_sa_defrag_extent *new = NULL; int compress_type = 0; int ret; bool nolock; @@ -1983,6 +2619,20 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) ordered_extent->file_offset + ordered_extent->len - 1, 0, &cached_state); + ret = test_range_bit(io_tree, ordered_extent->file_offset, + ordered_extent->file_offset + ordered_extent->len - 1, + EXTENT_DEFRAG, 1, cached_state); + if (ret) { + u64 last_snapshot = btrfs_root_last_snapshot(&root->root_item); + if (last_snapshot >= BTRFS_I(inode)->generation) + /* the inode is shared */ + new = record_old_file_extents(inode, ordered_extent); + + clear_extent_bit(io_tree, ordered_extent->file_offset, + ordered_extent->file_offset + ordered_extent->len - 1, + EXTENT_DEFRAG, 0, 0, &cached_state, GFP_NOFS); + } + if (nolock) trans = btrfs_join_transaction_nolock(root); else @@ -2064,6 +2714,10 @@ out: */ btrfs_remove_ordered_extent(inode, ordered_extent); + /* for snapshot-aware defrag */ + if (new) + relink_file_extents(new); + /* once for us */ btrfs_put_ordered_extent(ordered_extent); /* once for the tree */ -- cgit From 172a50497ffaf84d60dff37fbeb03894268fe5c2 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 21 Feb 2013 02:48:22 -0700 Subject: Btrfs: fix wrong outstanding_extents when doing DIO write When running the 083th case of xfstests on the filesystem with "compress-force=lzo", the following WARNINGs were triggered. WARNING: at fs/btrfs/inode.c:7908 WARNING: at fs/btrfs/inode.c:7909 WARNING: at fs/btrfs/inode.c:7911 WARNING: at fs/btrfs/extent-tree.c:4510 WARNING: at fs/btrfs/extent-tree.c:4511 This problem was introduced by the patch "Btrfs: fix deadlock due to unsubmitted". In this patch, there are two bugs which caused the above problem. The 1st one is a off-by-one bug, if the DIO write return 0, it is also a short write, we need release the reserved space for it. But we didn't do it in that patch. Fix it by change "ret > 0" to "ret >= 0". The 2nd one is ->outstanding_extents was increased twice when a short write happened. As we know, ->outstanding_extents is a counter to keep track of the number of extent items we may use duo to delalloc, when we reserve the free space for a delalloc write, we assume that the write will introduce just one extent item, so we increase ->outstanding_extents by 1 at that time. And then we will increase it every time we split the write, it is done at the beginning of btrfs_get_blocks_direct(). So when a short write happens, we needn't increase ->outstanding_extents again. But this patch done. In order to fix the 2nd problem, I re-write the logic for ->outstanding_extents operation. We don't increase it at the beginning of btrfs_get_blocks_direct(), instead, we just increase it when the split actually happens. Reported-by: Mitch Harder Signed-off-by: Miao Xie Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 20 +++++++++----------- 1 file changed, 9 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 4d0aec0cf5d8..40d49da5e846 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6708,12 +6708,9 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, int unlock_bits = EXTENT_LOCKED; int ret = 0; - if (create) { - spin_lock(&BTRFS_I(inode)->lock); - BTRFS_I(inode)->outstanding_extents++; - spin_unlock(&BTRFS_I(inode)->lock); + if (create) unlock_bits |= EXTENT_DELALLOC | EXTENT_DIRTY; - } else + else len = min_t(u64, len, root->sectorsize); lockstart = start; @@ -6855,6 +6852,10 @@ unlock: if (start + len > i_size_read(inode)) i_size_write(inode, start + len); + spin_lock(&BTRFS_I(inode)->lock); + BTRFS_I(inode)->outstanding_extents++; + spin_unlock(&BTRFS_I(inode)->lock); + ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockstart + len - 1, EXTENT_DELALLOC, NULL, &cached_state, GFP_NOFS); @@ -7362,14 +7363,11 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, if (rw & WRITE) { if (ret < 0 && ret != -EIOCBQUEUED) btrfs_delalloc_release_space(inode, count); - else if (ret > 0 && (size_t)ret < count) { - spin_lock(&BTRFS_I(inode)->lock); - BTRFS_I(inode)->outstanding_extents++; - spin_unlock(&BTRFS_I(inode)->lock); + else if (ret >= 0 && (size_t)ret < count) btrfs_delalloc_release_space(inode, count - (size_t)ret); - } - btrfs_delalloc_release_metadata(inode, 0); + else + btrfs_delalloc_release_metadata(inode, 0); } out: if (wakeup) -- cgit From dc81cdc58ad2f413b96b9004f8d681e5dc554473 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 20 Feb 2013 23:32:52 -0700 Subject: Btrfs: fix remount vs autodefrag If we remount the fs to close the auto defragment or make the fs R/O, we should stop the auto defragment. Signed-off-by: Miao Xie Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 2 ++ fs/btrfs/file.c | 5 +++++ fs/btrfs/super.c | 40 ++++++++++++++++++++++++++++++++++++++-- 3 files changed, 45 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 3dcedfe4f759..ae8dcc406805 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -339,6 +339,7 @@ static inline unsigned long btrfs_chunk_item_size(int num_stripes) * File system states */ #define BTRFS_FS_STATE_ERROR 0 +#define BTRFS_FS_STATE_REMOUNTING 1 /* Super block flags */ /* Errors detected */ @@ -1902,6 +1903,7 @@ struct btrfs_ioctl_defrag_range_args { #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) +#define btrfs_raw_test_opt(o, opt) ((o) & BTRFS_MOUNT_##opt) #define btrfs_test_opt(root, opt) ((root)->fs_info->mount_opt & \ BTRFS_MOUNT_##opt) /* diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 9f67e623206d..6e6dd8cdad92 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -374,6 +374,11 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info) atomic_inc(&fs_info->defrag_running); while(1) { + /* Pause the auto defragger. */ + if (test_bit(BTRFS_FS_STATE_REMOUNTING, + &fs_info->fs_state)) + break; + if (!__need_auto_defrag(fs_info->tree_root)) break; diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index db1ba9a2ed64..68a29a1ea068 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1202,6 +1202,38 @@ static void btrfs_resize_thread_pool(struct btrfs_fs_info *fs_info, new_pool_size); } +static inline void btrfs_remount_prepare(struct btrfs_fs_info *fs_info, + unsigned long old_opts, int flags) +{ + set_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state); + + if (btrfs_raw_test_opt(old_opts, AUTO_DEFRAG) && + (!btrfs_raw_test_opt(fs_info->mount_opt, AUTO_DEFRAG) || + (flags & MS_RDONLY))) { + /* wait for any defraggers to finish */ + wait_event(fs_info->transaction_wait, + (atomic_read(&fs_info->defrag_running) == 0)); + if (flags & MS_RDONLY) + sync_filesystem(fs_info->sb); + } +} + +static inline void btrfs_remount_cleanup(struct btrfs_fs_info *fs_info, + unsigned long old_opts) +{ + /* + * We need cleanup all defragable inodes if the autodefragment is + * close or the fs is R/O. + */ + if (btrfs_raw_test_opt(old_opts, AUTO_DEFRAG) && + (!btrfs_raw_test_opt(fs_info->mount_opt, AUTO_DEFRAG) || + (fs_info->sb->s_flags & MS_RDONLY))) { + btrfs_cleanup_defrag_inodes(fs_info); + } + + clear_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state); +} + static int btrfs_remount(struct super_block *sb, int *flags, char *data) { struct btrfs_fs_info *fs_info = btrfs_sb(sb); @@ -1215,6 +1247,8 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) unsigned int old_metadata_ratio = fs_info->metadata_ratio; int ret; + btrfs_remount_prepare(fs_info, old_opts, *flags); + ret = btrfs_parse_options(root, data); if (ret) { ret = -EINVAL; @@ -1225,7 +1259,7 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) fs_info->thread_pool_size, old_thread_pool_size); if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY)) - return 0; + goto out; if (*flags & MS_RDONLY) { /* @@ -1280,7 +1314,8 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) } sb->s_flags &= ~MS_RDONLY; } - +out: + btrfs_remount_cleanup(fs_info, old_opts); return 0; restore: @@ -1297,6 +1332,7 @@ restore: btrfs_resize_thread_pool(fs_info, old_thread_pool_size, fs_info->thread_pool_size); fs_info->metadata_ratio = old_metadata_ratio; + btrfs_remount_cleanup(fs_info, old_opts); return ret; } -- cgit From 676a0675cf9200ac047fb50825f80867b3bb733b Mon Sep 17 00:00:00 2001 From: Jim Somerville Date: Thu, 21 Feb 2013 16:41:59 -0800 Subject: inotify: remove broken mask checks causing unmount to be EINVAL Running the command: inotifywait -e unmount /mnt/disk immediately aborts with a -EINVAL return code. This is however a valid parameter. This abort occurs only if unmount is the sole event parameter. If other event parameters are supplied, then the unmount event wait will work. The problem was introduced by commit 44b350fc23e ("inotify: Fix mask checks"). In that commit, it states: The mask checks in inotify_update_existing_watch() and inotify_new_watch() are useless because inotify_arg_to_mask() sets FS_IN_IGNORED and FS_EVENT_ON_CHILD bits anyway. But instead of removing the useless checks, it did this: mask = inotify_arg_to_mask(arg); - if (unlikely(!mask)) + if (unlikely(!(mask & IN_ALL_EVENTS))) return -EINVAL; The problem is that IN_ALL_EVENTS doesn't include IN_UNMOUNT, and other parts of the code keep IN_UNMOUNT separate from IN_ALL_EVENTS. So the check should be: if (unlikely(!(mask & (IN_ALL_EVENTS | IN_UNMOUNT)))) But inotify_arg_to_mask(arg) always sets the IN_UNMOUNT bit in the mask anyway, so the check is always going to pass and thus should simply be removed. Also note that inotify_arg_to_mask completely controls what mask bits get set from arg, there's no way for invalid bits to get enabled there. Lets fix it by simply removing the useless broken checks. Signed-off-by: Jim Somerville Signed-off-by: Paul Gortmaker Cc: Jerome Marchand Cc: John McCutchan Cc: Robert Love Cc: Eric Paris Cc: [2.6.37+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/notify/inotify/inotify_user.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 228a2c2ad8d7..07f7a92fe88e 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -576,8 +576,6 @@ static int inotify_update_existing_watch(struct fsnotify_group *group, /* don't allow invalid bits: we don't want flags set */ mask = inotify_arg_to_mask(arg); - if (unlikely(!(mask & IN_ALL_EVENTS))) - return -EINVAL; fsn_mark = fsnotify_find_inode_mark(group, inode); if (!fsn_mark) @@ -629,8 +627,6 @@ static int inotify_new_watch(struct fsnotify_group *group, /* don't allow invalid bits: we don't want flags set */ mask = inotify_arg_to_mask(arg); - if (unlikely(!(mask & IN_ALL_EVENTS))) - return -EINVAL; tmp_i_mark = kmem_cache_alloc(inotify_inode_mark_cachep, GFP_KERNEL); if (unlikely(!tmp_i_mark)) -- cgit From 7630b661da330b35dd57b6f5d6d62b386f2dd751 Mon Sep 17 00:00:00 2001 From: MITSUNARI Shigeo Date: Thu, 21 Feb 2013 16:42:01 -0800 Subject: fs/block_dev.c: page cache wrongly left invalidated after revalidate_disk() We found that bdev->bd_invalidated was left set once revalidate_disk() is called, which results in page cache flush every time that device is open. Specifically, we found this problem in MD block device. Once we resize a MD device, mdadm --monitor periodically flush all page cache for that device every 60 or 1000 seconds when it opens the device. This bug lies since at least 3.2.0 till the latest kernel(3.6.2). Patch is attached. The following steps will reproduce the problem. 1. prepair a block device (eg /dev/sdb). 2. create two partitions: sudo parted /dev/sdb mklabel gpt mkpart primary 0% 50% mkpart primary 50% 100% 3. create a md device. sudo mdadm -C /dev/md/hoge -l 1 -n 2 -e 1.2 --assume-clean --auto=md --symlink=no /dev/sdb1 /dev/sdb2 4. create file system and mount it sudo mkfs.ext3 /dev/md/hoge sudo mkdir /mnt/test sudo mount /dev/md/hoge /mnt/test 5. try to resize the device sudo mdadm -G /dev/md/hoge --size=max 6. create a file to fill file cache. sudo dd if=/dev/urandom of=/mnt/test/data bs=1M count=10 and verify the current status of file by free command. 7. mdadm monitor will open the md device every 1000 seconds and you will find all file cache on the device are cleared. The timing can be reduced by the following steps. a) kill mdadm and restart it with --delay option /sbin/mdadm --monitor --delay=30 --pid-file /var/run/mdadm/monitor.pid --daemonise --scan --syslog or open the md device directly. sudo dd if=/dev/md/hoge of=/dev/null bs=4096 count=1 Signed-off-by: MITSUNARI Shigeo Cc: Al Viro Cc: Jeff Moyer Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/block_dev.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index 172f8491a2bd..78333a37f49d 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -994,6 +994,7 @@ int revalidate_disk(struct gendisk *disk) mutex_lock(&bdev->bd_mutex); check_disk_size_change(disk, bdev); + bdev->bd_invalidated = 0; mutex_unlock(&bdev->bd_mutex); bdput(bdev); return ret; -- cgit From 49deb4bc227cb9db5b8ebf9434367f8bed057c7a Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Thu, 21 Feb 2013 16:42:43 -0800 Subject: configfs: move the dereference below the NULL test The dereference should be moved below the NULL test. spatch with a semantic match is used to found this. (http://coccinelle.lip6.fr/) Signed-off-by: Wei Yongjun Cc: Joel Becker Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/configfs/dir.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 712b10f64c70..e9dcfa3c208c 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1037,10 +1037,11 @@ static int configfs_dump(struct configfs_dirent *sd, int level) static int configfs_depend_prep(struct dentry *origin, struct config_item *target) { - struct configfs_dirent *child_sd, *sd = origin->d_fsdata; + struct configfs_dirent *child_sd, *sd; int ret = 0; - BUG_ON(!origin || !sd); + BUG_ON(!origin || !origin->d_fsdata); + sd = origin->d_fsdata; if (sd->s_element == target) /* Boo-yah */ goto out; -- cgit From d787ab0977c58e2c421b8d0ab49e363893ddb814 Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Thu, 21 Feb 2013 16:42:44 -0800 Subject: ocfs2: remove kfree() redundant null checks smatch analysis indicates a number of redundant NULL checks before calling kfree(), eg: fs/ocfs2/alloc.c:6138 ocfs2_begin_truncate_log_recovery() info: redundant null check on *tl_copy calling kfree() fs/ocfs2/alloc.c:6755 ocfs2_zero_range_for_truncate() info: redundant null check on pages calling kfree() etc.... [akpm@linux-foundation.org: revert dubious change in ocfs2_begin_truncate_log_recovery()] Signed-off-by: Tim Gardner Cc: Mark Fasheh Acked-by: Joel Becker Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ocfs2/alloc.c | 3 +-- fs/ocfs2/cluster/heartbeat.c | 6 ++---- fs/ocfs2/cluster/tcp.c | 6 ++---- fs/ocfs2/dlm/dlmdomain.c | 4 +--- fs/ocfs2/extent_map.c | 3 +-- fs/ocfs2/journal.c | 10 +++------- fs/ocfs2/localalloc.c | 8 +++----- fs/ocfs2/stack_o2cb.c | 2 +- fs/ocfs2/super.c | 6 ++---- fs/ocfs2/sysfile.c | 3 +-- 10 files changed, 17 insertions(+), 34 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 31b9463fba1f..b8a9d87231b1 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -6751,8 +6751,7 @@ int ocfs2_zero_range_for_truncate(struct inode *inode, handle_t *handle, mlog_errno(ret); out: - if (pages) - kfree(pages); + kfree(pages); return ret; } diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index f7c648d7d6bf..42252bf64b51 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -1471,8 +1471,7 @@ static void o2hb_region_release(struct config_item *item) mlog(ML_HEARTBEAT, "hb region release (%s)\n", reg->hr_dev_name); - if (reg->hr_tmp_block) - kfree(reg->hr_tmp_block); + kfree(reg->hr_tmp_block); if (reg->hr_slot_data) { for (i = 0; i < reg->hr_num_pages; i++) { @@ -1486,8 +1485,7 @@ static void o2hb_region_release(struct config_item *item) if (reg->hr_bdev) blkdev_put(reg->hr_bdev, FMODE_READ|FMODE_WRITE); - if (reg->hr_slots) - kfree(reg->hr_slots); + kfree(reg->hr_slots); kfree(reg->hr_db_regnum); kfree(reg->hr_db_livenodes); diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 1bfe8802cc1e..f0edbd84f7bc 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -1165,10 +1165,8 @@ out: o2net_debug_del_nst(&nst); /* must be before dropping sc and node */ if (sc) sc_put(sc); - if (vec) - kfree(vec); - if (msg) - kfree(msg); + kfree(vec); + kfree(msg); o2net_complete_nsw(nn, &nsw, 0, 0, 0); return ret; } diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index 9e89d70df337..dbb17c07656a 100644 --- a/fs/ocfs2/dlm/dlmdomain.c +++ b/fs/ocfs2/dlm/dlmdomain.c @@ -319,9 +319,7 @@ static void dlm_free_ctxt_mem(struct dlm_ctxt *dlm) if (dlm->master_hash) dlm_free_pagevec((void **)dlm->master_hash, DLM_HASH_PAGES); - if (dlm->name) - kfree(dlm->name); - + kfree(dlm->name); kfree(dlm); } diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c index f487aa343442..1c39efb71bab 100644 --- a/fs/ocfs2/extent_map.c +++ b/fs/ocfs2/extent_map.c @@ -282,8 +282,7 @@ search: spin_unlock(&oi->ip_lock); out: - if (new_emi) - kfree(new_emi); + kfree(new_emi); } static int ocfs2_last_eb_is_empty(struct inode *inode, diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 2dd36af79e26..8eccfabcd12e 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -1234,11 +1234,8 @@ static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal, /* Though we wish to avoid it, we are in fact safe in * skipping local alloc cleanup as fsck.ocfs2 is more * than capable of reclaiming unused space. */ - if (la_dinode) - kfree(la_dinode); - - if (tl_dinode) - kfree(tl_dinode); + kfree(la_dinode); + kfree(tl_dinode); if (qrec) ocfs2_free_quota_recovery(qrec); @@ -1408,8 +1405,7 @@ bail: mutex_unlock(&osb->recovery_lock); - if (rm_quota) - kfree(rm_quota); + kfree(rm_quota); /* no one is callint kthread_stop() for us so the kthread() api * requires that we call do_exit(). And it isn't exported, but diff --git a/fs/ocfs2/localalloc.c b/fs/ocfs2/localalloc.c index a9f78c74d687..aebeacd807c3 100644 --- a/fs/ocfs2/localalloc.c +++ b/fs/ocfs2/localalloc.c @@ -476,8 +476,7 @@ out: if (local_alloc_inode) iput(local_alloc_inode); - if (alloc_copy) - kfree(alloc_copy); + kfree(alloc_copy); } /* @@ -534,7 +533,7 @@ int ocfs2_begin_local_alloc_recovery(struct ocfs2_super *osb, mlog_errno(status); bail: - if ((status < 0) && (*alloc_copy)) { + if (status < 0) { kfree(*alloc_copy); *alloc_copy = NULL; } @@ -1290,8 +1289,7 @@ bail: if (main_bm_inode) iput(main_bm_inode); - if (alloc_copy) - kfree(alloc_copy); + kfree(alloc_copy); if (ac) ocfs2_free_alloc_context(ac); diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c index 94368017edb3..bf1f8930456f 100644 --- a/fs/ocfs2/stack_o2cb.c +++ b/fs/ocfs2/stack_o2cb.c @@ -376,7 +376,7 @@ static int o2cb_cluster_connect(struct ocfs2_cluster_connection *conn) dlm_register_eviction_cb(dlm, &priv->op_eviction_cb); out_free: - if (rc && conn->cc_private) + if (rc) kfree(conn->cc_private); out: diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 0e91ec22a940..9b6910dec4ba 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -2525,8 +2525,7 @@ static int ocfs2_check_volume(struct ocfs2_super *osb) mlog_errno(status); finally: - if (local_alloc) - kfree(local_alloc); + kfree(local_alloc); if (status) mlog_errno(status); @@ -2553,8 +2552,7 @@ static void ocfs2_delete_osb(struct ocfs2_super *osb) * we free it here. */ kfree(osb->journal); - if (osb->local_alloc_copy) - kfree(osb->local_alloc_copy); + kfree(osb->local_alloc_copy); kfree(osb->uuid_str); ocfs2_put_dlm_debug(osb->osb_dlm_debug); memset(osb, 0, sizeof(struct ocfs2_super)); diff --git a/fs/ocfs2/sysfile.c b/fs/ocfs2/sysfile.c index 3d635f4bbb20..f053688d22a3 100644 --- a/fs/ocfs2/sysfile.c +++ b/fs/ocfs2/sysfile.c @@ -91,8 +91,7 @@ static struct inode **get_local_system_inode(struct ocfs2_super *osb, } else osb->local_system_inodes = local_system_inodes; spin_unlock(&osb->osb_lock); - if (unlikely(free)) - kfree(free); + kfree(free); } index = (slot * NUM_LOCAL_SYSTEM_INODES) + -- cgit From 3278bb748d2437eb1464765f36429e5d6aa91c38 Mon Sep 17 00:00:00 2001 From: Junxiao Bi Date: Thu, 21 Feb 2013 16:42:45 -0800 Subject: ocfs2: unlock super lock if lockres refresh failed If lockres refresh failed, the super lock will never be released which will cause some processes on other cluster nodes hung forever. Signed-off-by: Junxiao Bi Cc: Joel Becker Cc: Mark Fasheh Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ocfs2/dlmglue.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 4f7795fb5fc0..88577eb5d712 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -2545,6 +2545,7 @@ int ocfs2_super_lock(struct ocfs2_super *osb, * everything is up to the caller :) */ status = ocfs2_should_refresh_lock_res(lockres); if (status < 0) { + ocfs2_cluster_unlock(osb, lockres, level); mlog_errno(status); goto bail; } @@ -2553,8 +2554,10 @@ int ocfs2_super_lock(struct ocfs2_super *osb, ocfs2_complete_lock_res_refresh(lockres, status); - if (status < 0) + if (status < 0) { + ocfs2_cluster_unlock(osb, lockres, level); mlog_errno(status); + } ocfs2_track_lock_refresh(lockres); } bail: -- cgit From 1d1d1a767206fbe5d4c69493b7e6d2a8d08cc0a0 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 21 Feb 2013 16:42:51 -0800 Subject: mm: only enforce stable page writes if the backing device requires it Create a helper function to check if a backing device requires stable page writes and, if so, performs the necessary wait. Then, make it so that all points in the memory manager that handle making pages writable use the helper function. This should provide stable page write support to most filesystems, while eliminating unnecessary waiting for devices that don't require the feature. Before this patchset, all filesystems would block, regardless of whether or not it was necessary. ext3 would wait, but still generate occasional checksum errors. The network filesystems were left to do their own thing, so they'd wait too. After this patchset, all the disk filesystems except ext3 and btrfs will wait only if the hardware requires it. ext3 (if necessary) snapshots pages instead of blocking, and btrfs provides its own bdi so the mm will never wait. Network filesystems haven't been touched, so either they provide their own stable page guarantees or they don't block at all. The blocking behavior is back to what it was before 3.0 if you don't have a disk requiring stable page writes. Here's the result of using dbench to test latency on ext2: 3.8.0-rc3: Operation Count AvgLat MaxLat ---------------------------------------- WriteX 109347 0.028 59.817 ReadX 347180 0.004 3.391 Flush 15514 29.828 287.283 Throughput 57.429 MB/sec 4 clients 4 procs max_latency=287.290 ms 3.8.0-rc3 + patches: WriteX 105556 0.029 4.273 ReadX 335004 0.005 4.112 Flush 14982 30.540 298.634 Throughput 55.4496 MB/sec 4 clients 4 procs max_latency=298.650 ms As you can see, the maximum write latency drops considerably with this patch enabled. The other filesystems (ext3/ext4/xfs/btrfs) behave similarly, but see the cover letter for those results. Signed-off-by: Darrick J. Wong Acked-by: Steven Whitehouse Reviewed-by: Jan Kara Cc: Adrian Hunter Cc: Andy Lutomirski Cc: Artem Bityutskiy Cc: Joel Becker Cc: Mark Fasheh Cc: Jens Axboe Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 2 +- fs/ext4/inode.c | 2 +- fs/gfs2/file.c | 2 +- fs/nilfs2/file.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index 7a75c3e0fd58..2ea9cd44aeae 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -2359,7 +2359,7 @@ int __block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, if (unlikely(ret < 0)) goto out_unlock; set_page_dirty(page); - wait_on_page_writeback(page); + wait_for_stable_page(page); return 0; out_unlock: unlock_page(page); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index cbfe13bf5b2a..cd818d8bb221 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -4968,7 +4968,7 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) 0, len, NULL, ext4_bh_unmapped)) { /* Wait so that we don't change page under IO */ - wait_on_page_writeback(page); + wait_for_stable_page(page); ret = VM_FAULT_LOCKED; goto out; } diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 06b7092a3f25..2687f50d98cb 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -483,7 +483,7 @@ out: gfs2_holder_uninit(&gh); if (ret == 0) { set_page_dirty(page); - wait_on_page_writeback(page); + wait_for_stable_page(page); } sb_end_pagefault(inode->i_sb); return block_page_mkwrite_return(ret); diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c index 61946883025c..bec4af6eab13 100644 --- a/fs/nilfs2/file.c +++ b/fs/nilfs2/file.c @@ -126,7 +126,7 @@ static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) nilfs_transaction_commit(inode->i_sb); mapped: - wait_on_page_writeback(page); + wait_for_stable_page(page); out: sb_end_pagefault(inode->i_sb); return block_page_mkwrite_return(ret); -- cgit From 13575ca14fcdacd1ad914d00bc63eb4d96280986 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 21 Feb 2013 16:42:53 -0800 Subject: 9pfs: fix filesystem to wait for stable page writeback Fix up the ->page_mkwrite handler to provide stable page writes if necessary. Signed-off-by: Darrick J. Wong Cc: Adrian Hunter Cc: Andy Lutomirski Cc: Artem Bityutskiy Cc: Jan Kara Cc: Joel Becker Cc: Mark Fasheh Cc: Steven Whitehouse Cc: Jens Axboe Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/9p/vfs_file.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index 7a3399767570..c921ac92ea4c 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -616,6 +616,7 @@ v9fs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) lock_page(page); if (page->mapping != inode->i_mapping) goto out_unlock; + wait_for_stable_page(page); return VM_FAULT_LOCKED; out_unlock: -- cgit From ffecfd1a72fccfcee3dabb99b9ecba9735318f90 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 21 Feb 2013 16:42:55 -0800 Subject: block: optionally snapshot page contents to provide stable pages during write This provides a band-aid to provide stable page writes on jbd without needing to backport the fixed locking and page writeback bit handling schemes of jbd2. The band-aid works by using bounce buffers to snapshot page contents instead of waiting. For those wondering about the ext3 bandage -- fixing the jbd locking (which was done as part of ext4dev years ago) is a lot of surgery, and setting PG_writeback on data pages when we actually hold the page lock dropped ext3 performance by nearly an order of magnitude. If we're going to migrate iscsi and raid to use stable page writes, the complaints about high latency will likely return. We might as well centralize their page snapshotting thing to one place. Signed-off-by: Darrick J. Wong Tested-by: Andy Lutomirski Cc: Adrian Hunter Cc: Artem Bityutskiy Reviewed-by: Jan Kara Cc: Joel Becker Cc: Mark Fasheh Cc: Steven Whitehouse Cc: Jens Axboe Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext3/super.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 6e50223b3299..4ba2683c1d44 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -2065,6 +2065,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) test_opt(sb,DATA_FLAGS) == EXT3_MOUNT_JOURNAL_DATA ? "journal": test_opt(sb,DATA_FLAGS) == EXT3_MOUNT_ORDERED_DATA ? "ordered": "writeback"); + sb->s_flags |= MS_SNAP_STABLE; return 0; -- cgit From 1269529bda27823ed42798762e896c059ea5e486 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 21 Feb 2013 16:42:57 -0800 Subject: ocfs2: wait for page writeback to provide stable pages When stable pages are required, we have to wait if the page is just going to disk and we want to modify it. Add proper callback to ocfs2_grab_pages_for_write(). Signed-off-by: Jan Kara Signed-off-by: Darrick J. Wong Acked-by: Joel Becker Cc: Mark Fasheh Cc: Adrian Hunter Cc: Andy Lutomirski Cc: Artem Bityutskiy Cc: Steven Whitehouse Cc: Jens Axboe Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ocfs2/aops.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 657743254eb9..9796330d8f04 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -1194,6 +1194,7 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping, goto out; } } + wait_for_stable_page(wc->w_pages[i]); if (index == target_index) wc->w_target_page = wc->w_pages[i]; -- cgit From 182dcfd648aef0705fd04c42cbe507c09e00c25d Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 21 Feb 2013 16:42:59 -0800 Subject: ubifs: wait for page writeback to provide stable pages When stable pages are required, we have to wait if the page is just going to disk and we want to modify it. Add proper callback to ubifs_vm_page_mkwrite(). Signed-off-by: Jan Kara Signed-off-by: Darrick J. Wong Cc: Artem Bityutskiy Cc: Adrian Hunter Cc: Andy Lutomirski Cc: Joel Becker Cc: Mark Fasheh Cc: Steven Whitehouse Cc: Jens Axboe Cc: Eric Van Hensbergen Cc: Ron Minnich Cc: Latchesar Ionkov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ubifs/file.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 5bc77817f382..4f6493c130e0 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -1522,6 +1522,7 @@ static int ubifs_vm_page_mkwrite(struct vm_area_struct *vma, ubifs_release_dirty_inode_budget(c, ui); } + wait_for_stable_page(page); unlock_page(page); return 0; -- cgit From d3330cf08ccf9aef28a3e5740fd5cc7ac536db84 Mon Sep 17 00:00:00 2001 From: Zhang Yanfei Date: Thu, 21 Feb 2013 16:44:20 -0800 Subject: binfmt_elf: remove unused argument in fill_elf_header In fill_elf_header(), elf->e_ident[EI_OSABI] is always set to ELF_OSABI, so remove the unused argument 'osabi'. Signed-off-by: Zhang Yanfei Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/binfmt_elf.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 49d0b43458b7..ff9dbc630efa 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -1249,7 +1249,7 @@ static int writenote(struct memelfnote *men, struct file *file, #undef DUMP_WRITE static void fill_elf_header(struct elfhdr *elf, int segs, - u16 machine, u32 flags, u8 osabi) + u16 machine, u32 flags) { memset(elf, 0, sizeof(*elf)); @@ -1634,7 +1634,7 @@ static int fill_note_info(struct elfhdr *elf, int phdrs, * Initialize the ELF file header. */ fill_elf_header(elf, phdrs, - view->e_machine, view->e_flags, view->ei_osabi); + view->e_machine, view->e_flags); /* * Allocate a structure for each thread. @@ -1874,7 +1874,7 @@ static int fill_note_info(struct elfhdr *elf, int phdrs, elf_core_copy_regs(&info->prstatus->pr_reg, regs); /* Set up header */ - fill_elf_header(elf, phdrs, ELF_ARCH, ELF_CORE_EFLAGS, ELF_OSABI); + fill_elf_header(elf, phdrs, ELF_ARCH, ELF_CORE_EFLAGS); /* * Set up the notes in similar form to SVR4 core dumps made -- cgit From d646a02a9d44d1421f273ae3923d97b47b918176 Mon Sep 17 00:00:00 2001 From: Guo Chao Date: Thu, 21 Feb 2013 15:16:42 -0800 Subject: block: use i_size_write() in bd_set_size() blkdev_ioctl(GETBLKSIZE) uses i_size_read() to read size of block device. If we update block size directly, reader may see intermediate result in some machines and configurations. Use i_size_write() instead. Signed-off-by: Guo Chao Cc: Alexander Viro Cc: Guo Chao Cc: M. Hindess Cc: Nikanth Karthikesan Cc: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Jens Axboe --- fs/block_dev.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index 172f8491a2bd..82b7c9ad29ac 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1032,7 +1032,9 @@ void bd_set_size(struct block_device *bdev, loff_t size) { unsigned bsize = bdev_logical_block_size(bdev); - bdev->bd_inode->i_size = size; + mutex_lock(&bdev->bd_inode->i_mutex); + i_size_write(bdev->bd_inode, size); + mutex_unlock(&bdev->bd_inode->i_mutex); while (bsize < PAGE_CACHE_SIZE) { if (size & bsize) break; -- cgit From de33127d8d3f1d570aad8c2223cd81b206636bc1 Mon Sep 17 00:00:00 2001 From: Guo Chao Date: Thu, 21 Feb 2013 15:16:43 -0800 Subject: block: remove redundant check to bd_openers() bd_openers is stable under bd_mutex, no need to check it twice. Signed-off-by: Guo Chao Cc: Alexander Viro Cc: Guo Chao Cc: M. Hindess Cc: Nikanth Karthikesan Cc: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Jens Axboe --- fs/block_dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index 82b7c9ad29ac..2f7058421ffb 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1119,7 +1119,7 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) } } - if (!ret && !bdev->bd_openers) { + if (!ret) { bd_set_size(bdev,(loff_t)get_capacity(disk)<<9); bdi = blk_get_backing_dev_info(bdev); if (bdi == NULL) -- cgit From 5a7a613a47a715711b3f2d3322a0eac21d459166 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 22 Feb 2013 12:53:43 -0500 Subject: NFS: Don't allow NFS silly-renamed files to be deleted, no signal Commit 73ca100 broke the code that prevents the client from deleting a silly renamed dentry. This affected "delete on last close" semantics as after that commit, nothing prevented removal of silly-renamed files. As a result, a process holding a file open could easily get an ESTALE on the file in a directory where some other process issued 'rm -rf some_dir_containing_the_file' twice. Before the commit, any attempt at unlinking silly renamed files would fail inside may_delete() with -EBUSY because of the DCACHE_NFSFS_RENAMED flag. The following testcase demonstrates the problem: tail -f /nfsmnt/dir/file & rm -rf /nfsmnt/dir rm -rf /nfsmnt/dir # second removal does not fail, 'tail' process receives ESTALE The problem with the above commit is that it unhashes the old and new dentries from the lookup path, even in the normal case when a signal is not encountered and it would have been safe to call d_move. Unfortunately the old dentry has the special DCACHE_NFSFS_RENAMED flag set on it. Unhashing has the side-effect that future lookups call d_alloc(), allocating a new dentry without the special flag for any silly-renamed files. As a result, subsequent calls to unlink silly renamed files do not fail but allow the removal to go through. This will result in ESTALE errors for any other process doing operations on the file. To fix this, go back to using d_move on success. For the signal case, it's unclear what we may safely do beyond d_drop. Reported-by: Dave Wysochanski Signed-off-by: Trond Myklebust Acked-by: Jeff Layton Cc: stable@vger.kernel.org --- fs/nfs/unlink.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index d26a32f5b53b..1f1f38f0c5d5 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -335,20 +335,14 @@ static void nfs_async_rename_done(struct rpc_task *task, void *calldata) struct inode *old_dir = data->old_dir; struct inode *new_dir = data->new_dir; struct dentry *old_dentry = data->old_dentry; - struct dentry *new_dentry = data->new_dentry; if (!NFS_PROTO(old_dir)->rename_done(task, old_dir, new_dir)) { rpc_restart_call_prepare(task); return; } - if (task->tk_status != 0) { + if (task->tk_status != 0) nfs_cancel_async_unlink(old_dentry); - return; - } - - d_drop(old_dentry); - d_drop(new_dentry); } /** @@ -549,6 +543,18 @@ nfs_sillyrename(struct inode *dir, struct dentry *dentry) error = rpc_wait_for_completion_task(task); if (error == 0) error = task->tk_status; + switch (error) { + case 0: + /* The rename succeeded */ + nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); + d_move(dentry, sdentry); + break; + case -ERESTARTSYS: + /* The result of the rename is unknown. Play it safe by + * forcing a new lookup */ + d_drop(dentry); + d_drop(sdentry); + } rpc_put_task(task); out_dput: dput(sdentry); -- cgit From d43814721111041e26671a153e300e2054fb36fa Mon Sep 17 00:00:00 2001 From: Eryu Guan Date: Fri, 22 Feb 2013 15:27:47 -0500 Subject: ext4: no need to remove extent if len is 0 in ext4_es_remove_extent() len is 0 means no extent needs to be removed, so return immediately. Otherwise it could trigger the following BUG_ON() in ext4_es_remove_extent() end = lblk + len - 1; BUG_ON(end < lblk); This could be reproduced by a simple truncate(1) command by an unprivileged user truncate -s $(($((2**32 - 1)) * 4096)) /mnt/ext4/testfile The same is true for __es_insert_extent(). Patched kernel passed xfstests regression test. Signed-off-by: Eryu Guan Signed-off-by: "Theodore Ts'o" Reviewed-by: Zheng Liu --- fs/ext4/extents_status.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 9f1380e05474..f768f4a98a2b 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -456,6 +456,9 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, es_debug("add [%u/%u) %llu %llx to extent status tree of inode %lu\n", lblk, len, pblk, status, inode->i_ino); + if (!len) + return 0; + BUG_ON(end < lblk); newes.es_lblk = lblk; @@ -649,6 +652,9 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, es_debug("remove [%u/%u) from extent status tree of inode %lu\n", lblk, len, inode->i_ino); + if (!len) + return err; + end = lblk + len - 1; BUG_ON(end < lblk); -- cgit From 304e220f0879198b1f5309ad6f0be862b4009491 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Fri, 22 Feb 2013 15:27:52 -0500 Subject: ext4: fix free clusters calculation in bigalloc filesystem ext4_has_free_clusters() should tell us whether there is enough free clusters to allocate, however number of free clusters in the file system is converted to blocks using EXT4_C2B() which is not only wrong use of the macro (we should have used EXT4_NUM_B2C) but it's also completely wrong concept since everything else is in cluster units. Moreover when calculating number of root clusters we should be using macro EXT4_NUM_B2C() instead of EXT4_B2C() otherwise the result might be off by one. However r_blocks_count should always be a multiple of the cluster ratio so doing a plain bit shift should be enough here. We avoid using EXT4_B2C() because it's confusing. As a result of the first problem number of free clusters is much bigger than it should have been and ext4_has_free_clusters() would return 1 even if there is really not enough free clusters available. Fix this by removing the EXT4_C2B() conversion of free clusters and using bit shift when calculating number of root clusters. This bug affects number of xfstests tests covering file system ENOSPC situation handling. With this patch most of the ENOSPC problems with bigalloc file system disappear, especially the errors caused by delayed allocation not having enough space when the actual allocation is finally requested. Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/ext4/balloc.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 33938c120c85..2f2e0da1a6b7 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -484,11 +484,16 @@ static int ext4_has_free_clusters(struct ext4_sb_info *sbi, free_clusters = percpu_counter_read_positive(fcc); dirty_clusters = percpu_counter_read_positive(dcc); - root_clusters = EXT4_B2C(sbi, ext4_r_blocks_count(sbi->s_es)); + + /* + * r_blocks_count should always be multiple of the cluster ratio so + * we are safe to do a plane bit shift only. + */ + root_clusters = ext4_r_blocks_count(sbi->s_es) >> sbi->s_cluster_bits; if (free_clusters - (nclusters + root_clusters + dirty_clusters) < EXT4_FREECLUSTERS_WATERMARK) { - free_clusters = EXT4_C2B(sbi, percpu_counter_sum_positive(fcc)); + free_clusters = percpu_counter_sum_positive(fcc); dirty_clusters = percpu_counter_sum_positive(dcc); } /* Check whether we have space after accounting for current -- cgit From 92a49fb0f79f3300e6e50ddf56238e70678e4202 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 22 Feb 2013 15:31:00 -0800 Subject: ceph: fix statvfs fr_size Different versions of glibc are broken in different ways, but the short of it is that for the time being, frsize should == bsize, and be used as the multiple for the blocks, free, and available fields. This mirrors what is done for NFS. The previous reporting of the page size for frsize meant that newer glibc and df would report a very small value for the fs size. Fixes http://tracker.ceph.com/issues/3793. Signed-off-by: Sage Weil Reviewed-by: Greg Farnum --- fs/ceph/super.c | 7 ++++++- fs/ceph/super.h | 2 +- 2 files changed, 7 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ceph/super.c b/fs/ceph/super.c index e86aa9948124..9fe17c6c2876 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -71,8 +71,14 @@ static int ceph_statfs(struct dentry *dentry, struct kstatfs *buf) /* * express utilization in terms of large blocks to avoid * overflow on 32-bit machines. + * + * NOTE: for the time being, we make bsize == frsize to humor + * not-yet-ancient versions of glibc that are broken. + * Someday, we will probably want to report a real block + * size... whatever that may mean for a network file system! */ buf->f_bsize = 1 << CEPH_BLOCK_SHIFT; + buf->f_frsize = 1 << CEPH_BLOCK_SHIFT; buf->f_blocks = le64_to_cpu(st.kb) >> (CEPH_BLOCK_SHIFT-10); buf->f_bfree = le64_to_cpu(st.kb_avail) >> (CEPH_BLOCK_SHIFT-10); buf->f_bavail = le64_to_cpu(st.kb_avail) >> (CEPH_BLOCK_SHIFT-10); @@ -80,7 +86,6 @@ static int ceph_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_files = le64_to_cpu(st.num_objects); buf->f_ffree = -1; buf->f_namelen = NAME_MAX; - buf->f_frsize = PAGE_CACHE_SIZE; /* leave fsid little-endian, regardless of host endianness */ fsid = *(u64 *)(&monmap->fsid) ^ *((u64 *)&monmap->fsid + 1); diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 9861cce10a49..604526a0d6cd 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -21,7 +21,7 @@ /* large granularity for statfs utilization stats to facilitate * large volume sizes on 32-bit machines. */ -#define CEPH_BLOCK_SHIFT 20 /* 1 MB */ +#define CEPH_BLOCK_SHIFT 22 /* 4 MB */ #define CEPH_BLOCK (1 << CEPH_BLOCK_SHIFT) #define CEPH_MOUNT_OPT_DIRSTAT (1<<4) /* `cat dirname` for stats */ -- cgit From 9b40bc90abd126bcc5da5658059b8e72e285e559 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 22 Feb 2013 22:45:42 -0500 Subject: get rid of unprotected dereferencing of mnt->mnt_ns It's safe only under namespace_sem or vfsmount_lock; all places in fs/namespace.c that want mnt->mnt_ns->user_ns actually want to use current->nsproxy->mnt_ns->user_ns (note the calls of check_mnt() in there). Cc: stable@vger.kernel.org Signed-off-by: Al Viro --- fs/namespace.c | 29 +++++++++++++++++------------ 1 file changed, 17 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 55605c552787..edac42c6eff2 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1237,6 +1237,14 @@ static int do_umount(struct mount *mnt, int flags) return retval; } +/* + * Is the caller allowed to modify his namespace? + */ +static inline bool may_mount(void) +{ + return ns_capable(current->nsproxy->mnt_ns->user_ns, CAP_SYS_ADMIN); +} + /* * Now umount can handle mount points as well as block devices. * This is important for filesystems which use unnamed block devices. @@ -1255,6 +1263,9 @@ SYSCALL_DEFINE2(umount, char __user *, name, int, flags) if (flags & ~(MNT_FORCE | MNT_DETACH | MNT_EXPIRE | UMOUNT_NOFOLLOW)) return -EINVAL; + if (!may_mount()) + return -EPERM; + if (!(flags & UMOUNT_NOFOLLOW)) lookup_flags |= LOOKUP_FOLLOW; @@ -1268,10 +1279,6 @@ SYSCALL_DEFINE2(umount, char __user *, name, int, flags) if (!check_mnt(mnt)) goto dput_and_out; - retval = -EPERM; - if (!ns_capable(mnt->mnt_ns->user_ns, CAP_SYS_ADMIN)) - goto dput_and_out; - retval = do_umount(mnt, flags); dput_and_out: /* we mustn't call path_put() as that would clear mnt_expiry_mark */ @@ -1295,7 +1302,7 @@ SYSCALL_DEFINE1(oldumount, char __user *, name) static int mount_is_safe(struct path *path) { - if (ns_capable(real_mount(path->mnt)->mnt_ns->user_ns, CAP_SYS_ADMIN)) + if (may_mount()) return 0; return -EPERM; #ifdef notyet @@ -1633,7 +1640,7 @@ static int do_change_type(struct path *path, int flag) int type; int err = 0; - if (!ns_capable(mnt->mnt_ns->user_ns, CAP_SYS_ADMIN)) + if (!may_mount()) return -EPERM; if (path->dentry != path->mnt->mnt_root) @@ -1797,7 +1804,7 @@ static int do_move_mount(struct path *path, const char *old_name) struct mount *p; struct mount *old; int err = 0; - if (!ns_capable(real_mount(path->mnt)->mnt_ns->user_ns, CAP_SYS_ADMIN)) + if (!may_mount()) return -EPERM; if (!old_name || !*old_name) return -EINVAL; @@ -1933,16 +1940,14 @@ static int do_new_mount(struct path *path, const char *fstype, int flags, int mnt_flags, const char *name, void *data) { struct file_system_type *type; - struct user_namespace *user_ns; + struct user_namespace *user_ns = current->nsproxy->mnt_ns->user_ns; struct vfsmount *mnt; int err; if (!fstype) return -EINVAL; - /* we need capabilities... */ - user_ns = real_mount(path->mnt)->mnt_ns->user_ns; - if (!ns_capable(user_ns, CAP_SYS_ADMIN)) + if (!may_mount()) return -EPERM; type = get_fs_type(fstype); @@ -2567,7 +2572,7 @@ SYSCALL_DEFINE2(pivot_root, const char __user *, new_root, struct mount *new_mnt, *root_mnt; int error; - if (!ns_capable(current->nsproxy->mnt_ns->user_ns, CAP_SYS_ADMIN)) + if (!may_mount()) return -EPERM; error = user_path_dir(new_root, &new); -- cgit From 57eccb830f1cc93d4b506ba306d8dfa685e0c88f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 22 Feb 2013 22:49:10 -0500 Subject: mount: consolidate permission checks ... and ask for global CAP_SYS_ADMIN only for superblock-level remounts Signed-off-by: Al Viro --- fs/namespace.c | 40 +++++++--------------------------------- 1 file changed, 7 insertions(+), 33 deletions(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index edac42c6eff2..269919fa116d 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1300,24 +1300,6 @@ SYSCALL_DEFINE1(oldumount, char __user *, name) #endif -static int mount_is_safe(struct path *path) -{ - if (may_mount()) - return 0; - return -EPERM; -#ifdef notyet - if (S_ISLNK(path->dentry->d_inode->i_mode)) - return -EPERM; - if (path->dentry->d_inode->i_mode & S_ISVTX) { - if (current_uid() != path->dentry->d_inode->i_uid) - return -EPERM; - } - if (inode_permission(path->dentry->d_inode, MAY_WRITE)) - return -EPERM; - return 0; -#endif -} - static bool mnt_ns_loop(struct path *path) { /* Could bind mounting the mount namespace inode cause a @@ -1640,9 +1622,6 @@ static int do_change_type(struct path *path, int flag) int type; int err = 0; - if (!may_mount()) - return -EPERM; - if (path->dentry != path->mnt->mnt_root) return -EINVAL; @@ -1676,9 +1655,7 @@ static int do_loopback(struct path *path, const char *old_name, LIST_HEAD(umount_list); struct path old_path; struct mount *mnt = NULL, *old; - int err = mount_is_safe(path); - if (err) - return err; + int err; if (!old_name || !*old_name) return -EINVAL; err = kern_path(old_name, LOOKUP_FOLLOW|LOOKUP_AUTOMOUNT, &old_path); @@ -1755,9 +1732,6 @@ static int do_remount(struct path *path, int flags, int mnt_flags, struct super_block *sb = path->mnt->mnt_sb; struct mount *mnt = real_mount(path->mnt); - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; - if (!check_mnt(mnt)) return -EINVAL; @@ -1771,6 +1745,8 @@ static int do_remount(struct path *path, int flags, int mnt_flags, down_write(&sb->s_umount); if (flags & MS_BIND) err = change_mount_flags(path->mnt, flags); + else if (!capable(CAP_SYS_ADMIN)) + err = -EPERM; else err = do_remount_sb(sb, flags, data, 0); if (!err) { @@ -1803,9 +1779,7 @@ static int do_move_mount(struct path *path, const char *old_name) struct path old_path, parent_path; struct mount *p; struct mount *old; - int err = 0; - if (!may_mount()) - return -EPERM; + int err; if (!old_name || !*old_name) return -EINVAL; err = kern_path(old_name, LOOKUP_FOLLOW, &old_path); @@ -1947,9 +1921,6 @@ static int do_new_mount(struct path *path, const char *fstype, int flags, if (!fstype) return -EINVAL; - if (!may_mount()) - return -EPERM; - type = get_fs_type(fstype); if (!type) return -ENODEV; @@ -2263,6 +2234,9 @@ long do_mount(const char *dev_name, const char *dir_name, if (retval) goto dput_out; + if (!may_mount()) + return -EPERM; + /* Default to relatime unless overriden */ if (!(flags & MS_NOATIME)) mnt_flags |= MNT_RELATIME; -- cgit From 496ad9aa8ef448058e36ca7a787c61f2e63f0f54 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 23 Jan 2013 17:07:38 -0500 Subject: new helper: file_inode(file) Signed-off-by: Al Viro --- fs/9p/vfs_file.c | 10 ++++---- fs/adfs/dir.c | 2 +- fs/affs/dir.c | 2 +- fs/afs/dir.c | 4 +-- fs/afs/flock.c | 4 +-- fs/afs/write.c | 7 +++-- fs/autofs4/autofs_i.h | 2 +- fs/autofs4/dev-ioctl.c | 2 +- fs/autofs4/root.c | 4 +-- fs/befs/linuxvfs.c | 2 +- fs/bfs/dir.c | 2 +- fs/binfmt_aout.c | 4 +-- fs/binfmt_elf.c | 2 +- fs/binfmt_elf_fdpic.c | 4 +-- fs/binfmt_flat.c | 2 +- fs/binfmt_misc.c | 4 +-- fs/block_dev.c | 2 +- fs/btrfs/file.c | 8 +++--- fs/btrfs/inode.c | 4 +-- fs/btrfs/ioctl.c | 52 +++++++++++++++++++------------------- fs/btrfs/send.c | 2 +- fs/buffer.c | 4 +-- fs/ceph/addr.c | 12 ++++----- fs/ceph/dir.c | 6 ++--- fs/ceph/file.c | 10 ++++---- fs/ceph/ioctl.c | 16 ++++++------ fs/ceph/locks.c | 2 +- fs/cifs/cifsfs.c | 6 ++--- fs/cifs/file.c | 26 +++++++++---------- fs/cifs/inode.c | 8 +++--- fs/cifs/ioctl.c | 2 +- fs/cifs/readdir.c | 4 +-- fs/coda/dir.c | 2 +- fs/coda/file.c | 12 ++++----- fs/coda/inode.c | 2 +- fs/coda/pioctl.c | 2 +- fs/compat_ioctl.c | 2 +- fs/configfs/dir.c | 2 +- fs/coredump.c | 4 +-- fs/cramfs/inode.c | 2 +- fs/ecryptfs/file.c | 4 +-- fs/efs/dir.c | 2 +- fs/exec.c | 8 +++--- fs/exofs/dir.c | 2 +- fs/ext2/dir.c | 2 +- fs/ext2/ioctl.c | 2 +- fs/ext3/dir.c | 8 +++--- fs/ext3/ioctl.c | 2 +- fs/ext3/namei.c | 4 +-- fs/ext4/dir.c | 8 +++--- fs/ext4/extents.c | 4 +-- fs/ext4/file.c | 2 +- fs/ext4/inline.c | 2 +- fs/ext4/inode.c | 6 ++--- fs/ext4/ioctl.c | 2 +- fs/ext4/move_extent.c | 6 ++--- fs/ext4/namei.c | 2 +- fs/ext4/super.c | 2 +- fs/f2fs/dir.c | 2 +- fs/fat/dir.c | 6 ++--- fs/fat/file.c | 4 +-- fs/fcntl.c | 2 +- fs/file_table.c | 2 +- fs/freevxfs/vxfs_lookup.c | 2 +- fs/fuse/control.c | 2 +- fs/fuse/dir.c | 2 +- fs/gfs2/file.c | 17 ++++++------- fs/gfs2/rgrp.c | 2 +- fs/hfs/dir.c | 2 +- fs/hfs/inode.c | 2 +- fs/hfsplus/dir.c | 2 +- fs/hfsplus/inode.c | 2 +- fs/hfsplus/ioctl.c | 4 +-- fs/hostfs/hostfs_kern.c | 2 +- fs/hpfs/dir.c | 4 +-- fs/hpfs/file.c | 2 +- fs/hppfs/hppfs.c | 8 +++--- fs/hugetlbfs/inode.c | 2 +- fs/inode.c | 2 +- fs/ioctl.c | 12 ++++----- fs/isofs/compress.c | 2 +- fs/isofs/dir.c | 2 +- fs/jffs2/dir.c | 4 +-- fs/jfs/ioctl.c | 2 +- fs/jfs/jfs_dtree.c | 2 +- fs/lockd/clntlock.c | 2 +- fs/lockd/clntproc.c | 2 +- fs/lockd/svclock.c | 16 ++++++------ fs/lockd/svcsubs.c | 2 +- fs/locks.c | 24 +++++++++--------- fs/logfs/dir.c | 4 +-- fs/logfs/file.c | 2 +- fs/minix/dir.c | 2 +- fs/namei.c | 2 +- fs/namespace.c | 2 +- fs/ncpfs/inode.c | 4 +-- fs/ncpfs/ioctl.c | 2 +- fs/ncpfs/mmap.c | 2 +- fs/nfs/dir.c | 8 +++--- fs/nfs/file.c | 2 +- fs/nfs/idmap.c | 2 +- fs/nfs/inode.c | 4 +-- fs/nfs/nfs3proc.c | 2 +- fs/nfs/nfs4file.c | 2 +- fs/nfs/proc.c | 2 +- fs/nfsd/fault_inject.c | 6 ++--- fs/nfsd/nfsctl.c | 2 +- fs/nfsd/vfs.c | 6 ++--- fs/nilfs2/dir.c | 2 +- fs/nilfs2/file.c | 2 +- fs/nilfs2/ioctl.c | 2 +- fs/notify/dnotify/dnotify.c | 4 +-- fs/notify/fanotify/fanotify_user.c | 2 +- fs/ntfs/dir.c | 2 +- fs/ocfs2/aops.c | 4 +-- fs/ocfs2/dir.c | 2 +- fs/ocfs2/dlmfs/dlmfs.c | 6 ++--- fs/ocfs2/file.c | 10 ++++---- fs/ocfs2/ioctl.c | 4 +-- fs/ocfs2/mmap.c | 8 +++--- fs/ocfs2/move_extents.c | 2 +- fs/ocfs2/refcounttree.c | 4 +-- fs/omfs/dir.c | 4 +-- fs/open.c | 6 ++--- fs/openpromfs/inode.c | 2 +- fs/pipe.c | 16 ++++++------ fs/proc/base.c | 38 ++++++++++++++-------------- fs/proc/generic.c | 10 ++++---- fs/proc/inode.c | 14 +++++----- fs/proc/nommu.c | 2 +- fs/proc/proc_net.c | 2 +- fs/proc/proc_sysctl.c | 4 +-- fs/proc/task_mmu.c | 6 ++--- fs/proc/task_nommu.c | 2 +- fs/qnx4/dir.c | 2 +- fs/qnx6/dir.c | 2 +- fs/ramfs/file-nommu.c | 2 +- fs/read_write.c | 8 +++--- fs/readdir.c | 2 +- fs/reiserfs/file.c | 2 +- fs/reiserfs/ioctl.c | 2 +- fs/reiserfs/procfs.c | 2 +- fs/romfs/super.c | 2 +- fs/splice.c | 2 +- fs/squashfs/dir.c | 2 +- fs/sync.c | 2 +- fs/sysfs/bin.c | 6 ++--- fs/sysv/dir.c | 2 +- fs/ubifs/dir.c | 2 +- fs/ubifs/file.c | 2 +- fs/ubifs/ioctl.c | 2 +- fs/udf/dir.c | 2 +- fs/udf/file.c | 4 +-- fs/ufs/dir.c | 2 +- fs/xfs/xfs_dfrag.c | 8 +++--- fs/xfs/xfs_file.c | 4 +-- fs/xfs/xfs_ioctl.c | 6 ++--- fs/xfs/xfs_ioctl32.c | 2 +- 158 files changed, 373 insertions(+), 375 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index c2483e97beee..3356e3ed5115 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -133,7 +133,7 @@ out_error: static int v9fs_file_lock(struct file *filp, int cmd, struct file_lock *fl) { int res = 0; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); p9_debug(P9_DEBUG_VFS, "filp: %p lock: %p\n", filp, fl); @@ -302,7 +302,7 @@ static int v9fs_file_getlock(struct file *filp, struct file_lock *fl) static int v9fs_file_lock_dotl(struct file *filp, int cmd, struct file_lock *fl) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int ret = -ENOLCK; p9_debug(P9_DEBUG_VFS, "filp: %p cmd:%d lock: %p name: %s\n", @@ -338,7 +338,7 @@ out_err: static int v9fs_file_flock_dotl(struct file *filp, int cmd, struct file_lock *fl) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int ret = -ENOLCK; p9_debug(P9_DEBUG_VFS, "filp: %p cmd:%d lock: %p name: %s\n", @@ -529,7 +529,7 @@ v9fs_file_write(struct file *filp, const char __user * data, if (!count) goto out; - retval = v9fs_file_write_internal(filp->f_path.dentry->d_inode, + retval = v9fs_file_write_internal(file_inode(filp), filp->private_data, data, count, &origin, 1); /* update offset on successful write */ @@ -604,7 +604,7 @@ v9fs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) struct v9fs_inode *v9inode; struct page *page = vmf->page; struct file *filp = vma->vm_file; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); p9_debug(P9_DEBUG_VFS, "page %p fid %lx\n", diff --git a/fs/adfs/dir.c b/fs/adfs/dir.c index b3be2e7c5643..9cf874ce8336 100644 --- a/fs/adfs/dir.c +++ b/fs/adfs/dir.c @@ -19,7 +19,7 @@ static DEFINE_RWLOCK(adfs_dir_lock); static int adfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; struct adfs_dir_ops *ops = ADFS_SB(sb)->s_dir; struct object_info obj; diff --git a/fs/affs/dir.c b/fs/affs/dir.c index 8ca8f3a55599..fd11a6d608ee 100644 --- a/fs/affs/dir.c +++ b/fs/affs/dir.c @@ -42,7 +42,7 @@ const struct inode_operations affs_dir_inode_operations = { static int affs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; struct buffer_head *dir_bh; struct buffer_head *fh_bh; diff --git a/fs/afs/dir.c b/fs/afs/dir.c index db477906ba4f..7a465ed04444 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -393,12 +393,12 @@ static int afs_readdir(struct file *file, void *cookie, filldir_t filldir) int ret; _enter("{%Ld,{%lu}}", - file->f_pos, file->f_path.dentry->d_inode->i_ino); + file->f_pos, file_inode(file)->i_ino); ASSERT(file->private_data != NULL); fpos = file->f_pos; - ret = afs_dir_iterate(file->f_path.dentry->d_inode, &fpos, + ret = afs_dir_iterate(file_inode(file), &fpos, cookie, filldir, file->private_data); file->f_pos = fpos; diff --git a/fs/afs/flock.c b/fs/afs/flock.c index 757d664575dd..2497bf306c70 100644 --- a/fs/afs/flock.c +++ b/fs/afs/flock.c @@ -514,7 +514,7 @@ error: */ int afs_lock(struct file *file, int cmd, struct file_lock *fl) { - struct afs_vnode *vnode = AFS_FS_I(file->f_dentry->d_inode); + struct afs_vnode *vnode = AFS_FS_I(file_inode(file)); _enter("{%x:%u},%d,{t=%x,fl=%x,r=%Ld:%Ld}", vnode->fid.vid, vnode->fid.vnode, cmd, @@ -537,7 +537,7 @@ int afs_lock(struct file *file, int cmd, struct file_lock *fl) */ int afs_flock(struct file *file, int cmd, struct file_lock *fl) { - struct afs_vnode *vnode = AFS_FS_I(file->f_dentry->d_inode); + struct afs_vnode *vnode = AFS_FS_I(file_inode(file)); _enter("{%x:%u},%d,{t=%x,fl=%x}", vnode->fid.vid, vnode->fid.vnode, cmd, diff --git a/fs/afs/write.c b/fs/afs/write.c index 9aa52d93c73c..7e03eadb40c0 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -120,7 +120,7 @@ int afs_write_begin(struct file *file, struct address_space *mapping, struct page **pagep, void **fsdata) { struct afs_writeback *candidate, *wb; - struct afs_vnode *vnode = AFS_FS_I(file->f_dentry->d_inode); + struct afs_vnode *vnode = AFS_FS_I(file_inode(file)); struct page *page; struct key *key = file->private_data; unsigned from = pos & (PAGE_CACHE_SIZE - 1); @@ -245,7 +245,7 @@ int afs_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { - struct afs_vnode *vnode = AFS_FS_I(file->f_dentry->d_inode); + struct afs_vnode *vnode = AFS_FS_I(file_inode(file)); loff_t i_size, maybe_i_size; _enter("{%x:%u},{%lx}", @@ -627,8 +627,7 @@ void afs_pages_written_back(struct afs_vnode *vnode, struct afs_call *call) ssize_t afs_file_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { - struct dentry *dentry = iocb->ki_filp->f_path.dentry; - struct afs_vnode *vnode = AFS_FS_I(dentry->d_inode); + struct afs_vnode *vnode = AFS_FS_I(file_inode(iocb->ki_filp)); ssize_t result; size_t count = iov_length(iov, nr_segs); diff --git a/fs/autofs4/autofs_i.h b/fs/autofs4/autofs_i.h index b785e7707959..3f1128b37e46 100644 --- a/fs/autofs4/autofs_i.h +++ b/fs/autofs4/autofs_i.h @@ -273,7 +273,7 @@ static inline int autofs_prepare_pipe(struct file *pipe) { if (!pipe->f_op || !pipe->f_op->write) return -EINVAL; - if (!S_ISFIFO(pipe->f_dentry->d_inode->i_mode)) + if (!S_ISFIFO(file_inode(pipe)->i_mode)) return -EINVAL; /* We want a packet pipe */ pipe->f_flags |= O_DIRECT; diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index 9f68a37bb2b2..743c7c2c949d 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -159,7 +159,7 @@ static struct autofs_sb_info *autofs_dev_ioctl_sbi(struct file *f) struct inode *inode; if (f) { - inode = f->f_path.dentry->d_inode; + inode = file_inode(f); sbi = autofs4_sbi(inode->i_sb); } return sbi; diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index c93447604da8..3cdf835e8b49 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -874,7 +874,7 @@ static int autofs4_root_ioctl_unlocked(struct inode *inode, struct file *filp, static long autofs4_root_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); return autofs4_root_ioctl_unlocked(inode, filp, cmd, arg); } @@ -882,7 +882,7 @@ static long autofs4_root_ioctl(struct file *filp, static long autofs4_root_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int ret; if (cmd == AUTOFS_IOC_READY || cmd == AUTOFS_IOC_FAIL) diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 2b3bda8d5e68..c8f4e25eb9e2 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -213,7 +213,7 @@ befs_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) static int befs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; befs_data_stream *ds = &BEFS_I(inode)->i_data.ds; befs_off_t value; diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c index 2785ef91191a..3f422f6bb5ca 100644 --- a/fs/bfs/dir.c +++ b/fs/bfs/dir.c @@ -28,7 +28,7 @@ static struct buffer_head *bfs_find_entry(struct inode *dir, static int bfs_readdir(struct file *f, void *dirent, filldir_t filldir) { - struct inode *dir = f->f_path.dentry->d_inode; + struct inode *dir = file_inode(f); struct buffer_head *bh; struct bfs_dirent *de; struct bfs_sb_info *info = BFS_SB(dir->i_sb); diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c index 6043567b95c2..bbc8f8827eac 100644 --- a/fs/binfmt_aout.c +++ b/fs/binfmt_aout.c @@ -214,7 +214,7 @@ static int load_aout_binary(struct linux_binprm * bprm) if ((N_MAGIC(ex) != ZMAGIC && N_MAGIC(ex) != OMAGIC && N_MAGIC(ex) != QMAGIC && N_MAGIC(ex) != NMAGIC) || N_TRSIZE(ex) || N_DRSIZE(ex) || - i_size_read(bprm->file->f_path.dentry->d_inode) < ex.a_text+ex.a_data+N_SYMSIZE(ex)+N_TXTOFF(ex)) { + i_size_read(file_inode(bprm->file)) < ex.a_text+ex.a_data+N_SYMSIZE(ex)+N_TXTOFF(ex)) { return -ENOEXEC; } @@ -367,7 +367,7 @@ static int load_aout_library(struct file *file) int retval; struct exec ex; - inode = file->f_path.dentry->d_inode; + inode = file_inode(file); retval = -ENOEXEC; error = kernel_read(file, 0, (char *) &ex, sizeof(ex)); diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 0c42cdbabecf..11e078a747a5 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -1140,7 +1140,7 @@ static unsigned long vma_dump_size(struct vm_area_struct *vma, /* By default, dump shared memory if mapped from an anonymous file. */ if (vma->vm_flags & VM_SHARED) { - if (vma->vm_file->f_path.dentry->d_inode->i_nlink == 0 ? + if (file_inode(vma->vm_file)->i_nlink == 0 ? FILTER(ANON_SHARED) : FILTER(MAPPED_SHARED)) goto whole; return 0; diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index dc84732e554f..30de01ca3eeb 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -909,7 +909,7 @@ static int elf_fdpic_map_file(struct elf_fdpic_params *params, dynamic_error: printk("ELF FDPIC %s with invalid DYNAMIC section (inode=%lu)\n", - what, file->f_path.dentry->d_inode->i_ino); + what, file_inode(file)->i_ino); return -ELIBBAD; } @@ -1219,7 +1219,7 @@ static int maydump(struct vm_area_struct *vma, unsigned long mm_flags) /* By default, dump shared memory if mapped from an anonymous file. */ if (vma->vm_flags & VM_SHARED) { - if (vma->vm_file->f_path.dentry->d_inode->i_nlink == 0) { + if (file_inode(vma->vm_file)->i_nlink == 0) { dump_ok = test_bit(MMF_DUMP_ANON_SHARED, &mm_flags); kdcore("%08lx: %08lx: %s (share)", vma->vm_start, vma->vm_flags, dump_ok ? "yes" : "no"); diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index b56371981d16..2036d21baaef 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -438,7 +438,7 @@ static int load_flat_file(struct linux_binprm * bprm, int ret; hdr = ((struct flat_hdr *) bprm->buf); /* exec-header */ - inode = bprm->file->f_path.dentry->d_inode; + inode = file_inode(bprm->file); text_len = ntohl(hdr->data_start); data_len = ntohl(hdr->data_end) - ntohl(hdr->data_start); diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index 0c8869fdd14e..fecbbf3f8ff2 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -531,7 +531,7 @@ static void kill_node(Node *e) static ssize_t bm_entry_read(struct file * file, char __user * buf, size_t nbytes, loff_t *ppos) { - Node *e = file->f_path.dentry->d_inode->i_private; + Node *e = file_inode(file)->i_private; ssize_t res; char *page; @@ -550,7 +550,7 @@ static ssize_t bm_entry_write(struct file *file, const char __user *buffer, size_t count, loff_t *ppos) { struct dentry *root; - Node *e = file->f_path.dentry->d_inode->i_private; + Node *e = file_inode(file)->i_private; int res = parse_command(buffer, count); switch (res) { diff --git a/fs/block_dev.c b/fs/block_dev.c index 172f8491a2bd..7d6bdfc6b7bc 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -318,7 +318,7 @@ static int blkdev_write_end(struct file *file, struct address_space *mapping, /* * private llseek: - * for a block special file file->f_path.dentry->d_inode->i_size is zero + * for a block special file file_inode(file)->i_size is zero * so we compute the size by hand (just as in block_read/write above) */ static loff_t block_llseek(struct file *file, loff_t offset, int whence) diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 77061bf43edb..4118e0b6e339 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1211,7 +1211,7 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file, struct extent_state *cached_state = NULL; int i; unsigned long index = pos >> PAGE_CACHE_SHIFT; - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); gfp_t mask = btrfs_alloc_write_mask(inode->i_mapping); int err = 0; int faili = 0; @@ -1298,7 +1298,7 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, struct iov_iter *i, loff_t pos) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct page **pages = NULL; unsigned long first_index; @@ -1486,7 +1486,7 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, unsigned long nr_segs, loff_t pos) { struct file *file = iocb->ki_filp; - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; loff_t *ppos = &iocb->ki_pos; u64 start_pos; @@ -2087,7 +2087,7 @@ out: static long btrfs_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct extent_state *cached_state = NULL; u64 cur_offset; u64 last_byte; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 16d9e8e191e6..02d946a61ddd 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -4342,7 +4342,7 @@ unsigned char btrfs_filetype_table[] = { static int btrfs_real_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_item *item; struct btrfs_dir_item *di; @@ -6737,7 +6737,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset) int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = fdentry(vma->vm_file)->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct btrfs_root *root = BTRFS_I(inode)->root; struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; struct btrfs_ordered_extent *ordered; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 4b4516770f05..61045adc3075 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -152,7 +152,7 @@ void btrfs_inherit_iflags(struct inode *inode, struct inode *dir) static int btrfs_ioctl_getflags(struct file *file, void __user *arg) { - struct btrfs_inode *ip = BTRFS_I(file->f_path.dentry->d_inode); + struct btrfs_inode *ip = BTRFS_I(file_inode(file)); unsigned int flags = btrfs_flags_to_ioctl(ip->flags); if (copy_to_user(arg, &flags, sizeof(flags))) @@ -177,7 +177,7 @@ static int check_flags(unsigned int flags) static int btrfs_ioctl_setflags(struct file *file, void __user *arg) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct btrfs_inode *ip = BTRFS_I(inode); struct btrfs_root *root = ip->root; struct btrfs_trans_handle *trans; @@ -310,7 +310,7 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) static int btrfs_ioctl_getversion(struct file *file, int __user *arg) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); return put_user(inode->i_generation, arg); } @@ -1317,7 +1317,7 @@ static noinline int btrfs_ioctl_resize(struct file *file, u64 new_size; u64 old_size; u64 devid = 1; - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_ioctl_vol_args *vol_args; struct btrfs_trans_handle *trans; struct btrfs_device *device = NULL; @@ -1483,8 +1483,8 @@ static noinline int btrfs_ioctl_snap_create_transid(struct file *file, goto out_drop_write; } - src_inode = src.file->f_path.dentry->d_inode; - if (src_inode->i_sb != file->f_path.dentry->d_inode->i_sb) { + src_inode = file_inode(src.file); + if (src_inode->i_sb != file_inode(file)->i_sb) { printk(KERN_INFO "btrfs: Snapshot src from " "another FS\n"); ret = -EINVAL; @@ -1576,7 +1576,7 @@ out: static noinline int btrfs_ioctl_subvol_getflags(struct file *file, void __user *arg) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; int ret = 0; u64 flags = 0; @@ -1598,7 +1598,7 @@ static noinline int btrfs_ioctl_subvol_getflags(struct file *file, static noinline int btrfs_ioctl_subvol_setflags(struct file *file, void __user *arg) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_trans_handle *trans; u64 root_flags; @@ -1892,7 +1892,7 @@ static noinline int btrfs_ioctl_tree_search(struct file *file, if (IS_ERR(args)) return PTR_ERR(args); - inode = fdentry(file)->d_inode; + inode = file_inode(file); ret = search_ioctl(inode, args); if (ret == 0 && copy_to_user(argp, args, sizeof(*args))) ret = -EFAULT; @@ -2002,7 +2002,7 @@ static noinline int btrfs_ioctl_ino_lookup(struct file *file, if (IS_ERR(args)) return PTR_ERR(args); - inode = fdentry(file)->d_inode; + inode = file_inode(file); if (args->treeid == 0) args->treeid = BTRFS_I(inode)->root->root_key.objectid; @@ -2178,7 +2178,7 @@ out: static int btrfs_ioctl_defrag(struct file *file, void __user *argp) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_ioctl_defrag_range_args *range; int ret; @@ -2237,7 +2237,7 @@ static int btrfs_ioctl_defrag(struct file *file, void __user *argp) /* the rest are all set to zero by kzalloc */ range->len = (u64)-1; } - ret = btrfs_defrag_file(fdentry(file)->d_inode, file, + ret = btrfs_defrag_file(file_inode(file), file, range, 0, 0); if (ret > 0) ret = 0; @@ -2285,7 +2285,7 @@ out: static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_ioctl_vol_args *vol_args; int ret; @@ -2408,7 +2408,7 @@ out: static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, u64 off, u64 olen, u64 destoff) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct fd src_file; struct inode *src; @@ -2454,7 +2454,7 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, if (src_file.file->f_path.mnt != file->f_path.mnt) goto out_fput; - src = src_file.file->f_dentry->d_inode; + src = file_inode(src_file.file); ret = -EINVAL; if (src == inode) @@ -2816,7 +2816,7 @@ static long btrfs_ioctl_clone_range(struct file *file, void __user *argp) */ static long btrfs_ioctl_trans_start(struct file *file) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_trans_handle *trans; int ret; @@ -2856,7 +2856,7 @@ out: static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_root *new_root; struct btrfs_dir_item *di; @@ -3080,7 +3080,7 @@ out: */ long btrfs_ioctl_trans_end(struct file *file) { - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_trans_handle *trans; @@ -3142,7 +3142,7 @@ static noinline long btrfs_ioctl_wait_sync(struct btrfs_root *root, static long btrfs_ioctl_scrub(struct file *file, void __user *arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_ioctl_scrub_args *sa; int ret; @@ -3433,7 +3433,7 @@ void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock, static long btrfs_ioctl_balance(struct file *file, void __user *arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_fs_info *fs_info = root->fs_info; struct btrfs_ioctl_balance_args *bargs; struct btrfs_balance_control *bctl; @@ -3573,7 +3573,7 @@ out: static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_ioctl_quota_ctl_args *sa; struct btrfs_trans_handle *trans = NULL; int ret; @@ -3632,7 +3632,7 @@ drop_write: static long btrfs_ioctl_qgroup_assign(struct file *file, void __user *arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_ioctl_qgroup_assign_args *sa; struct btrfs_trans_handle *trans; int ret; @@ -3679,7 +3679,7 @@ drop_write: static long btrfs_ioctl_qgroup_create(struct file *file, void __user *arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_ioctl_qgroup_create_args *sa; struct btrfs_trans_handle *trans; int ret; @@ -3725,7 +3725,7 @@ drop_write: static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; struct btrfs_ioctl_qgroup_limit_args *sa; struct btrfs_trans_handle *trans; int ret; @@ -3775,7 +3775,7 @@ static long btrfs_ioctl_set_received_subvol(struct file *file, void __user *arg) { struct btrfs_ioctl_received_subvol_args *sa = NULL; - struct inode *inode = fdentry(file)->d_inode; + struct inode *inode = file_inode(file); struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_root_item *root_item = &root->root_item; struct btrfs_trans_handle *trans; @@ -3855,7 +3855,7 @@ out: long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { - struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root; + struct btrfs_root *root = BTRFS_I(file_inode(file))->root; void __user *argp = (void __user *)arg; switch (cmd) { diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index 54454542ad40..f80df6b04648 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -4542,7 +4542,7 @@ long btrfs_ioctl_send(struct file *mnt_file, void __user *arg_) if (!capable(CAP_SYS_ADMIN)) return -EPERM; - send_root = BTRFS_I(fdentry(mnt_file)->d_inode)->root; + send_root = BTRFS_I(file_inode(mnt_file))->root; fs_info = send_root->fs_info; arg = memdup_user(arg_, sizeof(*arg)); diff --git a/fs/buffer.c b/fs/buffer.c index 7a75c3e0fd58..b8a8b4d64d8c 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -2332,7 +2332,7 @@ int __block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, get_block_t get_block) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); unsigned long end; loff_t size; int ret; @@ -2371,7 +2371,7 @@ int block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, get_block_t get_block) { int ret; - struct super_block *sb = vma->vm_file->f_path.dentry->d_inode->i_sb; + struct super_block *sb = file_inode(vma->vm_file)->i_sb; sb_start_pagefault(sb); diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 064d1a68d2c1..d4f81edd9a5d 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -195,7 +195,7 @@ static int ceph_releasepage(struct page *page, gfp_t g) */ static int readpage_nounlock(struct file *filp, struct page *page) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_osd_client *osdc = &ceph_inode_to_client(inode)->client->osdc; @@ -370,7 +370,7 @@ out: static int ceph_readpages(struct file *file, struct address_space *mapping, struct list_head *page_list, unsigned nr_pages) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_fs_client *fsc = ceph_inode_to_client(inode); int rc = 0; int max = 0; @@ -977,7 +977,7 @@ static int ceph_update_writeable_page(struct file *file, loff_t pos, unsigned len, struct page *page) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; loff_t page_off = pos & PAGE_CACHE_MASK; @@ -1086,7 +1086,7 @@ static int ceph_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, struct page **pagep, void **fsdata) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_file_info *fi = file->private_data; struct page *page; @@ -1144,7 +1144,7 @@ static int ceph_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_fs_client *fsc = ceph_inode_to_client(inode); struct ceph_mds_client *mdsc = fsc->mdsc; @@ -1228,7 +1228,7 @@ const struct address_space_operations ceph_aops = { */ static int ceph_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { - struct inode *inode = vma->vm_file->f_dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct page *page = vmf->page; struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; loff_t off = page_offset(page); diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index 8c1aabe93b67..6d797f46d772 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -238,7 +238,7 @@ static int note_last_dentry(struct ceph_file_info *fi, const char *name, static int ceph_readdir(struct file *filp, void *dirent, filldir_t filldir) { struct ceph_file_info *fi = filp->private_data; - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_fs_client *fsc = ceph_inode_to_client(inode); struct ceph_mds_client *mdsc = fsc->mdsc; @@ -1138,7 +1138,7 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size, loff_t *ppos) { struct ceph_file_info *cf = file->private_data; - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); int left; const int bufsize = 1024; @@ -1188,7 +1188,7 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size, static int ceph_dir_fsync(struct file *file, loff_t start, loff_t end, int datasync) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct list_head *head = &ci->i_unsafe_dirops; struct ceph_mds_request *req; diff --git a/fs/ceph/file.c b/fs/ceph/file.c index e51558fca3a3..11b57c2c8f15 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -393,7 +393,7 @@ more: static ssize_t ceph_sync_read(struct file *file, char __user *data, unsigned len, loff_t *poff, int *checkeof) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct page **pages; u64 off = *poff; int num_pages, ret; @@ -466,7 +466,7 @@ static void sync_write_commit(struct ceph_osd_request *req, static ssize_t ceph_sync_write(struct file *file, const char __user *data, size_t left, loff_t *offset) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_fs_client *fsc = ceph_inode_to_client(inode); struct ceph_osd_request *req; @@ -483,7 +483,7 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data, int ret; struct timespec mtime = CURRENT_TIME; - if (ceph_snap(file->f_dentry->d_inode) != CEPH_NOSNAP) + if (ceph_snap(file_inode(file)) != CEPH_NOSNAP) return -EROFS; dout("sync_write on file %p %lld~%u %s\n", file, *offset, @@ -637,7 +637,7 @@ static ssize_t ceph_aio_read(struct kiocb *iocb, const struct iovec *iov, struct ceph_file_info *fi = filp->private_data; loff_t *ppos = &iocb->ki_pos; size_t len = iov->iov_len; - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct ceph_inode_info *ci = ceph_inode(inode); void __user *base = iov->iov_base; ssize_t ret; @@ -707,7 +707,7 @@ static ssize_t ceph_aio_write(struct kiocb *iocb, const struct iovec *iov, { struct file *file = iocb->ki_filp; struct ceph_file_info *fi = file->private_data; - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_osd_client *osdc = &ceph_sb_to_client(inode->i_sb)->client->osdc; diff --git a/fs/ceph/ioctl.c b/fs/ceph/ioctl.c index 36549a46e311..f5ed767806df 100644 --- a/fs/ceph/ioctl.c +++ b/fs/ceph/ioctl.c @@ -16,11 +16,11 @@ */ static long ceph_ioctl_get_layout(struct file *file, void __user *arg) { - struct ceph_inode_info *ci = ceph_inode(file->f_dentry->d_inode); + struct ceph_inode_info *ci = ceph_inode(file_inode(file)); struct ceph_ioctl_layout l; int err; - err = ceph_do_getattr(file->f_dentry->d_inode, CEPH_STAT_CAP_LAYOUT); + err = ceph_do_getattr(file_inode(file), CEPH_STAT_CAP_LAYOUT); if (!err) { l.stripe_unit = ceph_file_layout_su(ci->i_layout); l.stripe_count = ceph_file_layout_stripe_count(ci->i_layout); @@ -63,12 +63,12 @@ static long __validate_layout(struct ceph_mds_client *mdsc, static long ceph_ioctl_set_layout(struct file *file, void __user *arg) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct inode *parent_inode; struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; struct ceph_mds_request *req; struct ceph_ioctl_layout l; - struct ceph_inode_info *ci = ceph_inode(file->f_dentry->d_inode); + struct ceph_inode_info *ci = ceph_inode(file_inode(file)); struct ceph_ioctl_layout nl; int err; @@ -76,7 +76,7 @@ static long ceph_ioctl_set_layout(struct file *file, void __user *arg) return -EFAULT; /* validate changed params against current layout */ - err = ceph_do_getattr(file->f_dentry->d_inode, CEPH_STAT_CAP_LAYOUT); + err = ceph_do_getattr(file_inode(file), CEPH_STAT_CAP_LAYOUT); if (err) return err; @@ -136,7 +136,7 @@ static long ceph_ioctl_set_layout(struct file *file, void __user *arg) */ static long ceph_ioctl_set_layout_policy (struct file *file, void __user *arg) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_mds_request *req; struct ceph_ioctl_layout l; int err; @@ -179,7 +179,7 @@ static long ceph_ioctl_set_layout_policy (struct file *file, void __user *arg) static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) { struct ceph_ioctl_dataloc dl; - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_osd_client *osdc = &ceph_sb_to_client(inode->i_sb)->client->osdc; @@ -234,7 +234,7 @@ static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) static long ceph_ioctl_lazyio(struct file *file) { struct ceph_file_info *fi = file->private_data; - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); if ((fi->fmode & CEPH_FILE_MODE_LAZY) == 0) { diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c index 80576d05d687..202dd3d68be0 100644 --- a/fs/ceph/locks.c +++ b/fs/ceph/locks.c @@ -13,7 +13,7 @@ static int ceph_lock_message(u8 lock_type, u16 operation, struct file *file, int cmd, u8 wait, struct file_lock *fl) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; struct ceph_mds_request *req; diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index de7f9168a118..8b35365c70be 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -677,7 +677,7 @@ out_nls: static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { - struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(iocb->ki_filp); ssize_t written; int rc; @@ -701,7 +701,7 @@ static loff_t cifs_llseek(struct file *file, loff_t offset, int whence) */ if (whence != SEEK_SET && whence != SEEK_CUR) { int rc; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); /* * We need to be sure that all dirty pages are written and the @@ -733,7 +733,7 @@ static int cifs_setlease(struct file *file, long arg, struct file_lock **lease) { /* note that this is called by vfs setlease with lock_flocks held to protect *lease from going away */ - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct cifsFileInfo *cfile = file->private_data; if (!(S_ISREG(inode->i_mode))) diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 8ea6ca50a665..1a5c2911b043 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -947,7 +947,7 @@ static int cifs_posix_lock_test(struct file *file, struct file_lock *flock) { int rc = 0; - struct cifsInodeInfo *cinode = CIFS_I(file->f_path.dentry->d_inode); + struct cifsInodeInfo *cinode = CIFS_I(file_inode(file)); unsigned char saved_type = flock->fl_type; if ((flock->fl_flags & FL_POSIX) == 0) @@ -974,7 +974,7 @@ cifs_posix_lock_test(struct file *file, struct file_lock *flock) static int cifs_posix_lock_set(struct file *file, struct file_lock *flock) { - struct cifsInodeInfo *cinode = CIFS_I(file->f_path.dentry->d_inode); + struct cifsInodeInfo *cinode = CIFS_I(file_inode(file)); int rc = 1; if ((flock->fl_flags & FL_POSIX) == 0) @@ -1548,7 +1548,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *flock) cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); netfid = cfile->fid.netfid; - cinode = CIFS_I(file->f_path.dentry->d_inode); + cinode = CIFS_I(file_inode(file)); if (cap_unix(tcon->ses) && (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) && @@ -2171,7 +2171,7 @@ int cifs_strict_fsync(struct file *file, loff_t start, loff_t end, struct cifs_tcon *tcon; struct TCP_Server_Info *server; struct cifsFileInfo *smbfile = file->private_data; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); rc = filemap_write_and_wait_range(inode->i_mapping, start, end); @@ -2246,7 +2246,7 @@ int cifs_fsync(struct file *file, loff_t start, loff_t end, int datasync) */ int cifs_flush(struct file *file, fl_owner_t id) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); int rc = 0; if (file->f_mode & FMODE_WRITE) @@ -2480,7 +2480,7 @@ ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov, ssize_t written; struct inode *inode; - inode = iocb->ki_filp->f_path.dentry->d_inode; + inode = file_inode(iocb->ki_filp); /* * BB - optimize the way when signing is disabled. We can drop this @@ -2543,7 +2543,7 @@ ssize_t cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { - struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(iocb->ki_filp); struct cifsInodeInfo *cinode = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsFileInfo *cfile = (struct cifsFileInfo *) @@ -2915,7 +2915,7 @@ ssize_t cifs_strict_readv(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { - struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(iocb->ki_filp); struct cifsInodeInfo *cinode = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsFileInfo *cfile = (struct cifsFileInfo *) @@ -3063,7 +3063,7 @@ static struct vm_operations_struct cifs_file_vm_ops = { int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma) { int rc, xid; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); xid = get_xid(); @@ -3356,7 +3356,7 @@ static int cifs_readpage_worker(struct file *file, struct page *page, int rc; /* Is the page cached? */ - rc = cifs_readpage_from_fscache(file->f_path.dentry->d_inode, page); + rc = cifs_readpage_from_fscache(file_inode(file), page); if (rc == 0) goto read_complete; @@ -3371,8 +3371,8 @@ static int cifs_readpage_worker(struct file *file, struct page *page, else cFYI(1, "Bytes read %d", rc); - file->f_path.dentry->d_inode->i_atime = - current_fs_time(file->f_path.dentry->d_inode->i_sb); + file_inode(file)->i_atime = + current_fs_time(file_inode(file)->i_sb); if (PAGE_CACHE_SIZE > rc) memset(read_data + rc, 0, PAGE_CACHE_SIZE - rc); @@ -3381,7 +3381,7 @@ static int cifs_readpage_worker(struct file *file, struct page *page, SetPageUptodate(page); /* send this page to the cache */ - cifs_readpage_to_fscache(file->f_path.dentry->d_inode, page); + cifs_readpage_to_fscache(file_inode(file), page); rc = 0; diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index ed6208ff85a7..1fc864b92cf2 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -289,7 +289,7 @@ cifs_get_file_info_unix(struct file *filp) unsigned int xid; FILE_UNIX_BASIC_INFO find_data; struct cifs_fattr fattr; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsFileInfo *cfile = filp->private_data; struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); @@ -558,7 +558,7 @@ cifs_get_file_info(struct file *filp) unsigned int xid; FILE_ALL_INFO find_data; struct cifs_fattr fattr; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsFileInfo *cfile = filp->private_data; struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); @@ -1678,7 +1678,7 @@ cifs_invalidate_mapping(struct inode *inode) int cifs_revalidate_file_attr(struct file *filp) { int rc = 0; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct cifsFileInfo *cfile = (struct cifsFileInfo *) filp->private_data; if (!cifs_inode_needs_reval(inode)) @@ -1735,7 +1735,7 @@ out: int cifs_revalidate_file(struct file *filp) { int rc; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); rc = cifs_revalidate_file_attr(filp); if (rc) diff --git a/fs/cifs/ioctl.c b/fs/cifs/ioctl.c index fd5009d56f9f..6c9f1214cf0b 100644 --- a/fs/cifs/ioctl.c +++ b/fs/cifs/ioctl.c @@ -30,7 +30,7 @@ long cifs_ioctl(struct file *filep, unsigned int command, unsigned long arg) { - struct inode *inode = filep->f_dentry->d_inode; + struct inode *inode = file_inode(filep); int rc = -ENOTTY; /* strange error - but the precedent */ unsigned int xid; struct cifs_sb_info *cifs_sb; diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index cdd6ff48246b..7255b0c7aa7e 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -505,7 +505,7 @@ static int cifs_entry_is_dot(struct cifs_dirent *de, bool is_unicode) whether we can use the cached search results from the previous search */ static int is_dir_changed(struct file *file) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct cifsInodeInfo *cifsInfo = CIFS_I(inode); if (cifsInfo->time == 0) @@ -778,7 +778,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) switch ((int) file->f_pos) { case 0: if (filldir(direntry, ".", 1, file->f_pos, - file->f_path.dentry->d_inode->i_ino, DT_DIR) < 0) { + file_inode(file)->i_ino, DT_DIR) < 0) { cERROR(1, "Filldir for current dir failed"); rc = -ENOMEM; break; diff --git a/fs/coda/dir.c b/fs/coda/dir.c index 49fe52d25600..b7d3a05c062c 100644 --- a/fs/coda/dir.c +++ b/fs/coda/dir.c @@ -397,7 +397,7 @@ static int coda_readdir(struct file *coda_file, void *buf, filldir_t filldir) * We can't use vfs_readdir because we have to keep the file * position in sync between the coda_file and the host_file. * and as such we need grab the inode mutex. */ - struct inode *host_inode = host_file->f_path.dentry->d_inode; + struct inode *host_inode = file_inode(host_file); mutex_lock(&host_inode->i_mutex); host_file->f_pos = coda_file->f_pos; diff --git a/fs/coda/file.c b/fs/coda/file.c index 8edd404e6419..fa4c100bdc7d 100644 --- a/fs/coda/file.c +++ b/fs/coda/file.c @@ -66,7 +66,7 @@ coda_file_splice_read(struct file *coda_file, loff_t *ppos, static ssize_t coda_file_write(struct file *coda_file, const char __user *buf, size_t count, loff_t *ppos) { - struct inode *host_inode, *coda_inode = coda_file->f_path.dentry->d_inode; + struct inode *host_inode, *coda_inode = file_inode(coda_file); struct coda_file_info *cfi; struct file *host_file; ssize_t ret; @@ -78,7 +78,7 @@ coda_file_write(struct file *coda_file, const char __user *buf, size_t count, lo if (!host_file->f_op || !host_file->f_op->write) return -EINVAL; - host_inode = host_file->f_path.dentry->d_inode; + host_inode = file_inode(host_file); mutex_lock(&coda_inode->i_mutex); ret = host_file->f_op->write(host_file, buf, count, ppos); @@ -106,8 +106,8 @@ coda_file_mmap(struct file *coda_file, struct vm_area_struct *vma) if (!host_file->f_op || !host_file->f_op->mmap) return -ENODEV; - coda_inode = coda_file->f_path.dentry->d_inode; - host_inode = host_file->f_path.dentry->d_inode; + coda_inode = file_inode(coda_file); + host_inode = file_inode(host_file); cii = ITOC(coda_inode); spin_lock(&cii->c_lock); @@ -178,7 +178,7 @@ int coda_release(struct inode *coda_inode, struct file *coda_file) err = venus_close(coda_inode->i_sb, coda_i2f(coda_inode), coda_flags, coda_file->f_cred->fsuid); - host_inode = cfi->cfi_container->f_path.dentry->d_inode; + host_inode = file_inode(cfi->cfi_container); cii = ITOC(coda_inode); /* did we mmap this file? */ @@ -202,7 +202,7 @@ int coda_release(struct inode *coda_inode, struct file *coda_file) int coda_fsync(struct file *coda_file, loff_t start, loff_t end, int datasync) { struct file *host_file; - struct inode *coda_inode = coda_file->f_path.dentry->d_inode; + struct inode *coda_inode = file_inode(coda_file); struct coda_file_info *cfi; int err; diff --git a/fs/coda/inode.c b/fs/coda/inode.c index be2aa4909487..6df708c7b3e8 100644 --- a/fs/coda/inode.c +++ b/fs/coda/inode.c @@ -129,7 +129,7 @@ static int get_device_index(struct coda_mount_data *data) f = fdget(data->fd); if (!f.file) goto Ebadf; - inode = f.file->f_path.dentry->d_inode; + inode = file_inode(f.file); if (!S_ISCHR(inode->i_mode) || imajor(inode) != CODA_PSDEV_MAJOR) { fdput(f); goto Ebadf; diff --git a/fs/coda/pioctl.c b/fs/coda/pioctl.c index ee0981f1375b..3f5de96bbb58 100644 --- a/fs/coda/pioctl.c +++ b/fs/coda/pioctl.c @@ -52,7 +52,7 @@ static long coda_pioctl(struct file *filp, unsigned int cmd, struct path path; int error; struct PioctlData data; - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct inode *target_inode = NULL; struct coda_inode_info *cnp; diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index e2f57a007029..3ced75f765ca 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -1582,7 +1582,7 @@ asmlinkage long compat_sys_ioctl(unsigned int fd, unsigned int cmd, case FIBMAP: case FIGETBSZ: case FIONREAD: - if (S_ISREG(f.file->f_path.dentry->d_inode->i_mode)) + if (S_ISREG(file_inode(f.file)->i_mode)) break; /*FALL THROUGH*/ diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 712b10f64c70..90d222f11e36 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1625,7 +1625,7 @@ static loff_t configfs_dir_lseek(struct file *file, loff_t offset, int whence) if (offset >= 0) break; default: - mutex_unlock(&file->f_path.dentry->d_inode->i_mutex); + mutex_unlock(&file_inode(file)->i_mutex); return -EINVAL; } if (offset != file->f_pos) { diff --git a/fs/coredump.c b/fs/coredump.c index 177493272a61..69baf903d3bd 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -411,7 +411,7 @@ static void wait_for_dump_helpers(struct file *file) { struct pipe_inode_info *pipe; - pipe = file->f_path.dentry->d_inode->i_pipe; + pipe = file_inode(file)->i_pipe; pipe_lock(pipe); pipe->readers++; @@ -600,7 +600,7 @@ void do_coredump(siginfo_t *siginfo) if (IS_ERR(cprm.file)) goto fail_unlock; - inode = cprm.file->f_path.dentry->d_inode; + inode = file_inode(cprm.file); if (inode->i_nlink > 1) goto close_fail; if (d_unhashed(cprm.file->f_path.dentry)) diff --git a/fs/cramfs/inode.c b/fs/cramfs/inode.c index c6c3f91ecf06..3ceb9ec976e1 100644 --- a/fs/cramfs/inode.c +++ b/fs/cramfs/inode.c @@ -351,7 +351,7 @@ static int cramfs_statfs(struct dentry *dentry, struct kstatfs *buf) */ static int cramfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; char *buf; unsigned int offset; diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index d45ba4568128..53acc9d0c138 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -118,7 +118,7 @@ static int ecryptfs_readdir(struct file *file, void *dirent, filldir_t filldir) lower_file = ecryptfs_file_to_lower(file); lower_file->f_pos = file->f_pos; - inode = file->f_path.dentry->d_inode; + inode = file_inode(file); memset(&buf, 0, sizeof(buf)); buf.dirent = dirent; buf.dentry = file->f_path.dentry; @@ -133,7 +133,7 @@ static int ecryptfs_readdir(struct file *file, void *dirent, filldir_t filldir) goto out; if (rc >= 0) fsstack_copy_attr_atime(inode, - lower_file->f_path.dentry->d_inode); + file_inode(lower_file)); out: return rc; } diff --git a/fs/efs/dir.c b/fs/efs/dir.c index 7ee6f7e3a608..055a9e9ca747 100644 --- a/fs/efs/dir.c +++ b/fs/efs/dir.c @@ -20,7 +20,7 @@ const struct inode_operations efs_dir_inode_operations = { }; static int efs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct buffer_head *bh; struct efs_dir *dirblock; diff --git a/fs/exec.c b/fs/exec.c index 20df02c1cc70..7b6f4d59b26c 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -123,7 +123,7 @@ SYSCALL_DEFINE1(uselib, const char __user *, library) goto out; error = -EINVAL; - if (!S_ISREG(file->f_path.dentry->d_inode->i_mode)) + if (!S_ISREG(file_inode(file)->i_mode)) goto exit; error = -EACCES; @@ -764,7 +764,7 @@ struct file *open_exec(const char *name) goto out; err = -EACCES; - if (!S_ISREG(file->f_path.dentry->d_inode->i_mode)) + if (!S_ISREG(file_inode(file)->i_mode)) goto exit; if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) @@ -1098,7 +1098,7 @@ EXPORT_SYMBOL(flush_old_exec); void would_dump(struct linux_binprm *bprm, struct file *file) { - if (inode_permission(file->f_path.dentry->d_inode, MAY_READ) < 0) + if (inode_permission(file_inode(file), MAY_READ) < 0) bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP; } EXPORT_SYMBOL(would_dump); @@ -1270,7 +1270,7 @@ static int check_unsafe_exec(struct linux_binprm *bprm) int prepare_binprm(struct linux_binprm *bprm) { umode_t mode; - struct inode * inode = bprm->file->f_path.dentry->d_inode; + struct inode * inode = file_inode(bprm->file); int retval; mode = inode->i_mode; diff --git a/fs/exofs/dir.c b/fs/exofs/dir.c index c61e62ac231c..46375896cfc0 100644 --- a/fs/exofs/dir.c +++ b/fs/exofs/dir.c @@ -242,7 +242,7 @@ static int exofs_readdir(struct file *filp, void *dirent, filldir_t filldir) { loff_t pos = filp->f_pos; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); unsigned int offset = pos & ~PAGE_CACHE_MASK; unsigned long n = pos >> PAGE_CACHE_SHIFT; unsigned long npages = dir_pages(inode); diff --git a/fs/ext2/dir.c b/fs/ext2/dir.c index 0f4f5c929257..4237722bfd27 100644 --- a/fs/ext2/dir.c +++ b/fs/ext2/dir.c @@ -290,7 +290,7 @@ static int ext2_readdir (struct file * filp, void * dirent, filldir_t filldir) { loff_t pos = filp->f_pos; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; unsigned int offset = pos & ~PAGE_CACHE_MASK; unsigned long n = pos >> PAGE_CACHE_SHIFT; diff --git a/fs/ext2/ioctl.c b/fs/ext2/ioctl.c index 2de655f5d625..5d46c09863f0 100644 --- a/fs/ext2/ioctl.c +++ b/fs/ext2/ioctl.c @@ -19,7 +19,7 @@ long ext2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct ext2_inode_info *ei = EXT2_I(inode); unsigned int flags; unsigned short rsv_window_size; diff --git a/fs/ext3/dir.c b/fs/ext3/dir.c index dd91264ba94f..87eccbbca255 100644 --- a/fs/ext3/dir.c +++ b/fs/ext3/dir.c @@ -99,7 +99,7 @@ static int ext3_readdir(struct file * filp, int i, stored; struct ext3_dir_entry_2 *de; int err; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; int ret = 0; int dir_has_error = 0; @@ -114,7 +114,7 @@ static int ext3_readdir(struct file * filp, * We don't set the inode dirty flag since it's not * critical that it get flushed back to the disk. */ - EXT3_I(filp->f_path.dentry->d_inode)->i_flags &= ~EXT3_INDEX_FL; + EXT3_I(file_inode(filp))->i_flags &= ~EXT3_INDEX_FL; } stored = 0; offset = filp->f_pos & (sb->s_blocksize - 1); @@ -457,7 +457,7 @@ static int call_filldir(struct file * filp, void * dirent, { struct dir_private_info *info = filp->private_data; loff_t curr_pos; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block * sb; int error; @@ -487,7 +487,7 @@ static int ext3_dx_readdir(struct file * filp, void * dirent, filldir_t filldir) { struct dir_private_info *info = filp->private_data; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct fname *fname; int ret; diff --git a/fs/ext3/ioctl.c b/fs/ext3/ioctl.c index 677a5c27dc69..4d96e9a64532 100644 --- a/fs/ext3/ioctl.c +++ b/fs/ext3/ioctl.c @@ -14,7 +14,7 @@ long ext3_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct ext3_inode_info *ei = EXT3_I(inode); unsigned int flags; unsigned short rsv_window_size; diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 890b8947c546..61fa09eb2501 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -624,7 +624,7 @@ int ext3_htree_fill_tree(struct file *dir_file, __u32 start_hash, dxtrace(printk("In htree_fill_tree, start hash: %x:%x\n", start_hash, start_minor_hash)); - dir = dir_file->f_path.dentry->d_inode; + dir = file_inode(dir_file); if (!(EXT3_I(dir)->i_flags & EXT3_INDEX_FL)) { hinfo.hash_version = EXT3_SB(dir->i_sb)->s_def_hash_version; if (hinfo.hash_version <= DX_HASH_TEA) @@ -638,7 +638,7 @@ int ext3_htree_fill_tree(struct file *dir_file, __u32 start_hash, } hinfo.hash = start_hash; hinfo.minor_hash = 0; - frame = dx_probe(NULL, dir_file->f_path.dentry->d_inode, &hinfo, frames, &err); + frame = dx_probe(NULL, file_inode(dir_file), &hinfo, frames, &err); if (!frame) return err; diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index 80a28b297279..dc149d123de5 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -110,7 +110,7 @@ static int ext4_readdir(struct file *filp, int i, stored; struct ext4_dir_entry_2 *de; int err; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; int ret = 0; int dir_has_error = 0; @@ -133,7 +133,7 @@ static int ext4_readdir(struct file *filp, * We don't set the inode dirty flag since it's not * critical that it get flushed back to the disk. */ - ext4_clear_inode_flag(filp->f_path.dentry->d_inode, + ext4_clear_inode_flag(file_inode(filp), EXT4_INODE_INDEX); } stored = 0; @@ -494,7 +494,7 @@ static int call_filldir(struct file *filp, void *dirent, { struct dir_private_info *info = filp->private_data; loff_t curr_pos; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb; int error; @@ -526,7 +526,7 @@ static int ext4_dx_readdir(struct file *filp, void *dirent, filldir_t filldir) { struct dir_private_info *info = filp->private_data; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct fname *fname; int ret; diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 5ae1674ec12f..7817ca7c2bbf 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4386,7 +4386,7 @@ static void ext4_falloc_update_inode(struct inode *inode, */ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); handle_t *handle; loff_t new_size; unsigned int max_blocks; @@ -4643,7 +4643,7 @@ static int ext4_xattr_fiemap(struct inode *inode, */ int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct super_block *sb = inode->i_sb; ext4_lblk_t first_block, stop_block; struct address_space *mapping = inode->i_mapping; diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 405565a62277..c00ea7945eb5 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -167,7 +167,7 @@ static ssize_t ext4_file_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { - struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(iocb->ki_filp); ssize_t ret; /* diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index 387c47c6cda9..8106dca95456 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -1298,7 +1298,7 @@ int ext4_read_inline_dir(struct file *filp, int i, stored; struct ext4_dir_entry_2 *de; struct super_block *sb; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int ret, inline_size = 0; struct ext4_iloc iloc; void *dir_buf = NULL; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index cbfe13bf5b2a..521bd4ab8abe 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2959,7 +2959,7 @@ static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, ssize_t size, void *private, int ret, bool is_async) { - struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(iocb->ki_filp); ext4_io_end_t *io_end = iocb->private; /* if not async direct IO or dio with 0 bytes write, just return */ @@ -3553,7 +3553,7 @@ int ext4_can_truncate(struct inode *inode) int ext4_punch_hole(struct file *file, loff_t offset, loff_t length) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); if (!S_ISREG(inode->i_mode)) return -EOPNOTSUPP; @@ -4926,7 +4926,7 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) unsigned long len; int ret; struct file *file = vma->vm_file; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct address_space *mapping = inode->i_mapping; handle_t *handle; get_block_t *get_block; diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 5747f52f7c72..c2f8e060f636 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -22,7 +22,7 @@ long ext4_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; struct ext4_inode_info *ei = EXT4_I(inode); unsigned int flags; diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index d9cc5ee42f53..796f7ac03706 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -900,7 +900,7 @@ move_extent_per_page(struct file *o_filp, struct inode *donor_inode, pgoff_t orig_page_offset, int data_offset_in_page, int block_len_in_page, int uninit, int *err) { - struct inode *orig_inode = o_filp->f_dentry->d_inode; + struct inode *orig_inode = file_inode(o_filp); struct page *pagep[2] = {NULL, NULL}; handle_t *handle; ext4_lblk_t orig_blk_offset; @@ -1279,8 +1279,8 @@ ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 orig_start, __u64 donor_start, __u64 len, __u64 *moved_len) { - struct inode *orig_inode = o_filp->f_dentry->d_inode; - struct inode *donor_inode = d_filp->f_dentry->d_inode; + struct inode *orig_inode = file_inode(o_filp); + struct inode *donor_inode = file_inode(d_filp); struct ext4_ext_path *orig_path = NULL, *holecheck_path = NULL; struct ext4_extent *ext_prev, *ext_cur, *ext_dummy; ext4_lblk_t block_start = orig_start; diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index f9ed946a448e..bb97ad6905b2 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -937,7 +937,7 @@ int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash, dxtrace(printk(KERN_DEBUG "In htree_fill_tree, start hash: %x:%x\n", start_hash, start_minor_hash)); - dir = dir_file->f_path.dentry->d_inode; + dir = file_inode(dir_file); if (!(ext4_test_inode_flag(dir, EXT4_INODE_INDEX))) { hinfo.hash_version = EXT4_SB(dir->i_sb)->s_def_hash_version; if (hinfo.hash_version <= DX_HASH_TEA) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 3d4fb81bacd5..4df78dd3f523 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -553,7 +553,7 @@ void ext4_error_file(struct file *file, const char *function, va_list args; struct va_format vaf; struct ext4_super_block *es; - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); char pathname[80], *path; es = EXT4_SB(inode->i_sb)->s_es; diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 951ed52748f6..fda0bcc0907f 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -603,7 +603,7 @@ bool f2fs_empty_dir(struct inode *dir) static int f2fs_readdir(struct file *file, void *dirent, filldir_t filldir) { unsigned long pos = file->f_pos; - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); unsigned long npages = dir_blocks(inode); unsigned char *types = NULL; unsigned int bit_pos = 0, start_bit_pos = 0; diff --git a/fs/fat/dir.c b/fs/fat/dir.c index 58bf744dbf39..165012ef363a 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -698,7 +698,7 @@ out: static int fat_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); return __fat_readdir(inode, filp, dirent, filldir, 0, 0); } @@ -779,7 +779,7 @@ static int fat_ioctl_readdir(struct inode *inode, struct file *filp, static long fat_dir_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct __fat_dirent __user *d1 = (struct __fat_dirent __user *)arg; int short_only, both; @@ -819,7 +819,7 @@ FAT_IOCTL_FILLDIR_FUNC(fat_compat_ioctl_filldir, compat_dirent) static long fat_compat_dir_ioctl(struct file *filp, unsigned cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct compat_dirent __user *d1 = compat_ptr(arg); int short_only, both; diff --git a/fs/fat/file.c b/fs/fat/file.c index a62e0ecbe2db..3978f8ca1823 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -32,7 +32,7 @@ static int fat_ioctl_get_attributes(struct inode *inode, u32 __user *user_attr) static int fat_ioctl_set_attributes(struct file *file, u32 __user *user_attr) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct msdos_sb_info *sbi = MSDOS_SB(inode->i_sb); int is_dir = S_ISDIR(inode->i_mode); u32 attr, oldattr; @@ -116,7 +116,7 @@ out: long fat_generic_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); u32 __user *user_attr = (u32 __user *)arg; switch (cmd) { diff --git a/fs/fcntl.c b/fs/fcntl.c index 71a600a19f06..6599222536eb 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -30,7 +30,7 @@ static int setfl(int fd, struct file * filp, unsigned long arg) { - struct inode * inode = filp->f_path.dentry->d_inode; + struct inode * inode = file_inode(filp); int error = 0; /* diff --git a/fs/file_table.c b/fs/file_table.c index de9e9653d611..0f607ce89acc 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -447,7 +447,7 @@ void mark_files_ro(struct super_block *sb) lg_global_lock(&files_lglock); do_file_list_for_each_entry(sb, f) { - if (!S_ISREG(f->f_path.dentry->d_inode->i_mode)) + if (!S_ISREG(file_inode(f)->i_mode)) continue; if (!file_count(f)) continue; diff --git a/fs/freevxfs/vxfs_lookup.c b/fs/freevxfs/vxfs_lookup.c index bd447e88f208..664b07a53870 100644 --- a/fs/freevxfs/vxfs_lookup.c +++ b/fs/freevxfs/vxfs_lookup.c @@ -237,7 +237,7 @@ vxfs_lookup(struct inode *dip, struct dentry *dp, unsigned int flags) static int vxfs_readdir(struct file *fp, void *retp, filldir_t filler) { - struct inode *ip = fp->f_path.dentry->d_inode; + struct inode *ip = file_inode(fp); struct super_block *sbp = ip->i_sb; u_long bsize = sbp->s_blocksize; u_long page, npages, block, pblocks, nblocks, offset; diff --git a/fs/fuse/control.c b/fs/fuse/control.c index 75a20c092dd4..b7978b9f75ef 100644 --- a/fs/fuse/control.c +++ b/fs/fuse/control.c @@ -23,7 +23,7 @@ static struct fuse_conn *fuse_ctl_file_conn_get(struct file *file) { struct fuse_conn *fc; mutex_lock(&fuse_mutex); - fc = file->f_path.dentry->d_inode->i_private; + fc = file_inode(file)->i_private; if (fc) fc = fuse_conn_get(fc); mutex_unlock(&fuse_mutex); diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index b7c09f9eb40c..80ba3950c40d 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1160,7 +1160,7 @@ static int fuse_readdir(struct file *file, void *dstbuf, filldir_t filldir) int err; size_t nbytes; struct page *page; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_req *req; diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 991ab2d484dd..44543df9f400 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -157,7 +157,7 @@ static const u32 gfs2_to_fsflags[32] = { static int gfs2_get_flags(struct file *filp, u32 __user *ptr) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_holder gh; int error; @@ -217,7 +217,7 @@ void gfs2_set_inode_flags(struct inode *inode) */ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); struct buffer_head *bh; @@ -293,7 +293,7 @@ out_drop_write: static int gfs2_set_flags(struct file *filp, u32 __user *ptr) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); u32 fsflags, gfsflags; if (get_user(fsflags, ptr)) @@ -336,7 +336,7 @@ static long gfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) static void gfs2_size_hint(struct file *filep, loff_t offset, size_t size) { - struct inode *inode = filep->f_dentry->d_inode; + struct inode *inode = file_inode(filep); struct gfs2_sbd *sdp = GFS2_SB(inode); struct gfs2_inode *ip = GFS2_I(inode); size_t blks = (size + sdp->sd_sb.sb_bsize - 1) >> sdp->sd_sb.sb_bsize_shift; @@ -386,7 +386,7 @@ static int gfs2_allocate_page_backing(struct page *page) static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); unsigned long last_index; @@ -673,8 +673,7 @@ static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov, { struct file *file = iocb->ki_filp; size_t writesize = iov_length(iov, nr_segs); - struct dentry *dentry = file->f_dentry; - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); + struct gfs2_inode *ip = GFS2_I(file_inode(file)); int ret; ret = gfs2_rs_alloc(ip); @@ -772,7 +771,7 @@ static void calc_max_reserv(struct gfs2_inode *ip, loff_t max, loff_t *len, static long gfs2_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct gfs2_sbd *sdp = GFS2_SB(inode); struct gfs2_inode *ip = GFS2_I(inode); unsigned int data_blocks = 0, ind_blocks = 0, rblocks; @@ -938,7 +937,7 @@ static int do_flock(struct file *file, int cmd, struct file_lock *fl) { struct gfs2_file *fp = file->private_data; struct gfs2_holder *fl_gh = &fp->f_fl_gh; - struct gfs2_inode *ip = GFS2_I(file->f_path.dentry->d_inode); + struct gfs2_inode *ip = GFS2_I(file_inode(file)); struct gfs2_glock *gl; unsigned int state; int flags; diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index b7eff078fe90..04af1cf7ae34 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -1257,7 +1257,7 @@ fail: int gfs2_fitrim(struct file *filp, void __user *argp) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct gfs2_sbd *sdp = GFS2_SB(inode); struct request_queue *q = bdev_get_queue(sdp->sd_vfs->s_bdev); struct buffer_head *bh; diff --git a/fs/hfs/dir.c b/fs/hfs/dir.c index 422dde2ec0a1..5f7f1abd5f6d 100644 --- a/fs/hfs/dir.c +++ b/fs/hfs/dir.c @@ -51,7 +51,7 @@ done: */ static int hfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; int len, err; char strbuf[HFS_MAX_NAMELEN]; diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index d47f11658c17..3031dfdd2358 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -128,7 +128,7 @@ static ssize_t hfs_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; - struct inode *inode = file->f_path.dentry->d_inode->i_mapping->host; + struct inode *inode = file_inode(file)->i_mapping->host; ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, diff --git a/fs/hfsplus/dir.c b/fs/hfsplus/dir.c index 6b9f921ef2fa..074e04589248 100644 --- a/fs/hfsplus/dir.c +++ b/fs/hfsplus/dir.c @@ -122,7 +122,7 @@ fail: static int hfsplus_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; int len, err; char strbuf[HFSPLUS_MAX_STRLEN + 1]; diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index 799b336b59f9..dcd05be5344b 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -124,7 +124,7 @@ static ssize_t hfsplus_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; - struct inode *inode = file->f_path.dentry->d_inode->i_mapping->host; + struct inode *inode = file_inode(file)->i_mapping->host; ssize_t ret; ret = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs, diff --git a/fs/hfsplus/ioctl.c b/fs/hfsplus/ioctl.c index 09addc8615fa..e3c4c4209428 100644 --- a/fs/hfsplus/ioctl.c +++ b/fs/hfsplus/ioctl.c @@ -59,7 +59,7 @@ static int hfsplus_ioctl_bless(struct file *file, int __user *user_flags) static int hfsplus_ioctl_getflags(struct file *file, int __user *user_flags) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct hfsplus_inode_info *hip = HFSPLUS_I(inode); unsigned int flags = 0; @@ -75,7 +75,7 @@ static int hfsplus_ioctl_getflags(struct file *file, int __user *user_flags) static int hfsplus_ioctl_setflags(struct file *file, int __user *user_flags) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct hfsplus_inode_info *hip = HFSPLUS_I(inode); unsigned int flags; int err = 0; diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index 457addc5c91f..ba6de25771ac 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -30,7 +30,7 @@ static inline struct hostfs_inode_info *HOSTFS_I(struct inode *inode) return list_entry(inode, struct hostfs_inode_info, vfs_inode); } -#define FILE_HOSTFS_I(file) HOSTFS_I((file)->f_path.dentry->d_inode) +#define FILE_HOSTFS_I(file) HOSTFS_I(file_inode(file)) static int hostfs_d_delete(const struct dentry *dentry) { diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c index 78e12b2e0ea2..546f6d39713a 100644 --- a/fs/hpfs/dir.c +++ b/fs/hpfs/dir.c @@ -25,7 +25,7 @@ static loff_t hpfs_dir_lseek(struct file *filp, loff_t off, int whence) loff_t new_off = off + (whence == 1 ? filp->f_pos : 0); loff_t pos; struct quad_buffer_head qbh; - struct inode *i = filp->f_path.dentry->d_inode; + struct inode *i = file_inode(filp); struct hpfs_inode_info *hpfs_inode = hpfs_i(i); struct super_block *s = i->i_sb; @@ -57,7 +57,7 @@ fail: static int hpfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct hpfs_inode_info *hpfs_inode = hpfs_i(inode); struct quad_buffer_head qbh; struct hpfs_dirent *de; diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index fbfe2df5624b..9f9dbeceeee7 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -152,7 +152,7 @@ static ssize_t hpfs_file_write(struct file *file, const char __user *buf, retval = do_sync_write(file, buf, count, ppos); if (retval > 0) { hpfs_lock(file->f_path.dentry->d_sb); - hpfs_i(file->f_path.dentry->d_inode)->i_dirty = 1; + hpfs_i(file_inode(file))->i_dirty = 1; hpfs_unlock(file->f_path.dentry->d_sb); } return retval; diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c index 43b315f2002b..74f55703be49 100644 --- a/fs/hppfs/hppfs.c +++ b/fs/hppfs/hppfs.c @@ -180,7 +180,7 @@ static ssize_t read_proc(struct file *file, char __user *buf, ssize_t count, ssize_t (*read)(struct file *, char __user *, size_t, loff_t *); ssize_t n; - read = file->f_path.dentry->d_inode->i_fop->read; + read = file_inode(file)->i_fop->read; if (!is_user) set_fs(KERNEL_DS); @@ -288,7 +288,7 @@ static ssize_t hppfs_write(struct file *file, const char __user *buf, struct file *proc_file = data->proc_file; ssize_t (*write)(struct file *, const char __user *, size_t, loff_t *); - write = proc_file->f_path.dentry->d_inode->i_fop->write; + write = file_inode(proc_file)->i_fop->write; return (*write)(proc_file, buf, len, ppos); } @@ -513,7 +513,7 @@ static loff_t hppfs_llseek(struct file *file, loff_t off, int where) loff_t (*llseek)(struct file *, loff_t, int); loff_t ret; - llseek = proc_file->f_path.dentry->d_inode->i_fop->llseek; + llseek = file_inode(proc_file)->i_fop->llseek; if (llseek != NULL) { ret = (*llseek)(proc_file, off, where); if (ret < 0) @@ -561,7 +561,7 @@ static int hppfs_readdir(struct file *file, void *ent, filldir_t filldir) }); int err; - readdir = proc_file->f_path.dentry->d_inode->i_fop->readdir; + readdir = file_inode(proc_file)->i_fop->readdir; proc_file->f_pos = file->f_pos; err = (*readdir)(proc_file, &dirent, hppfs_filldir); diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 78bde32ea951..edb42ea60c76 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -97,7 +97,7 @@ static void huge_pagevec_release(struct pagevec *pvec) static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); loff_t len, vma_len; int ret; struct hstate *h = hstate_file(file); diff --git a/fs/inode.c b/fs/inode.c index 14084b72b259..67880e604399 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -1655,7 +1655,7 @@ EXPORT_SYMBOL(file_remove_suid); int file_update_time(struct file *file) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct timespec now; int sync_it = 0; int ret; diff --git a/fs/ioctl.c b/fs/ioctl.c index 3bdad6d1f268..fd507fb460f8 100644 --- a/fs/ioctl.c +++ b/fs/ioctl.c @@ -175,7 +175,7 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg) struct fiemap fiemap; struct fiemap __user *ufiemap = (struct fiemap __user *) arg; struct fiemap_extent_info fieinfo = { 0, }; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; u64 len; int error; @@ -424,7 +424,7 @@ EXPORT_SYMBOL(generic_block_fiemap); */ int ioctl_preallocate(struct file *filp, void __user *argp) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct space_resv sr; if (copy_from_user(&sr, argp, sizeof(sr))) @@ -449,7 +449,7 @@ int ioctl_preallocate(struct file *filp, void __user *argp) static int file_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int __user *p = (int __user *)arg; switch (cmd) { @@ -512,7 +512,7 @@ static int ioctl_fioasync(unsigned int fd, struct file *filp, static int ioctl_fsfreeze(struct file *filp) { - struct super_block *sb = filp->f_path.dentry->d_inode->i_sb; + struct super_block *sb = file_inode(filp)->i_sb; if (!capable(CAP_SYS_ADMIN)) return -EPERM; @@ -527,7 +527,7 @@ static int ioctl_fsfreeze(struct file *filp) static int ioctl_fsthaw(struct file *filp) { - struct super_block *sb = filp->f_path.dentry->d_inode->i_sb; + struct super_block *sb = file_inode(filp)->i_sb; if (!capable(CAP_SYS_ADMIN)) return -EPERM; @@ -548,7 +548,7 @@ int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd, { int error = 0; int __user *argp = (int __user *)arg; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); switch (cmd) { case FIOCLEX: diff --git a/fs/isofs/compress.c b/fs/isofs/compress.c index 0b3fa7974fa8..592e5115a561 100644 --- a/fs/isofs/compress.c +++ b/fs/isofs/compress.c @@ -296,7 +296,7 @@ static int zisofs_fill_pages(struct inode *inode, int full_page, int pcount, */ static int zisofs_readpage(struct file *file, struct page *page) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct address_space *mapping = inode->i_mapping; int err; int i, pcount, full_page; diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c index f20437c068a0..a7d5c3c3d4e6 100644 --- a/fs/isofs/dir.c +++ b/fs/isofs/dir.c @@ -253,7 +253,7 @@ static int isofs_readdir(struct file *filp, int result; char *tmpname; struct iso_directory_record *tmpde; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); tmpname = (char *)__get_free_page(GFP_KERNEL); if (tmpname == NULL) diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index ad7774d32095..acd46a4160cb 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -117,12 +117,12 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target, static int jffs2_readdir(struct file *filp, void *dirent, filldir_t filldir) { struct jffs2_inode_info *f; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct jffs2_full_dirent *fd; unsigned long offset, curofs; jffs2_dbg(1, "jffs2_readdir() for dir_i #%lu\n", - filp->f_path.dentry->d_inode->i_ino); + file_inode(filp)->i_ino); f = JFFS2_INODE_INFO(inode); diff --git a/fs/jfs/ioctl.c b/fs/jfs/ioctl.c index bc555ff417e9..93a1232894f6 100644 --- a/fs/jfs/ioctl.c +++ b/fs/jfs/ioctl.c @@ -58,7 +58,7 @@ static long jfs_map_ext2(unsigned long flags, int from) long jfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct jfs_inode_info *jfs_inode = JFS_IP(inode); unsigned int flags; diff --git a/fs/jfs/jfs_dtree.c b/fs/jfs/jfs_dtree.c index 9197a1b0d02d..0ddbeceafc62 100644 --- a/fs/jfs/jfs_dtree.c +++ b/fs/jfs/jfs_dtree.c @@ -3004,7 +3004,7 @@ static inline struct jfs_dirent *next_jfs_dirent(struct jfs_dirent *dirent) */ int jfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *ip = filp->f_path.dentry->d_inode; + struct inode *ip = file_inode(filp); struct nls_table *codepage = JFS_SBI(ip->i_sb)->nls_tab; int rc = 0; loff_t dtpos; /* legacy OS/2 style position */ diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c index ca0a08001449..a2717408c478 100644 --- a/fs/lockd/clntlock.c +++ b/fs/lockd/clntlock.c @@ -178,7 +178,7 @@ __be32 nlmclnt_grant(const struct sockaddr *addr, const struct nlm_lock *lock) continue; if (!rpc_cmp_addr(nlm_addr(block->b_host), addr)) continue; - if (nfs_compare_fh(NFS_FH(fl_blocked->fl_file->f_path.dentry->d_inode) ,fh) != 0) + if (nfs_compare_fh(NFS_FH(file_inode(fl_blocked->fl_file)) ,fh) != 0) continue; /* Alright, we found a lock. Set the return status * and wake up the caller diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 54f9e6ce0430..d7e1ec1c6827 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -127,7 +127,7 @@ static void nlmclnt_setlockargs(struct nlm_rqst *req, struct file_lock *fl) struct nlm_lock *lock = &argp->lock; nlmclnt_next_cookie(&argp->cookie); - memcpy(&lock->fh, NFS_FH(fl->fl_file->f_path.dentry->d_inode), sizeof(struct nfs_fh)); + memcpy(&lock->fh, NFS_FH(file_inode(fl->fl_file)), sizeof(struct nfs_fh)); lock->caller = utsname()->nodename; lock->oh.data = req->a_owner; lock->oh.len = snprintf(req->a_owner, sizeof(req->a_owner), "%u@%s", diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index 8d80c990dffd..e703318c41df 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -406,8 +406,8 @@ nlmsvc_lock(struct svc_rqst *rqstp, struct nlm_file *file, __be32 ret; dprintk("lockd: nlmsvc_lock(%s/%ld, ty=%d, pi=%d, %Ld-%Ld, bl=%d)\n", - file->f_file->f_path.dentry->d_inode->i_sb->s_id, - file->f_file->f_path.dentry->d_inode->i_ino, + file_inode(file->f_file)->i_sb->s_id, + file_inode(file->f_file)->i_ino, lock->fl.fl_type, lock->fl.fl_pid, (long long)lock->fl.fl_start, (long long)lock->fl.fl_end, @@ -513,8 +513,8 @@ nlmsvc_testlock(struct svc_rqst *rqstp, struct nlm_file *file, __be32 ret; dprintk("lockd: nlmsvc_testlock(%s/%ld, ty=%d, %Ld-%Ld)\n", - file->f_file->f_path.dentry->d_inode->i_sb->s_id, - file->f_file->f_path.dentry->d_inode->i_ino, + file_inode(file->f_file)->i_sb->s_id, + file_inode(file->f_file)->i_ino, lock->fl.fl_type, (long long)lock->fl.fl_start, (long long)lock->fl.fl_end); @@ -606,8 +606,8 @@ nlmsvc_unlock(struct net *net, struct nlm_file *file, struct nlm_lock *lock) int error; dprintk("lockd: nlmsvc_unlock(%s/%ld, pi=%d, %Ld-%Ld)\n", - file->f_file->f_path.dentry->d_inode->i_sb->s_id, - file->f_file->f_path.dentry->d_inode->i_ino, + file_inode(file->f_file)->i_sb->s_id, + file_inode(file->f_file)->i_ino, lock->fl.fl_pid, (long long)lock->fl.fl_start, (long long)lock->fl.fl_end); @@ -635,8 +635,8 @@ nlmsvc_cancel_blocked(struct net *net, struct nlm_file *file, struct nlm_lock *l int status = 0; dprintk("lockd: nlmsvc_cancel(%s/%ld, pi=%d, %Ld-%Ld)\n", - file->f_file->f_path.dentry->d_inode->i_sb->s_id, - file->f_file->f_path.dentry->d_inode->i_ino, + file_inode(file->f_file)->i_sb->s_id, + file_inode(file->f_file)->i_ino, lock->fl.fl_pid, (long long)lock->fl.fl_start, (long long)lock->fl.fl_end); diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index 0deb5f6c9dd4..b3a24b07d981 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -45,7 +45,7 @@ static inline void nlm_debug_print_fh(char *msg, struct nfs_fh *f) static inline void nlm_debug_print_file(char *msg, struct nlm_file *file) { - struct inode *inode = file->f_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file->f_file); dprintk("lockd: %s %s/%ld\n", msg, inode->i_sb->s_id, inode->i_ino); diff --git a/fs/locks.c b/fs/locks.c index a94e331a52a2..cb424a4fed71 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -334,7 +334,7 @@ static int flock_to_posix_lock(struct file *filp, struct file_lock *fl, start = filp->f_pos; break; case SEEK_END: - start = i_size_read(filp->f_path.dentry->d_inode); + start = i_size_read(file_inode(filp)); break; default: return -EINVAL; @@ -384,7 +384,7 @@ static int flock64_to_posix_lock(struct file *filp, struct file_lock *fl, start = filp->f_pos; break; case SEEK_END: - start = i_size_read(filp->f_path.dentry->d_inode); + start = i_size_read(file_inode(filp)); break; default: return -EINVAL; @@ -627,7 +627,7 @@ posix_test_lock(struct file *filp, struct file_lock *fl) struct file_lock *cfl; lock_flocks(); - for (cfl = filp->f_path.dentry->d_inode->i_flock; cfl; cfl = cfl->fl_next) { + for (cfl = file_inode(filp)->i_flock; cfl; cfl = cfl->fl_next) { if (!IS_POSIX(cfl)) continue; if (posix_locks_conflict(fl, cfl)) @@ -708,7 +708,7 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) { struct file_lock *new_fl = NULL; struct file_lock **before; - struct inode * inode = filp->f_path.dentry->d_inode; + struct inode * inode = file_inode(filp); int error = 0; int found = 0; @@ -1002,7 +1002,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str int posix_lock_file(struct file *filp, struct file_lock *fl, struct file_lock *conflock) { - return __posix_lock_file(filp->f_path.dentry->d_inode, fl, conflock); + return __posix_lock_file(file_inode(filp), fl, conflock); } EXPORT_SYMBOL(posix_lock_file); @@ -1326,8 +1326,8 @@ int fcntl_getlease(struct file *filp) int type = F_UNLCK; lock_flocks(); - time_out_leases(filp->f_path.dentry->d_inode); - for (fl = filp->f_path.dentry->d_inode->i_flock; fl && IS_LEASE(fl); + time_out_leases(file_inode(filp)); + for (fl = file_inode(filp)->i_flock; fl && IS_LEASE(fl); fl = fl->fl_next) { if (fl->fl_file == filp) { type = target_leasetype(fl); @@ -1843,7 +1843,7 @@ int fcntl_setlk(unsigned int fd, struct file *filp, unsigned int cmd, if (copy_from_user(&flock, l, sizeof(flock))) goto out; - inode = filp->f_path.dentry->d_inode; + inode = file_inode(filp); /* Don't allow mandatory locks on files that may be memory mapped * and shared. @@ -1961,7 +1961,7 @@ int fcntl_setlk64(unsigned int fd, struct file *filp, unsigned int cmd, if (copy_from_user(&flock, l, sizeof(flock))) goto out; - inode = filp->f_path.dentry->d_inode; + inode = file_inode(filp); /* Don't allow mandatory locks on files that may be memory mapped * and shared. @@ -2030,7 +2030,7 @@ void locks_remove_posix(struct file *filp, fl_owner_t owner) * posix_lock_file(). Another process could be setting a lock on this * file at the same time, but we wouldn't remove that lock anyway. */ - if (!filp->f_path.dentry->d_inode->i_flock) + if (!file_inode(filp)->i_flock) return; lock.fl_type = F_UNLCK; @@ -2056,7 +2056,7 @@ EXPORT_SYMBOL(locks_remove_posix); */ void locks_remove_flock(struct file *filp) { - struct inode * inode = filp->f_path.dentry->d_inode; + struct inode * inode = file_inode(filp); struct file_lock *fl; struct file_lock **before; @@ -2152,7 +2152,7 @@ static void lock_get_status(struct seq_file *f, struct file_lock *fl, fl_pid = fl->fl_pid; if (fl->fl_file != NULL) - inode = fl->fl_file->f_path.dentry->d_inode; + inode = file_inode(fl->fl_file); seq_printf(f, "%lld:%s ", id, pfx); if (IS_POSIX(fl)) { diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c index 26e4a941532f..b82751082112 100644 --- a/fs/logfs/dir.c +++ b/fs/logfs/dir.c @@ -284,7 +284,7 @@ static int logfs_rmdir(struct inode *dir, struct dentry *dentry) #define IMPLICIT_NODES 2 static int __logfs_readdir(struct file *file, void *buf, filldir_t filldir) { - struct inode *dir = file->f_dentry->d_inode; + struct inode *dir = file_inode(file); loff_t pos = file->f_pos - IMPLICIT_NODES; struct page *page; struct logfs_disk_dentry *dd; @@ -320,7 +320,7 @@ static int __logfs_readdir(struct file *file, void *buf, filldir_t filldir) static int logfs_readdir(struct file *file, void *buf, filldir_t filldir) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); ino_t pino = parent_ino(file->f_dentry); int err; diff --git a/fs/logfs/file.c b/fs/logfs/file.c index 3886cded283c..c2219a6dd3c8 100644 --- a/fs/logfs/file.c +++ b/fs/logfs/file.c @@ -183,7 +183,7 @@ static int logfs_releasepage(struct page *page, gfp_t only_xfs_uses_this) long logfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct logfs_inode *li = logfs_inode(inode); unsigned int oldflags, flags; int err; diff --git a/fs/minix/dir.c b/fs/minix/dir.c index 685b2d981b87..a9ed6f36e6ea 100644 --- a/fs/minix/dir.c +++ b/fs/minix/dir.c @@ -85,7 +85,7 @@ static inline void *minix_next_entry(void *de, struct minix_sb_info *sbi) static int minix_readdir(struct file * filp, void * dirent, filldir_t filldir) { unsigned long pos = filp->f_pos; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; unsigned offset = pos & ~PAGE_CACHE_MASK; unsigned long n = pos >> PAGE_CACHE_SHIFT; diff --git a/fs/namei.c b/fs/namei.c index 43a97ee1d4c8..df00b754631d 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2778,7 +2778,7 @@ retry_lookup: goto out; if ((*opened & FILE_CREATED) || - !S_ISREG(file->f_path.dentry->d_inode->i_mode)) + !S_ISREG(file_inode(file)->i_mode)) will_truncate = false; audit_inode(name, file->f_path.dentry, 0); diff --git a/fs/namespace.c b/fs/namespace.c index 269919fa116d..50ca17d3cb45 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -384,7 +384,7 @@ EXPORT_SYMBOL_GPL(mnt_clone_write); */ int __mnt_want_write_file(struct file *file) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); if (!(file->f_mode & FMODE_WRITE) || special_file(inode->i_mode)) return __mnt_want_write(file->f_path.mnt); diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c index 1acdad7fcec7..c41e02932542 100644 --- a/fs/ncpfs/inode.c +++ b/fs/ncpfs/inode.c @@ -525,7 +525,7 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) if (!ncp_filp) goto out; error = -ENOTSOCK; - sock_inode = ncp_filp->f_path.dentry->d_inode; + sock_inode = file_inode(ncp_filp); if (!S_ISSOCK(sock_inode->i_mode)) goto out_fput; sock = SOCKET_I(sock_inode); @@ -564,7 +564,7 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) if (!server->info_filp) goto out_bdi; error = -ENOTSOCK; - sock_inode = server->info_filp->f_path.dentry->d_inode; + sock_inode = file_inode(server->info_filp); if (!S_ISSOCK(sock_inode->i_mode)) goto out_fput2; info_sock = SOCKET_I(sock_inode); diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c index 6958adfaff08..5c1e9262219c 100644 --- a/fs/ncpfs/ioctl.c +++ b/fs/ncpfs/ioctl.c @@ -808,7 +808,7 @@ outrel: long ncp_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct ncp_server *server = NCP_SERVER(inode); uid_t uid = current_uid(); int need_drop_write = 0; diff --git a/fs/ncpfs/mmap.c b/fs/ncpfs/mmap.c index 63d14a99483d..ee24df5af1f9 100644 --- a/fs/ncpfs/mmap.c +++ b/fs/ncpfs/mmap.c @@ -105,7 +105,7 @@ static const struct vm_operations_struct ncp_file_mmap = /* This is used for a general mmap of a ncp file */ int ncp_mmap(struct file *file, struct vm_area_struct *vma) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); DPRINTK("ncp_mmap: called\n"); diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 1b2d7eb93796..a8bd28cde7e2 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -281,7 +281,7 @@ int nfs_readdir_search_for_cookie(struct nfs_cache_array *array, nfs_readdir_des for (i = 0; i < array->size; i++) { if (array->array[i].cookie == *desc->dir_cookie) { - struct nfs_inode *nfsi = NFS_I(desc->file->f_path.dentry->d_inode); + struct nfs_inode *nfsi = NFS_I(file_inode(desc->file)); struct nfs_open_dir_context *ctx = desc->file->private_data; new_pos = desc->current_index + i; @@ -629,7 +629,7 @@ out: static int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page* page) { - struct inode *inode = desc->file->f_path.dentry->d_inode; + struct inode *inode = file_inode(desc->file); int ret; ret = nfs_readdir_xdr_to_array(desc, page, inode); @@ -660,7 +660,7 @@ void cache_page_release(nfs_readdir_descriptor_t *desc) static struct page *get_cache_page(nfs_readdir_descriptor_t *desc) { - return read_cache_page(desc->file->f_path.dentry->d_inode->i_mapping, + return read_cache_page(file_inode(desc->file)->i_mapping, desc->page_index, (filler_t *)nfs_readdir_filler, desc); } @@ -764,7 +764,7 @@ int uncached_readdir(nfs_readdir_descriptor_t *desc, void *dirent, { struct page *page = NULL; int status; - struct inode *inode = desc->file->f_path.dentry->d_inode; + struct inode *inode = file_inode(desc->file); struct nfs_open_dir_context *ctx = desc->file->private_data; dfprintk(DIRCACHE, "NFS: uncached_readdir() searching for cookie %Lu\n", diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 3c2b893665ba..29f4a48a0ee6 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -292,7 +292,7 @@ static int nfs_file_fsync(struct file *file, loff_t start, loff_t end, int datasync) { int ret; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); do { ret = filemap_write_and_wait_range(inode->i_mapping, start, end); diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index bc3968fa81e5..2ad8deaf7dbf 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -764,7 +764,7 @@ out: static ssize_t idmap_pipe_downcall(struct file *filp, const char __user *src, size_t mlen) { - struct rpc_inode *rpci = RPC_I(filp->f_path.dentry->d_inode); + struct rpc_inode *rpci = RPC_I(file_inode(filp)); struct idmap *idmap = (struct idmap *)rpci->private; struct key_construction *cons; struct idmap_msg im; diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index ebeb94ce1b0b..548ae3113005 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -714,7 +714,7 @@ EXPORT_SYMBOL_GPL(put_nfs_open_context); */ void nfs_file_set_open_context(struct file *filp, struct nfs_open_context *ctx) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct nfs_inode *nfsi = NFS_I(inode); filp->private_data = get_nfs_open_context(ctx); @@ -747,7 +747,7 @@ struct nfs_open_context *nfs_find_open_context(struct inode *inode, struct rpc_c static void nfs_file_clear_open_context(struct file *filp) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct nfs_open_context *ctx = nfs_file_open_context(filp); if (ctx) { diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index 70efb63b1e42..43ea96ced28c 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -872,7 +872,7 @@ static void nfs3_proc_commit_setup(struct nfs_commit_data *data, struct rpc_mess static int nfs3_proc_lock(struct file *filp, int cmd, struct file_lock *fl) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); return nlmclnt_proc(NFS_SERVER(inode)->nlm_host, cmd, fl); } diff --git a/fs/nfs/nfs4file.c b/fs/nfs/nfs4file.c index 08ddcccb8887..13e6bb3e3fe5 100644 --- a/fs/nfs/nfs4file.c +++ b/fs/nfs/nfs4file.c @@ -94,7 +94,7 @@ static int nfs4_file_fsync(struct file *file, loff_t start, loff_t end, int datasync) { int ret; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); do { ret = filemap_write_and_wait_range(inode->i_mapping, start, end); diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index f084dac948e1..fc8de9016acf 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -662,7 +662,7 @@ nfs_proc_commit_setup(struct nfs_commit_data *data, struct rpc_message *msg) static int nfs_proc_lock(struct file *filp, int cmd, struct file_lock *fl) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); return nlmclnt_proc(NFS_SERVER(inode)->nlm_host, cmd, fl); } diff --git a/fs/nfsd/fault_inject.c b/fs/nfsd/fault_inject.c index e761ee95617f..497584c70366 100644 --- a/fs/nfsd/fault_inject.c +++ b/fs/nfsd/fault_inject.c @@ -101,7 +101,7 @@ static ssize_t fault_inject_read(struct file *file, char __user *buf, loff_t pos = *ppos; if (!pos) - nfsd_inject_get(file->f_dentry->d_inode->i_private, &val); + nfsd_inject_get(file_inode(file)->i_private, &val); size = scnprintf(read_buf, sizeof(read_buf), "%llu\n", val); if (pos < 0) @@ -133,10 +133,10 @@ static ssize_t fault_inject_write(struct file *file, const char __user *buf, size = rpc_pton(net, write_buf, size, (struct sockaddr *)&sa, sizeof(sa)); if (size > 0) - nfsd_inject_set_client(file->f_dentry->d_inode->i_private, &sa, size); + nfsd_inject_set_client(file_inode(file)->i_private, &sa, size); else { val = simple_strtoll(write_buf, NULL, 0); - nfsd_inject_set(file->f_dentry->d_inode->i_private, val); + nfsd_inject_set(file_inode(file)->i_private, val); } return len; /* on success, claim we got the whole input */ } diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 74934284d9a7..2db7021b01ae 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -85,7 +85,7 @@ static ssize_t (*write_op[])(struct file *, char *, size_t) = { static ssize_t nfsctl_transaction_write(struct file *file, const char __user *buf, size_t size, loff_t *pos) { - ino_t ino = file->f_path.dentry->d_inode->i_ino; + ino_t ino = file_inode(file)->i_ino; char *data; ssize_t rv; diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index d586117fa94a..a94245b4045f 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -979,7 +979,7 @@ static void kill_suid(struct dentry *dentry) */ static int wait_for_concurrent_writes(struct file *file) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); static ino_t last_ino; static dev_t last_dev; int err = 0; @@ -1070,7 +1070,7 @@ __be32 nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, if (err) return err; - inode = file->f_path.dentry->d_inode; + inode = file_inode(file); /* Get readahead parameters */ ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino); @@ -1957,7 +1957,7 @@ static __be32 nfsd_buffered_readdir(struct file *file, filldir_t func, offset = *offsetp; while (1) { - struct inode *dir_inode = file->f_path.dentry->d_inode; + struct inode *dir_inode = file_inode(file); unsigned int reclen; cdp->err = nfserr_eof; /* will be cleared on successful read */ diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index df1a7fb238d1..f30b017740a7 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -259,7 +259,7 @@ static void nilfs_set_de_type(struct nilfs_dir_entry *de, struct inode *inode) static int nilfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { loff_t pos = filp->f_pos; - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; unsigned int offset = pos & ~PAGE_CACHE_MASK; unsigned long n = pos >> PAGE_CACHE_SHIFT; diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c index 61946883025c..89dc0886387d 100644 --- a/fs/nilfs2/file.c +++ b/fs/nilfs2/file.c @@ -67,7 +67,7 @@ int nilfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct nilfs_transaction_info ti; int ret = 0; diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index fdb180769485..ef61c749641d 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -793,7 +793,7 @@ static int nilfs_ioctl_get_info(struct inode *inode, struct file *filp, long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); void __user *argp = (void __user *)arg; switch (cmd) { diff --git a/fs/notify/dnotify/dnotify.c b/fs/notify/dnotify/dnotify.c index 08b886f119ce..2bfe6dc413a0 100644 --- a/fs/notify/dnotify/dnotify.c +++ b/fs/notify/dnotify/dnotify.c @@ -174,7 +174,7 @@ void dnotify_flush(struct file *filp, fl_owner_t id) struct dnotify_struct **prev; struct inode *inode; - inode = filp->f_path.dentry->d_inode; + inode = file_inode(filp); if (!S_ISDIR(inode->i_mode)) return; @@ -296,7 +296,7 @@ int fcntl_dirnotify(int fd, struct file *filp, unsigned long arg) } /* dnotify only works on directories */ - inode = filp->f_path.dentry->d_inode; + inode = file_inode(filp); if (!S_ISDIR(inode->i_mode)) { error = -ENOTDIR; goto out_err; diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c index 9ff4a5ee6e20..5d8444268a16 100644 --- a/fs/notify/fanotify/fanotify_user.c +++ b/fs/notify/fanotify/fanotify_user.c @@ -466,7 +466,7 @@ static int fanotify_find_path(int dfd, const char __user *filename, ret = -ENOTDIR; if ((flags & FAN_MARK_ONLYDIR) && - !(S_ISDIR(f.file->f_path.dentry->d_inode->i_mode))) { + !(S_ISDIR(file_inode(f.file)->i_mode))) { fdput(f); goto out; } diff --git a/fs/ntfs/dir.c b/fs/ntfs/dir.c index 99e36107ff60..aa411c3f20e9 100644 --- a/fs/ntfs/dir.c +++ b/fs/ntfs/dir.c @@ -1101,7 +1101,7 @@ static int ntfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { s64 ia_pos, ia_start, prev_ia_pos, bmp_pos; loff_t fpos, i_size; - struct inode *bmp_vi, *vdir = filp->f_path.dentry->d_inode; + struct inode *bmp_vi, *vdir = file_inode(filp); struct super_block *sb = vdir->i_sb; ntfs_inode *ndir = NTFS_I(vdir); ntfs_volume *vol = NTFS_SB(sb); diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 657743254eb9..db1ad26e02a7 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -569,7 +569,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb, int ret, bool is_async) { - struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(iocb->ki_filp); int level; wait_queue_head_t *wq = ocfs2_ioend_wq(inode); @@ -626,7 +626,7 @@ static ssize_t ocfs2_direct_IO(int rw, unsigned long nr_segs) { struct file *file = iocb->ki_filp; - struct inode *inode = file->f_path.dentry->d_inode->i_mapping->host; + struct inode *inode = file_inode(file)->i_mapping->host; /* * Fallback to buffered I/O if we see an inode without diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index 8fe4e2892ab9..ac0d4a0e8a41 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -2015,7 +2015,7 @@ int ocfs2_dir_foreach(struct inode *inode, loff_t *f_pos, void *priv, int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir) { int error = 0; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int lock_level = 0; trace_ocfs2_readdir((unsigned long long)OCFS2_I(inode)->ip_blkno); diff --git a/fs/ocfs2/dlmfs/dlmfs.c b/fs/ocfs2/dlmfs/dlmfs.c index 16b712d260d4..4c5fc8d77dc2 100644 --- a/fs/ocfs2/dlmfs/dlmfs.c +++ b/fs/ocfs2/dlmfs/dlmfs.c @@ -224,7 +224,7 @@ static int dlmfs_file_setattr(struct dentry *dentry, struct iattr *attr) static unsigned int dlmfs_file_poll(struct file *file, poll_table *wait) { int event = 0; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct dlmfs_inode_private *ip = DLMFS_I(inode); poll_wait(file, &ip->ip_lockres.l_event, wait); @@ -245,7 +245,7 @@ static ssize_t dlmfs_file_read(struct file *filp, int bytes_left; ssize_t readlen, got; char *lvb_buf; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); mlog(0, "inode %lu, count = %zu, *ppos = %llu\n", inode->i_ino, count, *ppos); @@ -293,7 +293,7 @@ static ssize_t dlmfs_file_write(struct file *filp, int bytes_left; ssize_t writelen; char *lvb_buf; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); mlog(0, "inode %lu, count = %zu, *ppos = %llu\n", inode->i_ino, count, *ppos); diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 37d313ede159..04098af9dbc8 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1949,7 +1949,7 @@ out: int ocfs2_change_file_space(struct file *file, unsigned int cmd, struct ocfs2_space_resv *sr) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); int ret; @@ -1977,7 +1977,7 @@ int ocfs2_change_file_space(struct file *file, unsigned int cmd, static long ocfs2_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); struct ocfs2_space_resv sr; int change_size = 1; @@ -2232,7 +2232,7 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb, loff_t old_size, *ppos = &iocb->ki_pos; u32 old_clusters; struct file *file = iocb->ki_filp; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); int full_coherency = !(osb->s_mount_opt & OCFS2_MOUNT_COHERENCY_BUFFERED); @@ -2516,7 +2516,7 @@ static ssize_t ocfs2_file_splice_read(struct file *in, unsigned int flags) { int ret = 0, lock_level = 0; - struct inode *inode = in->f_path.dentry->d_inode; + struct inode *inode = file_inode(in); trace_ocfs2_file_splice_read(inode, in, in->f_path.dentry, (unsigned long long)OCFS2_I(inode)->ip_blkno, @@ -2546,7 +2546,7 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb, { int ret = 0, rw_level = -1, have_alloc_sem = 0, lock_level = 0; struct file *filp = iocb->ki_filp; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); trace_ocfs2_file_aio_read(inode, filp, filp->f_path.dentry, (unsigned long long)OCFS2_I(inode)->ip_blkno, diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c index f20edcbfe700..752f0b26221d 100644 --- a/fs/ocfs2/ioctl.c +++ b/fs/ocfs2/ioctl.c @@ -881,7 +881,7 @@ bail: long ocfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); unsigned int flags; int new_clusters; int status; @@ -994,7 +994,7 @@ long ocfs2_compat_ioctl(struct file *file, unsigned cmd, unsigned long arg) { bool preserve; struct reflink_arguments args; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct ocfs2_info info; void __user *argp = (void __user *)arg; diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c index 47a87dda54ce..07c585b85000 100644 --- a/fs/ocfs2/mmap.c +++ b/fs/ocfs2/mmap.c @@ -62,7 +62,7 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh, struct page *page) { int ret = VM_FAULT_NOPAGE; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct address_space *mapping = inode->i_mapping; loff_t pos = page_offset(page); unsigned int len = PAGE_CACHE_SIZE; @@ -131,7 +131,7 @@ out: static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct buffer_head *di_bh = NULL; sigset_t oldset; int ret; @@ -180,13 +180,13 @@ int ocfs2_mmap(struct file *file, struct vm_area_struct *vma) { int ret = 0, lock_level = 0; - ret = ocfs2_inode_lock_atime(file->f_dentry->d_inode, + ret = ocfs2_inode_lock_atime(file_inode(file), file->f_vfsmnt, &lock_level); if (ret < 0) { mlog_errno(ret); goto out; } - ocfs2_inode_unlock(file->f_dentry->d_inode, lock_level); + ocfs2_inode_unlock(file_inode(file), lock_level); out: vma->vm_ops = &ocfs2_file_vm_ops; return 0; diff --git a/fs/ocfs2/move_extents.c b/fs/ocfs2/move_extents.c index 6083432f667e..9f8dcadd9a50 100644 --- a/fs/ocfs2/move_extents.c +++ b/fs/ocfs2/move_extents.c @@ -1055,7 +1055,7 @@ int ocfs2_ioctl_move_extents(struct file *filp, void __user *argp) { int status; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct ocfs2_move_extents range; struct ocfs2_move_extents_context *context = NULL; diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index 30a055049e16..1baffaadda41 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -2927,7 +2927,7 @@ int ocfs2_duplicate_clusters_by_page(handle_t *handle, u32 new_cluster, u32 new_len) { int ret = 0, partial; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct ocfs2_caching_info *ci = INODE_CACHE(inode); struct super_block *sb = ocfs2_metadata_cache_get_super(ci); u64 new_block = ocfs2_clusters_to_blocks(sb, new_cluster); @@ -3020,7 +3020,7 @@ int ocfs2_duplicate_clusters_by_jbd(handle_t *handle, u32 new_cluster, u32 new_len) { int ret = 0; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct super_block *sb = inode->i_sb; struct ocfs2_caching_info *ci = INODE_CACHE(inode); int i, blocks = ocfs2_clusters_to_blocks(sb, new_len); diff --git a/fs/omfs/dir.c b/fs/omfs/dir.c index fb5b3ff79dc6..acbaebcad3a8 100644 --- a/fs/omfs/dir.c +++ b/fs/omfs/dir.c @@ -330,7 +330,7 @@ int omfs_is_bad(struct omfs_sb_info *sbi, struct omfs_header *header, static int omfs_fill_chain(struct file *filp, void *dirent, filldir_t filldir, u64 fsblock, int hindex) { - struct inode *dir = filp->f_dentry->d_inode; + struct inode *dir = file_inode(filp); struct buffer_head *bh; struct omfs_inode *oi; u64 self; @@ -405,7 +405,7 @@ out: static int omfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *dir = filp->f_dentry->d_inode; + struct inode *dir = file_inode(filp); struct buffer_head *bh; loff_t offset, res; unsigned int hchain, hindex; diff --git a/fs/open.c b/fs/open.c index 9b33c0cbfacf..e08643feb574 100644 --- a/fs/open.c +++ b/fs/open.c @@ -228,7 +228,7 @@ SYSCALL_ALIAS(sys_ftruncate64, SyS_ftruncate64); int do_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); long ret; if (offset < 0 || len <= 0) @@ -426,7 +426,7 @@ SYSCALL_DEFINE1(fchdir, unsigned int, fd) if (!f.file) goto out; - inode = f.file->f_path.dentry->d_inode; + inode = file_inode(f.file); error = -ENOTDIR; if (!S_ISDIR(inode->i_mode)) @@ -689,7 +689,7 @@ static int do_dentry_open(struct file *f, f->f_mode = FMODE_PATH; path_get(&f->f_path); - inode = f->f_path.dentry->d_inode; + inode = file_inode(f); if (f->f_mode & FMODE_WRITE) { error = __get_file_write_access(inode, f->f_path.mnt); if (error) diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c index 2ad080faca34..ae47fa7efb9d 100644 --- a/fs/openpromfs/inode.c +++ b/fs/openpromfs/inode.c @@ -262,7 +262,7 @@ found: static int openpromfs_readdir(struct file * filp, void * dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct op_inode_info *oi = OP_I(inode); struct device_node *dp = oi->u.node; struct device_node *child; diff --git a/fs/pipe.c b/fs/pipe.c index bd3479db4b62..39baf6c3ebb0 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -361,7 +361,7 @@ pipe_read(struct kiocb *iocb, const struct iovec *_iov, unsigned long nr_segs, loff_t pos) { struct file *filp = iocb->ki_filp; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct pipe_inode_info *pipe; int do_wakeup; ssize_t ret; @@ -486,7 +486,7 @@ pipe_write(struct kiocb *iocb, const struct iovec *_iov, unsigned long nr_segs, loff_t ppos) { struct file *filp = iocb->ki_filp; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct pipe_inode_info *pipe; ssize_t ret; int do_wakeup; @@ -677,7 +677,7 @@ bad_pipe_w(struct file *filp, const char __user *buf, size_t count, static long pipe_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct pipe_inode_info *pipe; int count, buf, nrbufs; @@ -705,7 +705,7 @@ static unsigned int pipe_poll(struct file *filp, poll_table *wait) { unsigned int mask; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct pipe_inode_info *pipe = inode->i_pipe; int nrbufs; @@ -758,7 +758,7 @@ pipe_release(struct inode *inode, int decr, int decw) static int pipe_read_fasync(int fd, struct file *filp, int on) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int retval; mutex_lock(&inode->i_mutex); @@ -772,7 +772,7 @@ pipe_read_fasync(int fd, struct file *filp, int on) static int pipe_write_fasync(int fd, struct file *filp, int on) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int retval; mutex_lock(&inode->i_mutex); @@ -786,7 +786,7 @@ pipe_write_fasync(int fd, struct file *filp, int on) static int pipe_rdwr_fasync(int fd, struct file *filp, int on) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct pipe_inode_info *pipe = inode->i_pipe; int retval; @@ -1226,7 +1226,7 @@ int pipe_proc_fn(struct ctl_table *table, int write, void __user *buf, */ struct pipe_inode_info *get_pipe_info(struct file *file) { - struct inode *i = file->f_path.dentry->d_inode; + struct inode *i = file_inode(file); return S_ISFIFO(i->i_mode) ? i->i_pipe : NULL; } diff --git a/fs/proc/base.c b/fs/proc/base.c index 9b43ff77a51e..760268d6cba6 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -383,7 +383,7 @@ static int lstats_open(struct inode *inode, struct file *file) static ssize_t lstats_write(struct file *file, const char __user *buf, size_t count, loff_t *offs) { - struct task_struct *task = get_proc_task(file->f_dentry->d_inode); + struct task_struct *task = get_proc_task(file_inode(file)); if (!task) return -ESRCH; @@ -602,7 +602,7 @@ static const struct inode_operations proc_def_inode_operations = { static ssize_t proc_info_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { - struct inode * inode = file->f_path.dentry->d_inode; + struct inode * inode = file_inode(file); unsigned long page; ssize_t length; struct task_struct *task = get_proc_task(inode); @@ -668,7 +668,7 @@ static const struct file_operations proc_single_file_operations = { static int __mem_open(struct inode *inode, struct file *file, unsigned int mode) { - struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); + struct task_struct *task = get_proc_task(file_inode(file)); struct mm_struct *mm; if (!task) @@ -869,7 +869,7 @@ static const struct file_operations proc_environ_operations = { static ssize_t oom_adj_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); + struct task_struct *task = get_proc_task(file_inode(file)); char buffer[PROC_NUMBUF]; int oom_adj = OOM_ADJUST_MIN; size_t len; @@ -916,7 +916,7 @@ static ssize_t oom_adj_write(struct file *file, const char __user *buf, goto out; } - task = get_proc_task(file->f_path.dentry->d_inode); + task = get_proc_task(file_inode(file)); if (!task) { err = -ESRCH; goto out; @@ -976,7 +976,7 @@ static const struct file_operations proc_oom_adj_operations = { static ssize_t oom_score_adj_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); + struct task_struct *task = get_proc_task(file_inode(file)); char buffer[PROC_NUMBUF]; short oom_score_adj = OOM_SCORE_ADJ_MIN; unsigned long flags; @@ -1019,7 +1019,7 @@ static ssize_t oom_score_adj_write(struct file *file, const char __user *buf, goto out; } - task = get_proc_task(file->f_path.dentry->d_inode); + task = get_proc_task(file_inode(file)); if (!task) { err = -ESRCH; goto out; @@ -1067,7 +1067,7 @@ static const struct file_operations proc_oom_score_adj_operations = { static ssize_t proc_loginuid_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { - struct inode * inode = file->f_path.dentry->d_inode; + struct inode * inode = file_inode(file); struct task_struct *task = get_proc_task(inode); ssize_t length; char tmpbuf[TMPBUFLEN]; @@ -1084,7 +1084,7 @@ static ssize_t proc_loginuid_read(struct file * file, char __user * buf, static ssize_t proc_loginuid_write(struct file * file, const char __user * buf, size_t count, loff_t *ppos) { - struct inode * inode = file->f_path.dentry->d_inode; + struct inode * inode = file_inode(file); char *page, *tmp; ssize_t length; uid_t loginuid; @@ -1142,7 +1142,7 @@ static const struct file_operations proc_loginuid_operations = { static ssize_t proc_sessionid_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { - struct inode * inode = file->f_path.dentry->d_inode; + struct inode * inode = file_inode(file); struct task_struct *task = get_proc_task(inode); ssize_t length; char tmpbuf[TMPBUFLEN]; @@ -1165,7 +1165,7 @@ static const struct file_operations proc_sessionid_operations = { static ssize_t proc_fault_inject_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { - struct task_struct *task = get_proc_task(file->f_dentry->d_inode); + struct task_struct *task = get_proc_task(file_inode(file)); char buffer[PROC_NUMBUF]; size_t len; int make_it_fail; @@ -1197,7 +1197,7 @@ static ssize_t proc_fault_inject_write(struct file * file, make_it_fail = simple_strtol(strstrip(buffer), &end, 0); if (*end) return -EINVAL; - task = get_proc_task(file->f_dentry->d_inode); + task = get_proc_task(file_inode(file)); if (!task) return -ESRCH; task->make_it_fail = make_it_fail; @@ -1237,7 +1237,7 @@ static ssize_t sched_write(struct file *file, const char __user *buf, size_t count, loff_t *offset) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct task_struct *p; p = get_proc_task(inode); @@ -1288,7 +1288,7 @@ static ssize_t sched_autogroup_write(struct file *file, const char __user *buf, size_t count, loff_t *offset) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct task_struct *p; char buffer[PROC_NUMBUF]; int nice; @@ -1343,7 +1343,7 @@ static const struct file_operations proc_pid_sched_autogroup_operations = { static ssize_t comm_write(struct file *file, const char __user *buf, size_t count, loff_t *offset) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct task_struct *p; char buffer[TASK_COMM_LEN]; @@ -2146,7 +2146,7 @@ out_no_task: static ssize_t proc_pid_attr_read(struct file * file, char __user * buf, size_t count, loff_t *ppos) { - struct inode * inode = file->f_path.dentry->d_inode; + struct inode * inode = file_inode(file); char *p = NULL; ssize_t length; struct task_struct *task = get_proc_task(inode); @@ -2167,7 +2167,7 @@ static ssize_t proc_pid_attr_read(struct file * file, char __user * buf, static ssize_t proc_pid_attr_write(struct file * file, const char __user * buf, size_t count, loff_t *ppos) { - struct inode * inode = file->f_path.dentry->d_inode; + struct inode * inode = file_inode(file); char *page; ssize_t length; struct task_struct *task = get_proc_task(inode); @@ -2256,7 +2256,7 @@ static const struct inode_operations proc_attr_dir_inode_operations = { static ssize_t proc_coredump_filter_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *task = get_proc_task(file->f_dentry->d_inode); + struct task_struct *task = get_proc_task(file_inode(file)); struct mm_struct *mm; char buffer[PROC_NUMBUF]; size_t len; @@ -2308,7 +2308,7 @@ static ssize_t proc_coredump_filter_write(struct file *file, goto out_no_task; ret = -ESRCH; - task = get_proc_task(file->f_dentry->d_inode); + task = get_proc_task(file_inode(file)); if (!task) goto out_no_task; diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 76ddae83daa5..7dfe548a28e8 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -42,7 +42,7 @@ static ssize_t __proc_file_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos) { - struct inode * inode = file->f_path.dentry->d_inode; + struct inode * inode = file_inode(file); char *page; ssize_t retval=0; int eof=0; @@ -188,7 +188,7 @@ static ssize_t proc_file_read(struct file *file, char __user *buf, size_t nbytes, loff_t *ppos) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); ssize_t rv = -EIO; spin_lock(&pde->pde_unload_lock); @@ -209,7 +209,7 @@ static ssize_t proc_file_write(struct file *file, const char __user *buffer, size_t count, loff_t *ppos) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); ssize_t rv = -EIO; if (pde->write_proc) { @@ -460,7 +460,7 @@ int proc_readdir_de(struct proc_dir_entry *de, struct file *filp, void *dirent, { unsigned int ino; int i; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); int ret = 0; ino = inode->i_ino; @@ -522,7 +522,7 @@ out: int proc_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); return proc_readdir_de(PDE(inode), filp, dirent, filldir); } diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 439ae6886507..38f5c119b806 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -144,7 +144,7 @@ void pde_users_dec(struct proc_dir_entry *pde) static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); loff_t rv = -EINVAL; loff_t (*llseek)(struct file *, loff_t, int); @@ -179,7 +179,7 @@ static loff_t proc_reg_llseek(struct file *file, loff_t offset, int whence) static ssize_t proc_reg_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); ssize_t rv = -EIO; ssize_t (*read)(struct file *, char __user *, size_t, loff_t *); @@ -201,7 +201,7 @@ static ssize_t proc_reg_read(struct file *file, char __user *buf, size_t count, static ssize_t proc_reg_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); ssize_t rv = -EIO; ssize_t (*write)(struct file *, const char __user *, size_t, loff_t *); @@ -223,7 +223,7 @@ static ssize_t proc_reg_write(struct file *file, const char __user *buf, size_t static unsigned int proc_reg_poll(struct file *file, struct poll_table_struct *pts) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); unsigned int rv = DEFAULT_POLLMASK; unsigned int (*poll)(struct file *, struct poll_table_struct *); @@ -245,7 +245,7 @@ static unsigned int proc_reg_poll(struct file *file, struct poll_table_struct *p static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); long rv = -ENOTTY; long (*ioctl)(struct file *, unsigned int, unsigned long); @@ -268,7 +268,7 @@ static long proc_reg_unlocked_ioctl(struct file *file, unsigned int cmd, unsigne #ifdef CONFIG_COMPAT static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); long rv = -ENOTTY; long (*compat_ioctl)(struct file *, unsigned int, unsigned long); @@ -291,7 +291,7 @@ static long proc_reg_compat_ioctl(struct file *file, unsigned int cmd, unsigned static int proc_reg_mmap(struct file *file, struct vm_area_struct *vma) { - struct proc_dir_entry *pde = PDE(file->f_path.dentry->d_inode); + struct proc_dir_entry *pde = PDE(file_inode(file)); int rv = -EIO; int (*mmap)(struct file *, struct vm_area_struct *); diff --git a/fs/proc/nommu.c b/fs/proc/nommu.c index b1822dde55c2..ccfd99bd1c5a 100644 --- a/fs/proc/nommu.c +++ b/fs/proc/nommu.c @@ -45,7 +45,7 @@ static int nommu_region_show(struct seq_file *m, struct vm_region *region) file = region->vm_file; if (file) { - struct inode *inode = region->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(region->vm_file); dev = inode->i_sb->s_dev; ino = inode->i_ino; } diff --git a/fs/proc/proc_net.c b/fs/proc/proc_net.c index fe72cd073dea..75df0d731110 100644 --- a/fs/proc/proc_net.c +++ b/fs/proc/proc_net.c @@ -163,7 +163,7 @@ static int proc_tgid_net_readdir(struct file *filp, void *dirent, struct net *net; ret = -EINVAL; - net = get_proc_task_net(filp->f_path.dentry->d_inode); + net = get_proc_task_net(file_inode(filp)); if (net != NULL) { ret = proc_readdir_de(net->proc_net, filp, dirent, filldir); put_net(net); diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index 1827d88ad58b..612df79cc6a1 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -478,7 +478,7 @@ out: static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, size_t count, loff_t *ppos, int write) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct ctl_table_header *head = grab_header(inode); struct ctl_table *table = PROC_I(inode)->sysctl_entry; ssize_t error; @@ -542,7 +542,7 @@ static int proc_sys_open(struct inode *inode, struct file *filp) static unsigned int proc_sys_poll(struct file *filp, poll_table *wait) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct ctl_table_header *head = grab_header(inode); struct ctl_table *table = PROC_I(inode)->sysctl_entry; unsigned int ret = DEFAULT_POLLMASK; diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index ca5ce7f9f800..3e636d864d56 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -271,7 +271,7 @@ show_map_vma(struct seq_file *m, struct vm_area_struct *vma, int is_pid) const char *name = NULL; if (file) { - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); dev = inode->i_sb->s_dev; ino = inode->i_ino; pgoff = ((loff_t)vma->vm_pgoff) << PAGE_SHIFT; @@ -743,7 +743,7 @@ static ssize_t clear_refs_write(struct file *file, const char __user *buf, return rv; if (type < CLEAR_REFS_ALL || type > CLEAR_REFS_MAPPED) return -EINVAL; - task = get_proc_task(file->f_path.dentry->d_inode); + task = get_proc_task(file_inode(file)); if (!task) return -ESRCH; mm = get_task_mm(task); @@ -1015,7 +1015,7 @@ static int pagemap_hugetlb_range(pte_t *pte, unsigned long hmask, static ssize_t pagemap_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) { - struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); + struct task_struct *task = get_proc_task(file_inode(file)); struct mm_struct *mm; struct pagemapread pm; int ret = -ESRCH; diff --git a/fs/proc/task_nommu.c b/fs/proc/task_nommu.c index 1ccfa537f5f5..56123a6f462e 100644 --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c @@ -149,7 +149,7 @@ static int nommu_vma_show(struct seq_file *m, struct vm_area_struct *vma, file = vma->vm_file; if (file) { - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); dev = inode->i_sb->s_dev; ino = inode->i_ino; pgoff = (loff_t)vma->vm_pgoff << PAGE_SHIFT; diff --git a/fs/qnx4/dir.c b/fs/qnx4/dir.c index 7b0329468a5d..28ce014b3cef 100644 --- a/fs/qnx4/dir.c +++ b/fs/qnx4/dir.c @@ -16,7 +16,7 @@ static int qnx4_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); unsigned int offset; struct buffer_head *bh; struct qnx4_inode_entry *de; diff --git a/fs/qnx6/dir.c b/fs/qnx6/dir.c index dc597353db3b..8798d065e400 100644 --- a/fs/qnx6/dir.c +++ b/fs/qnx6/dir.c @@ -117,7 +117,7 @@ static int qnx6_dir_longfilename(struct inode *inode, static int qnx6_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *s = inode->i_sb; struct qnx6_sb_info *sbi = QNX6_SB(s); loff_t pos = filp->f_pos & (QNX6_DIR_ENTRY_SIZE - 1); diff --git a/fs/ramfs/file-nommu.c b/fs/ramfs/file-nommu.c index d5378d028589..8d5b438cc188 100644 --- a/fs/ramfs/file-nommu.c +++ b/fs/ramfs/file-nommu.c @@ -202,7 +202,7 @@ unsigned long ramfs_nommu_get_unmapped_area(struct file *file, unsigned long pgoff, unsigned long flags) { unsigned long maxpages, lpages, nr, loop, ret; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct page **pages = NULL, **ptr, *page; loff_t isize; diff --git a/fs/read_write.c b/fs/read_write.c index bb34af315280..3ae6dbe828bf 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -163,7 +163,7 @@ EXPORT_SYMBOL(no_llseek); loff_t default_llseek(struct file *file, loff_t offset, int whence) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); loff_t retval; mutex_lock(&inode->i_mutex); @@ -290,7 +290,7 @@ int rw_verify_area(int read_write, struct file *file, loff_t *ppos, size_t count loff_t pos; int retval = -EINVAL; - inode = file->f_path.dentry->d_inode; + inode = file_inode(file); if (unlikely((ssize_t) count < 0)) return retval; pos = *ppos; @@ -901,8 +901,8 @@ ssize_t do_sendfile(int out_fd, int in_fd, loff_t *ppos, size_t count, if (!(out.file->f_mode & FMODE_WRITE)) goto fput_out; retval = -EINVAL; - in_inode = in.file->f_path.dentry->d_inode; - out_inode = out.file->f_path.dentry->d_inode; + in_inode = file_inode(in.file); + out_inode = file_inode(out.file); retval = rw_verify_area(WRITE, out.file, &out.file->f_pos, count); if (retval < 0) goto fput_out; diff --git a/fs/readdir.c b/fs/readdir.c index 5e69ef533b77..fee38e04fae4 100644 --- a/fs/readdir.c +++ b/fs/readdir.c @@ -22,7 +22,7 @@ int vfs_readdir(struct file *file, filldir_t filler, void *buf) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); int res = -ENOTDIR; if (!file->f_op || !file->f_op->readdir) goto out; diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c index 50302d6f8895..6165bd4784f6 100644 --- a/fs/reiserfs/file.c +++ b/fs/reiserfs/file.c @@ -268,7 +268,7 @@ static ssize_t reiserfs_file_write(struct file *file, /* the file we are going t * new current position before returning. */ ) { - struct inode *inode = file->f_path.dentry->d_inode; // Inode of the file that we are writing to. + struct inode *inode = file_inode(file); // Inode of the file that we are writing to. /* To simplify coding at this time, we store locked pages in array for now */ struct reiserfs_transaction_handle th; diff --git a/fs/reiserfs/ioctl.c b/fs/reiserfs/ioctl.c index 0c2185042d5f..15cb5fe6b425 100644 --- a/fs/reiserfs/ioctl.c +++ b/fs/reiserfs/ioctl.c @@ -21,7 +21,7 @@ */ long reiserfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); unsigned int flags; int err = 0; diff --git a/fs/reiserfs/procfs.c b/fs/reiserfs/procfs.c index e60e87035bb3..9cc0740adffa 100644 --- a/fs/reiserfs/procfs.c +++ b/fs/reiserfs/procfs.c @@ -281,7 +281,7 @@ static int show_oidmap(struct seq_file *m, struct super_block *sb) } #if defined( REISERFS_USE_OIDMAPF ) if (sb_info->oidmap.use_file && (sb_info->oidmap.mapf != NULL)) { - loff_t size = sb_info->oidmap.mapf->f_path.dentry->d_inode->i_size; + loff_t size = file_inode(sb_info->oidmap.mapf)->i_size; total_used += size / sizeof(reiserfs_oidinterval_d_t); } #endif diff --git a/fs/romfs/super.c b/fs/romfs/super.c index fd7c5f60b46b..7e8d3a80bdab 100644 --- a/fs/romfs/super.c +++ b/fs/romfs/super.c @@ -147,7 +147,7 @@ static const struct address_space_operations romfs_aops = { */ static int romfs_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *i = filp->f_dentry->d_inode; + struct inode *i = file_inode(filp); struct romfs_inode ri; unsigned long offset, maxoff; int j, ino, nextfh; diff --git a/fs/splice.c b/fs/splice.c index 6909d89d0da5..963213d56403 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -1170,7 +1170,7 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, * randomly drop data for eg socket -> socket splicing. Use the * piped splicing for that! */ - i_mode = in->f_path.dentry->d_inode->i_mode; + i_mode = file_inode(in)->i_mode; if (unlikely(!S_ISREG(i_mode) && !S_ISBLK(i_mode))) return -EINVAL; diff --git a/fs/squashfs/dir.c b/fs/squashfs/dir.c index b381305c9a47..57dc70ebbb19 100644 --- a/fs/squashfs/dir.c +++ b/fs/squashfs/dir.c @@ -102,7 +102,7 @@ static int get_dir_index_using_offset(struct super_block *sb, static int squashfs_readdir(struct file *file, void *dirent, filldir_t filldir) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info; u64 block = squashfs_i(inode)->start + msblk->directory_table; int offset = squashfs_i(inode)->offset, length, dir_count, size, diff --git a/fs/sync.c b/fs/sync.c index 14eefeb44636..2c5d6639a66a 100644 --- a/fs/sync.c +++ b/fs/sync.c @@ -332,7 +332,7 @@ SYSCALL_DEFINE(sync_file_range)(int fd, loff_t offset, loff_t nbytes, if (!f.file) goto out; - i_mode = f.file->f_path.dentry->d_inode->i_mode; + i_mode = file_inode(f.file)->i_mode; ret = -ESPIPE; if (!S_ISREG(i_mode) && !S_ISBLK(i_mode) && !S_ISDIR(i_mode) && !S_ISLNK(i_mode)) diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index 614b2b544880..2ce9a5db6ab5 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c @@ -70,7 +70,7 @@ static ssize_t read(struct file *file, char __user *userbuf, size_t bytes, loff_t *off) { struct bin_buffer *bb = file->private_data; - int size = file->f_path.dentry->d_inode->i_size; + int size = file_inode(file)->i_size; loff_t offs = *off; int count = min_t(size_t, bytes, PAGE_SIZE); char *temp; @@ -140,7 +140,7 @@ static ssize_t write(struct file *file, const char __user *userbuf, size_t bytes, loff_t *off) { struct bin_buffer *bb = file->private_data; - int size = file->f_path.dentry->d_inode->i_size; + int size = file_inode(file)->i_size; loff_t offs = *off; int count = min_t(size_t, bytes, PAGE_SIZE); char *temp; @@ -469,7 +469,7 @@ void unmap_bin_file(struct sysfs_dirent *attr_sd) mutex_lock(&sysfs_bin_lock); hlist_for_each_entry(bb, tmp, &attr_sd->s_bin_attr.buffers, list) { - struct inode *inode = bb->file->f_path.dentry->d_inode; + struct inode *inode = file_inode(bb->file); unmap_mapping_range(inode->i_mapping, 0, 0, 1); } diff --git a/fs/sysv/dir.c b/fs/sysv/dir.c index a77c42157620..3799e8dac3eb 100644 --- a/fs/sysv/dir.c +++ b/fs/sysv/dir.c @@ -68,7 +68,7 @@ static struct page * dir_get_page(struct inode *dir, unsigned long n) static int sysv_readdir(struct file * filp, void * dirent, filldir_t filldir) { unsigned long pos = filp->f_pos; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; unsigned offset = pos & ~PAGE_CACHE_MASK; unsigned long n = pos >> PAGE_CACHE_SHIFT; diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 8a574776a493..de08c92f2e23 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -352,7 +352,7 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) struct qstr nm; union ubifs_key key; struct ubifs_dent_node *dent; - struct inode *dir = file->f_path.dentry->d_inode; + struct inode *dir = file_inode(file); struct ubifs_info *c = dir->i_sb->s_fs_info; dbg_gen("dir ino %lu, f_pos %#llx", dir->i_ino, file->f_pos); diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 5bc77817f382..fa5b347ec729 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -1444,7 +1444,7 @@ static int ubifs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct ubifs_info *c = inode->i_sb->s_fs_info; struct timespec now = ubifs_current_time(inode); struct ubifs_budget_req req = { .new_page = 1 }; diff --git a/fs/ubifs/ioctl.c b/fs/ubifs/ioctl.c index 1a7e2d8bdbe9..648b143606cc 100644 --- a/fs/ubifs/ioctl.c +++ b/fs/ubifs/ioctl.c @@ -147,7 +147,7 @@ out_unlock: long ubifs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { int flags, err; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); switch (cmd) { case FS_IOC_GETFLAGS: diff --git a/fs/udf/dir.c b/fs/udf/dir.c index eb8bfe2b89a5..b3e93f5e17c3 100644 --- a/fs/udf/dir.c +++ b/fs/udf/dir.c @@ -186,7 +186,7 @@ out: static int udf_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct inode *dir = filp->f_path.dentry->d_inode; + struct inode *dir = file_inode(filp); int result; if (filp->f_pos == 0) { diff --git a/fs/udf/file.c b/fs/udf/file.c index 77b5953eaac8..4257a1f5302a 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -139,7 +139,7 @@ static ssize_t udf_file_aio_write(struct kiocb *iocb, const struct iovec *iov, { ssize_t retval; struct file *file = iocb->ki_filp; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); int err, pos; size_t count = iocb->ki_left; struct udf_inode_info *iinfo = UDF_I(inode); @@ -178,7 +178,7 @@ static ssize_t udf_file_aio_write(struct kiocb *iocb, const struct iovec *iov, long udf_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); long old_block, new_block; int result = -EINVAL; diff --git a/fs/ufs/dir.c b/fs/ufs/dir.c index dbc90994715a..3a75ca09c506 100644 --- a/fs/ufs/dir.c +++ b/fs/ufs/dir.c @@ -433,7 +433,7 @@ static int ufs_readdir(struct file *filp, void *dirent, filldir_t filldir) { loff_t pos = filp->f_pos; - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct super_block *sb = inode->i_sb; unsigned int offset = pos & ~PAGE_CACHE_MASK; unsigned long n = pos >> PAGE_CACHE_SHIFT; diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index d0e9c74d3d96..75d854b0c439 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -78,14 +78,14 @@ xfs_swapext( goto out_put_tmp_file; } - if (IS_SWAPFILE(f.file->f_path.dentry->d_inode) || - IS_SWAPFILE(tmp.file->f_path.dentry->d_inode)) { + if (IS_SWAPFILE(file_inode(f.file)) || + IS_SWAPFILE(file_inode(tmp.file))) { error = XFS_ERROR(EINVAL); goto out_put_tmp_file; } - ip = XFS_I(f.file->f_path.dentry->d_inode); - tip = XFS_I(tmp.file->f_path.dentry->d_inode); + ip = XFS_I(file_inode(f.file)); + tip = XFS_I(file_inode(tmp.file)); if (ip->i_mount != tip->i_mount) { error = XFS_ERROR(EINVAL); diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 67284edb84d7..f03bf1a456fb 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -811,7 +811,7 @@ xfs_file_fallocate( loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); long error; loff_t new_size = 0; xfs_flock64_t bf; @@ -912,7 +912,7 @@ xfs_file_readdir( void *dirent, filldir_t filldir) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); xfs_inode_t *ip = XFS_I(inode); int error; size_t bufsize; diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c index c1c3ef88a260..d681e34c2950 100644 --- a/fs/xfs/xfs_ioctl.c +++ b/fs/xfs/xfs_ioctl.c @@ -80,7 +80,7 @@ xfs_find_handle( f = fdget(hreq->fd); if (!f.file) return -EBADF; - inode = f.file->f_path.dentry->d_inode; + inode = file_inode(f.file); } else { error = user_lpath((const char __user *)hreq->path, &path); if (error) @@ -168,7 +168,7 @@ xfs_handle_to_dentry( /* * Only allow handle opens under a directory. */ - if (!S_ISDIR(parfilp->f_path.dentry->d_inode->i_mode)) + if (!S_ISDIR(file_inode(parfilp)->i_mode)) return ERR_PTR(-ENOTDIR); if (hlen != sizeof(xfs_handle_t)) @@ -1334,7 +1334,7 @@ xfs_file_ioctl( unsigned int cmd, unsigned long p) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct xfs_inode *ip = XFS_I(inode); struct xfs_mount *mp = ip->i_mount; void __user *arg = (void __user *)p; diff --git a/fs/xfs/xfs_ioctl32.c b/fs/xfs/xfs_ioctl32.c index 1244274a5674..63b8fc432151 100644 --- a/fs/xfs/xfs_ioctl32.c +++ b/fs/xfs/xfs_ioctl32.c @@ -530,7 +530,7 @@ xfs_file_compat_ioctl( unsigned cmd, unsigned long p) { - struct inode *inode = filp->f_path.dentry->d_inode; + struct inode *inode = file_inode(filp); struct xfs_inode *ip = XFS_I(inode); struct xfs_mount *mp = ip->i_mount; void __user *arg = (void __user *)p; -- cgit From 1afc99beaf0fca3767d9b67789a7ae91c4f7a9c9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 14 Feb 2013 20:41:04 -0500 Subject: propagate error from get_empty_filp() to its callers Based on parts from Anatol's patch (the rest is the next commit). Signed-off-by: Al Viro --- fs/file_table.c | 27 +++++++++++++-------------- fs/namei.c | 4 ++-- fs/open.c | 27 +++++++++++++-------------- 3 files changed, 28 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/file_table.c b/fs/file_table.c index 0f607ce89acc..cd76d4fdf4a4 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -94,8 +94,8 @@ int proc_nr_files(ctl_table *table, int write, #endif /* Find an unused file structure and return a pointer to it. - * Returns NULL, if there are no more free file structures or - * we run out of memory. + * Returns an error pointer if some error happend e.g. we over file + * structures limit, run out of memory or operation is not permitted. * * Be very careful using this. You are responsible for * getting write access to any mount that you might assign @@ -107,7 +107,8 @@ struct file *get_empty_filp(void) { const struct cred *cred = current_cred(); static long old_max; - struct file * f; + struct file *f; + int error; /* * Privileged users can go above max_files @@ -122,13 +123,16 @@ struct file *get_empty_filp(void) } f = kmem_cache_zalloc(filp_cachep, GFP_KERNEL); - if (f == NULL) - goto fail; + if (unlikely(!f)) + return ERR_PTR(-ENOMEM); percpu_counter_inc(&nr_files); f->f_cred = get_cred(cred); - if (security_file_alloc(f)) - goto fail_sec; + error = security_file_alloc(f); + if (unlikely(error)) { + file_free(f); + return ERR_PTR(error); + } INIT_LIST_HEAD(&f->f_u.fu_list); atomic_long_set(&f->f_count, 1); @@ -144,12 +148,7 @@ over: pr_info("VFS: file-max limit %lu reached\n", get_max_files()); old_max = get_nr_files(); } - goto fail; - -fail_sec: - file_free(f); -fail: - return NULL; + return ERR_PTR(-ENFILE); } /** @@ -173,7 +172,7 @@ struct file *alloc_file(struct path *path, fmode_t mode, struct file *file; file = get_empty_filp(); - if (!file) + if (IS_ERR(file)) return NULL; file->f_path = *path; diff --git a/fs/namei.c b/fs/namei.c index df00b754631d..e0a33f598cdf 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2941,8 +2941,8 @@ static struct file *path_openat(int dfd, struct filename *pathname, int error; file = get_empty_filp(); - if (!file) - return ERR_PTR(-ENFILE); + if (IS_ERR(file)) + return file; file->f_flags = op->open_flag; diff --git a/fs/open.c b/fs/open.c index e08643feb574..97a237f67b72 100644 --- a/fs/open.c +++ b/fs/open.c @@ -810,23 +810,22 @@ struct file *dentry_open(const struct path *path, int flags, /* We must always pass in a valid mount pointer. */ BUG_ON(!path->mnt); - error = -ENFILE; f = get_empty_filp(); - if (f == NULL) - return ERR_PTR(error); - - f->f_flags = flags; - f->f_path = *path; - error = do_dentry_open(f, NULL, cred); - if (!error) { - error = open_check_o_direct(f); - if (error) { - fput(f); + if (!IS_ERR(f)) { + f->f_flags = flags; + f->f_path = *path; + error = do_dentry_open(f, NULL, cred); + if (!error) { + /* from now on we need fput() to dispose of f */ + error = open_check_o_direct(f); + if (error) { + fput(f); + f = ERR_PTR(error); + } + } else { + put_filp(f); f = ERR_PTR(error); } - } else { - put_filp(f); - f = ERR_PTR(error); } return f; } -- cgit From 39b652527457452f09b35044fb4f8b3b0eabafdf Mon Sep 17 00:00:00 2001 From: Anatol Pomozov Date: Wed, 12 Sep 2012 20:11:55 -0700 Subject: fs: Preserve error code in get_empty_filp(), part 2 Allocating a file structure in function get_empty_filp() might fail because of several reasons: - not enough memory for file structures - operation is not allowed - user is over its limit Currently the function returns NULL in all cases and we loose the exact reason of the error. All callers of get_empty_filp() assume that the function can fail with ENFILE only. Return error through pointer. Change all callers to preserve this error code. [AV: cleaned up a bit, carved the get_empty_filp() part out into a separate commit (things remaining here deal with alloc_file()), removed pipe(2) behaviour change] Signed-off-by: Anatol Pomozov Reviewed-by: "Theodore Ts'o" Signed-off-by: Al Viro --- fs/anon_inodes.c | 8 +++----- fs/file_table.c | 2 +- fs/hugetlbfs/inode.c | 12 +++++------- fs/pipe.c | 4 ++-- 4 files changed, 11 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c index 28d39fb84ae3..a14eb0c1cd8c 100644 --- a/fs/anon_inodes.c +++ b/fs/anon_inodes.c @@ -131,7 +131,6 @@ struct file *anon_inode_getfile(const char *name, struct qstr this; struct path path; struct file *file; - int error; if (IS_ERR(anon_inode_inode)) return ERR_PTR(-ENODEV); @@ -143,7 +142,7 @@ struct file *anon_inode_getfile(const char *name, * Link the inode to a directory entry by creating a unique name * using the inode sequence number. */ - error = -ENOMEM; + file = ERR_PTR(-ENOMEM); this.name = name; this.len = strlen(name); this.hash = 0; @@ -160,9 +159,8 @@ struct file *anon_inode_getfile(const char *name, d_instantiate(path.dentry, anon_inode_inode); - error = -ENFILE; file = alloc_file(&path, OPEN_FMODE(flags), fops); - if (!file) + if (IS_ERR(file)) goto err_dput; file->f_mapping = anon_inode_inode->i_mapping; @@ -177,7 +175,7 @@ err_dput: path_put(&path); err_module: module_put(fops->owner); - return ERR_PTR(error); + return file; } EXPORT_SYMBOL_GPL(anon_inode_getfile); diff --git a/fs/file_table.c b/fs/file_table.c index cd76d4fdf4a4..aa07d3684a2e 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -173,7 +173,7 @@ struct file *alloc_file(struct path *path, fmode_t mode, file = get_empty_filp(); if (IS_ERR(file)) - return NULL; + return file; file->f_path = *path; file->f_mapping = path->dentry->d_inode->i_mapping; diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index edb42ea60c76..99fe7ef2f09b 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -923,8 +923,7 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, struct user_struct **user, int creat_flags, int page_size_log) { - int error = -ENOMEM; - struct file *file; + struct file *file = ERR_PTR(-ENOMEM); struct inode *inode; struct path path; struct dentry *root; @@ -964,7 +963,7 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, goto out_shm_unlock; path.mnt = mntget(hugetlbfs_vfsmount[hstate_idx]); - error = -ENOSPC; + file = ERR_PTR(-ENOSPC); inode = hugetlbfs_get_inode(root->d_sb, NULL, S_IFREG | S_IRWXUGO, 0); if (!inode) goto out_dentry; @@ -973,7 +972,7 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, size += addr & ~huge_page_mask(hstate); num_pages = ALIGN(size, huge_page_size(hstate)) >> huge_page_shift(hstate); - error = -ENOMEM; + file = ERR_PTR(-ENOMEM); if (hugetlb_reserve_pages(inode, 0, num_pages, NULL, acctflag)) goto out_inode; @@ -981,10 +980,9 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, inode->i_size = size; clear_nlink(inode); - error = -ENFILE; file = alloc_file(&path, FMODE_WRITE | FMODE_READ, &hugetlbfs_file_operations); - if (!file) + if (IS_ERR(file)) goto out_dentry; /* inode is already attached */ return file; @@ -998,7 +996,7 @@ out_shm_unlock: user_shm_unlock(size, *user); *user = NULL; } - return ERR_PTR(error); + return file; } static int __init init_hugetlbfs_fs(void) diff --git a/fs/pipe.c b/fs/pipe.c index 39baf6c3ebb0..64a494cef0a0 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -1037,13 +1037,13 @@ int create_pipe_files(struct file **res, int flags) err = -ENFILE; f = alloc_file(&path, FMODE_WRITE, &write_pipefifo_fops); - if (!f) + if (IS_ERR(f)) goto err_dentry; f->f_flags = O_WRONLY | (flags & (O_NONBLOCK | O_DIRECT)); res[0] = alloc_file(&path, FMODE_READ, &read_pipefifo_fops); - if (!res[0]) + if (IS_ERR(res[0])) goto err_file; path_get(&path); -- cgit From ad8ca3743cb7eef0eb8a2e87943c513201685eec Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 15 Jan 2013 12:54:29 -0500 Subject: vfs: remove d_path_with_unreachable The last caller was removed >2 years ago in commit 7b2a69ba7. Signed-off-by: Jeff Layton Signed-off-by: Al Viro --- fs/dcache.c | 31 ------------------------------- 1 file changed, 31 deletions(-) (limited to 'fs') diff --git a/fs/dcache.c b/fs/dcache.c index 19153a0a810c..8f15e8d71bba 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -2722,37 +2722,6 @@ char *d_path(const struct path *path, char *buf, int buflen) } EXPORT_SYMBOL(d_path); -/** - * d_path_with_unreachable - return the path of a dentry - * @path: path to report - * @buf: buffer to return value in - * @buflen: buffer length - * - * The difference from d_path() is that this prepends "(unreachable)" - * to paths which are unreachable from the current process' root. - */ -char *d_path_with_unreachable(const struct path *path, char *buf, int buflen) -{ - char *res = buf + buflen; - struct path root; - int error; - - if (path->dentry->d_op && path->dentry->d_op->d_dname) - return path->dentry->d_op->d_dname(path->dentry, buf, buflen); - - get_fs_root(current->fs, &root); - write_seqlock(&rename_lock); - error = path_with_deleted(path, &root, &res, &buflen); - if (error > 0) - error = prepend_unreachable(&res, &buflen); - write_sequnlock(&rename_lock); - path_put(&root); - if (error) - res = ERR_PTR(error); - - return res; -} - /* * Helper function for dentry_operations.d_dname() members */ -- cgit From 5f4a6a695075b79261793c25b1128e3b30007646 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 18:04:22 -0500 Subject: link_path_walk(): move assignments to nd->last/nd->last_type up ... and clean the main loop a bit Signed-off-by: Al Viro --- fs/namei.c | 22 ++++++++++------------ 1 file changed, 10 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index e0a33f598cdf..14eab848f7d8 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1802,8 +1802,11 @@ static int link_path_walk(const char *name, struct nameidata *nd) } } + nd->last = this; + nd->last_type = type; + if (!name[len]) - goto last_component; + return 0; /* * If it wasn't NUL, we know it was '/'. Skip that * slash, and continue until no more slashes. @@ -1812,7 +1815,8 @@ static int link_path_walk(const char *name, struct nameidata *nd) len++; } while (unlikely(name[len] == '/')); if (!name[len]) - goto last_component; + return 0; + name += len; err = walk_component(nd, &next, &this, type, LOOKUP_FOLLOW); @@ -1824,16 +1828,10 @@ static int link_path_walk(const char *name, struct nameidata *nd) if (err) return err; } - if (can_lookup(nd->inode)) - continue; - err = -ENOTDIR; - break; - /* here ends the main loop */ - -last_component: - nd->last = this; - nd->last_type = type; - return 0; + if (!can_lookup(nd->inode)) { + err = -ENOTDIR; + break; + } } terminate_walk(nd); return err; -- cgit From 21b9b073924aceb6b8d19c49f61daa86c0340e1a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 18:10:25 -0500 Subject: get rid of name and type arguments of walk_component() ... always can be found in nameidata now. Signed-off-by: Al Viro --- fs/namei.c | 18 ++++++++---------- 1 file changed, 8 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 14eab848f7d8..5f6da6cca394 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1519,7 +1519,7 @@ static inline int should_follow_link(struct inode *inode, int follow) } static inline int walk_component(struct nameidata *nd, struct path *path, - struct qstr *name, int type, int follow) + int follow) { struct inode *inode; int err; @@ -1528,14 +1528,14 @@ static inline int walk_component(struct nameidata *nd, struct path *path, * to be able to know about the current root directory and * parent relationships. */ - if (unlikely(type != LAST_NORM)) - return handle_dots(nd, type); - err = lookup_fast(nd, name, path, &inode); + if (unlikely(nd->last_type != LAST_NORM)) + return handle_dots(nd, nd->last_type); + err = lookup_fast(nd, &nd->last, path, &inode); if (unlikely(err)) { if (err < 0) goto out_err; - err = lookup_slow(nd, name, path); + err = lookup_slow(nd, &nd->last, path); if (err < 0) goto out_err; @@ -1594,8 +1594,7 @@ static inline int nested_symlink(struct path *path, struct nameidata *nd) res = follow_link(&link, nd, &cookie); if (res) break; - res = walk_component(nd, path, &nd->last, - nd->last_type, LOOKUP_FOLLOW); + res = walk_component(nd, path, LOOKUP_FOLLOW); put_link(nd, &link, cookie); } while (res > 0); @@ -1819,7 +1818,7 @@ static int link_path_walk(const char *name, struct nameidata *nd) name += len; - err = walk_component(nd, &next, &this, type, LOOKUP_FOLLOW); + err = walk_component(nd, &next, LOOKUP_FOLLOW); if (err < 0) return err; @@ -1930,8 +1929,7 @@ static inline int lookup_last(struct nameidata *nd, struct path *path) nd->flags |= LOOKUP_FOLLOW | LOOKUP_DIRECTORY; nd->flags &= ~LOOKUP_PARENT; - return walk_component(nd, path, &nd->last, nd->last_type, - nd->flags & LOOKUP_FOLLOW); + return walk_component(nd, path, nd->flags & LOOKUP_FOLLOW); } /* Returns 0 and nd will be valid on success; Retuns error, otherwise. */ -- cgit From e97cdc87be5804eb2922e169f6d81d3e214587ec Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 18:16:00 -0500 Subject: lookup_fast: get rid of name argument Signed-off-by: Al Viro --- fs/namei.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 5f6da6cca394..27823826a500 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1342,7 +1342,7 @@ static struct dentry *__lookup_hash(struct qstr *name, * small and for now I'd prefer to have fast path as straight as possible. * It _is_ time-critical. */ -static int lookup_fast(struct nameidata *nd, struct qstr *name, +static int lookup_fast(struct nameidata *nd, struct path *path, struct inode **inode) { struct vfsmount *mnt = nd->path.mnt; @@ -1358,7 +1358,7 @@ static int lookup_fast(struct nameidata *nd, struct qstr *name, */ if (nd->flags & LOOKUP_RCU) { unsigned seq; - dentry = __d_lookup_rcu(parent, name, &seq, nd->inode); + dentry = __d_lookup_rcu(parent, &nd->last, &seq, nd->inode); if (!dentry) goto unlazy; @@ -1400,7 +1400,7 @@ unlazy: if (unlazy_walk(nd, dentry)) return -ECHILD; } else { - dentry = __d_lookup(parent, name); + dentry = __d_lookup(parent, &nd->last); } if (unlikely(!dentry)) @@ -1530,7 +1530,7 @@ static inline int walk_component(struct nameidata *nd, struct path *path, */ if (unlikely(nd->last_type != LAST_NORM)) return handle_dots(nd, nd->last_type); - err = lookup_fast(nd, &nd->last, path, &inode); + err = lookup_fast(nd, path, &inode); if (unlikely(err)) { if (err < 0) goto out_err; @@ -2728,7 +2728,7 @@ static int do_last(struct nameidata *nd, struct path *path, if (open_flag & O_PATH && !(nd->flags & LOOKUP_FOLLOW)) symlink_ok = true; /* we _can_ be in RCU mode here */ - error = lookup_fast(nd, &nd->last, path, &inode); + error = lookup_fast(nd, path, &inode); if (likely(!error)) goto finish_lookup; -- cgit From cc2a5271155a108c9c6a8d70ec4e4f860f32cd07 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 18:19:49 -0500 Subject: lookup_slow: get rid of name argument Signed-off-by: Al Viro --- fs/namei.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 27823826a500..052c095c2808 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1436,8 +1436,7 @@ need_lookup: } /* Fast lookup failed, do it the slow way */ -static int lookup_slow(struct nameidata *nd, struct qstr *name, - struct path *path) +static int lookup_slow(struct nameidata *nd, struct path *path) { struct dentry *dentry, *parent; int err; @@ -1446,7 +1445,7 @@ static int lookup_slow(struct nameidata *nd, struct qstr *name, BUG_ON(nd->inode != parent->d_inode); mutex_lock(&parent->d_inode->i_mutex); - dentry = __lookup_hash(name, parent, nd->flags); + dentry = __lookup_hash(&nd->last, parent, nd->flags); mutex_unlock(&parent->d_inode->i_mutex); if (IS_ERR(dentry)) return PTR_ERR(dentry); @@ -1535,7 +1534,7 @@ static inline int walk_component(struct nameidata *nd, struct path *path, if (err < 0) goto out_err; - err = lookup_slow(nd, &nd->last, path); + err = lookup_slow(nd, path); if (err < 0) goto out_err; -- cgit From a713ca2ab9d14dc5c86634bc445ce1f66552c169 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 18:27:00 -0500 Subject: constify __d_lookup() arguments Signed-off-by: Al Viro --- fs/dcache.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/dcache.c b/fs/dcache.c index 8f15e8d71bba..93d9f4e1a8ed 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1919,7 +1919,7 @@ EXPORT_SYMBOL(d_lookup); * * __d_lookup callers must be commented. */ -struct dentry *__d_lookup(struct dentry *parent, struct qstr *name) +struct dentry *__d_lookup(const struct dentry *parent, const struct qstr *name) { unsigned int len = name->len; unsigned int hash = name->hash; -- cgit From da2d8455ed7eb22d7642ecee43dc463ac42a1256 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 18:29:34 -0500 Subject: constify d_lookup() arguments Signed-off-by: Al Viro --- fs/dcache.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/dcache.c b/fs/dcache.c index 93d9f4e1a8ed..0e4b5fa6c9f9 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1889,7 +1889,7 @@ seqretry: * dentry is returned. The caller must use dput to free the entry when it has * finished using it. %NULL is returned if the dentry does not exist. */ -struct dentry *d_lookup(struct dentry *parent, struct qstr *name) +struct dentry *d_lookup(const struct dentry *parent, const struct qstr *name) { struct dentry *dentry; unsigned seq; -- cgit From 54c807e71d5ac59dee56c685f2b66e27cd54c475 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 30 Jan 2013 00:28:01 +0100 Subject: fs: Fix possible use-after-free with AIO Running AIO is pinning inode in memory using file reference. Once AIO is completed using aio_complete(), file reference is put and inode can be freed from memory. So we have to be sure that calling aio_complete() is the last thing we do with the inode. CC: Christoph Hellwig CC: Jens Axboe CC: Jeff Moyer CC: stable@vger.kernel.org Acked-by: Jeff Moyer Signed-off-by: Jan Kara Signed-off-by: Al Viro --- fs/direct-io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/direct-io.c b/fs/direct-io.c index cf5b44b10c67..f853263cf74f 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -261,9 +261,9 @@ static ssize_t dio_complete(struct dio *dio, loff_t offset, ssize_t ret, bool is dio->end_io(dio->iocb, offset, transferred, dio->private, ret, is_async); } else { + inode_dio_done(dio->inode); if (is_async) aio_complete(dio->iocb, ret, 0); - inode_dio_done(dio->inode); } return ret; -- cgit From 740da42efa24ac00dfb40d2f02d9d7f7a485049b Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 30 Jan 2013 10:13:38 -0500 Subject: __d_materialise_unique() is too generic Its first argument is always non-root, while the second one is always root. Signed-off-by: Al Viro --- fs/dcache.c | 19 +++++-------------- 1 file changed, 5 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/dcache.c b/fs/dcache.c index 0e4b5fa6c9f9..ada6123414ae 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -2394,7 +2394,7 @@ out_err: */ static void __d_materialise_dentry(struct dentry *dentry, struct dentry *anon) { - struct dentry *dparent, *aparent; + struct dentry *dparent; dentry_lock_for_move(anon, dentry); @@ -2402,24 +2402,15 @@ static void __d_materialise_dentry(struct dentry *dentry, struct dentry *anon) write_seqcount_begin(&anon->d_seq); dparent = dentry->d_parent; - aparent = anon->d_parent; switch_names(dentry, anon); swap(dentry->d_name.hash, anon->d_name.hash); - dentry->d_parent = (aparent == anon) ? dentry : aparent; - list_del(&dentry->d_u.d_child); - if (!IS_ROOT(dentry)) - list_add(&dentry->d_u.d_child, &dentry->d_parent->d_subdirs); - else - INIT_LIST_HEAD(&dentry->d_u.d_child); - - anon->d_parent = (dparent == dentry) ? anon : dparent; + dentry->d_parent = dentry; + list_del_init(&dentry->d_u.d_child); + anon->d_parent = dparent; list_del(&anon->d_u.d_child); - if (!IS_ROOT(anon)) - list_add(&anon->d_u.d_child, &anon->d_parent->d_subdirs); - else - INIT_LIST_HEAD(&anon->d_u.d_child); + list_add(&anon->d_u.d_child, &dparent->d_subdirs); write_seqcount_end(&dentry->d_seq); write_seqcount_end(&anon->d_seq); -- cgit From 4e6b8973282829e0d5898d9721880556957f0f98 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 27 Jan 2013 16:51:34 -0500 Subject: hostfs: directory methods have no business in non-directory inode_operations Signed-off-by: Al Viro --- fs/hostfs/hostfs_kern.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'fs') diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index ba6de25771ac..fbabb906066f 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -861,14 +861,6 @@ int hostfs_setattr(struct dentry *dentry, struct iattr *attr) } static const struct inode_operations hostfs_iops = { - .create = hostfs_create, - .link = hostfs_link, - .unlink = hostfs_unlink, - .symlink = hostfs_symlink, - .mkdir = hostfs_mkdir, - .rmdir = hostfs_rmdir, - .mknod = hostfs_mknod, - .rename = hostfs_rename, .permission = hostfs_permission, .setattr = hostfs_setattr, }; -- cgit From bebeb3d68b24bb4132d452c5707fe321208bcbcd Mon Sep 17 00:00:00 2001 From: Michel Lespinasse Date: Fri, 22 Feb 2013 16:32:37 -0800 Subject: mm: introduce mm_populate() for populating new vmas When creating new mappings using the MAP_POPULATE / MAP_LOCKED flags (or with MCL_FUTURE in effect), we want to populate the pages within the newly created vmas. This may take a while as we may have to read pages from disk, so ideally we want to do this outside of the write-locked mmap_sem region. This change introduces mm_populate(), which is used to defer populating such mappings until after the mmap_sem write lock has been released. This is implemented as a generalization of the former do_mlock_pages(), which accomplished the same task but was using during mlock() / mlockall(). Signed-off-by: Michel Lespinasse Reported-by: Andy Lutomirski Acked-by: Rik van Riel Tested-by: Andy Lutomirski Cc: Greg Ungerer Cc: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/aio.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/aio.c b/fs/aio.c index 71f613cf4a85..82eec7c7b4bb 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -103,6 +103,7 @@ static int aio_setup_ring(struct kioctx *ctx) unsigned nr_events = ctx->max_reqs; unsigned long size; int nr_pages; + bool populate; /* Compensate for the ring buffer's head/tail overlap entry */ nr_events += 2; /* 1 is required, 2 for good luck */ @@ -129,7 +130,8 @@ static int aio_setup_ring(struct kioctx *ctx) down_write(&ctx->mm->mmap_sem); info->mmap_base = do_mmap_pgoff(NULL, 0, info->mmap_size, PROT_READ|PROT_WRITE, - MAP_ANONYMOUS|MAP_PRIVATE, 0); + MAP_ANONYMOUS|MAP_PRIVATE, 0, + &populate); if (IS_ERR((void *)info->mmap_base)) { up_write(&ctx->mm->mmap_sem); info->mmap_size = 0; @@ -147,6 +149,8 @@ static int aio_setup_ring(struct kioctx *ctx) aio_free_ring(ctx); return -EAGAIN; } + if (populate) + mm_populate(info->mmap_base, info->mmap_size); ctx->user_id = info->mmap_base; -- cgit From 41badc15cbad0350de34408c1b0c690f9df76d4b Mon Sep 17 00:00:00 2001 From: Michel Lespinasse Date: Fri, 22 Feb 2013 16:32:47 -0800 Subject: mm: make do_mmap_pgoff return populate as a size in bytes, not as a bool do_mmap_pgoff() rounds up the desired size to the next PAGE_SIZE multiple, however there was no equivalent code in mm_populate(), which caused issues. This could be fixed by introduced the same rounding in mm_populate(), however I think it's preferable to make do_mmap_pgoff() return populate as a size rather than as a boolean, so we don't have to duplicate the size rounding logic in mm_populate(). Signed-off-by: Michel Lespinasse Acked-by: Rik van Riel Tested-by: Andy Lutomirski Cc: Greg Ungerer Cc: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/aio.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/aio.c b/fs/aio.c index 82eec7c7b4bb..064bfbe37566 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -101,9 +101,8 @@ static int aio_setup_ring(struct kioctx *ctx) struct aio_ring *ring; struct aio_ring_info *info = &ctx->ring_info; unsigned nr_events = ctx->max_reqs; - unsigned long size; + unsigned long size, populate; int nr_pages; - bool populate; /* Compensate for the ring buffer's head/tail overlap entry */ nr_events += 2; /* 1 is required, 2 for good luck */ @@ -150,7 +149,7 @@ static int aio_setup_ring(struct kioctx *ctx) return -EAGAIN; } if (populate) - mm_populate(info->mmap_base, info->mmap_size); + mm_populate(info->mmap_base, populate); ctx->user_id = info->mmap_base; -- cgit From 293c07e31ab5a0b8df8c19b2a9e5c6fa30308849 Mon Sep 17 00:00:00 2001 From: Xishi Qiu Date: Fri, 22 Feb 2013 16:34:02 -0800 Subject: memory-failure: use num_poisoned_pages instead of mce_bad_pages Since MCE is an x86 concept, and this code is in mm/, it would be better to use the name num_poisoned_pages instead of mce_bad_pages. [akpm@linux-foundation.org: fix mm/sparse.c] Signed-off-by: Xishi Qiu Signed-off-by: Jiang Liu Suggested-by: Borislav Petkov Reviewed-by: Wanpeng Li Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/meminfo.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c index 80e4645f7990..c3dac611c3c0 100644 --- a/fs/proc/meminfo.c +++ b/fs/proc/meminfo.c @@ -158,7 +158,7 @@ static int meminfo_proc_show(struct seq_file *m, void *v) vmi.used >> 10, vmi.largest_chunk >> 10 #ifdef CONFIG_MEMORY_FAILURE - ,atomic_long_read(&mce_bad_pages) << (PAGE_SHIFT - 10) + ,atomic_long_read(&num_poisoned_pages) << (PAGE_SHIFT - 10) #endif #ifdef CONFIG_TRANSPARENT_HUGEPAGE ,K(global_page_state(NR_ANON_TRANSPARENT_HUGEPAGES) * -- cgit From 33806f06da654092182410d974b6d3c5396ea3eb Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Fri, 22 Feb 2013 16:34:37 -0800 Subject: swap: make each swap partition have one address_space When I use several fast SSD to do swap, swapper_space.tree_lock is heavily contended. This makes each swap partition have one address_space to reduce the lock contention. There is an array of address_space for swap. The swap entry type is the index to the array. In my test with 3 SSD, this increases the swapout throughput 20%. [akpm@linux-foundation.org: revert unneeded change to __add_to_swap_cache] Signed-off-by: Shaohua Li Cc: Hugh Dickins Acked-by: Rik van Riel Acked-by: Minchan Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/meminfo.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c index c3dac611c3c0..1efaaa19c4f3 100644 --- a/fs/proc/meminfo.c +++ b/fs/proc/meminfo.c @@ -40,7 +40,7 @@ static int meminfo_proc_show(struct seq_file *m, void *v) * sysctl_overcommit_ratio / 100) + total_swap_pages; cached = global_page_state(NR_FILE_PAGES) - - total_swapcache_pages - i.bufferram; + total_swapcache_pages() - i.bufferram; if (cached < 0) cached = 0; @@ -109,7 +109,7 @@ static int meminfo_proc_show(struct seq_file *m, void *v) K(i.freeram), K(i.bufferram), K(cached), - K(total_swapcache_pages), + K(total_swapcache_pages()), K(pages[LRU_ACTIVE_ANON] + pages[LRU_ACTIVE_FILE]), K(pages[LRU_INACTIVE_ANON] + pages[LRU_INACTIVE_FILE]), K(pages[LRU_ACTIVE_ANON]), -- cgit From 43be594a6b0ce34ab5fa3b13e08727a99fb95b91 Mon Sep 17 00:00:00 2001 From: Zhang Yanfei Date: Fri, 22 Feb 2013 16:35:46 -0800 Subject: fs/buffer.c: change type of max_buffer_heads to unsigned long max_buffer_heads is calculated from nr_free_buffer_pages(), so change its type to unsigned long in case of overflow. Signed-off-by: Zhang Yanfei Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index 2ea9cd44aeae..62169c192c21 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -3227,7 +3227,7 @@ static struct kmem_cache *bh_cachep __read_mostly; * Once the number of bh's in the machine exceeds this level, we start * stripping them in writeback. */ -static int max_buffer_heads; +static unsigned long max_buffer_heads; int buffer_heads_over_limit; @@ -3343,7 +3343,7 @@ EXPORT_SYMBOL(bh_submit_read); void __init buffer_init(void) { - int nrpages; + unsigned long nrpages; bh_cachep = kmem_cache_create("buffer_head", sizeof(struct buffer_head), 0, -- cgit From 697ce9be7d5c78d6a53dee15f5574f7398cba05c Mon Sep 17 00:00:00 2001 From: Zhang Yanfei Date: Fri, 22 Feb 2013 16:35:47 -0800 Subject: fs/nfsd: change type of max_delegations, nfsd_drc_max_mem and nfsd_drc_mem_used The three variables are calculated from nr_free_buffer_pages so change their types to unsigned long in case of overflow. Signed-off-by: Zhang Yanfei Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/nfsd/nfs4state.c | 6 +++--- fs/nfsd/nfsd.h | 6 +++--- fs/nfsd/nfssvc.c | 6 +++--- 3 files changed, 9 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index ac8ed96c4199..499e957510e7 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -151,7 +151,7 @@ get_nfs4_file(struct nfs4_file *fi) } static int num_delegations; -unsigned int max_delegations; +unsigned long max_delegations; /* * Open owner state (share locks) @@ -700,8 +700,8 @@ static int nfsd4_get_drc_mem(int slotsize, u32 num) num = min_t(u32, num, NFSD_MAX_SLOTS_PER_SESSION); spin_lock(&nfsd_drc_lock); - avail = min_t(int, NFSD_MAX_MEM_PER_SESSION, - nfsd_drc_max_mem - nfsd_drc_mem_used); + avail = min((unsigned long)NFSD_MAX_MEM_PER_SESSION, + nfsd_drc_max_mem - nfsd_drc_mem_used); num = min_t(int, num, avail / slotsize); nfsd_drc_mem_used += num * slotsize; spin_unlock(&nfsd_drc_lock); diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h index de23db255c69..07a473fd49bc 100644 --- a/fs/nfsd/nfsd.h +++ b/fs/nfsd/nfsd.h @@ -56,8 +56,8 @@ extern struct svc_version nfsd_version2, nfsd_version3, extern u32 nfsd_supported_minorversion; extern struct mutex nfsd_mutex; extern spinlock_t nfsd_drc_lock; -extern unsigned int nfsd_drc_max_mem; -extern unsigned int nfsd_drc_mem_used; +extern unsigned long nfsd_drc_max_mem; +extern unsigned long nfsd_drc_mem_used; extern const struct seq_operations nfs_exports_op; @@ -106,7 +106,7 @@ static inline int nfsd_v4client(struct svc_rqst *rq) * NFSv4 State */ #ifdef CONFIG_NFSD_V4 -extern unsigned int max_delegations; +extern unsigned long max_delegations; void nfs4_state_init(void); int nfsd4_init_slabs(void); void nfsd4_free_slabs(void); diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index cee62ab9d4a3..be7af509930c 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -59,8 +59,8 @@ DEFINE_MUTEX(nfsd_mutex); * nfsd_drc_pages_used tracks the current version 4.1 DRC memory usage. */ spinlock_t nfsd_drc_lock; -unsigned int nfsd_drc_max_mem; -unsigned int nfsd_drc_mem_used; +unsigned long nfsd_drc_max_mem; +unsigned long nfsd_drc_mem_used; #if defined(CONFIG_NFSD_V2_ACL) || defined(CONFIG_NFSD_V3_ACL) static struct svc_stat nfsd_acl_svcstats; @@ -342,7 +342,7 @@ static void set_max_drc(void) >> NFSD_DRC_SIZE_SHIFT) * PAGE_SIZE; nfsd_drc_mem_used = 0; spin_lock_init(&nfsd_drc_lock); - dprintk("%s nfsd_drc_max_mem %u \n", __func__, nfsd_drc_max_mem); + dprintk("%s nfsd_drc_max_mem %lu \n", __func__, nfsd_drc_max_mem); } static int nfsd_get_default_max_blksize(void) -- cgit From 78f33277f96430ea001c39e952f6b8200b2ab850 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Sun, 24 Feb 2013 09:55:57 -0500 Subject: pnfs: fix resend_to_mds for directio Pass the directio request on pageio_init to clean up the API. Percolate pg_dreq from original nfs_pageio_descriptor to the pnfs_{read,write}_done_resend_to_mds and use it on respective call to nfs_pageio_init_{read,write} on the newly created nfs_pageio_descriptor. Reproduced by command: mount -o vers=4.1 server:/ /mnt dd bs=128k count=8 if=/dev/zero of=/mnt/dd.out oflag=direct BUG: unable to handle kernel NULL pointer dereference at 0000000000000028 IP: [] atomic_inc+0x4/0x9 [nfs] PGD 34786067 PUD 34794067 PMD 0 Oops: 0002 [#1] SMP Modules linked in: nfs_layout_nfsv41_files nfsv4 nfs nfsd lockd nfs_acl auth_rpcgss exportfs sunrpc btrfs zlib_deflate libcrc32c ipv6 autofs4 CPU 1 Pid: 259, comm: kworker/1:2 Not tainted 3.8.0-rc6 #2 Bochs Bochs RIP: 0010:[] [] atomic_inc+0x4/0x9 [nfs] RSP: 0018:ffff880038f8fa68 EFLAGS: 00010206 RAX: ffffffffa021a6a9 RBX: ffff880038f8fb48 RCX: 00000000000a0000 RDX: ffffffffa021e616 RSI: ffff8800385e9a40 RDI: 0000000000000028 RBP: ffff880038f8fa68 R08: ffffffff81ad6720 R09: ffff8800385e9510 R10: ffffffffa0228450 R11: ffff880038e87418 R12: ffff8800385e9a40 R13: ffff8800385e9a70 R14: ffff880038f8fb38 R15: ffffffffa0148878 FS: 0000000000000000(0000) GS:ffff88003e400000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000000028 CR3: 0000000034789000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process kworker/1:2 (pid: 259, threadinfo ffff880038f8e000, task ffff880038302480) Stack: ffff880038f8fa78 ffffffffa021a6bf ffff880038f8fa88 ffffffffa021bb82 ffff880038f8fae8 ffffffffa021f454 ffff880038f8fae8 ffffffff8109689d ffff880038f8fab8 ffffffff00000006 0000000000000000 ffff880038f8fb48 Call Trace: [] nfs_direct_pgio_init+0x16/0x18 [nfs] [] nfs_pgheader_init+0x6a/0x6c [nfs] [] nfs_generic_pg_writepages+0x51/0xf8 [nfs] [] ? mark_held_locks+0x71/0x99 [] ? rpc_release_resources_task+0x37/0x37 [sunrpc] [] nfs_pageio_doio+0x1a/0x43 [nfs] [] nfs_pageio_complete+0x16/0x2c [nfs] [] pnfs_write_done_resend_to_mds+0x95/0xc5 [nfsv4] [] ? rpc_release_resources_task+0x37/0x37 [sunrpc] [] filelayout_reset_write+0x8c/0x99 [nfs_layout_nfsv41_files] [] filelayout_write_done_cb+0x4d/0xc1 [nfs_layout_nfsv41_files] [] nfs4_write_done+0x36/0x49 [nfsv4] [] nfs_writeback_done+0x53/0x1cc [nfs] [] nfs_writeback_done_common+0xe/0x10 [nfs] [] filelayout_write_call_done+0x28/0x2a [nfs_layout_nfsv41_files] [] rpc_exit_task+0x29/0x87 [sunrpc] [] __rpc_execute+0x11d/0x3cc [sunrpc] [] ? trace_hardirqs_on_caller+0x117/0x173 [] rpc_async_schedule+0x27/0x32 [sunrpc] [] ? __rpc_execute+0x3cc/0x3cc [sunrpc] [] process_one_work+0x226/0x422 [] ? process_one_work+0x159/0x422 [] ? lock_acquired+0x210/0x249 [] ? __rpc_execute+0x3cc/0x3cc [sunrpc] [] worker_thread+0x126/0x1c4 [] ? manage_workers+0x240/0x240 [] kthread+0xb1/0xb9 [] ? __kthread_parkme+0x65/0x65 [] ret_from_fork+0x7c/0xb0 [] ? __kthread_parkme+0x65/0x65 Code: 00 83 38 02 74 12 48 81 4b 50 00 00 01 00 c7 83 60 07 00 00 01 00 00 00 48 89 df e8 55 fe ff ff 5b 41 5c 5d c3 66 90 55 48 89 e5 ff 07 5d c3 55 48 89 e5 f0 ff 0f 0f 94 c0 84 c0 0f 95 c0 0f RIP [] atomic_inc+0x4/0x9 [nfs] RSP CR2: 0000000000000028 Signed-off-by: Benny Halevy Cc: stable@kernel.org [>= 3.6] Signed-off-by: Trond Myklebust --- fs/nfs/nfs4filelayout.c | 6 ++++-- fs/nfs/pnfs.c | 14 ++++++++++---- fs/nfs/pnfs.h | 6 ++++-- 3 files changed, 18 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4filelayout.c b/fs/nfs/nfs4filelayout.c index 194c48410336..49eeb044c109 100644 --- a/fs/nfs/nfs4filelayout.c +++ b/fs/nfs/nfs4filelayout.c @@ -99,7 +99,8 @@ static void filelayout_reset_write(struct nfs_write_data *data) task->tk_status = pnfs_write_done_resend_to_mds(hdr->inode, &hdr->pages, - hdr->completion_ops); + hdr->completion_ops, + hdr->dreq); } } @@ -119,7 +120,8 @@ static void filelayout_reset_read(struct nfs_read_data *data) task->tk_status = pnfs_read_done_resend_to_mds(hdr->inode, &hdr->pages, - hdr->completion_ops); + hdr->completion_ops, + hdr->dreq); } } diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 6be70f622b62..97767c8683f9 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -1422,13 +1422,15 @@ EXPORT_SYMBOL_GPL(pnfs_generic_pg_test); int pnfs_write_done_resend_to_mds(struct inode *inode, struct list_head *head, - const struct nfs_pgio_completion_ops *compl_ops) + const struct nfs_pgio_completion_ops *compl_ops, + struct nfs_direct_req *dreq) { struct nfs_pageio_descriptor pgio; LIST_HEAD(failed); /* Resend all requests through the MDS */ nfs_pageio_init_write(&pgio, inode, FLUSH_STABLE, compl_ops); + pgio.pg_dreq = dreq; while (!list_empty(head)) { struct nfs_page *req = nfs_list_entry(head->next); @@ -1463,7 +1465,8 @@ static void pnfs_ld_handle_write_error(struct nfs_write_data *data) if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) data->task.tk_status = pnfs_write_done_resend_to_mds(hdr->inode, &hdr->pages, - hdr->completion_ops); + hdr->completion_ops, + hdr->dreq); } /* @@ -1578,13 +1581,15 @@ EXPORT_SYMBOL_GPL(pnfs_generic_pg_writepages); int pnfs_read_done_resend_to_mds(struct inode *inode, struct list_head *head, - const struct nfs_pgio_completion_ops *compl_ops) + const struct nfs_pgio_completion_ops *compl_ops, + struct nfs_direct_req *dreq) { struct nfs_pageio_descriptor pgio; LIST_HEAD(failed); /* Resend all requests through the MDS */ nfs_pageio_init_read(&pgio, inode, compl_ops); + pgio.pg_dreq = dreq; while (!list_empty(head)) { struct nfs_page *req = nfs_list_entry(head->next); @@ -1615,7 +1620,8 @@ static void pnfs_ld_handle_read_error(struct nfs_read_data *data) if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) data->task.tk_status = pnfs_read_done_resend_to_mds(hdr->inode, &hdr->pages, - hdr->completion_ops); + hdr->completion_ops, + hdr->dreq); } /* diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index 97cb358bb882..94ba80417748 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -230,9 +230,11 @@ struct pnfs_layout_segment *pnfs_update_layout(struct inode *ino, void nfs4_deviceid_mark_client_invalid(struct nfs_client *clp); int pnfs_read_done_resend_to_mds(struct inode *inode, struct list_head *head, - const struct nfs_pgio_completion_ops *compl_ops); + const struct nfs_pgio_completion_ops *compl_ops, + struct nfs_direct_req *dreq); int pnfs_write_done_resend_to_mds(struct inode *inode, struct list_head *head, - const struct nfs_pgio_completion_ops *compl_ops); + const struct nfs_pgio_completion_ops *compl_ops, + struct nfs_direct_req *dreq); struct nfs4_threshold *pnfs_mdsthreshold_alloc(void); /* nfs4_deviceid_flags */ -- cgit From 561c6731978fa128f29342495f47fc3365898b3d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Feb 2013 10:52:26 -0500 Subject: switch lseek to COMPAT_SYSCALL_DEFINE Signed-off-by: Al Viro --- fs/read_write.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/read_write.c b/fs/read_write.c index bb34af315280..e57796cb7b59 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -15,6 +15,7 @@ #include #include #include +#include #include "read_write.h" #include @@ -247,6 +248,13 @@ SYSCALL_DEFINE3(lseek, unsigned int, fd, off_t, offset, unsigned int, whence) return retval; } +#ifdef CONFIG_COMPAT +COMPAT_SYSCALL_DEFINE3(lseek, unsigned int, fd, compat_off_t, offset, unsigned int, whence) +{ + return sys_lseek(fd, offset, whence); +} +#endif + #ifdef __ARCH_WANT_SYS_LLSEEK SYSCALL_DEFINE5(llseek, unsigned int, fd, unsigned long, offset_high, unsigned long, offset_low, loff_t __user *, result, @@ -278,7 +286,6 @@ out_putf: } #endif - /* * rw_verify_area doesn't like huge counts. We limit * them to something that fits in "int" so that others -- cgit From 3f6d078d4accfff8b114f968259a060bfdc7c682 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Feb 2013 13:49:08 -0500 Subject: fix compat truncate/ftruncate Signed-off-by: Al Viro --- fs/open.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'fs') diff --git a/fs/open.c b/fs/open.c index 9b33c0cbfacf..669ba0dd6667 100644 --- a/fs/open.c +++ b/fs/open.c @@ -30,6 +30,7 @@ #include #include #include +#include #include "internal.h" @@ -140,6 +141,13 @@ SYSCALL_DEFINE2(truncate, const char __user *, path, long, length) return do_sys_truncate(path, length); } +#ifdef CONFIG_COMPAT +COMPAT_SYSCALL_DEFINE2(truncate, const char __user *, path, compat_off_t, length) +{ + return do_sys_truncate(path, length); +} +#endif + static long do_sys_ftruncate(unsigned int fd, loff_t length, int small) { struct inode *inode; @@ -195,6 +203,13 @@ SYSCALL_DEFINE2(ftruncate, unsigned int, fd, unsigned long, length) return ret; } +#ifdef CONFIG_COMPAT +COMPAT_SYSCALL_DEFINE2(ftruncate, unsigned int, fd, compat_ulong_t, length) +{ + return do_sys_ftruncate(fd, length, 1); +} +#endif + /* LFS versions of truncate are only needed on 32 bit machines */ #if BITS_PER_LONG == 32 SYSCALL_DEFINE(truncate64)(const char __user * path, loff_t length) -- cgit From 2c3dd4ff595e604cd4c4c51cff7a208f23148c2d Mon Sep 17 00:00:00 2001 From: Alex Elder Date: Tue, 19 Feb 2013 12:25:56 -0600 Subject: ceph: eliminate sparse warnings in fs code Fix the causes for sparse warnings reported in the ceph file system code. Here there are only two (and they're sort of silly but they're easy to fix). This partially resolves: http://tracker.ceph.com/issues/4184 Reported-by: Fengguang Wu Signed-off-by: Alex Elder Reviewed-by: Josh Durgin --- fs/ceph/xattr.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 2135817e708d..9b6b2b6dd164 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -213,7 +213,7 @@ static struct ceph_vxattr ceph_dir_vxattrs[] = { XATTR_NAME_CEPH(dir, rsubdirs), XATTR_NAME_CEPH(dir, rbytes), XATTR_NAME_CEPH(dir, rctime), - { 0 } /* Required table terminator */ + { .name = NULL, 0 } /* Required table terminator */ }; static size_t ceph_dir_vxattrs_name_size; /* total size of all names */ @@ -232,7 +232,7 @@ static struct ceph_vxattr ceph_file_vxattrs[] = { XATTR_LAYOUT_FIELD(file, layout, stripe_count), XATTR_LAYOUT_FIELD(file, layout, object_size), XATTR_LAYOUT_FIELD(file, layout, pool), - { 0 } /* Required table terminator */ + { .name = NULL, 0 } /* Required table terminator */ }; static size_t ceph_file_vxattrs_name_size; /* total size of all names */ -- cgit From a47970ff7814718fec31b7d966747c6aa1a3545f Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Mon, 25 Feb 2013 21:27:33 -0500 Subject: NFSv4.1: Hold reference to layout hdr in layoutget This fixes an oops where a LAYOUTGET is in still in the rpciod queue, but the requesting processes has been killed. Without this, killing the process does the final pnfs_put_layout_hdr() and sets NFS_I(inode)->layout to NULL while the LAYOUTGET rpc task still references it. Example oops: BUG: unable to handle kernel NULL pointer dereference at 0000000000000080 IP: [] pnfs_choose_layoutget_stateid+0x37/0xef [nfsv4] PGD 7365b067 PUD 7365d067 PMD 0 Oops: 0000 [#1] SMP DEBUG_PAGEALLOC Modules linked in: nfs_layout_nfsv41_files nfsv4 auth_rpcgss nfs lockd sunrpc ipt_MASQUERADE ip6table_mangle ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 iptable_nat nf_nat_ipv4 nf_nat iptable_mangle ip6table_filter ip6_tables ppdev e1000 i2c_piix4 i2c_core shpchp parport_pc parport crc32c_intel aesni_intel xts aes_x86_64 lrw gf128mul ablk_helper cryptd mptspi scsi_transport_spi mptscsih mptbase floppy autofs4 CPU 0 Pid: 27, comm: kworker/0:1 Not tainted 3.8.0-dros_cthon2013+ #4 VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform RIP: 0010:[] [] pnfs_choose_layoutget_stateid+0x37/0xef [nfsv4] RSP: 0018:ffff88007b0c1c88 EFLAGS: 00010246 RAX: ffff88006ed36678 RBX: 0000000000000000 RCX: 0000000ea877e3bc RDX: ffff88007a729da8 RSI: 0000000000000000 RDI: ffff88007a72b958 RBP: ffff88007b0c1ca8 R08: 0000000000000002 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000000 R12: ffff88007a72b958 R13: ffff88007a729da8 R14: 0000000000000000 R15: ffffffffa011077e FS: 0000000000000000(0000) GS:ffff88007f600000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000080 CR3: 00000000735f8000 CR4: 00000000001407f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process kworker/0:1 (pid: 27, threadinfo ffff88007b0c0000, task ffff88007c2fa0c0) Stack: ffff88006fc05388 ffff88007a72b908 ffff88007b240900 ffff88006fc05388 ffff88007b0c1cd8 ffffffffa01a2170 ffff88007b240900 ffff88007b240900 ffff88007b240970 ffffffffa011077e ffff88007b0c1ce8 ffffffffa0110791 Call Trace: [] nfs4_layoutget_prepare+0x7b/0x92 [nfsv4] [] ? __rpc_atrun+0x15/0x15 [sunrpc] [] rpc_prepare_task+0x13/0x15 [sunrpc] Reported-by: Tigran Mkrtchyan Signed-off-by: Weston Andros Adamson Cc: stable@kernel.org Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index eae83bf96c6d..655237fc46df 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -6129,11 +6129,13 @@ static struct page **nfs4_alloc_pages(size_t size, gfp_t gfp_flags) static void nfs4_layoutget_release(void *calldata) { struct nfs4_layoutget *lgp = calldata; - struct nfs_server *server = NFS_SERVER(lgp->args.inode); + struct inode *inode = lgp->args.inode; + struct nfs_server *server = NFS_SERVER(inode); size_t max_pages = max_response_pages(server); dprintk("--> %s\n", __func__); nfs4_free_pages(lgp->args.layout.pages, max_pages); + pnfs_put_layout_hdr(NFS_I(inode)->layout); put_nfs_open_context(lgp->args.ctx); kfree(calldata); dprintk("<-- %s\n", __func__); @@ -6148,7 +6150,8 @@ static const struct rpc_call_ops nfs4_layoutget_call_ops = { struct pnfs_layout_segment * nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags) { - struct nfs_server *server = NFS_SERVER(lgp->args.inode); + struct inode *inode = lgp->args.inode; + struct nfs_server *server = NFS_SERVER(inode); size_t max_pages = max_response_pages(server); struct rpc_task *task; struct rpc_message msg = { @@ -6178,6 +6181,10 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags) lgp->res.layoutp = &lgp->args.layout; lgp->res.seq_res.sr_slot = NULL; nfs41_init_sequence(&lgp->args.seq_args, &lgp->res.seq_res, 0); + + /* nfs4_layoutget_release calls pnfs_put_layout_hdr */ + pnfs_get_layout_hdr(NFS_I(inode)->layout); + task = rpc_run_task(&task_setup_data); if (IS_ERR(task)) return ERR_CAST(task); -- cgit From 0df4d6e5bd008a94f1527aa751bbcf7160257c1d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 14 Feb 2013 22:39:53 -0500 Subject: hugetlb_file_setup(): use d_alloc_pseudo() Signed-off-by: Al Viro --- fs/hugetlbfs/inode.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 99fe7ef2f09b..7f94e0cbc69c 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -918,6 +918,16 @@ static int get_hstate_idx(int page_size_log) return h - hstates; } +static char *hugetlb_dname(struct dentry *dentry, char *buffer, int buflen) +{ + return dynamic_dname(dentry, buffer, buflen, "/%s (deleted)", + dentry->d_name.name); +} + +static struct dentry_operations anon_ops = { + .d_dname = hugetlb_dname +}; + struct file *hugetlb_file_setup(const char *name, unsigned long addr, size_t size, vm_flags_t acctflag, struct user_struct **user, @@ -926,7 +936,7 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, struct file *file = ERR_PTR(-ENOMEM); struct inode *inode; struct path path; - struct dentry *root; + struct super_block *sb; struct qstr quick_string; struct hstate *hstate; unsigned long num_pages; @@ -954,17 +964,18 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, } } - root = hugetlbfs_vfsmount[hstate_idx]->mnt_root; + sb = hugetlbfs_vfsmount[hstate_idx]->mnt_sb; quick_string.name = name; quick_string.len = strlen(quick_string.name); quick_string.hash = 0; - path.dentry = d_alloc(root, &quick_string); + path.dentry = d_alloc_pseudo(sb, &quick_string); if (!path.dentry) goto out_shm_unlock; + d_set_d_op(path.dentry, &anon_ops); path.mnt = mntget(hugetlbfs_vfsmount[hstate_idx]); file = ERR_PTR(-ENOSPC); - inode = hugetlbfs_get_inode(root->d_sb, NULL, S_IFREG | S_IRWXUGO, 0); + inode = hugetlbfs_get_inode(sb, NULL, S_IFREG | S_IRWXUGO, 0); if (!inode) goto out_dentry; -- cgit From 38baba9ea02780c2df35c7a02552fddeb8576e16 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 31 Jan 2013 12:34:58 -0500 Subject: 9p: add fid-based variant of v9fs_xattr_set() ... making v9fs_xattr_set() a wrapper for it. Signed-off-by: Al Viro --- fs/9p/xattr.c | 33 ++++++++++++++++++--------------- fs/9p/xattr.h | 2 ++ 2 files changed, 20 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/9p/xattr.c b/fs/9p/xattr.c index 29653b70a9c3..c45e016b190f 100644 --- a/fs/9p/xattr.c +++ b/fs/9p/xattr.c @@ -110,20 +110,27 @@ ssize_t v9fs_xattr_get(struct dentry *dentry, const char *name, */ int v9fs_xattr_set(struct dentry *dentry, const char *name, const void *value, size_t value_len, int flags) +{ + struct p9_fid *fid = v9fs_fid_lookup(dentry); + if (IS_ERR(fid)) + return PTR_ERR(fid); + return v9fs_fid_xattr_set(fid, name, value, value_len, flags); +} + +int v9fs_fid_xattr_set(struct p9_fid *fid, const char *name, + const void *value, size_t value_len, int flags) { u64 offset = 0; int retval, msize, write_count; - struct p9_fid *fid = NULL; p9_debug(P9_DEBUG_VFS, "name = %s value_len = %zu flags = %d\n", name, value_len, flags); - fid = v9fs_fid_clone(dentry); - if (IS_ERR(fid)) { - retval = PTR_ERR(fid); - fid = NULL; - goto error; - } + /* Clone it */ + fid = p9_client_walk(fid, 0, NULL, 1); + if (IS_ERR(fid)) + return PTR_ERR(fid); + /* * On success fid points to xattr */ @@ -131,7 +138,8 @@ int v9fs_xattr_set(struct dentry *dentry, const char *name, if (retval < 0) { p9_debug(P9_DEBUG_VFS, "p9_client_xattrcreate failed %d\n", retval); - goto error; + p9_client_clunk(fid); + return retval; } msize = fid->clnt->msize; while (value_len) { @@ -144,17 +152,12 @@ int v9fs_xattr_set(struct dentry *dentry, const char *name, if (write_count < 0) { /* error in xattr write */ retval = write_count; - goto error; + break; } offset += write_count; value_len -= write_count; } - /* Total read xattr bytes */ - retval = offset; -error: - if (fid) - retval = p9_client_clunk(fid); - return retval; + return p9_client_clunk(fid); } ssize_t v9fs_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size) diff --git a/fs/9p/xattr.h b/fs/9p/xattr.h index eaa837c53bd5..eec348a3df71 100644 --- a/fs/9p/xattr.h +++ b/fs/9p/xattr.h @@ -27,6 +27,8 @@ extern ssize_t v9fs_fid_xattr_get(struct p9_fid *, const char *, void *, size_t); extern ssize_t v9fs_xattr_get(struct dentry *, const char *, void *, size_t); +extern int v9fs_fid_xattr_set(struct p9_fid *, const char *, + const void *, size_t, int); extern int v9fs_xattr_set(struct dentry *, const char *, const void *, size_t, int); extern ssize_t v9fs_listxattr(struct dentry *, char *, size_t); -- cgit From 7f165aaa7dc898472f3b3fbf2231bb3b5623a3df Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 31 Jan 2013 12:46:55 -0500 Subject: 9p: lift the call of set_cached_acl() into the callers of v9fs_set_acl() Signed-off-by: Al Viro --- fs/9p/acl.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/9p/acl.c b/fs/9p/acl.c index 15b679166201..c5d8cd638e6e 100644 --- a/fs/9p/acl.c +++ b/fs/9p/acl.c @@ -119,10 +119,6 @@ static int v9fs_set_acl(struct dentry *dentry, int type, struct posix_acl *acl) char *name; size_t size; void *buffer; - struct inode *inode = dentry->d_inode; - - set_cached_acl(inode, type, acl); - if (!acl) return 0; @@ -163,6 +159,7 @@ int v9fs_acl_chmod(struct dentry *dentry) retval = posix_acl_chmod(&acl, GFP_KERNEL, inode->i_mode); if (retval) return retval; + set_cached_acl(inode, ACL_TYPE_ACCESS, acl); retval = v9fs_set_acl(dentry, ACL_TYPE_ACCESS, acl); posix_acl_release(acl); } @@ -173,7 +170,9 @@ int v9fs_set_create_acl(struct dentry *dentry, struct posix_acl **dpacl, struct posix_acl **pacl) { if (dentry) { + set_cached_acl(dentry->d_inode, ACL_TYPE_DEFAULT, *dpacl); v9fs_set_acl(dentry, ACL_TYPE_DEFAULT, *dpacl); + set_cached_acl(dentry->d_inode, ACL_TYPE_ACCESS, *pacl); v9fs_set_acl(dentry, ACL_TYPE_ACCESS, *pacl); } posix_acl_release(*dpacl); -- cgit From 0f235caeaed6bbde6d455e172a83fa693ca485a1 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 31 Jan 2013 12:54:47 -0500 Subject: 9p: switch v9fs_set_acl() from dentry to fid Signed-off-by: Al Viro --- fs/9p/acl.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/9p/acl.c b/fs/9p/acl.c index c5d8cd638e6e..506d8af26a1f 100644 --- a/fs/9p/acl.c +++ b/fs/9p/acl.c @@ -23,6 +23,7 @@ #include "acl.h" #include "v9fs.h" #include "v9fs_vfs.h" +#include "fid.h" static struct posix_acl *__v9fs_get_acl(struct p9_fid *fid, char *name) { @@ -113,7 +114,7 @@ struct posix_acl *v9fs_iop_get_acl(struct inode *inode, int type) } -static int v9fs_set_acl(struct dentry *dentry, int type, struct posix_acl *acl) +static int v9fs_set_acl(struct p9_fid *fid, int type, struct posix_acl *acl) { int retval; char *name; @@ -140,7 +141,7 @@ static int v9fs_set_acl(struct dentry *dentry, int type, struct posix_acl *acl) default: BUG(); } - retval = v9fs_xattr_set(dentry, name, buffer, size, 0); + retval = v9fs_fid_xattr_set(fid, name, buffer, size, 0); err_free_out: kfree(buffer); return retval; @@ -151,16 +152,19 @@ int v9fs_acl_chmod(struct dentry *dentry) int retval = 0; struct posix_acl *acl; struct inode *inode = dentry->d_inode; + struct p9_fid *fid = v9fs_fid_lookup(dentry); if (S_ISLNK(inode->i_mode)) return -EOPNOTSUPP; + if (IS_ERR(fid)) + return PTR_ERR(fid); acl = v9fs_get_cached_acl(inode, ACL_TYPE_ACCESS); if (acl) { retval = posix_acl_chmod(&acl, GFP_KERNEL, inode->i_mode); if (retval) return retval; set_cached_acl(inode, ACL_TYPE_ACCESS, acl); - retval = v9fs_set_acl(dentry, ACL_TYPE_ACCESS, acl); + retval = v9fs_set_acl(fid, ACL_TYPE_ACCESS, acl); posix_acl_release(acl); } return retval; @@ -170,10 +174,13 @@ int v9fs_set_create_acl(struct dentry *dentry, struct posix_acl **dpacl, struct posix_acl **pacl) { if (dentry) { + struct p9_fid *fid = v9fs_fid_lookup(dentry); set_cached_acl(dentry->d_inode, ACL_TYPE_DEFAULT, *dpacl); - v9fs_set_acl(dentry, ACL_TYPE_DEFAULT, *dpacl); set_cached_acl(dentry->d_inode, ACL_TYPE_ACCESS, *pacl); - v9fs_set_acl(dentry, ACL_TYPE_ACCESS, *pacl); + if (!IS_ERR(fid)) { + v9fs_set_acl(fid, ACL_TYPE_DEFAULT, *dpacl); + v9fs_set_acl(fid, ACL_TYPE_ACCESS, *pacl); + } } posix_acl_release(*dpacl); posix_acl_release(*pacl); -- cgit From be308f07964a597ea1806a0c02477a383b92f53a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 31 Jan 2013 12:58:16 -0500 Subject: 9p: switch v9fs_acl_chmod() from dentry to inode+fid caller has both, might as well pass them explicitly. Signed-off-by: Al Viro --- fs/9p/acl.c | 6 +----- fs/9p/acl.h | 4 ++-- fs/9p/vfs_inode_dotl.c | 19 ++++++++++--------- 3 files changed, 13 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/9p/acl.c b/fs/9p/acl.c index 506d8af26a1f..81ae143ff3f7 100644 --- a/fs/9p/acl.c +++ b/fs/9p/acl.c @@ -147,17 +147,13 @@ err_free_out: return retval; } -int v9fs_acl_chmod(struct dentry *dentry) +int v9fs_acl_chmod(struct inode *inode, struct p9_fid *fid) { int retval = 0; struct posix_acl *acl; - struct inode *inode = dentry->d_inode; - struct p9_fid *fid = v9fs_fid_lookup(dentry); if (S_ISLNK(inode->i_mode)) return -EOPNOTSUPP; - if (IS_ERR(fid)) - return PTR_ERR(fid); acl = v9fs_get_cached_acl(inode, ACL_TYPE_ACCESS); if (acl) { retval = posix_acl_chmod(&acl, GFP_KERNEL, inode->i_mode); diff --git a/fs/9p/acl.h b/fs/9p/acl.h index 559556411965..3a609c57b6ea 100644 --- a/fs/9p/acl.h +++ b/fs/9p/acl.h @@ -17,7 +17,7 @@ #ifdef CONFIG_9P_FS_POSIX_ACL extern int v9fs_get_acl(struct inode *, struct p9_fid *); extern struct posix_acl *v9fs_iop_get_acl(struct inode *inode, int type); -extern int v9fs_acl_chmod(struct dentry *); +extern int v9fs_acl_chmod(struct inode *, struct p9_fid *); extern int v9fs_set_create_acl(struct dentry *, struct posix_acl **, struct posix_acl **); extern int v9fs_acl_mode(struct inode *dir, umode_t *modep, @@ -28,7 +28,7 @@ static inline int v9fs_get_acl(struct inode *inode, struct p9_fid *fid) { return 0; } -static inline int v9fs_acl_chmod(struct dentry *dentry) +static inline int v9fs_acl_chmod(struct inode *inode, struct p9_fid *fid) { return 0; } diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index 40895546e103..44485f353f19 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -567,10 +567,11 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) struct v9fs_session_info *v9ses; struct p9_fid *fid; struct p9_iattr_dotl p9attr; + struct inode *inode = dentry->d_inode; p9_debug(P9_DEBUG_VFS, "\n"); - retval = inode_change_ok(dentry->d_inode, iattr); + retval = inode_change_ok(inode, iattr); if (retval) return retval; @@ -591,23 +592,23 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) return PTR_ERR(fid); /* Write all dirty data */ - if (S_ISREG(dentry->d_inode->i_mode)) - filemap_write_and_wait(dentry->d_inode->i_mapping); + if (S_ISREG(inode->i_mode)) + filemap_write_and_wait(inode->i_mapping); retval = p9_client_setattr(fid, &p9attr); if (retval < 0) return retval; if ((iattr->ia_valid & ATTR_SIZE) && - iattr->ia_size != i_size_read(dentry->d_inode)) - truncate_setsize(dentry->d_inode, iattr->ia_size); + iattr->ia_size != i_size_read(inode)) + truncate_setsize(inode, iattr->ia_size); - v9fs_invalidate_inode_attr(dentry->d_inode); - setattr_copy(dentry->d_inode, iattr); - mark_inode_dirty(dentry->d_inode); + v9fs_invalidate_inode_attr(inode); + setattr_copy(inode, iattr); + mark_inode_dirty(inode); if (iattr->ia_valid & ATTR_MODE) { /* We also want to update ACL when we update mode bits */ - retval = v9fs_acl_chmod(dentry); + retval = v9fs_acl_chmod(inode, fid); if (retval < 0) return retval; } -- cgit From 5fa6300ae0ccf76018775ea16bc3a061cadc39a6 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 31 Jan 2013 13:31:23 -0500 Subject: 9p: split dropping the acls from v9fs_set_create_acl() Signed-off-by: Al Viro --- fs/9p/acl.c | 26 ++++++++++++++------------ fs/9p/acl.h | 11 ++++++++--- fs/9p/vfs_inode_dotl.c | 12 ++++++------ 3 files changed, 28 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/9p/acl.c b/fs/9p/acl.c index 81ae143ff3f7..5b91689ac0b1 100644 --- a/fs/9p/acl.c +++ b/fs/9p/acl.c @@ -167,23 +167,25 @@ int v9fs_acl_chmod(struct inode *inode, struct p9_fid *fid) } int v9fs_set_create_acl(struct dentry *dentry, - struct posix_acl **dpacl, struct posix_acl **pacl) + struct posix_acl *dacl, struct posix_acl *acl) { - if (dentry) { - struct p9_fid *fid = v9fs_fid_lookup(dentry); - set_cached_acl(dentry->d_inode, ACL_TYPE_DEFAULT, *dpacl); - set_cached_acl(dentry->d_inode, ACL_TYPE_ACCESS, *pacl); - if (!IS_ERR(fid)) { - v9fs_set_acl(fid, ACL_TYPE_DEFAULT, *dpacl); - v9fs_set_acl(fid, ACL_TYPE_ACCESS, *pacl); - } + struct p9_fid *fid = v9fs_fid_lookup(dentry); + set_cached_acl(dentry->d_inode, ACL_TYPE_DEFAULT, dacl); + set_cached_acl(dentry->d_inode, ACL_TYPE_ACCESS, acl); + if (!IS_ERR(fid)) { + v9fs_set_acl(fid, ACL_TYPE_DEFAULT, dacl); + v9fs_set_acl(fid, ACL_TYPE_ACCESS, acl); } - posix_acl_release(*dpacl); - posix_acl_release(*pacl); - *dpacl = *pacl = NULL; return 0; } +void v9fs_put_acl(struct posix_acl *dacl, + struct posix_acl *acl) +{ + posix_acl_release(dacl); + posix_acl_release(acl); +} + int v9fs_acl_mode(struct inode *dir, umode_t *modep, struct posix_acl **dpacl, struct posix_acl **pacl) { diff --git a/fs/9p/acl.h b/fs/9p/acl.h index 3a609c57b6ea..cb7fc54081f7 100644 --- a/fs/9p/acl.h +++ b/fs/9p/acl.h @@ -19,9 +19,10 @@ extern int v9fs_get_acl(struct inode *, struct p9_fid *); extern struct posix_acl *v9fs_iop_get_acl(struct inode *inode, int type); extern int v9fs_acl_chmod(struct inode *, struct p9_fid *); extern int v9fs_set_create_acl(struct dentry *, - struct posix_acl **, struct posix_acl **); + struct posix_acl *, struct posix_acl *); extern int v9fs_acl_mode(struct inode *dir, umode_t *modep, struct posix_acl **dpacl, struct posix_acl **pacl); +extern void v9fs_put_acl(struct posix_acl *dacl, struct posix_acl *acl); #else #define v9fs_iop_get_acl NULL static inline int v9fs_get_acl(struct inode *inode, struct p9_fid *fid) @@ -33,11 +34,15 @@ static inline int v9fs_acl_chmod(struct inode *inode, struct p9_fid *fid) return 0; } static inline int v9fs_set_create_acl(struct dentry *dentry, - struct posix_acl **dpacl, - struct posix_acl **pacl) + struct posix_acl *dacl, + struct posix_acl *acl) { return 0; } +static inline void v9fs_put_acl(struct posix_acl *dacl, + struct posix_acl *acl) +{ +} static inline int v9fs_acl_mode(struct inode *dir, umode_t *modep, struct posix_acl **dpacl, struct posix_acl **pacl) diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index 44485f353f19..dd6355721fc7 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -331,7 +331,7 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, d_instantiate(dentry, inode); /* Now set the ACL based on the default value */ - v9fs_set_create_acl(dentry, &dacl, &pacl); + v9fs_set_create_acl(dentry, dacl, pacl); v9inode = V9FS_I(inode); mutex_lock(&v9inode->v_mutex); @@ -364,6 +364,7 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, #endif *opened |= FILE_CREATED; out: + v9fs_put_acl(dacl, pacl); dput(res); return err; @@ -373,7 +374,6 @@ error: err_clunk_old_fid: if (ofid) p9_client_clunk(ofid); - v9fs_set_create_acl(NULL, &dacl, &pacl); goto out; } @@ -467,13 +467,13 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, d_instantiate(dentry, inode); } /* Now set the ACL based on the default value */ - v9fs_set_create_acl(dentry, &dacl, &pacl); + v9fs_set_create_acl(dentry, dacl, pacl); inc_nlink(dir); v9fs_invalidate_inode_attr(dir); error: if (fid) p9_client_clunk(fid); - v9fs_set_create_acl(NULL, &dacl, &pacl); + v9fs_put_acl(dacl, pacl); return err; } @@ -912,11 +912,11 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode, d_instantiate(dentry, inode); } /* Now set the ACL based on the default value */ - v9fs_set_create_acl(dentry, &dacl, &pacl); + v9fs_set_create_acl(dentry, dacl, pacl); error: if (fid) p9_client_clunk(fid); - v9fs_set_create_acl(NULL, &dacl, &pacl); + v9fs_put_acl(dacl, pacl); return err; } -- cgit From 3592ac444017996f5a8ecf85856af0a8938e8fd1 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 31 Jan 2013 13:45:39 -0500 Subject: 9p: switch v9fs_set_create_acl() to inode+fid, do it before d_instantiate() Signed-off-by: Al Viro --- fs/9p/acl.c | 13 +++++-------- fs/9p/acl.h | 5 +++-- fs/9p/vfs_inode_dotl.c | 50 +++++++++++++++++++++++++------------------------- 3 files changed, 33 insertions(+), 35 deletions(-) (limited to 'fs') diff --git a/fs/9p/acl.c b/fs/9p/acl.c index 5b91689ac0b1..7af425f53bee 100644 --- a/fs/9p/acl.c +++ b/fs/9p/acl.c @@ -166,16 +166,13 @@ int v9fs_acl_chmod(struct inode *inode, struct p9_fid *fid) return retval; } -int v9fs_set_create_acl(struct dentry *dentry, +int v9fs_set_create_acl(struct inode *inode, struct p9_fid *fid, struct posix_acl *dacl, struct posix_acl *acl) { - struct p9_fid *fid = v9fs_fid_lookup(dentry); - set_cached_acl(dentry->d_inode, ACL_TYPE_DEFAULT, dacl); - set_cached_acl(dentry->d_inode, ACL_TYPE_ACCESS, acl); - if (!IS_ERR(fid)) { - v9fs_set_acl(fid, ACL_TYPE_DEFAULT, dacl); - v9fs_set_acl(fid, ACL_TYPE_ACCESS, acl); - } + set_cached_acl(inode, ACL_TYPE_DEFAULT, dacl); + set_cached_acl(inode, ACL_TYPE_ACCESS, acl); + v9fs_set_acl(fid, ACL_TYPE_DEFAULT, dacl); + v9fs_set_acl(fid, ACL_TYPE_ACCESS, acl); return 0; } diff --git a/fs/9p/acl.h b/fs/9p/acl.h index cb7fc54081f7..e4f7e882272b 100644 --- a/fs/9p/acl.h +++ b/fs/9p/acl.h @@ -18,7 +18,7 @@ extern int v9fs_get_acl(struct inode *, struct p9_fid *); extern struct posix_acl *v9fs_iop_get_acl(struct inode *inode, int type); extern int v9fs_acl_chmod(struct inode *, struct p9_fid *); -extern int v9fs_set_create_acl(struct dentry *, +extern int v9fs_set_create_acl(struct inode *, struct p9_fid *, struct posix_acl *, struct posix_acl *); extern int v9fs_acl_mode(struct inode *dir, umode_t *modep, struct posix_acl **dpacl, struct posix_acl **pacl); @@ -33,7 +33,8 @@ static inline int v9fs_acl_chmod(struct inode *inode, struct p9_fid *fid) { return 0; } -static inline int v9fs_set_create_acl(struct dentry *dentry, +static inline int v9fs_set_create_acl(struct inode *inode, + struct p9_fid *fid, struct posix_acl *dacl, struct posix_acl *acl) { diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index dd6355721fc7..7c295588150c 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -325,14 +325,14 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, p9_debug(P9_DEBUG_VFS, "inode creation failed %d\n", err); goto error; } + /* Now set the ACL based on the default value */ + v9fs_set_create_acl(inode, fid, dacl, pacl); + err = v9fs_fid_add(dentry, fid); if (err < 0) goto error; d_instantiate(dentry, inode); - /* Now set the ACL based on the default value */ - v9fs_set_create_acl(dentry, dacl, pacl); - v9inode = V9FS_I(inode); mutex_lock(&v9inode->v_mutex); if (v9ses->cache && !v9inode->writeback_fid && @@ -430,17 +430,17 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, if (err < 0) goto error; + fid = p9_client_walk(dfid, 1, &name, 1); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + p9_debug(P9_DEBUG_VFS, "p9_client_walk failed %d\n", + err); + fid = NULL; + goto error; + } + /* instantiate inode and assign the unopened fid to the dentry */ if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { - fid = p9_client_walk(dfid, 1, &name, 1); - if (IS_ERR(fid)) { - err = PTR_ERR(fid); - p9_debug(P9_DEBUG_VFS, "p9_client_walk failed %d\n", - err); - fid = NULL; - goto error; - } - inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); @@ -451,6 +451,7 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, err = v9fs_fid_add(dentry, fid); if (err < 0) goto error; + v9fs_set_create_acl(inode, fid, dacl, pacl); d_instantiate(dentry, inode); fid = NULL; } else { @@ -464,10 +465,9 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, err = PTR_ERR(inode); goto error; } + v9fs_set_create_acl(inode, fid, dacl, pacl); d_instantiate(dentry, inode); } - /* Now set the ACL based on the default value */ - v9fs_set_create_acl(dentry, dacl, pacl); inc_nlink(dir); v9fs_invalidate_inode_attr(dir); error: @@ -876,17 +876,17 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode, goto error; v9fs_invalidate_inode_attr(dir); + fid = p9_client_walk(dfid, 1, &name, 1); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + p9_debug(P9_DEBUG_VFS, "p9_client_walk failed %d\n", + err); + fid = NULL; + goto error; + } + /* instantiate inode and assign the unopened fid to the dentry */ if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { - fid = p9_client_walk(dfid, 1, &name, 1); - if (IS_ERR(fid)) { - err = PTR_ERR(fid); - p9_debug(P9_DEBUG_VFS, "p9_client_walk failed %d\n", - err); - fid = NULL; - goto error; - } - inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); @@ -894,6 +894,7 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode, err); goto error; } + v9fs_set_create_acl(inode, fid, dacl, pacl); err = v9fs_fid_add(dentry, fid); if (err < 0) goto error; @@ -909,10 +910,9 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode, err = PTR_ERR(inode); goto error; } + v9fs_set_create_acl(inode, fid, dacl, pacl); d_instantiate(dentry, inode); } - /* Now set the ACL based on the default value */ - v9fs_set_create_acl(dentry, dacl, pacl); error: if (fid) p9_client_clunk(fid); -- cgit From 4f522a247bc26d4ab5c8fc406ffffa8b3a77abe3 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 11 Feb 2013 23:20:37 -0500 Subject: d_hash_and_lookup(): export, switch open-coded instances * calling conventions change - ERR_PTR() is returned on ->d_hash() errors; NULL is just for dcache miss now. * exported, open-coded instances in ncpfs and cifs converted. Signed-off-by: Al Viro --- fs/cifs/readdir.c | 8 +++----- fs/dcache.c | 23 +++++++++++------------ fs/ncpfs/dir.c | 10 +++------- fs/proc/base.c | 1 + 4 files changed, 18 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 7255b0c7aa7e..df40cc5fd13a 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -82,12 +82,10 @@ cifs_prime_dcache(struct dentry *parent, struct qstr *name, cFYI(1, "%s: for %s", __func__, name->name); - if (parent->d_op && parent->d_op->d_hash) - parent->d_op->d_hash(parent, parent->d_inode, name); - else - name->hash = full_name_hash(name->name, name->len); + dentry = d_hash_and_lookup(parent, name); + if (unlikely(IS_ERR(dentry))) + return; - dentry = d_lookup(parent, name); if (dentry) { int err; diff --git a/fs/dcache.c b/fs/dcache.c index ada6123414ae..ebab049826c0 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1672,7 +1672,6 @@ EXPORT_SYMBOL(d_splice_alias); struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode, struct qstr *name) { - int error; struct dentry *found; struct dentry *new; @@ -1681,10 +1680,12 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode, * if not go ahead and create it now. */ found = d_hash_and_lookup(dentry->d_parent, name); + if (unlikely(IS_ERR(found))) + goto err_out; if (!found) { new = d_alloc(dentry->d_parent, name); if (!new) { - error = -ENOMEM; + found = ERR_PTR(-ENOMEM); goto err_out; } @@ -1725,7 +1726,7 @@ struct dentry *d_add_ci(struct dentry *dentry, struct inode *inode, err_out: iput(inode); - return ERR_PTR(error); + return found; } EXPORT_SYMBOL(d_add_ci); @@ -1997,12 +1998,10 @@ next: * @dir: Directory to search in * @name: qstr of name we wish to find * - * On hash failure or on lookup failure NULL is returned. + * On lookup failure NULL is returned; on bad name - ERR_PTR(-error) */ struct dentry *d_hash_and_lookup(struct dentry *dir, struct qstr *name) { - struct dentry *dentry = NULL; - /* * Check for a fs-specific hash function. Note that we must * calculate the standard hash first, as the d_op->d_hash() @@ -2010,13 +2009,13 @@ struct dentry *d_hash_and_lookup(struct dentry *dir, struct qstr *name) */ name->hash = full_name_hash(name->name, name->len); if (dir->d_flags & DCACHE_OP_HASH) { - if (dir->d_op->d_hash(dir, dir->d_inode, name) < 0) - goto out; + int err = dir->d_op->d_hash(dir, dir->d_inode, name); + if (unlikely(err < 0)) + return ERR_PTR(err); } - dentry = d_lookup(dir, name); -out: - return dentry; + return d_lookup(dir, name); } +EXPORT_SYMBOL(d_hash_and_lookup); /** * d_validate - verify dentry provided from insecure source (deprecated) @@ -2995,7 +2994,7 @@ ino_t find_inode_number(struct dentry *dir, struct qstr *name) ino_t ino = 0; dentry = d_hash_and_lookup(dir, name); - if (dentry) { + if (!IS_ERR_OR_NULL(dentry)) { if (dentry->d_inode) ino = dentry->d_inode->i_ino; dput(dentry); diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c index 4117e7b377bb..816326093656 100644 --- a/fs/ncpfs/dir.c +++ b/fs/ncpfs/dir.c @@ -593,14 +593,10 @@ ncp_fill_cache(struct file *filp, void *dirent, filldir_t filldir, return 1; /* I'm not sure */ qname.name = __name; - qname.hash = full_name_hash(qname.name, qname.len); - - if (dentry->d_op && dentry->d_op->d_hash) - if (dentry->d_op->d_hash(dentry, dentry->d_inode, &qname) != 0) - goto end_advance; - - newdent = d_lookup(dentry, &qname); + newdent = d_hash_and_lookup(dentry, &qname); + if (unlikely(IS_ERR(newdent))) + goto end_advance; if (!newdent) { newdent = d_alloc(dentry, &qname); if (!newdent) diff --git a/fs/proc/base.c b/fs/proc/base.c index 760268d6cba6..9d9625559727 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -2618,6 +2618,7 @@ static void proc_flush_task_mnt(struct vfsmount *mnt, pid_t pid, pid_t tgid) name.name = buf; name.len = snprintf(buf, sizeof(buf), "%d", pid); + /* no ->d_hash() rejects on procfs */ dentry = d_hash_and_lookup(mnt->mnt_root, &name); if (dentry) { shrink_dcache_parent(dentry); -- cgit From 79f9f99ad1e3e19d4ac300573b51289e3ee8ba86 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 29 Jan 2013 02:55:31 -0500 Subject: ceph: prepopulate inodes only when request is aborted If r_aborted is true, we do not hold the dir i_mutex, and cannot touch the dcache. However, we still need to update the inodes with the state returned by the MDS. Reported-by: Al Viro Signed-off-by: Sage Weil Signed-off-by: Al Viro --- fs/ceph/inode.c | 40 ++++++++++++++++++++++++++++++++++++++-- 1 file changed, 38 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index 2971eaa65cdc..4bc086a72475 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -1130,8 +1130,8 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req, req->r_request_started); dout(" final dn %p\n", dn); i++; - } else if (req->r_op == CEPH_MDS_OP_LOOKUPSNAP || - req->r_op == CEPH_MDS_OP_MKSNAP) { + } else if ((req->r_op == CEPH_MDS_OP_LOOKUPSNAP || + req->r_op == CEPH_MDS_OP_MKSNAP) && !req->r_aborted) { struct dentry *dn = req->r_dentry; /* fill out a snapdir LOOKUPSNAP dentry */ @@ -1195,6 +1195,39 @@ done: /* * Prepopulate our cache with readdir results, leases, etc. */ +static int readdir_prepopulate_inodes_only(struct ceph_mds_request *req, + struct ceph_mds_session *session) +{ + struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info; + int i, err = 0; + + for (i = 0; i < rinfo->dir_nr; i++) { + struct ceph_vino vino; + struct inode *in; + int rc; + + vino.ino = le64_to_cpu(rinfo->dir_in[i].in->ino); + vino.snap = le64_to_cpu(rinfo->dir_in[i].in->snapid); + + in = ceph_get_inode(req->r_dentry->d_sb, vino); + if (IS_ERR(in)) { + err = PTR_ERR(in); + dout("new_inode badness got %d\n", err); + continue; + } + rc = fill_inode(in, &rinfo->dir_in[i], NULL, session, + req->r_request_started, -1, + &req->r_caps_reservation); + if (rc < 0) { + pr_err("fill_inode badness on %p got %d\n", in, rc); + err = rc; + continue; + } + } + + return err; +} + int ceph_readdir_prepopulate(struct ceph_mds_request *req, struct ceph_mds_session *session) { @@ -1209,6 +1242,9 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req, u64 frag = le32_to_cpu(rhead->args.readdir.frag); struct ceph_dentry_info *di; + if (req->r_aborted) + return readdir_prepopulate_inodes_only(req, session); + if (le32_to_cpu(rinfo->head->op) == CEPH_MDS_OP_LSSNAP) { snapdir = ceph_get_snapdir(parent->d_inode); parent = d_find_alias(snapdir); -- cgit From 3dadecce20603aa380023c65e6f55f108fd5e952 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 02:18:08 -0500 Subject: switch vfs_getattr() to struct path Signed-off-by: Al Viro --- fs/ecryptfs/ecryptfs_kernel.h | 6 ++++++ fs/ecryptfs/inode.c | 3 +-- fs/nfsd/nfs3proc.c | 5 +---- fs/nfsd/nfs3xdr.c | 10 +++++----- fs/nfsd/nfs4xdr.c | 4 ++-- fs/nfsd/nfsproc.c | 12 +++--------- fs/nfsd/nfsxdr.c | 3 ++- fs/nfsd/vfs.h | 8 ++++++++ fs/stat.c | 13 ++++++------- 9 files changed, 34 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index cfb4b9fed520..7e2c6f5d7985 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -509,6 +509,12 @@ ecryptfs_dentry_to_lower_mnt(struct dentry *dentry) return ((struct ecryptfs_dentry_info *)dentry->d_fsdata)->lower_path.mnt; } +static inline struct path * +ecryptfs_dentry_to_lower_path(struct dentry *dentry) +{ + return &((struct ecryptfs_dentry_info *)dentry->d_fsdata)->lower_path; +} + static inline void ecryptfs_set_dentry_lower_mnt(struct dentry *dentry, struct vfsmount *lower_mnt) { diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index cc7709e7c508..e0f07fb6d56b 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -1027,8 +1027,7 @@ int ecryptfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat lower_stat; int rc; - rc = vfs_getattr(ecryptfs_dentry_to_lower_mnt(dentry), - ecryptfs_dentry_to_lower(dentry), &lower_stat); + rc = vfs_getattr(ecryptfs_dentry_to_lower_path(dentry), &lower_stat); if (!rc) { fsstack_copy_attr_all(dentry->d_inode, ecryptfs_inode_to_lower(dentry->d_inode)); diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c index 1fc02dfdc5c4..401289913130 100644 --- a/fs/nfsd/nfs3proc.c +++ b/fs/nfsd/nfs3proc.c @@ -43,7 +43,6 @@ static __be32 nfsd3_proc_getattr(struct svc_rqst *rqstp, struct nfsd_fhandle *argp, struct nfsd3_attrstat *resp) { - int err; __be32 nfserr; dprintk("nfsd: GETATTR(3) %s\n", @@ -55,9 +54,7 @@ nfsd3_proc_getattr(struct svc_rqst *rqstp, struct nfsd_fhandle *argp, if (nfserr) RETURN_STATUS(nfserr); - err = vfs_getattr(resp->fh.fh_export->ex_path.mnt, - resp->fh.fh_dentry, &resp->stat); - nfserr = nfserrno(err); + nfserr = fh_getattr(&resp->fh, &resp->stat); RETURN_STATUS(nfserr); } diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c index 324c0baf7cda..7af9417be88d 100644 --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -11,6 +11,7 @@ #include "xdr3.h" #include "auth.h" #include "netns.h" +#include "vfs.h" #define NFSDDBG_FACILITY NFSDDBG_XDR @@ -204,10 +205,10 @@ encode_post_op_attr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp) { struct dentry *dentry = fhp->fh_dentry; if (dentry && dentry->d_inode) { - int err; + __be32 err; struct kstat stat; - err = vfs_getattr(fhp->fh_export->ex_path.mnt, dentry, &stat); + err = fh_getattr(fhp, &stat); if (!err) { *p++ = xdr_one; /* attributes follow */ lease_get_mtime(dentry->d_inode, &stat.mtime); @@ -254,13 +255,12 @@ encode_wcc_data(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp) */ void fill_post_wcc(struct svc_fh *fhp) { - int err; + __be32 err; if (fhp->fh_post_saved) printk("nfsd: inode locked twice during operation.\n"); - err = vfs_getattr(fhp->fh_export->ex_path.mnt, fhp->fh_dentry, - &fhp->fh_post_attr); + err = fh_getattr(fhp, &fhp->fh_post_attr); fhp->fh_post_change = fhp->fh_dentry->d_inode->i_version; if (err) { fhp->fh_post_saved = 0; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 0dc11586682f..17e70dabe21c 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1997,7 +1997,7 @@ static int get_parent_attributes(struct svc_export *exp, struct kstat *stat) if (path.dentry != path.mnt->mnt_root) break; } - err = vfs_getattr(path.mnt, path.dentry, stat); + err = vfs_getattr(&path, stat); path_put(&path); return err; } @@ -2050,7 +2050,7 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, goto out; } - err = vfs_getattr(exp->ex_path.mnt, dentry, &stat); + err = vfs_getattr(&path, &stat); if (err) goto out_nfserr; if ((bmval0 & (FATTR4_WORD0_FILES_FREE | FATTR4_WORD0_FILES_TOTAL | diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c index aad6d457b9e8..54c6b3d3cc79 100644 --- a/fs/nfsd/nfsproc.c +++ b/fs/nfsd/nfsproc.c @@ -26,17 +26,13 @@ static __be32 nfsd_return_attrs(__be32 err, struct nfsd_attrstat *resp) { if (err) return err; - return nfserrno(vfs_getattr(resp->fh.fh_export->ex_path.mnt, - resp->fh.fh_dentry, - &resp->stat)); + return fh_getattr(&resp->fh, &resp->stat); } static __be32 nfsd_return_dirop(__be32 err, struct nfsd_diropres *resp) { if (err) return err; - return nfserrno(vfs_getattr(resp->fh.fh_export->ex_path.mnt, - resp->fh.fh_dentry, - &resp->stat)); + return fh_getattr(&resp->fh, &resp->stat); } /* * Get a file's attributes @@ -150,9 +146,7 @@ nfsd_proc_read(struct svc_rqst *rqstp, struct nfsd_readargs *argp, &resp->count); if (nfserr) return nfserr; - return nfserrno(vfs_getattr(resp->fh.fh_export->ex_path.mnt, - resp->fh.fh_dentry, - &resp->stat)); + return fh_getattr(&resp->fh, &resp->stat); } /* diff --git a/fs/nfsd/nfsxdr.c b/fs/nfsd/nfsxdr.c index 979b42106979..bf6d3bccdd98 100644 --- a/fs/nfsd/nfsxdr.c +++ b/fs/nfsd/nfsxdr.c @@ -4,6 +4,7 @@ * Copyright (C) 1995, 1996 Olaf Kirch */ +#include "vfs.h" #include "xdr.h" #include "auth.h" @@ -197,7 +198,7 @@ encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, __be32 *nfs2svc_encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp) { struct kstat stat; - vfs_getattr(fhp->fh_export->ex_path.mnt, fhp->fh_dentry, &stat); + fh_getattr(fhp, &stat); /* BUG */ return encode_fattr(rqstp, p, fhp, &stat); } diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h index 359594c393d2..5b5894159f22 100644 --- a/fs/nfsd/vfs.h +++ b/fs/nfsd/vfs.h @@ -6,6 +6,7 @@ #define LINUX_NFSD_VFS_H #include "nfsfh.h" +#include "nfsd.h" /* * Flags for nfsd_permission @@ -125,4 +126,11 @@ static inline void fh_drop_write(struct svc_fh *fh) } } +static inline __be32 fh_getattr(struct svc_fh *fh, struct kstat *stat) +{ + struct path p = {.mnt = fh->fh_export->ex_path.mnt, + .dentry = fh->fh_dentry}; + return nfserrno(vfs_getattr(&p, stat)); +} + #endif /* LINUX_NFSD_VFS_H */ diff --git a/fs/stat.c b/fs/stat.c index 14f45459c83d..04ce1ac20d20 100644 --- a/fs/stat.c +++ b/fs/stat.c @@ -37,17 +37,17 @@ void generic_fillattr(struct inode *inode, struct kstat *stat) EXPORT_SYMBOL(generic_fillattr); -int vfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) +int vfs_getattr(struct path *path, struct kstat *stat) { - struct inode *inode = dentry->d_inode; + struct inode *inode = path->dentry->d_inode; int retval; - retval = security_inode_getattr(mnt, dentry); + retval = security_inode_getattr(path->mnt, path->dentry); if (retval) return retval; if (inode->i_op->getattr) - return inode->i_op->getattr(mnt, dentry, stat); + return inode->i_op->getattr(path->mnt, path->dentry, stat); generic_fillattr(inode, stat); return 0; @@ -61,8 +61,7 @@ int vfs_fstat(unsigned int fd, struct kstat *stat) int error = -EBADF; if (f.file) { - error = vfs_getattr(f.file->f_path.mnt, f.file->f_path.dentry, - stat); + error = vfs_getattr(&f.file->f_path, stat); fdput(f); } return error; @@ -89,7 +88,7 @@ retry: if (error) goto out; - error = vfs_getattr(path.mnt, path.dentry, stat); + error = vfs_getattr(&path, stat); path_put(&path); if (retry_estale(error, lookup_flags)) { lookup_flags |= LOOKUP_REVAL; -- cgit From 4f4a4faddea0fe45bf508e723c3a810c5190ed62 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 1 Feb 2013 15:13:04 -0500 Subject: nfsd: handle vfs_getattr errors in acl protocol We're currently ignoring errors from vfs_getattr. The correct thing to do is to do the stat in the main service procedure not in the response encoding. Reported-by: Al Viro Signed-off-by: J. Bruce Fields Signed-off-by: Al Viro --- fs/nfsd/nfs2acl.c | 23 +++++++++++++++++++---- fs/nfsd/nfsxdr.c | 6 ++---- fs/nfsd/xdr.h | 2 +- fs/nfsd/xdr3.h | 2 ++ 4 files changed, 24 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs2acl.c b/fs/nfsd/nfs2acl.c index 9170861c804a..95d76dc6c5da 100644 --- a/fs/nfsd/nfs2acl.c +++ b/fs/nfsd/nfs2acl.c @@ -45,6 +45,10 @@ static __be32 nfsacld_proc_getacl(struct svc_rqst * rqstp, RETURN_STATUS(nfserr_inval); resp->mask = argp->mask; + nfserr = fh_getattr(fh, &resp->stat); + if (nfserr) + goto fail; + if (resp->mask & (NFS_ACL|NFS_ACLCNT)) { acl = nfsd_get_posix_acl(fh, ACL_TYPE_ACCESS); if (IS_ERR(acl)) { @@ -115,6 +119,9 @@ static __be32 nfsacld_proc_setacl(struct svc_rqst * rqstp, nfserr = nfserrno( nfsd_set_posix_acl( fh, ACL_TYPE_DEFAULT, argp->acl_default) ); } + if (!nfserr) { + nfserr = fh_getattr(fh, &resp->stat); + } /* argp->acl_{access,default} may have been allocated in nfssvc_decode_setaclargs. */ @@ -129,10 +136,15 @@ static __be32 nfsacld_proc_setacl(struct svc_rqst * rqstp, static __be32 nfsacld_proc_getattr(struct svc_rqst * rqstp, struct nfsd_fhandle *argp, struct nfsd_attrstat *resp) { + __be32 nfserr; dprintk("nfsd: GETATTR %s\n", SVCFH_fmt(&argp->fh)); fh_copy(&resp->fh, &argp->fh); - return fh_verify(rqstp, &resp->fh, 0, NFSD_MAY_NOP); + nfserr = fh_verify(rqstp, &resp->fh, 0, NFSD_MAY_NOP); + if (nfserr) + return nfserr; + nfserr = fh_getattr(&resp->fh, &resp->stat); + return nfserr; } /* @@ -150,6 +162,9 @@ static __be32 nfsacld_proc_access(struct svc_rqst *rqstp, struct nfsd3_accessarg fh_copy(&resp->fh, &argp->fh); resp->access = argp->access; nfserr = nfsd_access(rqstp, &resp->fh, &resp->access, NULL); + if (nfserr) + return nfserr; + nfserr = fh_getattr(&resp->fh, &resp->stat); return nfserr; } @@ -243,7 +258,7 @@ static int nfsaclsvc_encode_getaclres(struct svc_rqst *rqstp, __be32 *p, return 0; inode = dentry->d_inode; - p = nfs2svc_encode_fattr(rqstp, p, &resp->fh); + p = nfs2svc_encode_fattr(rqstp, p, &resp->fh, &resp->stat); *p++ = htonl(resp->mask); if (!xdr_ressize_check(rqstp, p)) return 0; @@ -274,7 +289,7 @@ static int nfsaclsvc_encode_getaclres(struct svc_rqst *rqstp, __be32 *p, static int nfsaclsvc_encode_attrstatres(struct svc_rqst *rqstp, __be32 *p, struct nfsd_attrstat *resp) { - p = nfs2svc_encode_fattr(rqstp, p, &resp->fh); + p = nfs2svc_encode_fattr(rqstp, p, &resp->fh, &resp->stat); return xdr_ressize_check(rqstp, p); } @@ -282,7 +297,7 @@ static int nfsaclsvc_encode_attrstatres(struct svc_rqst *rqstp, __be32 *p, static int nfsaclsvc_encode_accessres(struct svc_rqst *rqstp, __be32 *p, struct nfsd3_accessres *resp) { - p = nfs2svc_encode_fattr(rqstp, p, &resp->fh); + p = nfs2svc_encode_fattr(rqstp, p, &resp->fh, &resp->stat); *p++ = htonl(resp->access); return xdr_ressize_check(rqstp, p); } diff --git a/fs/nfsd/nfsxdr.c b/fs/nfsd/nfsxdr.c index bf6d3bccdd98..96e56192f984 100644 --- a/fs/nfsd/nfsxdr.c +++ b/fs/nfsd/nfsxdr.c @@ -195,11 +195,9 @@ encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, } /* Helper function for NFSv2 ACL code */ -__be32 *nfs2svc_encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp) +__be32 *nfs2svc_encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, struct kstat *stat) { - struct kstat stat; - fh_getattr(fhp, &stat); /* BUG */ - return encode_fattr(rqstp, p, fhp, &stat); + return encode_fattr(rqstp, p, fhp, stat); } /* diff --git a/fs/nfsd/xdr.h b/fs/nfsd/xdr.h index 53b1863dd8f6..4f0481d63804 100644 --- a/fs/nfsd/xdr.h +++ b/fs/nfsd/xdr.h @@ -167,7 +167,7 @@ int nfssvc_encode_entry(void *, const char *name, int nfssvc_release_fhandle(struct svc_rqst *, __be32 *, struct nfsd_fhandle *); /* Helper functions for NFSv2 ACL code */ -__be32 *nfs2svc_encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp); +__be32 *nfs2svc_encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp, struct kstat *stat); __be32 *nfs2svc_decode_fh(__be32 *p, struct svc_fh *fhp); #endif /* LINUX_NFSD_H */ diff --git a/fs/nfsd/xdr3.h b/fs/nfsd/xdr3.h index 7df980eb0562..b6d5542a4ac8 100644 --- a/fs/nfsd/xdr3.h +++ b/fs/nfsd/xdr3.h @@ -136,6 +136,7 @@ struct nfsd3_accessres { __be32 status; struct svc_fh fh; __u32 access; + struct kstat stat; }; struct nfsd3_readlinkres { @@ -225,6 +226,7 @@ struct nfsd3_getaclres { int mask; struct posix_acl *acl_access; struct posix_acl *acl_default; + struct kstat stat; }; /* dummy type for release */ -- cgit From ecf3d1f1aa74da0d632b651a2e05a911f60e92c0 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Wed, 20 Feb 2013 11:19:05 -0500 Subject: vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op The following set of operations on a NFS client and server will cause server# mkdir a client# cd a server# mv a a.bak client# sleep 30 # (or whatever the dir attrcache timeout is) client# stat . stat: cannot stat `.': Stale NFS file handle Obviously, we should not be getting an ESTALE error back there since the inode still exists on the server. The problem is that the lookup code will call d_revalidate on the dentry that "." refers to, because NFS has FS_REVAL_DOT set. nfs_lookup_revalidate will see that the parent directory has changed and will try to reverify the dentry by redoing a LOOKUP. That of course fails, so the lookup code returns ESTALE. The problem here is that d_revalidate is really a bad fit for this case. What we really want to know at this point is whether the inode is still good or not, but we don't really care what name it goes by or whether the dcache is still valid. Add a new d_op->d_weak_revalidate operation and have complete_walk call that instead of d_revalidate. The intent there is to allow for a "weaker" d_revalidate that just checks to see whether the inode is still good. This is also gives us an opportunity to kill off the FS_REVAL_DOT special casing. [AV: changed method name, added note in porting, fixed confusion re having it possibly called from RCU mode (it won't be)] Cc: NeilBrown Signed-off-by: Jeff Layton Signed-off-by: Al Viro --- fs/9p/vfs_dentry.c | 1 + fs/9p/vfs_super.c | 2 +- fs/dcache.c | 3 +++ fs/namei.c | 8 ++------ fs/nfs/dir.c | 40 ++++++++++++++++++++++++++++++++++++++++ fs/nfs/nfs4super.c | 6 +++--- fs/nfs/super.c | 6 +++--- 7 files changed, 53 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index 64600b5d0522..9ad68628522c 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -137,6 +137,7 @@ out_valid: const struct dentry_operations v9fs_cached_dentry_operations = { .d_revalidate = v9fs_lookup_revalidate, + .d_weak_revalidate = v9fs_lookup_revalidate, .d_delete = v9fs_cached_dentry_delete, .d_release = v9fs_dentry_release, }; diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 137d50396898..91dad63e5a2d 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -363,5 +363,5 @@ struct file_system_type v9fs_fs_type = { .mount = v9fs_mount, .kill_sb = v9fs_kill_super, .owner = THIS_MODULE, - .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT, + .fs_flags = FS_RENAME_DOES_D_MOVE, }; diff --git a/fs/dcache.c b/fs/dcache.c index ebab049826c0..68220dd0c135 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1358,6 +1358,7 @@ void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op) WARN_ON_ONCE(dentry->d_flags & (DCACHE_OP_HASH | DCACHE_OP_COMPARE | DCACHE_OP_REVALIDATE | + DCACHE_OP_WEAK_REVALIDATE | DCACHE_OP_DELETE )); dentry->d_op = op; if (!op) @@ -1368,6 +1369,8 @@ void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op) dentry->d_flags |= DCACHE_OP_COMPARE; if (op->d_revalidate) dentry->d_flags |= DCACHE_OP_REVALIDATE; + if (op->d_weak_revalidate) + dentry->d_flags |= DCACHE_OP_WEAK_REVALIDATE; if (op->d_delete) dentry->d_flags |= DCACHE_OP_DELETE; if (op->d_prune) diff --git a/fs/namei.c b/fs/namei.c index 052c095c2808..dc984fee5532 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -600,14 +600,10 @@ static int complete_walk(struct nameidata *nd) if (likely(!(nd->flags & LOOKUP_JUMPED))) return 0; - if (likely(!(dentry->d_flags & DCACHE_OP_REVALIDATE))) + if (likely(!(dentry->d_flags & DCACHE_OP_WEAK_REVALIDATE))) return 0; - if (likely(!(dentry->d_sb->s_type->fs_flags & FS_REVAL_DOT))) - return 0; - - /* Note: we do not d_invalidate() */ - status = d_revalidate(dentry, nd->flags); + status = dentry->d_op->d_weak_revalidate(dentry, nd->flags); if (status > 0) return 0; diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index a8bd28cde7e2..f23f455be42b 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1135,6 +1135,45 @@ out_error: return error; } +/* + * A weaker form of d_revalidate for revalidating just the dentry->d_inode + * when we don't really care about the dentry name. This is called when a + * pathwalk ends on a dentry that was not found via a normal lookup in the + * parent dir (e.g.: ".", "..", procfs symlinks or mountpoint traversals). + * + * In this situation, we just want to verify that the inode itself is OK + * since the dentry might have changed on the server. + */ +static int nfs_weak_revalidate(struct dentry *dentry, unsigned int flags) +{ + int error; + struct inode *inode = dentry->d_inode; + + /* + * I believe we can only get a negative dentry here in the case of a + * procfs-style symlink. Just assume it's correct for now, but we may + * eventually need to do something more here. + */ + if (!inode) { + dfprintk(LOOKUPCACHE, "%s: %s/%s has negative inode\n", + __func__, dentry->d_parent->d_name.name, + dentry->d_name.name); + return 1; + } + + if (is_bad_inode(inode)) { + dfprintk(LOOKUPCACHE, "%s: %s/%s has dud inode\n", + __func__, dentry->d_parent->d_name.name, + dentry->d_name.name); + return 0; + } + + error = nfs_revalidate_inode(NFS_SERVER(inode), inode); + dfprintk(LOOKUPCACHE, "NFS: %s: inode %lu is %s\n", + __func__, inode->i_ino, error ? "invalid" : "valid"); + return !error; +} + /* * This is called from dput() when d_count is going to 0. */ @@ -1202,6 +1241,7 @@ static void nfs_d_release(struct dentry *dentry) const struct dentry_operations nfs_dentry_operations = { .d_revalidate = nfs_lookup_revalidate, + .d_weak_revalidate = nfs_weak_revalidate, .d_delete = nfs_dentry_delete, .d_iput = nfs_dentry_iput, .d_automount = nfs_d_automount, diff --git a/fs/nfs/nfs4super.c b/fs/nfs/nfs4super.c index 84d2e9e2f313..569b166cc050 100644 --- a/fs/nfs/nfs4super.c +++ b/fs/nfs/nfs4super.c @@ -28,7 +28,7 @@ static struct file_system_type nfs4_remote_fs_type = { .name = "nfs4", .mount = nfs4_remote_mount, .kill_sb = nfs_kill_super, - .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; static struct file_system_type nfs4_remote_referral_fs_type = { @@ -36,7 +36,7 @@ static struct file_system_type nfs4_remote_referral_fs_type = { .name = "nfs4", .mount = nfs4_remote_referral_mount, .kill_sb = nfs_kill_super, - .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; struct file_system_type nfs4_referral_fs_type = { @@ -44,7 +44,7 @@ struct file_system_type nfs4_referral_fs_type = { .name = "nfs4", .mount = nfs4_referral_mount, .kill_sb = nfs_kill_super, - .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; static const struct super_operations nfs4_sops = { diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 2e7e8c878e5d..92acc26f9c5f 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -292,7 +292,7 @@ struct file_system_type nfs_fs_type = { .name = "nfs", .mount = nfs_fs_mount, .kill_sb = nfs_kill_super, - .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; EXPORT_SYMBOL_GPL(nfs_fs_type); @@ -301,7 +301,7 @@ struct file_system_type nfs_xdev_fs_type = { .name = "nfs", .mount = nfs_xdev_mount, .kill_sb = nfs_kill_super, - .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; const struct super_operations nfs_sops = { @@ -331,7 +331,7 @@ struct file_system_type nfs4_fs_type = { .name = "nfs4", .mount = nfs_fs_mount, .kill_sb = nfs_kill_super, - .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; EXPORT_SYMBOL_GPL(nfs4_fs_type); -- cgit From 182be684784334598eee1d90274e7f7aa0063616 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 24 Jan 2013 02:21:54 -0500 Subject: kill f_vfsmnt very few users left... Signed-off-by: Al Viro --- fs/ocfs2/dir.c | 2 +- fs/ocfs2/file.c | 4 ++-- fs/ocfs2/mmap.c | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index ac0d4a0e8a41..c87d0793bdec 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -2020,7 +2020,7 @@ int ocfs2_readdir(struct file * filp, void * dirent, filldir_t filldir) trace_ocfs2_readdir((unsigned long long)OCFS2_I(inode)->ip_blkno); - error = ocfs2_inode_lock_atime(inode, filp->f_vfsmnt, &lock_level); + error = ocfs2_inode_lock_atime(inode, filp->f_path.mnt, &lock_level); if (lock_level && error >= 0) { /* We release EX lock which used to update atime * and get PR lock again to reduce contention diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 04098af9dbc8..5bcd865905ef 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2526,7 +2526,7 @@ static ssize_t ocfs2_file_splice_read(struct file *in, /* * See the comment in ocfs2_file_aio_read() */ - ret = ocfs2_inode_lock_atime(inode, in->f_vfsmnt, &lock_level); + ret = ocfs2_inode_lock_atime(inode, in->f_path.mnt, &lock_level); if (ret < 0) { mlog_errno(ret); goto bail; @@ -2589,7 +2589,7 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb, * like i_size. This allows the checks down below * generic_file_aio_read() a chance of actually working. */ - ret = ocfs2_inode_lock_atime(inode, filp->f_vfsmnt, &lock_level); + ret = ocfs2_inode_lock_atime(inode, filp->f_path.mnt, &lock_level); if (ret < 0) { mlog_errno(ret); goto bail; diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c index 07c585b85000..10d66c75cecb 100644 --- a/fs/ocfs2/mmap.c +++ b/fs/ocfs2/mmap.c @@ -181,7 +181,7 @@ int ocfs2_mmap(struct file *file, struct vm_area_struct *vma) int ret = 0, lock_level = 0; ret = ocfs2_inode_lock_atime(file_inode(file), - file->f_vfsmnt, &lock_level); + file->f_path.mnt, &lock_level); if (ret < 0) { mlog_errno(ret); goto out; -- cgit From 94e07a7590ae855bae0536c42b3086fadc7c83a8 Mon Sep 17 00:00:00 2001 From: Namjae Jeon Date: Sun, 17 Feb 2013 15:48:11 +0900 Subject: fs: encode_fh: return FILEID_INVALID if invalid fid_type This patch is a follow up on below patch: [PATCH] exportfs: add FILEID_INVALID to indicate invalid fid_type commit: 216b6cbdcbd86b1db0754d58886b466ae31f5a63 Signed-off-by: Namjae Jeon Signed-off-by: Vivek Trivedi Acked-by: Steven Whitehouse Acked-by: Sage Weil Signed-off-by: Al Viro --- fs/btrfs/export.c | 4 ++-- fs/ceph/export.c | 4 ++-- fs/fuse/inode.c | 2 +- fs/gfs2/export.c | 4 ++-- fs/isofs/export.c | 4 ++-- fs/nilfs2/namei.c | 4 ++-- fs/ocfs2/export.c | 4 ++-- fs/reiserfs/inode.c | 4 ++-- fs/udf/namei.c | 4 ++-- fs/xfs/xfs_export.c | 4 ++-- 10 files changed, 19 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/export.c b/fs/btrfs/export.c index 614f34a899c2..81ee29eeb7ca 100644 --- a/fs/btrfs/export.c +++ b/fs/btrfs/export.c @@ -22,10 +22,10 @@ static int btrfs_encode_fh(struct inode *inode, u32 *fh, int *max_len, if (parent && (len < BTRFS_FID_SIZE_CONNECTABLE)) { *max_len = BTRFS_FID_SIZE_CONNECTABLE; - return 255; + return FILEID_INVALID; } else if (len < BTRFS_FID_SIZE_NON_CONNECTABLE) { *max_len = BTRFS_FID_SIZE_NON_CONNECTABLE; - return 255; + return FILEID_INVALID; } len = BTRFS_FID_SIZE_NON_CONNECTABLE; diff --git a/fs/ceph/export.c b/fs/ceph/export.c index ca3ab3f9ca70..16796be53ca5 100644 --- a/fs/ceph/export.c +++ b/fs/ceph/export.c @@ -81,7 +81,7 @@ static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len, if (parent_inode) { /* nfsd wants connectable */ *max_len = connected_handle_length; - type = 255; + type = FILEID_INVALID; } else { dout("encode_fh %p\n", dentry); fh->ino = ceph_ino(inode); @@ -90,7 +90,7 @@ static int ceph_encode_fh(struct inode *inode, u32 *rawfh, int *max_len, } } else { *max_len = handle_length; - type = 255; + type = FILEID_INVALID; } if (dentry) dput(dentry); diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 73ca6b72beaf..b730fda9aa4c 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -678,7 +678,7 @@ static int fuse_encode_fh(struct inode *inode, u32 *fh, int *max_len, if (*max_len < len) { *max_len = len; - return 255; + return FILEID_INVALID; } nodeid = get_fuse_inode(inode)->nodeid; diff --git a/fs/gfs2/export.c b/fs/gfs2/export.c index 4767774a5f3e..9973df4ff565 100644 --- a/fs/gfs2/export.c +++ b/fs/gfs2/export.c @@ -37,10 +37,10 @@ static int gfs2_encode_fh(struct inode *inode, __u32 *p, int *len, if (parent && (*len < GFS2_LARGE_FH_SIZE)) { *len = GFS2_LARGE_FH_SIZE; - return 255; + return FILEID_INVALID; } else if (*len < GFS2_SMALL_FH_SIZE) { *len = GFS2_SMALL_FH_SIZE; - return 255; + return FILEID_INVALID; } fh[0] = cpu_to_be32(ip->i_no_formal_ino >> 32); diff --git a/fs/isofs/export.c b/fs/isofs/export.c index 2b4f2358eadb..12088d8de3fa 100644 --- a/fs/isofs/export.c +++ b/fs/isofs/export.c @@ -125,10 +125,10 @@ isofs_export_encode_fh(struct inode *inode, */ if (parent && (len < 5)) { *max_len = 5; - return 255; + return FILEID_INVALID; } else if (len < 3) { *max_len = 3; - return 255; + return FILEID_INVALID; } len = 3; diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c index 1d0c0b84c5a3..9de78f08989e 100644 --- a/fs/nilfs2/namei.c +++ b/fs/nilfs2/namei.c @@ -517,11 +517,11 @@ static int nilfs_encode_fh(struct inode *inode, __u32 *fh, int *lenp, if (parent && *lenp < NILFS_FID_SIZE_CONNECTABLE) { *lenp = NILFS_FID_SIZE_CONNECTABLE; - return 255; + return FILEID_INVALID; } if (*lenp < NILFS_FID_SIZE_NON_CONNECTABLE) { *lenp = NILFS_FID_SIZE_NON_CONNECTABLE; - return 255; + return FILEID_INVALID; } fid->cno = root->cno; diff --git a/fs/ocfs2/export.c b/fs/ocfs2/export.c index 322216a5f0dd..29651167190d 100644 --- a/fs/ocfs2/export.c +++ b/fs/ocfs2/export.c @@ -195,11 +195,11 @@ static int ocfs2_encode_fh(struct inode *inode, u32 *fh_in, int *max_len, if (parent && (len < 6)) { *max_len = 6; - type = 255; + type = FILEID_INVALID; goto bail; } else if (len < 3) { *max_len = 3; - type = 255; + type = FILEID_INVALID; goto bail; } diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 95d7680ead47..ea5061fd4f3e 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -1603,10 +1603,10 @@ int reiserfs_encode_fh(struct inode *inode, __u32 * data, int *lenp, if (parent && (maxlen < 5)) { *lenp = 5; - return 255; + return FILEID_INVALID; } else if (maxlen < 3) { *lenp = 3; - return 255; + return FILEID_INVALID; } data[0] = inode->i_ino; diff --git a/fs/udf/namei.c b/fs/udf/namei.c index 95fee278ab9d..102c072c6bbf 100644 --- a/fs/udf/namei.c +++ b/fs/udf/namei.c @@ -1270,10 +1270,10 @@ static int udf_encode_fh(struct inode *inode, __u32 *fh, int *lenp, if (parent && (len < 5)) { *lenp = 5; - return 255; + return FILEID_INVALID; } else if (len < 3) { *lenp = 3; - return 255; + return FILEID_INVALID; } *lenp = 3; diff --git a/fs/xfs/xfs_export.c b/fs/xfs/xfs_export.c index a83611849cee..c585bc646395 100644 --- a/fs/xfs/xfs_export.c +++ b/fs/xfs/xfs_export.c @@ -48,7 +48,7 @@ static int xfs_fileid_length(int fileid_type) case FILEID_INO32_GEN_PARENT | XFS_FILEID_TYPE_64FLAG: return 6; } - return 255; /* invalid */ + return FILEID_INVALID; } STATIC int @@ -90,7 +90,7 @@ xfs_fs_encode_fh( len = xfs_fileid_length(fileid_type); if (*max_len < len) { *max_len = len; - return 255; + return FILEID_INVALID; } *max_len = len; -- cgit From 7bb307e894d51308aa0582a8c4cc5875bbc645b9 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 23 Feb 2013 14:51:48 -0500 Subject: export kernel_write(), convert open-coded instances Signed-off-by: Al Viro --- fs/ecryptfs/read_write.c | 6 +----- fs/splice.c | 5 +++-- 2 files changed, 4 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/read_write.c b/fs/ecryptfs/read_write.c index b2a34a192f4f..6a160539cd23 100644 --- a/fs/ecryptfs/read_write.c +++ b/fs/ecryptfs/read_write.c @@ -40,16 +40,12 @@ int ecryptfs_write_lower(struct inode *ecryptfs_inode, char *data, loff_t offset, size_t size) { struct file *lower_file; - mm_segment_t fs_save; ssize_t rc; lower_file = ecryptfs_inode_to_private(ecryptfs_inode)->lower_file; if (!lower_file) return -EIO; - fs_save = get_fs(); - set_fs(get_ds()); - rc = vfs_write(lower_file, data, size, &offset); - set_fs(fs_save); + rc = kernel_write(lower_file, data, size, offset); mark_inode_dirty_sync(ecryptfs_inode); return rc; } diff --git a/fs/splice.c b/fs/splice.c index 963213d56403..718bd0056384 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -569,7 +569,7 @@ static ssize_t kernel_readv(struct file *file, const struct iovec *vec, return res; } -static ssize_t kernel_write(struct file *file, const char *buf, size_t count, +ssize_t kernel_write(struct file *file, const char *buf, size_t count, loff_t pos) { mm_segment_t old_fs; @@ -578,11 +578,12 @@ static ssize_t kernel_write(struct file *file, const char *buf, size_t count, old_fs = get_fs(); set_fs(get_ds()); /* The cast to a user pointer is valid due to the set_fs() */ - res = vfs_write(file, (const char __user *)buf, count, &pos); + res = vfs_write(file, (__force const char __user *)buf, count, &pos); set_fs(old_fs); return res; } +EXPORT_SYMBOL(kernel_write); ssize_t default_file_splice_read(struct file *in, loff_t *ppos, struct pipe_inode_info *pipe, size_t len, -- cgit From 21d206819a6f111c77f0a9e920f5efca06dc65fd Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 23 Feb 2013 17:36:01 -0500 Subject: get_empty_filp()/alloc_file() leave both ->f_pos and ->f_version zero Signed-off-by: Al Viro --- fs/anon_inodes.c | 2 -- fs/open.c | 1 - 2 files changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c index a14eb0c1cd8c..47a65df8c871 100644 --- a/fs/anon_inodes.c +++ b/fs/anon_inodes.c @@ -164,9 +164,7 @@ struct file *anon_inode_getfile(const char *name, goto err_dput; file->f_mapping = anon_inode_inode->i_mapping; - file->f_pos = 0; file->f_flags = flags & (O_ACCMODE | O_NONBLOCK); - file->f_version = 0; file->private_data = priv; return file; diff --git a/fs/open.c b/fs/open.c index 97a237f67b72..62f907e3bc36 100644 --- a/fs/open.c +++ b/fs/open.c @@ -699,7 +699,6 @@ static int do_dentry_open(struct file *f, } f->f_mapping = inode->i_mapping; - f->f_pos = 0; file_sb_list_add(f, inode->i_sb); if (unlikely(f->f_mode & FMODE_PATH)) { -- cgit From 30b9c9e6ba289ba3bb67cc292efcc4122ea37ae5 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Fri, 3 Aug 2012 17:36:17 +0100 Subject: ocfs2: Fix oops in ocfs2_fast_symlink_readpage() code path Commit ea022dfb3c2a4680483b00eb2fecc9fc4f6091d1 was missing a var init. Reported-and-Tested-by: Vincent Etienne Signed-off-by: Sunil Mushran Signed-off-by: Joel Becker Signed-off-by: Al Viro --- fs/ocfs2/symlink.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/symlink.c b/fs/ocfs2/symlink.c index f1fbb4b552ad..66edce7ecfd7 100644 --- a/fs/ocfs2/symlink.c +++ b/fs/ocfs2/symlink.c @@ -57,7 +57,7 @@ static int ocfs2_fast_symlink_readpage(struct file *unused, struct page *page) { struct inode *inode = page->mapping->host; - struct buffer_head *bh; + struct buffer_head *bh = NULL; int status = ocfs2_read_inode_block(inode, &bh); struct ocfs2_dinode *fe; const char *link; -- cgit From 9b171e0c74ca0549d0610990a862dd895870f04a Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 20 Feb 2013 13:16:39 +1100 Subject: ocfs2: fix possible use-after-free with AIO Running AIO is pinning inode in memory using file reference. Once AIO is completed using aio_complete(), file reference is put and inode can be freed from memory. So we have to be sure that calling aio_complete() is the last thing we do with the inode. Signed-off-by: Jan Kara Acked-by: Jeff Moyer Acked-by: Joel Becker Cc: Mark Fasheh Cc: Al Viro Cc: Signed-off-by: Andrew Morton Signed-off-by: Al Viro --- fs/ocfs2/aops.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index db1ad26e02a7..50fe28b988c1 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -593,9 +593,9 @@ static void ocfs2_dio_end_io(struct kiocb *iocb, level = ocfs2_iocb_rw_locked_level(iocb); ocfs2_rw_unlock(inode, level); + inode_dio_done(inode); if (is_async) aio_complete(iocb, ret, 0); - inode_dio_done(inode); } /* -- cgit From b24ae0b54be4e46bb88398f29c5f79c5dad2a602 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Thu, 26 Jul 2012 16:05:05 +0300 Subject: ocfs2/dlm: use GFP_ATOMIC inside a spin_lock My static checker complains that this is called with a spin_lock held in dlm_master_requery_handler() from dlmrecovery.c. Probably the reason we have not received any bug reports about this is that recovery is not a common operation. Signed-off-by: Dan Carpenter Signed-off-by: Joel Becker Signed-off-by: Al Viro --- fs/ocfs2/dlm/dlmmaster.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 005261c333b0..33ecbe0e6734 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -2020,7 +2020,7 @@ int dlm_dispatch_assert_master(struct dlm_ctxt *dlm, int ignore_higher, u8 request_from, u32 flags) { struct dlm_work_item *item; - item = kzalloc(sizeof(*item), GFP_NOFS); + item = kzalloc(sizeof(*item), GFP_ATOMIC); if (!item) return -ENOMEM; -- cgit From 9cc64ceaa8b8e8c874519caee79e18cb35d3ce3e Mon Sep 17 00:00:00 2001 From: Yuanhan Liu Date: Wed, 20 Feb 2013 13:16:01 +1100 Subject: fs/exec.c: make bprm_mm_init() static There is only one user of bprm_mm_init, and it's inside the same file. Signed-off-by: Yuanhan Liu Signed-off-by: Andrew Morton Signed-off-by: Al Viro --- fs/exec.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/exec.c b/fs/exec.c index 7b6f4d59b26c..864c50df660a 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -355,7 +355,7 @@ static bool valid_arg_len(struct linux_binprm *bprm, long len) * flags, permissions, and offset, so we use temporary values. We'll update * them later in setup_arg_pages(). */ -int bprm_mm_init(struct linux_binprm *bprm) +static int bprm_mm_init(struct linux_binprm *bprm) { int err; struct mm_struct *mm = NULL; -- cgit From 41735818766c0ec215b9a69591e7eae642061954 Mon Sep 17 00:00:00 2001 From: Zhao Hongjiang Date: Wed, 20 Feb 2013 13:13:55 +1100 Subject: fs: change return values from -EACCES to -EPERM According to SUSv3: [EACCES] Permission denied. An attempt was made to access a file in a way forbidden by its file access permissions. [EPERM] Operation not permitted. An attempt was made to perform an operation limited to processes with appropriate privileges or to the owner of a file or other resource. So -EPERM should be returned if capability checks fails. Strictly speaking this is an API change since the error code user sees is altered. Signed-off-by: Zhao Hongjiang Acked-by: Jan Kara Acked-by: Steven Whitehouse Acked-by: Ian Kent Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Al Viro --- fs/autofs4/root.c | 2 +- fs/dlm/config.c | 2 +- fs/gfs2/sys.c | 18 +++++++++--------- fs/ncpfs/ioctl.c | 2 +- fs/proc/base.c | 6 +++--- fs/udf/file.c | 2 +- 6 files changed, 16 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index 3cdf835e8b49..230bd2aad4f4 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -587,7 +587,7 @@ static int autofs4_dir_unlink(struct inode *dir, struct dentry *dentry) /* This allows root to remove symlinks */ if (!autofs4_oz_mode(sbi) && !capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (atomic_dec_and_test(&ino->count)) { p_ino = autofs4_dentry_ino(dentry->d_parent); diff --git a/fs/dlm/config.c b/fs/dlm/config.c index a0387dd8b1f0..7d58d5b112b5 100644 --- a/fs/dlm/config.c +++ b/fs/dlm/config.c @@ -158,7 +158,7 @@ static ssize_t cluster_set(struct dlm_cluster *cl, unsigned int *cl_field, unsigned int x; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; x = simple_strtoul(buf, NULL, 0); diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 8056b7b7238e..0acbe2ff1e5d 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -107,7 +107,7 @@ static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len) int n = simple_strtol(buf, NULL, 0); if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; switch (n) { case 0: @@ -135,7 +135,7 @@ static ssize_t withdraw_show(struct gfs2_sbd *sdp, char *buf) static ssize_t withdraw_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (simple_strtol(buf, NULL, 0) != 1) return -EINVAL; @@ -150,7 +150,7 @@ static ssize_t statfs_sync_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (simple_strtol(buf, NULL, 0) != 1) return -EINVAL; @@ -163,7 +163,7 @@ static ssize_t quota_sync_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (simple_strtol(buf, NULL, 0) != 1) return -EINVAL; @@ -179,7 +179,7 @@ static ssize_t quota_refresh_user_store(struct gfs2_sbd *sdp, const char *buf, u32 id; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; id = simple_strtoul(buf, NULL, 0); @@ -194,7 +194,7 @@ static ssize_t quota_refresh_group_store(struct gfs2_sbd *sdp, const char *buf, u32 id; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; id = simple_strtoul(buf, NULL, 0); @@ -213,7 +213,7 @@ static ssize_t demote_rq_store(struct gfs2_sbd *sdp, const char *buf, size_t len int rv; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; rv = sscanf(buf, "%u:%llu %15s", &gltype, &glnum, mode); @@ -502,7 +502,7 @@ static ssize_t quota_scale_store(struct gfs2_sbd *sdp, const char *buf, unsigned int x, y; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; if (sscanf(buf, "%u %u", &x, &y) != 2 || !y) return -EINVAL; @@ -521,7 +521,7 @@ static ssize_t tune_set(struct gfs2_sbd *sdp, unsigned int *field, unsigned int x; if (!capable(CAP_SYS_ADMIN)) - return -EACCES; + return -EPERM; x = simple_strtoul(buf, NULL, 0); diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c index 5c1e9262219c..811d411418de 100644 --- a/fs/ncpfs/ioctl.c +++ b/fs/ncpfs/ioctl.c @@ -819,7 +819,7 @@ long ncp_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) case NCP_IOC_CONN_LOGGED_IN: case NCP_IOC_SETROOT: if (!capable(CAP_SYS_ADMIN)) { - ret = -EACCES; + ret = -EPERM; goto out; } break; diff --git a/fs/proc/base.c b/fs/proc/base.c index 9d9625559727..f3b133d79914 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -1711,7 +1711,7 @@ static int map_files_d_revalidate(struct dentry *dentry, unsigned int flags) return -ECHILD; if (!capable(CAP_SYS_ADMIN)) { - status = -EACCES; + status = -EPERM; goto out_notask; } @@ -1844,7 +1844,7 @@ static struct dentry *proc_map_files_lookup(struct inode *dir, struct dentry *result; struct mm_struct *mm; - result = ERR_PTR(-EACCES); + result = ERR_PTR(-EPERM); if (!capable(CAP_SYS_ADMIN)) goto out; @@ -1900,7 +1900,7 @@ proc_map_files_readdir(struct file *filp, void *dirent, filldir_t filldir) ino_t ino; int ret; - ret = -EACCES; + ret = -EPERM; if (!capable(CAP_SYS_ADMIN)) goto out; diff --git a/fs/udf/file.c b/fs/udf/file.c index 4257a1f5302a..29569dd08168 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -204,7 +204,7 @@ long udf_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) goto out; case UDF_RELOCATE_BLOCKS: if (!capable(CAP_SYS_ADMIN)) { - result = -EACCES; + result = -EPERM; goto out; } if (get_user(old_block, (long __user *)arg)) { -- cgit From 87e0aab37ff6c4284810a48d6034314fbf4eb319 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Wed, 20 Feb 2013 13:13:32 +1100 Subject: proc: avoid extra pde_put() in proc_fill_super() If proc_get_inode() succeeded, but d_make_root() failed, pde_put() for proc_root will be called twice: the first time due to iput() called from d_make_root() and the second time directly in the end of proc_fill_super(). Signed-off-by: Maxim Patlasov Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Al Viro --- fs/proc/inode.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 38f5c119b806..98a7d2870bef 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -486,6 +486,8 @@ struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) int proc_fill_super(struct super_block *s) { + struct inode *root_inode; + s->s_flags |= MS_NODIRATIME | MS_NOSUID | MS_NOEXEC; s->s_blocksize = 1024; s->s_blocksize_bits = 10; @@ -494,11 +496,18 @@ int proc_fill_super(struct super_block *s) s->s_time_gran = 1; pde_get(&proc_root); - s->s_root = d_make_root(proc_get_inode(s, &proc_root)); - if (s->s_root) - return 0; + root_inode = proc_get_inode(s, &proc_root); + if (!root_inode) { + printk(KERN_ERR "proc_fill_super: get root inode failed\n"); + pde_put(&proc_root); + return -ENOMEM; + } - printk("proc_read_super: get root inode failed\n"); - pde_put(&proc_root); - return -ENOMEM; + s->s_root = d_make_root(root_inode); + if (!s->s_root) { + printk(KERN_ERR "proc_fill_super: allocate dentry failed\n"); + return -ENOMEM; + } + + return 0; } -- cgit From d3d009cb965eae7e002ea5badf603ea8f4c34915 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 25 Jan 2013 20:11:22 -0500 Subject: saner proc_get_inode() calling conventions Make it drop the pde in *all* cases when no new reference to it is put into an inode - both when an inode had already been set up (as we were already doing) and when inode allocation has failed. Makes for simpler logics in callers... Signed-off-by: Al Viro --- fs/proc/generic.c | 21 +++++++-------------- fs/proc/inode.c | 10 +++------- 2 files changed, 10 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 7dfe548a28e8..2983dc52ca25 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -412,8 +412,7 @@ static const struct dentry_operations proc_dentry_operations = struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *dir, struct dentry *dentry) { - struct inode *inode = NULL; - int error = -ENOENT; + struct inode *inode; spin_lock(&proc_subdir_lock); for (de = de->subdir; de ; de = de->next) { @@ -422,22 +421,16 @@ struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *dir, if (!memcmp(dentry->d_name.name, de->name, de->namelen)) { pde_get(de); spin_unlock(&proc_subdir_lock); - error = -ENOMEM; inode = proc_get_inode(dir->i_sb, de); - goto out_unlock; + if (!inode) + return ERR_PTR(-ENOMEM); + d_set_d_op(dentry, &proc_dentry_operations); + d_add(dentry, inode); + return NULL; } } spin_unlock(&proc_subdir_lock); -out_unlock: - - if (inode) { - d_set_d_op(dentry, &proc_dentry_operations); - d_add(dentry, inode); - return NULL; - } - if (de) - pde_put(de); - return ERR_PTR(error); + return ERR_PTR(-ENOENT); } struct dentry *proc_lookup(struct inode *dir, struct dentry *dentry, diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 98a7d2870bef..70322e1a4f0f 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -445,12 +445,9 @@ static const struct file_operations proc_reg_file_ops_no_compat = { struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) { - struct inode * inode; + struct inode *inode = iget_locked(sb, de->low_ino); - inode = iget_locked(sb, de->low_ino); - if (!inode) - return NULL; - if (inode->i_state & I_NEW) { + if (inode && (inode->i_state & I_NEW)) { inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; PROC_I(inode)->pde = de; @@ -482,7 +479,7 @@ struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) } else pde_put(de); return inode; -} +} int proc_fill_super(struct super_block *s) { @@ -499,7 +496,6 @@ int proc_fill_super(struct super_block *s) root_inode = proc_get_inode(s, &proc_root); if (!root_inode) { printk(KERN_ERR "proc_fill_super: get root inode failed\n"); - pde_put(&proc_root); return -ENOMEM; } -- cgit From f2bdf9a8f79052e9577936cbe5696c0e232aa0e3 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 21 Feb 2013 15:28:28 -0500 Subject: Btrfs: make sure NODATACOW also gets NODATASUM set A user reported hitting the BUG_ON() in btrfs_finished_ordered_io() where we had csums on a NOCOW extent. This can happen if we have NODATACOW set but not NODATASUM set, which can happen in two cases, either we mount with -o nodatacow and then write into preallocated space, or chattr +C a directory and move a file into that directory. Liu has fixed the move case in a different place, but this fixes the mount -o nodatacow case. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 40d49da5e846..14c82cdbb696 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5574,7 +5574,8 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, if (btrfs_test_opt(root, NODATASUM)) BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM; if (btrfs_test_opt(root, NODATACOW)) - BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW; + BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW | + BTRFS_INODE_NODATASUM; } insert_inode_hash(inode); -- cgit From ca60ebfa303ca6d9f3cad5028c10f0a17d571a4c Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Thu, 21 Feb 2013 15:35:27 +0000 Subject: Btrfs: fix backref walking race with tree deletions When a subvolume is removed, we remove the root item from the root tree, while the tree blocks and backrefs remain for a while. When backref walking comes across one of those orphan tree blocks, it can find a backref for a no longer existing root. This is all good, we only must tolerate __resolve_indirect_ref returning an error and continue with the good refs found. Reported-by: Alex Lyakas Signed-off-by: Jan Schmidt Signed-off-by: Josef Bacik --- fs/btrfs/backref.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index 04edf69be875..bd605c87adfd 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -352,11 +352,8 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info, err = __resolve_indirect_ref(fs_info, search_commit_root, time_seq, ref, parents, extent_item_pos); - if (err) { - if (ret == 0) - ret = err; + if (err) continue; - } /* we put the first parent into the ref at hand */ ULIST_ITER_INIT(&uiter); -- cgit From a81cb9a2d939676dee6911993a37fa7e818ab72e Mon Sep 17 00:00:00 2001 From: Alexandre Oliva Date: Thu, 21 Feb 2013 21:15:14 +0000 Subject: clear chunk_alloc flag on retryable failure I've experienced filesystem freezes with permanent spikes in the active process count for quite a while, particularly on filesystems whose available raw space has already been fully allocated to chunks. While looking into this, I found a pretty obvious error in do_chunk_alloc: it sets space_info->chunk_alloc, but if btrfs_alloc_chunk returns an error other than ENOSPC, it returns leaving that flag set, which causes any other threads waiting for space_info->chunk_alloc to become zero to spin indefinitely. I haven't double-checked that this patch fixes the failure I've observed fully (it's not exactly trivial to trigger), but it surely is a bug and the fix is trivial, so... Please put it in :-) What I saw in that function also happens to explain why in some cases I see filesystems allocate a huge number of chunks that remain unused (leading to the scenario above, of not having more chunks to allocate). It happens for data and metadata, but not necessarily both. I'm guessing some thread sets the force_alloc flag on the corresponding space_info, and then several threads trying to get disk space end up attempting to allocate a new chunk concurrently. All of them will see the force_alloc flag and bump their local copy of force up to the level they see first, and they won't clear it even if another thread succeeds in allocating a chunk, thus clearing the force flag. Then each thread that observed the force flag will, on its turn, force the allocation of a new chunk. And any threads that come in while it does that will see the force flag still set and pick it up, and so on. This sounds like a problem to me, but... what should the correct behavior be? Clear force_flag once we copy it to a local force? Reset force to the incoming value on every loop? Set the flag to our incoming force if we have it at first, clear our local flag, and move it from the space_info when we determined that we are the thread that's going to perform the allocation? btrfs: clear chunk_alloc flag on retryable failure From: Alexandre Oliva If btrfs_alloc_chunk fails with e.g. ENOMEM, we exit do_chunk_alloc without clearing chunk_alloc in space_info. As a result, any further calls to do_chunk_alloc on that filesystem will start busy-waiting for chunk_alloc to be cleared, but it never will be. This patch adjusts do_chunk_alloc so that it clears this flag in case of an error. Signed-off-by: Alexandre Oliva Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index d2b3a5e9a621..8520354f086e 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3715,19 +3715,19 @@ again: ret = btrfs_alloc_chunk(trans, extent_root, flags); trans->allocating_chunk = false; - if (ret < 0 && ret != -ENOSPC) - goto out; spin_lock(&space_info->lock); + if (ret < 0 && ret != -ENOSPC) + goto out; if (ret) space_info->full = 1; else ret = 1; space_info->force_alloc = CHUNK_ALLOC_NO_FORCE; +out: space_info->chunk_alloc = 0; spin_unlock(&space_info->lock); -out: mutex_unlock(&fs_info->chunk_mutex); return ret; } -- cgit From 2382c5cc7ed0396b61a359765bf5ee125b0a2f46 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Fri, 22 Feb 2013 04:33:36 +0000 Subject: Btrfs: use reserved space for creating a snapshot While inserting dir index and updating inode for a snapshot, we'd add delayed items which consume trans->block_rsv, if we don't have any space reserved in this trans handle, we either just return or reserve space again. But before creating pending snapshots during committing transaction, we've done a release on this trans handle, so we don't have space reserved in it at this stage. What we're using is block_rsv of pending snapshots which has already reserved well enough space for both inserting dir index and updating inode, so we need to set trans handle to indicate that we have space now. Signed-off-by: Liu Bo Reviewed-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index a83d486cc70c..4330433b7b4f 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1123,6 +1123,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, rsv = trans->block_rsv; trans->block_rsv = &pending->block_rsv; + trans->bytes_reserved = trans->block_rsv->reserved; dentry = pending->dentry; parent = dget_parent(dentry); @@ -1276,6 +1277,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, fail: dput(parent); trans->block_rsv = rsv; + trans->bytes_reserved = 0; no_free_objectid: kfree(new_root_item); root_item_alloc_fail: -- cgit From 8c4ce81e911ab6c84e4f75e18d4ceb3fa555c35b Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 25 Feb 2013 04:04:42 +0000 Subject: Btrfs: do not change inode flags in rename Before we forced to change a file's NOCOW and COMPRESS flag due to the parent directory's, but this ends up a bad idea, because it confuses end users a lot about file's NOCOW status, eg. if someone change a file to NOCOW via 'chattr' and then rename it in the current directory which is without NOCOW attribute, the file will lose the NOCOW flag silently. This diables 'change flags in rename', so from now on we'll only inherit flags from the parent directory on creation stage while in other places we can use 'chattr' to set NOCOW or COMPRESS flags. Reported-by: Marios Titas Signed-off-by: Liu Bo Reviewed-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 25 ------------------------- 1 file changed, 25 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 14c82cdbb696..be09654e11b9 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -8060,29 +8060,6 @@ static int btrfs_getattr(struct vfsmount *mnt, return 0; } -/* - * If a file is moved, it will inherit the cow and compression flags of the new - * directory. - */ -static void fixup_inode_flags(struct inode *dir, struct inode *inode) -{ - struct btrfs_inode *b_dir = BTRFS_I(dir); - struct btrfs_inode *b_inode = BTRFS_I(inode); - - if (b_dir->flags & BTRFS_INODE_NODATACOW) - b_inode->flags |= BTRFS_INODE_NODATACOW; - else - b_inode->flags &= ~BTRFS_INODE_NODATACOW; - - if (b_dir->flags & BTRFS_INODE_COMPRESS) { - b_inode->flags |= BTRFS_INODE_COMPRESS; - b_inode->flags &= ~BTRFS_INODE_NOCOMPRESS; - } else { - b_inode->flags &= ~(BTRFS_INODE_COMPRESS | - BTRFS_INODE_NOCOMPRESS); - } -} - static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { @@ -8248,8 +8225,6 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, } } - fixup_inode_flags(new_dir, old_inode); - ret = btrfs_add_link(trans, new_dir, old_inode, new_dentry->d_name.name, new_dentry->d_name.len, 0, index); -- cgit From fda2832febb1928da0625b2c5d15559b29d7e740 Mon Sep 17 00:00:00 2001 From: Qu Wenruo Date: Tue, 26 Feb 2013 08:10:22 +0000 Subject: btrfs: cleanup for open-coded alignment Though most of the btrfs codes are using ALIGN macro for page alignment, there are still some codes using open-coded alignment like the following: ------ u64 mask = ((u64)root->stripesize - 1); u64 ret = (val + mask) & ~mask; ------ Or even hidden one: ------ num_bytes = (end - start + blocksize) & ~(blocksize - 1); ------ Sometimes these open-coded alignment is not so easy to understand for newbie like me. This commit changes the open-coded alignment to the ALIGN macro for a better readability. Also there is a previous patch from David Sterba with similar changes, but the patch is for 3.2 kernel and seems not merged. http://www.spinics.net/lists/linux-btrfs/msg12747.html Cc: David Sterba Signed-off-by: Qu Wenruo Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 9 +++------ fs/btrfs/extent_io.c | 8 ++++---- fs/btrfs/file.c | 3 +-- fs/btrfs/inode.c | 37 +++++++++++++++---------------------- fs/btrfs/tree-log.c | 3 +-- fs/btrfs/volumes.c | 3 +-- 6 files changed, 25 insertions(+), 38 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 8520354f086e..5681a91ed400 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3431,7 +3431,7 @@ int btrfs_check_data_free_space(struct inode *inode, u64 bytes) int ret = 0, committed = 0, alloc_chunk = 1; /* make sure bytes are sectorsize aligned */ - bytes = (bytes + root->sectorsize - 1) & ~((u64)root->sectorsize - 1); + bytes = ALIGN(bytes, root->sectorsize); if (root == root->fs_info->tree_root || BTRFS_I(inode)->location.objectid == BTRFS_FREE_INO_OBJECTID) { @@ -3526,7 +3526,7 @@ void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes) struct btrfs_space_info *data_sinfo; /* make sure bytes are sectorsize aligned */ - bytes = (bytes + root->sectorsize - 1) & ~((u64)root->sectorsize - 1); + bytes = ALIGN(bytes, root->sectorsize); data_sinfo = root->fs_info->data_sinfo; spin_lock(&data_sinfo->lock); @@ -5607,10 +5607,7 @@ static u64 stripe_align(struct btrfs_root *root, struct btrfs_block_group_cache *cache, u64 val, u64 num_bytes) { - u64 mask; - u64 ret; - mask = ((u64)root->stripesize - 1); - ret = (val + mask) & ~mask; + u64 ret = ALIGN(val, root->stripesize); return ret; } diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 66f999b97cbb..597ab8966c80 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -2686,7 +2686,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree, iosize = min(extent_map_end(em) - cur, end - cur + 1); cur_end = min(extent_map_end(em) - 1, end); - iosize = (iosize + blocksize - 1) & ~((u64)blocksize - 1); + iosize = ALIGN(iosize, blocksize); if (this_bio_flag & EXTENT_BIO_COMPRESSED) { disk_io_size = em->block_len; sector = em->block_start >> 9; @@ -2977,7 +2977,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, BUG_ON(extent_map_end(em) <= cur); BUG_ON(end < cur); iosize = min(extent_map_end(em) - cur, end - cur + 1); - iosize = (iosize + blocksize - 1) & ~((u64)blocksize - 1); + iosize = ALIGN(iosize, blocksize); sector = (em->block_start + extent_offset) >> 9; bdev = em->bdev; block_start = em->block_start; @@ -3664,7 +3664,7 @@ int extent_invalidatepage(struct extent_io_tree *tree, u64 end = start + PAGE_CACHE_SIZE - 1; size_t blocksize = page->mapping->host->i_sb->s_blocksize; - start += (offset + blocksize - 1) & ~(blocksize - 1); + start += ALIGN(offset, blocksize); if (start > end) return 0; @@ -3783,7 +3783,7 @@ static struct extent_map *get_extent_skip_holes(struct inode *inode, len = last - offset; if (len == 0) break; - len = (len + sectorsize - 1) & ~(sectorsize - 1); + len = ALIGN(len, sectorsize); em = get_extent(inode, NULL, 0, offset, len, 0); if (IS_ERR_OR_NULL(em)) return em; diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 6e6dd8cdad92..83c790d84038 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -510,8 +510,7 @@ int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode, loff_t isize = i_size_read(inode); start_pos = pos & ~((u64)root->sectorsize - 1); - num_bytes = (write_bytes + pos - start_pos + - root->sectorsize - 1) & ~((u64)root->sectorsize - 1); + num_bytes = ALIGN(write_bytes + pos - start_pos, root->sectorsize); end_of_last_block = start_pos + num_bytes - 1; err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block, diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index be09654e11b9..9ef7a5b1b77e 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -233,8 +233,7 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans, u64 isize = i_size_read(inode); u64 actual_end = min(end + 1, isize); u64 inline_len = actual_end - start; - u64 aligned_end = (end + root->sectorsize - 1) & - ~((u64)root->sectorsize - 1); + u64 aligned_end = ALIGN(end, root->sectorsize); u64 data_len = inline_len; int ret; @@ -391,7 +390,7 @@ again: * a compressed extent to 128k. */ total_compressed = min(total_compressed, max_uncompressed); - num_bytes = (end - start + blocksize) & ~(blocksize - 1); + num_bytes = ALIGN(end - start + 1, blocksize); num_bytes = max(blocksize, num_bytes); total_in = 0; ret = 0; @@ -490,15 +489,13 @@ cont: * up to a block size boundary so the allocator does sane * things */ - total_compressed = (total_compressed + blocksize - 1) & - ~(blocksize - 1); + total_compressed = ALIGN(total_compressed, blocksize); /* * one last check to make sure the compression is really a * win, compare the page count read with the blocks on disk */ - total_in = (total_in + PAGE_CACHE_SIZE - 1) & - ~(PAGE_CACHE_SIZE - 1); + total_in = ALIGN(total_in, PAGE_CACHE_SIZE); if (total_compressed >= total_in) { will_compress = 0; } else { @@ -856,7 +853,7 @@ static noinline int __cow_file_range(struct btrfs_trans_handle *trans, BUG_ON(btrfs_is_free_space_inode(inode)); - num_bytes = (end - start + blocksize) & ~(blocksize - 1); + num_bytes = ALIGN(end - start + 1, blocksize); num_bytes = max(blocksize, num_bytes); disk_num_bytes = num_bytes; @@ -4015,7 +4012,6 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans, u64 extent_num_bytes = 0; u64 extent_offset = 0; u64 item_end = 0; - u64 mask = root->sectorsize - 1; u32 found_type = (u8)-1; int found_extent; int del_item; @@ -4039,7 +4035,8 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans, * extent just the way it is. */ if (root->ref_cows || root == root->fs_info->tree_root) - btrfs_drop_extent_cache(inode, (new_size + mask) & (~mask), (u64)-1, 0); + btrfs_drop_extent_cache(inode, ALIGN(new_size, + root->sectorsize), (u64)-1, 0); /* * This function is also used to drop the items in the log tree before @@ -4118,10 +4115,9 @@ search_again: if (!del_item) { u64 orig_num_bytes = btrfs_file_extent_num_bytes(leaf, fi); - extent_num_bytes = new_size - - found_key.offset + root->sectorsize - 1; - extent_num_bytes = extent_num_bytes & - ~((u64)root->sectorsize - 1); + extent_num_bytes = ALIGN(new_size - + found_key.offset, + root->sectorsize); btrfs_set_file_extent_num_bytes(leaf, fi, extent_num_bytes); num_dec = (orig_num_bytes - @@ -4357,9 +4353,8 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size) struct extent_map *em = NULL; struct extent_state *cached_state = NULL; struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; - u64 mask = root->sectorsize - 1; - u64 hole_start = (oldsize + mask) & ~mask; - u64 block_end = (size + mask) & ~mask; + u64 hole_start = ALIGN(oldsize, root->sectorsize); + u64 block_end = ALIGN(size, root->sectorsize); u64 last_byte; u64 cur_offset; u64 hole_size; @@ -4392,7 +4387,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size) break; } last_byte = min(extent_map_end(em), block_end); - last_byte = (last_byte + mask) & ~mask; + last_byte = ALIGN(last_byte , root->sectorsize); if (!test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) { struct extent_map *hole_em; hole_size = last_byte - cur_offset; @@ -6111,8 +6106,7 @@ again: } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { size_t size; size = btrfs_file_extent_inline_len(leaf, item); - extent_end = (extent_start + size + root->sectorsize - 1) & - ~((u64)root->sectorsize - 1); + extent_end = ALIGN(extent_start + size, root->sectorsize); } if (start >= extent_end) { @@ -6184,8 +6178,7 @@ again: copy_size = min_t(u64, PAGE_CACHE_SIZE - pg_offset, size - extent_offset); em->start = extent_start + extent_offset; - em->len = (copy_size + root->sectorsize - 1) & - ~((u64)root->sectorsize - 1); + em->len = ALIGN(copy_size, root->sectorsize); em->orig_block_len = em->len; em->orig_start = em->start; if (compress_type) { diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 1a79087c4575..e8b7a68e1b37 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -484,7 +484,6 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, struct btrfs_key *key) { int found_type; - u64 mask = root->sectorsize - 1; u64 extent_end; u64 start = key->offset; u64 saved_nbytes; @@ -501,7 +500,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, extent_end = start + btrfs_file_extent_num_bytes(eb, item); else if (found_type == BTRFS_FILE_EXTENT_INLINE) { size = btrfs_file_extent_inline_len(eb, item); - extent_end = (start + size + mask) & ~mask; + extent_end = ALIGN(start + size, root->sectorsize); } else { ret = 0; goto out; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 538c5cfa005f..db72e0cc6f87 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -4556,8 +4556,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, num_stripes = 1; stripe_index = 0; stripe_nr_orig = stripe_nr; - stripe_nr_end = (offset + *length + map->stripe_len - 1) & - (~(map->stripe_len - 1)); + stripe_nr_end = ALIGN(offset + *length, map->stripe_len); do_div(stripe_nr_end, map->stripe_len); stripe_end_offset = stripe_nr_end * map->stripe_len - (offset + *length); -- cgit From 12979354a1d6ef25d86f381e4d5f9e103f29913a Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 8 Jan 2013 09:15:10 -0800 Subject: libceph: rename ceph_pg -> ceph_pg_v1 Rename the old version this type to distinguish it from the new version. Signed-off-by: Sage Weil Reviewed-by: Alex Elder --- fs/ceph/ioctl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ceph/ioctl.c b/fs/ceph/ioctl.c index 3b22150d3e19..e831436d6e68 100644 --- a/fs/ceph/ioctl.c +++ b/fs/ceph/ioctl.c @@ -186,7 +186,7 @@ static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) u64 len = 1, olen; u64 tmp; struct ceph_object_layout ol; - struct ceph_pg pgid; + struct ceph_pg_v1 pgid; int r; /* copy and validate */ -- cgit From 5b191d9914eb68257f47de9d5bfe099b77f0687c Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sat, 23 Feb 2013 10:38:16 -0800 Subject: libceph: decode into cpu-native ceph_pg type Always decode data into our cpu-native ceph_pg type that has the correct field widths. Limit any remaining uses of ceph_pg_v1 to dealing with the legacy protocol. Signed-off-by: Sage Weil Reviewed-by: Alex Elder --- fs/ceph/ioctl.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ceph/ioctl.c b/fs/ceph/ioctl.c index e831436d6e68..fb036ed3e129 100644 --- a/fs/ceph/ioctl.c +++ b/fs/ceph/ioctl.c @@ -186,7 +186,7 @@ static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) u64 len = 1, olen; u64 tmp; struct ceph_object_layout ol; - struct ceph_pg_v1 pgid; + struct ceph_pg pgid; int r; /* copy and validate */ @@ -212,7 +212,8 @@ static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) ceph_calc_object_layout(&ol, dl.object_name, &ci->i_layout, osdc->osdmap); - pgid = ol.ol_pgid; + pgid.pool = le32_to_cpu(ol.ol_pgid.pool); + pgid.seed = le16_to_cpu(ol.ol_pgid.ps); dl.osd = ceph_calc_pg_primary(osdc->osdmap, pgid); if (dl.osd >= 0) { struct ceph_entity_addr *a = -- cgit From 4f6a7e5ee1393ec4b243b39dac9f36992d161540 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Sat, 23 Feb 2013 10:41:09 -0800 Subject: ceph: update support for PGID64, PGPOOL3, OSDENC protocol features Support (and require) the PGID64, PGPOOL3, and OSDENC protocol features. These have been present in ceph.git since v0.42, Feb 2012. Require these features to simplify support; nobody is running older userspace. Note that the new request and reply encoding is still not in place, so the new code is not yet functional. Signed-off-by: Sage Weil Reviewed-by: Alex Elder --- fs/ceph/mdsmap.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ceph/mdsmap.c b/fs/ceph/mdsmap.c index 73b7d44e8a35..0d3c9240c61b 100644 --- a/fs/ceph/mdsmap.c +++ b/fs/ceph/mdsmap.c @@ -59,6 +59,10 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) return ERR_PTR(-ENOMEM); ceph_decode_16_safe(p, end, version, bad); + if (version > 3) { + pr_warning("got mdsmap version %d > 3, failing", version); + goto bad; + } ceph_decode_need(p, end, 8*sizeof(u32) + sizeof(u64), bad); m->m_epoch = ceph_decode_32(p); @@ -144,13 +148,13 @@ struct ceph_mdsmap *ceph_mdsmap_decode(void **p, void *end) /* pg_pools */ ceph_decode_32_safe(p, end, n, bad); m->m_num_data_pg_pools = n; - m->m_data_pg_pools = kcalloc(n, sizeof(u32), GFP_NOFS); + m->m_data_pg_pools = kcalloc(n, sizeof(u64), GFP_NOFS); if (!m->m_data_pg_pools) goto badmem; - ceph_decode_need(p, end, sizeof(u32)*(n+1), bad); + ceph_decode_need(p, end, sizeof(u64)*(n+1), bad); for (i = 0; i < n; i++) - m->m_data_pg_pools[i] = ceph_decode_32(p); - m->m_cas_pg_pool = ceph_decode_32(p); + m->m_data_pg_pools[i] = ceph_decode_64(p); + m->m_cas_pg_pool = ceph_decode_64(p); /* ok, we don't care about the rest. */ dout("mdsmap_decode success epoch %u\n", m->m_epoch); -- cgit From 2169aea649c08374bec7d220a3b8f64712275356 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Mon, 25 Feb 2013 16:13:08 -0800 Subject: libceph: calculate placement based on the internal data types Instead of using the old ceph_object_layout struct, update our internal ceph_calc_object_layout method to use the ceph_pg type. This allows us to pass the full 32-bit precision of the pgid.seed to the callers. It also allows some callers to avoid reaching into the request structures for the struct ceph_object_layout fields. Signed-off-by: Sage Weil Reviewed-by: Alex Elder --- fs/ceph/ioctl.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ceph/ioctl.c b/fs/ceph/ioctl.c index fb036ed3e129..7d85991fd647 100644 --- a/fs/ceph/ioctl.c +++ b/fs/ceph/ioctl.c @@ -185,7 +185,6 @@ static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) &ceph_sb_to_client(inode->i_sb)->client->osdc; u64 len = 1, olen; u64 tmp; - struct ceph_object_layout ol; struct ceph_pg pgid; int r; @@ -209,11 +208,9 @@ static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg) snprintf(dl.object_name, sizeof(dl.object_name), "%llx.%08llx", ceph_ino(inode), dl.object_no); - ceph_calc_object_layout(&ol, dl.object_name, &ci->i_layout, + ceph_calc_object_layout(&pgid, dl.object_name, &ci->i_layout, osdc->osdmap); - pgid.pool = le32_to_cpu(ol.ol_pgid.pool); - pgid.seed = le16_to_cpu(ol.ol_pgid.ps); dl.osd = ceph_calc_pg_primary(osdc->osdmap, pgid); if (dl.osd >= 0) { struct ceph_entity_addr *a = -- cgit From 1b83bef24c6746a146d39915a18fb5425f2facb0 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Mon, 25 Feb 2013 16:11:12 -0800 Subject: libceph: update osd request/reply encoding Use the new version of the encoding for osd requests and replies. In the process, update the way we are tracking request ops and reply lengths and results in the struct ceph_osd_request. Update the rbd and fs/ceph users appropriately. The main changes are: - we keep pointers into the request memory for fields we need to update each time the request is sent out over the wire - we keep information about the result in an array in the request struct where the users can easily get at it. Signed-off-by: Sage Weil Reviewed-by: Alex Elder --- fs/ceph/addr.c | 31 ++++++------------------------- 1 file changed, 6 insertions(+), 25 deletions(-) (limited to 'fs') diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index fc613715af46..cfef3e01a9b3 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -236,16 +236,10 @@ static int ceph_readpage(struct file *filp, struct page *page) static void finish_read(struct ceph_osd_request *req, struct ceph_msg *msg) { struct inode *inode = req->r_inode; - struct ceph_osd_reply_head *replyhead; - int rc, bytes; + int rc = req->r_result; + int bytes = le32_to_cpu(msg->hdr.data_len); int i; - /* parse reply */ - replyhead = msg->front.iov_base; - WARN_ON(le32_to_cpu(replyhead->num_ops) == 0); - rc = le32_to_cpu(replyhead->result); - bytes = le32_to_cpu(msg->hdr.data_len); - dout("finish_read %p req %p rc %d bytes %d\n", inode, req, rc, bytes); /* unlock all pages, zeroing any data we didn't read */ @@ -553,27 +547,18 @@ static void writepages_finish(struct ceph_osd_request *req, struct ceph_msg *msg) { struct inode *inode = req->r_inode; - struct ceph_osd_reply_head *replyhead; - struct ceph_osd_op *op; struct ceph_inode_info *ci = ceph_inode(inode); unsigned wrote; struct page *page; int i; struct ceph_snap_context *snapc = req->r_snapc; struct address_space *mapping = inode->i_mapping; - __s32 rc = -EIO; - u64 bytes = 0; + int rc = req->r_result; + u64 bytes = le64_to_cpu(req->r_request_ops[0].extent.length); struct ceph_fs_client *fsc = ceph_inode_to_client(inode); long writeback_stat; unsigned issued = ceph_caps_issued(ci); - /* parse reply */ - replyhead = msg->front.iov_base; - WARN_ON(le32_to_cpu(replyhead->num_ops) == 0); - op = (void *)(replyhead + 1); - rc = le32_to_cpu(replyhead->result); - bytes = le64_to_cpu(op->extent.length); - if (rc >= 0) { /* * Assume we wrote the pages we originally sent. The @@ -740,8 +725,6 @@ retry: struct page *page; int want; u64 offset, len; - struct ceph_osd_request_head *reqhead; - struct ceph_osd_op *op; long writeback_stat; next = 0; @@ -905,10 +888,8 @@ get_more_pages: /* revise final length, page count */ req->r_num_pages = locked_pages; - reqhead = req->r_request->front.iov_base; - op = (void *)(reqhead + 1); - op->extent.length = cpu_to_le64(len); - op->payload_len = cpu_to_le32(len); + req->r_request_ops[0].extent.length = cpu_to_le64(len); + req->r_request_ops[0].payload_len = cpu_to_le32(len); req->r_request->hdr.data_len = cpu_to_le32(len); rc = ceph_osdc_start_request(&fsc->client->osdc, req, true); -- cgit From 1111eae90fb64a9d9ed133e410712f1e34fdce4a Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Wed, 27 Feb 2013 11:37:48 -0800 Subject: eCryptfs: Fix redundant error check on ecryptfs_find_daemon_by_euid() It is sufficient to check the return code of ecryptfs_find_daemon_by_euid(). If it returns 0, it always sets the daemon pointer to point to a valid ecryptfs_daemon. Signed-off-by: Tyler Hicks Reported-by: Kees Cook --- fs/ecryptfs/messaging.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ecryptfs/messaging.c b/fs/ecryptfs/messaging.c index d5c7297c5816..474051bd8d92 100644 --- a/fs/ecryptfs/messaging.c +++ b/fs/ecryptfs/messaging.c @@ -283,7 +283,7 @@ ecryptfs_send_message_locked(char *data, int data_len, u8 msg_type, int rc; rc = ecryptfs_find_daemon_by_euid(&daemon); - if (rc || !daemon) { + if (rc) { rc = -ENOTCONN; goto out; } -- cgit From 8e919d13048cd5acaadb2b15b48acbfb8832d3c2 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 27 Feb 2013 14:54:37 -0500 Subject: ext4: fix extent status tree regression for file systems > 512GB This fixes a regression introduced by commit f7fec032aa782. The problem was that the extents status flags caused us to mask out block numbers smaller than 2**28 blocks. Since we didn't test with file systems smaller than 512GB, we didn't notice this during the development cycle. A typical failure looks like this: EXT4-fs error (device sdb1): htree_dirblock_to_tree:919: inode #172235804: block 152052301: comm ls: bad entry in directory: rec_len is smaller than minimal - offset=0(0), inode=0, rec_len=0, name_len=0 ... where 'debugfs -R "stat <172235804>" /dev/sdb1' reports that the inode has block number 688923213. When viewed in hex, block number 152052301 (from the syslog) is 0x910224D, while block number 688923213 is 0x2910224D. Note the missing "0x20000000" in the block number. Reported-by: Markus Trippelsdorf Verified-by: Markus Trippelsdorf Reported-by: Dave Jones Verified-by: Dave Jones Cc: Zheng Liu Signed-off-by: "Theodore Ts'o" --- fs/ext4/extents_status.h | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index cf83e77b16cb..f190dfe969da 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -20,10 +20,13 @@ #define es_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) #endif -#define EXTENT_STATUS_WRITTEN 0x80000000 /* written extent */ -#define EXTENT_STATUS_UNWRITTEN 0x40000000 /* unwritten extent */ -#define EXTENT_STATUS_DELAYED 0x20000000 /* delayed extent */ -#define EXTENT_STATUS_HOLE 0x10000000 /* hole */ +/* + * These flags live in the high bits of extent_status.es_pblk + */ +#define EXTENT_STATUS_WRITTEN (1ULL << 63) +#define EXTENT_STATUS_UNWRITTEN (1ULL << 62) +#define EXTENT_STATUS_DELAYED (1ULL << 61) +#define EXTENT_STATUS_HOLE (1ULL << 60) #define EXTENT_STATUS_FLAGS (EXTENT_STATUS_WRITTEN | \ EXTENT_STATUS_UNWRITTEN | \ @@ -58,22 +61,22 @@ extern int ext4_es_lookup_extent(struct inode *inode, ext4_lblk_t lblk, static inline int ext4_es_is_written(struct extent_status *es) { - return (es->es_pblk & EXTENT_STATUS_WRITTEN); + return (es->es_pblk & EXTENT_STATUS_WRITTEN) != 0; } static inline int ext4_es_is_unwritten(struct extent_status *es) { - return (es->es_pblk & EXTENT_STATUS_UNWRITTEN); + return (es->es_pblk & EXTENT_STATUS_UNWRITTEN) != 0; } static inline int ext4_es_is_delayed(struct extent_status *es) { - return (es->es_pblk & EXTENT_STATUS_DELAYED); + return (es->es_pblk & EXTENT_STATUS_DELAYED) != 0; } static inline int ext4_es_is_hole(struct extent_status *es) { - return (es->es_pblk & EXTENT_STATUS_HOLE); + return (es->es_pblk & EXTENT_STATUS_HOLE) != 0; } static inline ext4_fsblk_t ext4_es_status(struct extent_status *es) -- cgit From 6131ffaa1f091415b7a24abb01f033d9c0a727f4 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 27 Feb 2013 16:59:05 -0500 Subject: more file_inode() open-coded instances Signed-off-by: Al Viro --- fs/ext4/indirect.c | 2 +- fs/f2fs/file.c | 6 +++--- fs/fuse/dev.c | 2 +- fs/fuse/file.c | 24 ++++++++++++------------ fs/seq_file.c | 2 +- 5 files changed, 18 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index c541ab8b64dd..b505a145a593 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -1606,7 +1606,7 @@ err: int ext4_ind_punch_hole(struct file *file, loff_t offset, loff_t length) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct super_block *sb = inode->i_sb; ext4_lblk_t first_block, stop_block; struct address_space *mapping = inode->i_mapping; diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index b7a053d4c6d3..958a46da19ae 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -29,7 +29,7 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { struct page *page = vmf->page; - struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct inode *inode = file_inode(vma->vm_file); struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb); block_t old_blk_addr; struct dnode_of_data dn; @@ -544,7 +544,7 @@ static int expand_inode_data(struct inode *inode, loff_t offset, static long f2fs_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); long ret; if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) @@ -577,7 +577,7 @@ static inline __u32 f2fs_mask_flags(umode_t mode, __u32 flags) long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { - struct inode *inode = filp->f_dentry->d_inode; + struct inode *inode = file_inode(filp); struct f2fs_inode_info *fi = F2FS_I(inode); unsigned int flags; int ret; diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index e9bdec0b16d9..11dfa0c3fb46 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -532,7 +532,7 @@ void fuse_request_send_background_locked(struct fuse_conn *fc, void fuse_force_forget(struct file *file, u64 nodeid) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_req *req; struct fuse_forget_in inarg; diff --git a/fs/fuse/file.c b/fs/fuse/file.c index c8071768b950..34b80ba95bad 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -355,7 +355,7 @@ static int fuse_wait_on_page_writeback(struct inode *inode, pgoff_t index) static int fuse_flush(struct file *file, fl_owner_t id) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_file *ff = file->private_data; struct fuse_req *req; @@ -1215,7 +1215,7 @@ static ssize_t __fuse_direct_read(struct file *file, const struct iovec *iov, unsigned long nr_segs, loff_t *ppos) { ssize_t res; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); if (is_bad_inode(inode)) return -EIO; @@ -1238,7 +1238,7 @@ static ssize_t fuse_direct_read(struct file *file, char __user *buf, static ssize_t __fuse_direct_write(struct file *file, const struct iovec *iov, unsigned long nr_segs, loff_t *ppos) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); size_t count = iov_length(iov, nr_segs); ssize_t res; @@ -1258,7 +1258,7 @@ static ssize_t fuse_direct_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { struct iovec iov = { .iov_base = (void __user *)buf, .iov_len = count }; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); ssize_t res; if (is_bad_inode(inode)) @@ -1485,7 +1485,7 @@ static const struct vm_operations_struct fuse_file_vm_ops = { static int fuse_file_mmap(struct file *file, struct vm_area_struct *vma) { if ((vma->vm_flags & VM_SHARED) && (vma->vm_flags & VM_MAYWRITE)) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_inode *fi = get_fuse_inode(inode); struct fuse_file *ff = file->private_data; @@ -1543,7 +1543,7 @@ static void fuse_lk_fill(struct fuse_req *req, struct file *file, const struct file_lock *fl, int opcode, pid_t pid, int flock) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_file *ff = file->private_data; struct fuse_lk_in *arg = &req->misc.lk_in; @@ -1565,7 +1565,7 @@ static void fuse_lk_fill(struct fuse_req *req, struct file *file, static int fuse_getlk(struct file *file, struct file_lock *fl) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_req *req; struct fuse_lk_out outarg; @@ -1590,7 +1590,7 @@ static int fuse_getlk(struct file *file, struct file_lock *fl) static int fuse_setlk(struct file *file, struct file_lock *fl, int flock) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_req *req; int opcode = (fl->fl_flags & FL_SLEEP) ? FUSE_SETLKW : FUSE_SETLK; @@ -1622,7 +1622,7 @@ static int fuse_setlk(struct file *file, struct file_lock *fl, int flock) static int fuse_file_lock(struct file *file, int cmd, struct file_lock *fl) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); int err; @@ -1645,7 +1645,7 @@ static int fuse_file_lock(struct file *file, int cmd, struct file_lock *fl) static int fuse_file_flock(struct file *file, int cmd, struct file_lock *fl) { - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); int err; @@ -1702,7 +1702,7 @@ static sector_t fuse_bmap(struct address_space *mapping, sector_t block) static loff_t fuse_file_llseek(struct file *file, loff_t offset, int whence) { loff_t retval; - struct inode *inode = file->f_path.dentry->d_inode; + struct inode *inode = file_inode(file); /* No i_mutex protection necessary for SEEK_CUR and SEEK_SET */ if (whence == SEEK_CUR || whence == SEEK_SET) @@ -2079,7 +2079,7 @@ EXPORT_SYMBOL_GPL(fuse_do_ioctl); long fuse_ioctl_common(struct file *file, unsigned int cmd, unsigned long arg, unsigned int flags) { - struct inode *inode = file->f_dentry->d_inode; + struct inode *inode = file_inode(file); struct fuse_conn *fc = get_fuse_conn(inode); if (!fuse_allow_current_process(fc)) diff --git a/fs/seq_file.c b/fs/seq_file.c index f2bc3dfd0b88..f9538ea2a756 100644 --- a/fs/seq_file.c +++ b/fs/seq_file.c @@ -339,7 +339,7 @@ EXPORT_SYMBOL(seq_lseek); /** * seq_release - free the structures associated with sequential file. * @file: file in question - * @inode: file->f_path.dentry->d_inode + * @inode: its inode * * Frees the structures associated with sequential file; can be used * as ->f_op->release() if you don't have private data to destroy. -- cgit From ce2ac52105aa663056dfc17966ebed1bf93e6e64 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 1 Feb 2013 15:11:01 -0500 Subject: cifs: ensure that cifs_get_root() only traverses directories Kjell Braden reported this oops: [ 833.211970] BUG: unable to handle kernel NULL pointer dereference at (null) [ 833.212816] IP: [< (null)>] (null) [ 833.213280] PGD 1b9b2067 PUD e9f7067 PMD 0 [ 833.213874] Oops: 0010 [#1] SMP [ 833.214344] CPU 0 [ 833.214458] Modules linked in: des_generic md4 nls_utf8 cifs vboxvideo drm snd_intel8x0 snd_ac97_codec ac97_bus snd_pcm snd_seq_midi snd_rawmidi snd_seq_midi_event snd_seq bnep rfcomm snd_timer bluetooth snd_seq_device ppdev snd vboxguest parport_pc joydev mac_hid soundcore snd_page_alloc psmouse i2c_piix4 serio_raw lp parport usbhid hid e1000 [ 833.215629] [ 833.215629] Pid: 1752, comm: mount.cifs Not tainted 3.0.0-rc7-bisectcifs-fec11dd9a0+ #18 innotek GmbH VirtualBox/VirtualBox [ 833.215629] RIP: 0010:[<0000000000000000>] [< (null)>] (null) [ 833.215629] RSP: 0018:ffff8800119c9c50 EFLAGS: 00010282 [ 833.215629] RAX: ffffffffa02186c0 RBX: ffff88000c427780 RCX: 0000000000000000 [ 833.215629] RDX: 0000000000000000 RSI: ffff88000c427780 RDI: ffff88000c4362e8 [ 833.215629] RBP: ffff8800119c9c88 R08: ffff88001fc15e30 R09: 00000000d69515c7 [ 833.215629] R10: ffffffffa0201972 R11: ffff88000e8f6a28 R12: ffff88000c4362e8 [ 833.215629] R13: 0000000000000000 R14: 0000000000000000 R15: ffff88001181aaa6 [ 833.215629] FS: 00007f2986171700(0000) GS:ffff88001fc00000(0000) knlGS:0000000000000000 [ 833.215629] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 833.215629] CR2: 0000000000000000 CR3: 000000001b982000 CR4: 00000000000006f0 [ 833.215629] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 833.215629] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 833.215629] Process mount.cifs (pid: 1752, threadinfo ffff8800119c8000, task ffff88001c1c16f0) [ 833.215629] Stack: [ 833.215629] ffffffff8116a9b5 ffff8800119c9c88 ffffffff81178075 0000000000000286 [ 833.215629] 0000000000000000 ffff88000c4276c0 ffff8800119c9ce8 ffff8800119c9cc8 [ 833.215629] ffffffff8116b06e ffff88001bc6fc00 ffff88000c4276c0 ffff88000c4276c0 [ 833.215629] Call Trace: [ 833.215629] [] ? d_alloc_and_lookup+0x45/0x90 [ 833.215629] [] ? d_lookup+0x35/0x60 [ 833.215629] [] __lookup_hash.part.14+0x9e/0xc0 [ 833.215629] [] lookup_one_len+0x146/0x1e0 [ 833.215629] [] ? _raw_spin_lock+0xe/0x20 [ 833.215629] [] cifs_do_mount+0x26d/0x500 [cifs] [ 833.215629] [] mount_fs+0x43/0x1b0 [ 833.215629] [] vfs_kern_mount+0x6a/0xd0 [ 833.215629] [] do_kern_mount+0x54/0x110 [ 833.215629] [] do_mount+0x262/0x840 [ 833.215629] [] ? __get_free_pages+0xe/0x50 [ 833.215629] [] ? copy_mount_options+0x3a/0x180 [ 833.215629] [] sys_mount+0x8d/0xe0 [ 833.215629] [] system_call_fastpath+0x16/0x1b [ 833.215629] Code: Bad RIP value. [ 833.215629] RIP [< (null)>] (null) [ 833.215629] RSP [ 833.215629] CR2: 0000000000000000 [ 833.238525] ---[ end trace ec00758b8d44f529 ]--- When walking down the path on the server, it's possible to hit a symlink. The path walking code assumes that the caller will handle that situation properly, but cifs_get_root() isn't set up for it. This patch prevents the oops by simply returning an error. A better solution would be to try and chase the symlinks here, but that's fairly complicated to handle. Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=53221 Reported-and-tested-by: Kjell Braden Cc: stable Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 4bad7b16271f..1a052c0eee8e 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -564,6 +564,11 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb) dentry = ERR_PTR(-ENOENT); break; } + if (!S_ISDIR(dir->i_mode)) { + dput(dentry); + dentry = ERR_PTR(-ENOTDIR); + break; + } /* skip separators */ while (*s == sep) -- cgit From 07b92d0d570c903aeef4f3b76ebbdd728bc72805 Mon Sep 17 00:00:00 2001 From: Steve French Date: Mon, 18 Feb 2013 10:34:26 -0600 Subject: POSIX extensions disabled on client due to illegal O_EXCL flag sent to Samba Samba rejected libreoffice's attempt to open a file with illegal O_EXCL (without O_CREAT). Mask this flag off (as the local linux file system case does) for this case, so that we don't have disable Unix Extensions unnecessarily due to the Samba error (Samba server is also being fixed). See https://bugzilla.samba.org/show_bug.cgi?id=9519 Reviewed-by: Jeff Layton Signed-off-by: Steve French Signed-off-by: Steve French --- fs/cifs/file.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/cifs/file.c b/fs/cifs/file.c index c16d2a018ab8..8c0d85577314 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -43,6 +43,7 @@ #include "cifs_fs_sb.h" #include "fscache.h" + static inline int cifs_convert_flags(unsigned int flags) { if ((flags & O_ACCMODE) == O_RDONLY) @@ -72,10 +73,15 @@ static u32 cifs_posix_convert_flags(unsigned int flags) else if ((flags & O_ACCMODE) == O_RDWR) posix_flags = SMB_O_RDWR; - if (flags & O_CREAT) + if (flags & O_CREAT) { posix_flags |= SMB_O_CREAT; - if (flags & O_EXCL) - posix_flags |= SMB_O_EXCL; + if (flags & O_EXCL) + posix_flags |= SMB_O_EXCL; + } else if (flags & O_EXCL) + cFYI(1, "Application %s pid %d has incorrectly set O_EXCL flag" + "but not O_CREAT on file open. Ignoring O_EXCL", + current->comm, current->tgid); + if (flags & O_TRUNC) posix_flags |= SMB_O_TRUNC; /* be safe and imply O_SYNC for O_DSYNC */ -- cgit From 0b7bc84000d71f3647ca33ab1bf5bd928535c846 Mon Sep 17 00:00:00 2001 From: Martijn de Gouw Date: Wed, 24 Oct 2012 11:45:46 +0200 Subject: cifs: set MAY_SIGN when sec=krb5 Setting this secFlg allows usage of dfs where some servers require signing and others don't. Signed-off-by: Martijn de Gouw Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 4474a57f30ab..54125e04fd0c 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1031,7 +1031,7 @@ static int cifs_parse_security_flavors(char *value, switch (match_token(value, cifs_secflavor_tokens, args)) { case Opt_sec_krb5: - vol->secFlg |= CIFSSEC_MAY_KRB5; + vol->secFlg |= CIFSSEC_MAY_KRB5 | CIFSSEC_MAY_SIGN; break; case Opt_sec_krb5i: vol->secFlg |= CIFSSEC_MAY_KRB5 | CIFSSEC_MUST_SIGN; -- cgit From f6488c9ba51d65410e2dbc4345413c0d9120971e Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Wed, 27 Feb 2013 20:10:34 -0500 Subject: nfs: don't allow nfs_find_actor to match inodes of the wrong type Benny Halevy reported the following oops when testing RHEL6: <7>nfs_update_inode: inode 892950 mode changed, 0040755 to 0100644 <1>BUG: unable to handle kernel NULL pointer dereference at (null) <1>IP: [] nfs_closedir+0x15/0x30 [nfs] <4>PGD 81448a067 PUD 831632067 PMD 0 <4>Oops: 0000 [#1] SMP <4>last sysfs file: /sys/kernel/mm/redhat_transparent_hugepage/enabled <4>CPU 6 <4>Modules linked in: fuse bonding 8021q garp ebtable_nat ebtables be2iscsi iscsi_boot_sysfs bnx2i cnic uio cxgb4i cxgb4 cxgb3i libcxgbi cxgb3 mdio ib_iser rdma_cm ib_cm iw_cm ib_sa ib_mad ib_core ib_addr iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi softdog bridge stp llc xt_physdev ipt_REJECT nf_conntrack_ipv4 nf_defrag_ipv4 xt_multiport iptable_filter ip_tables ip6t_REJECT nf_conntrack_ipv6 nf_defrag_ipv6 xt_state nf_conntrack ip6table_filter ip6_tables ipv6 dm_round_robin dm_multipath objlayoutdriver2(U) nfs(U) lockd fscache auth_rpcgss nfs_acl sunrpc vhost_net macvtap macvlan tun kvm_intel kvm be2net igb dca ptp pps_core microcode serio_raw sg iTCO_wdt iTCO_vendor_support i7core_edac edac_core shpchp ext4 mbcache jbd2 sd_mod crc_t10dif ahci dm_mirror dm_region_hash dm_log dm_mod [last unloaded: scsi_wait_scan] <4> <4>Pid: 6332, comm: dd Not tainted 2.6.32-358.el6.x86_64 #1 HP ProLiant DL170e G6 /ProLiant DL170e G6 <4>RIP: 0010:[] [] nfs_closedir+0x15/0x30 [nfs] <4>RSP: 0018:ffff88081458bb98 EFLAGS: 00010292 <4>RAX: ffffffffa02a52b0 RBX: 0000000000000000 RCX: 0000000000000003 <4>RDX: ffffffffa02e45a0 RSI: ffff88081440b300 RDI: ffff88082d5f5760 <4>RBP: ffff88081458bba8 R08: 0000000000000000 R09: 0000000000000000 <4>R10: 0000000000000772 R11: 0000000000400004 R12: 0000000040000008 <4>R13: ffff88082d5f5760 R14: ffff88082d6e8800 R15: ffff88082f12d780 <4>FS: 00007f728f37e700(0000) GS:ffff8800456c0000(0000) knlGS:0000000000000000 <4>CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b <4>CR2: 0000000000000000 CR3: 0000000831279000 CR4: 00000000000007e0 <4>DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 <4>DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 <4>Process dd (pid: 6332, threadinfo ffff88081458a000, task ffff88082fa0e040) <4>Stack: <4> 0000000040000008 ffff88081440b300 ffff88081458bbf8 ffffffff81182745 <4> ffff88082d5f5760 ffff88082d6e8800 ffff88081458bbf8 ffffffffffffffea <4> ffff88082f12d780 ffff88082d6e8800 ffffffffa02a50a0 ffff88082d5f5760 <4>Call Trace: <4> [] __fput+0xf5/0x210 <4> [] ? do_open+0x0/0x20 [nfs] <4> [] fput+0x25/0x30 <4> [] __dentry_open+0x27e/0x360 <4> [] ? inotify_d_instantiate+0x2a/0x60 <4> [] lookup_instantiate_filp+0x69/0x90 <4> [] nfs_intent_set_file+0x59/0x90 [nfs] <4> [] nfs_atomic_lookup+0x1bb/0x310 [nfs] <4> [] __lookup_hash+0x102/0x160 <4> [] ? selinux_inode_permission+0x72/0xb0 <4> [] lookup_hash+0x3a/0x50 <4> [] do_filp_open+0x2eb/0xdd0 <4> [] ? __do_page_fault+0x1ec/0x480 <4> [] ? alloc_fd+0x92/0x160 <4> [] do_sys_open+0x69/0x140 <4> [] ? sys_lseek+0x66/0x80 <4> [] sys_open+0x20/0x30 <4> [] system_call_fastpath+0x16/0x1b <4>Code: 65 48 8b 04 25 c8 cb 00 00 83 a8 44 e0 ff ff 01 5b 41 5c c9 c3 90 55 48 89 e5 53 48 83 ec 08 0f 1f 44 00 00 48 8b 9e a0 00 00 00 <48> 8b 3b e8 13 0c f7 ff 48 89 df e8 ab 3d ec e0 48 83 c4 08 31 <1>RIP [] nfs_closedir+0x15/0x30 [nfs] <4> RSP <4>CR2: 0000000000000000 I think this is ultimately due to a bug on the server. The client had previously found a directory dentry. It then later tried to do an atomic open on a new (regular file) dentry. The attributes it got back had the same filehandle as the previously found directory inode. It then tried to put the filp because it failed the aops tests for O_DIRECT opens, and oopsed here because the ctx was still NULL. Obviously the root cause here is a server issue, but we can take steps to mitigate this on the client. When nfs_fhget is called, we always know what type of inode it is. In the event that there's a broken or malicious server on the other end of the wire, the client can end up crashing because the wrong ops are set on it. Have nfs_find_actor check that the inode type is correct after checking the fileid. The fileid check should rarely ever match, so it should only rarely ever get to this check. In the case where we have a broken server, we may see two different inodes with the same i_ino, but the client should be able to cope with them without crashing. This should fix the oops reported here: https://bugzilla.redhat.com/show_bug.cgi?id=913660 Reported-by: Benny Halevy Signed-off-by: Jeff Layton Signed-off-by: Trond Myklebust --- fs/nfs/inode.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 6acc73c80d7f..f52c99f4785b 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -237,6 +237,8 @@ nfs_find_actor(struct inode *inode, void *opaque) if (NFS_FILEID(inode) != fattr->fileid) return 0; + if ((S_IFMT & inode->i_mode) != (S_IFMT & fattr->mode)) + return 0; if (nfs_compare_fh(NFS_FH(inode), fh)) return 0; if (is_bad_inode(inode) || NFS_STALE(inode)) -- cgit From 32918dd9f19e5960af4cdfa41190bb843fb2247b Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Wed, 27 Feb 2013 17:02:48 -0800 Subject: ocfs2: fix ocfs2_init_security_and_acl() to initialize acl correctly We need to re-initialize the security for a new reflinked inode with its parent dirs if it isn't specified to be preserved for ocfs2_reflink(). However, the code logic is broken at ocfs2_init_security_and_acl() although ocfs2_init_security_get() succeed. As a result, ocfs2_acl_init() does not involked and therefore the default ACL of parent dir was missing on the new inode. Note this was introduced by 9d8f13ba3 ("security: new security_inode_init_security API adds function callback") To reproduce: set default ACL for the parent dir(ocfs2 in this case): $ setfacl -m default:user:jeff:rwx ../ocfs2/ $ getfacl ../ocfs2/ # file: ../ocfs2/ # owner: jeff # group: jeff user::rwx group::r-x other::r-x default:user::rwx default:user:jeff:rwx default:group::r-x default:mask::rwx default:other::r-x $ touch a $ getfacl a # file: a # owner: jeff # group: jeff user::rw- group::rw- other::r-- Before patching, create reflink file b from a, the user default ACL entry(user:jeff:rwx)was missing: $ ./ocfs2_reflink a b $ getfacl b # file: b # owner: jeff # group: jeff user::rw- group::rw- other::r-- In this case, the end user can also observed an error message at syslog: (ocfs2_reflink,3229,2):ocfs2_init_security_and_acl:7193 ERROR: status = 0 After applying this patch, create reflink file c from a: $ ./ocfs2_reflink a c $ getfacl c # file: c # owner: jeff # group: jeff user::rw- user:jeff:rwx #effective:rw- group::r-x #effective:r-- mask::rw- other::r-- Test program: /* Usage: reflink */ #include #include #include #include #include #include #include #include #include static int reflink_file(char const *src_name, char const *dst_name, bool preserve_attrs) { int fd; #ifndef REFLINK_ATTR_NONE # define REFLINK_ATTR_NONE 0 #endif #ifndef REFLINK_ATTR_PRESERVE # define REFLINK_ATTR_PRESERVE 1 #endif #ifndef OCFS2_IOC_REFLINK struct reflink_arguments { uint64_t old_path; uint64_t new_path; uint64_t preserve; }; # define OCFS2_IOC_REFLINK _IOW ('o', 4, struct reflink_arguments) #endif struct reflink_arguments args = { .old_path = (unsigned long) src_name, .new_path = (unsigned long) dst_name, .preserve = preserve_attrs ? REFLINK_ATTR_PRESERVE : REFLINK_ATTR_NONE, }; fd = open(src_name, O_RDONLY); if (fd < 0) { fprintf(stderr, "Failed to open %s: %s\n", src_name, strerror(errno)); return -1; } if (ioctl(fd, OCFS2_IOC_REFLINK, &args) < 0) { fprintf(stderr, "Failed to reflink %s to %s: %s\n", src_name, dst_name, strerror(errno)); return -1; } } int main(int argc, char *argv[]) { if (argc != 3) { fprintf(stdout, "Usage: %s source dest\n", argv[0]); return 1; } return reflink_file(argv[1], argv[2], 0); } Signed-off-by: Jie Liu Reviewed-by: Tao Ma Cc: Mimi Zohar Cc: Joel Becker Cc: Mark Fasheh Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ocfs2/xattr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index 0ba9ea1e7961..2e3ea308c144 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -7189,7 +7189,7 @@ int ocfs2_init_security_and_acl(struct inode *dir, struct buffer_head *dir_bh = NULL; ret = ocfs2_init_security_get(inode, dir, qstr, NULL); - if (!ret) { + if (ret) { mlog_errno(ret); goto leave; } -- cgit From 309a85b6861fedbb48a22d45e0e079d1be993b3a Mon Sep 17 00:00:00 2001 From: "Xiaowei.Hu" Date: Wed, 27 Feb 2013 17:02:49 -0800 Subject: ocfs2: ac->ac_allow_chain_relink=0 won't disable group relink ocfs2_block_group_alloc_discontig() disables chain relink by setting ac->ac_allow_chain_relink = 0 because it grabs clusters from multiple cluster groups. It doesn't keep the credits for all chain relink,but ocfs2_claim_suballoc_bits overrides this in this call trace: ocfs2_block_group_claim_bits()->ocfs2_claim_clusters()-> __ocfs2_claim_clusters()->ocfs2_claim_suballoc_bits() ocfs2_claim_suballoc_bits set ac->ac_allow_chain_relink = 1; then call ocfs2_search_chain() one time and disable it again, and then we run out of credits. Fix is to allow relink by default and disable it in ocfs2_block_group_alloc_discontig. Without this patch, End-users will run into a crash due to run out of credits, backtrace like this: RIP: 0010:[] [] jbd2_journal_dirty_metadata+0x164/0x170 [jbd2] RSP: 0018:ffff8801b919b5b8 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff88022139ddc0 RCX: ffff880159f652d0 RDX: ffff880178aa3000 RSI: ffff880159f652d0 RDI: ffff880087f09bf8 RBP: ffff8801b919b5e8 R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000001e00 R11: 00000000000150b0 R12: ffff880159f652d0 R13: ffff8801a0cae908 R14: ffff880087f09bf8 R15: ffff88018d177800 FS: 00007fc9b0b6b6e0(0000) GS:ffff88022fd40000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 000000000040819c CR3: 0000000184017000 CR4: 00000000000006e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process dd (pid: 9945, threadinfo ffff8801b919a000, task ffff880149a264c0) Call Trace: ocfs2_journal_dirty+0x2f/0x70 [ocfs2] ocfs2_relink_block_group+0x111/0x480 [ocfs2] ocfs2_search_chain+0x455/0x9a0 [ocfs2] ... Signed-off-by: Xiaowei.Hu Reviewed-by: Srinivas Eeda Cc: Mark Fasheh Cc: Joel Becker Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ocfs2/suballoc.c | 7 +++---- fs/ocfs2/suballoc.h | 2 +- 2 files changed, 4 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index f169da4624fd..b7e74b580c0f 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -642,7 +642,7 @@ ocfs2_block_group_alloc_discontig(handle_t *handle, * cluster groups will be staying in cache for the duration of * this operation. */ - ac->ac_allow_chain_relink = 0; + ac->ac_disable_chain_relink = 1; /* Claim the first region */ status = ocfs2_block_group_claim_bits(osb, handle, ac, min_bits, @@ -1823,7 +1823,7 @@ static int ocfs2_search_chain(struct ocfs2_alloc_context *ac, * Do this *after* figuring out how many bits we're taking out * of our target group. */ - if (ac->ac_allow_chain_relink && + if (!ac->ac_disable_chain_relink && (prev_group_bh) && (ocfs2_block_group_reasonably_empty(bg, res->sr_bits))) { status = ocfs2_relink_block_group(handle, alloc_inode, @@ -1928,7 +1928,6 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_alloc_context *ac, victim = ocfs2_find_victim_chain(cl); ac->ac_chain = victim; - ac->ac_allow_chain_relink = 1; status = ocfs2_search_chain(ac, handle, bits_wanted, min_bits, res, &bits_left); @@ -1947,7 +1946,7 @@ static int ocfs2_claim_suballoc_bits(struct ocfs2_alloc_context *ac, * searching each chain in order. Don't allow chain relinking * because we only calculate enough journal credits for one * relink per alloc. */ - ac->ac_allow_chain_relink = 0; + ac->ac_disable_chain_relink = 1; for (i = 0; i < le16_to_cpu(cl->cl_next_free_rec); i ++) { if (i == victim) continue; diff --git a/fs/ocfs2/suballoc.h b/fs/ocfs2/suballoc.h index b8afabfeede4..a36d0aa50911 100644 --- a/fs/ocfs2/suballoc.h +++ b/fs/ocfs2/suballoc.h @@ -49,7 +49,7 @@ struct ocfs2_alloc_context { /* these are used by the chain search */ u16 ac_chain; - int ac_allow_chain_relink; + int ac_disable_chain_relink; group_search_t *ac_group_search; u64 ac_last_group; -- cgit From 9ed083d8ccc6186448c3558c6f40b40ba0b1568a Mon Sep 17 00:00:00 2001 From: Vyacheslav Dubeyko Date: Wed, 27 Feb 2013 17:03:00 -0800 Subject: hfsplus: add on-disk layout declarations related to attributes tree Add all necessary on-disk layout declarations related to attributes file. Signed-off-by: Vyacheslav Dubeyko Reported-by: Hin-Tak Leung Cc: Al Viro Cc: Christoph Hellwig Cc: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hfsplus/hfsplus_raw.h | 68 ++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 66 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/hfsplus/hfsplus_raw.h b/fs/hfsplus/hfsplus_raw.h index 921967e5abb1..452ede01b036 100644 --- a/fs/hfsplus/hfsplus_raw.h +++ b/fs/hfsplus/hfsplus_raw.h @@ -52,13 +52,23 @@ typedef __be32 hfsplus_cnid; typedef __be16 hfsplus_unichr; +#define HFSPLUS_MAX_STRLEN 255 +#define HFSPLUS_ATTR_MAX_STRLEN 127 + /* A "string" as used in filenames, etc. */ struct hfsplus_unistr { __be16 length; - hfsplus_unichr unicode[255]; + hfsplus_unichr unicode[HFSPLUS_MAX_STRLEN]; } __packed; -#define HFSPLUS_MAX_STRLEN 255 +/* + * A "string" is used in attributes file + * for name of extended attribute + */ +struct hfsplus_attr_unistr { + __be16 length; + hfsplus_unichr unicode[HFSPLUS_ATTR_MAX_STRLEN]; +} __packed; /* POSIX permissions */ struct hfsplus_perm { @@ -291,6 +301,8 @@ struct hfsplus_cat_file { /* File attribute bits */ #define HFSPLUS_FILE_LOCKED 0x0001 #define HFSPLUS_FILE_THREAD_EXISTS 0x0002 +#define HFSPLUS_XATTR_EXISTS 0x0004 +#define HFSPLUS_ACL_EXISTS 0x0008 /* HFS+ catalog thread (part of a cat_entry) */ struct hfsplus_cat_thread { @@ -327,11 +339,63 @@ struct hfsplus_ext_key { #define HFSPLUS_EXT_KEYLEN sizeof(struct hfsplus_ext_key) +#define HFSPLUS_XATTR_FINDER_INFO_NAME "com.apple.FinderInfo" +#define HFSPLUS_XATTR_ACL_NAME "com.apple.system.Security" + +#define HFSPLUS_ATTR_INLINE_DATA 0x10 +#define HFSPLUS_ATTR_FORK_DATA 0x20 +#define HFSPLUS_ATTR_EXTENTS 0x30 + +/* HFS+ attributes tree key */ +struct hfsplus_attr_key { + __be16 key_len; + __be16 pad; + hfsplus_cnid cnid; + __be32 start_block; + struct hfsplus_attr_unistr key_name; +} __packed; + +#define HFSPLUS_ATTR_KEYLEN sizeof(struct hfsplus_attr_key) + +/* HFS+ fork data attribute */ +struct hfsplus_attr_fork_data { + __be32 record_type; + __be32 reserved; + struct hfsplus_fork_raw the_fork; +} __packed; + +/* HFS+ extension attribute */ +struct hfsplus_attr_extents { + __be32 record_type; + __be32 reserved; + struct hfsplus_extent extents; +} __packed; + +#define HFSPLUS_MAX_INLINE_DATA_SIZE 3802 + +/* HFS+ attribute inline data */ +struct hfsplus_attr_inline_data { + __be32 record_type; + __be32 reserved1; + u8 reserved2[6]; + __be16 length; + u8 raw_bytes[HFSPLUS_MAX_INLINE_DATA_SIZE]; +} __packed; + +/* A data record in the attributes tree */ +typedef union { + __be32 record_type; + struct hfsplus_attr_fork_data fork_data; + struct hfsplus_attr_extents extents; + struct hfsplus_attr_inline_data inline_data; +} __packed hfsplus_attr_entry; + /* HFS+ generic BTree key */ typedef union { __be16 key_len; struct hfsplus_cat_key cat; struct hfsplus_ext_key ext; + struct hfsplus_attr_key attr; } __packed hfsplus_btree_key; #endif -- cgit From 3e05ca20fb570b456bd9841b5ff489d865e8c563 Mon Sep 17 00:00:00 2001 From: Vyacheslav Dubeyko Date: Wed, 27 Feb 2013 17:03:01 -0800 Subject: hfsplus: add functionality of manipulating by records in attributes tree Add functionality of manipulating by records in attributes tree. Signed-off-by: Vyacheslav Dubeyko Reported-by: Hin-Tak Leung Cc: Al Viro Cc: Christoph Hellwig Cc: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hfsplus/attributes.c | 399 ++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 399 insertions(+) create mode 100644 fs/hfsplus/attributes.c (limited to 'fs') diff --git a/fs/hfsplus/attributes.c b/fs/hfsplus/attributes.c new file mode 100644 index 000000000000..8d691f124714 --- /dev/null +++ b/fs/hfsplus/attributes.c @@ -0,0 +1,399 @@ +/* + * linux/fs/hfsplus/attributes.c + * + * Vyacheslav Dubeyko + * + * Handling of records in attributes tree + */ + +#include "hfsplus_fs.h" +#include "hfsplus_raw.h" + +static struct kmem_cache *hfsplus_attr_tree_cachep; + +int hfsplus_create_attr_tree_cache(void) +{ + if (hfsplus_attr_tree_cachep) + return -EEXIST; + + hfsplus_attr_tree_cachep = + kmem_cache_create("hfsplus_attr_cache", + sizeof(hfsplus_attr_entry), 0, + SLAB_HWCACHE_ALIGN, NULL); + if (!hfsplus_attr_tree_cachep) + return -ENOMEM; + + return 0; +} + +void hfsplus_destroy_attr_tree_cache(void) +{ + kmem_cache_destroy(hfsplus_attr_tree_cachep); +} + +int hfsplus_attr_bin_cmp_key(const hfsplus_btree_key *k1, + const hfsplus_btree_key *k2) +{ + __be32 k1_cnid, k2_cnid; + + k1_cnid = k1->attr.cnid; + k2_cnid = k2->attr.cnid; + if (k1_cnid != k2_cnid) + return be32_to_cpu(k1_cnid) < be32_to_cpu(k2_cnid) ? -1 : 1; + + return hfsplus_strcmp( + (const struct hfsplus_unistr *)&k1->attr.key_name, + (const struct hfsplus_unistr *)&k2->attr.key_name); +} + +int hfsplus_attr_build_key(struct super_block *sb, hfsplus_btree_key *key, + u32 cnid, const char *name) +{ + int len; + + memset(key, 0, sizeof(struct hfsplus_attr_key)); + key->attr.cnid = cpu_to_be32(cnid); + if (name) { + len = strlen(name); + if (len > HFSPLUS_ATTR_MAX_STRLEN) { + printk(KERN_ERR "hfs: invalid xattr name's length\n"); + return -EINVAL; + } + hfsplus_asc2uni(sb, + (struct hfsplus_unistr *)&key->attr.key_name, + HFSPLUS_ATTR_MAX_STRLEN, name, len); + len = be16_to_cpu(key->attr.key_name.length); + } else { + key->attr.key_name.length = 0; + len = 0; + } + + /* The length of the key, as stored in key_len field, does not include + * the size of the key_len field itself. + * So, offsetof(hfsplus_attr_key, key_name) is a trick because + * it takes into consideration key_len field (__be16) of + * hfsplus_attr_key structure instead of length field (__be16) of + * hfsplus_attr_unistr structure. + */ + key->key_len = + cpu_to_be16(offsetof(struct hfsplus_attr_key, key_name) + + 2 * len); + + return 0; +} + +void hfsplus_attr_build_key_uni(hfsplus_btree_key *key, + u32 cnid, + struct hfsplus_attr_unistr *name) +{ + int ustrlen; + + memset(key, 0, sizeof(struct hfsplus_attr_key)); + ustrlen = be16_to_cpu(name->length); + key->attr.cnid = cpu_to_be32(cnid); + key->attr.key_name.length = cpu_to_be16(ustrlen); + ustrlen *= 2; + memcpy(key->attr.key_name.unicode, name->unicode, ustrlen); + + /* The length of the key, as stored in key_len field, does not include + * the size of the key_len field itself. + * So, offsetof(hfsplus_attr_key, key_name) is a trick because + * it takes into consideration key_len field (__be16) of + * hfsplus_attr_key structure instead of length field (__be16) of + * hfsplus_attr_unistr structure. + */ + key->key_len = + cpu_to_be16(offsetof(struct hfsplus_attr_key, key_name) + + ustrlen); +} + +hfsplus_attr_entry *hfsplus_alloc_attr_entry(void) +{ + return kmem_cache_alloc(hfsplus_attr_tree_cachep, GFP_KERNEL); +} + +void hfsplus_destroy_attr_entry(hfsplus_attr_entry *entry) +{ + if (entry) + kmem_cache_free(hfsplus_attr_tree_cachep, entry); +} + +#define HFSPLUS_INVALID_ATTR_RECORD -1 + +static int hfsplus_attr_build_record(hfsplus_attr_entry *entry, int record_type, + u32 cnid, const void *value, size_t size) +{ + if (record_type == HFSPLUS_ATTR_FORK_DATA) { + /* + * Mac OS X supports only inline data attributes. + * Do nothing + */ + memset(entry, 0, sizeof(*entry)); + return sizeof(struct hfsplus_attr_fork_data); + } else if (record_type == HFSPLUS_ATTR_EXTENTS) { + /* + * Mac OS X supports only inline data attributes. + * Do nothing. + */ + memset(entry, 0, sizeof(*entry)); + return sizeof(struct hfsplus_attr_extents); + } else if (record_type == HFSPLUS_ATTR_INLINE_DATA) { + u16 len; + + memset(entry, 0, sizeof(struct hfsplus_attr_inline_data)); + entry->inline_data.record_type = cpu_to_be32(record_type); + if (size <= HFSPLUS_MAX_INLINE_DATA_SIZE) + len = size; + else + return HFSPLUS_INVALID_ATTR_RECORD; + entry->inline_data.length = cpu_to_be16(len); + memcpy(entry->inline_data.raw_bytes, value, len); + /* + * Align len on two-byte boundary. + * It needs to add pad byte if we have odd len. + */ + len = round_up(len, 2); + return offsetof(struct hfsplus_attr_inline_data, raw_bytes) + + len; + } else /* invalid input */ + memset(entry, 0, sizeof(*entry)); + + return HFSPLUS_INVALID_ATTR_RECORD; +} + +int hfsplus_find_attr(struct super_block *sb, u32 cnid, + const char *name, struct hfs_find_data *fd) +{ + int err = 0; + + dprint(DBG_ATTR_MOD, "find_attr: %s,%d\n", name ? name : NULL, cnid); + + if (!HFSPLUS_SB(sb)->attr_tree) { + printk(KERN_ERR "hfs: attributes file doesn't exist\n"); + return -EINVAL; + } + + if (name) { + err = hfsplus_attr_build_key(sb, fd->search_key, cnid, name); + if (err) + goto failed_find_attr; + err = hfs_brec_find(fd, hfs_find_rec_by_key); + if (err) + goto failed_find_attr; + } else { + err = hfsplus_attr_build_key(sb, fd->search_key, cnid, NULL); + if (err) + goto failed_find_attr; + err = hfs_brec_find(fd, hfs_find_1st_rec_by_cnid); + if (err) + goto failed_find_attr; + } + +failed_find_attr: + return err; +} + +int hfsplus_attr_exists(struct inode *inode, const char *name) +{ + int err = 0; + struct super_block *sb = inode->i_sb; + struct hfs_find_data fd; + + if (!HFSPLUS_SB(sb)->attr_tree) + return 0; + + err = hfs_find_init(HFSPLUS_SB(sb)->attr_tree, &fd); + if (err) + return 0; + + err = hfsplus_find_attr(sb, inode->i_ino, name, &fd); + if (err) + goto attr_not_found; + + hfs_find_exit(&fd); + return 1; + +attr_not_found: + hfs_find_exit(&fd); + return 0; +} + +int hfsplus_create_attr(struct inode *inode, + const char *name, + const void *value, size_t size) +{ + struct super_block *sb = inode->i_sb; + struct hfs_find_data fd; + hfsplus_attr_entry *entry_ptr; + int entry_size; + int err; + + dprint(DBG_ATTR_MOD, "create_attr: %s,%ld\n", + name ? name : NULL, inode->i_ino); + + if (!HFSPLUS_SB(sb)->attr_tree) { + printk(KERN_ERR "hfs: attributes file doesn't exist\n"); + return -EINVAL; + } + + entry_ptr = hfsplus_alloc_attr_entry(); + if (!entry_ptr) + return -ENOMEM; + + err = hfs_find_init(HFSPLUS_SB(sb)->attr_tree, &fd); + if (err) + goto failed_init_create_attr; + + if (name) { + err = hfsplus_attr_build_key(sb, fd.search_key, + inode->i_ino, name); + if (err) + goto failed_create_attr; + } else { + err = -EINVAL; + goto failed_create_attr; + } + + /* Mac OS X supports only inline data attributes. */ + entry_size = hfsplus_attr_build_record(entry_ptr, + HFSPLUS_ATTR_INLINE_DATA, + inode->i_ino, + value, size); + if (entry_size == HFSPLUS_INVALID_ATTR_RECORD) { + err = -EINVAL; + goto failed_create_attr; + } + + err = hfs_brec_find(&fd, hfs_find_rec_by_key); + if (err != -ENOENT) { + if (!err) + err = -EEXIST; + goto failed_create_attr; + } + + err = hfs_brec_insert(&fd, entry_ptr, entry_size); + if (err) + goto failed_create_attr; + + hfsplus_mark_inode_dirty(inode, HFSPLUS_I_ATTR_DIRTY); + +failed_create_attr: + hfs_find_exit(&fd); + +failed_init_create_attr: + hfsplus_destroy_attr_entry(entry_ptr); + return err; +} + +static int __hfsplus_delete_attr(struct inode *inode, u32 cnid, + struct hfs_find_data *fd) +{ + int err = 0; + __be32 found_cnid, record_type; + + hfs_bnode_read(fd->bnode, &found_cnid, + fd->keyoffset + + offsetof(struct hfsplus_attr_key, cnid), + sizeof(__be32)); + if (cnid != be32_to_cpu(found_cnid)) + return -ENOENT; + + hfs_bnode_read(fd->bnode, &record_type, + fd->entryoffset, sizeof(record_type)); + + switch (be32_to_cpu(record_type)) { + case HFSPLUS_ATTR_INLINE_DATA: + /* All is OK. Do nothing. */ + break; + case HFSPLUS_ATTR_FORK_DATA: + case HFSPLUS_ATTR_EXTENTS: + printk(KERN_ERR "hfs: only inline data xattr are supported\n"); + return -EOPNOTSUPP; + default: + printk(KERN_ERR "hfs: invalid extended attribute record\n"); + return -ENOENT; + } + + err = hfs_brec_remove(fd); + if (err) + return err; + + hfsplus_mark_inode_dirty(inode, HFSPLUS_I_ATTR_DIRTY); + return err; +} + +int hfsplus_delete_attr(struct inode *inode, const char *name) +{ + int err = 0; + struct super_block *sb = inode->i_sb; + struct hfs_find_data fd; + + dprint(DBG_ATTR_MOD, "delete_attr: %s,%ld\n", + name ? name : NULL, inode->i_ino); + + if (!HFSPLUS_SB(sb)->attr_tree) { + printk(KERN_ERR "hfs: attributes file doesn't exist\n"); + return -EINVAL; + } + + err = hfs_find_init(HFSPLUS_SB(sb)->attr_tree, &fd); + if (err) + return err; + + if (name) { + err = hfsplus_attr_build_key(sb, fd.search_key, + inode->i_ino, name); + if (err) + goto out; + } else { + printk(KERN_ERR "hfs: invalid extended attribute name\n"); + err = -EINVAL; + goto out; + } + + err = hfs_brec_find(&fd, hfs_find_rec_by_key); + if (err) + goto out; + + err = __hfsplus_delete_attr(inode, inode->i_ino, &fd); + if (err) + goto out; + +out: + hfs_find_exit(&fd); + return err; +} + +int hfsplus_delete_all_attrs(struct inode *dir, u32 cnid) +{ + int err = 0; + struct hfs_find_data fd; + + dprint(DBG_ATTR_MOD, "delete_all_attrs: %d\n", cnid); + + if (!HFSPLUS_SB(dir->i_sb)->attr_tree) { + printk(KERN_ERR "hfs: attributes file doesn't exist\n"); + return -EINVAL; + } + + err = hfs_find_init(HFSPLUS_SB(dir->i_sb)->attr_tree, &fd); + if (err) + return err; + + for (;;) { + err = hfsplus_find_attr(dir->i_sb, cnid, NULL, &fd); + if (err) { + if (err != -ENOENT) + printk(KERN_ERR "hfs: xattr search failed.\n"); + goto end_delete_all; + } + + err = __hfsplus_delete_attr(dir, cnid, &fd); + if (err) + goto end_delete_all; + } + +end_delete_all: + hfs_find_exit(&fd); + return err; +} -- cgit From 127e5f5ae51eff44c9bff673f24e8587caa4e29f Mon Sep 17 00:00:00 2001 From: Vyacheslav Dubeyko Date: Wed, 27 Feb 2013 17:03:03 -0800 Subject: hfsplus: rework functionality of getting, setting and deleting of extended attributes Rework functionality of getting, setting and deleting of extended attributes. Signed-off-by: Vyacheslav Dubeyko Reported-by: Hin-Tak Leung Cc: Al Viro Cc: Christoph Hellwig Cc: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hfsplus/xattr.c | 709 ++++++++++++++++++++++++++++++++++++++++++++ fs/hfsplus/xattr.h | 60 ++++ fs/hfsplus/xattr_security.c | 104 +++++++ fs/hfsplus/xattr_trusted.c | 63 ++++ fs/hfsplus/xattr_user.c | 63 ++++ 5 files changed, 999 insertions(+) create mode 100644 fs/hfsplus/xattr.c create mode 100644 fs/hfsplus/xattr.h create mode 100644 fs/hfsplus/xattr_security.c create mode 100644 fs/hfsplus/xattr_trusted.c create mode 100644 fs/hfsplus/xattr_user.c (limited to 'fs') diff --git a/fs/hfsplus/xattr.c b/fs/hfsplus/xattr.c new file mode 100644 index 000000000000..e8a4b0815c61 --- /dev/null +++ b/fs/hfsplus/xattr.c @@ -0,0 +1,709 @@ +/* + * linux/fs/hfsplus/xattr.c + * + * Vyacheslav Dubeyko + * + * Logic of processing extended attributes + */ + +#include "hfsplus_fs.h" +#include "xattr.h" + +const struct xattr_handler *hfsplus_xattr_handlers[] = { + &hfsplus_xattr_osx_handler, + &hfsplus_xattr_user_handler, + &hfsplus_xattr_trusted_handler, + &hfsplus_xattr_security_handler, + NULL +}; + +static int strcmp_xattr_finder_info(const char *name) +{ + if (name) { + return strncmp(name, HFSPLUS_XATTR_FINDER_INFO_NAME, + sizeof(HFSPLUS_XATTR_FINDER_INFO_NAME)); + } + return -1; +} + +static int strcmp_xattr_acl(const char *name) +{ + if (name) { + return strncmp(name, HFSPLUS_XATTR_ACL_NAME, + sizeof(HFSPLUS_XATTR_ACL_NAME)); + } + return -1; +} + +static inline int is_known_namespace(const char *name) +{ + if (strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN) && + strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN) && + strncmp(name, XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN) && + strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN)) + return false; + + return true; +} + +static int can_set_xattr(struct inode *inode, const char *name, + const void *value, size_t value_len) +{ + if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN)) + return -EOPNOTSUPP; /* TODO: implement ACL support */ + + if (!strncmp(name, XATTR_MAC_OSX_PREFIX, XATTR_MAC_OSX_PREFIX_LEN)) { + /* + * This makes sure that we aren't trying to set an + * attribute in a different namespace by prefixing it + * with "osx." + */ + if (is_known_namespace(name + XATTR_MAC_OSX_PREFIX_LEN)) + return -EOPNOTSUPP; + + return 0; + } + + /* + * Don't allow setting an attribute in an unknown namespace. + */ + if (strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN) && + strncmp(name, XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN) && + strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN)) + return -EOPNOTSUPP; + + return 0; +} + +int __hfsplus_setxattr(struct inode *inode, const char *name, + const void *value, size_t size, int flags) +{ + int err = 0; + struct hfs_find_data cat_fd; + hfsplus_cat_entry entry; + u16 cat_entry_flags, cat_entry_type; + u16 folder_finderinfo_len = sizeof(struct DInfo) + + sizeof(struct DXInfo); + u16 file_finderinfo_len = sizeof(struct FInfo) + + sizeof(struct FXInfo); + + if ((!S_ISREG(inode->i_mode) && + !S_ISDIR(inode->i_mode)) || + HFSPLUS_IS_RSRC(inode)) + return -EOPNOTSUPP; + + err = can_set_xattr(inode, name, value, size); + if (err) + return err; + + if (strncmp(name, XATTR_MAC_OSX_PREFIX, + XATTR_MAC_OSX_PREFIX_LEN) == 0) + name += XATTR_MAC_OSX_PREFIX_LEN; + + if (value == NULL) { + value = ""; + size = 0; + } + + err = hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &cat_fd); + if (err) { + printk(KERN_ERR "hfs: can't init xattr find struct\n"); + return err; + } + + err = hfsplus_find_cat(inode->i_sb, inode->i_ino, &cat_fd); + if (err) { + printk(KERN_ERR "hfs: catalog searching failed\n"); + goto end_setxattr; + } + + if (!strcmp_xattr_finder_info(name)) { + if (flags & XATTR_CREATE) { + printk(KERN_ERR "hfs: xattr exists yet\n"); + err = -EOPNOTSUPP; + goto end_setxattr; + } + hfs_bnode_read(cat_fd.bnode, &entry, cat_fd.entryoffset, + sizeof(hfsplus_cat_entry)); + if (be16_to_cpu(entry.type) == HFSPLUS_FOLDER) { + if (size == folder_finderinfo_len) { + memcpy(&entry.folder.user_info, value, + folder_finderinfo_len); + hfs_bnode_write(cat_fd.bnode, &entry, + cat_fd.entryoffset, + sizeof(struct hfsplus_cat_folder)); + hfsplus_mark_inode_dirty(inode, + HFSPLUS_I_CAT_DIRTY); + } else { + err = -ERANGE; + goto end_setxattr; + } + } else if (be16_to_cpu(entry.type) == HFSPLUS_FILE) { + if (size == file_finderinfo_len) { + memcpy(&entry.file.user_info, value, + file_finderinfo_len); + hfs_bnode_write(cat_fd.bnode, &entry, + cat_fd.entryoffset, + sizeof(struct hfsplus_cat_file)); + hfsplus_mark_inode_dirty(inode, + HFSPLUS_I_CAT_DIRTY); + } else { + err = -ERANGE; + goto end_setxattr; + } + } else { + err = -EOPNOTSUPP; + goto end_setxattr; + } + goto end_setxattr; + } + + if (!HFSPLUS_SB(inode->i_sb)->attr_tree) { + err = -EOPNOTSUPP; + goto end_setxattr; + } + + if (hfsplus_attr_exists(inode, name)) { + if (flags & XATTR_CREATE) { + printk(KERN_ERR "hfs: xattr exists yet\n"); + err = -EOPNOTSUPP; + goto end_setxattr; + } + err = hfsplus_delete_attr(inode, name); + if (err) + goto end_setxattr; + err = hfsplus_create_attr(inode, name, value, size); + if (err) + goto end_setxattr; + } else { + if (flags & XATTR_REPLACE) { + printk(KERN_ERR "hfs: cannot replace xattr\n"); + err = -EOPNOTSUPP; + goto end_setxattr; + } + err = hfsplus_create_attr(inode, name, value, size); + if (err) + goto end_setxattr; + } + + cat_entry_type = hfs_bnode_read_u16(cat_fd.bnode, cat_fd.entryoffset); + if (cat_entry_type == HFSPLUS_FOLDER) { + cat_entry_flags = hfs_bnode_read_u16(cat_fd.bnode, + cat_fd.entryoffset + + offsetof(struct hfsplus_cat_folder, flags)); + cat_entry_flags |= HFSPLUS_XATTR_EXISTS; + if (!strcmp_xattr_acl(name)) + cat_entry_flags |= HFSPLUS_ACL_EXISTS; + hfs_bnode_write_u16(cat_fd.bnode, cat_fd.entryoffset + + offsetof(struct hfsplus_cat_folder, flags), + cat_entry_flags); + hfsplus_mark_inode_dirty(inode, HFSPLUS_I_CAT_DIRTY); + } else if (cat_entry_type == HFSPLUS_FILE) { + cat_entry_flags = hfs_bnode_read_u16(cat_fd.bnode, + cat_fd.entryoffset + + offsetof(struct hfsplus_cat_file, flags)); + cat_entry_flags |= HFSPLUS_XATTR_EXISTS; + if (!strcmp_xattr_acl(name)) + cat_entry_flags |= HFSPLUS_ACL_EXISTS; + hfs_bnode_write_u16(cat_fd.bnode, cat_fd.entryoffset + + offsetof(struct hfsplus_cat_file, flags), + cat_entry_flags); + hfsplus_mark_inode_dirty(inode, HFSPLUS_I_CAT_DIRTY); + } else { + printk(KERN_ERR "hfs: invalid catalog entry type\n"); + err = -EIO; + goto end_setxattr; + } + +end_setxattr: + hfs_find_exit(&cat_fd); + return err; +} + +static inline int is_osx_xattr(const char *xattr_name) +{ + return !is_known_namespace(xattr_name); +} + +static int name_len(const char *xattr_name, int xattr_name_len) +{ + int len = xattr_name_len + 1; + + if (is_osx_xattr(xattr_name)) + len += XATTR_MAC_OSX_PREFIX_LEN; + + return len; +} + +static int copy_name(char *buffer, const char *xattr_name, int name_len) +{ + int len = name_len; + int offset = 0; + + if (is_osx_xattr(xattr_name)) { + strncpy(buffer, XATTR_MAC_OSX_PREFIX, XATTR_MAC_OSX_PREFIX_LEN); + offset += XATTR_MAC_OSX_PREFIX_LEN; + len += XATTR_MAC_OSX_PREFIX_LEN; + } + + strncpy(buffer + offset, xattr_name, name_len); + memset(buffer + offset + name_len, 0, 1); + len += 1; + + return len; +} + +static ssize_t hfsplus_getxattr_finder_info(struct dentry *dentry, + void *value, size_t size) +{ + ssize_t res = 0; + struct inode *inode = dentry->d_inode; + struct hfs_find_data fd; + u16 entry_type; + u16 folder_rec_len = sizeof(struct DInfo) + sizeof(struct DXInfo); + u16 file_rec_len = sizeof(struct FInfo) + sizeof(struct FXInfo); + u16 record_len = max(folder_rec_len, file_rec_len); + u8 folder_finder_info[sizeof(struct DInfo) + sizeof(struct DXInfo)]; + u8 file_finder_info[sizeof(struct FInfo) + sizeof(struct FXInfo)]; + + if (size >= record_len) { + res = hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &fd); + if (res) { + printk(KERN_ERR "hfs: can't init xattr find struct\n"); + return res; + } + res = hfsplus_find_cat(inode->i_sb, inode->i_ino, &fd); + if (res) + goto end_getxattr_finder_info; + entry_type = hfs_bnode_read_u16(fd.bnode, fd.entryoffset); + + if (entry_type == HFSPLUS_FOLDER) { + hfs_bnode_read(fd.bnode, folder_finder_info, + fd.entryoffset + + offsetof(struct hfsplus_cat_folder, user_info), + folder_rec_len); + memcpy(value, folder_finder_info, folder_rec_len); + res = folder_rec_len; + } else if (entry_type == HFSPLUS_FILE) { + hfs_bnode_read(fd.bnode, file_finder_info, + fd.entryoffset + + offsetof(struct hfsplus_cat_file, user_info), + file_rec_len); + memcpy(value, file_finder_info, file_rec_len); + res = file_rec_len; + } else { + res = -EOPNOTSUPP; + goto end_getxattr_finder_info; + } + } else + res = size ? -ERANGE : record_len; + +end_getxattr_finder_info: + if (size >= record_len) + hfs_find_exit(&fd); + return res; +} + +ssize_t hfsplus_getxattr(struct dentry *dentry, const char *name, + void *value, size_t size) +{ + struct inode *inode = dentry->d_inode; + struct hfs_find_data fd; + hfsplus_attr_entry *entry; + __be32 xattr_record_type; + u32 record_type; + u16 record_length = 0; + ssize_t res = 0; + + if ((!S_ISREG(inode->i_mode) && + !S_ISDIR(inode->i_mode)) || + HFSPLUS_IS_RSRC(inode)) + return -EOPNOTSUPP; + + if (strncmp(name, XATTR_MAC_OSX_PREFIX, + XATTR_MAC_OSX_PREFIX_LEN) == 0) { + /* skip "osx." prefix */ + name += XATTR_MAC_OSX_PREFIX_LEN; + /* + * Don't allow retrieving properly prefixed attributes + * by prepending them with "osx." + */ + if (is_known_namespace(name)) + return -EOPNOTSUPP; + } + + if (!strcmp_xattr_finder_info(name)) + return hfsplus_getxattr_finder_info(dentry, value, size); + + if (!HFSPLUS_SB(inode->i_sb)->attr_tree) + return -EOPNOTSUPP; + + entry = hfsplus_alloc_attr_entry(); + if (!entry) { + printk(KERN_ERR "hfs: can't allocate xattr entry\n"); + return -ENOMEM; + } + + res = hfs_find_init(HFSPLUS_SB(inode->i_sb)->attr_tree, &fd); + if (res) { + printk(KERN_ERR "hfs: can't init xattr find struct\n"); + goto failed_getxattr_init; + } + + res = hfsplus_find_attr(inode->i_sb, inode->i_ino, name, &fd); + if (res) { + if (res == -ENOENT) + res = -ENODATA; + else + printk(KERN_ERR "hfs: xattr searching failed\n"); + goto out; + } + + hfs_bnode_read(fd.bnode, &xattr_record_type, + fd.entryoffset, sizeof(xattr_record_type)); + record_type = be32_to_cpu(xattr_record_type); + if (record_type == HFSPLUS_ATTR_INLINE_DATA) { + record_length = hfs_bnode_read_u16(fd.bnode, + fd.entryoffset + + offsetof(struct hfsplus_attr_inline_data, + length)); + if (record_length > HFSPLUS_MAX_INLINE_DATA_SIZE) { + printk(KERN_ERR "hfs: invalid xattr record size\n"); + res = -EIO; + goto out; + } + } else if (record_type == HFSPLUS_ATTR_FORK_DATA || + record_type == HFSPLUS_ATTR_EXTENTS) { + printk(KERN_ERR "hfs: only inline data xattr are supported\n"); + res = -EOPNOTSUPP; + goto out; + } else { + printk(KERN_ERR "hfs: invalid xattr record\n"); + res = -EIO; + goto out; + } + + if (size) { + hfs_bnode_read(fd.bnode, entry, fd.entryoffset, + offsetof(struct hfsplus_attr_inline_data, + raw_bytes) + record_length); + } + + if (size >= record_length) { + memcpy(value, entry->inline_data.raw_bytes, record_length); + res = record_length; + } else + res = size ? -ERANGE : record_length; + +out: + hfs_find_exit(&fd); + +failed_getxattr_init: + hfsplus_destroy_attr_entry(entry); + return res; +} + +static inline int can_list(const char *xattr_name) +{ + if (!xattr_name) + return 0; + + return strncmp(xattr_name, XATTR_TRUSTED_PREFIX, + XATTR_TRUSTED_PREFIX_LEN) || + capable(CAP_SYS_ADMIN); +} + +static ssize_t hfsplus_listxattr_finder_info(struct dentry *dentry, + char *buffer, size_t size) +{ + ssize_t res = 0; + struct inode *inode = dentry->d_inode; + struct hfs_find_data fd; + u16 entry_type; + u8 folder_finder_info[sizeof(struct DInfo) + sizeof(struct DXInfo)]; + u8 file_finder_info[sizeof(struct FInfo) + sizeof(struct FXInfo)]; + unsigned long len, found_bit; + int xattr_name_len, symbols_count; + + res = hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &fd); + if (res) { + printk(KERN_ERR "hfs: can't init xattr find struct\n"); + return res; + } + + res = hfsplus_find_cat(inode->i_sb, inode->i_ino, &fd); + if (res) + goto end_listxattr_finder_info; + + entry_type = hfs_bnode_read_u16(fd.bnode, fd.entryoffset); + if (entry_type == HFSPLUS_FOLDER) { + len = sizeof(struct DInfo) + sizeof(struct DXInfo); + hfs_bnode_read(fd.bnode, folder_finder_info, + fd.entryoffset + + offsetof(struct hfsplus_cat_folder, user_info), + len); + found_bit = find_first_bit((void *)folder_finder_info, len*8); + } else if (entry_type == HFSPLUS_FILE) { + len = sizeof(struct FInfo) + sizeof(struct FXInfo); + hfs_bnode_read(fd.bnode, file_finder_info, + fd.entryoffset + + offsetof(struct hfsplus_cat_file, user_info), + len); + found_bit = find_first_bit((void *)file_finder_info, len*8); + } else { + res = -EOPNOTSUPP; + goto end_listxattr_finder_info; + } + + if (found_bit >= (len*8)) + res = 0; + else { + symbols_count = sizeof(HFSPLUS_XATTR_FINDER_INFO_NAME) - 1; + xattr_name_len = + name_len(HFSPLUS_XATTR_FINDER_INFO_NAME, symbols_count); + if (!buffer || !size) { + if (can_list(HFSPLUS_XATTR_FINDER_INFO_NAME)) + res = xattr_name_len; + } else if (can_list(HFSPLUS_XATTR_FINDER_INFO_NAME)) { + if (size < xattr_name_len) + res = -ERANGE; + else { + res = copy_name(buffer, + HFSPLUS_XATTR_FINDER_INFO_NAME, + symbols_count); + } + } + } + +end_listxattr_finder_info: + hfs_find_exit(&fd); + + return res; +} + +ssize_t hfsplus_listxattr(struct dentry *dentry, char *buffer, size_t size) +{ + ssize_t err; + ssize_t res = 0; + struct inode *inode = dentry->d_inode; + struct hfs_find_data fd; + u16 key_len = 0; + struct hfsplus_attr_key attr_key; + char strbuf[HFSPLUS_ATTR_MAX_STRLEN + + XATTR_MAC_OSX_PREFIX_LEN + 1] = {0}; + int xattr_name_len; + + if ((!S_ISREG(inode->i_mode) && + !S_ISDIR(inode->i_mode)) || + HFSPLUS_IS_RSRC(inode)) + return -EOPNOTSUPP; + + res = hfsplus_listxattr_finder_info(dentry, buffer, size); + if (res < 0) + return res; + else if (!HFSPLUS_SB(inode->i_sb)->attr_tree) + return (res == 0) ? -EOPNOTSUPP : res; + + err = hfs_find_init(HFSPLUS_SB(inode->i_sb)->attr_tree, &fd); + if (err) { + printk(KERN_ERR "hfs: can't init xattr find struct\n"); + return err; + } + + err = hfsplus_find_attr(inode->i_sb, inode->i_ino, NULL, &fd); + if (err) { + if (err == -ENOENT) { + if (res == 0) + res = -ENODATA; + goto end_listxattr; + } else { + res = err; + goto end_listxattr; + } + } + + for (;;) { + key_len = hfs_bnode_read_u16(fd.bnode, fd.keyoffset); + if (key_len == 0 || key_len > fd.tree->max_key_len) { + printk(KERN_ERR "hfs: invalid xattr key length: %d\n", + key_len); + res = -EIO; + goto end_listxattr; + } + + hfs_bnode_read(fd.bnode, &attr_key, + fd.keyoffset, key_len + sizeof(key_len)); + + if (be32_to_cpu(attr_key.cnid) != inode->i_ino) + goto end_listxattr; + + xattr_name_len = HFSPLUS_ATTR_MAX_STRLEN; + if (hfsplus_uni2asc(inode->i_sb, + (const struct hfsplus_unistr *)&fd.key->attr.key_name, + strbuf, &xattr_name_len)) { + printk(KERN_ERR "hfs: unicode conversion failed\n"); + res = -EIO; + goto end_listxattr; + } + + if (!buffer || !size) { + if (can_list(strbuf)) + res += name_len(strbuf, xattr_name_len); + } else if (can_list(strbuf)) { + if (size < (res + name_len(strbuf, xattr_name_len))) { + res = -ERANGE; + goto end_listxattr; + } else + res += copy_name(buffer + res, + strbuf, xattr_name_len); + } + + if (hfs_brec_goto(&fd, 1)) + goto end_listxattr; + } + +end_listxattr: + hfs_find_exit(&fd); + return res; +} + +int hfsplus_removexattr(struct dentry *dentry, const char *name) +{ + int err = 0; + struct inode *inode = dentry->d_inode; + struct hfs_find_data cat_fd; + u16 flags; + u16 cat_entry_type; + int is_xattr_acl_deleted = 0; + int is_all_xattrs_deleted = 0; + + if ((!S_ISREG(inode->i_mode) && + !S_ISDIR(inode->i_mode)) || + HFSPLUS_IS_RSRC(inode)) + return -EOPNOTSUPP; + + if (!HFSPLUS_SB(inode->i_sb)->attr_tree) + return -EOPNOTSUPP; + + err = can_set_xattr(inode, name, NULL, 0); + if (err) + return err; + + if (strncmp(name, XATTR_MAC_OSX_PREFIX, + XATTR_MAC_OSX_PREFIX_LEN) == 0) + name += XATTR_MAC_OSX_PREFIX_LEN; + + if (!strcmp_xattr_finder_info(name)) + return -EOPNOTSUPP; + + err = hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &cat_fd); + if (err) { + printk(KERN_ERR "hfs: can't init xattr find struct\n"); + return err; + } + + err = hfsplus_find_cat(inode->i_sb, inode->i_ino, &cat_fd); + if (err) { + printk(KERN_ERR "hfs: catalog searching failed\n"); + goto end_removexattr; + } + + err = hfsplus_delete_attr(inode, name); + if (err) + goto end_removexattr; + + is_xattr_acl_deleted = !strcmp_xattr_acl(name); + is_all_xattrs_deleted = !hfsplus_attr_exists(inode, NULL); + + if (!is_xattr_acl_deleted && !is_all_xattrs_deleted) + goto end_removexattr; + + cat_entry_type = hfs_bnode_read_u16(cat_fd.bnode, cat_fd.entryoffset); + + if (cat_entry_type == HFSPLUS_FOLDER) { + flags = hfs_bnode_read_u16(cat_fd.bnode, cat_fd.entryoffset + + offsetof(struct hfsplus_cat_folder, flags)); + if (is_xattr_acl_deleted) + flags &= ~HFSPLUS_ACL_EXISTS; + if (is_all_xattrs_deleted) + flags &= ~HFSPLUS_XATTR_EXISTS; + hfs_bnode_write_u16(cat_fd.bnode, cat_fd.entryoffset + + offsetof(struct hfsplus_cat_folder, flags), + flags); + hfsplus_mark_inode_dirty(inode, HFSPLUS_I_CAT_DIRTY); + } else if (cat_entry_type == HFSPLUS_FILE) { + flags = hfs_bnode_read_u16(cat_fd.bnode, cat_fd.entryoffset + + offsetof(struct hfsplus_cat_file, flags)); + if (is_xattr_acl_deleted) + flags &= ~HFSPLUS_ACL_EXISTS; + if (is_all_xattrs_deleted) + flags &= ~HFSPLUS_XATTR_EXISTS; + hfs_bnode_write_u16(cat_fd.bnode, cat_fd.entryoffset + + offsetof(struct hfsplus_cat_file, flags), + flags); + hfsplus_mark_inode_dirty(inode, HFSPLUS_I_CAT_DIRTY); + } else { + printk(KERN_ERR "hfs: invalid catalog entry type\n"); + err = -EIO; + goto end_removexattr; + } + +end_removexattr: + hfs_find_exit(&cat_fd); + return err; +} + +static int hfsplus_osx_getxattr(struct dentry *dentry, const char *name, + void *buffer, size_t size, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + + XATTR_MAC_OSX_PREFIX_LEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_MAC_OSX_PREFIX); + strcpy(xattr_name + XATTR_MAC_OSX_PREFIX_LEN, name); + + return hfsplus_getxattr(dentry, xattr_name, buffer, size); +} + +static int hfsplus_osx_setxattr(struct dentry *dentry, const char *name, + const void *buffer, size_t size, int flags, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + + XATTR_MAC_OSX_PREFIX_LEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_MAC_OSX_PREFIX); + strcpy(xattr_name + XATTR_MAC_OSX_PREFIX_LEN, name); + + return hfsplus_setxattr(dentry, xattr_name, buffer, size, flags); +} + +static size_t hfsplus_osx_listxattr(struct dentry *dentry, char *list, + size_t list_size, const char *name, size_t name_len, int type) +{ + /* + * This method is not used. + * It is used hfsplus_listxattr() instead of generic_listxattr(). + */ + return -EOPNOTSUPP; +} + +const struct xattr_handler hfsplus_xattr_osx_handler = { + .prefix = XATTR_MAC_OSX_PREFIX, + .list = hfsplus_osx_listxattr, + .get = hfsplus_osx_getxattr, + .set = hfsplus_osx_setxattr, +}; diff --git a/fs/hfsplus/xattr.h b/fs/hfsplus/xattr.h new file mode 100644 index 000000000000..847b695b984d --- /dev/null +++ b/fs/hfsplus/xattr.h @@ -0,0 +1,60 @@ +/* + * linux/fs/hfsplus/xattr.h + * + * Vyacheslav Dubeyko + * + * Logic of processing extended attributes + */ + +#ifndef _LINUX_HFSPLUS_XATTR_H +#define _LINUX_HFSPLUS_XATTR_H + +#include + +extern const struct xattr_handler hfsplus_xattr_osx_handler; +extern const struct xattr_handler hfsplus_xattr_user_handler; +extern const struct xattr_handler hfsplus_xattr_trusted_handler; +/*extern const struct xattr_handler hfsplus_xattr_acl_access_handler;*/ +/*extern const struct xattr_handler hfsplus_xattr_acl_default_handler;*/ +extern const struct xattr_handler hfsplus_xattr_security_handler; + +extern const struct xattr_handler *hfsplus_xattr_handlers[]; + +int __hfsplus_setxattr(struct inode *inode, const char *name, + const void *value, size_t size, int flags); + +static inline int hfsplus_setxattr(struct dentry *dentry, const char *name, + const void *value, size_t size, int flags) +{ + return __hfsplus_setxattr(dentry->d_inode, name, value, size, flags); +} + +ssize_t hfsplus_getxattr(struct dentry *dentry, const char *name, + void *value, size_t size); + +ssize_t hfsplus_listxattr(struct dentry *dentry, char *buffer, size_t size); + +int hfsplus_removexattr(struct dentry *dentry, const char *name); + +int hfsplus_init_security(struct inode *inode, struct inode *dir, + const struct qstr *qstr); + +static inline int hfsplus_init_acl(struct inode *inode, struct inode *dir) +{ + /*TODO: implement*/ + return 0; +} + +static inline int hfsplus_init_inode_security(struct inode *inode, + struct inode *dir, + const struct qstr *qstr) +{ + int err; + + err = hfsplus_init_acl(inode, dir); + if (!err) + err = hfsplus_init_security(inode, dir, qstr); + return err; +} + +#endif diff --git a/fs/hfsplus/xattr_security.c b/fs/hfsplus/xattr_security.c new file mode 100644 index 000000000000..83b842f113c5 --- /dev/null +++ b/fs/hfsplus/xattr_security.c @@ -0,0 +1,104 @@ +/* + * linux/fs/hfsplus/xattr_trusted.c + * + * Vyacheslav Dubeyko + * + * Handler for storing security labels as extended attributes. + */ + +#include +#include "hfsplus_fs.h" +#include "xattr.h" + +static int hfsplus_security_getxattr(struct dentry *dentry, const char *name, + void *buffer, size_t size, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len + XATTR_SECURITY_PREFIX_LEN > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_SECURITY_PREFIX); + strcpy(xattr_name + XATTR_SECURITY_PREFIX_LEN, name); + + return hfsplus_getxattr(dentry, xattr_name, buffer, size); +} + +static int hfsplus_security_setxattr(struct dentry *dentry, const char *name, + const void *buffer, size_t size, int flags, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len + XATTR_SECURITY_PREFIX_LEN > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_SECURITY_PREFIX); + strcpy(xattr_name + XATTR_SECURITY_PREFIX_LEN, name); + + return hfsplus_setxattr(dentry, xattr_name, buffer, size, flags); +} + +static size_t hfsplus_security_listxattr(struct dentry *dentry, char *list, + size_t list_size, const char *name, size_t name_len, int type) +{ + /* + * This method is not used. + * It is used hfsplus_listxattr() instead of generic_listxattr(). + */ + return -EOPNOTSUPP; +} + +static int hfsplus_initxattrs(struct inode *inode, + const struct xattr *xattr_array, + void *fs_info) +{ + const struct xattr *xattr; + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + 1] = {0}; + size_t xattr_name_len; + int err = 0; + + for (xattr = xattr_array; xattr->name != NULL; xattr++) { + xattr_name_len = strlen(xattr->name); + + if (xattr_name_len == 0) + continue; + + if (xattr_name_len + XATTR_SECURITY_PREFIX_LEN > + HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_SECURITY_PREFIX); + strcpy(xattr_name + + XATTR_SECURITY_PREFIX_LEN, xattr->name); + memset(xattr_name + + XATTR_SECURITY_PREFIX_LEN + xattr_name_len, 0, 1); + + err = __hfsplus_setxattr(inode, xattr_name, + xattr->value, xattr->value_len, 0); + if (err) + break; + } + return err; +} + +int hfsplus_init_security(struct inode *inode, struct inode *dir, + const struct qstr *qstr) +{ + return security_inode_init_security(inode, dir, qstr, + &hfsplus_initxattrs, NULL); +} + +const struct xattr_handler hfsplus_xattr_security_handler = { + .prefix = XATTR_SECURITY_PREFIX, + .list = hfsplus_security_listxattr, + .get = hfsplus_security_getxattr, + .set = hfsplus_security_setxattr, +}; diff --git a/fs/hfsplus/xattr_trusted.c b/fs/hfsplus/xattr_trusted.c new file mode 100644 index 000000000000..426cee277542 --- /dev/null +++ b/fs/hfsplus/xattr_trusted.c @@ -0,0 +1,63 @@ +/* + * linux/fs/hfsplus/xattr_trusted.c + * + * Vyacheslav Dubeyko + * + * Handler for trusted extended attributes. + */ + +#include "hfsplus_fs.h" +#include "xattr.h" + +static int hfsplus_trusted_getxattr(struct dentry *dentry, const char *name, + void *buffer, size_t size, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len + XATTR_TRUSTED_PREFIX_LEN > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_TRUSTED_PREFIX); + strcpy(xattr_name + XATTR_TRUSTED_PREFIX_LEN, name); + + return hfsplus_getxattr(dentry, xattr_name, buffer, size); +} + +static int hfsplus_trusted_setxattr(struct dentry *dentry, const char *name, + const void *buffer, size_t size, int flags, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len + XATTR_TRUSTED_PREFIX_LEN > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_TRUSTED_PREFIX); + strcpy(xattr_name + XATTR_TRUSTED_PREFIX_LEN, name); + + return hfsplus_setxattr(dentry, xattr_name, buffer, size, flags); +} + +static size_t hfsplus_trusted_listxattr(struct dentry *dentry, char *list, + size_t list_size, const char *name, size_t name_len, int type) +{ + /* + * This method is not used. + * It is used hfsplus_listxattr() instead of generic_listxattr(). + */ + return -EOPNOTSUPP; +} + +const struct xattr_handler hfsplus_xattr_trusted_handler = { + .prefix = XATTR_TRUSTED_PREFIX, + .list = hfsplus_trusted_listxattr, + .get = hfsplus_trusted_getxattr, + .set = hfsplus_trusted_setxattr, +}; diff --git a/fs/hfsplus/xattr_user.c b/fs/hfsplus/xattr_user.c new file mode 100644 index 000000000000..e34016561ae0 --- /dev/null +++ b/fs/hfsplus/xattr_user.c @@ -0,0 +1,63 @@ +/* + * linux/fs/hfsplus/xattr_user.c + * + * Vyacheslav Dubeyko + * + * Handler for user extended attributes. + */ + +#include "hfsplus_fs.h" +#include "xattr.h" + +static int hfsplus_user_getxattr(struct dentry *dentry, const char *name, + void *buffer, size_t size, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len + XATTR_USER_PREFIX_LEN > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_USER_PREFIX); + strcpy(xattr_name + XATTR_USER_PREFIX_LEN, name); + + return hfsplus_getxattr(dentry, xattr_name, buffer, size); +} + +static int hfsplus_user_setxattr(struct dentry *dentry, const char *name, + const void *buffer, size_t size, int flags, int type) +{ + char xattr_name[HFSPLUS_ATTR_MAX_STRLEN + 1] = {0}; + size_t len = strlen(name); + + if (!strcmp(name, "")) + return -EINVAL; + + if (len + XATTR_USER_PREFIX_LEN > HFSPLUS_ATTR_MAX_STRLEN) + return -EOPNOTSUPP; + + strcpy(xattr_name, XATTR_USER_PREFIX); + strcpy(xattr_name + XATTR_USER_PREFIX_LEN, name); + + return hfsplus_setxattr(dentry, xattr_name, buffer, size, flags); +} + +static size_t hfsplus_user_listxattr(struct dentry *dentry, char *list, + size_t list_size, const char *name, size_t name_len, int type) +{ + /* + * This method is not used. + * It is used hfsplus_listxattr() instead of generic_listxattr(). + */ + return -EOPNOTSUPP; +} + +const struct xattr_handler hfsplus_xattr_user_handler = { + .prefix = XATTR_USER_PREFIX, + .list = hfsplus_user_listxattr, + .get = hfsplus_user_getxattr, + .set = hfsplus_user_setxattr, +}; -- cgit From 324ef39a8a4f693035d63527f16100ed27310ecc Mon Sep 17 00:00:00 2001 From: Vyacheslav Dubeyko Date: Wed, 27 Feb 2013 17:03:04 -0800 Subject: hfsplus: add support of manipulation by attributes file Add support of manipulation by attributes file. Signed-off-by: Vyacheslav Dubeyko Reported-by: Hin-Tak Leung Cc: Al Viro Cc: Christoph Hellwig Cc: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hfsplus/Makefile | 4 +- fs/hfsplus/bfind.c | 93 +++++++++++++++++++++++++++++++++++------ fs/hfsplus/bnode.c | 6 ++- fs/hfsplus/brec.c | 23 +++++++---- fs/hfsplus/btree.c | 8 ++++ fs/hfsplus/catalog.c | 36 +++++++++------- fs/hfsplus/dir.c | 55 +++++++++++++++++------- fs/hfsplus/extents.c | 4 +- fs/hfsplus/hfsplus_fs.h | 52 +++++++++++++++++++---- fs/hfsplus/inode.c | 18 +++++++- fs/hfsplus/ioctl.c | 108 ------------------------------------------------ fs/hfsplus/super.c | 56 +++++++++++++++++++++++-- fs/hfsplus/unicode.c | 7 ++-- 13 files changed, 287 insertions(+), 183 deletions(-) (limited to 'fs') diff --git a/fs/hfsplus/Makefile b/fs/hfsplus/Makefile index 3cc0df730156..09d278bb7b91 100644 --- a/fs/hfsplus/Makefile +++ b/fs/hfsplus/Makefile @@ -5,5 +5,5 @@ obj-$(CONFIG_HFSPLUS_FS) += hfsplus.o hfsplus-objs := super.o options.o inode.o ioctl.o extents.o catalog.o dir.o btree.o \ - bnode.o brec.o bfind.o tables.o unicode.o wrapper.o bitmap.o part_tbl.o - + bnode.o brec.o bfind.o tables.o unicode.o wrapper.o bitmap.o part_tbl.o \ + attributes.o xattr.o xattr_user.o xattr_security.o xattr_trusted.o diff --git a/fs/hfsplus/bfind.c b/fs/hfsplus/bfind.c index 5d799c13205f..d73c98d1ee99 100644 --- a/fs/hfsplus/bfind.c +++ b/fs/hfsplus/bfind.c @@ -24,7 +24,19 @@ int hfs_find_init(struct hfs_btree *tree, struct hfs_find_data *fd) fd->key = ptr + tree->max_key_len + 2; dprint(DBG_BNODE_REFS, "find_init: %d (%p)\n", tree->cnid, __builtin_return_address(0)); - mutex_lock(&tree->tree_lock); + switch (tree->cnid) { + case HFSPLUS_CAT_CNID: + mutex_lock_nested(&tree->tree_lock, CATALOG_BTREE_MUTEX); + break; + case HFSPLUS_EXT_CNID: + mutex_lock_nested(&tree->tree_lock, EXTENTS_BTREE_MUTEX); + break; + case HFSPLUS_ATTR_CNID: + mutex_lock_nested(&tree->tree_lock, ATTR_BTREE_MUTEX); + break; + default: + BUG(); + } return 0; } @@ -38,15 +50,73 @@ void hfs_find_exit(struct hfs_find_data *fd) fd->tree = NULL; } -/* Find the record in bnode that best matches key (not greater than...)*/ -int __hfs_brec_find(struct hfs_bnode *bnode, struct hfs_find_data *fd) +int hfs_find_1st_rec_by_cnid(struct hfs_bnode *bnode, + struct hfs_find_data *fd, + int *begin, + int *end, + int *cur_rec) +{ + __be32 cur_cnid, search_cnid; + + if (bnode->tree->cnid == HFSPLUS_EXT_CNID) { + cur_cnid = fd->key->ext.cnid; + search_cnid = fd->search_key->ext.cnid; + } else if (bnode->tree->cnid == HFSPLUS_CAT_CNID) { + cur_cnid = fd->key->cat.parent; + search_cnid = fd->search_key->cat.parent; + } else if (bnode->tree->cnid == HFSPLUS_ATTR_CNID) { + cur_cnid = fd->key->attr.cnid; + search_cnid = fd->search_key->attr.cnid; + } else + BUG(); + + if (cur_cnid == search_cnid) { + (*end) = (*cur_rec); + if ((*begin) == (*end)) + return 1; + } else { + if (be32_to_cpu(cur_cnid) < be32_to_cpu(search_cnid)) + (*begin) = (*cur_rec) + 1; + else + (*end) = (*cur_rec) - 1; + } + + return 0; +} + +int hfs_find_rec_by_key(struct hfs_bnode *bnode, + struct hfs_find_data *fd, + int *begin, + int *end, + int *cur_rec) { int cmpval; + + cmpval = bnode->tree->keycmp(fd->key, fd->search_key); + if (!cmpval) { + (*end) = (*cur_rec); + return 1; + } + if (cmpval < 0) + (*begin) = (*cur_rec) + 1; + else + *(end) = (*cur_rec) - 1; + + return 0; +} + +/* Find the record in bnode that best matches key (not greater than...)*/ +int __hfs_brec_find(struct hfs_bnode *bnode, struct hfs_find_data *fd, + search_strategy_t rec_found) +{ u16 off, len, keylen; int rec; int b, e; int res; + if (!rec_found) + BUG(); + b = 0; e = bnode->num_recs - 1; res = -ENOENT; @@ -59,17 +129,12 @@ int __hfs_brec_find(struct hfs_bnode *bnode, struct hfs_find_data *fd) goto fail; } hfs_bnode_read(bnode, fd->key, off, keylen); - cmpval = bnode->tree->keycmp(fd->key, fd->search_key); - if (!cmpval) { - e = rec; + if (rec_found(bnode, fd, &b, &e, &rec)) { res = 0; goto done; } - if (cmpval < 0) - b = rec + 1; - else - e = rec - 1; } while (b <= e); + if (rec != e && e >= 0) { len = hfs_brec_lenoff(bnode, e, &off); keylen = hfs_brec_keylen(bnode, e); @@ -79,19 +144,21 @@ int __hfs_brec_find(struct hfs_bnode *bnode, struct hfs_find_data *fd) } hfs_bnode_read(bnode, fd->key, off, keylen); } + done: fd->record = e; fd->keyoffset = off; fd->keylength = keylen; fd->entryoffset = off + keylen; fd->entrylength = len - keylen; + fail: return res; } /* Traverse a B*Tree from the root to a leaf finding best fit to key */ /* Return allocated copy of node found, set recnum to best record */ -int hfs_brec_find(struct hfs_find_data *fd) +int hfs_brec_find(struct hfs_find_data *fd, search_strategy_t do_key_compare) { struct hfs_btree *tree; struct hfs_bnode *bnode; @@ -122,7 +189,7 @@ int hfs_brec_find(struct hfs_find_data *fd) goto invalid; bnode->parent = parent; - res = __hfs_brec_find(bnode, fd); + res = __hfs_brec_find(bnode, fd, do_key_compare); if (!height) break; if (fd->record < 0) @@ -149,7 +216,7 @@ int hfs_brec_read(struct hfs_find_data *fd, void *rec, int rec_len) { int res; - res = hfs_brec_find(fd); + res = hfs_brec_find(fd, hfs_find_rec_by_key); if (res) return res; if (fd->entrylength > rec_len) diff --git a/fs/hfsplus/bnode.c b/fs/hfsplus/bnode.c index 1c42cc5b899f..5c125ce6bd72 100644 --- a/fs/hfsplus/bnode.c +++ b/fs/hfsplus/bnode.c @@ -62,7 +62,8 @@ void hfs_bnode_read_key(struct hfs_bnode *node, void *key, int off) tree = node->tree; if (node->type == HFS_NODE_LEAF || - tree->attributes & HFS_TREE_VARIDXKEYS) + tree->attributes & HFS_TREE_VARIDXKEYS || + node->tree->cnid == HFSPLUS_ATTR_CNID) key_len = hfs_bnode_read_u16(node, off) + 2; else key_len = tree->max_key_len + 2; @@ -314,7 +315,8 @@ void hfs_bnode_dump(struct hfs_bnode *node) if (i && node->type == HFS_NODE_INDEX) { int tmp; - if (node->tree->attributes & HFS_TREE_VARIDXKEYS) + if (node->tree->attributes & HFS_TREE_VARIDXKEYS || + node->tree->cnid == HFSPLUS_ATTR_CNID) tmp = hfs_bnode_read_u16(node, key_off) + 2; else tmp = node->tree->max_key_len + 2; diff --git a/fs/hfsplus/brec.c b/fs/hfsplus/brec.c index 2a734cfccc92..298d4e45604b 100644 --- a/fs/hfsplus/brec.c +++ b/fs/hfsplus/brec.c @@ -36,7 +36,8 @@ u16 hfs_brec_keylen(struct hfs_bnode *node, u16 rec) return 0; if ((node->type == HFS_NODE_INDEX) && - !(node->tree->attributes & HFS_TREE_VARIDXKEYS)) { + !(node->tree->attributes & HFS_TREE_VARIDXKEYS) && + (node->tree->cnid != HFSPLUS_ATTR_CNID)) { retval = node->tree->max_key_len + 2; } else { recoff = hfs_bnode_read_u16(node, @@ -151,12 +152,13 @@ skip: /* get index key */ hfs_bnode_read_key(new_node, fd->search_key, 14); - __hfs_brec_find(fd->bnode, fd); + __hfs_brec_find(fd->bnode, fd, hfs_find_rec_by_key); hfs_bnode_put(new_node); new_node = NULL; - if (tree->attributes & HFS_TREE_VARIDXKEYS) + if ((tree->attributes & HFS_TREE_VARIDXKEYS) || + (tree->cnid == HFSPLUS_ATTR_CNID)) key_len = be16_to_cpu(fd->search_key->key_len) + 2; else { fd->search_key->key_len = @@ -201,7 +203,7 @@ again: hfs_bnode_put(node); node = fd->bnode = parent; - __hfs_brec_find(node, fd); + __hfs_brec_find(node, fd, hfs_find_rec_by_key); goto again; } hfs_bnode_write_u16(node, @@ -367,12 +369,13 @@ again: parent = hfs_bnode_find(tree, node->parent); if (IS_ERR(parent)) return PTR_ERR(parent); - __hfs_brec_find(parent, fd); + __hfs_brec_find(parent, fd, hfs_find_rec_by_key); hfs_bnode_dump(parent); rec = fd->record; /* size difference between old and new key */ - if (tree->attributes & HFS_TREE_VARIDXKEYS) + if ((tree->attributes & HFS_TREE_VARIDXKEYS) || + (tree->cnid == HFSPLUS_ATTR_CNID)) newkeylen = hfs_bnode_read_u16(node, 14) + 2; else fd->keylength = newkeylen = tree->max_key_len + 2; @@ -427,7 +430,7 @@ skip: hfs_bnode_read_key(new_node, fd->search_key, 14); cnid = cpu_to_be32(new_node->this); - __hfs_brec_find(fd->bnode, fd); + __hfs_brec_find(fd->bnode, fd, hfs_find_rec_by_key); hfs_brec_insert(fd, &cnid, sizeof(cnid)); hfs_bnode_put(fd->bnode); hfs_bnode_put(new_node); @@ -495,13 +498,15 @@ static int hfs_btree_inc_height(struct hfs_btree *tree) /* insert old root idx into new root */ node->parent = tree->root; if (node->type == HFS_NODE_LEAF || - tree->attributes & HFS_TREE_VARIDXKEYS) + tree->attributes & HFS_TREE_VARIDXKEYS || + tree->cnid == HFSPLUS_ATTR_CNID) key_size = hfs_bnode_read_u16(node, 14) + 2; else key_size = tree->max_key_len + 2; hfs_bnode_copy(new_node, 14, node, 14, key_size); - if (!(tree->attributes & HFS_TREE_VARIDXKEYS)) { + if (!(tree->attributes & HFS_TREE_VARIDXKEYS) && + (tree->cnid != HFSPLUS_ATTR_CNID)) { key_size = tree->max_key_len + 2; hfs_bnode_write_u16(new_node, 14, tree->max_key_len); } diff --git a/fs/hfsplus/btree.c b/fs/hfsplus/btree.c index 685d07d0ed18..efb689c21a95 100644 --- a/fs/hfsplus/btree.c +++ b/fs/hfsplus/btree.c @@ -98,6 +98,14 @@ struct hfs_btree *hfs_btree_open(struct super_block *sb, u32 id) set_bit(HFSPLUS_SB_CASEFOLD, &HFSPLUS_SB(sb)->flags); } break; + case HFSPLUS_ATTR_CNID: + if (tree->max_key_len != HFSPLUS_ATTR_KEYLEN - sizeof(u16)) { + printk(KERN_ERR "hfs: invalid attributes max_key_len %d\n", + tree->max_key_len); + goto fail_page; + } + tree->keycmp = hfsplus_attr_bin_cmp_key; + break; default: printk(KERN_ERR "hfs: unknown B*Tree requested\n"); goto fail_page; diff --git a/fs/hfsplus/catalog.c b/fs/hfsplus/catalog.c index 798d9c4c5e71..840d71edd193 100644 --- a/fs/hfsplus/catalog.c +++ b/fs/hfsplus/catalog.c @@ -45,7 +45,8 @@ void hfsplus_cat_build_key(struct super_block *sb, hfsplus_btree_key *key, key->cat.parent = cpu_to_be32(parent); if (str) { - hfsplus_asc2uni(sb, &key->cat.name, str->name, str->len); + hfsplus_asc2uni(sb, &key->cat.name, HFSPLUS_MAX_STRLEN, + str->name, str->len); len = be16_to_cpu(key->cat.name.length); } else { key->cat.name.length = 0; @@ -167,7 +168,8 @@ static int hfsplus_fill_cat_thread(struct super_block *sb, entry->type = cpu_to_be16(type); entry->thread.reserved = 0; entry->thread.parentID = cpu_to_be32(parentid); - hfsplus_asc2uni(sb, &entry->thread.nodeName, str->name, str->len); + hfsplus_asc2uni(sb, &entry->thread.nodeName, HFSPLUS_MAX_STRLEN, + str->name, str->len); return 10 + be16_to_cpu(entry->thread.nodeName.length) * 2; } @@ -198,7 +200,7 @@ int hfsplus_find_cat(struct super_block *sb, u32 cnid, hfsplus_cat_build_key_uni(fd->search_key, be32_to_cpu(tmp.thread.parentID), &tmp.thread.nodeName); - return hfs_brec_find(fd); + return hfs_brec_find(fd, hfs_find_rec_by_key); } int hfsplus_create_cat(u32 cnid, struct inode *dir, @@ -221,7 +223,7 @@ int hfsplus_create_cat(u32 cnid, struct inode *dir, S_ISDIR(inode->i_mode) ? HFSPLUS_FOLDER_THREAD : HFSPLUS_FILE_THREAD, dir->i_ino, str); - err = hfs_brec_find(&fd); + err = hfs_brec_find(&fd, hfs_find_rec_by_key); if (err != -ENOENT) { if (!err) err = -EEXIST; @@ -233,7 +235,7 @@ int hfsplus_create_cat(u32 cnid, struct inode *dir, hfsplus_cat_build_key(sb, fd.search_key, dir->i_ino, str); entry_size = hfsplus_cat_build_record(&entry, cnid, inode); - err = hfs_brec_find(&fd); + err = hfs_brec_find(&fd, hfs_find_rec_by_key); if (err != -ENOENT) { /* panic? */ if (!err) @@ -253,7 +255,7 @@ int hfsplus_create_cat(u32 cnid, struct inode *dir, err1: hfsplus_cat_build_key(sb, fd.search_key, cnid, NULL); - if (!hfs_brec_find(&fd)) + if (!hfs_brec_find(&fd, hfs_find_rec_by_key)) hfs_brec_remove(&fd); err2: hfs_find_exit(&fd); @@ -279,7 +281,7 @@ int hfsplus_delete_cat(u32 cnid, struct inode *dir, struct qstr *str) int len; hfsplus_cat_build_key(sb, fd.search_key, cnid, NULL); - err = hfs_brec_find(&fd); + err = hfs_brec_find(&fd, hfs_find_rec_by_key); if (err) goto out; @@ -296,7 +298,7 @@ int hfsplus_delete_cat(u32 cnid, struct inode *dir, struct qstr *str) } else hfsplus_cat_build_key(sb, fd.search_key, dir->i_ino, str); - err = hfs_brec_find(&fd); + err = hfs_brec_find(&fd, hfs_find_rec_by_key); if (err) goto out; @@ -326,7 +328,7 @@ int hfsplus_delete_cat(u32 cnid, struct inode *dir, struct qstr *str) goto out; hfsplus_cat_build_key(sb, fd.search_key, cnid, NULL); - err = hfs_brec_find(&fd); + err = hfs_brec_find(&fd, hfs_find_rec_by_key); if (err) goto out; @@ -337,6 +339,12 @@ int hfsplus_delete_cat(u32 cnid, struct inode *dir, struct qstr *str) dir->i_size--; dir->i_mtime = dir->i_ctime = CURRENT_TIME_SEC; hfsplus_mark_inode_dirty(dir, HFSPLUS_I_CAT_DIRTY); + + if (type == HFSPLUS_FILE || type == HFSPLUS_FOLDER) { + if (HFSPLUS_SB(sb)->attr_tree) + hfsplus_delete_all_attrs(dir, cnid); + } + out: hfs_find_exit(&fd); @@ -363,7 +371,7 @@ int hfsplus_rename_cat(u32 cnid, /* find the old dir entry and read the data */ hfsplus_cat_build_key(sb, src_fd.search_key, src_dir->i_ino, src_name); - err = hfs_brec_find(&src_fd); + err = hfs_brec_find(&src_fd, hfs_find_rec_by_key); if (err) goto out; if (src_fd.entrylength > sizeof(entry) || src_fd.entrylength < 0) { @@ -376,7 +384,7 @@ int hfsplus_rename_cat(u32 cnid, /* create new dir entry with the data from the old entry */ hfsplus_cat_build_key(sb, dst_fd.search_key, dst_dir->i_ino, dst_name); - err = hfs_brec_find(&dst_fd); + err = hfs_brec_find(&dst_fd, hfs_find_rec_by_key); if (err != -ENOENT) { if (!err) err = -EEXIST; @@ -391,7 +399,7 @@ int hfsplus_rename_cat(u32 cnid, /* finally remove the old entry */ hfsplus_cat_build_key(sb, src_fd.search_key, src_dir->i_ino, src_name); - err = hfs_brec_find(&src_fd); + err = hfs_brec_find(&src_fd, hfs_find_rec_by_key); if (err) goto out; err = hfs_brec_remove(&src_fd); @@ -402,7 +410,7 @@ int hfsplus_rename_cat(u32 cnid, /* remove old thread entry */ hfsplus_cat_build_key(sb, src_fd.search_key, cnid, NULL); - err = hfs_brec_find(&src_fd); + err = hfs_brec_find(&src_fd, hfs_find_rec_by_key); if (err) goto out; type = hfs_bnode_read_u16(src_fd.bnode, src_fd.entryoffset); @@ -414,7 +422,7 @@ int hfsplus_rename_cat(u32 cnid, hfsplus_cat_build_key(sb, dst_fd.search_key, cnid, NULL); entry_size = hfsplus_fill_cat_thread(sb, &entry, type, dst_dir->i_ino, dst_name); - err = hfs_brec_find(&dst_fd); + err = hfs_brec_find(&dst_fd, hfs_find_rec_by_key); if (err != -ENOENT) { if (!err) err = -EEXIST; diff --git a/fs/hfsplus/dir.c b/fs/hfsplus/dir.c index 074e04589248..031c24e50521 100644 --- a/fs/hfsplus/dir.c +++ b/fs/hfsplus/dir.c @@ -15,6 +15,7 @@ #include "hfsplus_fs.h" #include "hfsplus_raw.h" +#include "xattr.h" static inline void hfsplus_instantiate(struct dentry *dentry, struct inode *inode, u32 cnid) @@ -138,7 +139,7 @@ static int hfsplus_readdir(struct file *filp, void *dirent, filldir_t filldir) if (err) return err; hfsplus_cat_build_key(sb, fd.search_key, inode->i_ino, NULL); - err = hfs_brec_find(&fd); + err = hfs_brec_find(&fd, hfs_find_rec_by_key); if (err) goto out; @@ -421,6 +422,15 @@ static int hfsplus_symlink(struct inode *dir, struct dentry *dentry, if (res) goto out_err; + res = hfsplus_init_inode_security(inode, dir, &dentry->d_name); + if (res == -EOPNOTSUPP) + res = 0; /* Operation is not supported. */ + else if (res) { + /* Try to delete anyway without error analysis. */ + hfsplus_delete_cat(inode->i_ino, dir, &dentry->d_name); + goto out_err; + } + hfsplus_instantiate(dentry, inode, inode->i_ino); mark_inode_dirty(inode); goto out; @@ -450,15 +460,26 @@ static int hfsplus_mknod(struct inode *dir, struct dentry *dentry, init_special_inode(inode, mode, rdev); res = hfsplus_create_cat(inode->i_ino, dir, &dentry->d_name, inode); - if (res) { - clear_nlink(inode); - hfsplus_delete_inode(inode); - iput(inode); - goto out; + if (res) + goto failed_mknod; + + res = hfsplus_init_inode_security(inode, dir, &dentry->d_name); + if (res == -EOPNOTSUPP) + res = 0; /* Operation is not supported. */ + else if (res) { + /* Try to delete anyway without error analysis. */ + hfsplus_delete_cat(inode->i_ino, dir, &dentry->d_name); + goto failed_mknod; } hfsplus_instantiate(dentry, inode, inode->i_ino); mark_inode_dirty(inode); + goto out; + +failed_mknod: + clear_nlink(inode); + hfsplus_delete_inode(inode); + iput(inode); out: mutex_unlock(&sbi->vh_mutex); return res; @@ -499,15 +520,19 @@ static int hfsplus_rename(struct inode *old_dir, struct dentry *old_dentry, } const struct inode_operations hfsplus_dir_inode_operations = { - .lookup = hfsplus_lookup, - .create = hfsplus_create, - .link = hfsplus_link, - .unlink = hfsplus_unlink, - .mkdir = hfsplus_mkdir, - .rmdir = hfsplus_rmdir, - .symlink = hfsplus_symlink, - .mknod = hfsplus_mknod, - .rename = hfsplus_rename, + .lookup = hfsplus_lookup, + .create = hfsplus_create, + .link = hfsplus_link, + .unlink = hfsplus_unlink, + .mkdir = hfsplus_mkdir, + .rmdir = hfsplus_rmdir, + .symlink = hfsplus_symlink, + .mknod = hfsplus_mknod, + .rename = hfsplus_rename, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .listxattr = hfsplus_listxattr, + .removexattr = hfsplus_removexattr, }; const struct file_operations hfsplus_dir_operations = { diff --git a/fs/hfsplus/extents.c b/fs/hfsplus/extents.c index eba76eab6d62..a94f0f779d5e 100644 --- a/fs/hfsplus/extents.c +++ b/fs/hfsplus/extents.c @@ -95,7 +95,7 @@ static void __hfsplus_ext_write_extent(struct inode *inode, HFSPLUS_IS_RSRC(inode) ? HFSPLUS_TYPE_RSRC : HFSPLUS_TYPE_DATA); - res = hfs_brec_find(fd); + res = hfs_brec_find(fd, hfs_find_rec_by_key); if (hip->extent_state & HFSPLUS_EXT_NEW) { if (res != -ENOENT) return; @@ -154,7 +154,7 @@ static inline int __hfsplus_ext_read_extent(struct hfs_find_data *fd, hfsplus_ext_build_key(fd->search_key, cnid, block, type); fd->key->ext.cnid = 0; - res = hfs_brec_find(fd); + res = hfs_brec_find(fd, hfs_find_rec_by_key); if (res && res != -ENOENT) return res; if (fd->key->ext.cnid != fd->search_key->ext.cnid || diff --git a/fs/hfsplus/hfsplus_fs.h b/fs/hfsplus/hfsplus_fs.h index a6da86b1b4c1..05b11f36024c 100644 --- a/fs/hfsplus/hfsplus_fs.h +++ b/fs/hfsplus/hfsplus_fs.h @@ -23,6 +23,7 @@ #define DBG_SUPER 0x00000010 #define DBG_EXTENT 0x00000020 #define DBG_BITMAP 0x00000040 +#define DBG_ATTR_MOD 0x00000080 #if 0 #define DBG_MASK (DBG_EXTENT|DBG_INODE|DBG_BNODE_MOD) @@ -46,6 +47,13 @@ typedef int (*btree_keycmp)(const hfsplus_btree_key *, #define NODE_HASH_SIZE 256 +/* B-tree mutex nested subclasses */ +enum hfsplus_btree_mutex_classes { + CATALOG_BTREE_MUTEX, + EXTENTS_BTREE_MUTEX, + ATTR_BTREE_MUTEX, +}; + /* An HFS+ BTree held in memory */ struct hfs_btree { struct super_block *sb; @@ -223,6 +231,7 @@ struct hfsplus_inode_info { #define HFSPLUS_I_CAT_DIRTY 1 /* has changes in the catalog tree */ #define HFSPLUS_I_EXT_DIRTY 2 /* has changes in the extent tree */ #define HFSPLUS_I_ALLOC_DIRTY 3 /* has changes in the allocation file */ +#define HFSPLUS_I_ATTR_DIRTY 4 /* has changes in the attributes tree */ #define HFSPLUS_IS_RSRC(inode) \ test_bit(HFSPLUS_I_RSRC, &HFSPLUS_I(inode)->flags) @@ -302,7 +311,7 @@ static inline unsigned short hfsplus_min_io_size(struct super_block *sb) #define hfs_brec_remove hfsplus_brec_remove #define hfs_find_init hfsplus_find_init #define hfs_find_exit hfsplus_find_exit -#define __hfs_brec_find __hplusfs_brec_find +#define __hfs_brec_find __hfsplus_brec_find #define hfs_brec_find hfsplus_brec_find #define hfs_brec_read hfsplus_brec_read #define hfs_brec_goto hfsplus_brec_goto @@ -324,10 +333,33 @@ static inline unsigned short hfsplus_min_io_size(struct super_block *sb) */ #define HFSPLUS_IOC_BLESS _IO('h', 0x80) +typedef int (*search_strategy_t)(struct hfs_bnode *, + struct hfs_find_data *, + int *, int *, int *); + /* * Functions in any *.c used in other files */ +/* attributes.c */ +int hfsplus_create_attr_tree_cache(void); +void hfsplus_destroy_attr_tree_cache(void); +hfsplus_attr_entry *hfsplus_alloc_attr_entry(void); +void hfsplus_destroy_attr_entry(hfsplus_attr_entry *entry_p); +int hfsplus_attr_bin_cmp_key(const hfsplus_btree_key *, + const hfsplus_btree_key *); +int hfsplus_attr_build_key(struct super_block *, hfsplus_btree_key *, + u32, const char *); +void hfsplus_attr_build_key_uni(hfsplus_btree_key *key, + u32 cnid, + struct hfsplus_attr_unistr *name); +int hfsplus_find_attr(struct super_block *, u32, + const char *, struct hfs_find_data *); +int hfsplus_attr_exists(struct inode *inode, const char *name); +int hfsplus_create_attr(struct inode *, const char *, const void *, size_t); +int hfsplus_delete_attr(struct inode *, const char *); +int hfsplus_delete_all_attrs(struct inode *dir, u32 cnid); + /* bitmap.c */ int hfsplus_block_allocate(struct super_block *, u32, u32, u32 *); int hfsplus_block_free(struct super_block *, u32, u32); @@ -369,8 +401,15 @@ int hfs_brec_remove(struct hfs_find_data *); /* bfind.c */ int hfs_find_init(struct hfs_btree *, struct hfs_find_data *); void hfs_find_exit(struct hfs_find_data *); -int __hfs_brec_find(struct hfs_bnode *, struct hfs_find_data *); -int hfs_brec_find(struct hfs_find_data *); +int hfs_find_1st_rec_by_cnid(struct hfs_bnode *, + struct hfs_find_data *, + int *, int *, int *); +int hfs_find_rec_by_key(struct hfs_bnode *, + struct hfs_find_data *, + int *, int *, int *); +int __hfs_brec_find(struct hfs_bnode *, struct hfs_find_data *, + search_strategy_t); +int hfs_brec_find(struct hfs_find_data *, search_strategy_t); int hfs_brec_read(struct hfs_find_data *, void *, int); int hfs_brec_goto(struct hfs_find_data *, int); @@ -417,11 +456,6 @@ int hfsplus_file_fsync(struct file *file, loff_t start, loff_t end, /* ioctl.c */ long hfsplus_ioctl(struct file *filp, unsigned int cmd, unsigned long arg); -int hfsplus_setxattr(struct dentry *dentry, const char *name, - const void *value, size_t size, int flags); -ssize_t hfsplus_getxattr(struct dentry *dentry, const char *name, - void *value, size_t size); -ssize_t hfsplus_listxattr(struct dentry *dentry, char *buffer, size_t size); /* options.c */ int hfsplus_parse_options(char *, struct hfsplus_sb_info *); @@ -446,7 +480,7 @@ int hfsplus_strcmp(const struct hfsplus_unistr *, int hfsplus_uni2asc(struct super_block *, const struct hfsplus_unistr *, char *, int *); int hfsplus_asc2uni(struct super_block *, - struct hfsplus_unistr *, const char *, int); + struct hfsplus_unistr *, int, const char *, int); int hfsplus_hash_dentry(const struct dentry *dentry, const struct inode *inode, struct qstr *str); int hfsplus_compare_dentry(const struct dentry *parent, diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index dcd05be5344b..160ccc9cdb4b 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -17,6 +17,7 @@ #include "hfsplus_fs.h" #include "hfsplus_raw.h" +#include "xattr.h" static int hfsplus_readpage(struct file *file, struct page *page) { @@ -348,6 +349,18 @@ int hfsplus_file_fsync(struct file *file, loff_t start, loff_t end, error = error2; } + if (test_and_clear_bit(HFSPLUS_I_ATTR_DIRTY, &hip->flags)) { + if (sbi->attr_tree) { + error2 = + filemap_write_and_wait( + sbi->attr_tree->inode->i_mapping); + if (!error) + error = error2; + } else { + printk(KERN_ERR "hfs: sync non-existent attributes tree\n"); + } + } + if (test_and_clear_bit(HFSPLUS_I_ALLOC_DIRTY, &hip->flags)) { error2 = filemap_write_and_wait(sbi->alloc_file->i_mapping); if (!error) @@ -365,9 +378,10 @@ int hfsplus_file_fsync(struct file *file, loff_t start, loff_t end, static const struct inode_operations hfsplus_file_inode_operations = { .lookup = hfsplus_file_lookup, .setattr = hfsplus_setattr, - .setxattr = hfsplus_setxattr, - .getxattr = hfsplus_getxattr, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, .listxattr = hfsplus_listxattr, + .removexattr = hfsplus_removexattr, }; static const struct file_operations hfsplus_file_operations = { diff --git a/fs/hfsplus/ioctl.c b/fs/hfsplus/ioctl.c index e3c4c4209428..d3ff5cc317d7 100644 --- a/fs/hfsplus/ioctl.c +++ b/fs/hfsplus/ioctl.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include "hfsplus_fs.h" @@ -151,110 +150,3 @@ long hfsplus_ioctl(struct file *file, unsigned int cmd, unsigned long arg) return -ENOTTY; } } - -int hfsplus_setxattr(struct dentry *dentry, const char *name, - const void *value, size_t size, int flags) -{ - struct inode *inode = dentry->d_inode; - struct hfs_find_data fd; - hfsplus_cat_entry entry; - struct hfsplus_cat_file *file; - int res; - - if (!S_ISREG(inode->i_mode) || HFSPLUS_IS_RSRC(inode)) - return -EOPNOTSUPP; - - res = hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &fd); - if (res) - return res; - res = hfsplus_find_cat(inode->i_sb, inode->i_ino, &fd); - if (res) - goto out; - hfs_bnode_read(fd.bnode, &entry, fd.entryoffset, - sizeof(struct hfsplus_cat_file)); - file = &entry.file; - - if (!strcmp(name, "hfs.type")) { - if (size == 4) - memcpy(&file->user_info.fdType, value, 4); - else - res = -ERANGE; - } else if (!strcmp(name, "hfs.creator")) { - if (size == 4) - memcpy(&file->user_info.fdCreator, value, 4); - else - res = -ERANGE; - } else - res = -EOPNOTSUPP; - if (!res) { - hfs_bnode_write(fd.bnode, &entry, fd.entryoffset, - sizeof(struct hfsplus_cat_file)); - hfsplus_mark_inode_dirty(inode, HFSPLUS_I_CAT_DIRTY); - } -out: - hfs_find_exit(&fd); - return res; -} - -ssize_t hfsplus_getxattr(struct dentry *dentry, const char *name, - void *value, size_t size) -{ - struct inode *inode = dentry->d_inode; - struct hfs_find_data fd; - hfsplus_cat_entry entry; - struct hfsplus_cat_file *file; - ssize_t res = 0; - - if (!S_ISREG(inode->i_mode) || HFSPLUS_IS_RSRC(inode)) - return -EOPNOTSUPP; - - if (size) { - res = hfs_find_init(HFSPLUS_SB(inode->i_sb)->cat_tree, &fd); - if (res) - return res; - res = hfsplus_find_cat(inode->i_sb, inode->i_ino, &fd); - if (res) - goto out; - hfs_bnode_read(fd.bnode, &entry, fd.entryoffset, - sizeof(struct hfsplus_cat_file)); - } - file = &entry.file; - - if (!strcmp(name, "hfs.type")) { - if (size >= 4) { - memcpy(value, &file->user_info.fdType, 4); - res = 4; - } else - res = size ? -ERANGE : 4; - } else if (!strcmp(name, "hfs.creator")) { - if (size >= 4) { - memcpy(value, &file->user_info.fdCreator, 4); - res = 4; - } else - res = size ? -ERANGE : 4; - } else - res = -EOPNOTSUPP; -out: - if (size) - hfs_find_exit(&fd); - return res; -} - -#define HFSPLUS_ATTRLIST_SIZE (sizeof("hfs.creator")+sizeof("hfs.type")) - -ssize_t hfsplus_listxattr(struct dentry *dentry, char *buffer, size_t size) -{ - struct inode *inode = dentry->d_inode; - - if (!S_ISREG(inode->i_mode) || HFSPLUS_IS_RSRC(inode)) - return -EOPNOTSUPP; - - if (!buffer || !size) - return HFSPLUS_ATTRLIST_SIZE; - if (size < HFSPLUS_ATTRLIST_SIZE) - return -ERANGE; - strcpy(buffer, "hfs.type"); - strcpy(buffer + sizeof("hfs.type"), "hfs.creator"); - - return HFSPLUS_ATTRLIST_SIZE; -} diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 796198d26553..974c26f96fae 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -20,6 +20,7 @@ static struct inode *hfsplus_alloc_inode(struct super_block *sb); static void hfsplus_destroy_inode(struct inode *inode); #include "hfsplus_fs.h" +#include "xattr.h" static int hfsplus_system_read_inode(struct inode *inode) { @@ -118,6 +119,7 @@ static int hfsplus_system_write_inode(struct inode *inode) case HFSPLUS_ATTR_CNID: fork = &vhdr->attr_file; tree = sbi->attr_tree; + break; default: return -EIO; } @@ -191,6 +193,12 @@ static int hfsplus_sync_fs(struct super_block *sb, int wait) error2 = filemap_write_and_wait(sbi->ext_tree->inode->i_mapping); if (!error) error = error2; + if (sbi->attr_tree) { + error2 = + filemap_write_and_wait(sbi->attr_tree->inode->i_mapping); + if (!error) + error = error2; + } error2 = filemap_write_and_wait(sbi->alloc_file->i_mapping); if (!error) error = error2; @@ -281,6 +289,7 @@ static void hfsplus_put_super(struct super_block *sb) hfsplus_sync_fs(sb, 1); } + hfs_btree_close(sbi->attr_tree); hfs_btree_close(sbi->cat_tree); hfs_btree_close(sbi->ext_tree); iput(sbi->alloc_file); @@ -477,12 +486,20 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) printk(KERN_ERR "hfs: failed to load catalog file\n"); goto out_close_ext_tree; } + if (vhdr->attr_file.total_blocks != 0) { + sbi->attr_tree = hfs_btree_open(sb, HFSPLUS_ATTR_CNID); + if (!sbi->attr_tree) { + printk(KERN_ERR "hfs: failed to load attributes file\n"); + goto out_close_cat_tree; + } + } + sb->s_xattr = hfsplus_xattr_handlers; inode = hfsplus_iget(sb, HFSPLUS_ALLOC_CNID); if (IS_ERR(inode)) { printk(KERN_ERR "hfs: failed to load allocation file\n"); err = PTR_ERR(inode); - goto out_close_cat_tree; + goto out_close_attr_tree; } sbi->alloc_file = inode; @@ -542,10 +559,27 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) } err = hfsplus_create_cat(sbi->hidden_dir->i_ino, root, &str, sbi->hidden_dir); - mutex_unlock(&sbi->vh_mutex); - if (err) + if (err) { + mutex_unlock(&sbi->vh_mutex); + goto out_put_hidden_dir; + } + + err = hfsplus_init_inode_security(sbi->hidden_dir, + root, &str); + if (err == -EOPNOTSUPP) + err = 0; /* Operation is not supported. */ + else if (err) { + /* + * Try to delete anyway without + * error analysis. + */ + hfsplus_delete_cat(sbi->hidden_dir->i_ino, + root, &str); + mutex_unlock(&sbi->vh_mutex); goto out_put_hidden_dir; + } + mutex_unlock(&sbi->vh_mutex); hfsplus_mark_inode_dirty(sbi->hidden_dir, HFSPLUS_I_CAT_DIRTY); } @@ -562,6 +596,8 @@ out_put_root: sb->s_root = NULL; out_put_alloc_file: iput(sbi->alloc_file); +out_close_attr_tree: + hfs_btree_close(sbi->attr_tree); out_close_cat_tree: hfs_btree_close(sbi->cat_tree); out_close_ext_tree: @@ -635,9 +671,20 @@ static int __init init_hfsplus_fs(void) hfsplus_init_once); if (!hfsplus_inode_cachep) return -ENOMEM; + err = hfsplus_create_attr_tree_cache(); + if (err) + goto destroy_inode_cache; err = register_filesystem(&hfsplus_fs_type); if (err) - kmem_cache_destroy(hfsplus_inode_cachep); + goto destroy_attr_tree_cache; + return 0; + +destroy_attr_tree_cache: + hfsplus_destroy_attr_tree_cache(); + +destroy_inode_cache: + kmem_cache_destroy(hfsplus_inode_cachep); + return err; } @@ -650,6 +697,7 @@ static void __exit exit_hfsplus_fs(void) * destroy cache. */ rcu_barrier(); + hfsplus_destroy_attr_tree_cache(); kmem_cache_destroy(hfsplus_inode_cachep); } diff --git a/fs/hfsplus/unicode.c b/fs/hfsplus/unicode.c index a32998f29f0b..2c2e47dcfdd8 100644 --- a/fs/hfsplus/unicode.c +++ b/fs/hfsplus/unicode.c @@ -295,7 +295,8 @@ static inline u16 *decompose_unichar(wchar_t uc, int *size) return hfsplus_decompose_table + (off / 4); } -int hfsplus_asc2uni(struct super_block *sb, struct hfsplus_unistr *ustr, +int hfsplus_asc2uni(struct super_block *sb, + struct hfsplus_unistr *ustr, int max_unistr_len, const char *astr, int len) { int size, dsize, decompose; @@ -303,7 +304,7 @@ int hfsplus_asc2uni(struct super_block *sb, struct hfsplus_unistr *ustr, wchar_t c; decompose = !test_bit(HFSPLUS_SB_NODECOMPOSE, &HFSPLUS_SB(sb)->flags); - while (outlen < HFSPLUS_MAX_STRLEN && len > 0) { + while (outlen < max_unistr_len && len > 0) { size = asc2unichar(sb, astr, len, &c); if (decompose) @@ -311,7 +312,7 @@ int hfsplus_asc2uni(struct super_block *sb, struct hfsplus_unistr *ustr, else dstr = NULL; if (dstr) { - if (outlen + dsize > HFSPLUS_MAX_STRLEN) + if (outlen + dsize > max_unistr_len) break; do { ustr->unicode[outlen++] = cpu_to_be16(*dstr++); -- cgit From 899bed05e9f6bbb21776f9ebd88f5631987f987a Mon Sep 17 00:00:00 2001 From: Vyacheslav Dubeyko Date: Wed, 27 Feb 2013 17:03:06 -0800 Subject: hfsplus: fix issue with unzeroed unused b-tree nodes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The fsck_hfs (under MacOS X) complains about unzeroed unused b-tree nodes after deletion of folders' tree under Linux. SYMPTOMS: Running Disk Utiltiy's "Verify Disk" on "test" gives the following: Verifying volume “Test” Checking file systemChecking Journaled HFS Plus volume. Checking extents overflow file. Checking catalog file. Unused node is not erased (node = 3111) Checking multi-linked files. Checking catalog hierarchy. Checking extended attributes file. Checking volume bitmap. Checking volume information. The volume Test was found corrupt and needs to be repaired. Error: This disk needs to be repaired. Click Repair Disk. REPRODUCING PATH: 1. Prepare HFS+ (non-case sensitive) partition (for example, 5GB) under MacOS X. 2. Copy linux kernel source tree (for example, 3.7-rc6 version) on this partition under MacOS X. 3. Then switch to Linux and mount this prepared partition. 4. Execute `sudo rm -r` under prepared directory with linux kernel source tree. 5. Unmount and boot back into OS X. 6. Open up Disk Utility and verify partition. REPRODUCIBILITY: 100% FIX: It is added code of node clearing in hfs_bnode_put() method for the case when node has flag HFS_BNODE_DELETED. Signed-off-by: Vyacheslav Dubeyko Reported-by: Kyle Laracey Acked-by: Hin-Tak Leung Cc: Al Viro Cc: Christoph Hellwig Cc: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hfsplus/bnode.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/hfsplus/bnode.c b/fs/hfsplus/bnode.c index 5c125ce6bd72..f31ac6f404f1 100644 --- a/fs/hfsplus/bnode.c +++ b/fs/hfsplus/bnode.c @@ -648,6 +648,8 @@ void hfs_bnode_put(struct hfs_bnode *node) if (test_bit(HFS_BNODE_DELETED, &node->flags)) { hfs_bnode_unhash(node); spin_unlock(&tree->hash_lock); + hfs_bnode_clear(node, 0, + PAGE_CACHE_SIZE * tree->pages_per_bnode); hfs_bmap_free(node); hfs_bnode_free(node); return; -- cgit From 6b46419b0462ae565880f02e9cd0baf9b25ea71f Mon Sep 17 00:00:00 2001 From: Oleksij Rempel Date: Wed, 27 Feb 2013 17:03:07 -0800 Subject: fat: add extended fileds to struct fat_boot_sector Later we will need "state" field to check if volume was cleanly unmounted. Signed-off-by: Oleksij Rempel Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fat/inode.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/fat/inode.c b/fs/fat/inode.c index f8f491677a4a..4b4d4ef910f3 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -1298,17 +1298,17 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, int isvfat, sbi->prev_free = FAT_START_ENT; sb->s_maxbytes = 0xffffffff; - if (!sbi->fat_length && b->fat32_length) { + if (!sbi->fat_length && b->fat32.length) { struct fat_boot_fsinfo *fsinfo; struct buffer_head *fsinfo_bh; /* Must be FAT32 */ sbi->fat_bits = 32; - sbi->fat_length = le32_to_cpu(b->fat32_length); - sbi->root_cluster = le32_to_cpu(b->root_cluster); + sbi->fat_length = le32_to_cpu(b->fat32.length); + sbi->root_cluster = le32_to_cpu(b->fat32.root_cluster); /* MC - if info_sector is 0, don't multiply by 0 */ - sbi->fsinfo_sector = le16_to_cpu(b->info_sector); + sbi->fsinfo_sector = le16_to_cpu(b->fat32.info_sector); if (sbi->fsinfo_sector == 0) sbi->fsinfo_sector = 1; -- cgit From b88a105802e9aeb6e234e8106659f5d1271081bb Mon Sep 17 00:00:00 2001 From: Oleksij Rempel Date: Wed, 27 Feb 2013 17:03:09 -0800 Subject: fat: mark fs as dirty on mount and clean on umount There is no documented methods to mark FAT as dirty. Unofficially MS started to use reserved Byte in boot sector for this purpose, at least since Win 2000. With Win 7 user is warned if fs is dirty and asked to clean it. Different versions of Win, handle it in different ways, but always have same meaning: - Win 2000 and XP, set it on write operations and remove it after operation was finnished - Win 7, set dirty flag on first write and remove it on umount. We will do it as follows: - set dirty flag on mount. If fs was initially dirty, warn user, remember it and do not do any changes to boot sector. - clean it on umount. If fs was initially dirty, leave it dirty. - do not do any thing if fs mounted read-only. - TODO: leave fs dirty if we found some error after mount. Signed-off-by: Oleksij Rempel Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fat/fat.h | 2 ++ fs/fat/inode.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 68 insertions(+) (limited to 'fs') diff --git a/fs/fat/fat.h b/fs/fat/fat.h index 12701a567752..e9cc3f0d58e2 100644 --- a/fs/fat/fat.h +++ b/fs/fat/fat.h @@ -95,6 +95,8 @@ struct msdos_sb_info { spinlock_t dir_hash_lock; struct hlist_head dir_hashtable[FAT_HASH_SIZE]; + + unsigned int dirty; /* fs state before mount */ }; #define FAT_CACHE_VALID 0 /* special case for valid cache */ diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 4b4d4ef910f3..780e20806346 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -488,10 +488,59 @@ static void fat_evict_inode(struct inode *inode) fat_detach(inode); } +static void fat_set_state(struct super_block *sb, + unsigned int set, unsigned int force) +{ + struct buffer_head *bh; + struct fat_boot_sector *b; + struct msdos_sb_info *sbi = sb->s_fs_info; + + /* do not change any thing if mounted read only */ + if ((sb->s_flags & MS_RDONLY) && !force) + return; + + /* do not change state if fs was dirty */ + if (sbi->dirty) { + /* warn only on set (mount). */ + if (set) + fat_msg(sb, KERN_WARNING, "Volume was not properly " + "unmounted. Some data may be corrupt. " + "Please run fsck."); + return; + } + + bh = sb_bread(sb, 0); + if (bh == NULL) { + fat_msg(sb, KERN_ERR, "unable to read boot sector " + "to mark fs as dirty"); + return; + } + + b = (struct fat_boot_sector *) bh->b_data; + + if (sbi->fat_bits == 32) { + if (set) + b->fat32.state |= FAT_STATE_DIRTY; + else + b->fat32.state &= ~FAT_STATE_DIRTY; + } else /* fat 16 and 12 */ { + if (set) + b->fat16.state |= FAT_STATE_DIRTY; + else + b->fat16.state &= ~FAT_STATE_DIRTY; + } + + mark_buffer_dirty(bh); + sync_dirty_buffer(bh); + brelse(bh); +} + static void fat_put_super(struct super_block *sb) { struct msdos_sb_info *sbi = MSDOS_SB(sb); + fat_set_state(sb, 0, 0); + iput(sbi->fsinfo_inode); iput(sbi->fat_inode); @@ -566,8 +615,18 @@ static void __exit fat_destroy_inodecache(void) static int fat_remount(struct super_block *sb, int *flags, char *data) { + int new_rdonly; struct msdos_sb_info *sbi = MSDOS_SB(sb); *flags |= MS_NODIRATIME | (sbi->options.isvfat ? 0 : MS_NOATIME); + + /* make sure we update state on remount. */ + new_rdonly = *flags & MS_RDONLY; + if (new_rdonly != (sb->s_flags & MS_RDONLY)) { + if (new_rdonly) + fat_set_state(sb, 0, 0); + else + fat_set_state(sb, 1, 1); + } return 0; } @@ -1362,6 +1421,12 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, int isvfat, if (sbi->fat_bits != 32) sbi->fat_bits = (total_clusters > MAX_FAT12) ? 16 : 12; + /* some OSes set FAT_STATE_DIRTY and clean it on unmount. */ + if (sbi->fat_bits == 32) + sbi->dirty = b->fat32.state & FAT_STATE_DIRTY; + else /* fat 16 or 12 */ + sbi->dirty = b->fat16.state & FAT_STATE_DIRTY; + /* check that FAT table does not overflow */ fat_clusters = sbi->fat_length * sb->s_blocksize * 8 / sbi->fat_bits; total_clusters = min(total_clusters, fat_clusters - FAT_START_ENT); @@ -1456,6 +1521,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, int isvfat, "the device does not support discard"); } + fat_set_state(sb, 1, 0); return 0; out_invalid: -- cgit From e579d2c259be42b6f29458327e5153b22414b031 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Wed, 27 Feb 2013 17:03:15 -0800 Subject: coredump: remove redundant defines for dumpable states The existing SUID_DUMP_* defines duplicate the newer SUID_DUMPABLE_* defines introduced in 54b501992dd2 ("coredump: warn about unsafe suid_dumpable / core_pattern combo"). Remove the new ones, and use the prior values instead. Signed-off-by: Kees Cook Reported-by: Chen Gang Cc: Alexander Viro Cc: Alan Cox Cc: "Eric W. Biederman" Cc: Doug Ledford Cc: Serge Hallyn Cc: James Morris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/coredump.c | 2 +- fs/exec.c | 10 +++++----- fs/proc/internal.h | 3 ++- 3 files changed, 8 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/coredump.c b/fs/coredump.c index 69baf903d3bd..c6479658d487 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -501,7 +501,7 @@ void do_coredump(siginfo_t *siginfo) * so we dump it as root in mode 2, and only into a controlled * environment (pipe handler or fully qualified path). */ - if (__get_dumpable(cprm.mm_flags) == SUID_DUMPABLE_SAFE) { + if (__get_dumpable(cprm.mm_flags) == SUID_DUMP_ROOT) { /* Setuid core dump mode */ flag = O_EXCL; /* Stop rewrite attacks */ cred->fsuid = GLOBAL_ROOT_UID; /* Dump root private */ diff --git a/fs/exec.c b/fs/exec.c index 864c50df660a..a96a4885bbbf 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1111,7 +1111,7 @@ void setup_new_exec(struct linux_binprm * bprm) current->sas_ss_sp = current->sas_ss_size = 0; if (uid_eq(current_euid(), current_uid()) && gid_eq(current_egid(), current_gid())) - set_dumpable(current->mm, SUID_DUMPABLE_ENABLED); + set_dumpable(current->mm, SUID_DUMP_USER); else set_dumpable(current->mm, suid_dumpable); @@ -1639,17 +1639,17 @@ EXPORT_SYMBOL(set_binfmt); void set_dumpable(struct mm_struct *mm, int value) { switch (value) { - case SUID_DUMPABLE_DISABLED: + case SUID_DUMP_DISABLE: clear_bit(MMF_DUMPABLE, &mm->flags); smp_wmb(); clear_bit(MMF_DUMP_SECURELY, &mm->flags); break; - case SUID_DUMPABLE_ENABLED: + case SUID_DUMP_USER: set_bit(MMF_DUMPABLE, &mm->flags); smp_wmb(); clear_bit(MMF_DUMP_SECURELY, &mm->flags); break; - case SUID_DUMPABLE_SAFE: + case SUID_DUMP_ROOT: set_bit(MMF_DUMP_SECURELY, &mm->flags); smp_wmb(); set_bit(MMF_DUMPABLE, &mm->flags); @@ -1662,7 +1662,7 @@ int __get_dumpable(unsigned long mm_flags) int ret; ret = mm_flags & MMF_DUMPABLE_MASK; - return (ret > SUID_DUMPABLE_ENABLED) ? SUID_DUMPABLE_SAFE : ret; + return (ret > SUID_DUMP_USER) ? SUID_DUMP_ROOT : ret; } int get_dumpable(struct mm_struct *mm) diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 252544c05207..85ff3a4598b3 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -11,6 +11,7 @@ #include #include +#include struct ctl_table_header; struct mempolicy; @@ -108,7 +109,7 @@ static inline int task_dumpable(struct task_struct *task) if (mm) dumpable = get_dumpable(mm); task_unlock(task); - if (dumpable == SUID_DUMPABLE_ENABLED) + if (dumpable == SUID_DUMP_USER) return 1; return 0; } -- cgit From 87ebdc00eeb474615496d5f10eed46709e25c707 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Wed, 27 Feb 2013 17:03:16 -0800 Subject: fs/proc: clean up printks - use pr_foo() throughout - remove a couple of duplicated KERN_WARNINGs, via WARN(KERN_WARNING "...") - nuke a few warnings which I've never seen happen, ever. Cc: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/base.c | 3 ++- fs/proc/generic.c | 27 +++++++++------------------ fs/proc/inode.c | 5 +++-- fs/proc/kcore.c | 3 ++- fs/proc/proc_devtree.c | 13 +++++++------ fs/proc/proc_sysctl.c | 19 ++++++++++--------- fs/proc/vmcore.c | 15 ++++++--------- 7 files changed, 39 insertions(+), 46 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index f3b133d79914..69078c7cef1f 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -73,6 +73,7 @@ #include #include #include +#include #include #include #include @@ -952,7 +953,7 @@ static ssize_t oom_adj_write(struct file *file, const char __user *buf, * /proc/pid/oom_adj is provided for legacy purposes, ask users to use * /proc/pid/oom_score_adj instead. */ - printk_once(KERN_WARNING "%s (%d): /proc/%d/oom_adj is deprecated, please use /proc/%d/oom_score_adj instead.\n", + pr_warn_once("%s (%d): /proc/%d/oom_adj is deprecated, please use /proc/%d/oom_score_adj instead.\n", current->comm, task_pid_nr(current), task_pid_nr(task), task_pid_nr(task)); diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 2983dc52ca25..4b3b3ffb52f1 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include #include @@ -132,11 +133,8 @@ __proc_file_read(struct file *file, char __user *buf, size_t nbytes, } if (start == NULL) { - if (n > PAGE_SIZE) { - printk(KERN_ERR - "proc_file_read: Apparent buffer overflow!\n"); + if (n > PAGE_SIZE) /* Apparent buffer overflow */ n = PAGE_SIZE; - } n -= *ppos; if (n <= 0) break; @@ -144,26 +142,19 @@ __proc_file_read(struct file *file, char __user *buf, size_t nbytes, n = count; start = page + *ppos; } else if (start < page) { - if (n > PAGE_SIZE) { - printk(KERN_ERR - "proc_file_read: Apparent buffer overflow!\n"); + if (n > PAGE_SIZE) /* Apparent buffer overflow */ n = PAGE_SIZE; - } if (n > count) { /* * Don't reduce n because doing so might * cut off part of a data block. */ - printk(KERN_WARNING - "proc_file_read: Read count exceeded\n"); + pr_warn("proc_file_read: count exceeded\n"); } } else /* start >= page */ { unsigned long startoff = (unsigned long)(start - page); - if (n > (PAGE_SIZE - startoff)) { - printk(KERN_ERR - "proc_file_read: Apparent buffer overflow!\n"); + if (n > (PAGE_SIZE - startoff)) /* buffer overflow? */ n = PAGE_SIZE - startoff; - } if (n > count) n = count; } @@ -569,7 +560,7 @@ static int proc_register(struct proc_dir_entry * dir, struct proc_dir_entry * dp for (tmp = dir->subdir; tmp; tmp = tmp->next) if (strcmp(tmp->name, dp->name) == 0) { - WARN(1, KERN_WARNING "proc_dir_entry '%s/%s' already registered\n", + WARN(1, "proc_dir_entry '%s/%s' already registered\n", dir->name, dp->name); break; } @@ -830,9 +821,9 @@ void remove_proc_entry(const char *name, struct proc_dir_entry *parent) if (S_ISDIR(de->mode)) parent->nlink--; de->nlink = 0; - WARN(de->subdir, KERN_WARNING "%s: removing non-empty directory " - "'%s/%s', leaking at least '%s'\n", __func__, - de->parent->name, de->name, de->subdir->name); + WARN(de->subdir, "%s: removing non-empty directory " + "'%s/%s', leaking at least '%s'\n", __func__, + de->parent->name, de->name, de->subdir->name); pde_put(de); } EXPORT_SYMBOL(remove_proc_entry); diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 70322e1a4f0f..a86aebc9ba7c 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include @@ -495,13 +496,13 @@ int proc_fill_super(struct super_block *s) pde_get(&proc_root); root_inode = proc_get_inode(s, &proc_root); if (!root_inode) { - printk(KERN_ERR "proc_fill_super: get root inode failed\n"); + pr_err("proc_fill_super: get root inode failed\n"); return -ENOMEM; } s->s_root = d_make_root(root_inode); if (!s->s_root) { - printk(KERN_ERR "proc_fill_super: allocate dentry failed\n"); + pr_err("proc_fill_super: allocate dentry failed\n"); return -ENOMEM; } diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c index e96d4f18ca3a..eda6f017f272 100644 --- a/fs/proc/kcore.c +++ b/fs/proc/kcore.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include @@ -619,7 +620,7 @@ static int __init proc_kcore_init(void) proc_root_kcore = proc_create("kcore", S_IRUSR, NULL, &proc_kcore_operations); if (!proc_root_kcore) { - printk(KERN_ERR "couldn't create /proc/kcore\n"); + pr_err("couldn't create /proc/kcore\n"); return 0; /* Always returns 0. */ } /* Store text area if it's special */ diff --git a/fs/proc/proc_devtree.c b/fs/proc/proc_devtree.c index de20ec480fa0..30b590f5bd35 100644 --- a/fs/proc/proc_devtree.c +++ b/fs/proc/proc_devtree.c @@ -8,6 +8,7 @@ #include #include #include +#include #include #include #include @@ -110,8 +111,8 @@ void proc_device_tree_update_prop(struct proc_dir_entry *pde, if (ent->data == oldprop) break; if (ent == NULL) { - printk(KERN_WARNING "device-tree: property \"%s\" " - " does not exist\n", oldprop->name); + pr_warn("device-tree: property \"%s\" does not exist\n", + oldprop->name); } else { ent->data = newprop; ent->size = newprop->length; @@ -153,8 +154,8 @@ static const char *fixup_name(struct device_node *np, struct proc_dir_entry *de, realloc: fixed_name = kmalloc(fixup_len, GFP_KERNEL); if (fixed_name == NULL) { - printk(KERN_ERR "device-tree: Out of memory trying to fixup " - "name \"%s\"\n", name); + pr_err("device-tree: Out of memory trying to fixup " + "name \"%s\"\n", name); return name; } @@ -175,8 +176,8 @@ retry: goto retry; } - printk(KERN_WARNING "device-tree: Duplicate name in %s, " - "renamed to \"%s\"\n", np->full_name, fixed_name); + pr_warn("device-tree: Duplicate name in %s, renamed to \"%s\"\n", + np->full_name, fixed_name); return fixed_name; } diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index 612df79cc6a1..ac05f33a0dde 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -5,6 +5,7 @@ #include #include #include +#include #include #include #include @@ -57,7 +58,7 @@ static void sysctl_print_dir(struct ctl_dir *dir) { if (dir->header.parent) sysctl_print_dir(dir->header.parent); - printk(KERN_CONT "%s/", dir->header.ctl_table[0].procname); + pr_cont("%s/", dir->header.ctl_table[0].procname); } static int namecmp(const char *name1, int len1, const char *name2, int len2) @@ -134,9 +135,9 @@ static int insert_entry(struct ctl_table_header *head, struct ctl_table *entry) else if (cmp > 0) p = &(*p)->rb_right; else { - printk(KERN_ERR "sysctl duplicate entry: "); + pr_err("sysctl duplicate entry: "); sysctl_print_dir(head->parent); - printk(KERN_CONT "/%s\n", entry->procname); + pr_cont("/%s\n", entry->procname); return -EEXIST; } } @@ -927,9 +928,9 @@ found: subdir->header.nreg++; failed: if (unlikely(IS_ERR(subdir))) { - printk(KERN_ERR "sysctl could not get directory: "); + pr_err("sysctl could not get directory: "); sysctl_print_dir(dir); - printk(KERN_CONT "/%*.*s %ld\n", + pr_cont("/%*.*s %ld\n", namelen, namelen, name, PTR_ERR(subdir)); } drop_sysctl_table(&dir->header); @@ -995,8 +996,8 @@ static int sysctl_err(const char *path, struct ctl_table *table, char *fmt, ...) vaf.fmt = fmt; vaf.va = &args; - printk(KERN_ERR "sysctl table check failed: %s/%s %pV\n", - path, table->procname, &vaf); + pr_err("sysctl table check failed: %s/%s %pV\n", + path, table->procname, &vaf); va_end(args); return -EINVAL; @@ -1510,9 +1511,9 @@ static void put_links(struct ctl_table_header *header) drop_sysctl_table(link_head); } else { - printk(KERN_ERR "sysctl link missing during unregister: "); + pr_err("sysctl link missing during unregister: "); sysctl_print_dir(parent); - printk(KERN_CONT "/%s\n", name); + pr_cont("/%s\n", name); } } } diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 0d5071d29985..41dd018f72dd 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include #include @@ -553,8 +554,7 @@ static int __init parse_crash_elf64_headers(void) ehdr.e_ehsize != sizeof(Elf64_Ehdr) || ehdr.e_phentsize != sizeof(Elf64_Phdr) || ehdr.e_phnum == 0) { - printk(KERN_WARNING "Warning: Core image elf header is not" - "sane\n"); + pr_warn("Warning: Core image elf header is not sane\n"); return -EINVAL; } @@ -609,8 +609,7 @@ static int __init parse_crash_elf32_headers(void) ehdr.e_ehsize != sizeof(Elf32_Ehdr) || ehdr.e_phentsize != sizeof(Elf32_Phdr) || ehdr.e_phnum == 0) { - printk(KERN_WARNING "Warning: Core image elf header is not" - "sane\n"); + pr_warn("Warning: Core image elf header is not sane\n"); return -EINVAL; } @@ -653,8 +652,7 @@ static int __init parse_crash_elf_headers(void) if (rc < 0) return rc; if (memcmp(e_ident, ELFMAG, SELFMAG) != 0) { - printk(KERN_WARNING "Warning: Core image elf header" - " not found\n"); + pr_warn("Warning: Core image elf header not found\n"); return -EINVAL; } @@ -673,8 +671,7 @@ static int __init parse_crash_elf_headers(void) /* Determine vmcore size. */ vmcore_size = get_vmcore_size_elf32(elfcorebuf); } else { - printk(KERN_WARNING "Warning: Core image elf header is not" - " sane\n"); + pr_warn("Warning: Core image elf header is not sane\n"); return -EINVAL; } return 0; @@ -690,7 +687,7 @@ static int __init vmcore_init(void) return rc; rc = parse_crash_elf_headers(); if (rc) { - printk(KERN_WARNING "Kdump: vmcore not initialized\n"); + pr_warn("Kdump: vmcore not initialized\n"); return rc; } -- cgit From c2c1b089b44b783bd50fae4bccaa6f367f92e492 Mon Sep 17 00:00:00 2001 From: Zhang Yanfei Date: Wed, 27 Feb 2013 17:03:17 -0800 Subject: fs/proc/vmcore.c: put if tests in the top of the while loop to reduce duplication In read_vmcore() two `if' tests are duplicated. Change the position of them could reduce the duplication. This change does not affect the behaviour of the function. [akpm@linux-foundation.org: avoid `if (foo = bar)' thing, use min_t()] [akpm@linux-foundation.org: s/max_t/min_t/] Signed-off-by: Zhang Yanfei Cc: "Eric W. Biederman" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/vmcore.c | 20 +++++++------------- 1 file changed, 7 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 41dd018f72dd..b870f740ab5a 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -176,15 +176,15 @@ static ssize_t read_vmcore(struct file *file, char __user *buffer, start = map_offset_to_paddr(*fpos, &vmcore_list, &curr_m); if (!curr_m) return -EINVAL; - if ((tsz = (PAGE_SIZE - (start & ~PAGE_MASK))) > buflen) - tsz = buflen; - - /* Calculate left bytes in current memory segment. */ - nr_bytes = (curr_m->size - (start - curr_m->paddr)); - if (tsz > nr_bytes) - tsz = nr_bytes; while (buflen) { + tsz = min_t(size_t, buflen, PAGE_SIZE - (start & ~PAGE_MASK)); + + /* Calculate left bytes in current memory segment. */ + nr_bytes = (curr_m->size - (start - curr_m->paddr)); + if (tsz > nr_bytes) + tsz = nr_bytes; + tmp = read_from_oldmem(buffer, tsz, &start, 1); if (tmp < 0) return tmp; @@ -199,12 +199,6 @@ static ssize_t read_vmcore(struct file *file, char __user *buffer, struct vmcore, list); start = curr_m->paddr; } - if ((tsz = (PAGE_SIZE - (start & ~PAGE_MASK))) > buflen) - tsz = buflen; - /* Calculate left bytes in current memory segment. */ - nr_bytes = (curr_m->size - (start - curr_m->paddr)); - if (tsz > nr_bytes) - tsz = nr_bytes; } return acc; } -- cgit From 80de7f7ae013b1e287059f39eaba2a12219681cf Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Wed, 27 Feb 2013 17:03:21 -0800 Subject: seq-file: use SEEK_ macros instead of hardcoded numbers Signed-off-by: Cyrill Gorcunov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/seq_file.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/seq_file.c b/fs/seq_file.c index f2bc3dfd0b88..11ba05608818 100644 --- a/fs/seq_file.c +++ b/fs/seq_file.c @@ -308,9 +308,9 @@ loff_t seq_lseek(struct file *file, loff_t offset, int whence) mutex_lock(&m->lock); m->version = file->f_version; switch (whence) { - case 1: + case SEEK_CUR: offset += file->f_pos; - case 0: + case SEEK_SET: if (offset < 0) break; retval = offset; -- cgit From 5e62adef9e1d3e8e26dfd6c95be484b955ad2927 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Wed, 27 Feb 2013 17:03:22 -0800 Subject: fs/seq_file.c:seq_lseek(): fix switch statement indenting [akpm@linux-foundation.org: checkpatch fixes] Cc: Cyrill Gorcunov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/seq_file.c | 40 ++++++++++++++++++++-------------------- 1 file changed, 20 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/seq_file.c b/fs/seq_file.c index 11ba05608818..15c6304bab71 100644 --- a/fs/seq_file.c +++ b/fs/seq_file.c @@ -308,27 +308,27 @@ loff_t seq_lseek(struct file *file, loff_t offset, int whence) mutex_lock(&m->lock); m->version = file->f_version; switch (whence) { - case SEEK_CUR: - offset += file->f_pos; - case SEEK_SET: - if (offset < 0) - break; - retval = offset; - if (offset != m->read_pos) { - while ((retval=traverse(m, offset)) == -EAGAIN) - ; - if (retval) { - /* with extreme prejudice... */ - file->f_pos = 0; - m->read_pos = 0; - m->version = 0; - m->index = 0; - m->count = 0; - } else { - m->read_pos = offset; - retval = file->f_pos = offset; - } + case SEEK_CUR: + offset += file->f_pos; + case SEEK_SET: + if (offset < 0) + break; + retval = offset; + if (offset != m->read_pos) { + while ((retval = traverse(m, offset)) == -EAGAIN) + ; + if (retval) { + /* with extreme prejudice... */ + file->f_pos = 0; + m->read_pos = 0; + m->version = 0; + m->index = 0; + m->count = 0; + } else { + m->read_pos = offset; + retval = file->f_pos = offset; } + } } file->f_version = m->version; mutex_unlock(&m->lock); -- cgit From cda95406c8fe491c01b3abc422c25a613e9f36ea Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:03:44 -0800 Subject: dlm: use idr_for_each_entry() in recover_idr_clear() error path Convert recover_idr_clear() to use idr_for_each_entry() instead of idr_for_each(). It's somewhat less efficient this way but it shouldn't matter in an error path. This is to help with deprecation of idr_remove_all(). Signed-off-by: Tejun Heo Cc: Christine Caulfield Cc: David Teigland Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/dlm/recover.c | 23 ++++++++++------------- 1 file changed, 10 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/dlm/recover.c b/fs/dlm/recover.c index aedea28a86a1..b2856e701641 100644 --- a/fs/dlm/recover.c +++ b/fs/dlm/recover.c @@ -351,23 +351,20 @@ static struct dlm_rsb *recover_idr_find(struct dlm_ls *ls, uint64_t id) return r; } -static int recover_idr_clear_rsb(int id, void *p, void *data) +static void recover_idr_clear(struct dlm_ls *ls) { - struct dlm_ls *ls = data; - struct dlm_rsb *r = p; + struct dlm_rsb *r; + int id; - r->res_id = 0; - r->res_recover_locks_count = 0; - ls->ls_recover_list_count--; + spin_lock(&ls->ls_recover_idr_lock); - dlm_put_rsb(r); - return 0; -} + idr_for_each_entry(&ls->ls_recover_idr, r, id) { + r->res_id = 0; + r->res_recover_locks_count = 0; + ls->ls_recover_list_count--; -static void recover_idr_clear(struct dlm_ls *ls) -{ - spin_lock(&ls->ls_recover_idr_lock); - idr_for_each(&ls->ls_recover_idr, recover_idr_clear_rsb, ls); + dlm_put_rsb(r); + } idr_remove_all(&ls->ls_recover_idr); if (ls->ls_recover_list_count != 0) { -- cgit From a67a380e6f8534b9fed3b3b663d108bf35af0132 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:03:45 -0800 Subject: dlm: don't use idr_remove_all() idr_destroy() can destroy idr by itself and idr_remove_all() is being deprecated. The conversion isn't completely trivial for recover_idr_clear() as it's the only place in kernel which makes legitimate use of idr_remove_all() w/o idr_destroy(). Replace it with idr_remove() call inside idr_for_each_entry() loop. It goes on top so that it matches the operation order in recover_idr_del(). Signed-off-by: Tejun Heo Cc: Christine Caulfield Cc: David Teigland Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/dlm/lockspace.c | 1 - fs/dlm/recover.c | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c index 2e99fb0c9737..3ca79d3253b9 100644 --- a/fs/dlm/lockspace.c +++ b/fs/dlm/lockspace.c @@ -796,7 +796,6 @@ static int release_lockspace(struct dlm_ls *ls, int force) */ idr_for_each(&ls->ls_lkbidr, lkb_idr_free, ls); - idr_remove_all(&ls->ls_lkbidr); idr_destroy(&ls->ls_lkbidr); /* diff --git a/fs/dlm/recover.c b/fs/dlm/recover.c index b2856e701641..236d1085d6a2 100644 --- a/fs/dlm/recover.c +++ b/fs/dlm/recover.c @@ -359,13 +359,13 @@ static void recover_idr_clear(struct dlm_ls *ls) spin_lock(&ls->ls_recover_idr_lock); idr_for_each_entry(&ls->ls_recover_idr, r, id) { + idr_remove(&ls->ls_recover_idr, id); r->res_id = 0; r->res_recover_locks_count = 0; ls->ls_recover_list_count--; dlm_put_rsb(r); } - idr_remove_all(&ls->ls_recover_idr); if (ls->ls_recover_list_count != 0) { log_error(ls, "warning: recover_list_count %d", -- cgit From d97bec801d726abb14ab0e18d6aef5293580a467 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:03:46 -0800 Subject: nfs: idr_destroy() no longer needs idr_remove_all() idr_destroy() can destroy idr by itself and idr_remove_all() is being deprecated. Drop reference to idr_remove_all(). Note that the code wasn't completely correct before because idr_remove() on all entries doesn't necessarily release all idr_layers which could lead to memory leak. Signed-off-by: Tejun Heo Cc: "J. Bruce Fields" Cc: Trond Myklebust Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/nfs/client.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 9f3c66438d0e..84d8eae203a7 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -197,7 +197,6 @@ error_0: EXPORT_SYMBOL_GPL(nfs_alloc_client); #if IS_ENABLED(CONFIG_NFS_V4) -/* idr_remove_all is not needed as all id's are removed by nfs_put_client */ void nfs_cleanup_cb_ident_idr(struct net *net) { struct nfs_net *nn = net_generic(net, nfs_net_id); -- cgit From 644e1b90ef06611268b48a0a8d14a0ab990ab3f4 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:03:48 -0800 Subject: inotify: don't use idr_remove_all() idr_destroy() can destroy idr by itself and idr_remove_all() is being deprecated. Drop its usage. Signed-off-by: Tejun Heo Cc: John McCutchan Cc: Robert Love Cc: Eric Paris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/notify/inotify/inotify_fsnotify.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_fsnotify.c b/fs/notify/inotify/inotify_fsnotify.c index 871569c7d609..4216308b81b4 100644 --- a/fs/notify/inotify/inotify_fsnotify.c +++ b/fs/notify/inotify/inotify_fsnotify.c @@ -197,7 +197,6 @@ static void inotify_free_group_priv(struct fsnotify_group *group) { /* ideally the idr is empty and we won't hit the BUG in the callback */ idr_for_each(&group->inotify_data.idr, idr_callback, group); - idr_remove_all(&group->inotify_data.idr); idr_destroy(&group->inotify_data.idr); atomic_dec(&group->inotify_data.user->inotify_devs); free_uid(group->inotify_data.user); -- cgit From 2a86b3e74f12bcdd13ceb1bf333bc2d5f43b3c02 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:04:49 -0800 Subject: dlm: convert to idr_alloc() Convert to the much saner new idr interface. Error return values from recover_idr_add() mix -1 and -errno. The conversion doesn't change that but it looks iffy. Signed-off-by: Tejun Heo Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/dlm/lock.c | 18 ++++++------------ fs/dlm/recover.c | 27 +++++++++++++-------------- 2 files changed, 19 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index f7501651762d..1b1146670c4b 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -1183,7 +1183,7 @@ static void detach_lkb(struct dlm_lkb *lkb) static int create_lkb(struct dlm_ls *ls, struct dlm_lkb **lkb_ret) { struct dlm_lkb *lkb; - int rv, id; + int rv; lkb = dlm_allocate_lkb(ls); if (!lkb) @@ -1199,19 +1199,13 @@ static int create_lkb(struct dlm_ls *ls, struct dlm_lkb **lkb_ret) mutex_init(&lkb->lkb_cb_mutex); INIT_WORK(&lkb->lkb_cb_work, dlm_callback_work); - retry: - rv = idr_pre_get(&ls->ls_lkbidr, GFP_NOFS); - if (!rv) - return -ENOMEM; - + idr_preload(GFP_NOFS); spin_lock(&ls->ls_lkbidr_spin); - rv = idr_get_new_above(&ls->ls_lkbidr, lkb, 1, &id); - if (!rv) - lkb->lkb_id = id; + rv = idr_alloc(&ls->ls_lkbidr, lkb, 1, 0, GFP_NOWAIT); + if (rv >= 0) + lkb->lkb_id = rv; spin_unlock(&ls->ls_lkbidr_spin); - - if (rv == -EAGAIN) - goto retry; + idr_preload_end(); if (rv < 0) { log_error(ls, "create_lkb idr error %d", rv); diff --git a/fs/dlm/recover.c b/fs/dlm/recover.c index 236d1085d6a2..a6bc63f6e31b 100644 --- a/fs/dlm/recover.c +++ b/fs/dlm/recover.c @@ -305,27 +305,26 @@ static int recover_idr_empty(struct dlm_ls *ls) static int recover_idr_add(struct dlm_rsb *r) { struct dlm_ls *ls = r->res_ls; - int rv, id; - - rv = idr_pre_get(&ls->ls_recover_idr, GFP_NOFS); - if (!rv) - return -ENOMEM; + int rv; + idr_preload(GFP_NOFS); spin_lock(&ls->ls_recover_idr_lock); if (r->res_id) { - spin_unlock(&ls->ls_recover_idr_lock); - return -1; - } - rv = idr_get_new_above(&ls->ls_recover_idr, r, 1, &id); - if (rv) { - spin_unlock(&ls->ls_recover_idr_lock); - return rv; + rv = -1; + goto out_unlock; } - r->res_id = id; + rv = idr_alloc(&ls->ls_recover_idr, r, 1, 0, GFP_NOWAIT); + if (rv < 0) + goto out_unlock; + + r->res_id = rv; ls->ls_recover_list_count++; dlm_hold_rsb(r); + rv = 0; +out_unlock: spin_unlock(&ls->ls_recover_idr_lock); - return 0; + idr_preload_end(); + return rv; } static void recover_idr_del(struct dlm_rsb *r) -- cgit From 4542da631ad210716d097aa803a0828f9fed5e87 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:04:50 -0800 Subject: inotify: convert to idr_alloc() Convert to the much saner new idr interface. Note that the adhoc cyclic id allocation is buggy. If wraparound happens, the previous code with idr_get_new_above() may segfault and the converted code will trigger WARN and return -EINVAL. Even if it's fixed to wrap to zero, the code will be prone to unnecessary -ENOSPC failures after the first wraparound. We probably need to implement proper cyclic support in idr. Signed-off-by: Tejun Heo Cc: John McCutchan Cc: Robert Love Cc: Eric Paris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/notify/inotify/inotify_user.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 07f7a92fe88e..e0f7c1241a6a 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -364,22 +364,20 @@ static int inotify_add_to_idr(struct idr *idr, spinlock_t *idr_lock, { int ret; - do { - if (unlikely(!idr_pre_get(idr, GFP_KERNEL))) - return -ENOMEM; + idr_preload(GFP_KERNEL); + spin_lock(idr_lock); - spin_lock(idr_lock); - ret = idr_get_new_above(idr, i_mark, *last_wd + 1, - &i_mark->wd); + ret = idr_alloc(idr, i_mark, *last_wd + 1, 0, GFP_NOWAIT); + if (ret >= 0) { /* we added the mark to the idr, take a reference */ - if (!ret) { - *last_wd = i_mark->wd; - fsnotify_get_mark(&i_mark->fsn_mark); - } - spin_unlock(idr_lock); - } while (ret == -EAGAIN); + i_mark->wd = ret; + *last_wd = i_mark->wd; + fsnotify_get_mark(&i_mark->fsn_mark); + } - return ret; + spin_unlock(idr_lock); + idr_preload_end(); + return ret < 0 ? ret : 0; } static struct inotify_inode_mark *inotify_idr_find_locked(struct fsnotify_group *group, -- cgit From 6b207ba3ebe7428a18878c58130e13a3e3bccef9 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:04:51 -0800 Subject: ocfs2: convert to idr_alloc() Convert to the much saner new idr interface. Signed-off-by: Tejun Heo Cc: Mark Fasheh Acked-by: Joel Becker Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ocfs2/cluster/tcp.c | 32 +++++++++++++------------------- 1 file changed, 13 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 0d2bf566e39a..aa88bd8bcedc 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -304,28 +304,22 @@ static u8 o2net_num_from_nn(struct o2net_node *nn) static int o2net_prep_nsw(struct o2net_node *nn, struct o2net_status_wait *nsw) { - int ret = 0; - - do { - if (!idr_pre_get(&nn->nn_status_idr, GFP_ATOMIC)) { - ret = -EAGAIN; - break; - } - spin_lock(&nn->nn_lock); - ret = idr_get_new(&nn->nn_status_idr, nsw, &nsw->ns_id); - if (ret == 0) - list_add_tail(&nsw->ns_node_item, - &nn->nn_status_list); - spin_unlock(&nn->nn_lock); - } while (ret == -EAGAIN); + int ret; - if (ret == 0) { - init_waitqueue_head(&nsw->ns_wq); - nsw->ns_sys_status = O2NET_ERR_NONE; - nsw->ns_status = 0; + spin_lock(&nn->nn_lock); + ret = idr_alloc(&nn->nn_status_idr, nsw, 0, 0, GFP_ATOMIC); + if (ret >= 0) { + nsw->ns_id = ret; + list_add_tail(&nsw->ns_node_item, &nn->nn_status_list); } + spin_unlock(&nn->nn_lock); + if (ret < 0) + return ret; - return ret; + init_waitqueue_head(&nsw->ns_wq); + nsw->ns_sys_status = O2NET_ERR_NONE; + nsw->ns_status = 0; + return 0; } static void o2net_complete_nsw_locked(struct o2net_node *nn, -- cgit From d6870312659d8c6e80419ddb09d366bdd0f17ab6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:05:01 -0800 Subject: nfs4client: convert to idr_alloc() Convert to the much saner new idr interface. Signed-off-by: Tejun Heo Cc: Trond Myklebust Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/nfs/nfs4client.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index 2e9779b58b7a..47d100872390 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -29,15 +29,14 @@ static int nfs_get_cb_ident_idr(struct nfs_client *clp, int minorversion) if (clp->rpc_ops->version != 4 || minorversion != 0) return ret; -retry: - if (!idr_pre_get(&nn->cb_ident_idr, GFP_KERNEL)) - return -ENOMEM; + idr_preload(GFP_KERNEL); spin_lock(&nn->nfs_client_lock); - ret = idr_get_new(&nn->cb_ident_idr, clp, &clp->cl_cb_ident); + ret = idr_alloc(&nn->cb_ident_idr, clp, 0, 0, GFP_NOWAIT); + if (ret >= 0) + clp->cl_cb_ident = ret; spin_unlock(&nn->nfs_client_lock); - if (ret == -EAGAIN) - goto retry; - return ret; + idr_preload_end(); + return ret < 0 ? ret : 0; } #ifdef CONFIG_NFS_V4_1 -- cgit From e8c8d1bc063bc88cfa1356266027b5075d3a82d7 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 27 Feb 2013 17:05:04 -0800 Subject: idr: remove MAX_IDR_MASK and move left MAX_IDR_* into idr.c MAX_IDR_MASK is another weirdness in the idr interface. As idr covers whole positive integer range, it's defined as 0x7fffffff or INT_MAX. Its usage in idr_find(), idr_replace() and idr_remove() is bizarre. They basically mask off the sign bit and operate on the rest, so if the caller, by accident, passes in a negative number, the sign bit will be masked off and the remaining part will be used as if that was the input, which is worse than crashing. The constant is visible in idr.h and there are several users in the kernel. * drivers/i2c/i2c-core.c:i2c_add_numbered_adapter() Basically used to test if adap->nr is a negative number which isn't -1 and returns -EINVAL if so. idr_alloc() already has negative @start checking (w/ WARN_ON_ONCE), so this can go away. * drivers/infiniband/core/cm.c:cm_alloc_id() drivers/infiniband/hw/mlx4/cm.c:id_map_alloc() Used to wrap cyclic @start. Can be replaced with max(next, 0). Note that this type of cyclic allocation using idr is buggy. These are prone to spurious -ENOSPC failure after the first wraparound. * fs/super.c:get_anon_bdev() The ID allocated from ida is masked off before being tested whether it's inside valid range. ida allocated ID can never be a negative number and the masking is unnecessary. Update idr_*() functions to fail with -EINVAL when negative @id is specified and update other MAX_IDR_MASK users as described above. This leaves MAX_IDR_MASK without any user, remove it and relocate other MAX_IDR_* constants to lib/idr.c. Signed-off-by: Tejun Heo Cc: Jean Delvare Cc: Roland Dreier Cc: Sean Hefty Cc: Hal Rosenstock Cc: "Marciniszyn, Mike" Cc: Jack Morgenstein Cc: Or Gerlitz Cc: Al Viro Acked-by: Wolfram Sang Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/super.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/super.c b/fs/super.c index 12f123712161..df6c2f4c6b59 100644 --- a/fs/super.c +++ b/fs/super.c @@ -842,7 +842,7 @@ int get_anon_bdev(dev_t *p) else if (error) return -EAGAIN; - if ((dev & MAX_IDR_MASK) == (1 << MINORBITS)) { + if (dev == (1 << MINORBITS)) { spin_lock(&unnamed_dev_lock); ida_remove(&unnamed_dev_ida, dev); if (unnamed_dev_start > dev) -- cgit From b67bfe0d42cac56c512dd5da4b1b347a23f4b70a Mon Sep 17 00:00:00 2001 From: Sasha Levin Date: Wed, 27 Feb 2013 17:06:00 -0800 Subject: hlist: drop the node parameter from iterators I'm not sure why, but the hlist for each entry iterators were conceived list_for_each_entry(pos, head, member) The hlist ones were greedy and wanted an extra parameter: hlist_for_each_entry(tpos, pos, head, member) Why did they need an extra pos parameter? I'm not quite sure. Not only they don't really need it, it also prevents the iterator from looking exactly like the list iterator, which is unfortunate. Besides the semantic patch, there was some manual work required: - Fix up the actual hlist iterators in linux/list.h - Fix up the declaration of other iterators based on the hlist ones. - A very small amount of places were using the 'node' parameter, this was modified to use 'obj->member' instead. - Coccinelle didn't handle the hlist_for_each_entry_safe iterator properly, so those had to be fixed up manually. The semantic patch which is mostly the work of Peter Senna Tschudin is here: @@ iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host; type T; expression a,c,d,e; identifier b; statement S; @@ -T b; <+... when != b ( hlist_for_each_entry(a, - b, c, d) S | hlist_for_each_entry_continue(a, - b, c) S | hlist_for_each_entry_from(a, - b, c) S | hlist_for_each_entry_rcu(a, - b, c, d) S | hlist_for_each_entry_rcu_bh(a, - b, c, d) S | hlist_for_each_entry_continue_rcu_bh(a, - b, c) S | for_each_busy_worker(a, c, - b, d) S | ax25_uid_for_each(a, - b, c) S | ax25_for_each(a, - b, c) S | inet_bind_bucket_for_each(a, - b, c) S | sctp_for_each_hentry(a, - b, c) S | sk_for_each(a, - b, c) S | sk_for_each_rcu(a, - b, c) S | sk_for_each_from -(a, b) +(a) S + sk_for_each_from(a) S | sk_for_each_safe(a, - b, c, d) S | sk_for_each_bound(a, - b, c) S | hlist_for_each_entry_safe(a, - b, c, d, e) S | hlist_for_each_entry_continue_rcu(a, - b, c) S | nr_neigh_for_each(a, - b, c) S | nr_neigh_for_each_safe(a, - b, c, d) S | nr_node_for_each(a, - b, c) S | nr_node_for_each_safe(a, - b, c, d) S | - for_each_gfn_sp(a, c, d, b) S + for_each_gfn_sp(a, c, d) S | - for_each_gfn_indirect_valid_sp(a, c, d, b) S + for_each_gfn_indirect_valid_sp(a, c, d) S | for_each_host(a, - b, c) S | for_each_host_safe(a, - b, c, d) S | for_each_mesh_entry(a, - b, c, d) S ) ...+> [akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c] [akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c] [akpm@linux-foundation.org: checkpatch fixes] [akpm@linux-foundation.org: fix warnings] [akpm@linux-foudnation.org: redo intrusive kvm changes] Tested-by: Peter Senna Tschudin Acked-by: Paul E. McKenney Signed-off-by: Sasha Levin Cc: Wu Fengguang Cc: Marcelo Tosatti Cc: Gleb Natapov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/affs/amigaffs.c | 3 +-- fs/aio.c | 3 +-- fs/cifs/inode.c | 3 +-- fs/dcache.c | 9 +++------ fs/dlm/lowcomms.c | 11 ++++------- fs/ecryptfs/messaging.c | 6 ++---- fs/exportfs/expfs.c | 3 +-- fs/fat/inode.c | 3 +-- fs/fat/nfs.c | 3 +-- fs/fscache/cookie.c | 11 ++++------- fs/inode.c | 19 +++++++------------ fs/lockd/host.c | 29 ++++++++++++----------------- fs/lockd/svcsubs.c | 7 +++---- fs/nfs/pnfs_dev.c | 9 +++------ fs/nfsd/nfscache.c | 3 +-- fs/notify/fsnotify.c | 3 +-- fs/notify/inode_mark.c | 19 ++++++++----------- fs/notify/vfsmount_mark.c | 19 ++++++++----------- fs/ocfs2/dcache.c | 3 +-- fs/ocfs2/dlm/dlmrecovery.c | 6 ++---- fs/super.c | 6 ++---- fs/sysfs/bin.c | 3 +-- fs/xfs/xfs_log_recover.c | 3 +-- 23 files changed, 69 insertions(+), 115 deletions(-) (limited to 'fs') diff --git a/fs/affs/amigaffs.c b/fs/affs/amigaffs.c index eb82ee53ee0b..d9a43674cb94 100644 --- a/fs/affs/amigaffs.c +++ b/fs/affs/amigaffs.c @@ -125,9 +125,8 @@ static void affs_fix_dcache(struct inode *inode, u32 entry_ino) { struct dentry *dentry; - struct hlist_node *p; spin_lock(&inode->i_lock); - hlist_for_each_entry(dentry, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(dentry, &inode->i_dentry, d_alias) { if (entry_ino == (u32)(long)dentry->d_fsdata) { dentry->d_fsdata = (void *)inode->i_ino; break; diff --git a/fs/aio.c b/fs/aio.c index 064bfbe37566..3f941f2a3059 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -591,11 +591,10 @@ static struct kioctx *lookup_ioctx(unsigned long ctx_id) { struct mm_struct *mm = current->mm; struct kioctx *ctx, *ret = NULL; - struct hlist_node *n; rcu_read_lock(); - hlist_for_each_entry_rcu(ctx, n, &mm->ioctx_list, list) { + hlist_for_each_entry_rcu(ctx, &mm->ioctx_list, list) { /* * RCU protects us against accessing freed memory but * we have to be careful not to get a reference when the diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index d2a833999bcc..83f2606c76d0 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -816,10 +816,9 @@ static bool inode_has_hashed_dentries(struct inode *inode) { struct dentry *dentry; - struct hlist_node *p; spin_lock(&inode->i_lock); - hlist_for_each_entry(dentry, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(dentry, &inode->i_dentry, d_alias) { if (!d_unhashed(dentry) || IS_ROOT(dentry)) { spin_unlock(&inode->i_lock); return true; diff --git a/fs/dcache.c b/fs/dcache.c index 68220dd0c135..fbfae008ba44 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -675,11 +675,10 @@ EXPORT_SYMBOL(dget_parent); static struct dentry *__d_find_alias(struct inode *inode, int want_discon) { struct dentry *alias, *discon_alias; - struct hlist_node *p; again: discon_alias = NULL; - hlist_for_each_entry(alias, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(alias, &inode->i_dentry, d_alias) { spin_lock(&alias->d_lock); if (S_ISDIR(inode->i_mode) || !d_unhashed(alias)) { if (IS_ROOT(alias) && @@ -730,10 +729,9 @@ EXPORT_SYMBOL(d_find_alias); void d_prune_aliases(struct inode *inode) { struct dentry *dentry; - struct hlist_node *p; restart: spin_lock(&inode->i_lock); - hlist_for_each_entry(dentry, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(dentry, &inode->i_dentry, d_alias) { spin_lock(&dentry->d_lock); if (!dentry->d_count) { __dget_dlock(dentry); @@ -1443,14 +1441,13 @@ static struct dentry *__d_instantiate_unique(struct dentry *entry, int len = entry->d_name.len; const char *name = entry->d_name.name; unsigned int hash = entry->d_name.hash; - struct hlist_node *p; if (!inode) { __d_instantiate(entry, NULL); return NULL; } - hlist_for_each_entry(alias, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(alias, &inode->i_dentry, d_alias) { /* * Don't need alias->d_lock here, because aliases with * d_parent == entry->d_parent are not subject to name or diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index dd87a31bcc21..4f5ad246582f 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -177,12 +177,11 @@ static inline int nodeid_hash(int nodeid) static struct connection *__find_con(int nodeid) { int r; - struct hlist_node *h; struct connection *con; r = nodeid_hash(nodeid); - hlist_for_each_entry(con, h, &connection_hash[r], list) { + hlist_for_each_entry(con, &connection_hash[r], list) { if (con->nodeid == nodeid) return con; } @@ -232,13 +231,12 @@ static struct connection *__nodeid2con(int nodeid, gfp_t alloc) static void foreach_conn(void (*conn_func)(struct connection *c)) { int i; - struct hlist_node *h, *n; + struct hlist_node *n; struct connection *con; for (i = 0; i < CONN_HASH_SIZE; i++) { - hlist_for_each_entry_safe(con, h, n, &connection_hash[i], list){ + hlist_for_each_entry_safe(con, n, &connection_hash[i], list) conn_func(con); - } } } @@ -257,13 +255,12 @@ static struct connection *nodeid2con(int nodeid, gfp_t allocation) static struct connection *assoc2con(int assoc_id) { int i; - struct hlist_node *h; struct connection *con; mutex_lock(&connections_lock); for (i = 0 ; i < CONN_HASH_SIZE; i++) { - hlist_for_each_entry(con, h, &connection_hash[i], list) { + hlist_for_each_entry(con, &connection_hash[i], list) { if (con->sctp_assoc == assoc_id) { mutex_unlock(&connections_lock); return con; diff --git a/fs/ecryptfs/messaging.c b/fs/ecryptfs/messaging.c index 5fa2471796c2..8d7a577ae497 100644 --- a/fs/ecryptfs/messaging.c +++ b/fs/ecryptfs/messaging.c @@ -115,10 +115,9 @@ void ecryptfs_msg_ctx_alloc_to_free(struct ecryptfs_msg_ctx *msg_ctx) */ int ecryptfs_find_daemon_by_euid(struct ecryptfs_daemon **daemon) { - struct hlist_node *elem; int rc; - hlist_for_each_entry(*daemon, elem, + hlist_for_each_entry(*daemon, &ecryptfs_daemon_hash[ecryptfs_current_euid_hash()], euid_chain) { if (uid_eq((*daemon)->file->f_cred->euid, current_euid())) { @@ -445,7 +444,6 @@ void ecryptfs_release_messaging(void) mutex_unlock(&ecryptfs_msg_ctx_lists_mux); } if (ecryptfs_daemon_hash) { - struct hlist_node *elem; struct ecryptfs_daemon *daemon; int i; @@ -453,7 +451,7 @@ void ecryptfs_release_messaging(void) for (i = 0; i < (1 << ecryptfs_hash_bits); i++) { int rc; - hlist_for_each_entry(daemon, elem, + hlist_for_each_entry(daemon, &ecryptfs_daemon_hash[i], euid_chain) { rc = ecryptfs_exorcise_daemon(daemon); diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c index 5df4bb4aab14..262fc9940982 100644 --- a/fs/exportfs/expfs.c +++ b/fs/exportfs/expfs.c @@ -44,14 +44,13 @@ find_acceptable_alias(struct dentry *result, { struct dentry *dentry, *toput = NULL; struct inode *inode; - struct hlist_node *p; if (acceptable(context, result)) return result; inode = result->d_inode; spin_lock(&inode->i_lock); - hlist_for_each_entry(dentry, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(dentry, &inode->i_dentry, d_alias) { dget(dentry); spin_unlock(&inode->i_lock); if (toput) diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 780e20806346..acf6e479b443 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -341,12 +341,11 @@ struct inode *fat_iget(struct super_block *sb, loff_t i_pos) { struct msdos_sb_info *sbi = MSDOS_SB(sb); struct hlist_head *head = sbi->inode_hashtable + fat_hash(i_pos); - struct hlist_node *_p; struct msdos_inode_info *i; struct inode *inode = NULL; spin_lock(&sbi->inode_hash_lock); - hlist_for_each_entry(i, _p, head, i_fat_hash) { + hlist_for_each_entry(i, head, i_fat_hash) { BUG_ON(i->vfs_inode.i_sb != sb); if (i->i_pos != i_pos) continue; diff --git a/fs/fat/nfs.c b/fs/fat/nfs.c index ef4b5faba87b..499c10438ca2 100644 --- a/fs/fat/nfs.c +++ b/fs/fat/nfs.c @@ -21,13 +21,12 @@ static struct inode *fat_dget(struct super_block *sb, int i_logstart) { struct msdos_sb_info *sbi = MSDOS_SB(sb); struct hlist_head *head; - struct hlist_node *_p; struct msdos_inode_info *i; struct inode *inode = NULL; head = sbi->dir_hashtable + fat_dir_hash(i_logstart); spin_lock(&sbi->dir_hash_lock); - hlist_for_each_entry(i, _p, head, i_dir_hash) { + hlist_for_each_entry(i, head, i_dir_hash) { BUG_ON(i->vfs_inode.i_sb != sb); if (i->i_logstart != i_logstart) continue; diff --git a/fs/fscache/cookie.c b/fs/fscache/cookie.c index 8dcb114758e3..e2cba1f60c21 100644 --- a/fs/fscache/cookie.c +++ b/fs/fscache/cookie.c @@ -237,13 +237,12 @@ static int fscache_alloc_object(struct fscache_cache *cache, struct fscache_cookie *cookie) { struct fscache_object *object; - struct hlist_node *_n; int ret; _enter("%p,%p{%s}", cache, cookie, cookie->def->name); spin_lock(&cookie->lock); - hlist_for_each_entry(object, _n, &cookie->backing_objects, + hlist_for_each_entry(object, &cookie->backing_objects, cookie_link) { if (object->cache == cache) goto object_already_extant; @@ -311,7 +310,6 @@ static int fscache_attach_object(struct fscache_cookie *cookie, { struct fscache_object *p; struct fscache_cache *cache = object->cache; - struct hlist_node *_n; int ret; _enter("{%s},{OBJ%x}", cookie->def->name, object->debug_id); @@ -321,7 +319,7 @@ static int fscache_attach_object(struct fscache_cookie *cookie, /* there may be multiple initial creations of this object, but we only * want one */ ret = -EEXIST; - hlist_for_each_entry(p, _n, &cookie->backing_objects, cookie_link) { + hlist_for_each_entry(p, &cookie->backing_objects, cookie_link) { if (p->cache == object->cache) { if (p->state >= FSCACHE_OBJECT_DYING) ret = -ENOBUFS; @@ -331,7 +329,7 @@ static int fscache_attach_object(struct fscache_cookie *cookie, /* pin the parent object */ spin_lock_nested(&cookie->parent->lock, 1); - hlist_for_each_entry(p, _n, &cookie->parent->backing_objects, + hlist_for_each_entry(p, &cookie->parent->backing_objects, cookie_link) { if (p->cache == object->cache) { if (p->state >= FSCACHE_OBJECT_DYING) { @@ -435,7 +433,6 @@ EXPORT_SYMBOL(__fscache_wait_on_invalidate); void __fscache_update_cookie(struct fscache_cookie *cookie) { struct fscache_object *object; - struct hlist_node *_p; fscache_stat(&fscache_n_updates); @@ -452,7 +449,7 @@ void __fscache_update_cookie(struct fscache_cookie *cookie) spin_lock(&cookie->lock); /* update the index entry on disk in each cache backing this cookie */ - hlist_for_each_entry(object, _p, + hlist_for_each_entry(object, &cookie->backing_objects, cookie_link) { fscache_raise_event(object, FSCACHE_OBJECT_EV_UPDATE); } diff --git a/fs/inode.c b/fs/inode.c index 67880e604399..f5f7c06c36fb 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -798,11 +798,10 @@ static struct inode *find_inode(struct super_block *sb, int (*test)(struct inode *, void *), void *data) { - struct hlist_node *node; struct inode *inode = NULL; repeat: - hlist_for_each_entry(inode, node, head, i_hash) { + hlist_for_each_entry(inode, head, i_hash) { spin_lock(&inode->i_lock); if (inode->i_sb != sb) { spin_unlock(&inode->i_lock); @@ -830,11 +829,10 @@ repeat: static struct inode *find_inode_fast(struct super_block *sb, struct hlist_head *head, unsigned long ino) { - struct hlist_node *node; struct inode *inode = NULL; repeat: - hlist_for_each_entry(inode, node, head, i_hash) { + hlist_for_each_entry(inode, head, i_hash) { spin_lock(&inode->i_lock); if (inode->i_ino != ino) { spin_unlock(&inode->i_lock); @@ -1132,11 +1130,10 @@ EXPORT_SYMBOL(iget_locked); static int test_inode_iunique(struct super_block *sb, unsigned long ino) { struct hlist_head *b = inode_hashtable + hash(sb, ino); - struct hlist_node *node; struct inode *inode; spin_lock(&inode_hash_lock); - hlist_for_each_entry(inode, node, b, i_hash) { + hlist_for_each_entry(inode, b, i_hash) { if (inode->i_ino == ino && inode->i_sb == sb) { spin_unlock(&inode_hash_lock); return 0; @@ -1291,10 +1288,9 @@ int insert_inode_locked(struct inode *inode) struct hlist_head *head = inode_hashtable + hash(sb, ino); while (1) { - struct hlist_node *node; struct inode *old = NULL; spin_lock(&inode_hash_lock); - hlist_for_each_entry(old, node, head, i_hash) { + hlist_for_each_entry(old, head, i_hash) { if (old->i_ino != ino) continue; if (old->i_sb != sb) @@ -1306,7 +1302,7 @@ int insert_inode_locked(struct inode *inode) } break; } - if (likely(!node)) { + if (likely(!old)) { spin_lock(&inode->i_lock); inode->i_state |= I_NEW; hlist_add_head(&inode->i_hash, head); @@ -1334,11 +1330,10 @@ int insert_inode_locked4(struct inode *inode, unsigned long hashval, struct hlist_head *head = inode_hashtable + hash(sb, hashval); while (1) { - struct hlist_node *node; struct inode *old = NULL; spin_lock(&inode_hash_lock); - hlist_for_each_entry(old, node, head, i_hash) { + hlist_for_each_entry(old, head, i_hash) { if (old->i_sb != sb) continue; if (!test(old, data)) @@ -1350,7 +1345,7 @@ int insert_inode_locked4(struct inode *inode, unsigned long hashval, } break; } - if (likely(!node)) { + if (likely(!old)) { spin_lock(&inode->i_lock); inode->i_state |= I_NEW; hlist_add_head(&inode->i_hash, head); diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 0e17090c310f..abdd75d44dd4 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -32,15 +32,15 @@ static struct hlist_head nlm_server_hosts[NLM_HOST_NRHASH]; static struct hlist_head nlm_client_hosts[NLM_HOST_NRHASH]; -#define for_each_host(host, pos, chain, table) \ +#define for_each_host(host, chain, table) \ for ((chain) = (table); \ (chain) < (table) + NLM_HOST_NRHASH; ++(chain)) \ - hlist_for_each_entry((host), (pos), (chain), h_hash) + hlist_for_each_entry((host), (chain), h_hash) -#define for_each_host_safe(host, pos, next, chain, table) \ +#define for_each_host_safe(host, next, chain, table) \ for ((chain) = (table); \ (chain) < (table) + NLM_HOST_NRHASH; ++(chain)) \ - hlist_for_each_entry_safe((host), (pos), (next), \ + hlist_for_each_entry_safe((host), (next), \ (chain), h_hash) static unsigned long nrhosts; @@ -225,7 +225,6 @@ struct nlm_host *nlmclnt_lookup_host(const struct sockaddr *sap, .net = net, }; struct hlist_head *chain; - struct hlist_node *pos; struct nlm_host *host; struct nsm_handle *nsm = NULL; struct lockd_net *ln = net_generic(net, lockd_net_id); @@ -237,7 +236,7 @@ struct nlm_host *nlmclnt_lookup_host(const struct sockaddr *sap, mutex_lock(&nlm_host_mutex); chain = &nlm_client_hosts[nlm_hash_address(sap)]; - hlist_for_each_entry(host, pos, chain, h_hash) { + hlist_for_each_entry(host, chain, h_hash) { if (host->net != net) continue; if (!rpc_cmp_addr(nlm_addr(host), sap)) @@ -322,7 +321,6 @@ struct nlm_host *nlmsvc_lookup_host(const struct svc_rqst *rqstp, const size_t hostname_len) { struct hlist_head *chain; - struct hlist_node *pos; struct nlm_host *host = NULL; struct nsm_handle *nsm = NULL; struct sockaddr *src_sap = svc_daddr(rqstp); @@ -350,7 +348,7 @@ struct nlm_host *nlmsvc_lookup_host(const struct svc_rqst *rqstp, nlm_gc_hosts(net); chain = &nlm_server_hosts[nlm_hash_address(ni.sap)]; - hlist_for_each_entry(host, pos, chain, h_hash) { + hlist_for_each_entry(host, chain, h_hash) { if (host->net != net) continue; if (!rpc_cmp_addr(nlm_addr(host), ni.sap)) @@ -515,10 +513,9 @@ static struct nlm_host *next_host_state(struct hlist_head *cache, { struct nlm_host *host; struct hlist_head *chain; - struct hlist_node *pos; mutex_lock(&nlm_host_mutex); - for_each_host(host, pos, chain, cache) { + for_each_host(host, chain, cache) { if (host->h_nsmhandle == nsm && host->h_nsmstate != info->state) { host->h_nsmstate = info->state; @@ -570,7 +567,6 @@ void nlm_host_rebooted(const struct nlm_reboot *info) static void nlm_complain_hosts(struct net *net) { struct hlist_head *chain; - struct hlist_node *pos; struct nlm_host *host; if (net) { @@ -587,7 +583,7 @@ static void nlm_complain_hosts(struct net *net) dprintk("lockd: %lu hosts left:\n", nrhosts); } - for_each_host(host, pos, chain, nlm_server_hosts) { + for_each_host(host, chain, nlm_server_hosts) { if (net && host->net != net) continue; dprintk(" %s (cnt %d use %d exp %ld net %p)\n", @@ -600,14 +596,13 @@ void nlm_shutdown_hosts_net(struct net *net) { struct hlist_head *chain; - struct hlist_node *pos; struct nlm_host *host; mutex_lock(&nlm_host_mutex); /* First, make all hosts eligible for gc */ dprintk("lockd: nuking all hosts in net %p...\n", net); - for_each_host(host, pos, chain, nlm_server_hosts) { + for_each_host(host, chain, nlm_server_hosts) { if (net && host->net != net) continue; host->h_expires = jiffies - 1; @@ -644,11 +639,11 @@ static void nlm_gc_hosts(struct net *net) { struct hlist_head *chain; - struct hlist_node *pos, *next; + struct hlist_node *next; struct nlm_host *host; dprintk("lockd: host garbage collection for net %p\n", net); - for_each_host(host, pos, chain, nlm_server_hosts) { + for_each_host(host, chain, nlm_server_hosts) { if (net && host->net != net) continue; host->h_inuse = 0; @@ -657,7 +652,7 @@ nlm_gc_hosts(struct net *net) /* Mark all hosts that hold locks, blocks or shares */ nlmsvc_mark_resources(net); - for_each_host_safe(host, pos, next, chain, nlm_server_hosts) { + for_each_host_safe(host, next, chain, nlm_server_hosts) { if (net && host->net != net) continue; if (atomic_read(&host->h_count) || host->h_inuse diff --git a/fs/lockd/svcsubs.c b/fs/lockd/svcsubs.c index b3a24b07d981..d17bb62b06d6 100644 --- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -84,7 +84,6 @@ __be32 nlm_lookup_file(struct svc_rqst *rqstp, struct nlm_file **result, struct nfs_fh *f) { - struct hlist_node *pos; struct nlm_file *file; unsigned int hash; __be32 nfserr; @@ -96,7 +95,7 @@ nlm_lookup_file(struct svc_rqst *rqstp, struct nlm_file **result, /* Lock file table */ mutex_lock(&nlm_file_mutex); - hlist_for_each_entry(file, pos, &nlm_files[hash], f_list) + hlist_for_each_entry(file, &nlm_files[hash], f_list) if (!nfs_compare_fh(&file->f_handle, f)) goto found; @@ -248,13 +247,13 @@ static int nlm_traverse_files(void *data, nlm_host_match_fn_t match, int (*is_failover_file)(void *data, struct nlm_file *file)) { - struct hlist_node *pos, *next; + struct hlist_node *next; struct nlm_file *file; int i, ret = 0; mutex_lock(&nlm_file_mutex); for (i = 0; i < FILE_NRHASH; i++) { - hlist_for_each_entry_safe(file, pos, next, &nlm_files[i], f_list) { + hlist_for_each_entry_safe(file, next, &nlm_files[i], f_list) { if (is_failover_file && !is_failover_file(data, file)) continue; file->f_count++; diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c index d35b62e83ea6..6da209bd9408 100644 --- a/fs/nfs/pnfs_dev.c +++ b/fs/nfs/pnfs_dev.c @@ -77,9 +77,8 @@ _lookup_deviceid(const struct pnfs_layoutdriver_type *ld, long hash) { struct nfs4_deviceid_node *d; - struct hlist_node *n; - hlist_for_each_entry_rcu(d, n, &nfs4_deviceid_cache[hash], node) + hlist_for_each_entry_rcu(d, &nfs4_deviceid_cache[hash], node) if (d->ld == ld && d->nfs_client == clp && !memcmp(&d->deviceid, id, sizeof(*id))) { if (atomic_read(&d->ref)) @@ -248,12 +247,11 @@ static void _deviceid_purge_client(const struct nfs_client *clp, long hash) { struct nfs4_deviceid_node *d; - struct hlist_node *n; HLIST_HEAD(tmp); spin_lock(&nfs4_deviceid_lock); rcu_read_lock(); - hlist_for_each_entry_rcu(d, n, &nfs4_deviceid_cache[hash], node) + hlist_for_each_entry_rcu(d, &nfs4_deviceid_cache[hash], node) if (d->nfs_client == clp && atomic_read(&d->ref)) { hlist_del_init_rcu(&d->node); hlist_add_head(&d->tmpnode, &tmp); @@ -291,12 +289,11 @@ void nfs4_deviceid_mark_client_invalid(struct nfs_client *clp) { struct nfs4_deviceid_node *d; - struct hlist_node *n; int i; rcu_read_lock(); for (i = 0; i < NFS4_DEVICE_ID_HASH_SIZE; i ++){ - hlist_for_each_entry_rcu(d, n, &nfs4_deviceid_cache[i], node) + hlist_for_each_entry_rcu(d, &nfs4_deviceid_cache[i], node) if (d->nfs_client == clp) set_bit(NFS_DEVICEID_INVALID, &d->flags); } diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 2cbac34a55da..da3dbd0f8979 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -120,7 +120,6 @@ hash_refile(struct svc_cacherep *rp) int nfsd_cache_lookup(struct svc_rqst *rqstp) { - struct hlist_node *hn; struct hlist_head *rh; struct svc_cacherep *rp; __be32 xid = rqstp->rq_xid; @@ -141,7 +140,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) rtn = RC_DOIT; rh = &cache_hash[request_hash(xid)]; - hlist_for_each_entry(rp, hn, rh, c_hash) { + hlist_for_each_entry(rp, rh, c_hash) { if (rp->c_state != RC_UNUSED && xid == rp->c_xid && proc == rp->c_proc && proto == rp->c_prot && vers == rp->c_vers && diff --git a/fs/notify/fsnotify.c b/fs/notify/fsnotify.c index 6baadb5a8430..4bb21d67d9b1 100644 --- a/fs/notify/fsnotify.c +++ b/fs/notify/fsnotify.c @@ -52,7 +52,6 @@ void __fsnotify_vfsmount_delete(struct vfsmount *mnt) void __fsnotify_update_child_dentry_flags(struct inode *inode) { struct dentry *alias; - struct hlist_node *p; int watched; if (!S_ISDIR(inode->i_mode)) @@ -64,7 +63,7 @@ void __fsnotify_update_child_dentry_flags(struct inode *inode) spin_lock(&inode->i_lock); /* run all of the dentries associated with this inode. Since this is a * directory, there damn well better only be one item on this list */ - hlist_for_each_entry(alias, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(alias, &inode->i_dentry, d_alias) { struct dentry *child; /* run all of the children of the original inode and fix their diff --git a/fs/notify/inode_mark.c b/fs/notify/inode_mark.c index f31e90fc050d..74825be65b7b 100644 --- a/fs/notify/inode_mark.c +++ b/fs/notify/inode_mark.c @@ -36,12 +36,11 @@ static void fsnotify_recalc_inode_mask_locked(struct inode *inode) { struct fsnotify_mark *mark; - struct hlist_node *pos; __u32 new_mask = 0; assert_spin_locked(&inode->i_lock); - hlist_for_each_entry(mark, pos, &inode->i_fsnotify_marks, i.i_list) + hlist_for_each_entry(mark, &inode->i_fsnotify_marks, i.i_list) new_mask |= mark->mask; inode->i_fsnotify_mask = new_mask; } @@ -87,11 +86,11 @@ void fsnotify_destroy_inode_mark(struct fsnotify_mark *mark) void fsnotify_clear_marks_by_inode(struct inode *inode) { struct fsnotify_mark *mark, *lmark; - struct hlist_node *pos, *n; + struct hlist_node *n; LIST_HEAD(free_list); spin_lock(&inode->i_lock); - hlist_for_each_entry_safe(mark, pos, n, &inode->i_fsnotify_marks, i.i_list) { + hlist_for_each_entry_safe(mark, n, &inode->i_fsnotify_marks, i.i_list) { list_add(&mark->i.free_i_list, &free_list); hlist_del_init_rcu(&mark->i.i_list); fsnotify_get_mark(mark); @@ -129,11 +128,10 @@ static struct fsnotify_mark *fsnotify_find_inode_mark_locked( struct inode *inode) { struct fsnotify_mark *mark; - struct hlist_node *pos; assert_spin_locked(&inode->i_lock); - hlist_for_each_entry(mark, pos, &inode->i_fsnotify_marks, i.i_list) { + hlist_for_each_entry(mark, &inode->i_fsnotify_marks, i.i_list) { if (mark->group == group) { fsnotify_get_mark(mark); return mark; @@ -194,8 +192,7 @@ int fsnotify_add_inode_mark(struct fsnotify_mark *mark, struct fsnotify_group *group, struct inode *inode, int allow_dups) { - struct fsnotify_mark *lmark; - struct hlist_node *node, *last = NULL; + struct fsnotify_mark *lmark, *last = NULL; int ret = 0; mark->flags |= FSNOTIFY_MARK_FLAG_INODE; @@ -214,8 +211,8 @@ int fsnotify_add_inode_mark(struct fsnotify_mark *mark, } /* should mark be in the middle of the current list? */ - hlist_for_each_entry(lmark, node, &inode->i_fsnotify_marks, i.i_list) { - last = node; + hlist_for_each_entry(lmark, &inode->i_fsnotify_marks, i.i_list) { + last = lmark; if ((lmark->group == group) && !allow_dups) { ret = -EEXIST; @@ -235,7 +232,7 @@ int fsnotify_add_inode_mark(struct fsnotify_mark *mark, BUG_ON(last == NULL); /* mark should be the last entry. last is the current last entry */ - hlist_add_after_rcu(last, &mark->i.i_list); + hlist_add_after_rcu(&last->i.i_list, &mark->i.i_list); out: fsnotify_recalc_inode_mask_locked(inode); spin_unlock(&inode->i_lock); diff --git a/fs/notify/vfsmount_mark.c b/fs/notify/vfsmount_mark.c index 4df58b8ea64a..68ca5a8704b5 100644 --- a/fs/notify/vfsmount_mark.c +++ b/fs/notify/vfsmount_mark.c @@ -33,12 +33,12 @@ void fsnotify_clear_marks_by_mount(struct vfsmount *mnt) { struct fsnotify_mark *mark, *lmark; - struct hlist_node *pos, *n; + struct hlist_node *n; struct mount *m = real_mount(mnt); LIST_HEAD(free_list); spin_lock(&mnt->mnt_root->d_lock); - hlist_for_each_entry_safe(mark, pos, n, &m->mnt_fsnotify_marks, m.m_list) { + hlist_for_each_entry_safe(mark, n, &m->mnt_fsnotify_marks, m.m_list) { list_add(&mark->m.free_m_list, &free_list); hlist_del_init_rcu(&mark->m.m_list); fsnotify_get_mark(mark); @@ -71,12 +71,11 @@ static void fsnotify_recalc_vfsmount_mask_locked(struct vfsmount *mnt) { struct mount *m = real_mount(mnt); struct fsnotify_mark *mark; - struct hlist_node *pos; __u32 new_mask = 0; assert_spin_locked(&mnt->mnt_root->d_lock); - hlist_for_each_entry(mark, pos, &m->mnt_fsnotify_marks, m.m_list) + hlist_for_each_entry(mark, &m->mnt_fsnotify_marks, m.m_list) new_mask |= mark->mask; m->mnt_fsnotify_mask = new_mask; } @@ -114,11 +113,10 @@ static struct fsnotify_mark *fsnotify_find_vfsmount_mark_locked(struct fsnotify_ { struct mount *m = real_mount(mnt); struct fsnotify_mark *mark; - struct hlist_node *pos; assert_spin_locked(&mnt->mnt_root->d_lock); - hlist_for_each_entry(mark, pos, &m->mnt_fsnotify_marks, m.m_list) { + hlist_for_each_entry(mark, &m->mnt_fsnotify_marks, m.m_list) { if (mark->group == group) { fsnotify_get_mark(mark); return mark; @@ -153,8 +151,7 @@ int fsnotify_add_vfsmount_mark(struct fsnotify_mark *mark, int allow_dups) { struct mount *m = real_mount(mnt); - struct fsnotify_mark *lmark; - struct hlist_node *node, *last = NULL; + struct fsnotify_mark *lmark, *last = NULL; int ret = 0; mark->flags |= FSNOTIFY_MARK_FLAG_VFSMOUNT; @@ -173,8 +170,8 @@ int fsnotify_add_vfsmount_mark(struct fsnotify_mark *mark, } /* should mark be in the middle of the current list? */ - hlist_for_each_entry(lmark, node, &m->mnt_fsnotify_marks, m.m_list) { - last = node; + hlist_for_each_entry(lmark, &m->mnt_fsnotify_marks, m.m_list) { + last = lmark; if ((lmark->group == group) && !allow_dups) { ret = -EEXIST; @@ -194,7 +191,7 @@ int fsnotify_add_vfsmount_mark(struct fsnotify_mark *mark, BUG_ON(last == NULL); /* mark should be the last entry. last is the current last entry */ - hlist_add_after_rcu(last, &mark->m.m_list); + hlist_add_after_rcu(&last->m.m_list, &mark->m.m_list); out: fsnotify_recalc_vfsmount_mask_locked(mnt); spin_unlock(&mnt->mnt_root->d_lock); diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c index 8db4b58b2e4b..ef999729e274 100644 --- a/fs/ocfs2/dcache.c +++ b/fs/ocfs2/dcache.c @@ -169,11 +169,10 @@ struct dentry *ocfs2_find_local_alias(struct inode *inode, u64 parent_blkno, int skip_unhashed) { - struct hlist_node *p; struct dentry *dentry; spin_lock(&inode->i_lock); - hlist_for_each_entry(dentry, p, &inode->i_dentry, d_alias) { + hlist_for_each_entry(dentry, &inode->i_dentry, d_alias) { spin_lock(&dentry->d_lock); if (ocfs2_match_dentry(dentry, parent_blkno, skip_unhashed)) { trace_ocfs2_find_local_alias(dentry->d_name.len, diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index 01ebfd0bdad7..eeac97bb3bfa 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -2083,7 +2083,6 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm, u8 dead_node, u8 new_master) { int i; - struct hlist_node *hash_iter; struct hlist_head *bucket; struct dlm_lock_resource *res, *next; @@ -2114,7 +2113,7 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm, * if necessary */ for (i = 0; i < DLM_HASH_BUCKETS; i++) { bucket = dlm_lockres_hash(dlm, i); - hlist_for_each_entry(res, hash_iter, bucket, hash_node) { + hlist_for_each_entry(res, bucket, hash_node) { if (!(res->state & DLM_LOCK_RES_RECOVERING)) continue; @@ -2273,7 +2272,6 @@ static void dlm_free_dead_locks(struct dlm_ctxt *dlm, static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node) { - struct hlist_node *iter; struct dlm_lock_resource *res; int i; struct hlist_head *bucket; @@ -2299,7 +2297,7 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node) */ for (i = 0; i < DLM_HASH_BUCKETS; i++) { bucket = dlm_lockres_hash(dlm, i); - hlist_for_each_entry(res, iter, bucket, hash_node) { + hlist_for_each_entry(res, bucket, hash_node) { /* always prune any $RECOVERY entries for dead nodes, * otherwise hangs can occur during later recovery */ if (dlm_is_recovery_lock(res->lockname.name, diff --git a/fs/super.c b/fs/super.c index df6c2f4c6b59..7465d4364208 100644 --- a/fs/super.c +++ b/fs/super.c @@ -447,14 +447,13 @@ struct super_block *sget(struct file_system_type *type, void *data) { struct super_block *s = NULL; - struct hlist_node *node; struct super_block *old; int err; retry: spin_lock(&sb_lock); if (test) { - hlist_for_each_entry(old, node, &type->fs_supers, s_instances) { + hlist_for_each_entry(old, &type->fs_supers, s_instances) { if (!test(old, data)) continue; if (!grab_super(old)) @@ -554,10 +553,9 @@ void iterate_supers_type(struct file_system_type *type, void (*f)(struct super_block *, void *), void *arg) { struct super_block *sb, *p = NULL; - struct hlist_node *node; spin_lock(&sb_lock); - hlist_for_each_entry(sb, node, &type->fs_supers, s_instances) { + hlist_for_each_entry(sb, &type->fs_supers, s_instances) { sb->s_count++; spin_unlock(&sb_lock); diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index 2ce9a5db6ab5..15c68f9489ae 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c @@ -461,14 +461,13 @@ const struct file_operations bin_fops = { void unmap_bin_file(struct sysfs_dirent *attr_sd) { struct bin_buffer *bb; - struct hlist_node *tmp; if (sysfs_type(attr_sd) != SYSFS_KOBJ_BIN_ATTR) return; mutex_lock(&sysfs_bin_lock); - hlist_for_each_entry(bb, tmp, &attr_sd->s_bin_attr.buffers, list) { + hlist_for_each_entry(bb, &attr_sd->s_bin_attr.buffers, list) { struct inode *inode = file_inode(bb->file); unmap_mapping_range(inode->i_mapping, 0, 0, 1); diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 96fcbb85ff83..d1dba7ce75ae 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -1442,9 +1442,8 @@ xlog_recover_find_tid( xlog_tid_t tid) { xlog_recover_t *trans; - struct hlist_node *n; - hlist_for_each_entry(trans, n, head, r_list) { + hlist_for_each_entry(trans, head, r_list) { if (trans->r_log_tid == tid) return trans; } -- cgit From 634095dab2a2001844fc8b26673c0cb14a766cdf Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 27 Feb 2013 22:37:21 -0500 Subject: 9p: don't bother with private lock in ->d_fsdata; dentry->d_lock will do just fine Signed-off-by: Al Viro --- fs/9p/fid.c | 9 ++++----- fs/9p/fid.h | 4 ++-- 2 files changed, 6 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index afd4724b2d92..71bc36a03e72 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -54,14 +54,13 @@ int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) if (!dent) return -ENOMEM; - spin_lock_init(&dent->lock); INIT_LIST_HEAD(&dent->fidlist); dentry->d_fsdata = dent; } - spin_lock(&dent->lock); + spin_lock(&dentry->d_lock); list_add(&fid->dlist, &dent->fidlist); - spin_unlock(&dent->lock); + spin_unlock(&dentry->d_lock); return 0; } @@ -85,14 +84,14 @@ static struct p9_fid *v9fs_fid_find(struct dentry *dentry, kuid_t uid, int any) dent = (struct v9fs_dentry *) dentry->d_fsdata; ret = NULL; if (dent) { - spin_lock(&dent->lock); + spin_lock(&dentry->d_lock); list_for_each_entry(fid, &dent->fidlist, dlist) { if (any || uid_eq(fid->uid, uid)) { ret = fid; break; } } - spin_unlock(&dent->lock); + spin_unlock(&dentry->d_lock); } return ret; diff --git a/fs/9p/fid.h b/fs/9p/fid.h index bb0b6e7f58fc..469b5d517694 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -25,7 +25,6 @@ /** * struct v9fs_dentry - 9p private data stored in dentry d_fsdata - * @lock: protects the fidlist * @fidlist: list of FIDs currently associated with this dentry * * This structure defines the 9p private data associated with @@ -35,11 +34,12 @@ * inodes in order to more closely map functionality to the Plan 9 * expected behavior for FID reclaimation and tracking. * + * Protected by ->d_lock of dentry it belongs to. + * * See Also: Mapping FIDs to Linux VFS model in * Design and Implementation of the Linux 9P File System documentation */ struct v9fs_dentry { - spinlock_t lock; /* protect fidlist */ struct list_head fidlist; }; -- cgit From c4d30967f3020cda9df9ee22af79cd1f2c284244 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 27 Feb 2013 22:51:08 -0500 Subject: 9p: turn fid->dlist into hlist Signed-off-by: Al Viro --- fs/9p/fid.c | 7 ++++--- fs/9p/fid.h | 2 +- fs/9p/vfs_dentry.c | 9 +++------ 3 files changed, 8 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 71bc36a03e72..49de4264db9a 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -54,12 +54,12 @@ int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) if (!dent) return -ENOMEM; - INIT_LIST_HEAD(&dent->fidlist); + INIT_HLIST_HEAD(&dent->fidlist); dentry->d_fsdata = dent; } spin_lock(&dentry->d_lock); - list_add(&fid->dlist, &dent->fidlist); + hlist_add_head(&fid->dlist, &dent->fidlist); spin_unlock(&dentry->d_lock); return 0; @@ -84,8 +84,9 @@ static struct p9_fid *v9fs_fid_find(struct dentry *dentry, kuid_t uid, int any) dent = (struct v9fs_dentry *) dentry->d_fsdata; ret = NULL; if (dent) { + struct hlist_node *n; spin_lock(&dentry->d_lock); - list_for_each_entry(fid, &dent->fidlist, dlist) { + hlist_for_each_entry(fid, n, &dent->fidlist, dlist) { if (any || uid_eq(fid->uid, uid)) { ret = fid; break; diff --git a/fs/9p/fid.h b/fs/9p/fid.h index 469b5d517694..86eeb34ce46f 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -40,7 +40,7 @@ * Design and Implementation of the Linux 9P File System documentation */ struct v9fs_dentry { - struct list_head fidlist; + struct hlist_head fidlist; }; struct p9_fid *v9fs_fid_lookup(struct dentry *dentry); diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index 9ad68628522c..fcd49833ef80 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -84,16 +84,13 @@ static int v9fs_cached_dentry_delete(const struct dentry *dentry) static void v9fs_dentry_release(struct dentry *dentry) { struct v9fs_dentry *dent; - struct p9_fid *temp, *current_fid; - p9_debug(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name, dentry); dent = dentry->d_fsdata; if (dent) { - list_for_each_entry_safe(current_fid, temp, &dent->fidlist, - dlist) { - p9_client_clunk(current_fid); - } + struct hlist_node *p, *n; + hlist_for_each_safe(p, n, &dent->fidlist) + p9_client_clunk(hlist_entry(p, struct p9_fid, dlist)); kfree(dent); dentry->d_fsdata = NULL; -- cgit From aaeb7ecfb48ad4c8942a26874322d8918524a04f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 28 Feb 2013 01:13:19 -0500 Subject: v9fs: get rid of v9fs_dentry ->d_fsdata can act as hlist_head... Signed-off-by: Al Viro --- fs/9p/fid.c | 26 +++++--------------------- fs/9p/fid.h | 20 -------------------- fs/9p/vfs_dentry.c | 14 ++++---------- 3 files changed, 9 insertions(+), 51 deletions(-) (limited to 'fs') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 49de4264db9a..ddf618936d8a 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -43,25 +43,9 @@ int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) { - struct v9fs_dentry *dent; - - p9_debug(P9_DEBUG_VFS, "fid %d dentry %s\n", - fid->fid, dentry->d_name.name); - - dent = dentry->d_fsdata; - if (!dent) { - dent = kmalloc(sizeof(struct v9fs_dentry), GFP_KERNEL); - if (!dent) - return -ENOMEM; - - INIT_HLIST_HEAD(&dent->fidlist); - dentry->d_fsdata = dent; - } - spin_lock(&dentry->d_lock); - hlist_add_head(&fid->dlist, &dent->fidlist); + hlist_add_head(&fid->dlist, (struct hlist_head *)&dentry->d_fsdata); spin_unlock(&dentry->d_lock); - return 0; } @@ -75,18 +59,18 @@ int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) static struct p9_fid *v9fs_fid_find(struct dentry *dentry, kuid_t uid, int any) { - struct v9fs_dentry *dent; struct p9_fid *fid, *ret; p9_debug(P9_DEBUG_VFS, " dentry: %s (%p) uid %d any %d\n", dentry->d_name.name, dentry, from_kuid(&init_user_ns, uid), any); - dent = (struct v9fs_dentry *) dentry->d_fsdata; ret = NULL; - if (dent) { + /* we'll recheck under lock if there's anything to look in */ + if (dentry->d_fsdata) { + struct hlist_head *h = (struct hlist_head *)&dentry->d_fsdata; struct hlist_node *n; spin_lock(&dentry->d_lock); - hlist_for_each_entry(fid, n, &dent->fidlist, dlist) { + hlist_for_each_entry(fid, n, h, dlist) { if (any || uid_eq(fid->uid, uid)) { ret = fid; break; diff --git a/fs/9p/fid.h b/fs/9p/fid.h index 86eeb34ce46f..377eb6f2e7f7 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -23,26 +23,6 @@ #define FS_9P_FID_H #include -/** - * struct v9fs_dentry - 9p private data stored in dentry d_fsdata - * @fidlist: list of FIDs currently associated with this dentry - * - * This structure defines the 9p private data associated with - * a particular dentry. In particular, this private data is used - * to lookup which 9P FID handle should be used for a particular VFS - * operation. FID handles are associated with dentries instead of - * inodes in order to more closely map functionality to the Plan 9 - * expected behavior for FID reclaimation and tracking. - * - * Protected by ->d_lock of dentry it belongs to. - * - * See Also: Mapping FIDs to Linux VFS model in - * Design and Implementation of the Linux 9P File System documentation - */ -struct v9fs_dentry { - struct hlist_head fidlist; -}; - struct p9_fid *v9fs_fid_lookup(struct dentry *dentry); struct p9_fid *v9fs_fid_clone(struct dentry *dentry); int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid); diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index fcd49833ef80..f039b104a98e 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -83,18 +83,12 @@ static int v9fs_cached_dentry_delete(const struct dentry *dentry) static void v9fs_dentry_release(struct dentry *dentry) { - struct v9fs_dentry *dent; + struct hlist_node *p, *n; p9_debug(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name, dentry); - dent = dentry->d_fsdata; - if (dent) { - struct hlist_node *p, *n; - hlist_for_each_safe(p, n, &dent->fidlist) - p9_client_clunk(hlist_entry(p, struct p9_fid, dlist)); - - kfree(dent); - dentry->d_fsdata = NULL; - } + hlist_for_each_safe(p, n, (struct hlist_head *)&dentry->d_fsdata) + p9_client_clunk(hlist_entry(p, struct p9_fid, dlist)); + dentry->d_fsdata = NULL; } static int v9fs_lookup_revalidate(struct dentry *dentry, unsigned int flags) -- cgit From 2ea03e1d62d56c37737e43550c360b43a5e40a32 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 28 Feb 2013 01:18:14 -0500 Subject: 9p: v9fs_fid_add() can't fail now Signed-off-by: Al Viro --- fs/9p/fid.c | 3 +-- fs/9p/fid.h | 2 +- fs/9p/vfs_inode.c | 9 ++------- fs/9p/vfs_inode_dotl.c | 19 +++++++------------ 4 files changed, 11 insertions(+), 22 deletions(-) (limited to 'fs') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index ddf618936d8a..73ca55042cb0 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -41,12 +41,11 @@ * */ -int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) +void v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) { spin_lock(&dentry->d_lock); hlist_add_head(&fid->dlist, (struct hlist_head *)&dentry->d_fsdata); spin_unlock(&dentry->d_lock); - return 0; } /** diff --git a/fs/9p/fid.h b/fs/9p/fid.h index 377eb6f2e7f7..2b6787fcb626 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -25,6 +25,6 @@ struct p9_fid *v9fs_fid_lookup(struct dentry *dentry); struct p9_fid *v9fs_fid_clone(struct dentry *dentry); -int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid); +void v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid); struct p9_fid *v9fs_writeback_fid(struct dentry *dentry); #endif diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index b5340c829de1..ce601d71a427 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -692,9 +692,7 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, "inode creation failed %d\n", err); goto error; } - err = v9fs_fid_add(dentry, fid); - if (err < 0) - goto error; + v9fs_fid_add(dentry, fid); d_instantiate(dentry, inode); } return ofid; @@ -830,9 +828,7 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, inode = NULL; goto error; } - result = v9fs_fid_add(dentry, fid); - if (result < 0) - goto error_iput; + v9fs_fid_add(dentry, fid); inst_out: /* * If we had a rename on the server and a parallel lookup @@ -845,7 +841,6 @@ inst_out: if (!IS_ERR(res)) return res; result = PTR_ERR(res); -error_iput: iput(inode); error: p9_client_clunk(fid); diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index 61e4fa70a6fa..53687bbf2296 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -333,9 +333,7 @@ v9fs_vfs_atomic_open_dotl(struct inode *dir, struct dentry *dentry, /* Now set the ACL based on the default value */ v9fs_set_create_acl(inode, fid, dacl, pacl); - err = v9fs_fid_add(dentry, fid); - if (err < 0) - goto error; + v9fs_fid_add(dentry, fid); d_instantiate(dentry, inode); v9inode = V9FS_I(inode); @@ -453,12 +451,11 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, err); goto error; } - err = v9fs_fid_add(dentry, fid); - if (err < 0) - goto error; + v9fs_fid_add(dentry, fid); v9fs_set_create_acl(inode, fid, dacl, pacl); d_instantiate(dentry, inode); fid = NULL; + err = 0; } else { /* * Not in cached mode. No need to populate @@ -747,11 +744,10 @@ v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry, err); goto error; } - err = v9fs_fid_add(dentry, fid); - if (err < 0) - goto error; + v9fs_fid_add(dentry, fid); d_instantiate(dentry, inode); fid = NULL; + err = 0; } else { /* Not in cached mode. No need to populate inode with stat */ inode = v9fs_get_inode(dir->i_sb, S_IFLNK, 0); @@ -900,11 +896,10 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode, goto error; } v9fs_set_create_acl(inode, fid, dacl, pacl); - err = v9fs_fid_add(dentry, fid); - if (err < 0) - goto error; + v9fs_fid_add(dentry, fid); d_instantiate(dentry, inode); fid = NULL; + err = 0; } else { /* * Not in cached mode. No need to populate inode with stat. -- cgit From 3509b678a6bb93a49d9603c9c8028c8d95019539 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 28 Feb 2013 01:21:38 -0500 Subject: 9p: double iput() in ->lookup() if d_materialise_unique() fails d_materialise_unique() does iput() itself. Signed-off-by: Al Viro --- fs/9p/vfs_inode.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index ce601d71a427..cbee5ec4039d 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -841,7 +841,6 @@ inst_out: if (!IS_ERR(res)) return res; result = PTR_ERR(res); - iput(inode); error: p9_client_clunk(fid); -- cgit From 7b5be621993567d39f09a5190c4d651241be296f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 28 Feb 2013 01:28:21 -0500 Subject: 9p: untangle ->lookup() a bit Signed-off-by: Al Viro --- fs/9p/vfs_inode.c | 27 +++++++++------------------ 1 file changed, 9 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index cbee5ec4039d..80ff01bcae3d 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -788,7 +788,6 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, struct p9_fid *dfid, *fid; struct inode *inode; char *name; - int result = 0; p9_debug(P9_DEBUG_VFS, "dir: %p dentry: (%s) %p flags: %x\n", dir, dentry->d_name.name, dentry, flags); @@ -806,13 +805,11 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, name = (char *) dentry->d_name.name; fid = p9_client_walk(dfid, 1, &name, 1); if (IS_ERR(fid)) { - result = PTR_ERR(fid); - if (result == -ENOENT) { - inode = NULL; - goto inst_out; + if (fid == ERR_PTR(-ENOENT)) { + d_add(dentry, NULL); + return NULL; } - - return ERR_PTR(result); + return ERR_CAST(fid); } /* * Make sure we don't use a wrong inode due to parallel @@ -824,12 +821,10 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, else inode = v9fs_get_new_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { - result = PTR_ERR(inode); - inode = NULL; - goto error; + p9_client_clunk(fid); + return ERR_CAST(inode); } v9fs_fid_add(dentry, fid); -inst_out: /* * If we had a rename on the server and a parallel lookup * for the new name, then make sure we instantiate with @@ -838,13 +833,9 @@ inst_out: * k/b. */ res = d_materialise_unique(dentry, inode); - if (!IS_ERR(res)) - return res; - result = PTR_ERR(res); -error: - p9_client_clunk(fid); - - return ERR_PTR(result); + if (IS_ERR(res)) + p9_client_clunk(fid); + return res; } static int -- cgit From a3b2157e72e321fa313389ac744bbf6d6cb6986d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 28 Feb 2013 01:29:48 -0500 Subject: 9p: make sure ->lookup() adds fid to the right dentry Signed-off-by: Al Viro --- fs/9p/vfs_inode.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 80ff01bcae3d..d86edc8d3fd0 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -824,7 +824,6 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, p9_client_clunk(fid); return ERR_CAST(inode); } - v9fs_fid_add(dentry, fid); /* * If we had a rename on the server and a parallel lookup * for the new name, then make sure we instantiate with @@ -833,7 +832,11 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, * k/b. */ res = d_materialise_unique(dentry, inode); - if (IS_ERR(res)) + if (!res) + v9fs_fid_add(dentry, fid); + else if (!IS_ERR(res)) + v9fs_fid_add(res, fid); + else p9_client_clunk(fid); return res; } -- cgit From 5e608671dfbfd6a9556c31df65a4f147439eed59 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 28 Feb 2013 01:50:20 -0500 Subject: 9p: if v9fs_fid_lookup() gets to asking server, it'd better have hashed dentry ... otherwise the path we'd built isn't worth much. Don't accept such fids obtained from paths unless dentry is still alived by the end of the work. Signed-off-by: Al Viro --- fs/9p/fid.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 73ca55042cb0..616abaf1c6cd 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -41,10 +41,15 @@ * */ +static inline void __add_fid(struct dentry *dentry, struct p9_fid *fid) +{ + hlist_add_head(&fid->dlist, (struct hlist_head *)&dentry->d_fsdata); +} + void v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid) { spin_lock(&dentry->d_lock); - hlist_add_head(&fid->dlist, (struct hlist_head *)&dentry->d_fsdata); + __add_fid(dentry, fid); spin_unlock(&dentry->d_lock); } @@ -198,8 +203,17 @@ static struct p9_fid *v9fs_fid_lookup_with_uid(struct dentry *dentry, } kfree(wnames); fid_out: - if (!IS_ERR(fid)) - v9fs_fid_add(dentry, fid); + if (!IS_ERR(fid)) { + spin_lock(&dentry->d_lock); + if (d_unhashed(dentry)) { + spin_unlock(&dentry->d_lock); + p9_client_clunk(fid); + fid = ERR_PTR(-ENOENT); + } else { + __add_fid(dentry, fid); + spin_unlock(&dentry->d_lock); + } + } err_out: up_read(&v9ses->rename_sem); return fid; -- cgit From c51bb0ea40ca038da26b1fa7d450f4078124af03 Mon Sep 17 00:00:00 2001 From: Ouyang Maochun Date: Mon, 18 Feb 2013 09:54:52 -0600 Subject: cifs: bugfix for unreclaimed writeback pages in cifs_writev_requeue() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Pages get the PG_writeback flag set before cifs sends its request to SMB server in cifs_writepages(), if the SMB service goes down, cifs may try to recommit the writing requests in cifs_writev_requeue(). However, it does not clean its PG_writeback flag and relaimed the pages even if it fails again in cifs_writev_requeue(), which may lead to the hanging of the processes accessing the cifs directory. This patch just cleans the PG_writeback flags and reclaims the pages under that circumstances. Steps to reproduce the bug(trying serveral times may trigger the issue): 1.Write from cifs client continuously.(e.g dd if=/dev/zero of=) 2.Stop SMB service from server.(e.g service smb stop) 3.Wait for two minutes， and then start SMB service from server.(e.g service smb start) 4.The processes which are accessing cifs directory may hang up. Signed-off-by: Ouyang Maochun Signed-off-by: Jiang Yong Tested-by: Zhang Xianwei Reviewed-by: Wang Liang Reviewed-by: Cai Qu Reviewed-by: Jiang Biao Reviewed-by: Jeff Layton Reviewed-by: Pavel Shilovsky Signed-off-by: Steve French --- fs/cifs/cifssmb.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 00e12f2d626b..7353bc5d73d7 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1909,8 +1909,11 @@ cifs_writev_requeue(struct cifs_writedata *wdata) } while (rc == -EAGAIN); for (i = 0; i < wdata->nr_pages; i++) { - if (rc != 0) + if (rc != 0) { SetPageError(wdata->pages[i]); + end_page_writeback(wdata->pages[i]); + page_cache_release(wdata->pages[i]); + } unlock_page(wdata->pages[i]); } -- cgit From bdc20e67e82cfc4901d3a5a0d79104b0e2296d83 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 28 Feb 2013 13:23:38 -0500 Subject: Btrfs: copy everything if we've created an inline extent I noticed while looking into a tree logging bug that we aren't logging inline extents properly. Since this requires copying and it shouldn't happen too often just force us to copy everything for the inode into the tree log when we have an inline extent. With this patch we have valid data after a crash when we write an inline extent. Thanks, Cc: stable@vger.kernel.org Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 9ef7a5b1b77e..ecd9c4cdb0db 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -266,6 +266,7 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans, return 1; } + set_bit(BTRFS_INODE_NEEDS_FULL_SYNC, &BTRFS_I(inode)->runtime_flags); btrfs_delalloc_release_metadata(inode, end + 1 - start); btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0); return 0; -- cgit From 3321719ed67440bba1b0c5ae19c30d640263ccc8 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Wed, 27 Feb 2013 13:28:24 +0000 Subject: Btrfs: fix memory leak of log roots When we abort a transaction while fsyncing, we'll skip freeing log roots part of committing a transaction, which leads to memory leak. This adds a 'free log roots' in putting super when no more users hold references on log roots, so it's safe and clean. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 5 +++++ fs/btrfs/tree-log.c | 6 ++++-- 2 files changed, 9 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index eb7c14308521..5031e6dd5938 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3253,6 +3253,11 @@ void btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root) if (btrfs_root_refs(&root->root_item) == 0) synchronize_srcu(&fs_info->subvol_srcu); + if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { + btrfs_free_log(NULL, root); + btrfs_free_log_root_tree(NULL, fs_info); + } + __btrfs_remove_free_space_cache(root->free_ino_pinned); __btrfs_remove_free_space_cache(root->free_ino_ctl); free_fs_root(root); diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index e8b7a68e1b37..8e85e0e4333d 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -2466,8 +2466,10 @@ static void free_log_tree(struct btrfs_trans_handle *trans, .process_func = process_one_buffer }; - ret = walk_log_tree(trans, log, &wc); - BUG_ON(ret); + if (trans) { + ret = walk_log_tree(trans, log, &wc); + BUG_ON(ret); + } while (1) { ret = find_first_extent_bit(&log->dirty_log_pages, -- cgit From f094ac32aba3a51c00e970a2ea029339af2ca048 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Wed, 27 Feb 2013 13:28:25 +0000 Subject: Btrfs: fix NULL pointer after aborting a transaction While doing cleanup work on an aborted transaction, we've set the global running transaction pointer to NULL _before_ waiting all other transaction handles to finish, so others'd hit NULL pointer crash when referencing the global running transaction pointer. This first sets a hint to avoid new transaction handle joining, then waits other existing handles to abort or finish so that we can safely set the above global pointer to NULL. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 4330433b7b4f..3733c4939a27 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1447,6 +1447,7 @@ static void cleanup_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root, int err) { struct btrfs_transaction *cur_trans = trans->transaction; + DEFINE_WAIT(wait); WARN_ON(trans->use_count > 1); @@ -1455,8 +1456,13 @@ static void cleanup_transaction(struct btrfs_trans_handle *trans, spin_lock(&root->fs_info->trans_lock); list_del_init(&cur_trans->list); if (cur_trans == root->fs_info->running_transaction) { + root->fs_info->trans_no_join = 1; + spin_unlock(&root->fs_info->trans_lock); + wait_event(cur_trans->writer_wait, + atomic_read(&cur_trans->num_writers) == 1); + + spin_lock(&root->fs_info->trans_lock); root->fs_info->running_transaction = NULL; - root->fs_info->trans_no_join = 0; } spin_unlock(&root->fs_info->trans_lock); -- cgit From 2d8946c59743bc635a5e5701e6ef8e71e0a16ab7 Mon Sep 17 00:00:00 2001 From: David Sterba Date: Wed, 27 Feb 2013 23:13:55 +0000 Subject: btrfs: remove a printk from scan_one_device Dave pointed out that he saw messages from btrfs although there was no such filesystem on his computers. The automatic device scan is called on every new blockdevice if the usual distro udev rule set is used. The printk introduced in 6f60cbd3ae442c was a remainder from copying portions of code from btrfs_get_bdev_and_sb which is used under different conditions and the warning makes sense there. Reported-by: Dave Chinner Signed-off-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/volumes.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index db72e0cc6f87..35bb2d4ed29f 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -829,7 +829,6 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, if (IS_ERR(bdev)) { ret = PTR_ERR(bdev); - printk(KERN_INFO "btrfs: open %s failed\n", path); goto error; } -- cgit From e9662f701c85ebc99f532bf8bb53208c0648846a Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 28 Feb 2013 10:01:15 +0000 Subject: Btrfs: remove unnecessary dget_parent/dput when creating the pending snapshot Since we have grabbed the parent inode at the beginning of the snapshot creation, and both sync and async snapshot creation release it after the pending snapshots are actually created, it is safe to access the parent inode directly during the snapshot creation, we needn't use dget_parent/dput to fix the parent dentry and get the dir inode. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 10 ++++++---- fs/btrfs/transaction.c | 5 +---- fs/btrfs/transaction.h | 1 + 3 files changed, 8 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 059546aa8fdf..8dcd4ff0c3a5 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -530,9 +530,10 @@ fail: return ret; } -static int create_snapshot(struct btrfs_root *root, struct dentry *dentry, - char *name, int namelen, u64 *async_transid, - bool readonly, struct btrfs_qgroup_inherit *inherit) +static int create_snapshot(struct btrfs_root *root, struct inode *dir, + struct dentry *dentry, char *name, int namelen, + u64 *async_transid, bool readonly, + struct btrfs_qgroup_inherit *inherit) { struct inode *inode; struct btrfs_pending_snapshot *pending_snapshot; @@ -551,6 +552,7 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry, pending_snapshot->dentry = dentry; pending_snapshot->root = root; pending_snapshot->readonly = readonly; + pending_snapshot->dir = dir; pending_snapshot->inherit = inherit; trans = btrfs_start_transaction(root->fs_info->extent_root, 6); @@ -728,7 +730,7 @@ static noinline int btrfs_mksubvol(struct path *parent, goto out_up_read; if (snap_src) { - error = create_snapshot(snap_src, dentry, name, namelen, + error = create_snapshot(snap_src, dir, dentry, name, namelen, async_transid, readonly, inherit); } else { error = create_subvol(BTRFS_I(dir)->root, dentry, diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 3733c4939a27..71de435a291e 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1068,7 +1068,6 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, struct inode *parent_inode; struct btrfs_path *path; struct btrfs_dir_item *dir_item; - struct dentry *parent; struct dentry *dentry; struct extent_buffer *tmp; struct extent_buffer *old; @@ -1126,8 +1125,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, trans->bytes_reserved = trans->block_rsv->reserved; dentry = pending->dentry; - parent = dget_parent(dentry); - parent_inode = parent->d_inode; + parent_inode = pending->dir; parent_root = BTRFS_I(parent_inode)->root; record_root_in_trans(trans, parent_root); @@ -1275,7 +1273,6 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, if (ret) btrfs_abort_transaction(trans, root, ret); fail: - dput(parent); trans->block_rsv = rsv; trans->bytes_reserved = 0; no_free_objectid: diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 5afd7b1dceac..5f67fba07ab4 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -84,6 +84,7 @@ struct btrfs_trans_handle { struct btrfs_pending_snapshot { struct dentry *dentry; + struct inode *dir; struct btrfs_root *root; struct btrfs_root *snap; struct btrfs_qgroup_inherit *inherit; -- cgit From d5c1207017cd8387b4d3224dd7ab6cf5cd7f1c9a Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 28 Feb 2013 10:04:33 +0000 Subject: Btrfs: fix wrong reserved space in qgroup during snap/subv creation There are two problems in the space reservation of the snapshot/ subvolume creation. - don't reserve the space for the root item insertion - the space which is reserved in the qgroup is different with the free space reservation. we need reserve free space for 7 items, but in qgroup reservation, we need reserve space only for 3 items. So we implement new metadata reservation functions for the snapshot/subvolume creation. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 9 +++++-- fs/btrfs/extent-tree.c | 65 ++++++++++++++++++++++++++++++++++++++++---------- fs/btrfs/ioctl.c | 62 ++++++++++++++++++++++++++++++++--------------- fs/btrfs/transaction.c | 4 +--- fs/btrfs/transaction.h | 1 + 5 files changed, 105 insertions(+), 36 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index ae8dcc406805..0d82922179db 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3106,8 +3106,13 @@ void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans, int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans, struct inode *inode); void btrfs_orphan_release_metadata(struct inode *inode); -int btrfs_snap_reserve_metadata(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending); +int btrfs_subvolume_reserve_metadata(struct btrfs_root *root, + struct btrfs_block_rsv *rsv, + int nitems, + u64 *qgroup_reserved); +void btrfs_subvolume_release_metadata(struct btrfs_root *root, + struct btrfs_block_rsv *rsv, + u64 qgroup_reserved); int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes); void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes); int btrfs_delalloc_reserve_space(struct inode *inode, u64 num_bytes); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 5681a91ed400..7cb9d734f6e5 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4560,19 +4560,60 @@ void btrfs_orphan_release_metadata(struct inode *inode) btrfs_block_rsv_release(root, root->orphan_block_rsv, num_bytes); } -int btrfs_snap_reserve_metadata(struct btrfs_trans_handle *trans, - struct btrfs_pending_snapshot *pending) +/* + * btrfs_subvolume_reserve_metadata() - reserve space for subvolume operation + * root: the root of the parent directory + * rsv: block reservation + * items: the number of items that we need do reservation + * qgroup_reserved: used to return the reserved size in qgroup + * + * This function is used to reserve the space for snapshot/subvolume + * creation and deletion. Those operations are different with the + * common file/directory operations, they change two fs/file trees + * and root tree, the number of items that the qgroup reserves is + * different with the free space reservation. So we can not use + * the space reseravtion mechanism in start_transaction(). + */ +int btrfs_subvolume_reserve_metadata(struct btrfs_root *root, + struct btrfs_block_rsv *rsv, + int items, + u64 *qgroup_reserved) { - struct btrfs_root *root = pending->root; - struct btrfs_block_rsv *src_rsv = get_block_rsv(trans, root); - struct btrfs_block_rsv *dst_rsv = &pending->block_rsv; - /* - * two for root back/forward refs, two for directory entries, - * one for root of the snapshot and one for parent inode. - */ - u64 num_bytes = btrfs_calc_trans_metadata_size(root, 6); - dst_rsv->space_info = src_rsv->space_info; - return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes); + u64 num_bytes; + int ret; + + if (root->fs_info->quota_enabled) { + /* One for parent inode, two for dir entries */ + num_bytes = 3 * root->leafsize; + ret = btrfs_qgroup_reserve(root, num_bytes); + if (ret) + return ret; + } else { + num_bytes = 0; + } + + *qgroup_reserved = num_bytes; + + num_bytes = btrfs_calc_trans_metadata_size(root, items); + rsv->space_info = __find_space_info(root->fs_info, + BTRFS_BLOCK_GROUP_METADATA); + ret = btrfs_block_rsv_add(root, rsv, num_bytes, + BTRFS_RESERVE_FLUSH_ALL); + if (ret) { + if (*qgroup_reserved) + btrfs_qgroup_free(root, *qgroup_reserved); + } + + return ret; +} + +void btrfs_subvolume_release_metadata(struct btrfs_root *root, + struct btrfs_block_rsv *rsv, + u64 qgroup_reserved) +{ + btrfs_block_rsv_release(root, rsv, (u64)-1); + if (qgroup_reserved) + btrfs_qgroup_free(root, qgroup_reserved); } /** diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 8dcd4ff0c3a5..56d92549389c 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -363,7 +363,7 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg) return 0; } -static noinline int create_subvol(struct btrfs_root *root, +static noinline int create_subvol(struct inode *dir, struct dentry *dentry, char *name, int namelen, u64 *async_transid, @@ -374,32 +374,39 @@ static noinline int create_subvol(struct btrfs_root *root, struct btrfs_root_item root_item; struct btrfs_inode_item *inode_item; struct extent_buffer *leaf; + struct btrfs_root *root = BTRFS_I(dir)->root; struct btrfs_root *new_root; - struct dentry *parent = dentry->d_parent; - struct inode *dir; + struct btrfs_block_rsv block_rsv; struct timespec cur_time = CURRENT_TIME; int ret; int err; u64 objectid; u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID; u64 index = 0; + u64 qgroup_reserved; uuid_le new_uuid; ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid); if (ret) return ret; - dir = parent->d_inode; - + btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP); /* - * 1 - inode item - * 2 - refs - * 1 - root item - * 2 - dir items + * The same as the snapshot creation, please see the comment + * of create_snapshot(). */ - trans = btrfs_start_transaction(root, 6); - if (IS_ERR(trans)) - return PTR_ERR(trans); + ret = btrfs_subvolume_reserve_metadata(root, &block_rsv, + 7, &qgroup_reserved); + if (ret) + return ret; + + trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto out; + } + trans->block_rsv = &block_rsv; + trans->bytes_reserved = block_rsv.size; ret = btrfs_qgroup_inherit(trans, root->fs_info, 0, objectid, inherit); if (ret) @@ -515,6 +522,8 @@ static noinline int create_subvol(struct btrfs_root *root, BUG_ON(ret); fail: + trans->block_rsv = NULL; + trans->bytes_reserved = 0; if (async_transid) { *async_transid = trans->transid; err = btrfs_commit_transaction_async(trans, root, 1); @@ -526,7 +535,8 @@ fail: if (!ret) d_instantiate(dentry, btrfs_lookup_dentry(dir, dentry)); - +out: + btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved); return ret; } @@ -549,21 +559,31 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir, btrfs_init_block_rsv(&pending_snapshot->block_rsv, BTRFS_BLOCK_RSV_TEMP); + /* + * 1 - parent dir inode + * 2 - dir entries + * 1 - root item + * 2 - root ref/backref + * 1 - root of snapshot + */ + ret = btrfs_subvolume_reserve_metadata(BTRFS_I(dir)->root, + &pending_snapshot->block_rsv, 7, + &pending_snapshot->qgroup_reserved); + if (ret) + goto out; + pending_snapshot->dentry = dentry; pending_snapshot->root = root; pending_snapshot->readonly = readonly; pending_snapshot->dir = dir; pending_snapshot->inherit = inherit; - trans = btrfs_start_transaction(root->fs_info->extent_root, 6); + trans = btrfs_start_transaction(root, 0); if (IS_ERR(trans)) { ret = PTR_ERR(trans); goto fail; } - ret = btrfs_snap_reserve_metadata(trans, pending_snapshot); - BUG_ON(ret); - spin_lock(&root->fs_info->trans_lock); list_add(&pending_snapshot->list, &trans->transaction->pending_snapshots); @@ -600,6 +620,10 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir, d_instantiate(dentry, inode); ret = 0; fail: + btrfs_subvolume_release_metadata(BTRFS_I(dir)->root, + &pending_snapshot->block_rsv, + pending_snapshot->qgroup_reserved); +out: kfree(pending_snapshot); return ret; } @@ -733,8 +757,8 @@ static noinline int btrfs_mksubvol(struct path *parent, error = create_snapshot(snap_src, dir, dentry, name, namelen, async_transid, readonly, inherit); } else { - error = create_subvol(BTRFS_I(dir)->root, dentry, - name, namelen, async_transid, inherit); + error = create_subvol(dir, dentry, name, namelen, + async_transid, inherit); } if (!error) fsnotify_mkdir(dir, dentry); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 71de435a291e..f11c2e0a3746 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1082,7 +1082,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, path = btrfs_alloc_path(); if (!path) { ret = pending->error = -ENOMEM; - goto path_alloc_fail; + return ret; } new_root_item = kmalloc(sizeof(*new_root_item), GFP_NOFS); @@ -1279,8 +1279,6 @@ no_free_objectid: kfree(new_root_item); root_item_alloc_fail: btrfs_free_path(path); -path_alloc_fail: - btrfs_block_rsv_release(root, &pending->block_rsv, (u64)-1); return ret; } diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 5f67fba07ab4..3c8e0d25c8e4 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -90,6 +90,7 @@ struct btrfs_pending_snapshot { struct btrfs_qgroup_inherit *inherit; /* block reservation for the operation */ struct btrfs_block_rsv block_rsv; + u64 qgroup_reserved; /* extra metadata reseration for relocation */ int error; bool readonly; -- cgit From c58aaad2ac46f8dbce9ab7f7a7a3726e332a56ee Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 28 Feb 2013 10:05:36 +0000 Subject: Btrfs: fix wrong reserved space when deleting a snapshot/subvolume When deleting a snapshot/subvolume, we need remove root ref/backref, dir entries and update the dir inode, so we must reserve free space for those operations. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 21 +++++++++++++++++++-- 1 file changed, 19 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 56d92549389c..b908960c9746 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2064,6 +2064,8 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file, struct btrfs_root *dest = NULL; struct btrfs_ioctl_vol_args *vol_args; struct btrfs_trans_handle *trans; + struct btrfs_block_rsv block_rsv; + u64 qgroup_reserved; int namelen; int ret; int err = 0; @@ -2153,12 +2155,23 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file, if (err) goto out_up_write; + btrfs_init_block_rsv(&block_rsv, BTRFS_BLOCK_RSV_TEMP); + /* + * One for dir inode, two for dir entries, two for root + * ref/backref. + */ + err = btrfs_subvolume_reserve_metadata(root, &block_rsv, + 5, &qgroup_reserved); + if (err) + goto out_up_write; + trans = btrfs_start_transaction(root, 0); if (IS_ERR(trans)) { err = PTR_ERR(trans); - goto out_up_write; + goto out_release; } - trans->block_rsv = &root->fs_info->global_block_rsv; + trans->block_rsv = &block_rsv; + trans->bytes_reserved = block_rsv.size; ret = btrfs_unlink_subvol(trans, root, dir, dest->root_key.objectid, @@ -2188,10 +2201,14 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file, } } out_end_trans: + trans->block_rsv = NULL; + trans->bytes_reserved = 0; ret = btrfs_end_transaction(trans, root); if (ret && !err) err = ret; inode->i_flags |= S_DEAD; +out_release: + btrfs_subvolume_release_metadata(root, &block_rsv, qgroup_reserved); out_up_write: up_write(&root->fs_info->subvol_sem); out_unlock: -- cgit From b8dae3138876080d4dd98cc438ff759338d632ef Mon Sep 17 00:00:00 2001 From: David Sterba Date: Thu, 28 Feb 2013 14:54:18 +0000 Subject: btrfs: use only inline_pages from extent buffer The nodesize is capped at 64k and there are enough pages preallocated in extent_buffer::inline_pages. The fallback to kmalloc never happened because even on the smallest page size considered (4k) inline_pages covered the needs. Signed-off-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/extent_io.c | 21 ++++++--------------- fs/btrfs/extent_io.h | 3 +-- 2 files changed, 7 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 597ab8966c80..0306665cb1d7 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3981,8 +3981,6 @@ static void __free_extent_buffer(struct extent_buffer *eb) list_del(&eb->leak_list); spin_unlock_irqrestore(&leak_lock, flags); #endif - if (eb->pages && eb->pages != eb->inline_pages) - kfree(eb->pages); kmem_cache_free(extent_buffer_cache, eb); } @@ -4023,19 +4021,12 @@ static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree, atomic_set(&eb->refs, 1); atomic_set(&eb->io_pages, 0); - if (len > MAX_INLINE_EXTENT_BUFFER_SIZE) { - struct page **pages; - int num_pages = (len + PAGE_CACHE_SIZE - 1) >> - PAGE_CACHE_SHIFT; - pages = kzalloc(num_pages, mask); - if (!pages) { - __free_extent_buffer(eb); - return NULL; - } - eb->pages = pages; - } else { - eb->pages = eb->inline_pages; - } + /* + * Sanity checks, currently the maximum is 64k covered by 16x 4k pages + */ + BUILD_BUG_ON(BTRFS_MAX_METADATA_BLOCKSIZE + > MAX_INLINE_EXTENT_BUFFER_SIZE); + BUG_ON(len > MAX_INLINE_EXTENT_BUFFER_SIZE); return eb; } diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index dc81868d975a..6068a1985560 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -158,8 +158,7 @@ struct extent_buffer { */ wait_queue_head_t read_lock_wq; wait_queue_head_t lock_wq; - struct page *inline_pages[INLINE_EXTENT_BUFFER_PAGES]; - struct page **pages; + struct page *pages[INLINE_EXTENT_BUFFER_PAGES]; }; static inline void extent_set_compress_type(unsigned long *bio_flags, -- cgit From 7aa262b522ee5a7f1bf9b8fe34539b262ba6da0a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 28 Feb 2013 16:19:59 -0800 Subject: NFSv4: Fix another open/open_recovery deadlock If we don't release the open seqid before we wait for state recovery, then we may end up deadlocking the state recovery thread. This patch addresses a new deadlock that was introduced by commit c21443c2c792cd9b463646d982b0fe48aa6feb0f (NFSv4: Fix a reboot recovery race when opening a file) Reported-by: Andy Adamson Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 655237fc46df..37eb38566fbc 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1158,6 +1158,7 @@ _nfs4_opendata_to_nfs4_state(struct nfs4_opendata *data) data->o_arg.fmode); iput(inode); out: + nfs_release_seqid(data->o_arg.seqid); return state; err_put_inode: iput(inode); -- cgit From eb97cbb459c656b542101091f67ce112b3e56183 Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Thu, 28 Feb 2013 20:30:08 -0500 Subject: PNFS: set the default DS timeout to 60 seconds The client should have 60 second default timeouts for DS operations, not 6 seconds. NFS4_DEF_DS_TIMEO is used as "timeout in tenths of a second" in nfs_init_timeout_values (and is not used anywhere else). This matches up with the description of the module param dataserver_timeo. Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust --- fs/nfs/nfs4filelayout.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4filelayout.h b/fs/nfs/nfs4filelayout.h index 8c07241fe52b..b8da95548d3d 100644 --- a/fs/nfs/nfs4filelayout.h +++ b/fs/nfs/nfs4filelayout.h @@ -36,7 +36,7 @@ * Default data server connection timeout and retrans vaules. * Set by module paramters dataserver_timeo and dataserver_retrans. */ -#define NFS4_DEF_DS_TIMEO 60 +#define NFS4_DEF_DS_TIMEO 600 /* in tenths of a second */ #define NFS4_DEF_DS_RETRANS 5 /* -- cgit From 3000512137602b84d1ad5fd89d62984993a19bb6 Mon Sep 17 00:00:00 2001 From: Weston Andros Adamson Date: Thu, 28 Feb 2013 20:30:10 -0500 Subject: NFSv4.1: LAYOUTGET EDELAY loops timeout to the MDS The client will currently try LAYOUTGETs forever if a server is returning NFS4ERR_LAYOUTTRYLATER or NFS4ERR_RECALLCONFLICT - even if the client no longer needs the layout (ie process killed, unmounted). This patch uses the DS timeout value (module parameter 'dataserver_timeo' via rpc layer) to set an upper limit of how long the client tries LATOUTGETs in this situation. Once the timeout is reached, IO is redirected to the MDS. This also changes how the client checks if a layout is on the clp list to avoid a double list_add. Signed-off-by: Weston Andros Adamson Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 9 ++++++++- fs/nfs/pnfs.c | 7 +++---- 2 files changed, 11 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 37eb38566fbc..b2671cb0f901 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -93,6 +93,8 @@ static int nfs4_map_errors(int err) return err; switch (err) { case -NFS4ERR_RESOURCE: + case -NFS4ERR_LAYOUTTRYLATER: + case -NFS4ERR_RECALLCONFLICT: return -EREMOTEIO; case -NFS4ERR_WRONGSEC: return -EPERM; @@ -6046,6 +6048,7 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata) struct nfs_server *server = NFS_SERVER(inode); struct pnfs_layout_hdr *lo; struct nfs4_state *state = NULL; + unsigned long timeo, giveup; dprintk("--> %s\n", __func__); @@ -6057,7 +6060,10 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata) goto out; case -NFS4ERR_LAYOUTTRYLATER: case -NFS4ERR_RECALLCONFLICT: - task->tk_status = -NFS4ERR_DELAY; + timeo = rpc_get_timeout(task->tk_client); + giveup = lgp->args.timestamp + timeo; + if (time_after(giveup, jiffies)) + task->tk_status = -NFS4ERR_DELAY; break; case -NFS4ERR_EXPIRED: case -NFS4ERR_BAD_STATEID: @@ -6178,6 +6184,7 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags) return ERR_PTR(-ENOMEM); } lgp->args.layout.pglen = max_pages * PAGE_SIZE; + lgp->args.timestamp = jiffies; lgp->res.layoutp = &lgp->args.layout; lgp->res.seq_res.sr_slot = NULL; diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 97767c8683f9..48ac5aad6258 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -1181,7 +1181,7 @@ pnfs_update_layout(struct inode *ino, struct nfs_client *clp = server->nfs_client; struct pnfs_layout_hdr *lo; struct pnfs_layout_segment *lseg = NULL; - bool first = false; + bool first; if (!pnfs_enabled_sb(NFS_SERVER(ino))) goto out; @@ -1215,10 +1215,9 @@ pnfs_update_layout(struct inode *ino, goto out_unlock; atomic_inc(&lo->plh_outstanding); - if (list_empty(&lo->plh_segs)) - first = true; - + first = list_empty(&lo->plh_layouts) ? true : false; spin_unlock(&ino->i_lock); + if (first) { /* The lo must be on the clp list if there is any * chance of a CB_LAYOUTRECALL(FILE) coming in. -- cgit From 246307745c406379996e6ed6411f0e20f1ce1449 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Thu, 28 Feb 2013 23:58:56 -0500 Subject: ext4: optimize ext4_es_shrink() When the system is under memory pressure, ext4_es_srhink() will get called very often. So optimize returning the number of items in the file system's extent status cache by keeping a per-filesystem count, instead of calculating it each time by scanning all of the inodes in the extent status cache. Also rename the slab used for the extent status cache to be "ext4_extent_status" so it's obviousl the slab in question is created by ext4. Signed-off-by: "Theodore Ts'o" Cc: Zheng Liu --- fs/ext4/ext4.h | 1 + fs/ext4/extents_status.c | 39 +++++++++++++-------------------------- 2 files changed, 14 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 6e16c1867959..96c10934bb96 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1268,6 +1268,7 @@ struct ext4_sb_info { atomic_t s_mb_preallocated; atomic_t s_mb_discarded; atomic_t s_lock_busy; + atomic_t s_extent_cache_cnt; /* locality groups */ struct ext4_locality_group __percpu *s_locality_groups; diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index f768f4a98a2b..27fcdd2b2607 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -147,11 +147,12 @@ static int __es_remove_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t end); static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei, int nr_to_scan); -static int ext4_es_reclaim_extents_count(struct super_block *sb); int __init ext4_init_es(void) { - ext4_es_cachep = KMEM_CACHE(extent_status, SLAB_RECLAIM_ACCOUNT); + ext4_es_cachep = kmem_cache_create("ext4_extent_status", + sizeof(struct extent_status), + 0, (SLAB_RECLAIM_ACCOUNT), NULL); if (ext4_es_cachep == NULL) return -ENOMEM; return 0; @@ -302,8 +303,10 @@ ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len, /* * We don't count delayed extent because we never try to reclaim them */ - if (!ext4_es_is_delayed(es)) + if (!ext4_es_is_delayed(es)) { EXT4_I(inode)->i_es_lru_nr++; + atomic_inc(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt); + } return es; } @@ -314,6 +317,7 @@ static void ext4_es_free_extent(struct inode *inode, struct extent_status *es) if (!ext4_es_is_delayed(es)) { BUG_ON(EXT4_I(inode)->i_es_lru_nr == 0); EXT4_I(inode)->i_es_lru_nr--; + atomic_dec(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt); } kmem_cache_free(ext4_es_cachep, es); @@ -674,10 +678,11 @@ static int ext4_es_shrink(struct shrinker *shrink, struct shrink_control *sc) int nr_to_scan = sc->nr_to_scan; int ret, nr_shrunk = 0; - trace_ext4_es_shrink_enter(sbi->s_sb, nr_to_scan); + ret = atomic_read(&sbi->s_extent_cache_cnt); + trace_ext4_es_shrink_enter(sbi->s_sb, nr_to_scan, ret); if (!nr_to_scan) - return ext4_es_reclaim_extents_count(sbi->s_sb); + return ret; INIT_LIST_HEAD(&scanned); @@ -705,9 +710,10 @@ static int ext4_es_shrink(struct shrinker *shrink, struct shrink_control *sc) } list_splice_tail(&scanned, &sbi->s_es_lru); spin_unlock(&sbi->s_es_lru_lock); - trace_ext4_es_shrink_exit(sbi->s_sb, nr_shrunk); - return ext4_es_reclaim_extents_count(sbi->s_sb); + ret = atomic_read(&sbi->s_extent_cache_cnt); + trace_ext4_es_shrink_exit(sbi->s_sb, nr_shrunk, ret); + return ret; } void ext4_es_register_shrinker(struct super_block *sb) @@ -751,25 +757,6 @@ void ext4_es_lru_del(struct inode *inode) spin_unlock(&sbi->s_es_lru_lock); } -static int ext4_es_reclaim_extents_count(struct super_block *sb) -{ - struct ext4_sb_info *sbi = EXT4_SB(sb); - struct ext4_inode_info *ei; - struct list_head *cur; - int nr_cached = 0; - - spin_lock(&sbi->s_es_lru_lock); - list_for_each(cur, &sbi->s_es_lru) { - ei = list_entry(cur, struct ext4_inode_info, i_es_lru); - read_lock(&ei->i_es_lock); - nr_cached += ei->i_es_lru_nr; - read_unlock(&ei->i_es_lock); - } - spin_unlock(&sbi->s_es_lru_lock); - trace_ext4_es_reclaim_extents_count(sb, nr_cached); - return nr_cached; -} - static int __es_try_to_reclaim_extents(struct ext4_inode_info *ei, int nr_to_scan) { -- cgit From 06b3a860dcf596bdc2bb1cca3252d3907b581938 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Wed, 27 Feb 2013 11:16:57 +0000 Subject: Btrfs: fix missing deleted items in btrfs_clean_quota_tree Steps to reproduce: i=0 ncases=100 mkfs.btrfs mount btrfs quota enable btrfs qgroup create 2/1 while [ $i -le $ncases ] do btrfs qgroup create 1/$i btrfs qgroup assign 1/$i 2/1 i=$(($i+1)) done btrfs quota disable umount btrfsck You can also use the commands: btrfs-debug-tree | grep QGROUP You will find there are still items existed.The reasons why this happens is because the original code just checks slots[0]==0 and returns. We try to fix it by deleting the leaf one by one. Signed-off-by: Wang Shilong Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/qgroup.c | 34 +++++++++++++++++++++------------- 1 file changed, 21 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 88ab785bbd73..f011d9b8ffb1 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -732,7 +732,9 @@ static int btrfs_clean_quota_tree(struct btrfs_trans_handle *trans, { struct btrfs_path *path; struct btrfs_key key; + struct extent_buffer *leaf = NULL; int ret; + int nr = 0; if (!root) return -EINVAL; @@ -741,24 +743,30 @@ static int btrfs_clean_quota_tree(struct btrfs_trans_handle *trans, if (!path) return -ENOMEM; - while (1) { - key.objectid = 0; - key.offset = 0; - key.type = 0; + path->leave_spinning = 1; - path->leave_spinning = 1; + key.objectid = 0; + key.offset = 0; + key.type = 0; + + while (1) { ret = btrfs_search_slot(trans, root, &key, path, -1, 1); - if (ret > 0) { - if (path->slots[0] == 0) - break; - path->slots[0]--; - } else if (ret < 0) { + if (ret < 0) + goto out; + leaf = path->nodes[0]; + nr = btrfs_header_nritems(leaf); + if (!nr) break; - } - - ret = btrfs_del_item(trans, root, path); + /* + * delete the leaf one by one + * since the whole tree is going + * to be deleted. + */ + path->slots[0] = 0; + ret = btrfs_del_items(trans, root, path, 0, nr); if (ret) goto out; + btrfs_release_path(path); } ret = 0; -- cgit From 84cbe2f725f59b72fffb48ab7161e771eeee5103 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Wed, 27 Feb 2013 11:20:56 +0000 Subject: Btrfs: return ENOMEM rather than use BUG_ON when btrfs_alloc_path fails Return ENOMEM rather trigger BUG_ON, fix it. Signed-off-by: Wang Shilong Reviewed-by: Miao Xie Reviewed-by: Zach Brown Signed-off-by: Josef Bacik --- fs/btrfs/qgroup.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index f011d9b8ffb1..69fa487e29ce 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -620,7 +620,9 @@ static int update_qgroup_limit_item(struct btrfs_trans_handle *trans, key.offset = qgroupid; path = btrfs_alloc_path(); - BUG_ON(!path); + if (!path) + return -ENOMEM; + ret = btrfs_search_slot(trans, root, &key, path, 0, 1); if (ret > 0) ret = -ENOENT; @@ -661,7 +663,9 @@ static int update_qgroup_info_item(struct btrfs_trans_handle *trans, key.offset = qgroup->qgroupid; path = btrfs_alloc_path(); - BUG_ON(!path); + if (!path) + return -ENOMEM; + ret = btrfs_search_slot(trans, root, &key, path, 0, 1); if (ret > 0) ret = -ENOENT; @@ -702,7 +706,9 @@ static int update_qgroup_status_item(struct btrfs_trans_handle *trans, key.offset = 0; path = btrfs_alloc_path(); - BUG_ON(!path); + if (!path) + return -ENOMEM; + ret = btrfs_search_slot(trans, root, &key, path, 0, 1); if (ret > 0) ret = -ENOENT; -- cgit From 235fdb8ef2a313749cf3a42e70156c618263a388 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Wed, 27 Feb 2013 11:21:40 +0000 Subject: Btrfs: remove reduplicate check about root in the function btrfs_clean_quota_tree The check work has been done just before the function btrfs_clean_quota_tree is called, it is not necessary to check it again, remove it. Signed-off-by: Wang Shilong Signed-off-by: Josef Bacik --- fs/btrfs/qgroup.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 69fa487e29ce..aee4b1cc3d98 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -742,9 +742,6 @@ static int btrfs_clean_quota_tree(struct btrfs_trans_handle *trans, int ret; int nr = 0; - if (!root) - return -EINVAL; - path = btrfs_alloc_path(); if (!path) return -ENOMEM; -- cgit From a9870c0e031527fbfa382019f30d2e9b98124a0d Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Fri, 1 Mar 2013 11:33:01 +0000 Subject: Btrfs: don't call btrfs_qgroup_free if just btrfs_qgroup_reserve fails commit eb6b88d92c6df083dd09a8c471011e3788dfd7c6 leads into another bug. If it is just because qgroup_reserve fails, the function btrfs_qgroup_free should not be called, otherwise, it will cause the wrong quota accounting. Signed-off-by: Wang Shilong Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 7cb9d734f6e5..075854e7a044 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4772,9 +4772,14 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) * ret != 0 here means the qgroup reservation failed, we go straight to * the shared error handling then. */ - if (ret == 0) + if (ret == 0) { ret = reserve_metadata_bytes(root, block_rsv, to_reserve, flush); + if (ret && root->fs_info->quota_enabled) { + btrfs_qgroup_free(root, num_bytes + + nr_extents * root->leafsize); + } + } if (ret) { u64 to_free = 0; @@ -4805,10 +4810,6 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) btrfs_ino(inode), to_free, 0); } - if (root->fs_info->quota_enabled) { - btrfs_qgroup_free(root, num_bytes + - nr_extents * root->leafsize); - } if (delalloc_lock) mutex_unlock(&BTRFS_I(inode)->delalloc_mutex); return ret; -- cgit From 88e081bf82ffcb9be3ad33d04c829051d66da564 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Fri, 1 Mar 2013 11:36:01 +0000 Subject: Btrfs: cleanup to make the function btrfs_delalloc_reserve_metadata more logic The original code is a little confusing and not clear, The right way to deal with the kernel code like this: [...] if (ret) goto out; [...] So i move the common clean_up code to the place labeled with out_fail, this will be easier to maintain. Signed-off-by: Wang Shilong Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 82 +++++++++++++++++++++++--------------------------- 1 file changed, 38 insertions(+), 44 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 075854e7a044..aaee2b7fee78 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4721,6 +4721,8 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) enum btrfs_reserve_flush_enum flush = BTRFS_RESERVE_FLUSH_ALL; int ret = 0; bool delalloc_lock = true; + u64 to_free = 0; + unsigned dropped; /* If we are a free space inode we need to not flush since we will be in * the middle of a transaction commit. We also don't need the delalloc @@ -4764,55 +4766,19 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) csum_bytes = BTRFS_I(inode)->csum_bytes; spin_unlock(&BTRFS_I(inode)->lock); - if (root->fs_info->quota_enabled) + if (root->fs_info->quota_enabled) { ret = btrfs_qgroup_reserve(root, num_bytes + nr_extents * root->leafsize); + if (ret) + goto out_fail; + } - /* - * ret != 0 here means the qgroup reservation failed, we go straight to - * the shared error handling then. - */ - if (ret == 0) { - ret = reserve_metadata_bytes(root, block_rsv, - to_reserve, flush); - if (ret && root->fs_info->quota_enabled) { + ret = reserve_metadata_bytes(root, block_rsv, to_reserve, flush); + if (unlikely(ret)) { + if (root->fs_info->quota_enabled) btrfs_qgroup_free(root, num_bytes + nr_extents * root->leafsize); - } - } - - if (ret) { - u64 to_free = 0; - unsigned dropped; - - spin_lock(&BTRFS_I(inode)->lock); - dropped = drop_outstanding_extent(inode); - /* - * If the inodes csum_bytes is the same as the original - * csum_bytes then we know we haven't raced with any free()ers - * so we can just reduce our inodes csum bytes and carry on. - * Otherwise we have to do the normal free thing to account for - * the case that the free side didn't free up its reserve - * because of this outstanding reservation. - */ - if (BTRFS_I(inode)->csum_bytes == csum_bytes) - calc_csum_metadata_size(inode, num_bytes, 0); - else - to_free = calc_csum_metadata_size(inode, num_bytes, 0); - spin_unlock(&BTRFS_I(inode)->lock); - if (dropped) - to_free += btrfs_calc_trans_metadata_size(root, dropped); - - if (to_free) { - btrfs_block_rsv_release(root, block_rsv, to_free); - trace_btrfs_space_reservation(root->fs_info, - "delalloc", - btrfs_ino(inode), - to_free, 0); - } - if (delalloc_lock) - mutex_unlock(&BTRFS_I(inode)->delalloc_mutex); - return ret; + goto out_fail; } spin_lock(&BTRFS_I(inode)->lock); @@ -4833,6 +4799,34 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) block_rsv_add_bytes(block_rsv, to_reserve, 1); return 0; + +out_fail: + spin_lock(&BTRFS_I(inode)->lock); + dropped = drop_outstanding_extent(inode); + /* + * If the inodes csum_bytes is the same as the original + * csum_bytes then we know we haven't raced with any free()ers + * so we can just reduce our inodes csum bytes and carry on. + * Otherwise we have to do the normal free thing to account for + * the case that the free side didn't free up its reserve + * because of this outstanding reservation. + */ + if (BTRFS_I(inode)->csum_bytes == csum_bytes) + calc_csum_metadata_size(inode, num_bytes, 0); + else + to_free = calc_csum_metadata_size(inode, num_bytes, 0); + spin_unlock(&BTRFS_I(inode)->lock); + if (dropped) + to_free += btrfs_calc_trans_metadata_size(root, dropped); + + if (to_free) { + btrfs_block_rsv_release(root, block_rsv, to_free); + trace_btrfs_space_reservation(root->fs_info, "delalloc", + btrfs_ino(inode), to_free, 0); + } + if (delalloc_lock) + mutex_unlock(&BTRFS_I(inode)->delalloc_mutex); + return ret; } /** -- cgit From 83c8266acc1d19debbf353a16aabbd892ef99462 Mon Sep 17 00:00:00 2001 From: David Sterba Date: Fri, 1 Mar 2013 15:03:00 +0000 Subject: btrfs: try harder to allocate raid56 stripe cache The stripe hash table is large, starting with allocation order 4 and can go as high as order 7 in case lock debugging is turned on and structure padding happens. Observed mount failure: mount: page allocation failure: order:7, mode:0x200050 Pid: 8234, comm: mount Tainted: G W 3.8.0-default+ #267 Call Trace: [] warn_alloc_failed+0xf3/0x140 [] ? __alloc_pages_direct_compact+0x92/0x250 [] __alloc_pages_nodemask+0x733/0x9d0 [] ? cache_alloc_refill+0x3f8/0x840 [] cache_alloc_refill+0x43c/0x840 [] ? is_kernel_percpu_address+0x4b/0x90 [] ? btrfs_alloc_stripe_hash_table+0x5c/0x130 [btrfs] [] kmem_cache_alloc_trace+0x247/0x270 [] btrfs_alloc_stripe_hash_table+0x5c/0x130 [btrfs] [] open_ctree+0xb2f/0x1f90 [btrfs] [] ? string+0x49/0xe0 [] ? vsnprintf+0x443/0x5d0 [] btrfs_mount+0x526/0x600 [btrfs] [] ? cache_alloc_debugcheck_after+0x4c/0x200 [] mount_fs+0x20/0xe0 [] vfs_kern_mount+0x76/0x120 [] do_mount+0x386/0x980 [] ? strndup_user+0x5b/0x80 [] sys_mount+0x90/0xe0 [] system_call_fastpath+0x16/0x1b Signed-off-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 2 +- fs/btrfs/raid56.c | 31 +++++++++++++++++++++++++------ 2 files changed, 26 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 5031e6dd5938..02369a3c162e 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2197,7 +2197,7 @@ int open_ctree(struct super_block *sb, ret = btrfs_alloc_stripe_hash_table(fs_info); if (ret) { - err = -ENOMEM; + err = ret; goto fail_alloc; } diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c index e34e568534d9..07222053c7d8 100644 --- a/fs/btrfs/raid56.c +++ b/fs/btrfs/raid56.c @@ -188,13 +188,25 @@ int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info) struct btrfs_stripe_hash *h; int num_entries = 1 << BTRFS_STRIPE_HASH_TABLE_BITS; int i; + int table_size; if (info->stripe_hash_table) return 0; - table = kzalloc(sizeof(*table) + sizeof(*h) * num_entries, GFP_NOFS); - if (!table) - return -ENOMEM; + /* + * The table is large, starting with order 4 and can go as high as + * order 7 in case lock debugging is turned on. + * + * Try harder to allocate and fallback to vmalloc to lower the chance + * of a failing mount. + */ + table_size = sizeof(*table) + sizeof(*h) * num_entries; + table = kzalloc(table_size, GFP_KERNEL | __GFP_NOWARN | __GFP_REPEAT); + if (!table) { + table = vzalloc(table_size); + if (!table) + return -ENOMEM; + } spin_lock_init(&table->cache_lock); INIT_LIST_HEAD(&table->stripe_cache); @@ -209,8 +221,12 @@ int btrfs_alloc_stripe_hash_table(struct btrfs_fs_info *info) } x = cmpxchg(&info->stripe_hash_table, NULL, table); - if (x) - kfree(x); + if (x) { + if (is_vmalloc_addr(x)) + vfree(x); + else + kfree(x); + } return 0; } @@ -420,7 +436,10 @@ void btrfs_free_stripe_hash_table(struct btrfs_fs_info *info) if (!info->stripe_hash_table) return; btrfs_clear_rbio_cache(info); - kfree(info->stripe_hash_table); + if (is_vmalloc_addr(info->stripe_hash_table)) + vfree(info->stripe_hash_table); + else + kfree(info->stripe_hash_table); info->stripe_hash_table = NULL; } -- cgit From 124fe663f93162d17b7e391705cac122101e93d8 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 1 Mar 2013 11:47:21 -0500 Subject: Btrfs: delete inline extents when we find them during logging Apparently when we do inline extents we allow the data to overlap the last chunk of the btrfs_file_extent_item, which means that we can possibly have a btrfs_file_extent_item that isn't actually as large as a btrfs_file_extent_item. This messes with us when we try to overwrite the extent when logging new extents since we expect for it to be the right size. To fix this just delete the item and try to do the insert again which will give us the proper sized btrfs_file_extent_item. This fixes a panic where map_private_extent_buffer would blow up because we're trying to write past the end of the leaf. Thanks, Cc: stable@vger.kernel.org Signed-off-by: Josef Bacik --- fs/btrfs/tree-log.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 8e85e0e4333d..c7ef569eb22a 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3300,6 +3300,7 @@ static int log_one_extent(struct btrfs_trans_handle *trans, int index = log->log_transid % 2; bool skip_csum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM; +insert: INIT_LIST_HEAD(&ordered_sums); btrfs_init_map_token(&token); key.objectid = btrfs_ino(inode); @@ -3315,6 +3316,23 @@ static int log_one_extent(struct btrfs_trans_handle *trans, leaf = path->nodes[0]; fi = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item); + + /* + * If we are overwriting an inline extent with a real one then we need + * to just delete the inline extent as it may not be large enough to + * have the entire file_extent_item. + */ + if (ret && btrfs_token_file_extent_type(leaf, fi, &token) == + BTRFS_FILE_EXTENT_INLINE) { + ret = btrfs_del_item(trans, log, path); + btrfs_release_path(path); + if (ret) { + path->really_keep_locks = 0; + return ret; + } + goto insert; + } + btrfs_set_token_file_extent_generation(leaf, fi, em->generation, &token); if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) { -- cgit From 180e001cd5fc2950dc6a7997dde5b65c954d0e79 Mon Sep 17 00:00:00 2001 From: Paul Gortmaker Date: Thu, 14 Feb 2013 13:50:15 -0700 Subject: btrfs: fixup/remove module.h usage as required We want to avoid module.h where posible, since it in turn includes nearly all of header space. This means removing it where it is not required, and using export.h where we are only exporting symbols via EXPORT_SYMBOL and friends. Signed-off-by: Paul Gortmaker Signed-off-by: Chris Mason --- fs/btrfs/extent_io.c | 1 - fs/btrfs/extent_map.c | 1 - fs/btrfs/sysfs.c | 1 - fs/btrfs/ulist.c | 2 +- 4 files changed, 1 insertion(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 0306665cb1d7..f173c5af6461 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -4,7 +4,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index fdb7a8db3b57..2834ca5768ea 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -1,6 +1,5 @@ #include #include -#include #include #include #include "ctree.h" diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c index daac9ae6d731..5b326cd60a4a 100644 --- a/fs/btrfs/sysfs.c +++ b/fs/btrfs/sysfs.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include "ctree.h" diff --git a/fs/btrfs/ulist.c b/fs/btrfs/ulist.c index 99be4c138db6..ddc61cad0080 100644 --- a/fs/btrfs/ulist.c +++ b/fs/btrfs/ulist.c @@ -5,7 +5,7 @@ */ #include -#include +#include #include "ulist.h" /* -- cgit From 9d8072e7c33991a08b29338f3f04dfef7c24afd5 Mon Sep 17 00:00:00 2001 From: Peter Huewe Date: Fri, 1 Mar 2013 19:46:40 +0800 Subject: autofs - Fix sparse warning: context imbalance in autofs4_d_automount() different lock contexts for basic block Sparse complains: fs/autofs4/root.c:409:9: sparse: context imbalance in 'autofs4_d_automount' - different lock contexts for basic block This was introduced by commit f55fb0c24386 ("autofs4 - dont clear DCACHE_NEED_AUTOMOUNT on rootless mount") The function autofs4_d_automount can be left with the (&sbi->fs_lock) held if sbi->version <= 4 and simple_empty(dentry) == false so the warning seems valid. --> Add an spin_unlock in this case before we jump to done Unfortunately compile tested only. Reported-by: Fengguang Wu Signed-off-by: Peter Huewe Acked-by: Ian Kent Signed-off-by: Linus Torvalds --- fs/autofs4/root.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index 230bd2aad4f4..9bd16255dd9c 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -383,8 +383,10 @@ static struct vfsmount *autofs4_d_automount(struct path *path) goto done; } } else { - if (!simple_empty(dentry)) + if (!simple_empty(dentry)) { + spin_unlock(&sbi->fs_lock); goto done; + } } ino->flags |= AUTOFS_INF_PENDING; spin_unlock(&sbi->fs_lock); -- cgit From 5140a8ceaa369d362d0ab7a438a5d99c75ee098d Mon Sep 17 00:00:00 2001 From: Tim Gardner Date: Fri, 1 Mar 2013 19:46:48 +0800 Subject: autofs4 - autofs4_catatonic_mode(): remove redundant null check on kfree() smatch analysis: fs/autofs4/waitq.c:46 autofs4_catatonic_mode() info: redundant null check on wq->name.name calling kfree() Signed-off-by: Tim Gardner Signed-off-by: Ian Kent Cc: autofs@vger.kernel.org Signed-off-by: Linus Torvalds --- fs/autofs4/waitq.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/autofs4/waitq.c b/fs/autofs4/waitq.c index 03bc1d347d8e..3db70dae40d3 100644 --- a/fs/autofs4/waitq.c +++ b/fs/autofs4/waitq.c @@ -42,10 +42,8 @@ void autofs4_catatonic_mode(struct autofs_sb_info *sbi) while (wq) { nwq = wq->next; wq->status = -ENOENT; /* Magic is gone - report failure */ - if (wq->name.name) { - kfree(wq->name.name); - wq->name.name = NULL; - } + kfree(wq->name.name); + wq->name.name = NULL; wq->wait_ctr--; wake_up_interruptible(&wq->queue); wq = nwq; -- cgit From dd37978c50bc8b354e5c4633f69387f16572fdac Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 1 Mar 2013 19:48:30 -0500 Subject: cache the value of file_inode() in struct file Note that this thing does *not* contribute to inode refcount; it's pinned down by dentry. Signed-off-by: Al Viro --- fs/file_table.c | 2 ++ fs/open.c | 3 ++- 2 files changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/file_table.c b/fs/file_table.c index aa07d3684a2e..cd4d87a82951 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -176,6 +176,7 @@ struct file *alloc_file(struct path *path, fmode_t mode, return file; file->f_path = *path; + file->f_inode = path->dentry->d_inode; file->f_mapping = path->dentry->d_inode->i_mapping; file->f_mode = mode; file->f_op = fop; @@ -258,6 +259,7 @@ static void __fput(struct file *file) drop_file_write_access(file); file->f_path.dentry = NULL; file->f_path.mnt = NULL; + file->f_inode = NULL; file_free(file); dput(dentry); mntput(mnt); diff --git a/fs/open.c b/fs/open.c index 62f907e3bc36..806d4589559f 100644 --- a/fs/open.c +++ b/fs/open.c @@ -689,7 +689,7 @@ static int do_dentry_open(struct file *f, f->f_mode = FMODE_PATH; path_get(&f->f_path); - inode = file_inode(f); + inode = f->f_inode = f->f_path.dentry->d_inode; if (f->f_mode & FMODE_WRITE) { error = __get_file_write_access(inode, f->f_path.mnt); if (error) @@ -752,6 +752,7 @@ cleanup_file: path_put(&f->f_path); f->f_path.mnt = NULL; f->f_path.dentry = NULL; + f->f_inode = NULL; return error; } -- cgit From dcf787f39162ce32ca325b3e784aba2d2444619a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 1 Mar 2013 23:51:07 -0500 Subject: constify path_get/path_put and fs_struct.c stuff Signed-off-by: Al Viro --- fs/fs_struct.c | 6 +++--- fs/internal.h | 2 +- fs/namei.c | 4 ++-- 3 files changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/fs_struct.c b/fs/fs_struct.c index fe6ca583bbc0..d8ac61d0c932 100644 --- a/fs/fs_struct.c +++ b/fs/fs_struct.c @@ -10,7 +10,7 @@ * Replace the fs->{rootmnt,root} with {mnt,dentry}. Put the old values. * It can block. */ -void set_fs_root(struct fs_struct *fs, struct path *path) +void set_fs_root(struct fs_struct *fs, const struct path *path) { struct path old_root; @@ -29,7 +29,7 @@ void set_fs_root(struct fs_struct *fs, struct path *path) * Replace the fs->{pwdmnt,pwd} with {mnt,dentry}. Put the old values. * It can block. */ -void set_fs_pwd(struct fs_struct *fs, struct path *path) +void set_fs_pwd(struct fs_struct *fs, const struct path *path) { struct path old_pwd; @@ -53,7 +53,7 @@ static inline int replace_path(struct path *p, const struct path *old, const str return 1; } -void chroot_fs_refs(struct path *old_root, struct path *new_root) +void chroot_fs_refs(const struct path *old_root, const struct path *new_root) { struct task_struct *g, *p; struct fs_struct *fs; diff --git a/fs/internal.h b/fs/internal.h index 2f6af7f645eb..507141fceb99 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -69,7 +69,7 @@ extern void __mnt_drop_write_file(struct file *); /* * fs_struct.c */ -extern void chroot_fs_refs(struct path *, struct path *); +extern void chroot_fs_refs(const struct path *, const struct path *); /* * file_table.c diff --git a/fs/namei.c b/fs/namei.c index dc984fee5532..961bc1268366 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -451,7 +451,7 @@ int inode_permission(struct inode *inode, int mask) * * Given a path increment the reference count to the dentry and the vfsmount. */ -void path_get(struct path *path) +void path_get(const struct path *path) { mntget(path->mnt); dget(path->dentry); @@ -464,7 +464,7 @@ EXPORT_SYMBOL(path_get); * * Given a path decrement the reference count to the dentry and the vfsmount. */ -void path_put(struct path *path) +void path_put(const struct path *path) { dput(path->dentry); mntput(path->mnt); -- cgit From 0e803bafbb7d1b8a9031104f1a982a01b45da4c6 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Sat, 2 Mar 2013 12:26:30 +0100 Subject: compat: restore timerfd settime and gettime compat syscalls Both compat syscalls got lost with 9d94b9e2 "switch timerfd compat syscalls to COMPAT_SYSCALL_DEFINE" because of a typo: COMPAT instead of CONFIG_COMPAT. Signed-off-by: Heiko Carstens Signed-off-by: Al Viro --- fs/timerfd.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/timerfd.c b/fs/timerfd.c index 0e606b12a59d..32b644f03690 100644 --- a/fs/timerfd.c +++ b/fs/timerfd.c @@ -383,10 +383,10 @@ SYSCALL_DEFINE2(timerfd_gettime, int, ufd, struct itimerspec __user *, otmr) return copy_to_user(otmr, &kotmr, sizeof(kotmr)) ? -EFAULT: 0; } -#ifdef COMPAT +#ifdef CONFIG_COMPAT COMPAT_SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, - const struct itimerspec __user *, utmr, - struct itimerspec __user *, otmr) + const struct compat_itimerspec __user *, utmr, + struct compat_itimerspec __user *, otmr) { struct itimerspec new, old; int ret; @@ -402,12 +402,12 @@ COMPAT_SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, } COMPAT_SYSCALL_DEFINE2(timerfd_gettime, int, ufd, - struct itimerspec __user *, otmr) + struct compat_itimerspec __user *, otmr) { struct itimerspec kotmr; int ret = do_timerfd_gettime(ufd, &kotmr); if (ret) return ret; - return put_compat_itimerspec(otmr, &t) ? -EFAULT: 0; + return put_compat_itimerspec(otmr, &kotmr) ? -EFAULT: 0; } #endif -- cgit From 1ac6466f253ef7bd063b7877fb056afe1820841c Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 2 Mar 2013 10:27:46 -0500 Subject: ext4: use percpu counter for extent cache count Use a percpu counter rather than atomic types for shrinker accounting. There's no need for ultimate accuracy in the shrinker, so this should come a little more cheaply. The percpu struct is somewhat large, but there was a big gap before the cache-aligned s_es_lru_lock anyway, and it fits nicely in there. Signed-off-by: Eric Sandeen Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 2 +- fs/ext4/extents_status.c | 8 ++++---- fs/ext4/super.c | 5 +++++ 3 files changed, 10 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 96c10934bb96..4a01ba315262 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1268,7 +1268,6 @@ struct ext4_sb_info { atomic_t s_mb_preallocated; atomic_t s_mb_discarded; atomic_t s_lock_busy; - atomic_t s_extent_cache_cnt; /* locality groups */ struct ext4_locality_group __percpu *s_locality_groups; @@ -1310,6 +1309,7 @@ struct ext4_sb_info { /* Reclaim extents from extent status tree */ struct shrinker s_es_shrinker; struct list_head s_es_lru; + struct percpu_counter s_extent_cache_cnt; spinlock_t s_es_lru_lock ____cacheline_aligned_in_smp; }; diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 27fcdd2b2607..95796a1b7522 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -305,7 +305,7 @@ ext4_es_alloc_extent(struct inode *inode, ext4_lblk_t lblk, ext4_lblk_t len, */ if (!ext4_es_is_delayed(es)) { EXT4_I(inode)->i_es_lru_nr++; - atomic_inc(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt); + percpu_counter_inc(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt); } return es; @@ -317,7 +317,7 @@ static void ext4_es_free_extent(struct inode *inode, struct extent_status *es) if (!ext4_es_is_delayed(es)) { BUG_ON(EXT4_I(inode)->i_es_lru_nr == 0); EXT4_I(inode)->i_es_lru_nr--; - atomic_dec(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt); + percpu_counter_dec(&EXT4_SB(inode->i_sb)->s_extent_cache_cnt); } kmem_cache_free(ext4_es_cachep, es); @@ -678,7 +678,7 @@ static int ext4_es_shrink(struct shrinker *shrink, struct shrink_control *sc) int nr_to_scan = sc->nr_to_scan; int ret, nr_shrunk = 0; - ret = atomic_read(&sbi->s_extent_cache_cnt); + ret = percpu_counter_read_positive(&sbi->s_extent_cache_cnt); trace_ext4_es_shrink_enter(sbi->s_sb, nr_to_scan, ret); if (!nr_to_scan) @@ -711,7 +711,7 @@ static int ext4_es_shrink(struct shrinker *shrink, struct shrink_control *sc) list_splice_tail(&scanned, &sbi->s_es_lru); spin_unlock(&sbi->s_es_lru_lock); - ret = atomic_read(&sbi->s_extent_cache_cnt); + ret = percpu_counter_read_positive(&sbi->s_extent_cache_cnt); trace_ext4_es_shrink_exit(sbi->s_sb, nr_shrunk, ret); return ret; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 373d46cd5d3f..1ae5860b30a3 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -783,6 +783,7 @@ static void ext4_put_super(struct super_block *sb) percpu_counter_destroy(&sbi->s_freeinodes_counter); percpu_counter_destroy(&sbi->s_dirs_counter); percpu_counter_destroy(&sbi->s_dirtyclusters_counter); + percpu_counter_destroy(&sbi->s_extent_cache_cnt); brelse(sbi->s_sbh); #ifdef CONFIG_QUOTA for (i = 0; i < MAXQUOTAS; i++) @@ -3688,6 +3689,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) if (!err) { err = percpu_counter_init(&sbi->s_dirtyclusters_counter, 0); } + if (!err) { + err = percpu_counter_init(&sbi->s_extent_cache_cnt, 0); + } if (err) { ext4_msg(sb, KERN_ERR, "insufficient memory"); goto failed_mount3; @@ -3993,6 +3997,7 @@ failed_mount3: percpu_counter_destroy(&sbi->s_freeinodes_counter); percpu_counter_destroy(&sbi->s_dirs_counter); percpu_counter_destroy(&sbi->s_dirtyclusters_counter); + percpu_counter_destroy(&sbi->s_extent_cache_cnt); if (sbi->s_mmp_tsk) kthread_stop(sbi->s_mmp_tsk); failed_mount2: -- cgit From c07380beafabee343a7f298bacd22598d72c7c87 Mon Sep 17 00:00:00 2001 From: James Hogan Date: Mon, 9 May 2011 10:58:40 +0100 Subject: Revert some of "binfmt_elf: cleanups" The commit "binfmt_elf: cleanups" (f670d0ecda73b7438eec9ed108680bc5f5362ad8) removed an ifndef elf_map but this breaks compilation for metag which does define elf_map. This adds the ifndef back in as it was before, but does not affect the other cleanups made by that patch. Signed-off-by: James Hogan Cc: Alexander Viro Cc: linux-fsdevel@vger.kernel.org Acked-by: Mikael Pettersson --- fs/binfmt_elf.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 0c42cdbabecf..cfc22c9d75bc 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -321,6 +321,8 @@ create_elf_tables(struct linux_binprm *bprm, struct elfhdr *exec, return 0; } +#ifndef elf_map + static unsigned long elf_map(struct file *filep, unsigned long addr, struct elf_phdr *eppnt, int prot, int type, unsigned long total_size) @@ -355,6 +357,8 @@ static unsigned long elf_map(struct file *filep, unsigned long addr, return(map_addr); } +#endif /* !elf_map */ + static unsigned long total_mapping_size(struct elf_phdr *cmds, int nr) { int i, first_idx = -1, last_idx = -1; -- cgit From df05c1b85a88888b92725ef27ba70086f38fb083 Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Sat, 2 Mar 2013 17:08:46 -0500 Subject: jbd2: fix ERR_PTR dereference in jbd2__journal_start If start_this_handle() failed handle will be initialized to ERR_PTR() and can not be dereferenced. paging request at fffffffffffffff6 IP: [] jbd2__journal_start+0x18f/0x290 PGD 200e067 PUD 200f067 PMD 0 Oops: 0000 [#1] SMP Modules linked in: cpufreq_ondemand acpi_cpufreq freq_table mperf coretemp kvm_intel kvm crc32c_intel ghash_clmulni_intel microcode sg xhci_hcd button sd_mod crc_t10dif aesni_intel ablk_helper cryptd lrw aes_x86_64 xts gf128mul ahci libahci pata_acpi ata_generic dm_mirror dm_region_hash dm_log dm_mod CPU 0 journal commit I/O error Pid: 2694, comm: fio Not tainted 3.8.0-rc3+ #79 /DQ67SW RIP: 0010:[] [] jbd2__journal_start+0x18f/0x290 RSP: 0018:ffff880233b8ba58 EFLAGS: 00010292 RAX: 00000000ffffffe2 RBX: ffffffffffffffe2 RCX: 0000000000000006 RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffffffff82128f48 RBP: ffff880233b8ba98 R08: 0000000000000000 R09: ffff88021440a6e0 Signed-off-by: Dmitry Monakhov Signed-off-by: "Theodore Ts'o" --- fs/jbd2/transaction.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index b7e2385c6e92..d6ee5aed56b1 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -382,7 +382,7 @@ handle_t *jbd2__journal_start(journal_t *journal, int nblocks, gfp_t gfp_mask, if (err < 0) { jbd2_free_handle(handle); current->journal_info = NULL; - handle = ERR_PTR(err); + return ERR_PTR(err); } handle->h_type = type; handle->h_line_no = line_no; -- cgit From 3e36a16375eee390a9d80f99499fe630efa08128 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Sat, 2 Mar 2013 17:13:55 -0500 Subject: ext4: fix possible memory leak in ext4_remount() 'orig_data' is malloced in ext4_remount() and should be freed before leaving from the error handling cases, otherwise it will cause memory leak. Signed-off-by: Wei Yongjun Signed-off-by: "Theodore Ts'o" Reviewed-by: Lukas Czerner Cc: stable@vger.kernel.org --- fs/ext4/super.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 1ae5860b30a3..6ac9c8a3eb08 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4543,6 +4543,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) if (!old_opts.s_qf_names[i]) { for (j = 0; j < i; j++) kfree(old_opts.s_qf_names[j]); + kfree(orig_data); return -ENOMEM; } } else -- cgit From 810da240f221d64bf90020f25941b05b378186fe Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Sat, 2 Mar 2013 17:18:58 -0500 Subject: ext4: convert number of blocks to clusters properly We're using macro EXT4_B2C() to convert number of blocks to number of clusters for bigalloc file systems. However, we should be using EXT4_NUM_B2C(). Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/ext4/balloc.c | 2 +- fs/ext4/mballoc.c | 8 ++++---- fs/ext4/resize.c | 6 +++--- fs/ext4/super.c | 2 +- 4 files changed, 9 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 2f2e0da1a6b7..92e68b33fffd 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -635,7 +635,7 @@ ext4_fsblk_t ext4_count_free_clusters(struct super_block *sb) brelse(bitmap_bh); printk(KERN_DEBUG "ext4_count_free_clusters: stored = %llu" ", computed = %llu, %llu\n", - EXT4_B2C(EXT4_SB(sb), ext4_free_blocks_count(es)), + EXT4_NUM_B2C(EXT4_SB(sb), ext4_free_blocks_count(es)), desc_count, bitmap_count); return bitmap_count; #else diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 6540ebe058e3..7bb713a46fe4 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -3419,7 +3419,7 @@ ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) win = offs; ac->ac_b_ex.fe_logical = ac->ac_o_ex.fe_logical - - EXT4_B2C(sbi, win); + EXT4_NUM_B2C(sbi, win); BUG_ON(ac->ac_o_ex.fe_logical < ac->ac_b_ex.fe_logical); BUG_ON(ac->ac_o_ex.fe_len > ac->ac_b_ex.fe_len); } @@ -4565,7 +4565,7 @@ do_more: EXT4_BLOCKS_PER_GROUP(sb); count -= overflow; } - count_clusters = EXT4_B2C(sbi, count); + count_clusters = EXT4_NUM_B2C(sbi, count); bitmap_bh = ext4_read_block_bitmap(sb, block_group); if (!bitmap_bh) { err = -EIO; @@ -4807,11 +4807,11 @@ int ext4_group_add_blocks(handle_t *handle, struct super_block *sb, ext4_group_desc_csum_set(sb, block_group, desc); ext4_unlock_group(sb, block_group); percpu_counter_add(&sbi->s_freeclusters_counter, - EXT4_B2C(sbi, blocks_freed)); + EXT4_NUM_B2C(sbi, blocks_freed)); if (sbi->s_log_groups_per_flex) { ext4_group_t flex_group = ext4_flex_group(sbi, block_group); - atomic_add(EXT4_B2C(sbi, blocks_freed), + atomic_add(EXT4_NUM_B2C(sbi, blocks_freed), &sbi->s_flex_groups[flex_group].free_clusters); } diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index c7f4d7584669..b2c8ee56eb98 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -1247,7 +1247,7 @@ static int ext4_setup_new_descs(handle_t *handle, struct super_block *sb, ext4_inode_table_set(sb, gdp, group_data->inode_table); ext4_free_group_clusters_set(sb, gdp, - EXT4_B2C(sbi, group_data->free_blocks_count)); + EXT4_NUM_B2C(sbi, group_data->free_blocks_count)); ext4_free_inodes_set(sb, gdp, EXT4_INODES_PER_GROUP(sb)); if (ext4_has_group_desc_csum(sb)) ext4_itable_unused_set(sb, gdp, @@ -1349,7 +1349,7 @@ static void ext4_update_super(struct super_block *sb, /* Update the free space counts */ percpu_counter_add(&sbi->s_freeclusters_counter, - EXT4_B2C(sbi, free_blocks)); + EXT4_NUM_B2C(sbi, free_blocks)); percpu_counter_add(&sbi->s_freeinodes_counter, EXT4_INODES_PER_GROUP(sb) * flex_gd->count); @@ -1360,7 +1360,7 @@ static void ext4_update_super(struct super_block *sb, sbi->s_log_groups_per_flex) { ext4_group_t flex_group; flex_group = ext4_flex_group(sbi, group_data[0].group); - atomic_add(EXT4_B2C(sbi, free_blocks), + atomic_add(EXT4_NUM_B2C(sbi, free_blocks), &sbi->s_flex_groups[flex_group].free_clusters); atomic_add(EXT4_INODES_PER_GROUP(sb) * flex_gd->count, &sbi->s_flex_groups[flex_group].free_inodes); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 6ac9c8a3eb08..990ceab2e2bc 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3162,7 +3162,7 @@ int ext4_calculate_overhead(struct super_block *sb) } /* Add the journal blocks as well */ if (sbi->s_journal) - overhead += EXT4_B2C(sbi, sbi->s_journal->j_maxlen); + overhead += EXT4_NUM_B2C(sbi, sbi->s_journal->j_maxlen); sbi->s_overhead = overhead; smp_wmb(); -- cgit From d4e439549127d3ca544482551f5f1af4f114debd Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Sat, 2 Mar 2013 17:24:05 -0500 Subject: ext4: fix a warning from sparse check for ext4_dir_llseek ext4_dir_llseek is only used as a callback function, and no one calls it directly. So make it as a static function in order to remove a warning message from sparse check. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" --- fs/ext4/dir.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index 3882fbc5e215..b8eeaf15b034 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -334,7 +334,7 @@ static inline loff_t ext4_get_htree_eof(struct file *filp) * * For non-htree, ext4_llseek already chooses the proper max offset. */ -loff_t ext4_dir_llseek(struct file *file, loff_t offset, int whence) +static loff_t ext4_dir_llseek(struct file *file, loff_t offset, int whence) { struct inode *inode = file->f_mapping->host; int dx_dir = is_dx_dir(inode); -- cgit From 262b4662f42787bff24453ddd3e657265b5d0039 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sat, 2 Mar 2013 17:57:08 -0500 Subject: ext4: don't allow quota mount options when quota feature enabled So far we silently ignored when quota mount options were set while quota feature was enabled. But this can create confusion in userspace when mount options are set but silently ignored and also creates opportunities for bugs when we don't properly test all quota types. Actually ext4_mark_dquot_dirty() forgets to test for quota feature so it was dependent on journaled quota options being set. OTOH ext4_orphan_cleanup() tries to enable journaled quota when quota options are specified which is wrong when quota feature is enabled. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/super.c | 33 ++++++++++++++++++++++++++------- 1 file changed, 26 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 990ceab2e2bc..b7deba7d6a44 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1248,6 +1248,11 @@ static int set_qf_name(struct super_block *sb, int qtype, substring_t *args) "quota options when quota turned on"); return -1; } + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA)) { + ext4_msg(sb, KERN_ERR, "Cannot set journaled quota options " + "when QUOTA feature is enabled"); + return -1; + } qname = match_strdup(args); if (!qname) { ext4_msg(sb, KERN_ERR, @@ -1545,6 +1550,13 @@ static int handle_mount_opt(struct super_block *sb, char *opt, int token, "quota options when quota turned on"); return -1; } + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, + EXT4_FEATURE_RO_COMPAT_QUOTA)) { + ext4_msg(sb, KERN_ERR, + "Cannot set journaled quota options " + "when QUOTA feature is enabled"); + return -1; + } sbi->s_jquota_fmt = m->mount_opt; #endif } else { @@ -1593,6 +1605,12 @@ static int parse_options(char *options, struct super_block *sb, return 0; } #ifdef CONFIG_QUOTA + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA) && + (test_opt(sb, USRQUOTA) || test_opt(sb, GRPQUOTA))) { + ext4_msg(sb, KERN_ERR, "Cannot set quota options when QUOTA " + "feature is enabled"); + return 0; + } if (sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) { if (test_opt(sb, USRQUOTA) && sbi->s_qf_names[USRQUOTA]) clear_opt(sb, USRQUOTA); @@ -3715,13 +3733,11 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sb->s_export_op = &ext4_export_ops; sb->s_xattr = ext4_xattr_handlers; #ifdef CONFIG_QUOTA - sb->s_qcop = &ext4_qctl_operations; sb->dq_op = &ext4_quota_operations; - - if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA)) { - /* Use qctl operations for hidden quota files. */ + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA)) sb->s_qcop = &ext4_qctl_sysfile_operations; - } + else + sb->s_qcop = &ext4_qctl_operations; #endif memcpy(sb->s_uuid, es->s_uuid, sizeof(es->s_uuid)); @@ -4822,9 +4838,12 @@ static int ext4_release_dquot(struct dquot *dquot) static int ext4_mark_dquot_dirty(struct dquot *dquot) { + struct super_block *sb = dquot->dq_sb; + struct ext4_sb_info *sbi = EXT4_SB(sb); + /* Are we journaling quotas? */ - if (EXT4_SB(dquot->dq_sb)->s_qf_names[USRQUOTA] || - EXT4_SB(dquot->dq_sb)->s_qf_names[GRPQUOTA]) { + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA) || + sbi->s_qf_names[USRQUOTA] || sbi->s_qf_names[GRPQUOTA]) { dquot_mark_dquot_dirty(dquot); return ext4_write_dquot(dquot); } else { -- cgit From 9b2ff35753c0512bc8c6adae9e9c87cbeee86f82 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sat, 2 Mar 2013 18:22:38 -0500 Subject: ext4: enable quotas before orphan cleanup When using quota feature we need to enable quotas before orphan cleanup so that changes happening during it are properly reflected in quota accounting. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/super.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index b7deba7d6a44..9379b7fbfd92 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3933,6 +3933,16 @@ no_journal: if (err) goto failed_mount7; +#ifdef CONFIG_QUOTA + /* Enable quota usage during mount. */ + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA) && + !(sb->s_flags & MS_RDONLY)) { + err = ext4_enable_quotas(sb); + if (err) + goto failed_mount8; + } +#endif /* CONFIG_QUOTA */ + EXT4_SB(sb)->s_mount_state |= EXT4_ORPHAN_FS; ext4_orphan_cleanup(sb, es); EXT4_SB(sb)->s_mount_state &= ~EXT4_ORPHAN_FS; @@ -3950,16 +3960,6 @@ no_journal: } else descr = "out journal"; -#ifdef CONFIG_QUOTA - /* Enable quota usage during mount. */ - if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA) && - !(sb->s_flags & MS_RDONLY)) { - err = ext4_enable_quotas(sb); - if (err) - goto failed_mount8; - } -#endif /* CONFIG_QUOTA */ - if (test_opt(sb, DISCARD)) { struct request_queue *q = bdev_get_queue(sb->s_bdev); if (!blk_queue_discard(q)) -- cgit From d7011f5b9dd3c88feb5cebad7d0b719ed6816ebd Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Sun, 3 Mar 2013 04:44:41 -0700 Subject: btrfs/raid56: Add missing #include tilegx_defconfig: fs/btrfs/raid56.c: In function 'btrfs_alloc_stripe_hash_table': fs/btrfs/raid56.c:206:3: error: implicit declaration of function 'vzalloc' [-Werror=implicit-function-declaration] fs/btrfs/raid56.c:206:9: warning: assignment makes pointer from integer without a cast [enabled by default] fs/btrfs/raid56.c:226:4: error: implicit declaration of function 'vfree' [-Werror=implicit-function-declaration] Signed-off-by: Geert Uytterhoeven Signed-off-by: Chris Mason --- fs/btrfs/raid56.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c index 07222053c7d8..9a79fb790adb 100644 --- a/fs/btrfs/raid56.c +++ b/fs/btrfs/raid56.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include "compat.h" #include "ctree.h" -- cgit From 7f78e0351394052e1a6293e175825eb5c7869507 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 2 Mar 2013 19:39:14 -0800 Subject: fs: Limit sys_mount to only request filesystem modules. Modify the request_module to prefix the file system type with "fs-" and add aliases to all of the filesystems that can be built as modules to match. A common practice is to build all of the kernel code and leave code that is not commonly needed as modules, with the result that many users are exposed to any bug anywhere in the kernel. Looking for filesystems with a fs- prefix limits the pool of possible modules that can be loaded by mount to just filesystems trivially making things safer with no real cost. Using aliases means user space can control the policy of which filesystem modules are auto-loaded by editing /etc/modprobe.d/*.conf with blacklist and alias directives. Allowing simple, safe, well understood work-arounds to known problematic software. This also addresses a rare but unfortunate problem where the filesystem name is not the same as it's module name and module auto-loading would not work. While writing this patch I saw a handful of such cases. The most significant being autofs that lives in the module autofs4. This is relevant to user namespaces because we can reach the request module in get_fs_type() without having any special permissions, and people get uncomfortable when a user specified string (in this case the filesystem type) goes all of the way to request_module. After having looked at this issue I don't think there is any particular reason to perform any filtering or permission checks beyond making it clear in the module request that we want a filesystem module. The common pattern in the kernel is to call request_module() without regards to the users permissions. In general all a filesystem module does once loaded is call register_filesystem() and go to sleep. Which means there is not much attack surface exposed by loading a filesytem module unless the filesystem is mounted. In a user namespace filesystems are not mounted unless .fs_flags = FS_USERNS_MOUNT, which most filesystems do not set today. Acked-by: Serge Hallyn Acked-by: Kees Cook Reported-by: Kees Cook Signed-off-by: "Eric W. Biederman" --- fs/9p/vfs_super.c | 1 + fs/adfs/super.c | 1 + fs/affs/super.c | 1 + fs/afs/super.c | 1 + fs/autofs4/init.c | 1 + fs/befs/linuxvfs.c | 1 + fs/bfs/inode.c | 1 + fs/binfmt_misc.c | 1 + fs/btrfs/super.c | 1 + fs/ceph/super.c | 1 + fs/coda/inode.c | 1 + fs/configfs/mount.c | 1 + fs/cramfs/inode.c | 1 + fs/debugfs/inode.c | 1 + fs/devpts/inode.c | 1 + fs/ecryptfs/main.c | 1 + fs/efs/super.c | 1 + fs/exofs/super.c | 1 + fs/ext2/super.c | 1 + fs/ext3/super.c | 1 + fs/ext4/super.c | 5 +++-- fs/f2fs/super.c | 1 + fs/fat/namei_msdos.c | 1 + fs/fat/namei_vfat.c | 1 + fs/filesystems.c | 2 +- fs/freevxfs/vxfs_super.c | 2 +- fs/fuse/control.c | 1 + fs/fuse/inode.c | 2 ++ fs/gfs2/ops_fstype.c | 4 +++- fs/hfs/super.c | 1 + fs/hfsplus/super.c | 1 + fs/hppfs/hppfs.c | 1 + fs/hugetlbfs/inode.c | 1 + fs/isofs/inode.c | 3 +-- fs/jffs2/super.c | 1 + fs/jfs/super.c | 1 + fs/logfs/super.c | 1 + fs/minix/inode.c | 1 + fs/ncpfs/inode.c | 1 + fs/nfs/super.c | 3 ++- fs/nfsd/nfsctl.c | 1 + fs/nilfs2/super.c | 1 + fs/ntfs/super.c | 1 + fs/ocfs2/dlmfs/dlmfs.c | 1 + fs/omfs/inode.c | 1 + fs/openpromfs/inode.c | 1 + fs/qnx4/inode.c | 1 + fs/qnx6/inode.c | 1 + fs/reiserfs/super.c | 1 + fs/romfs/super.c | 1 + fs/sysv/super.c | 3 ++- fs/ubifs/super.c | 1 + fs/ufs/super.c | 1 + fs/xfs/xfs_super.c | 1 + 54 files changed, 61 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 91dad63e5a2d..2756dcd5de6e 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -365,3 +365,4 @@ struct file_system_type v9fs_fs_type = { .owner = THIS_MODULE, .fs_flags = FS_RENAME_DOES_D_MOVE, }; +MODULE_ALIAS_FS("9p"); diff --git a/fs/adfs/super.c b/fs/adfs/super.c index d57122935793..0ff4bae2c2a2 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -524,6 +524,7 @@ static struct file_system_type adfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("adfs"); static int __init init_adfs_fs(void) { diff --git a/fs/affs/super.c b/fs/affs/super.c index b84dc7352502..45161a832bbc 100644 --- a/fs/affs/super.c +++ b/fs/affs/super.c @@ -622,6 +622,7 @@ static struct file_system_type affs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("affs"); static int __init init_affs_fs(void) { diff --git a/fs/afs/super.c b/fs/afs/super.c index 7c31ec399575..c4861557e385 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -45,6 +45,7 @@ struct file_system_type afs_fs_type = { .kill_sb = afs_kill_super, .fs_flags = 0, }; +MODULE_ALIAS_FS("afs"); static const struct super_operations afs_super_ops = { .statfs = afs_statfs, diff --git a/fs/autofs4/init.c b/fs/autofs4/init.c index cddc74b9cdb2..b3db517e89ec 100644 --- a/fs/autofs4/init.c +++ b/fs/autofs4/init.c @@ -26,6 +26,7 @@ static struct file_system_type autofs_fs_type = { .mount = autofs_mount, .kill_sb = autofs4_kill_sb, }; +MODULE_ALIAS_FS("autofs"); static int __init init_autofs4_fs(void) { diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index c8f4e25eb9e2..8615ee89ab55 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -951,6 +951,7 @@ static struct file_system_type befs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("befs"); static int __init init_befs_fs(void) diff --git a/fs/bfs/inode.c b/fs/bfs/inode.c index 737aaa3f7090..5e376bb93419 100644 --- a/fs/bfs/inode.c +++ b/fs/bfs/inode.c @@ -473,6 +473,7 @@ static struct file_system_type bfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("bfs"); static int __init init_bfs_fs(void) { diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index fecbbf3f8ff2..751df5e4f61a 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -720,6 +720,7 @@ static struct file_system_type bm_fs_type = { .mount = bm_mount, .kill_sb = kill_litter_super, }; +MODULE_ALIAS_FS("binfmt_misc"); static int __init init_misc_binfmt(void) { diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 68a29a1ea068..f6b88595f858 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1558,6 +1558,7 @@ static struct file_system_type btrfs_fs_type = { .kill_sb = btrfs_kill_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("btrfs"); /* * used by btrfsctl to scan devices when no FS is mounted diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 9fe17c6c2876..6ddc0bca56b2 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -952,6 +952,7 @@ static struct file_system_type ceph_fs_type = { .kill_sb = ceph_kill_sb, .fs_flags = FS_RENAME_DOES_D_MOVE, }; +MODULE_ALIAS_FS("ceph"); #define _STRINGIFY(x) #x #define STRINGIFY(x) _STRINGIFY(x) diff --git a/fs/coda/inode.c b/fs/coda/inode.c index dada9d0abede..4dcc0d81a7aa 100644 --- a/fs/coda/inode.c +++ b/fs/coda/inode.c @@ -329,4 +329,5 @@ struct file_system_type coda_fs_type = { .kill_sb = kill_anon_super, .fs_flags = FS_BINARY_MOUNTDATA, }; +MODULE_ALIAS_FS("coda"); diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index aee0a7ebbd8e..7f26c3cf75ae 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -114,6 +114,7 @@ static struct file_system_type configfs_fs_type = { .mount = configfs_do_mount, .kill_sb = kill_litter_super, }; +MODULE_ALIAS_FS("configfs"); struct dentry *configfs_pin_fs(void) { diff --git a/fs/cramfs/inode.c b/fs/cramfs/inode.c index 3ceb9ec976e1..35b1c7bd18b7 100644 --- a/fs/cramfs/inode.c +++ b/fs/cramfs/inode.c @@ -573,6 +573,7 @@ static struct file_system_type cramfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("cramfs"); static int __init init_cramfs_fs(void) { diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 0c4f80b447fb..4888cb3fdef7 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -299,6 +299,7 @@ static struct file_system_type debug_fs_type = { .mount = debug_mount, .kill_sb = kill_litter_super, }; +MODULE_ALIAS_FS("debugfs"); static struct dentry *__create_file(const char *name, umode_t mode, struct dentry *parent, void *data, diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 073d30b9d1ac..79b662985efe 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -510,6 +510,7 @@ static struct file_system_type devpts_fs_type = { .fs_flags = FS_USERNS_MOUNT | FS_USERNS_DEV_MOUNT, #endif }; +MODULE_ALIAS_FS("devpts"); /* * The normal naming convention is simply /dev/pts/; this conforms diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index 4e0886c9e5c4..e924cf45aad9 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -629,6 +629,7 @@ static struct file_system_type ecryptfs_fs_type = { .kill_sb = ecryptfs_kill_block_super, .fs_flags = 0 }; +MODULE_ALIAS_FS("ecryptfs"); /** * inode_info_init_once diff --git a/fs/efs/super.c b/fs/efs/super.c index 2002431ef9a0..c6f57a74a559 100644 --- a/fs/efs/super.c +++ b/fs/efs/super.c @@ -33,6 +33,7 @@ static struct file_system_type efs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("efs"); static struct pt_types sgi_pt_types[] = { {0x00, "SGI vh"}, diff --git a/fs/exofs/super.c b/fs/exofs/super.c index 5e59280d42d7..9d9763328734 100644 --- a/fs/exofs/super.c +++ b/fs/exofs/super.c @@ -1010,6 +1010,7 @@ static struct file_system_type exofs_type = { .mount = exofs_mount, .kill_sb = generic_shutdown_super, }; +MODULE_ALIAS_FS("exofs"); static int __init init_exofs(void) { diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 7f68c8114026..288534920fe5 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -1536,6 +1536,7 @@ static struct file_system_type ext2_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("ext2"); static int __init init_ext2_fs(void) { diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 5546ca225ffe..1d6e2ed85322 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -3068,6 +3068,7 @@ static struct file_system_type ext3_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("ext3"); static int __init init_ext3_fs(void) { diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 5e6c87836193..34e855219231 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -90,6 +90,7 @@ static struct file_system_type ext2_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("ext2"); #define IS_EXT2_SB(sb) ((sb)->s_bdev->bd_holder == &ext2_fs_type) #else #define IS_EXT2_SB(sb) (0) @@ -104,6 +105,7 @@ static struct file_system_type ext3_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("ext3"); #define IS_EXT3_SB(sb) ((sb)->s_bdev->bd_holder == &ext3_fs_type) #else #define IS_EXT3_SB(sb) (0) @@ -5152,7 +5154,6 @@ static inline int ext2_feature_set_ok(struct super_block *sb) return 0; return 1; } -MODULE_ALIAS("ext2"); #else static inline void register_as_ext2(void) { } static inline void unregister_as_ext2(void) { } @@ -5185,7 +5186,6 @@ static inline int ext3_feature_set_ok(struct super_block *sb) return 0; return 1; } -MODULE_ALIAS("ext3"); #else static inline void register_as_ext3(void) { } static inline void unregister_as_ext3(void) { } @@ -5199,6 +5199,7 @@ static struct file_system_type ext4_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("ext4"); static int __init ext4_init_feat_adverts(void) { diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 8c117649a035..fea6e582a2ed 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -687,6 +687,7 @@ static struct file_system_type f2fs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("f2fs"); static int __init init_inodecache(void) { diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c index e2cfda94a28d..081b759cff83 100644 --- a/fs/fat/namei_msdos.c +++ b/fs/fat/namei_msdos.c @@ -668,6 +668,7 @@ static struct file_system_type msdos_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("msdos"); static int __init init_msdos_fs(void) { diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c index ac959d655e7d..2da952036a3d 100644 --- a/fs/fat/namei_vfat.c +++ b/fs/fat/namei_vfat.c @@ -1073,6 +1073,7 @@ static struct file_system_type vfat_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("vfat"); static int __init init_vfat_fs(void) { diff --git a/fs/filesystems.c b/fs/filesystems.c index da165f6adcbf..92567d95ba6a 100644 --- a/fs/filesystems.c +++ b/fs/filesystems.c @@ -273,7 +273,7 @@ struct file_system_type *get_fs_type(const char *name) int len = dot ? dot - name : strlen(name); fs = __get_fs_type(name, len); - if (!fs && (request_module("%.*s", len, name) == 0)) + if (!fs && (request_module("fs-%.*s", len, name) == 0)) fs = __get_fs_type(name, len); if (dot && fs && !(fs->fs_flags & FS_HAS_SUBTYPE)) { diff --git a/fs/freevxfs/vxfs_super.c b/fs/freevxfs/vxfs_super.c index fed2c8afb3a9..455074308069 100644 --- a/fs/freevxfs/vxfs_super.c +++ b/fs/freevxfs/vxfs_super.c @@ -52,7 +52,6 @@ MODULE_AUTHOR("Christoph Hellwig"); MODULE_DESCRIPTION("Veritas Filesystem (VxFS) driver"); MODULE_LICENSE("Dual BSD/GPL"); -MODULE_ALIAS("vxfs"); /* makes mount -t vxfs autoload the module */ static void vxfs_put_super(struct super_block *); @@ -258,6 +257,7 @@ static struct file_system_type vxfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("vxfs"); /* makes mount -t vxfs autoload the module */ static int __init vxfs_init(void) diff --git a/fs/fuse/control.c b/fs/fuse/control.c index b7978b9f75ef..a0b0855d00a9 100644 --- a/fs/fuse/control.c +++ b/fs/fuse/control.c @@ -341,6 +341,7 @@ static struct file_system_type fuse_ctl_fs_type = { .mount = fuse_ctl_mount, .kill_sb = fuse_ctl_kill_sb, }; +MODULE_ALIAS_FS("fusectl"); int __init fuse_ctl_init(void) { diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index df00993ed108..137185c3884f 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -1117,6 +1117,7 @@ static struct file_system_type fuse_fs_type = { .mount = fuse_mount, .kill_sb = fuse_kill_sb_anon, }; +MODULE_ALIAS_FS("fuse"); #ifdef CONFIG_BLOCK static struct dentry *fuse_mount_blk(struct file_system_type *fs_type, @@ -1146,6 +1147,7 @@ static struct file_system_type fuseblk_fs_type = { .kill_sb = fuse_kill_sb_blk, .fs_flags = FS_REQUIRES_DEV | FS_HAS_SUBTYPE, }; +MODULE_ALIAS_FS("fuseblk"); static inline int register_fuseblk(void) { diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 1b612be4b873..60ede2a0f43f 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -20,6 +20,7 @@ #include #include #include +#include #include "gfs2.h" #include "incore.h" @@ -1425,6 +1426,7 @@ struct file_system_type gfs2_fs_type = { .kill_sb = gfs2_kill_sb, .owner = THIS_MODULE, }; +MODULE_ALIAS_FS("gfs2"); struct file_system_type gfs2meta_fs_type = { .name = "gfs2meta", @@ -1432,4 +1434,4 @@ struct file_system_type gfs2meta_fs_type = { .mount = gfs2_mount_meta, .owner = THIS_MODULE, }; - +MODULE_ALIAS_FS("gfs2meta"); diff --git a/fs/hfs/super.c b/fs/hfs/super.c index e93ddaadfd1e..bbaaa8a4ee64 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -466,6 +466,7 @@ static struct file_system_type hfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("hfs"); static void hfs_init_once(void *p) { diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 974c26f96fae..7b87284e46dc 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -654,6 +654,7 @@ static struct file_system_type hfsplus_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("hfsplus"); static void hfsplus_init_once(void *p) { diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c index 74f55703be49..126d3c2e2dee 100644 --- a/fs/hppfs/hppfs.c +++ b/fs/hppfs/hppfs.c @@ -748,6 +748,7 @@ static struct file_system_type hppfs_type = { .kill_sb = kill_anon_super, .fs_flags = 0, }; +MODULE_ALIAS_FS("hppfs"); static int __init init_hppfs(void) { diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 7f94e0cbc69c..84e3d856e91d 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -896,6 +896,7 @@ static struct file_system_type hugetlbfs_fs_type = { .mount = hugetlbfs_mount, .kill_sb = kill_litter_super, }; +MODULE_ALIAS_FS("hugetlbfs"); static struct vfsmount *hugetlbfs_vfsmount[HUGE_MAX_HSTATE]; diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index 67ce52507d7d..a67f16e846a2 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -1556,6 +1556,7 @@ static struct file_system_type iso9660_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("iso9660"); static int __init init_iso9660_fs(void) { @@ -1593,5 +1594,3 @@ static void __exit exit_iso9660_fs(void) module_init(init_iso9660_fs) module_exit(exit_iso9660_fs) MODULE_LICENSE("GPL"); -/* Actual filesystem name is iso9660, as requested in filesystems.c */ -MODULE_ALIAS("iso9660"); diff --git a/fs/jffs2/super.c b/fs/jffs2/super.c index d3d8799e2187..0defb1cc2a35 100644 --- a/fs/jffs2/super.c +++ b/fs/jffs2/super.c @@ -356,6 +356,7 @@ static struct file_system_type jffs2_fs_type = { .mount = jffs2_mount, .kill_sb = jffs2_kill_sb, }; +MODULE_ALIAS_FS("jffs2"); static int __init init_jffs2_fs(void) { diff --git a/fs/jfs/super.c b/fs/jfs/super.c index 060ba638becb..2003e830ed1c 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -833,6 +833,7 @@ static struct file_system_type jfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("jfs"); static void init_once(void *foo) { diff --git a/fs/logfs/super.c b/fs/logfs/super.c index 345c24b8a6f8..54360293bcb5 100644 --- a/fs/logfs/super.c +++ b/fs/logfs/super.c @@ -608,6 +608,7 @@ static struct file_system_type logfs_fs_type = { .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("logfs"); static int __init logfs_init(void) { diff --git a/fs/minix/inode.c b/fs/minix/inode.c index 99541cceb584..df122496f328 100644 --- a/fs/minix/inode.c +++ b/fs/minix/inode.c @@ -660,6 +660,7 @@ static struct file_system_type minix_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("minix"); static int __init init_minix_fs(void) { diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c index 7dafd6899a62..26910c8154da 100644 --- a/fs/ncpfs/inode.c +++ b/fs/ncpfs/inode.c @@ -1051,6 +1051,7 @@ static struct file_system_type ncp_fs_type = { .kill_sb = kill_anon_super, .fs_flags = FS_BINARY_MOUNTDATA, }; +MODULE_ALIAS_FS("ncpfs"); static int __init init_ncp_fs(void) { diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 17b32b722457..95cdcb208dfb 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -294,6 +294,7 @@ struct file_system_type nfs_fs_type = { .kill_sb = nfs_kill_super, .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; +MODULE_ALIAS_FS("nfs"); EXPORT_SYMBOL_GPL(nfs_fs_type); struct file_system_type nfs_xdev_fs_type = { @@ -333,6 +334,7 @@ struct file_system_type nfs4_fs_type = { .kill_sb = nfs_kill_super, .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; +MODULE_ALIAS_FS("nfs4"); EXPORT_SYMBOL_GPL(nfs4_fs_type); static int __init register_nfs4_fs(void) @@ -2717,6 +2719,5 @@ module_param(send_implementation_id, ushort, 0644); MODULE_PARM_DESC(send_implementation_id, "Send implementation ID with NFSv4.1 exchange_id"); MODULE_PARM_DESC(nfs4_unique_id, "nfs_client_id4 uniquifier string"); -MODULE_ALIAS("nfs4"); #endif /* CONFIG_NFS_V4 */ diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 13a21c8fca49..f33455b4d957 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1090,6 +1090,7 @@ static struct file_system_type nfsd_fs_type = { .mount = nfsd_mount, .kill_sb = nfsd_umount, }; +MODULE_ALIAS_FS("nfsd"); #ifdef CONFIG_PROC_FS static int create_proc_exports_entry(void) diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 3c991dc84f2f..c7d1f9f18b09 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -1361,6 +1361,7 @@ struct file_system_type nilfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("nilfs2"); static void nilfs_inode_init_once(void *obj) { diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c index 4a8289f8b16c..82650d52d916 100644 --- a/fs/ntfs/super.c +++ b/fs/ntfs/super.c @@ -3079,6 +3079,7 @@ static struct file_system_type ntfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("ntfs"); /* Stable names for the slab caches. */ static const char ntfs_index_ctx_cache_name[] = "ntfs_index_ctx_cache"; diff --git a/fs/ocfs2/dlmfs/dlmfs.c b/fs/ocfs2/dlmfs/dlmfs.c index 4c5fc8d77dc2..12bafb7265ce 100644 --- a/fs/ocfs2/dlmfs/dlmfs.c +++ b/fs/ocfs2/dlmfs/dlmfs.c @@ -640,6 +640,7 @@ static struct file_system_type dlmfs_fs_type = { .mount = dlmfs_mount, .kill_sb = kill_litter_super, }; +MODULE_ALIAS_FS("ocfs2_dlmfs"); static int __init init_dlmfs_fs(void) { diff --git a/fs/omfs/inode.c b/fs/omfs/inode.c index 25d715c7c87a..d8b0afde2179 100644 --- a/fs/omfs/inode.c +++ b/fs/omfs/inode.c @@ -572,6 +572,7 @@ static struct file_system_type omfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("omfs"); static int __init init_omfs_fs(void) { diff --git a/fs/openpromfs/inode.c b/fs/openpromfs/inode.c index ae47fa7efb9d..75885ffde44e 100644 --- a/fs/openpromfs/inode.c +++ b/fs/openpromfs/inode.c @@ -432,6 +432,7 @@ static struct file_system_type openprom_fs_type = { .mount = openprom_mount, .kill_sb = kill_anon_super, }; +MODULE_ALIAS_FS("openpromfs"); static void op_inode_init_once(void *data) { diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c index 43098bb5723a..2e8caa62da78 100644 --- a/fs/qnx4/inode.c +++ b/fs/qnx4/inode.c @@ -412,6 +412,7 @@ static struct file_system_type qnx4_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("qnx4"); static int __init init_qnx4_fs(void) { diff --git a/fs/qnx6/inode.c b/fs/qnx6/inode.c index 57199a52a351..8d941edfefa1 100644 --- a/fs/qnx6/inode.c +++ b/fs/qnx6/inode.c @@ -672,6 +672,7 @@ static struct file_system_type qnx6_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("qnx6"); static int __init init_qnx6_fs(void) { diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 418bdc3a57da..194113b1b11b 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -2434,6 +2434,7 @@ struct file_system_type reiserfs_fs_type = { .kill_sb = reiserfs_kill_sb, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("reiserfs"); MODULE_DESCRIPTION("ReiserFS journaled filesystem"); MODULE_AUTHOR("Hans Reiser "); diff --git a/fs/romfs/super.c b/fs/romfs/super.c index 7e8d3a80bdab..15cbc41ee365 100644 --- a/fs/romfs/super.c +++ b/fs/romfs/super.c @@ -599,6 +599,7 @@ static struct file_system_type romfs_fs_type = { .kill_sb = romfs_kill_sb, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("romfs"); /* * inode storage initialiser diff --git a/fs/sysv/super.c b/fs/sysv/super.c index a38e87bdd78d..a39938b1feea 100644 --- a/fs/sysv/super.c +++ b/fs/sysv/super.c @@ -545,6 +545,7 @@ static struct file_system_type sysv_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("sysv"); static struct file_system_type v7_fs_type = { .owner = THIS_MODULE, @@ -553,6 +554,7 @@ static struct file_system_type v7_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("v7"); static int __init init_sysv_fs(void) { @@ -586,5 +588,4 @@ static void __exit exit_sysv_fs(void) module_init(init_sysv_fs) module_exit(exit_sysv_fs) -MODULE_ALIAS("v7"); MODULE_LICENSE("GPL"); diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index ddc0f6ae65e9..ac838b844936 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -2174,6 +2174,7 @@ static struct file_system_type ubifs_fs_type = { .mount = ubifs_mount, .kill_sb = kill_ubifs_super, }; +MODULE_ALIAS_FS("ubifs"); /* * Inode slab cache constructor. diff --git a/fs/ufs/super.c b/fs/ufs/super.c index dc8e3a861d0f..329f2f53b7ed 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -1500,6 +1500,7 @@ static struct file_system_type ufs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("ufs"); static int __init init_ufs_fs(void) { diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index c407121873b4..ea341cea68cb 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -1561,6 +1561,7 @@ static struct file_system_type xfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("xfs"); STATIC int __init xfs_init_zones(void) -- cgit From 357b66fdc8ad4cea6e6336956a70742f961f0a4d Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Mon, 4 Mar 2013 00:34:34 -0500 Subject: ext4: ext4_split_extent should take care of extent zeroout When ext4_split_extent_at() ends up doing zeroout & conversion to initialized instead of split & conversion, ext4_split_extent() gets confused and can wrongly mark the extent back as uninitialized resulting in end IO code getting confused from large unwritten extents and may result in data loss. The example of problematic behavior is: lblk len lblk len ext4_split_extent() (ex=[1000,30,uninit], map=[1010,10]) ext4_split_extent_at() (split [1000,30,uninit] at 1020) ext4_ext_insert_extent() -> ENOSPC ext4_ext_zeroout() -> extent [1000,30] is now initialized ext4_split_extent_at() (split [1000,30,init] at 1010, MARK_UNINIT1 | MARK_UNINIT2) -> extent is split and parts marked as uninitialized Fix the problem by rechecking extent type after the first ext4_split_extent_at() returns. None of split_flags can not be applied to initialized extent so this patch also add BUG_ON to prevent similar issues in future. TESTCASE: https://github.com/dmonakhov/xfstests/commit/b8a55eb5ce28c6ff29e620ab090902fcd5833597 Signed-off-by: Dmitry Monakhov Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/extents.c | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 372b2cbee07e..bef194a14437 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -2943,6 +2943,10 @@ static int ext4_split_extent_at(handle_t *handle, newblock = split - ee_block + ext4_ext_pblock(ex); BUG_ON(split < ee_block || split >= (ee_block + ee_len)); + BUG_ON(!ext4_ext_is_uninitialized(ex) && + split_flag & (EXT4_EXT_MAY_ZEROOUT | + EXT4_EXT_MARK_UNINIT1 | + EXT4_EXT_MARK_UNINIT2)); err = ext4_ext_get_access(handle, inode, path + depth); if (err) @@ -3061,19 +3065,26 @@ static int ext4_split_extent(handle_t *handle, if (err) goto out; } - + /* + * Update path is required because previous ext4_split_extent_at() may + * result in split of original leaf or extent zeroout. + */ ext4_ext_drop_refs(path); path = ext4_ext_find_extent(inode, map->m_lblk, path); if (IS_ERR(path)) return PTR_ERR(path); + depth = ext_depth(inode); + ex = path[depth].p_ext; + uninitialized = ext4_ext_is_uninitialized(ex); + split_flag1 = 0; if (map->m_lblk >= ee_block) { - split_flag1 = split_flag & (EXT4_EXT_MAY_ZEROOUT | - EXT4_EXT_DATA_VALID2); - if (uninitialized) + split_flag1 = split_flag & EXT4_EXT_DATA_VALID2; + if (uninitialized) { split_flag1 |= EXT4_EXT_MARK_UNINIT1; - if (split_flag & EXT4_EXT_MARK_UNINIT2) - split_flag1 |= EXT4_EXT_MARK_UNINIT2; + split_flag1 |= split_flag & (EXT4_EXT_MAY_ZEROOUT | + EXT4_EXT_MARK_UNINIT2); + } err = ext4_split_extent_at(handle, inode, path, map->m_lblk, split_flag1, flags); if (err) -- cgit From ec22ba8edb507395c95fbc617eea26a6b2d98797 Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Mon, 4 Mar 2013 00:36:06 -0500 Subject: ext4: disable merging of uninitialized extents Derived from Jan's patch:http://permalink.gmane.org/gmane.comp.file-systems.ext4/36470 Merging of uninitialized extents creates all sorts of interesting race possibilities when writeback / DIO races with fallocate. Thus ext4_convert_unwritten_extents_endio() has to deal with a case where extent to be converted needs to be split out first. That isn't nice for two reasons: 1) It may need allocation of extent tree block so ENOSPC is possible. 2) It complicates end_io handling code So we disable merging of uninitialized extents which allows us to simplify the code. Extents will get merged after they are converted to initialized ones. Signed-off-by: Dmitry Monakhov Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/extents.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index bef194a14437..60818ed1f6a9 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -1584,10 +1584,12 @@ ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1, unsigned short ext1_ee_len, ext2_ee_len, max_len; /* - * Make sure that either both extents are uninitialized, or - * both are _not_. + * Make sure that both extents are initialized. We don't merge + * uninitialized extents so that we can be sure that end_io code has + * the extent that was written properly split out and conversion to + * initialized is trivial. */ - if (ext4_ext_is_uninitialized(ex1) ^ ext4_ext_is_uninitialized(ex2)) + if (ext4_ext_is_uninitialized(ex1) || ext4_ext_is_uninitialized(ex2)) return 0; if (ext4_ext_is_uninitialized(ex1)) -- cgit From ff95ec22cd7faa0d8b58dcc4207f21502df7b00b Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Mon, 4 Mar 2013 00:41:05 -0500 Subject: ext4: add warning to ext4_convert_unwritten_extents_endio Splitting extents inside endio is a bad thing, but unfortunately it is still possible. In fact we are pretty close to the moment when all related issues will be fixed. Let's warn developer if it still the case. Signed-off-by: Dmitry Monakhov Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/extents.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 60818ed1f6a9..265cb0e50c51 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3387,8 +3387,19 @@ static int ext4_convert_unwritten_extents_endio(handle_t *handle, "block %llu, max_blocks %u\n", inode->i_ino, (unsigned long long)ee_block, ee_len); - /* If extent is larger than requested then split is required */ + /* If extent is larger than requested it is a clear sign that we still + * have some extent state machine issues left. So extent_split is still + * required. + * TODO: Once all related issues will be fixed this situation should be + * illegal. + */ if (ee_block != map->m_lblk || ee_len > map->m_len) { +#ifdef EXT4_DEBUG + ext4_warning("Inode (%ld) finished: extent logical block %llu," + " len %u; IO logical block %llu, len %u\n", + inode->i_ino, (unsigned long long)ee_block, ee_len, + (unsigned long long)map->m_lblk, map->m_len); +#endif err = ext4_split_unwritten_extents(handle, inode, map, path, EXT4_GET_BLOCKS_CONVERT); if (err < 0) -- cgit From de99fcce1da7933a90198b80a2e896754ea3bdc8 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 4 Mar 2013 00:43:32 -0500 Subject: ext4: remove unnecessary wait for extent conversion in ext4_fallocate() Now that we don't merge uninitialized extents anymore, ext4_fallocate() is free to operate on the inode while there are still some extent conversions pending - it won't disturb them in any way. Reviewed-by: Zheng Liu Reviewed-by: Dmitry Monakhov Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/extents.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 265cb0e50c51..25c86aaa38d6 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4392,8 +4392,6 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) if (len <= EXT_UNINIT_MAX_LEN << blkbits) flags |= EXT4_GET_BLOCKS_NO_NORMALIZE; - /* Prevent race condition between unwritten */ - ext4_flush_unwritten_io(inode); retry: while (ret >= 0 && ret < max_blocks) { map.m_lblk = map.m_lblk + ret; -- cgit From 6ca470d7b5e7639b7925b3202e796282703b6d5d Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Mon, 4 Mar 2013 00:50:47 -0500 Subject: ext4: invalidate extent status tree during extent migration mext_replace_branches() will change inode's extents layout so we have to drop corresponding cache. TESTCASE: 301'th xfstest was not yet accepted to official xfstest's branch and can be found here: https://github.com/dmonakhov/xfstests/commit/7b7efeee30a41109201e2040034e71db9b66ddc0 Signed-off-by: Dmitry Monakhov Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara --- fs/ext4/move_extent.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index d78c33eed7e5..c1f15b203e98 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -666,6 +666,14 @@ mext_replace_branches(handle_t *handle, struct inode *orig_inode, int replaced_count = 0; int dext_alen; + *err = ext4_es_remove_extent(orig_inode, from, count); + if (*err) + goto out; + + *err = ext4_es_remove_extent(donor_inode, from, count); + if (*err) + goto out; + /* Get the original extent for the block "orig_off" */ *err = get_ext_path(orig_inode, orig_off, &orig_path); if (*err) -- cgit From 290502bee239062499297916bb7d21d205e99d62 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 28 Feb 2013 00:39:37 -0800 Subject: eCryptfs: allow userspace messaging to be disabled When the userspace messaging (for the less common case of userspace key wrap/unwrap via ecryptfsd) is not needed, allow eCryptfs to build with it removed. This saves on kernel code size and reduces potential attack surface by removing the /dev/ecryptfs node. Signed-off-by: Kees Cook Signed-off-by: Tyler Hicks --- fs/ecryptfs/Kconfig | 8 ++++++++ fs/ecryptfs/Makefile | 7 +++++-- fs/ecryptfs/ecryptfs_kernel.h | 40 ++++++++++++++++++++++++++++++++++++++-- fs/ecryptfs/keystore.c | 4 ++-- 4 files changed, 53 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/ecryptfs/Kconfig b/fs/ecryptfs/Kconfig index cc16562654de..1f63120b669a 100644 --- a/fs/ecryptfs/Kconfig +++ b/fs/ecryptfs/Kconfig @@ -12,3 +12,11 @@ config ECRYPT_FS To compile this file system support as a module, choose M here: the module will be called ecryptfs. + +config ECRYPT_FS_MESSAGING + bool "Enable notifications for userspace key wrap/unwrap" + depends on ECRYPT_FS + help + Enables the /dev/ecryptfs entry for use by ecryptfsd. This allows + for userspace to wrap/unwrap file encryption keys by other + backends, like OpenSSL. diff --git a/fs/ecryptfs/Makefile b/fs/ecryptfs/Makefile index 2cc9ee4ad2eb..49678a69947d 100644 --- a/fs/ecryptfs/Makefile +++ b/fs/ecryptfs/Makefile @@ -1,7 +1,10 @@ # -# Makefile for the Linux 2.6 eCryptfs +# Makefile for the Linux eCryptfs # obj-$(CONFIG_ECRYPT_FS) += ecryptfs.o -ecryptfs-objs := dentry.o file.o inode.o main.o super.o mmap.o read_write.o crypto.o keystore.o messaging.o miscdev.o kthread.o debug.o +ecryptfs-y := dentry.o file.o inode.o main.o super.o mmap.o read_write.o \ + crypto.o keystore.o kthread.o debug.o + +ecryptfs-$(CONFIG_ECRYPT_FS_MESSAGING) += messaging.o miscdev.o diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index cfb4b9fed520..a9df69efadc2 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -172,6 +172,19 @@ ecryptfs_get_key_payload_data(struct key *key) #define ECRYPTFS_FNEK_ENCRYPTED_FILENAME_PREFIX_SIZE 24 #define ECRYPTFS_ENCRYPTED_DENTRY_NAME_LEN (18 + 1 + 4 + 1 + 32) +#ifdef CONFIG_ECRYPT_FS_MESSAGING +# define ECRYPTFS_VERSIONING_MASK_MESSAGING (ECRYPTFS_VERSIONING_DEVMISC \ + | ECRYPTFS_VERSIONING_PUBKEY) +#else +# define ECRYPTFS_VERSIONING_MASK_MESSAGING 0 +#endif + +#define ECRYPTFS_VERSIONING_MASK (ECRYPTFS_VERSIONING_PASSPHRASE \ + | ECRYPTFS_VERSIONING_PLAINTEXT_PASSTHROUGH \ + | ECRYPTFS_VERSIONING_XATTR \ + | ECRYPTFS_VERSIONING_MULTKEY \ + | ECRYPTFS_VERSIONING_MASK_MESSAGING \ + | ECRYPTFS_VERSIONING_FILENAME_ENCRYPTION) struct ecryptfs_key_sig { struct list_head crypt_stat_list; char keysig[ECRYPTFS_SIG_SIZE_HEX + 1]; @@ -399,7 +412,9 @@ struct ecryptfs_daemon { struct hlist_node euid_chain; }; +#ifdef CONFIG_ECRYPT_FS_MESSAGING extern struct mutex ecryptfs_daemon_hash_mux; +#endif static inline size_t ecryptfs_lower_header_size(struct ecryptfs_crypt_stat *crypt_stat) @@ -604,6 +619,7 @@ int ecryptfs_setxattr(struct dentry *dentry, const char *name, const void *value, size_t size, int flags); int ecryptfs_read_xattr_region(char *page_virt, struct inode *ecryptfs_inode); +#ifdef CONFIG_ECRYPT_FS_MESSAGING int ecryptfs_process_response(struct ecryptfs_daemon *daemon, struct ecryptfs_message *msg, u32 seq); int ecryptfs_send_message(char *data, int data_len, @@ -612,6 +628,24 @@ int ecryptfs_wait_for_response(struct ecryptfs_msg_ctx *msg_ctx, struct ecryptfs_message **emsg); int ecryptfs_init_messaging(void); void ecryptfs_release_messaging(void); +#else +static inline int ecryptfs_init_messaging(void) +{ + return 0; +} +static inline void ecryptfs_release_messaging(void) +{ } +static inline int ecryptfs_send_message(char *data, int data_len, + struct ecryptfs_msg_ctx **msg_ctx) +{ + return -ENOTCONN; +} +static inline int ecryptfs_wait_for_response(struct ecryptfs_msg_ctx *msg_ctx, + struct ecryptfs_message **emsg) +{ + return -ENOMSG; +} +#endif void ecryptfs_write_header_metadata(char *virt, @@ -649,12 +683,11 @@ int ecryptfs_read_lower_page_segment(struct page *page_for_ecryptfs, size_t offset_in_page, size_t size, struct inode *ecryptfs_inode); struct page *ecryptfs_get_locked_page(struct inode *inode, loff_t index); -int ecryptfs_exorcise_daemon(struct ecryptfs_daemon *daemon); -int ecryptfs_find_daemon_by_euid(struct ecryptfs_daemon **daemon); int ecryptfs_parse_packet_length(unsigned char *data, size_t *size, size_t *length_size); int ecryptfs_write_packet_length(char *dest, size_t size, size_t *packet_size_length); +#ifdef CONFIG_ECRYPT_FS_MESSAGING int ecryptfs_init_ecryptfs_miscdev(void); void ecryptfs_destroy_ecryptfs_miscdev(void); int ecryptfs_send_miscdev(char *data, size_t data_size, @@ -663,6 +696,9 @@ int ecryptfs_send_miscdev(char *data, size_t data_size, void ecryptfs_msg_ctx_alloc_to_free(struct ecryptfs_msg_ctx *msg_ctx); int ecryptfs_spawn_daemon(struct ecryptfs_daemon **daemon, struct file *file); +int ecryptfs_exorcise_daemon(struct ecryptfs_daemon *daemon); +int ecryptfs_find_daemon_by_euid(struct ecryptfs_daemon **daemon); +#endif int ecryptfs_init_kthread(void); void ecryptfs_destroy_kthread(void); int ecryptfs_privileged_open(struct file **lower_file, diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index 5aceff202dc0..7d52806c2119 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -1168,7 +1168,7 @@ decrypt_pki_encrypted_session_key(struct ecryptfs_auth_tok *auth_tok, rc = ecryptfs_send_message(payload, payload_len, &msg_ctx); if (rc) { ecryptfs_printk(KERN_ERR, "Error sending message to " - "ecryptfsd\n"); + "ecryptfsd: %d\n", rc); goto out; } rc = ecryptfs_wait_for_response(msg_ctx, &msg); @@ -1988,7 +1988,7 @@ pki_encrypt_session_key(struct key *auth_tok_key, rc = ecryptfs_send_message(payload, payload_len, &msg_ctx); if (rc) { ecryptfs_printk(KERN_ERR, "Error sending message to " - "ecryptfsd\n"); + "ecryptfsd: %d\n", rc); goto out; } rc = ecryptfs_wait_for_response(msg_ctx, &msg); -- cgit From 9bf7a4890518186238d2579be16ecc5190a707c0 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 1 Mar 2013 13:35:47 -0500 Subject: Btrfs: use set_nlink if our i_nlink is 0 We need to inc the nlink of deleted entries when running replay so we can do the unlink on the fs_root and get everything cleaned up and then have the orphan cleanup do the right thing. The problem is inc_nlink complains about this, even thought it still does the right thing. So use set_nlink() if our i_nlink is 0 to keep users from seeing the warnings during log replay. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/tree-log.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index c7ef569eb22a..451fad96ecd1 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -1382,7 +1382,10 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans, btrfs_release_path(path); if (ret == 0) { - btrfs_inc_nlink(inode); + if (!inode->i_nlink) + set_nlink(inode, 1); + else + btrfs_inc_nlink(inode); ret = btrfs_update_inode(trans, root, inode); } else if (ret == -EEXIST) { ret = 0; -- cgit From aec8030a8745221c8658f2033b22c98528897b13 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Mon, 4 Mar 2013 09:44:29 +0000 Subject: Btrfs: fix wrong handle at error path of create_snapshot() when the commit fails There are several bugs at error path of create_snapshot() when the transaction commitment failed. - access the freed transaction handler. At the end of the transaction commitment, the transaction handler was freed, so we should not access it after the transaction commitment. - we were not aware of the error which happened during the snapshot creation if we submitted a async transaction commitment. - pending snapshot access vs pending snapshot free. when something wrong happened after we submitted a async transaction commitment, the transaction committer would cleanup the pending snapshots and free them. But the snapshot creators were not aware of it, they would access the freed pending snapshots. This patch fixes the above problems by: - remove the dangerous code that accessed the freed handler - assign ->error if the error happens during the snapshot creation - the transaction committer doesn't free the pending snapshots, just assigns the error number and evicts them before we unblock the transaction. Reported-by: Dan Carpenter Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 16 ++++++-------- fs/btrfs/ioctl.c | 6 +----- fs/btrfs/transaction.c | 58 ++++++++++++++++++++++++++++---------------------- 3 files changed, 41 insertions(+), 39 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 02369a3c162e..7d84651e850b 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -62,7 +62,7 @@ static void btrfs_destroy_ordered_operations(struct btrfs_transaction *t, static void btrfs_destroy_ordered_extents(struct btrfs_root *root); static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, struct btrfs_root *root); -static void btrfs_destroy_pending_snapshots(struct btrfs_transaction *t); +static void btrfs_evict_pending_snapshots(struct btrfs_transaction *t); static void btrfs_destroy_delalloc_inodes(struct btrfs_root *root); static int btrfs_destroy_marked_extents(struct btrfs_root *root, struct extent_io_tree *dirty_pages, @@ -3687,7 +3687,7 @@ int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, return ret; } -static void btrfs_destroy_pending_snapshots(struct btrfs_transaction *t) +static void btrfs_evict_pending_snapshots(struct btrfs_transaction *t) { struct btrfs_pending_snapshot *snapshot; struct list_head splice; @@ -3700,10 +3700,8 @@ static void btrfs_destroy_pending_snapshots(struct btrfs_transaction *t) snapshot = list_entry(splice.next, struct btrfs_pending_snapshot, list); - + snapshot->error = -ECANCELED; list_del_init(&snapshot->list); - - kfree(snapshot); } } @@ -3840,6 +3838,8 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans, cur_trans->blocked = 1; wake_up(&root->fs_info->transaction_blocked_wait); + btrfs_evict_pending_snapshots(cur_trans); + cur_trans->blocked = 0; wake_up(&root->fs_info->transaction_wait); @@ -3849,8 +3849,6 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans, btrfs_destroy_delayed_inodes(root); btrfs_assert_delayed_root_empty(root); - btrfs_destroy_pending_snapshots(cur_trans); - btrfs_destroy_marked_extents(root, &cur_trans->dirty_pages, EXTENT_DIRTY); btrfs_destroy_pinned_extent(root, @@ -3894,6 +3892,8 @@ int btrfs_cleanup_transaction(struct btrfs_root *root) if (waitqueue_active(&root->fs_info->transaction_blocked_wait)) wake_up(&root->fs_info->transaction_blocked_wait); + btrfs_evict_pending_snapshots(t); + t->blocked = 0; smp_mb(); if (waitqueue_active(&root->fs_info->transaction_wait)) @@ -3907,8 +3907,6 @@ int btrfs_cleanup_transaction(struct btrfs_root *root) btrfs_destroy_delayed_inodes(root); btrfs_assert_delayed_root_empty(root); - btrfs_destroy_pending_snapshots(t); - btrfs_destroy_delalloc_inodes(root); spin_lock(&root->fs_info->trans_lock); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index b908960c9746..94c0e42dfa1e 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -596,12 +596,8 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir, ret = btrfs_commit_transaction(trans, root->fs_info->extent_root); } - if (ret) { - /* cleanup_transaction has freed this for us */ - if (trans->aborted) - pending_snapshot = NULL; + if (ret) goto fail; - } ret = pending_snapshot->error; if (ret) diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index f11c2e0a3746..d8fce6fe9cf8 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1053,7 +1053,12 @@ int btrfs_defrag_root(struct btrfs_root *root) /* * new snapshots need to be created at a very specific time in the - * transaction commit. This does the actual creation + * transaction commit. This does the actual creation. + * + * Note: + * If the error which may affect the commitment of the current transaction + * happens, we should return the error number. If the error which just affect + * the creation of the pending snapshots, just return 0. */ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info, @@ -1072,7 +1077,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, struct extent_buffer *tmp; struct extent_buffer *old; struct timespec cur_time = CURRENT_TIME; - int ret; + int ret = 0; u64 to_reserve = 0; u64 index = 0; u64 objectid; @@ -1081,40 +1086,36 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, path = btrfs_alloc_path(); if (!path) { - ret = pending->error = -ENOMEM; - return ret; + pending->error = -ENOMEM; + return 0; } new_root_item = kmalloc(sizeof(*new_root_item), GFP_NOFS); if (!new_root_item) { - ret = pending->error = -ENOMEM; + pending->error = -ENOMEM; goto root_item_alloc_fail; } - ret = btrfs_find_free_objectid(tree_root, &objectid); - if (ret) { - pending->error = ret; + pending->error = btrfs_find_free_objectid(tree_root, &objectid); + if (pending->error) goto no_free_objectid; - } btrfs_reloc_pre_snapshot(trans, pending, &to_reserve); if (to_reserve > 0) { - ret = btrfs_block_rsv_add(root, &pending->block_rsv, - to_reserve, - BTRFS_RESERVE_NO_FLUSH); - if (ret) { - pending->error = ret; + pending->error = btrfs_block_rsv_add(root, + &pending->block_rsv, + to_reserve, + BTRFS_RESERVE_NO_FLUSH); + if (pending->error) goto no_free_objectid; - } } - ret = btrfs_qgroup_inherit(trans, fs_info, root->root_key.objectid, - objectid, pending->inherit); - if (ret) { - pending->error = ret; + pending->error = btrfs_qgroup_inherit(trans, fs_info, + root->root_key.objectid, + objectid, pending->inherit); + if (pending->error) goto no_free_objectid; - } key.objectid = objectid; key.offset = (u64)-1; @@ -1142,7 +1143,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, dentry->d_name.len, 0); if (dir_item != NULL && !IS_ERR(dir_item)) { pending->error = -EEXIST; - goto fail; + goto dir_item_existed; } else if (IS_ERR(dir_item)) { ret = PTR_ERR(dir_item); btrfs_abort_transaction(trans, root, ret); @@ -1273,6 +1274,8 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, if (ret) btrfs_abort_transaction(trans, root, ret); fail: + pending->error = ret; +dir_item_existed: trans->block_rsv = rsv; trans->bytes_reserved = 0; no_free_objectid: @@ -1288,12 +1291,17 @@ root_item_alloc_fail: static noinline int create_pending_snapshots(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info) { - struct btrfs_pending_snapshot *pending; + struct btrfs_pending_snapshot *pending, *next; struct list_head *head = &trans->transaction->pending_snapshots; + int ret = 0; - list_for_each_entry(pending, head, list) - create_pending_snapshot(trans, fs_info, pending); - return 0; + list_for_each_entry_safe(pending, next, head, list) { + list_del(&pending->list); + ret = create_pending_snapshot(trans, fs_info, pending); + if (ret) + break; + } + return ret; } static void update_super_roots(struct btrfs_root *root) -- cgit From 00d71c9c17b1fd28fa54f323a29a0e23c6d3de40 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Mon, 4 Mar 2013 09:45:06 +0000 Subject: Btrfs: fix unclosed transaction handler when the async transaction commitment fails If the async transaction commitment failed, we need close the current transaction handler, or the current transaction will be blocked to commit because of this orphan handler. We fix the problem by doing sync transaction commitment, that is to invoke btrfs_commit_transaction(). Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 94c0e42dfa1e..3fdfabcc1aaa 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -527,6 +527,8 @@ fail: if (async_transid) { *async_transid = trans->transid; err = btrfs_commit_transaction_async(trans, root, 1); + if (err) + err = btrfs_commit_transaction(trans, root); } else { err = btrfs_commit_transaction(trans, root); } @@ -592,6 +594,8 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir, *async_transid = trans->transid; ret = btrfs_commit_transaction_async(trans, root->fs_info->extent_root, 1); + if (ret) + ret = btrfs_commit_transaction(trans, root); } else { ret = btrfs_commit_transaction(trans, root->fs_info->extent_root); -- cgit From 8f71f3e0e4fb5a2445fb93d3057a33aefc4aa30d Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 4 Mar 2013 16:25:36 +0000 Subject: Btrfs: check for NULL pointer in updating reloc roots Add a check for NULL pointer to avoid invalid reference. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/relocation.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index ba5a3210da9a..16e0c6fbdbed 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -1269,6 +1269,8 @@ static int __update_reloc_root(struct btrfs_root *root, int del) } spin_unlock(&rc->reloc_root_tree.lock); + if (!node) + return 0; BUG_ON((struct btrfs_root *)node->data != root); if (!del) { -- cgit From aca1bba6f9bc550a33312b28e98b3de5ddb3bb15 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 4 Mar 2013 16:25:37 +0000 Subject: Btrfs: build up error handling for merge_reloc_roots We first use btrfs_std_error hook to replace with BUG_ON, and we also need to cleanup what is left, including reloc roots rbtree and reloc roots list. Here we use a helper function to cleanup both rbtree and list, and since this function can also be used in the balance recover path, we also make the change as well to keep code simple. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/relocation.c | 47 +++++++++++++++++++++++++++++++++++------------ 1 file changed, 35 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 16e0c6fbdbed..0f001c14eaf4 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -2239,6 +2239,21 @@ again: return err; } +static noinline_for_stack +void free_reloc_roots(struct list_head *list) +{ + struct btrfs_root *reloc_root; + + while (!list_empty(list)) { + reloc_root = list_entry(list->next, struct btrfs_root, + root_list); + __update_reloc_root(reloc_root, 1); + free_extent_buffer(reloc_root->node); + free_extent_buffer(reloc_root->commit_root); + kfree(reloc_root); + } +} + static noinline_for_stack int merge_reloc_roots(struct reloc_control *rc) { @@ -2246,7 +2261,7 @@ int merge_reloc_roots(struct reloc_control *rc) struct btrfs_root *reloc_root; LIST_HEAD(reloc_roots); int found = 0; - int ret; + int ret = 0; again: root = rc->extent_root; @@ -2272,20 +2287,33 @@ again: BUG_ON(root->reloc_root != reloc_root); ret = merge_reloc_root(rc, root); - BUG_ON(ret); + if (ret) + goto out; } else { list_del_init(&reloc_root->root_list); } ret = btrfs_drop_snapshot(reloc_root, rc->block_rsv, 0, 1); - BUG_ON(ret < 0); + if (ret < 0) { + if (list_empty(&reloc_root->root_list)) + list_add_tail(&reloc_root->root_list, + &reloc_roots); + goto out; + } } if (found) { found = 0; goto again; } +out: + if (ret) { + btrfs_std_error(root->fs_info, ret); + if (!list_empty(&reloc_roots)) + free_reloc_roots(&reloc_roots); + } + BUG_ON(!RB_EMPTY_ROOT(&rc->reloc_root_tree.rb_root)); - return 0; + return ret; } static void free_block_list(struct rb_root *blocks) @@ -4266,14 +4294,9 @@ int btrfs_recover_relocation(struct btrfs_root *root) out_free: kfree(rc); out: - while (!list_empty(&reloc_roots)) { - reloc_root = list_entry(reloc_roots.next, - struct btrfs_root, root_list); - list_del(&reloc_root->root_list); - free_extent_buffer(reloc_root->node); - free_extent_buffer(reloc_root->commit_root); - kfree(reloc_root); - } + if (!list_empty(&reloc_roots)) + free_reloc_roots(&reloc_roots); + btrfs_free_path(path); if (err == 0) { -- cgit From e1a1267054ea9dcafd01636e52d441f809efad5e Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 4 Mar 2013 16:25:38 +0000 Subject: Btrfs: free all recorded tree blocks on error We've missed the 'free blocks' part on ENOMEM error. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/relocation.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 0f001c14eaf4..7d1654cf9d48 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -2848,8 +2848,10 @@ int relocate_tree_blocks(struct btrfs_trans_handle *trans, int err = 0; path = btrfs_alloc_path(); - if (!path) - return -ENOMEM; + if (!path) { + err = -ENOMEM; + goto out_path; + } rb_node = rb_first(blocks); while (rb_node) { @@ -2888,10 +2890,11 @@ int relocate_tree_blocks(struct btrfs_trans_handle *trans, rb_node = rb_next(rb_node); } out: - free_block_list(blocks); err = finish_pending_nodes(trans, rc, path, err); btrfs_free_path(path); +out_path: + free_block_list(blocks); return err; } -- cgit From 288189471d7e12fb22c37870e151833f438deea8 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 4 Mar 2013 16:25:39 +0000 Subject: Btrfs: do not BUG_ON in prepare_to_reloc We can bail out from here gracefully instead of a cold BUG_ON. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/relocation.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 7d1654cf9d48..9d13786eec73 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -3731,7 +3731,15 @@ int prepare_to_relocate(struct reloc_control *rc) set_reloc_control(rc); trans = btrfs_join_transaction(rc->extent_root); - BUG_ON(IS_ERR(trans)); + if (IS_ERR(trans)) { + unset_reloc_control(rc); + /* + * extent tree is not a ref_cow tree and has no reloc_root to + * cleanup. And callers are responsible to free the above + * block rsv. + */ + return PTR_ERR(trans); + } btrfs_commit_transaction(trans, rc->extent_root); return 0; } -- cgit From 0f788c58194e4ccc5b3ab23f872c5e18542335e4 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 4 Mar 2013 16:25:40 +0000 Subject: Btrfs: do not BUG_ON on aborted situation Btrfs balance can easily hit BUG_ON in these places, but we want to it bail out gracefully after we force the whole filesystem to readonly. So we use btrfs_std_error hook in place of BUG_ON. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/relocation.c | 6 +++++- fs/btrfs/volumes.c | 9 +++++++-- 2 files changed, 12 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 9d13786eec73..3ebe87977aae 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -3771,7 +3771,11 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc) while (1) { progress++; trans = btrfs_start_transaction(rc->extent_root, 0); - BUG_ON(IS_ERR(trans)); + if (IS_ERR(trans)) { + err = PTR_ERR(trans); + trans = NULL; + break; + } restart: if (update_backref_cache(trans, &rc->backref_cache)) { btrfs_end_transaction(trans, rc->extent_root); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 35bb2d4ed29f..9ff454df6756 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2379,7 +2379,11 @@ static int btrfs_relocate_chunk(struct btrfs_root *root, return ret; trans = btrfs_start_transaction(root, 0); - BUG_ON(IS_ERR(trans)); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + btrfs_std_error(root->fs_info, ret); + return ret; + } lock_chunks(root); @@ -3050,7 +3054,8 @@ static void __cancel_balance(struct btrfs_fs_info *fs_info) unset_balance_control(fs_info); ret = del_balance_item(fs_info->tree_root); - BUG_ON(ret); + if (ret) + btrfs_std_error(fs_info, ret); atomic_set(&fs_info->mutually_exclusive_operation_running, 0); } -- cgit From 66b6135b7cf741f6f44ba938b27583ea3b83bd12 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 4 Mar 2013 16:25:41 +0000 Subject: Btrfs: avoid deadlock on transaction waiting list Only let one trans handle to wait for other handles, otherwise we will get ABBA issues. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik --- fs/btrfs/transaction.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index d8fce6fe9cf8..fedede1fe178 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1457,6 +1457,13 @@ static void cleanup_transaction(struct btrfs_trans_handle *trans, btrfs_abort_transaction(trans, root, err); spin_lock(&root->fs_info->trans_lock); + + if (list_empty(&cur_trans->list)) { + spin_unlock(&root->fs_info->trans_lock); + btrfs_end_transaction(trans, root); + return; + } + list_del_init(&cur_trans->list); if (cur_trans == root->fs_info->running_transaction) { root->fs_info->trans_no_join = 1; -- cgit From 9b53157aac7366cea413ee29b629f83225829e87 Mon Sep 17 00:00:00 2001 From: Stefan Behrens Date: Mon, 4 Mar 2013 17:28:38 +0000 Subject: Btrfs: allow running defrag in parallel to administrative tasks Commit 5ac00add added a testnset mutex and code that disallows running administrative tasks in parallel. It is prevented that the device add/delete/balance/replace/resize operations are started in parallel. By mistake, the defragmentation operation was included in the check for mutually exclusiveness as well. This is fixed with this commit. Signed-off-by: Stefan Behrens Signed-off-by: Josef Bacik --- fs/btrfs/ioctl.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 3fdfabcc1aaa..898c5729e7e5 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2245,13 +2245,6 @@ static int btrfs_ioctl_defrag(struct file *file, void __user *argp) if (ret) return ret; - if (atomic_xchg(&root->fs_info->mutually_exclusive_operation_running, - 1)) { - pr_info("btrfs: dev add/delete/balance/replace/resize operation in progress\n"); - mnt_drop_write_file(file); - return -EINVAL; - } - if (btrfs_root_readonly(root)) { ret = -EROFS; goto out; @@ -2306,7 +2299,6 @@ static int btrfs_ioctl_defrag(struct file *file, void __user *argp) ret = -EINVAL; } out: - atomic_set(&root->fs_info->mutually_exclusive_operation_running, 0); mnt_drop_write_file(file); return ret; } -- cgit From 154ea2893002618bc3f9a1e2d8186c65490968b1 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 5 Mar 2013 11:11:26 -0500 Subject: Btrfs: enforce min_bytes parameter during extent allocation Commit 24542bf7ea5e4fdfdb5157ff544c093fa4dcb536 changed preallocation of extents to cap the max size we try to allocate. It's a valid change, but the extent reservation code is also used by balance, and that can't tolerate a smaller extent being allocated. __btrfs_prealloc_file_range already has a min_size parameter, which is used by relocation to request a specific extent size. This commit adds an extra check to enforce that minimum extent size. Signed-off-by: Chris Mason Reported-by: Stefan Behrens --- fs/btrfs/inode.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index ecd9c4cdb0db..13ab4de0a400 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -8502,6 +8502,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode, struct btrfs_key ins; u64 cur_offset = start; u64 i_size; + u64 cur_bytes; int ret = 0; bool own_trans = true; @@ -8516,8 +8517,9 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode, } } - ret = btrfs_reserve_extent(trans, root, - min(num_bytes, 256ULL * 1024 * 1024), + cur_bytes = min(num_bytes, 256ULL * 1024 * 1024); + cur_bytes = max(cur_bytes, min_size); + ret = btrfs_reserve_extent(trans, root, cur_bytes, min_size, 0, *alloc_hint, &ins, 1); if (ret) { if (own_trans) -- cgit From 72d282dc5109e5dc0d963be020604e0cc82f7ed7 Mon Sep 17 00:00:00 2001 From: Sachin Prabhu Date: Tue, 5 Mar 2013 19:25:55 +0000 Subject: cifs: Fix bug when checking error condition in cifs_rename_pending_delete() Fix check for error condition after setting attributes with CIFSSMBSetFileInfo(). Signed-off-by: Sachin Prabhu Reviewed-by: Jeff Layton Reviewed-by: Pavel Shilovsky Signed-off-by: Steve French --- fs/cifs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 83f2606c76d0..e7ae45cd94ef 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1023,7 +1023,7 @@ cifs_rename_pending_delete(const char *full_path, struct dentry *dentry, current->tgid); /* although we would like to mark the file hidden if that fails we will still try to rename it */ - if (rc != 0) + if (!rc) cifsInode->cifsAttrs = dosattr; else dosattr = origattr; /* since not able to change them */ -- cgit From c483a9841df1de327e01af7deb6ba349210e5f82 Mon Sep 17 00:00:00 2001 From: Sachin Prabhu Date: Tue, 5 Mar 2013 19:25:56 +0000 Subject: cifs: Check server capability before attempting silly rename cifs_rename_pending_delete() attempts to silly rename file using CIFSSMBRenameOpenFile(). This uses the SET_FILE_INFORMATION TRANS2 command with information level set to the passthru info-level SMB_SET_FILE_RENAME_INFORMATION. We need to check to make sure that the server support passthru info-levels before attempting the silly rename or else we will fail to rename the file. Signed-off-by: Sachin Prabhu Reviewed-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/inode.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index e7ae45cd94ef..0079696305c9 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -995,6 +995,15 @@ cifs_rename_pending_delete(const char *full_path, struct dentry *dentry, return PTR_ERR(tlink); tcon = tlink_tcon(tlink); + /* + * We cannot rename the file if the server doesn't support + * CAP_INFOLEVEL_PASSTHRU + */ + if (!(tcon->ses->capabilities & CAP_INFOLEVEL_PASSTHRU)) { + rc = -EBUSY; + goto out; + } + rc = CIFSSMBOpen(xid, tcon, full_path, FILE_OPEN, DELETE|FILE_WRITE_ATTRIBUTES, CREATE_NOT_DIR, &netfid, &oplock, NULL, cifs_sb->local_nls, -- cgit From 25189643a165a5ccad48ee3c116b55905299fd3d Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 21 Feb 2013 06:32:59 -0500 Subject: cifs: remove the sockopt= mount option ...as promised for 3.9. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 16 +--------------- 1 file changed, 1 insertion(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 54125e04fd0c..991c63c6bdd0 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -97,7 +97,7 @@ enum { Opt_user, Opt_pass, Opt_ip, Opt_unc, Opt_domain, Opt_srcaddr, Opt_prefixpath, - Opt_iocharset, Opt_sockopt, + Opt_iocharset, Opt_netbiosname, Opt_servern, Opt_ver, Opt_vers, Opt_sec, Opt_cache, @@ -202,7 +202,6 @@ static const match_table_t cifs_mount_option_tokens = { { Opt_srcaddr, "srcaddr=%s" }, { Opt_prefixpath, "prefixpath=%s" }, { Opt_iocharset, "iocharset=%s" }, - { Opt_sockopt, "sockopt=%s" }, { Opt_netbiosname, "netbiosname=%s" }, { Opt_servern, "servern=%s" }, { Opt_ver, "ver=%s" }, @@ -1752,19 +1751,6 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, */ cFYI(1, "iocharset set to %s", string); break; - case Opt_sockopt: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - - if (strnicmp(string, "TCP_NODELAY", 11) == 0) { - printk(KERN_WARNING "CIFS: the " - "sockopt=TCP_NODELAY option has been " - "deprecated and will be removed " - "in 3.9\n"); - vol->sockopt_tcp_nodelay = 1; - } - break; case Opt_netbiosname: string = match_strdup(args); if (string == NULL) -- cgit From 94e18007688a13e5da1a2f1b7f52f15cc56c9f5e Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 4 Mar 2013 15:18:25 -0500 Subject: cifs: don't try to unlock pagecache page after releasing it We had a recent fix to fix the release of pagecache pages when cifs_writev_requeue writes fail. Unfortunately, it releases the page before trying to unlock it. At that point, the page might be gone by the time the unlock comes in. Unlock the page first before checking the value of "rc", and only then end writeback and release the pages. The page lock isn't required for any of those operations so this should be safe. Reported-by: Anton Altaparmakov Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifssmb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 7353bc5d73d7..8e2e799e7a24 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1909,12 +1909,12 @@ cifs_writev_requeue(struct cifs_writedata *wdata) } while (rc == -EAGAIN); for (i = 0; i < wdata->nr_pages; i++) { + unlock_page(wdata->pages[i]); if (rc != 0) { SetPageError(wdata->pages[i]); end_page_writeback(wdata->pages[i]); page_cache_release(wdata->pages[i]); } - unlock_page(wdata->pages[i]); } mapping_set_error(inode->i_mapping, rc); -- cgit From 067785c40e52089993757afa28988c05f3cb2694 Mon Sep 17 00:00:00 2001 From: Pavel Shilovsky Date: Wed, 6 Mar 2013 19:38:36 +0400 Subject: CIFS: Fix missing of oplock_read value in smb30_values structure Cc: stable@vger.kernel.org Signed-off-by: Pavel Shilovsky Signed-off-by: Steve French --- fs/cifs/smb2ops.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c index c9c7aa7ed966..bceffe7b8f8d 100644 --- a/fs/cifs/smb2ops.c +++ b/fs/cifs/smb2ops.c @@ -744,4 +744,5 @@ struct smb_version_values smb30_values = { .cap_unix = 0, .cap_nt_find = SMB2_NT_FIND, .cap_large_files = SMB2_LARGE_FILES, + .oplock_read = SMB2_OPLOCK_LEVEL_II, }; -- cgit From 3a01aa7a25274308fe813a6237f678aed901cea3 Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Wed, 6 Mar 2013 01:57:55 -0700 Subject: Btrfs: fix a mismerge in btrfs_balance() Raid56 merge (merge commit e942f88) had mistakenly removed a call to __cancel_balance(), which resulted in balance not cleaning up after itself after a successful finish. (Cleanup includes switching the state, removing the balance item and releasing mut_ex_op testnset lock.) Bring it back. Reported-by: David Sterba Signed-off-by: Ilya Dryomov Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 9ff454df6756..6b9cff42265d 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -3235,6 +3235,11 @@ int btrfs_balance(struct btrfs_balance_control *bctl, update_ioctl_balance_args(fs_info, 0, bargs); } + if ((ret && ret != -ECANCELED && ret != -ENOSPC) || + balance_need_close(fs_info)) { + __cancel_balance(fs_info); + } + wake_up(&fs_info->balance_wait_q); return ret; -- cgit From 9141770548d529b9d32d5b08d59b65ee65afe0d4 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 7 Mar 2013 01:08:55 -0800 Subject: fs: Limit sys_mount to only request filesystem modules (Part 2). Add missing MODULE_ALIAS_FS("ocfs2") how did I miss that? Remove unnecessary MODULE_ALIAS_FS("devpts") devpts can not be modular. Signed-off-by: "Eric W. Biederman" --- fs/devpts/inode.c | 1 - fs/ocfs2/super.c | 1 + 2 files changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 79b662985efe..073d30b9d1ac 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -510,7 +510,6 @@ static struct file_system_type devpts_fs_type = { .fs_flags = FS_USERNS_MOUNT | FS_USERNS_DEV_MOUNT, #endif }; -MODULE_ALIAS_FS("devpts"); /* * The normal naming convention is simply /dev/pts/; this conforms diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 9b6910dec4ba..01b85165552b 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1266,6 +1266,7 @@ static struct file_system_type ocfs2_fs_type = { .fs_flags = FS_REQUIRES_DEV|FS_RENAME_DOES_D_MOVE, .next = NULL }; +MODULE_ALIAS_FS("ocfs2"); static int ocfs2_check_set_options(struct super_block *sb, struct mount_options *options) -- cgit From de3cb945db4d8eb3b046dc7a5ea89a893372750c Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Mon, 4 Mar 2013 17:13:31 -0500 Subject: Btrfs: improve the delayed inode throttling The delayed inode code batches up changes to the btree in hopes of doing them in bulk. As the changes build up, processes kick off worker threads and wait for them to make progress. The current code kicks off an async work queue item for each delayed node, which creates a lot of churn. It also uses a fixed 1 HZ waiting period for the throttle, which allows us to build a lot of pending work and can slow down the commit. This changes us to watch a sequence counter as it is bumped during the operations. We kick off fewer work items and have each work item do more work. Signed-off-by: Chris Mason --- fs/btrfs/delayed-inode.c | 151 ++++++++++++++++++++++++++++------------------- fs/btrfs/delayed-inode.h | 2 + 2 files changed, 92 insertions(+), 61 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index 0b278b117cbe..14fce27b4780 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -22,8 +22,9 @@ #include "disk-io.h" #include "transaction.h" -#define BTRFS_DELAYED_WRITEBACK 400 -#define BTRFS_DELAYED_BACKGROUND 100 +#define BTRFS_DELAYED_WRITEBACK 512 +#define BTRFS_DELAYED_BACKGROUND 128 +#define BTRFS_DELAYED_BATCH 16 static struct kmem_cache *delayed_node_cache; @@ -494,6 +495,15 @@ static int __btrfs_add_delayed_deletion_item(struct btrfs_delayed_node *node, BTRFS_DELAYED_DELETION_ITEM); } +static void finish_one_item(struct btrfs_delayed_root *delayed_root) +{ + int seq = atomic_inc_return(&delayed_root->items_seq); + if ((atomic_dec_return(&delayed_root->items) < + BTRFS_DELAYED_BACKGROUND || seq % BTRFS_DELAYED_BATCH == 0) && + waitqueue_active(&delayed_root->wait)) + wake_up(&delayed_root->wait); +} + static void __btrfs_remove_delayed_item(struct btrfs_delayed_item *delayed_item) { struct rb_root *root; @@ -512,10 +522,8 @@ static void __btrfs_remove_delayed_item(struct btrfs_delayed_item *delayed_item) rb_erase(&delayed_item->rb_node, root); delayed_item->delayed_node->count--; - if (atomic_dec_return(&delayed_root->items) < - BTRFS_DELAYED_BACKGROUND && - waitqueue_active(&delayed_root->wait)) - wake_up(&delayed_root->wait); + + finish_one_item(delayed_root); } static void btrfs_release_delayed_item(struct btrfs_delayed_item *item) @@ -1056,10 +1064,7 @@ static void btrfs_release_delayed_inode(struct btrfs_delayed_node *delayed_node) delayed_node->count--; delayed_root = delayed_node->root->fs_info->delayed_root; - if (atomic_dec_return(&delayed_root->items) < - BTRFS_DELAYED_BACKGROUND && - waitqueue_active(&delayed_root->wait)) - wake_up(&delayed_root->wait); + finish_one_item(delayed_root); } } @@ -1304,35 +1309,44 @@ void btrfs_remove_delayed_node(struct inode *inode) btrfs_release_delayed_node(delayed_node); } -struct btrfs_async_delayed_node { - struct btrfs_root *root; - struct btrfs_delayed_node *delayed_node; +struct btrfs_async_delayed_work { + struct btrfs_delayed_root *delayed_root; + int nr; struct btrfs_work work; }; -static void btrfs_async_run_delayed_node_done(struct btrfs_work *work) +static void btrfs_async_run_delayed_root(struct btrfs_work *work) { - struct btrfs_async_delayed_node *async_node; + struct btrfs_async_delayed_work *async_work; + struct btrfs_delayed_root *delayed_root; struct btrfs_trans_handle *trans; struct btrfs_path *path; struct btrfs_delayed_node *delayed_node = NULL; struct btrfs_root *root; struct btrfs_block_rsv *block_rsv; - int need_requeue = 0; + int total_done = 0; - async_node = container_of(work, struct btrfs_async_delayed_node, work); + async_work = container_of(work, struct btrfs_async_delayed_work, work); + delayed_root = async_work->delayed_root; path = btrfs_alloc_path(); if (!path) goto out; - path->leave_spinning = 1; - delayed_node = async_node->delayed_node; +again: + if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND / 2) + goto free_path; + + delayed_node = btrfs_first_prepared_delayed_node(delayed_root); + if (!delayed_node) + goto free_path; + + path->leave_spinning = 1; root = delayed_node->root; trans = btrfs_join_transaction(root); if (IS_ERR(trans)) - goto free_path; + goto release_path; block_rsv = trans->block_rsv; trans->block_rsv = &root->fs_info->delayed_block_rsv; @@ -1363,57 +1377,47 @@ static void btrfs_async_run_delayed_node_done(struct btrfs_work *work) * Task1 will sleep until the transaction is commited. */ mutex_lock(&delayed_node->mutex); - if (delayed_node->count) - need_requeue = 1; - else - btrfs_dequeue_delayed_node(root->fs_info->delayed_root, - delayed_node); + btrfs_dequeue_delayed_node(root->fs_info->delayed_root, delayed_node); mutex_unlock(&delayed_node->mutex); trans->block_rsv = block_rsv; btrfs_end_transaction_dmeta(trans, root); btrfs_btree_balance_dirty_nodelay(root); + +release_path: + btrfs_release_path(path); + total_done++; + + btrfs_release_prepared_delayed_node(delayed_node); + if (async_work->nr == 0 || total_done < async_work->nr) + goto again; + free_path: btrfs_free_path(path); out: - if (need_requeue) - btrfs_requeue_work(&async_node->work); - else { - btrfs_release_prepared_delayed_node(delayed_node); - kfree(async_node); - } + wake_up(&delayed_root->wait); + kfree(async_work); } + static int btrfs_wq_run_delayed_node(struct btrfs_delayed_root *delayed_root, - struct btrfs_root *root, int all) + struct btrfs_root *root, int nr) { - struct btrfs_async_delayed_node *async_node; - struct btrfs_delayed_node *curr; - int count = 0; + struct btrfs_async_delayed_work *async_work; -again: - curr = btrfs_first_prepared_delayed_node(delayed_root); - if (!curr) + if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND) return 0; - async_node = kmalloc(sizeof(*async_node), GFP_NOFS); - if (!async_node) { - btrfs_release_prepared_delayed_node(curr); + async_work = kmalloc(sizeof(*async_work), GFP_NOFS); + if (!async_work) return -ENOMEM; - } - - async_node->root = root; - async_node->delayed_node = curr; - - async_node->work.func = btrfs_async_run_delayed_node_done; - async_node->work.flags = 0; - btrfs_queue_worker(&root->fs_info->delayed_workers, &async_node->work); - count++; - - if (all || count < 4) - goto again; + async_work->delayed_root = delayed_root; + async_work->work.func = btrfs_async_run_delayed_root; + async_work->work.flags = 0; + async_work->nr = nr; + btrfs_queue_worker(&root->fs_info->delayed_workers, &async_work->work); return 0; } @@ -1424,30 +1428,55 @@ void btrfs_assert_delayed_root_empty(struct btrfs_root *root) WARN_ON(btrfs_first_delayed_node(delayed_root)); } +static int refs_newer(struct btrfs_delayed_root *delayed_root, + int seq, int count) +{ + int val = atomic_read(&delayed_root->items_seq); + + if (val < seq || val >= seq + count) + return 1; + return 0; +} + void btrfs_balance_delayed_items(struct btrfs_root *root) { struct btrfs_delayed_root *delayed_root; + int seq; delayed_root = btrfs_get_delayed_root(root); if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND) return; + seq = atomic_read(&delayed_root->items_seq); + if (atomic_read(&delayed_root->items) >= BTRFS_DELAYED_WRITEBACK) { int ret; - ret = btrfs_wq_run_delayed_node(delayed_root, root, 1); + DEFINE_WAIT(__wait); + + ret = btrfs_wq_run_delayed_node(delayed_root, root, 0); if (ret) return; - wait_event_interruptible_timeout( - delayed_root->wait, - (atomic_read(&delayed_root->items) < - BTRFS_DELAYED_BACKGROUND), - HZ); - return; + while (1) { + prepare_to_wait(&delayed_root->wait, &__wait, + TASK_INTERRUPTIBLE); + + if (refs_newer(delayed_root, seq, + BTRFS_DELAYED_BATCH) || + atomic_read(&delayed_root->items) < + BTRFS_DELAYED_BACKGROUND) { + break; + } + if (!signal_pending(current)) + schedule(); + else + break; + } + finish_wait(&delayed_root->wait, &__wait); } - btrfs_wq_run_delayed_node(delayed_root, root, 0); + btrfs_wq_run_delayed_node(delayed_root, root, BTRFS_DELAYED_BATCH); } /* Will return 0 or -ENOMEM */ diff --git a/fs/btrfs/delayed-inode.h b/fs/btrfs/delayed-inode.h index 78b6ad0fc669..1d5c5f7abe3e 100644 --- a/fs/btrfs/delayed-inode.h +++ b/fs/btrfs/delayed-inode.h @@ -43,6 +43,7 @@ struct btrfs_delayed_root { */ struct list_head prepare_list; atomic_t items; /* for delayed items */ + atomic_t items_seq; /* for delayed items */ int nodes; /* for delayed nodes */ wait_queue_head_t wait; }; @@ -86,6 +87,7 @@ static inline void btrfs_init_delayed_root( struct btrfs_delayed_root *delayed_root) { atomic_set(&delayed_root->items, 0); + atomic_set(&delayed_root->items_seq, 0); delayed_root->nodes = 0; spin_lock_init(&delayed_root->lock); init_waitqueue_head(&delayed_root->wait); -- cgit From 7b54c165a0c012edbaeaa73c5c87cb73721eb580 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 8 Mar 2013 09:03:07 -0800 Subject: vfs: don't BUG_ON() if following a /proc fd pseudo-symlink results in a symlink It's "normal" - it can happen if the file descriptor you followed was opened with O_NOFOLLOW. Reported-by: Dave Jones Cc: Al Viro Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- fs/namei.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 961bc1268366..57ae9c8c66bf 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -689,8 +689,6 @@ void nd_jump_link(struct nameidata *nd, struct path *path) nd->path = *path; nd->inode = nd->path.dentry->d_inode; nd->flags |= LOOKUP_JUMPED; - - BUG_ON(nd->inode->i_op->follow_link); } static inline void put_link(struct nameidata *nd, struct path *link, void *cookie) -- cgit From db04dc679bcc780ad6907943afe24a30de974a1b Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 9 Mar 2013 00:14:45 -0800 Subject: proc: Use nd_jump_link in proc_ns_follow_link Update proc_ns_follow_link to use nd_jump_link instead of just manually updating nd.path.dentry. This fixes the BUG_ON(nd->inode != parent->d_inode) reported by Dave Jones and reproduced trivially with mkdir /proc/self/ns/uts/a. Sigh it looks like the VFS change to require use of nd_jump_link happend while proc_ns_follow_link was baking and since the common case of proc_ns_follow_link continued to work without problems the need for making this change was overlooked. Cc: stable@vger.kernel.org Signed-off-by: "Eric W. Biederman" --- fs/proc/namespaces.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/proc/namespaces.c b/fs/proc/namespaces.c index b7a47196c8c3..66b51c0383da 100644 --- a/fs/proc/namespaces.c +++ b/fs/proc/namespaces.c @@ -118,7 +118,7 @@ static void *proc_ns_follow_link(struct dentry *dentry, struct nameidata *nd) struct super_block *sb = inode->i_sb; struct proc_inode *ei = PROC_I(inode); struct task_struct *task; - struct dentry *ns_dentry; + struct path ns_path; void *error = ERR_PTR(-EACCES); task = get_proc_task(inode); @@ -128,14 +128,14 @@ static void *proc_ns_follow_link(struct dentry *dentry, struct nameidata *nd) if (!ptrace_may_access(task, PTRACE_MODE_READ)) goto out_put_task; - ns_dentry = proc_ns_get_dentry(sb, task, ei->ns_ops); - if (IS_ERR(ns_dentry)) { - error = ERR_CAST(ns_dentry); + ns_path.dentry = proc_ns_get_dentry(sb, task, ei->ns_ops); + if (IS_ERR(ns_path.dentry)) { + error = ERR_CAST(ns_path.dentry); goto out_put_task; } - dput(nd->path.dentry); - nd->path.dentry = ns_dentry; + ns_path.mnt = mntget(nd->path.mnt); + nd_jump_link(nd, &ns_path); error = NULL; out_put_task: -- cgit From bd384364c1185ecb01f3b8242c915ccb5921c60d Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Sun, 10 Mar 2013 20:48:59 -0400 Subject: ext4: avoid a potential overflow in ext4_es_can_be_merged() Check the length of an extent to avoid a potential overflow in ext4_es_can_be_merged(). Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Dmitry Monakhov --- fs/ext4/extents_status.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 95796a1b7522..37f9a2d8fd04 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -333,17 +333,27 @@ static void ext4_es_free_extent(struct inode *inode, struct extent_status *es) static int ext4_es_can_be_merged(struct extent_status *es1, struct extent_status *es2) { - if (es1->es_lblk + es1->es_len != es2->es_lblk) + if (ext4_es_status(es1) != ext4_es_status(es2)) return 0; - if (ext4_es_status(es1) != ext4_es_status(es2)) + if (((__u64) es1->es_len) + es2->es_len > 0xFFFFFFFFULL) return 0; - if ((ext4_es_is_written(es1) || ext4_es_is_unwritten(es1)) && - (ext4_es_pblock(es1) + es1->es_len != ext4_es_pblock(es2))) + if (((__u64) es1->es_lblk) + es1->es_len != es2->es_lblk) return 0; - return 1; + if ((ext4_es_is_written(es1) || ext4_es_is_unwritten(es1)) && + (ext4_es_pblock(es1) + es1->es_len == ext4_es_pblock(es2))) + return 1; + + if (ext4_es_is_hole(es1)) + return 1; + + /* we need to check delayed extent is without unwritten status */ + if (ext4_es_is_delayed(es1) && !ext4_es_is_unwritten(es1)) + return 1; + + return 0; } static struct extent_status * -- cgit From 921f266bc6bfe6ebb599c559f10443af314c19ec Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Sun, 10 Mar 2013 21:01:03 -0400 Subject: ext4: add self-testing infrastructure to do a sanity check This commit adds a self-testing infrastructure like extent tree does to do a sanity check for extent status tree. After status tree is as a extent cache, we'd better to make sure that it caches right result. After applied this commit, we will get a lot of messages when we run xfstests as below. ... kernel: ES len assertation failed for inode: 230 retval 1 != map->m_len 3 in ext4_map_blocks (allocation) ... kernel: ES cache assertation failed for inode: 230 es_cached ex [974/2/4781/20] != found ex [974/1/4781/1000] ... kernel: ES insert assertation failed for inode: 635 ex_status [0/45/21388/w] != es_status [44/1/21432/u] ... Signed-off-by: Dmitry Monakhov Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" --- fs/ext4/extents_status.c | 175 +++++++++++++++++++++++++++++++++++++++++++++++ fs/ext4/extents_status.h | 6 ++ fs/ext4/inode.c | 96 ++++++++++++++++++++++++++ 3 files changed, 277 insertions(+) (limited to 'fs') diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index 37f9a2d8fd04..d2a8cb74676b 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -399,6 +399,179 @@ ext4_es_try_to_merge_right(struct inode *inode, struct extent_status *es) return es; } +#ifdef ES_AGGRESSIVE_TEST +static void ext4_es_insert_extent_ext_check(struct inode *inode, + struct extent_status *es) +{ + struct ext4_ext_path *path = NULL; + struct ext4_extent *ex; + ext4_lblk_t ee_block; + ext4_fsblk_t ee_start; + unsigned short ee_len; + int depth, ee_status, es_status; + + path = ext4_ext_find_extent(inode, es->es_lblk, NULL); + if (IS_ERR(path)) + return; + + depth = ext_depth(inode); + ex = path[depth].p_ext; + + if (ex) { + + ee_block = le32_to_cpu(ex->ee_block); + ee_start = ext4_ext_pblock(ex); + ee_len = ext4_ext_get_actual_len(ex); + + ee_status = ext4_ext_is_uninitialized(ex) ? 1 : 0; + es_status = ext4_es_is_unwritten(es) ? 1 : 0; + + /* + * Make sure ex and es are not overlap when we try to insert + * a delayed/hole extent. + */ + if (!ext4_es_is_written(es) && !ext4_es_is_unwritten(es)) { + if (in_range(es->es_lblk, ee_block, ee_len)) { + pr_warn("ES insert assertation failed for " + "inode: %lu we can find an extent " + "at block [%d/%d/%llu/%c], but we " + "want to add an delayed/hole extent " + "[%d/%d/%llu/%llx]\n", + inode->i_ino, ee_block, ee_len, + ee_start, ee_status ? 'u' : 'w', + es->es_lblk, es->es_len, + ext4_es_pblock(es), ext4_es_status(es)); + } + goto out; + } + + /* + * We don't check ee_block == es->es_lblk, etc. because es + * might be a part of whole extent, vice versa. + */ + if (es->es_lblk < ee_block || + ext4_es_pblock(es) != ee_start + es->es_lblk - ee_block) { + pr_warn("ES insert assertation failed for inode: %lu " + "ex_status [%d/%d/%llu/%c] != " + "es_status [%d/%d/%llu/%c]\n", inode->i_ino, + ee_block, ee_len, ee_start, + ee_status ? 'u' : 'w', es->es_lblk, es->es_len, + ext4_es_pblock(es), es_status ? 'u' : 'w'); + goto out; + } + + if (ee_status ^ es_status) { + pr_warn("ES insert assertation failed for inode: %lu " + "ex_status [%d/%d/%llu/%c] != " + "es_status [%d/%d/%llu/%c]\n", inode->i_ino, + ee_block, ee_len, ee_start, + ee_status ? 'u' : 'w', es->es_lblk, es->es_len, + ext4_es_pblock(es), es_status ? 'u' : 'w'); + } + } else { + /* + * We can't find an extent on disk. So we need to make sure + * that we don't want to add an written/unwritten extent. + */ + if (!ext4_es_is_delayed(es) && !ext4_es_is_hole(es)) { + pr_warn("ES insert assertation failed for inode: %lu " + "can't find an extent at block %d but we want " + "to add an written/unwritten extent " + "[%d/%d/%llu/%llx]\n", inode->i_ino, + es->es_lblk, es->es_lblk, es->es_len, + ext4_es_pblock(es), ext4_es_status(es)); + } + } +out: + if (path) { + ext4_ext_drop_refs(path); + kfree(path); + } +} + +static void ext4_es_insert_extent_ind_check(struct inode *inode, + struct extent_status *es) +{ + struct ext4_map_blocks map; + int retval; + + /* + * Here we call ext4_ind_map_blocks to lookup a block mapping because + * 'Indirect' structure is defined in indirect.c. So we couldn't + * access direct/indirect tree from outside. It is too dirty to define + * this function in indirect.c file. + */ + + map.m_lblk = es->es_lblk; + map.m_len = es->es_len; + + retval = ext4_ind_map_blocks(NULL, inode, &map, 0); + if (retval > 0) { + if (ext4_es_is_delayed(es) || ext4_es_is_hole(es)) { + /* + * We want to add a delayed/hole extent but this + * block has been allocated. + */ + pr_warn("ES insert assertation failed for inode: %lu " + "We can find blocks but we want to add a " + "delayed/hole extent [%d/%d/%llu/%llx]\n", + inode->i_ino, es->es_lblk, es->es_len, + ext4_es_pblock(es), ext4_es_status(es)); + return; + } else if (ext4_es_is_written(es)) { + if (retval != es->es_len) { + pr_warn("ES insert assertation failed for " + "inode: %lu retval %d != es_len %d\n", + inode->i_ino, retval, es->es_len); + return; + } + if (map.m_pblk != ext4_es_pblock(es)) { + pr_warn("ES insert assertation failed for " + "inode: %lu m_pblk %llu != " + "es_pblk %llu\n", + inode->i_ino, map.m_pblk, + ext4_es_pblock(es)); + return; + } + } else { + /* + * We don't need to check unwritten extent because + * indirect-based file doesn't have it. + */ + BUG_ON(1); + } + } else if (retval == 0) { + if (ext4_es_is_written(es)) { + pr_warn("ES insert assertation failed for inode: %lu " + "We can't find the block but we want to add " + "an written extent [%d/%d/%llu/%llx]\n", + inode->i_ino, es->es_lblk, es->es_len, + ext4_es_pblock(es), ext4_es_status(es)); + return; + } + } +} + +static inline void ext4_es_insert_extent_check(struct inode *inode, + struct extent_status *es) +{ + /* + * We don't need to worry about the race condition because + * caller takes i_data_sem locking. + */ + BUG_ON(!rwsem_is_locked(&EXT4_I(inode)->i_data_sem)); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + ext4_es_insert_extent_ext_check(inode, es); + else + ext4_es_insert_extent_ind_check(inode, es); +} +#else +static inline void ext4_es_insert_extent_check(struct inode *inode, + struct extent_status *es) +{ +} +#endif + static int __es_insert_extent(struct inode *inode, struct extent_status *newes) { struct ext4_es_tree *tree = &EXT4_I(inode)->i_es_tree; @@ -481,6 +654,8 @@ int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t lblk, ext4_es_store_status(&newes, status); trace_ext4_es_insert_extent(inode, &newes); + ext4_es_insert_extent_check(inode, &newes); + write_lock(&EXT4_I(inode)->i_es_lock); err = __es_remove_extent(inode, lblk, end); if (err != 0) diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index f190dfe969da..56140ad4150b 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -20,6 +20,12 @@ #define es_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) #endif +/* + * With ES_AGGRESSIVE_TEST defined, the result of es caching will be + * checked with old map_block's result. + */ +#define ES_AGGRESSIVE_TEST__ + /* * These flags live in the high bits of extent_status.es_pblk */ diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 95a0c62c5683..3186a43fa4b0 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -482,6 +482,58 @@ static pgoff_t ext4_num_dirty_pages(struct inode *inode, pgoff_t idx, return num; } +#ifdef ES_AGGRESSIVE_TEST +static void ext4_map_blocks_es_recheck(handle_t *handle, + struct inode *inode, + struct ext4_map_blocks *es_map, + struct ext4_map_blocks *map, + int flags) +{ + int retval; + + map->m_flags = 0; + /* + * There is a race window that the result is not the same. + * e.g. xfstests #223 when dioread_nolock enables. The reason + * is that we lookup a block mapping in extent status tree with + * out taking i_data_sem. So at the time the unwritten extent + * could be converted. + */ + if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) + down_read((&EXT4_I(inode)->i_data_sem)); + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + retval = ext4_ext_map_blocks(handle, inode, map, flags & + EXT4_GET_BLOCKS_KEEP_SIZE); + } else { + retval = ext4_ind_map_blocks(handle, inode, map, flags & + EXT4_GET_BLOCKS_KEEP_SIZE); + } + if (!(flags & EXT4_GET_BLOCKS_NO_LOCK)) + up_read((&EXT4_I(inode)->i_data_sem)); + /* + * Clear EXT4_MAP_FROM_CLUSTER and EXT4_MAP_BOUNDARY flag + * because it shouldn't be marked in es_map->m_flags. + */ + map->m_flags &= ~(EXT4_MAP_FROM_CLUSTER | EXT4_MAP_BOUNDARY); + + /* + * We don't check m_len because extent will be collpased in status + * tree. So the m_len might not equal. + */ + if (es_map->m_lblk != map->m_lblk || + es_map->m_flags != map->m_flags || + es_map->m_pblk != map->m_pblk) { + printk("ES cache assertation failed for inode: %lu " + "es_cached ex [%d/%d/%llu/%x] != " + "found ex [%d/%d/%llu/%x] retval %d flags %x\n", + inode->i_ino, es_map->m_lblk, es_map->m_len, + es_map->m_pblk, es_map->m_flags, map->m_lblk, + map->m_len, map->m_pblk, map->m_flags, + retval, flags); + } +} +#endif /* ES_AGGRESSIVE_TEST */ + /* * The ext4_map_blocks() function tries to look up the requested blocks, * and returns if the blocks are already mapped. @@ -509,6 +561,11 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, { struct extent_status es; int retval; +#ifdef ES_AGGRESSIVE_TEST + struct ext4_map_blocks orig_map; + + memcpy(&orig_map, map, sizeof(*map)); +#endif map->m_flags = 0; ext_debug("ext4_map_blocks(): inode %lu, flag %d, max_blocks %u," @@ -531,6 +588,10 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, } else { BUG_ON(1); } +#ifdef ES_AGGRESSIVE_TEST + ext4_map_blocks_es_recheck(handle, inode, map, + &orig_map, flags); +#endif goto found; } @@ -551,6 +612,15 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, int ret; unsigned long long status; +#ifdef ES_AGGRESSIVE_TEST + if (retval != map->m_len) { + printk("ES len assertation failed for inode: %lu " + "retval %d != map->m_len %d " + "in %s (lookup)\n", inode->i_ino, retval, + map->m_len, __func__); + } +#endif + status = map->m_flags & EXT4_MAP_UNWRITTEN ? EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) && @@ -643,6 +713,15 @@ found: int ret; unsigned long long status; +#ifdef ES_AGGRESSIVE_TEST + if (retval != map->m_len) { + printk("ES len assertation failed for inode: %lu " + "retval %d != map->m_len %d " + "in %s (allocation)\n", inode->i_ino, retval, + map->m_len, __func__); + } +#endif + status = map->m_flags & EXT4_MAP_UNWRITTEN ? EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) && @@ -1768,6 +1847,11 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, struct extent_status es; int retval; sector_t invalid_block = ~((sector_t) 0xffff); +#ifdef ES_AGGRESSIVE_TEST + struct ext4_map_blocks orig_map; + + memcpy(&orig_map, map, sizeof(*map)); +#endif if (invalid_block < ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es)) invalid_block = ~0; @@ -1809,6 +1893,9 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, else BUG_ON(1); +#ifdef ES_AGGRESSIVE_TEST + ext4_map_blocks_es_recheck(NULL, inode, map, &orig_map, 0); +#endif return retval; } @@ -1873,6 +1960,15 @@ add_delayed: int ret; unsigned long long status; +#ifdef ES_AGGRESSIVE_TEST + if (retval != map->m_len) { + printk("ES len assertation failed for inode: %lu " + "retval %d != map->m_len %d " + "in %s (lookup)\n", inode->i_ino, retval, + map->m_len, __func__); + } +#endif + status = map->m_flags & EXT4_MAP_UNWRITTEN ? EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, -- cgit From cdee78433c138c2f2018a6884673739af2634787 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Sun, 10 Mar 2013 21:08:52 -0400 Subject: ext4: fix wrong m_len value after unwritten extent conversion The ext4_ext_handle_uninitialized_extents() function was assuming the return value of ext4_ext_map_blocks() is equal to map->m_len. This incorrect assumption was harmless until we started use status tree as a extent cache because we need to update status tree according to 'm_len' value. Meanwhile this commit marks EXT4_MAP_MAPPED flag after unwritten extent conversion. It shouldn't cause a bug because we update status tree according to checking EXT4_MAP_UNWRITTEN flag. But it should be fixed. After applied this commit, the following error message from self-testing infrastructure disappears. ... kernel: ES len assertation failed for inode: 230 retval 1 != map->m_len 3 in ext4_map_blocks (allocation) ... Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Dmitry Monakhov --- fs/ext4/extents.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 25c86aaa38d6..110e85a1f82a 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3650,6 +3650,10 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, path, map->m_len); } else err = ret; + map->m_flags |= EXT4_MAP_MAPPED; + if (allocated > map->m_len) + allocated = map->m_len; + map->m_len = allocated; goto out2; } /* buffered IO case */ -- cgit From adb2355104b2109e06ba5276485d187d023b2fd2 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Sun, 10 Mar 2013 21:13:05 -0400 Subject: ext4: update extent status tree after an extent is zeroed out When we try to split an extent, this extent could be zeroed out and mark as initialized. But we don't know this in ext4_map_blocks because it only returns a length of allocated extent. Meanwhile we will mark this extent as uninitialized because we only check m_flags. This commit update extent status tree when we try to split an unwritten extent. We don't need to worry about the status of this extent because we always mark it as initialized. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Dmitry Monakhov --- fs/ext4/extents.c | 35 +++++++++++++++++++++++++++++++---- fs/ext4/extents_status.c | 17 +++++++++++++++++ fs/ext4/extents_status.h | 3 +++ fs/ext4/inode.c | 10 ++++++++++ 4 files changed, 61 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 110e85a1f82a..7e37018d1753 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -2925,7 +2925,7 @@ static int ext4_split_extent_at(handle_t *handle, { ext4_fsblk_t newblock; ext4_lblk_t ee_block; - struct ext4_extent *ex, newex, orig_ex; + struct ext4_extent *ex, newex, orig_ex, zero_ex; struct ext4_extent *ex2 = NULL; unsigned int ee_len, depth; int err = 0; @@ -2996,12 +2996,26 @@ static int ext4_split_extent_at(handle_t *handle, err = ext4_ext_insert_extent(handle, inode, path, &newex, flags); if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) { if (split_flag & (EXT4_EXT_DATA_VALID1|EXT4_EXT_DATA_VALID2)) { - if (split_flag & EXT4_EXT_DATA_VALID1) + if (split_flag & EXT4_EXT_DATA_VALID1) { err = ext4_ext_zeroout(inode, ex2); - else + zero_ex.ee_block = ex2->ee_block; + zero_ex.ee_len = ext4_ext_get_actual_len(ex2); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(ex2)); + } else { err = ext4_ext_zeroout(inode, ex); - } else + zero_ex.ee_block = ex->ee_block; + zero_ex.ee_len = ext4_ext_get_actual_len(ex); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(ex)); + } + } else { err = ext4_ext_zeroout(inode, &orig_ex); + zero_ex.ee_block = orig_ex.ee_block; + zero_ex.ee_len = ext4_ext_get_actual_len(&orig_ex); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(&orig_ex)); + } if (err) goto fix_extent_len; @@ -3009,6 +3023,12 @@ static int ext4_split_extent_at(handle_t *handle, ex->ee_len = cpu_to_le16(ee_len); ext4_ext_try_to_merge(handle, inode, path, ex); err = ext4_ext_dirty(handle, inode, path + path->p_depth); + if (err) + goto fix_extent_len; + + /* update extent status tree */ + err = ext4_es_zeroout(inode, &zero_ex); + goto out; } else if (err) goto fix_extent_len; @@ -3150,6 +3170,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle, ee_block = le32_to_cpu(ex->ee_block); ee_len = ext4_ext_get_actual_len(ex); allocated = ee_len - (map->m_lblk - ee_block); + zero_ex.ee_len = 0; trace_ext4_ext_convert_to_initialized_enter(inode, map, ex); @@ -3247,6 +3268,9 @@ static int ext4_ext_convert_to_initialized(handle_t *handle, err = ext4_ext_zeroout(inode, ex); if (err) goto out; + zero_ex.ee_block = ex->ee_block; + zero_ex.ee_len = ext4_ext_get_actual_len(ex); + ext4_ext_store_pblock(&zero_ex, ext4_ext_pblock(ex)); err = ext4_ext_get_access(handle, inode, path + depth); if (err) @@ -3305,6 +3329,9 @@ static int ext4_ext_convert_to_initialized(handle_t *handle, err = allocated; out: + /* If we have gotten a failure, don't zero out status tree */ + if (!err) + err = ext4_es_zeroout(inode, &zero_ex); return err ? err : allocated; } diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c index d2a8cb74676b..fe3337a85ede 100644 --- a/fs/ext4/extents_status.c +++ b/fs/ext4/extents_status.c @@ -854,6 +854,23 @@ int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t lblk, return err; } +int ext4_es_zeroout(struct inode *inode, struct ext4_extent *ex) +{ + ext4_lblk_t ee_block; + ext4_fsblk_t ee_pblock; + unsigned int ee_len; + + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + ee_pblock = ext4_ext_pblock(ex); + + if (ee_len == 0) + return 0; + + return ext4_es_insert_extent(inode, ee_block, ee_len, ee_pblock, + EXTENT_STATUS_WRITTEN); +} + static int ext4_es_shrink(struct shrinker *shrink, struct shrink_control *sc) { struct ext4_sb_info *sbi = container_of(shrink, diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h index 56140ad4150b..d8e2d4dc311e 100644 --- a/fs/ext4/extents_status.h +++ b/fs/ext4/extents_status.h @@ -39,6 +39,8 @@ EXTENT_STATUS_DELAYED | \ EXTENT_STATUS_HOLE) +struct ext4_extent; + struct extent_status { struct rb_node rb_node; ext4_lblk_t es_lblk; /* first logical block extent covers */ @@ -64,6 +66,7 @@ extern void ext4_es_find_delayed_extent(struct inode *inode, ext4_lblk_t lblk, struct extent_status *es); extern int ext4_es_lookup_extent(struct inode *inode, ext4_lblk_t lblk, struct extent_status *es); +extern int ext4_es_zeroout(struct inode *inode, struct ext4_extent *ex); static inline int ext4_es_is_written(struct extent_status *es) { diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 3186a43fa4b0..4f1d54a88d8c 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -722,6 +722,15 @@ found: } #endif + /* + * If the extent has been zeroed out, we don't need to update + * extent status tree. + */ + if ((flags & EXT4_GET_BLOCKS_PRE_IO) && + ext4_es_lookup_extent(inode, map->m_lblk, &es)) { + if (ext4_es_is_written(&es)) + goto has_zeroout; + } status = map->m_flags & EXT4_MAP_UNWRITTEN ? EXTENT_STATUS_UNWRITTEN : EXTENT_STATUS_WRITTEN; if (!(flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) && @@ -734,6 +743,7 @@ found: retval = ret; } +has_zeroout: up_write((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { int ret = check_block_validity(inode, map); -- cgit From 3a2256702e47f68f921dfad41b1764d05c572329 Mon Sep 17 00:00:00 2001 From: Zheng Liu Date: Sun, 10 Mar 2013 21:20:23 -0400 Subject: ext4: fix the wrong number of the allocated blocks in ext4_split_extent() This commit fixes a wrong return value of the number of the allocated blocks in ext4_split_extent. When the length of blocks we want to allocate is greater than the length of the current extent, we return a wrong number. Let's see what happens in the following case when we call ext4_split_extent(). map: [48, 72] ex: [32, 64, u] 'ex' will be split into two parts: ex1: [32, 47, u] ex2: [48, 64, w] 'map->m_len' is returned from this function, and the value is 24. But the real length is 16. So it should be fixed. Meanwhile in this commit we use right length of the allocated blocks when get_reserved_cluster_alloc in ext4_ext_handle_uninitialized_extents is called. Signed-off-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Cc: Dmitry Monakhov Cc: stable@vger.kernel.org --- fs/ext4/extents.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 7e37018d1753..69df02ff96aa 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3067,6 +3067,7 @@ static int ext4_split_extent(handle_t *handle, int err = 0; int uninitialized; int split_flag1, flags1; + int allocated = map->m_len; depth = ext_depth(inode); ex = path[depth].p_ext; @@ -3086,6 +3087,8 @@ static int ext4_split_extent(handle_t *handle, map->m_lblk + map->m_len, split_flag1, flags1); if (err) goto out; + } else { + allocated = ee_len - (map->m_lblk - ee_block); } /* * Update path is required because previous ext4_split_extent_at() may @@ -3115,7 +3118,7 @@ static int ext4_split_extent(handle_t *handle, ext4_ext_show_leaf(inode, path); out: - return err ? err : map->m_len; + return err ? err : allocated; } /* @@ -3730,6 +3733,7 @@ out: allocated - map->m_len); allocated = map->m_len; } + map->m_len = allocated; /* * If we have done fallocate with the offset that is already -- cgit From e1c36595bedc2e1b4112f01256cb30f4d9f9ae46 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sun, 10 Mar 2013 22:19:00 -0400 Subject: ext4: fix WARN_ON from ext4_releasepage() ext4_releasepage() warns when it is passed a page with PageChecked set. However this can correctly happen when invalidate_inode_pages2_range() invalidates pages - and we should fail the release in that case. Since the page was dirty anyway, it won't be discarded and no harm has happened but it's good to be safe. Also remove bogus page_has_buffers() check - we are guaranteed page has buffers in this function. Reported-by: Zheng Liu Tested-by: Zheng Liu Signed-off-by: "Theodore Ts'o" Signed-off-by: Jan Kara --- fs/ext4/inode.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 4f1d54a88d8c..117a9e7aa4a0 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3018,8 +3018,8 @@ static int ext4_releasepage(struct page *page, gfp_t wait) trace_ext4_releasepage(page); - WARN_ON(PageChecked(page)); - if (!page_has_buffers(page)) + /* Page has dirty journalled data -> cannot release */ + if (PageChecked(page)) return 0; if (journal) return jbd2_journal_try_to_free_buffers(journal, page, wait); -- cgit From e3d85c366089015805f175324bb1780249f44669 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Sun, 10 Mar 2013 22:21:49 -0400 Subject: ext4: remove unused variable in ext4_free_blocks() Remove unused variable 'freed' in ext4_free_blocks(). Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" --- fs/ext4/mballoc.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 7bb713a46fe4..75e05f3a730f 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4464,7 +4464,6 @@ void ext4_free_blocks(handle_t *handle, struct inode *inode, struct buffer_head *bitmap_bh = NULL; struct super_block *sb = inode->i_sb; struct ext4_group_desc *gdp; - unsigned long freed = 0; unsigned int overflow; ext4_grpblk_t bit; struct buffer_head *gd_bh; @@ -4672,8 +4671,6 @@ do_more: ext4_mb_unload_buddy(&e4b); - freed += count; - if (!(flags & EXT4_FREE_BLOCKS_NO_QUOT_UPDATE)) dquot_free_block(inode, EXT4_C2B(sbi, count_clusters)); -- cgit From bb8b20ed94bc69120e31399c43cb336300dea109 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Sun, 10 Mar 2013 22:28:09 -0400 Subject: ext4: do not use yield() Using yield() is strongly discouraged (see sched/core.c) especially since we can just use cond_resched(). Replace all use of yield() with cond_resched(). Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 2 +- fs/ext4/mballoc.c | 8 ++------ 2 files changed, 3 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 117a9e7aa4a0..48fc023ab0a2 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1352,7 +1352,7 @@ repeat: ei->i_da_metadata_calc_last_lblock = save_last_lblock; spin_unlock(&ei->i_block_reservation_lock); if (ext4_should_retry_alloc(inode->i_sb, &retries)) { - yield(); + cond_resched(); goto repeat; } dquot_release_reservation_block(inode, EXT4_C2B(sbi, 1)); diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 75e05f3a730f..8b2ea9f75004 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -3692,11 +3692,7 @@ repeat: if (free < needed && busy) { busy = 0; ext4_unlock_group(sb, group); - /* - * Yield the CPU here so that we don't get soft lockup - * in non preempt case. - */ - yield(); + cond_resched(); goto repeat; } @@ -4246,7 +4242,7 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, ext4_claim_free_clusters(sbi, ar->len, ar->flags)) { /* let others to free the space */ - yield(); + cond_resched(); ar->len = ar->len >> 1; } if (!ar->len) { -- cgit From 232ec8720d4e45405e37144c67053042c6b886d3 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Sun, 10 Mar 2013 22:46:30 -0400 Subject: ext4: update reserved space after the 'correction' Currently in ext4_ext_map_blocks() in delayed allocation writeback we would update the reservation and after that check whether we claimed cluster outside of the range of the allocation and if so, we'll give the block back to the reservation pool. However this also means that if the number of reserved data block dropped to zero before the correction, we would release all the metadata reservation as well, however we might still need it because the we're not done with the delayed allocation and there might be more blocks to come. This will result in error messages such as: EXT4-fs warning (device sdb): ext4_da_update_reserve_space:361: ino 12, allocated 1 with only 0 reserved metadata blocks (releasing 1 blocks with reserved 1 data blocks) This will only happen on bigalloc file system and it can be easily reproduced using fiemap-tester from xfstests like this: ./src/fiemap-tester -m DHDHDHDHD -S -p0 /mnt/test/file Or using xfstests such as 225. Fix this by doing the correction first and updating the reservation after that so that we do not accidentally decrease i_reserved_data_blocks to zero. Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" --- fs/ext4/extents.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 69df02ff96aa..bd69e906bd91 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4165,9 +4165,6 @@ got_allocated_blocks: } } else { BUG_ON(allocated_clusters < reserved_clusters); - /* We will claim quota for all newly allocated blocks.*/ - ext4_da_update_reserve_space(inode, allocated_clusters, - 1); if (reserved_clusters < allocated_clusters) { struct ext4_inode_info *ei = EXT4_I(inode); int reservation = allocated_clusters - @@ -4218,6 +4215,15 @@ got_allocated_blocks: ei->i_reserved_data_blocks += reservation; spin_unlock(&ei->i_block_reservation_lock); } + /* + * We will claim quota for all newly allocated blocks. + * We're updating the reserved space *after* the + * correction above so we do not accidentally free + * all the metadata reservation because we might + * actually need it later on. + */ + ext4_da_update_reserve_space(inode, allocated_clusters, + 1); } } -- cgit From 386ad67c9ac043890121c066186883d1640348a4 Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Sun, 10 Mar 2013 22:50:00 -0400 Subject: ext4: reserve metadata block for every delayed write Currently we only reserve space (data+metadata) in delayed allocation if we're allocating from new cluster (which is always in non-bigalloc file system) which is ok for data blocks, because we reserve the whole cluster. However we have to reserve metadata for every delayed block we're going to write because every block could potentially require metedata block when we need to grow the extent tree. Signed-off-by: Lukas Czerner --- fs/ext4/inode.c | 63 +++++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 61 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 48fc023ab0a2..65bbc9339aca 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1304,6 +1304,55 @@ static int ext4_journalled_write_end(struct file *file, return ret ? ret : copied; } +/* + * Reserve a metadata for a single block located at lblock + */ +static int ext4_da_reserve_metadata(struct inode *inode, ext4_lblk_t lblock) +{ + int retries = 0; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct ext4_inode_info *ei = EXT4_I(inode); + unsigned int md_needed; + ext4_lblk_t save_last_lblock; + int save_len; + + /* + * recalculate the amount of metadata blocks to reserve + * in order to allocate nrblocks + * worse case is one extent per block + */ +repeat: + spin_lock(&ei->i_block_reservation_lock); + /* + * ext4_calc_metadata_amount() has side effects, which we have + * to be prepared undo if we fail to claim space. + */ + save_len = ei->i_da_metadata_calc_len; + save_last_lblock = ei->i_da_metadata_calc_last_lblock; + md_needed = EXT4_NUM_B2C(sbi, + ext4_calc_metadata_amount(inode, lblock)); + trace_ext4_da_reserve_space(inode, md_needed); + + /* + * We do still charge estimated metadata to the sb though; + * we cannot afford to run out of free blocks. + */ + if (ext4_claim_free_clusters(sbi, md_needed, 0)) { + ei->i_da_metadata_calc_len = save_len; + ei->i_da_metadata_calc_last_lblock = save_last_lblock; + spin_unlock(&ei->i_block_reservation_lock); + if (ext4_should_retry_alloc(inode->i_sb, &retries)) { + cond_resched(); + goto repeat; + } + return -ENOSPC; + } + ei->i_reserved_meta_blocks += md_needed; + spin_unlock(&ei->i_block_reservation_lock); + + return 0; /* success */ +} + /* * Reserve a single cluster located at lblock */ @@ -1940,8 +1989,11 @@ add_delayed: * XXX: __block_prepare_write() unmaps passed block, * is it OK? */ - /* If the block was allocated from previously allocated cluster, - * then we dont need to reserve it again. */ + /* + * If the block was allocated from previously allocated cluster, + * then we don't need to reserve it again. However we still need + * to reserve metadata for every block we're going to write. + */ if (!(map->m_flags & EXT4_MAP_FROM_CLUSTER)) { ret = ext4_da_reserve_space(inode, iblock); if (ret) { @@ -1949,6 +2001,13 @@ add_delayed: retval = ret; goto out_unlock; } + } else { + ret = ext4_da_reserve_metadata(inode, iblock); + if (ret) { + /* not enough space to reserve */ + retval = ret; + goto out_unlock; + } } ret = ext4_es_insert_extent(inode, map->m_lblk, map->m_len, -- cgit From bc077320f8f6449197f4010d6774afab7bb998b2 Mon Sep 17 00:00:00 2001 From: Marco Stornelli Date: Sat, 20 Oct 2012 12:02:59 +0200 Subject: hostfs: fix a not needed double check With the commit 3be2be0a32c18b0fd6d623cda63174a332ca0de1 we removed vmtruncate, but actaully there is no need to call inode_newsize_ok() because the checks are already done in inode_change_ok() at the begin of the function. Signed-off-by: Marco Stornelli Signed-off-by: Richard Weinberger --- fs/hostfs/hostfs_kern.c | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index fbabb906066f..178b90c229b5 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -845,15 +845,8 @@ int hostfs_setattr(struct dentry *dentry, struct iattr *attr) return err; if ((attr->ia_valid & ATTR_SIZE) && - attr->ia_size != i_size_read(inode)) { - int error; - - error = inode_newsize_ok(inode, attr->ia_size); - if (error) - return error; - + attr->ia_size != i_size_read(inode)) truncate_setsize(inode, attr->ia_size); - } setattr_copy(inode, attr); mark_inode_dirty(inode); -- cgit From 3e64fe5b21852375f2f53e7244ba697f1fee2fcf Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 11 Mar 2013 07:05:42 -0700 Subject: fs: Limit sys_mount to only request filesystem modules. (Part 3) Somehow I failed to add the MODULE_ALIAS_FS for cifs, hostfs, hpfs, squashfs, and udf despite what I thought were my careful checks :( Add them now. Signed-off-by: "Eric W. Biederman" --- fs/cifs/cifsfs.c | 1 + fs/hostfs/hostfs_kern.c | 1 + fs/hpfs/super.c | 1 + fs/squashfs/super.c | 1 + fs/udf/super.c | 1 + 5 files changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 1a052c0eee8e..3cf8a15af916 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -777,6 +777,7 @@ struct file_system_type cifs_fs_type = { .kill_sb = cifs_kill_sb, /* .fs_flags */ }; +MODULE_ALIAS_FS("cifs"); const struct inode_operations cifs_dir_inode_ops = { .create = cifs_create, .atomic_open = cifs_atomic_open, diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index fbabb906066f..e3c6d504a2b7 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -993,6 +993,7 @@ static struct file_system_type hostfs_type = { .kill_sb = hostfs_kill_sb, .fs_flags = 0, }; +MODULE_ALIAS_FS("hostfs"); static int __init init_hostfs(void) { diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c index a3076228523d..a0617e706957 100644 --- a/fs/hpfs/super.c +++ b/fs/hpfs/super.c @@ -688,6 +688,7 @@ static struct file_system_type hpfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("hpfs"); static int __init init_hpfs_fs(void) { diff --git a/fs/squashfs/super.c b/fs/squashfs/super.c index 260e3928d4f5..60553a9053ca 100644 --- a/fs/squashfs/super.c +++ b/fs/squashfs/super.c @@ -489,6 +489,7 @@ static struct file_system_type squashfs_fs_type = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV }; +MODULE_ALIAS_FS("squashfs"); static const struct super_operations squashfs_super_ops = { .alloc_inode = squashfs_alloc_inode, diff --git a/fs/udf/super.c b/fs/udf/super.c index bc5b30a819e8..9ac4057a86c9 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -118,6 +118,7 @@ static struct file_system_type udf_fstype = { .kill_sb = kill_block_super, .fs_flags = FS_REQUIRES_DEV, }; +MODULE_ALIAS_FS("udf"); static struct kmem_cache *udf_inode_cachep; -- cgit From ad56edad089b56300fd13bb9eeb7d0424d978239 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 11 Mar 2013 13:24:56 -0400 Subject: jbd2: fix use after free in jbd2_journal_dirty_metadata() jbd2_journal_dirty_metadata() didn't get a reference to journal_head it was working with. This is OK in most of the cases since the journal head should be attached to a transaction but in rare occasions when we are journalling data, __ext4_journalled_writepage() can race with jbd2_journal_invalidatepage() stripping buffers from a page and thus journal head can be freed under hands of jbd2_journal_dirty_metadata(). Fix the problem by getting own journal head reference in jbd2_journal_dirty_metadata() (and also in jbd2_journal_set_triggers() which can possibly have the same issue). Reported-by: Zheng Liu Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/jbd2/transaction.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index d6ee5aed56b1..325bc019ed88 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -1065,9 +1065,12 @@ out: void jbd2_journal_set_triggers(struct buffer_head *bh, struct jbd2_buffer_trigger_type *type) { - struct journal_head *jh = bh2jh(bh); + struct journal_head *jh = jbd2_journal_grab_journal_head(bh); + if (WARN_ON(!jh)) + return; jh->b_triggers = type; + jbd2_journal_put_journal_head(jh); } void jbd2_buffer_frozen_trigger(struct journal_head *jh, void *mapped_data, @@ -1119,17 +1122,18 @@ int jbd2_journal_dirty_metadata(handle_t *handle, struct buffer_head *bh) { transaction_t *transaction = handle->h_transaction; journal_t *journal = transaction->t_journal; - struct journal_head *jh = bh2jh(bh); + struct journal_head *jh; int ret = 0; - jbd_debug(5, "journal_head %p\n", jh); - JBUFFER_TRACE(jh, "entry"); if (is_handle_aborted(handle)) goto out; - if (!buffer_jbd(bh)) { + jh = jbd2_journal_grab_journal_head(bh); + if (!jh) { ret = -EUCLEAN; goto out; } + jbd_debug(5, "journal_head %p\n", jh); + JBUFFER_TRACE(jh, "entry"); jbd_lock_bh_state(bh); @@ -1220,6 +1224,7 @@ int jbd2_journal_dirty_metadata(handle_t *handle, struct buffer_head *bh) spin_unlock(&journal->j_list_lock); out_unlock_bh: jbd_unlock_bh_state(bh); + jbd2_journal_put_journal_head(jh); out: JBUFFER_TRACE(jh, "exit"); WARN_ON(ret); /* All errors are bugs, so dump the stack */ -- cgit From 68ac8bfb6aa4ce1a146b5a291f03fc2dd6d17cf2 Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Wed, 27 Feb 2013 14:43:09 -0500 Subject: quota: add missing use of dq_data_lock in __dquot_initialize The bulk of __dquot_initialize runs under the dqptr_sem which protects the inode->i_dquot pointers. It doesn't protect the dereferenced contents, though. Those are protected by the dq_data_lock, which is missing around the dquot_resv_space call. Signed-off-by: Jeff Mahoney Signed-off-by: Jan Kara --- fs/quota/dquot.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 05ae3c97f7a5..3e64169ef527 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -1439,8 +1439,11 @@ static void __dquot_initialize(struct inode *inode, int type) * did a write before quota was turned on */ rsv = inode_get_rsv_space(inode); - if (unlikely(rsv)) + if (unlikely(rsv)) { + spin_lock(&dq_data_lock); dquot_resv_space(inode->i_dquot[cnt], rsv); + spin_unlock(&dq_data_lock); + } } } out_err: -- cgit From 8d0c2d10dd72c5292eda7a06231056a4c972e4cc Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Sat, 9 Mar 2013 15:28:44 +0100 Subject: ext3: Fix format string issues ext3_msg() takes the printk prefix as the second parameter and the format string as the third parameter. Two callers of ext3_msg omit the prefix and pass the format string as the second parameter and the first parameter to the format string as the third parameter. In both cases this string comes from an arbitrary source. Which means the string may contain format string characters, which will lead to undefined and potentially harmful behavior. The issue was introduced in commit 4cf46b67eb("ext3: Unify log messages in ext3") and is fixed by this patch. CC: stable@vger.kernel.org Signed-off-by: Lars-Peter Clausen Signed-off-by: Jan Kara --- fs/ext3/super.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 1d6e2ed85322..fb5120a5505c 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -353,7 +353,7 @@ static struct block_device *ext3_blkdev_get(dev_t dev, struct super_block *sb) return bdev; fail: - ext3_msg(sb, "error: failed to open journal device %s: %ld", + ext3_msg(sb, KERN_ERR, "error: failed to open journal device %s: %ld", __bdevname(dev, b), PTR_ERR(bdev)); return NULL; @@ -887,7 +887,7 @@ static ext3_fsblk_t get_sb_block(void **data, struct super_block *sb) /*todo: use simple_strtoll with >32bit ext3 */ sb_block = simple_strtoul(options, &options, 0); if (*options && *options != ',') { - ext3_msg(sb, "error: invalid sb specification: %s", + ext3_msg(sb, KERN_ERR, "error: invalid sb specification: %s", (char *) *data); return 1; } -- cgit From af591ad896ef75585752ac2eab4fba9437f23322 Mon Sep 17 00:00:00 2001 From: Ionut-Gabriel Radu Date: Sun, 10 Mar 2013 15:06:23 +0200 Subject: reiserfs: Use kstrdup instead of kmalloc/strcpy Signed-off-by: Ionut-Gabriel Radu Signed-off-by: Jan Kara --- fs/reiserfs/super.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 194113b1b11b..f8a23c3078f8 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -1147,8 +1147,7 @@ static int reiserfs_parse_options(struct super_block *s, char *options, /* strin "on filesystem root."); return 0; } - qf_names[qtype] = - kmalloc(strlen(arg) + 1, GFP_KERNEL); + qf_names[qtype] = kstrdup(arg, GFP_KERNEL); if (!qf_names[qtype]) { reiserfs_warning(s, "reiserfs-2502", "not enough memory " @@ -1156,7 +1155,6 @@ static int reiserfs_parse_options(struct super_block *s, char *options, /* strin "quotafile name."); return 0; } - strcpy(qf_names[qtype], arg); if (qtype == USRQUOTA) *mount_options |= 1 << REISERFS_USRQUOTA; else -- cgit From 90ba983f6889e65a3b506b30dc606aa9d1d46cd2 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 11 Mar 2013 23:39:59 -0400 Subject: ext4: use atomic64_t for the per-flexbg free_clusters count A user who was using a 8TB+ file system and with a very large flexbg size (> 65536) could cause the atomic_t used in the struct flex_groups to overflow. This was detected by PaX security patchset: http://forums.grsecurity.net/viewtopic.php?f=3&t=3289&p=12551#p12551 This bug was introduced in commit 9f24e4208f7e, so it's been around since 2.6.30. :-( Fix this by using an atomic64_t for struct orlav_stats's free_clusters. Signed-off-by: "Theodore Ts'o" Reviewed-by: Lukas Czerner Cc: stable@vger.kernel.org --- fs/ext4/ext4.h | 6 +++--- fs/ext4/ialloc.c | 4 ++-- fs/ext4/mballoc.c | 12 ++++++------ fs/ext4/resize.c | 4 ++-- fs/ext4/super.c | 4 ++-- 5 files changed, 15 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 4a01ba315262..167ff564bbfa 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -335,9 +335,9 @@ struct ext4_group_desc */ struct flex_groups { - atomic_t free_inodes; - atomic_t free_clusters; - atomic_t used_dirs; + atomic64_t free_clusters; + atomic_t free_inodes; + atomic_t used_dirs; }; #define EXT4_BG_INODE_UNINIT 0x0001 /* Inode table/bitmap not in use */ diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 32fd2b9075dd..6c5bb8d993fe 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -324,8 +324,8 @@ error_return: } struct orlov_stats { + __u64 free_clusters; __u32 free_inodes; - __u32 free_clusters; __u32 used_dirs; }; @@ -342,7 +342,7 @@ static void get_orlov_stats(struct super_block *sb, ext4_group_t g, if (flex_size > 1) { stats->free_inodes = atomic_read(&flex_group[g].free_inodes); - stats->free_clusters = atomic_read(&flex_group[g].free_clusters); + stats->free_clusters = atomic64_read(&flex_group[g].free_clusters); stats->used_dirs = atomic_read(&flex_group[g].used_dirs); return; } diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 8b2ea9f75004..ee6614bdb639 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2804,8 +2804,8 @@ ext4_mb_mark_diskspace_used(struct ext4_allocation_context *ac, if (sbi->s_log_groups_per_flex) { ext4_group_t flex_group = ext4_flex_group(sbi, ac->ac_b_ex.fe_group); - atomic_sub(ac->ac_b_ex.fe_len, - &sbi->s_flex_groups[flex_group].free_clusters); + atomic64_sub(ac->ac_b_ex.fe_len, + &sbi->s_flex_groups[flex_group].free_clusters); } err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); @@ -4661,8 +4661,8 @@ do_more: if (sbi->s_log_groups_per_flex) { ext4_group_t flex_group = ext4_flex_group(sbi, block_group); - atomic_add(count_clusters, - &sbi->s_flex_groups[flex_group].free_clusters); + atomic64_add(count_clusters, + &sbi->s_flex_groups[flex_group].free_clusters); } ext4_mb_unload_buddy(&e4b); @@ -4804,8 +4804,8 @@ int ext4_group_add_blocks(handle_t *handle, struct super_block *sb, if (sbi->s_log_groups_per_flex) { ext4_group_t flex_group = ext4_flex_group(sbi, block_group); - atomic_add(EXT4_NUM_B2C(sbi, blocks_freed), - &sbi->s_flex_groups[flex_group].free_clusters); + atomic64_add(EXT4_NUM_B2C(sbi, blocks_freed), + &sbi->s_flex_groups[flex_group].free_clusters); } ext4_mb_unload_buddy(&e4b); diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index b2c8ee56eb98..c169477a62c9 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -1360,8 +1360,8 @@ static void ext4_update_super(struct super_block *sb, sbi->s_log_groups_per_flex) { ext4_group_t flex_group; flex_group = ext4_flex_group(sbi, group_data[0].group); - atomic_add(EXT4_NUM_B2C(sbi, free_blocks), - &sbi->s_flex_groups[flex_group].free_clusters); + atomic64_add(EXT4_NUM_B2C(sbi, free_blocks), + &sbi->s_flex_groups[flex_group].free_clusters); atomic_add(EXT4_INODES_PER_GROUP(sb) * flex_gd->count, &sbi->s_flex_groups[flex_group].free_inodes); } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 9379b7fbfd92..d1ee6a84338a 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1923,8 +1923,8 @@ static int ext4_fill_flex_info(struct super_block *sb) flex_group = ext4_flex_group(sbi, i); atomic_add(ext4_free_inodes_count(sb, gdp), &sbi->s_flex_groups[flex_group].free_inodes); - atomic_add(ext4_free_group_clusters(sb, gdp), - &sbi->s_flex_groups[flex_group].free_clusters); + atomic64_add(ext4_free_group_clusters(sb, gdp), + &sbi->s_flex_groups[flex_group].free_clusters); atomic_add(ext4_used_dirs_count(sb, gdp), &sbi->s_flex_groups[flex_group].used_dirs); } -- cgit From a930d8790552658140d7d0d2e316af4f0d76a512 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 12 Mar 2013 02:59:49 +0000 Subject: vfs: fix pipe counter breakage If you open a pipe for neither read nor write, the pipe code will not add any usage counters to the pipe, causing the 'struct pipe_inode_info" to be potentially released early. That doesn't normally matter, since you cannot actually use the pipe, but the pipe release code - particularly fasync handling - still expects the actual pipe infrastructure to all be there. And rather than adding NULL pointer checks, let's just disallow this case, the same way we already do for the named pipe ("fifo") case. This is ancient going back to pre-2.4 days, and until trinity, nobody naver noticed. Reported-by: Dave Jones Signed-off-by: Linus Torvalds --- fs/pipe.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/pipe.c b/fs/pipe.c index 64a494cef0a0..2234f3f61f8d 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -863,6 +863,9 @@ pipe_rdwr_open(struct inode *inode, struct file *filp) { int ret = -ENOENT; + if (!(filp->f_mode & (FMODE_READ|FMODE_WRITE))) + return -EINVAL; + mutex_lock(&inode->i_mutex); if (inode->i_pipe) { -- cgit From 4f42f80a8f08d4c3f52c4267361241885d5dee3a Mon Sep 17 00:00:00 2001 From: Lukas Czerner Date: Tue, 12 Mar 2013 12:40:04 -0400 Subject: ext4: use s_extent_max_zeroout_kb value as number of kb Currently when converting extent to initialized, we have to decide whether to zeroout part/all of the uninitialized extent in order to avoid extent tree growing rapidly. The decision is made by comparing the size of the extent with the configurable value s_extent_max_zeroout_kb which is in kibibytes units. However when converting it to number of blocks we currently use it as it was in bytes. This is obviously bug and it will result in ext4 _never_ zeroout extents, but rather always split and convert parts to initialized while leaving the rest uninitialized in default setting. Fix this by using s_extent_max_zeroout_kb as kibibytes. Signed-off-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" Cc: stable@vger.kernel.org --- fs/ext4/extents.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index bd69e906bd91..e2bb929bea93 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3264,7 +3264,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle, if (EXT4_EXT_MAY_ZEROOUT & split_flag) max_zeroout = sbi->s_extent_max_zeroout_kb >> - inode->i_sb->s_blocksize_bits; + (inode->i_sb->s_blocksize_bits - 10); /* If extent is less than s_max_zeroout_kb, zeroout directly */ if (max_zeroout && (ee_len <= max_zeroout)) { -- cgit From 8aec0f5d4137532de14e6554fd5dd201ff3a3c49 Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 25 Feb 2013 10:20:36 -0500 Subject: Fix: compat_rw_copy_check_uvector() misuse in aio, readv, writev, and security keys Looking at mm/process_vm_access.c:process_vm_rw() and comparing it to compat_process_vm_rw() shows that the compatibility code requires an explicit "access_ok()" check before calling compat_rw_copy_check_uvector(). The same difference seems to appear when we compare fs/read_write.c:do_readv_writev() to fs/compat.c:compat_do_readv_writev(). This subtle difference between the compat and non-compat requirements should probably be debated, as it seems to be error-prone. In fact, there are two others sites that use this function in the Linux kernel, and they both seem to get it wrong: Now shifting our attention to fs/aio.c, we see that aio_setup_iocb() also ends up calling compat_rw_copy_check_uvector() through aio_setup_vectored_rw(). Unfortunately, the access_ok() check appears to be missing. Same situation for security/keys/compat.c:compat_keyctl_instantiate_key_iov(). I propose that we add the access_ok() check directly into compat_rw_copy_check_uvector(), so callers don't have to worry about it, and it therefore makes the compat call code similar to its non-compat counterpart. Place the access_ok() check in the same location where copy_from_user() can trigger a -EFAULT error in the non-compat code, so the ABI behaviors are alike on both compat and non-compat. While we are here, fix compat_do_readv_writev() so it checks for compat_rw_copy_check_uvector() negative return values. And also, fix a memory leak in compat_keyctl_instantiate_key_iov() error handling. Acked-by: Linus Torvalds Acked-by: Al Viro Signed-off-by: Mathieu Desnoyers Signed-off-by: Linus Torvalds --- fs/compat.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/compat.c b/fs/compat.c index fe40fde29111..d487985dd0ea 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -558,6 +558,10 @@ ssize_t compat_rw_copy_check_uvector(int type, } *ret_pointer = iov; + ret = -EFAULT; + if (!access_ok(VERIFY_READ, uvector, nr_segs*sizeof(*uvector))) + goto out; + /* * Single unix specification: * We should -EINVAL if an element length is not >= 0 and fitting an @@ -1080,17 +1084,12 @@ static ssize_t compat_do_readv_writev(int type, struct file *file, if (!file->f_op) goto out; - ret = -EFAULT; - if (!access_ok(VERIFY_READ, uvector, nr_segs*sizeof(*uvector))) - goto out; - - tot_len = compat_rw_copy_check_uvector(type, uvector, nr_segs, + ret = compat_rw_copy_check_uvector(type, uvector, nr_segs, UIO_FASTIOV, iovstack, &iov); - if (tot_len == 0) { - ret = 0; + if (ret <= 0) goto out; - } + tot_len = ret; ret = rw_verify_area(type, file, pos, tot_len); if (ret < 0) goto out; -- cgit From fa7614ddd6c2368b8cd54cc67ab4b767af0a2a50 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 12 Mar 2013 18:27:41 -0700 Subject: fs: Readd the fs module aliases. I had assumed that the only use of module aliases for filesystems prior to "fs: Limit sys_mount to only request filesystem modules." was in request_module. It turns out I was wrong. At least mkinitcpio in Arch linux uses these aliases. So readd the preexising aliases, to keep from breaking userspace. Userspace eventually will have to follow and use the same aliases the kernel does. So at some point we may be delete these aliases without problems. However that day is not today. Signed-off-by: "Eric W. Biederman" --- fs/ext4/super.c | 2 ++ fs/freevxfs/vxfs_super.c | 1 + fs/isofs/inode.c | 1 + fs/nfs/super.c | 1 + fs/sysv/super.c | 1 + 5 files changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 34e855219231..b3818b48f418 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -91,6 +91,7 @@ static struct file_system_type ext2_fs_type = { .fs_flags = FS_REQUIRES_DEV, }; MODULE_ALIAS_FS("ext2"); +MODULE_ALIAS("ext2"); #define IS_EXT2_SB(sb) ((sb)->s_bdev->bd_holder == &ext2_fs_type) #else #define IS_EXT2_SB(sb) (0) @@ -106,6 +107,7 @@ static struct file_system_type ext3_fs_type = { .fs_flags = FS_REQUIRES_DEV, }; MODULE_ALIAS_FS("ext3"); +MODULE_ALIAS("ext3"); #define IS_EXT3_SB(sb) ((sb)->s_bdev->bd_holder == &ext3_fs_type) #else #define IS_EXT3_SB(sb) (0) diff --git a/fs/freevxfs/vxfs_super.c b/fs/freevxfs/vxfs_super.c index 455074308069..e37eb274e492 100644 --- a/fs/freevxfs/vxfs_super.c +++ b/fs/freevxfs/vxfs_super.c @@ -258,6 +258,7 @@ static struct file_system_type vxfs_fs_type = { .fs_flags = FS_REQUIRES_DEV, }; MODULE_ALIAS_FS("vxfs"); /* makes mount -t vxfs autoload the module */ +MODULE_ALIAS("vxfs"); static int __init vxfs_init(void) diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index a67f16e846a2..d9b8aebdeb22 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -1557,6 +1557,7 @@ static struct file_system_type iso9660_fs_type = { .fs_flags = FS_REQUIRES_DEV, }; MODULE_ALIAS_FS("iso9660"); +MODULE_ALIAS("iso9660"); static int __init init_iso9660_fs(void) { diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 95cdcb208dfb..2f8a29db0f1b 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -335,6 +335,7 @@ struct file_system_type nfs4_fs_type = { .fs_flags = FS_RENAME_DOES_D_MOVE|FS_BINARY_MOUNTDATA, }; MODULE_ALIAS_FS("nfs4"); +MODULE_ALIAS("nfs4"); EXPORT_SYMBOL_GPL(nfs4_fs_type); static int __init register_nfs4_fs(void) diff --git a/fs/sysv/super.c b/fs/sysv/super.c index a39938b1feea..d0c6a007ce83 100644 --- a/fs/sysv/super.c +++ b/fs/sysv/super.c @@ -555,6 +555,7 @@ static struct file_system_type v7_fs_type = { .fs_flags = FS_REQUIRES_DEV, }; MODULE_ALIAS_FS("v7"); +MODULE_ALIAS("v7"); static int __init init_sysv_fs(void) { -- cgit From c288d2969627be7ffc90904ac8c6aae0295fbf9f Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 13 Mar 2013 12:57:08 +0100 Subject: ext2: Fix BUG_ON in evict() on inode deletion Commit 8e3dffc6 introduced a regression where deleting inode with large extended attributes leads to triggering BUG_ON(inode->i_state != (I_FREEING | I_CLEAR)) in fs/inode.c:evict(). That happens because freeing of xattr block dirtied the inode and it happened after clear_inode() has been called. Fix the issue by moving removal of xattr block into ext2_evict_inode() before clear_inode() call close to a place where data blocks are truncated. That is also more logical place and removes surprising requirement that ext2_free_blocks() mustn't dirty the inode. Reported-by: Tyler Hicks Signed-off-by: Jan Kara --- fs/ext2/ialloc.c | 1 - fs/ext2/inode.c | 2 ++ 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext2/ialloc.c b/fs/ext2/ialloc.c index 8f370e012e61..7cadd823bb31 100644 --- a/fs/ext2/ialloc.c +++ b/fs/ext2/ialloc.c @@ -118,7 +118,6 @@ void ext2_free_inode (struct inode * inode) * as writing the quota to disk may need the lock as well. */ /* Quota is already initialized in iput() */ - ext2_xattr_delete_inode(inode); dquot_free_inode(inode); dquot_drop(inode); diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index c3881e56662e..fe60cc1117d8 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -34,6 +34,7 @@ #include "ext2.h" #include "acl.h" #include "xip.h" +#include "xattr.h" static int __ext2_write_inode(struct inode *inode, int do_sync); @@ -88,6 +89,7 @@ void ext2_evict_inode(struct inode * inode) inode->i_size = 0; if (inode->i_blocks) ext2_truncate_blocks(inode, 0); + ext2_xattr_delete_inode(inode); } invalidate_inode_buffers(inode); -- cgit From 47c78f4a70d791ff44cab3254b489605a52e3181 Mon Sep 17 00:00:00 2001 From: Sachin Prabhu Date: Mon, 11 Mar 2013 13:08:49 +0000 Subject: cifs: map NT_STATUS_SHARING_VIOLATION to EBUSY instead of ETXTBSY NT_SHARING_VIOLATION errors are mapped to ETXTBSY which is unexpected for operations such as unlink where we can hit these errors. The patch maps the error NT_SHARING_VIOLATION to EBUSY instead. The patch also replaces all instances of ETXTBSY in cifs_rename_pending_delete() with EBUSY. Signed-off-by: Sachin Prabhu Reviewed-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/inode.c | 10 ++++------ fs/cifs/netmisc.c | 2 +- 2 files changed, 5 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 0079696305c9..20887bf63121 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1043,7 +1043,7 @@ cifs_rename_pending_delete(const char *full_path, struct dentry *dentry, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc != 0) { - rc = -ETXTBSY; + rc = -EBUSY; goto undo_setattr; } @@ -1062,7 +1062,7 @@ cifs_rename_pending_delete(const char *full_path, struct dentry *dentry, if (rc == -ENOENT) rc = 0; else if (rc != 0) { - rc = -ETXTBSY; + rc = -EBUSY; goto undo_rename; } cifsInode->delete_pending = true; @@ -1169,15 +1169,13 @@ psx_del_no_retry: cifs_drop_nlink(inode); } else if (rc == -ENOENT) { d_drop(dentry); - } else if (rc == -ETXTBSY) { + } else if (rc == -EBUSY) { if (server->ops->rename_pending_delete) { rc = server->ops->rename_pending_delete(full_path, dentry, xid); if (rc == 0) cifs_drop_nlink(inode); } - if (rc == -ETXTBSY) - rc = -EBUSY; } else if ((rc == -EACCES) && (dosattr == 0) && inode) { attrs = kzalloc(sizeof(*attrs), GFP_KERNEL); if (attrs == NULL) { @@ -1518,7 +1516,7 @@ cifs_do_rename(const unsigned int xid, struct dentry *from_dentry, * source. Note that cross directory moves do not work with * rename by filehandle to various Windows servers. */ - if (rc == 0 || rc != -ETXTBSY) + if (rc == 0 || rc != -EBUSY) goto do_rename_exit; /* open-file renames don't work across directories */ diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index a82bc51fdc82..c0b25b28be6c 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -62,7 +62,7 @@ static const struct smb_to_posix_error mapping_table_ERRDOS[] = { {ERRdiffdevice, -EXDEV}, {ERRnofiles, -ENOENT}, {ERRwriteprot, -EROFS}, - {ERRbadshare, -ETXTBSY}, + {ERRbadshare, -EBUSY}, {ERRlock, -EACCES}, {ERRunsup, -EINVAL}, {ERRnosuchshare, -ENXIO}, -- cgit From 24261fc23db950951760d00c188ba63cc756b932 Mon Sep 17 00:00:00 2001 From: Mateusz Guzik Date: Fri, 8 Mar 2013 16:30:03 +0100 Subject: cifs: delay super block destruction until all cifsFileInfo objects are gone cifsFileInfo objects hold references to dentries and it is possible that these will still be around in workqueues when VFS decides to kill super block during unmount. This results in panics like this one: BUG: Dentry ffff88001f5e76c0{i=66b4a,n=1M-2} still in use (1) [unmount of cifs cifs] ------------[ cut here ]------------ kernel BUG at fs/dcache.c:943! [..] Process umount (pid: 1781, threadinfo ffff88003d6e8000, task ffff880035eeaec0) [..] Call Trace: [] shrink_dcache_for_umount+0x33/0x60 [] generic_shutdown_super+0x2c/0xe0 [] kill_anon_super+0x16/0x30 [] cifs_kill_sb+0x1a/0x30 [cifs] [] deactivate_locked_super+0x57/0x80 [] deactivate_super+0x4e/0x70 [] mntput_no_expire+0xd7/0x130 [] sys_umount+0x9c/0x3c0 [] system_call_fastpath+0x16/0x1b Fix this by making each cifsFileInfo object hold a reference to cifs super block, which implicitly keeps VFS super block around as well. Signed-off-by: Mateusz Guzik Reviewed-by: Jeff Layton Cc: Reported-and-Tested-by: Ben Greear Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 24 ++++++++++++++++++++++++ fs/cifs/cifsfs.h | 4 ++++ fs/cifs/file.c | 6 +++++- 3 files changed, 33 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 1a052c0eee8e..054b90b682a7 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -91,6 +91,30 @@ struct workqueue_struct *cifsiod_wq; __u8 cifs_client_guid[SMB2_CLIENT_GUID_SIZE]; #endif +/* + * Bumps refcount for cifs super block. + * Note that it should be only called if a referece to VFS super block is + * already held, e.g. in open-type syscalls context. Otherwise it can race with + * atomic_dec_and_test in deactivate_locked_super. + */ +void +cifs_sb_active(struct super_block *sb) +{ + struct cifs_sb_info *server = CIFS_SB(sb); + + if (atomic_inc_return(&server->active) == 1) + atomic_inc(&sb->s_active); +} + +void +cifs_sb_deactive(struct super_block *sb) +{ + struct cifs_sb_info *server = CIFS_SB(sb); + + if (atomic_dec_and_test(&server->active)) + deactivate_super(sb); +} + static int cifs_read_super(struct super_block *sb) { diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index 7163419cecd9..0e32c3446ce9 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -41,6 +41,10 @@ extern struct file_system_type cifs_fs_type; extern const struct address_space_operations cifs_addr_ops; extern const struct address_space_operations cifs_addr_ops_smallbuf; +/* Functions related to super block operations */ +extern void cifs_sb_active(struct super_block *sb); +extern void cifs_sb_deactive(struct super_block *sb); + /* Functions related to inodes */ extern const struct inode_operations cifs_dir_inode_ops; extern struct inode *cifs_root_iget(struct super_block *); diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 8c0d85577314..7a0dd99e4507 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -300,6 +300,8 @@ cifs_new_fileinfo(struct cifs_fid *fid, struct file *file, INIT_WORK(&cfile->oplock_break, cifs_oplock_break); mutex_init(&cfile->fh_mutex); + cifs_sb_active(inode->i_sb); + /* * If the server returned a read oplock and we have mandatory brlocks, * set oplock level to None. @@ -349,7 +351,8 @@ void cifsFileInfo_put(struct cifsFileInfo *cifs_file) struct cifs_tcon *tcon = tlink_tcon(cifs_file->tlink); struct TCP_Server_Info *server = tcon->ses->server; struct cifsInodeInfo *cifsi = CIFS_I(inode); - struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); + struct super_block *sb = inode->i_sb; + struct cifs_sb_info *cifs_sb = CIFS_SB(sb); struct cifsLockInfo *li, *tmp; struct cifs_fid fid; struct cifs_pending_open open; @@ -414,6 +417,7 @@ void cifsFileInfo_put(struct cifsFileInfo *cifs_file) cifs_put_tlink(cifs_file->tlink); dput(cifs_file->dentry); + cifs_sb_deactive(sb); kfree(cifs_file); } -- cgit From 801cb2d62d5f673ac671c01397956d8b77a5215b Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 13 Mar 2013 14:59:36 -0700 Subject: nfsd: remove unused get_new_stid() get_new_stid() is no longer used since commit 3abdb607125 ("nfsd4: simplify idr allocation"). Remove it. Signed-off-by: Tejun Heo Acked-by: J. Bruce Fields Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/nfsd/nfs4state.c | 31 ------------------------------- 1 file changed, 31 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 16d39c6c4fbb..d91d6dbf698a 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -230,37 +230,6 @@ static void nfs4_file_put_access(struct nfs4_file *fp, int oflag) __nfs4_file_put_access(fp, oflag); } -static inline int get_new_stid(struct nfs4_stid *stid) -{ - static int min_stateid = 0; - struct idr *stateids = &stid->sc_client->cl_stateids; - int new_stid; - int error; - - error = idr_get_new_above(stateids, stid, min_stateid, &new_stid); - /* - * Note: the necessary preallocation was done in - * nfs4_alloc_stateid(). The idr code caps the number of - * preallocations that can exist at a time, but the state lock - * prevents anyone from using ours before we get here: - */ - WARN_ON_ONCE(error); - /* - * It shouldn't be a problem to reuse an opaque stateid value. - * I don't think it is for 4.1. But with 4.0 I worry that, for - * example, a stray write retransmission could be accepted by - * the server when it should have been rejected. Therefore, - * adopt a trick from the sctp code to attempt to maximize the - * amount of time until an id is reused, by ensuring they always - * "increase" (mod INT_MAX): - */ - - min_stateid = new_stid+1; - if (min_stateid == INT_MAX) - min_stateid = 0; - return new_stid; -} - static struct nfs4_stid *nfs4_alloc_stid(struct nfs4_client *cl, struct kmem_cache *slab) { -- cgit From ebd6c70714f5eda9cd1b60d23754ffd1d62481f6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 13 Mar 2013 14:59:37 -0700 Subject: nfsd: convert to idr_alloc() idr_get_new*() and friends are about to be deprecated. Convert to the new idr_alloc() interface. Only compile-tested. Signed-off-by: Tejun Heo Acked-by: J. Bruce Fields Tested-by: J. Bruce Fields Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/nfsd/nfs4state.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d91d6dbf698a..2e27430b9070 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -242,9 +242,8 @@ kmem_cache *slab) if (!stid) return NULL; - if (!idr_pre_get(stateids, GFP_KERNEL)) - goto out_free; - if (idr_get_new_above(stateids, stid, min_stateid, &new_id)) + new_id = idr_alloc(stateids, stid, min_stateid, 0, GFP_KERNEL); + if (new_id < 0) goto out_free; stid->sc_client = cl; stid->sc_type = 0; -- cgit From a09a0a705dd6c80bc96b5e6f18dc103d4e1a7d63 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 11 Mar 2013 09:20:58 +0000 Subject: Btrfs: get better concurrency for snapshot-aware defrag work Using spinning case instead of blocking will result in better concurrency overall. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 13ab4de0a400..1f26888825e2 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2312,6 +2312,7 @@ again: key.type = BTRFS_EXTENT_DATA_KEY; key.offset = start; + path->leave_spinning = 1; if (merge) { struct btrfs_file_extent_item *fi; u64 extent_len; @@ -2368,6 +2369,7 @@ again: btrfs_mark_buffer_dirty(leaf); inode_add_bytes(inode, len); + btrfs_release_path(path); ret = btrfs_inc_extent_ref(trans, root, new->bytenr, new->disk_len, 0, @@ -2381,6 +2383,7 @@ again: ret = 1; out_free_path: btrfs_release_path(path); + path->leave_spinning = 0; btrfs_end_transaction(trans, root); out_unlock: unlock_extent_cached(&BTRFS_I(inode)->io_tree, lock_start, lock_end, -- cgit From d340d2475c6e394013325f83f499594628a9e558 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 11 Mar 2013 09:37:45 +0000 Subject: Btrfs: remove btrfs_try_spin_lock Remove a useless function declaration Signed-off-by: Liu Bo Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/locking.h | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/locking.h b/fs/btrfs/locking.h index ca52681e5f40..b81e0e9a4894 100644 --- a/fs/btrfs/locking.h +++ b/fs/btrfs/locking.h @@ -26,7 +26,6 @@ void btrfs_tree_lock(struct extent_buffer *eb); void btrfs_tree_unlock(struct extent_buffer *eb); -int btrfs_try_spin_lock(struct extent_buffer *eb); void btrfs_tree_read_lock(struct extent_buffer *eb); void btrfs_tree_read_unlock(struct extent_buffer *eb); -- cgit From bc178622d40d87e75abc131007342429c9b03351 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Sat, 9 Mar 2013 15:18:39 +0000 Subject: btrfs: use rcu_barrier() to wait for bdev puts at unmount Doing this would reliably fail with -EBUSY for me: # mount /dev/sdb2 /mnt/scratch; umount /mnt/scratch; mkfs.btrfs -f /dev/sdb2 ... unable to open /dev/sdb2: Device or resource busy because mkfs.btrfs tries to open the device O_EXCL, and somebody still has it. Using systemtap to track bdev gets & puts shows a kworker thread doing a blkdev put after mkfs attempts a get; this is left over from the unmount path: btrfs_close_devices __btrfs_close_devices call_rcu(&device->rcu, free_device); free_device INIT_WORK(&device->rcu_work, __free_device); schedule_work(&device->rcu_work); so unmount might complete before __free_device fires & does its blkdev_put. Adding an rcu_barrier() to btrfs_close_devices() causes unmount to wait until all blkdev_put()s are done, and the device is truly free once unmount completes. Cc: stable@vger.kernel.org Signed-off-by: Eric Sandeen Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 6b9cff42265d..5989a92236f7 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -684,6 +684,12 @@ int btrfs_close_devices(struct btrfs_fs_devices *fs_devices) __btrfs_close_devices(fs_devices); free_fs_devices(fs_devices); } + /* + * Wait for rcu kworkers under __btrfs_close_devices + * to finish all blkdev_puts so device is really + * free when umount is done. + */ + rcu_barrier(); return ret; } -- cgit From 492104c866cb1b62a11393adccb477f5cd2c7768 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 8 Mar 2013 15:41:02 -0500 Subject: Btrfs: return EIO if we have extent tree corruption The callers of lookup_inline_extent_info all handle getting an error back properly, so return an error if we have corruption instead of being a jerk and panicing. Still WARN_ON() since this is kind of crucial and I've been seeing it a bit too much recently for my taste, I think we're doing something wrong somewhere. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index aaee2b7fee78..350b9b18140c 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -1467,8 +1467,11 @@ int lookup_inline_extent_backref(struct btrfs_trans_handle *trans, if (ret && !insert) { err = -ENOENT; goto out; + } else if (ret) { + err = -EIO; + WARN_ON(1); + goto out; } - BUG_ON(ret); /* Corruption */ leaf = path->nodes[0]; item_size = btrfs_item_size_nr(leaf, path->slots[0]); -- cgit From 720f1e2060138855b4a1b1e8aa642f9c7feb6750 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Wed, 6 Mar 2013 11:51:47 +0000 Subject: Btrfs: return as soon as possible when edquot happens If one of qgroup fails to reserve firstly, we should return immediately, it is unnecessary to continue check. Signed-off-by: Wang Shilong Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/qgroup.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index aee4b1cc3d98..5471e47d6559 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -1525,21 +1525,23 @@ int btrfs_qgroup_reserve(struct btrfs_root *root, u64 num_bytes) if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_RFER) && qg->reserved + qg->rfer + num_bytes > - qg->max_rfer) + qg->max_rfer) { ret = -EDQUOT; + goto out; + } if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_EXCL) && qg->reserved + qg->excl + num_bytes > - qg->max_excl) + qg->max_excl) { ret = -EDQUOT; + goto out; + } list_for_each_entry(glist, &qg->groups, next_group) { ulist_add(ulist, glist->group->qgroupid, (uintptr_t)glist->group, GFP_ATOMIC); } } - if (ret) - goto out; /* * no limits exceeded, now record the reservation into all qgroups -- cgit From 7c2ec3f0730729f4829d01f7c19538d135f86712 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Wed, 13 Mar 2013 07:43:03 -0600 Subject: Btrfs: fix warning when creating snapshots Creating snapshot passes extent_root to commit its transaction, but it can lead to the warning of checking root for quota in the __btrfs_end_transaction() when someone else is committing the current transaction. Since we've recorded the needed root in trans_handle, just use it to get rid of the warning. Signed-off-by: Liu Bo Signed-off-by: Chris Mason --- fs/btrfs/transaction.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index fedede1fe178..c4a1531c52d8 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -626,14 +626,13 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, btrfs_trans_release_metadata(trans, root); trans->block_rsv = NULL; - /* - * the same root has to be passed to start_transaction and - * end_transaction. Subvolume quota depends on this. - */ - WARN_ON(trans->root != root); if (trans->qgroup_reserved) { - btrfs_qgroup_free(root, trans->qgroup_reserved); + /* + * the same root has to be passed here between start_transaction + * and end_transaction. Subvolume quota depends on this. + */ + btrfs_qgroup_free(trans->root, trans->qgroup_reserved); trans->qgroup_reserved = 0; } -- cgit From 3b2775942d6ccb14342f3aae55f22fbbfea8db14 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Fri, 15 Mar 2013 08:46:39 -0600 Subject: Btrfs: fix warning of free_extent_map Users report that an extent map's list is still linked when it's actually going to be freed from cache. The story is that a) when we're going to drop an extent map and may split this large one into smaller ems, and if this large one is flagged as EXTENT_FLAG_LOGGING which means that it's on the list to be logged, then the smaller ems split from it will also be flagged as EXTENT_FLAG_LOGGING, and this is _not_ expected. b) we'll keep ems from unlinking the list and freeing when they are flagged with EXTENT_FLAG_LOGGING, because the log code holds one reference. The end result is the warning, but the truth is that we set the flag EXTENT_FLAG_LOGGING only during fsync. So clear flag EXTENT_FLAG_LOGGING for extent maps split from a large one. Reported-by: Johannes Hirte Reported-by: Darrick J. Wong Signed-off-by: Liu Bo Signed-off-by: Chris Mason --- fs/btrfs/file.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 83c790d84038..7bdb47faa12e 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -591,6 +591,7 @@ void btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end, } compressed = test_bit(EXTENT_FLAG_COMPRESSED, &em->flags); clear_bit(EXTENT_FLAG_PINNED, &em->flags); + clear_bit(EXTENT_FLAG_LOGGING, &flags); remove_extent_mapping(em_tree, em); if (no_splits) goto next; -- cgit From 0e401101db49959f5783f6ee9e676124b5a183ac Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Mon, 18 Mar 2013 11:40:19 -0400 Subject: ext4: fix memory leakage in mext_check_coverage Regression was introduced by following commit 8c854473 TESTCASE (git://oss.sgi.com/xfs/cmds/xfstests.git): #while true;do ./check 301 || break ;done Also fix potential memory leakage in get_ext_path() once ext4_ext_find_extent() have failed. Signed-off-by: Dmitry Monakhov Signed-off-by: "Theodore Ts'o" --- fs/ext4/move_extent.c | 35 +++++++++++++++++++---------------- 1 file changed, 19 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index c1f15b203e98..bbae4ed15c3d 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -32,16 +32,18 @@ */ static inline int get_ext_path(struct inode *inode, ext4_lblk_t lblock, - struct ext4_ext_path **path) + struct ext4_ext_path **orig_path) { int ret = 0; + struct ext4_ext_path *path; - *path = ext4_ext_find_extent(inode, lblock, *path); - if (IS_ERR(*path)) { - ret = PTR_ERR(*path); - *path = NULL; - } else if ((*path)[ext_depth(inode)].p_ext == NULL) + path = ext4_ext_find_extent(inode, lblock, *orig_path); + if (IS_ERR(path)) + ret = PTR_ERR(path); + else if (path[ext_depth(inode)].p_ext == NULL) ret = -ENODATA; + else + *orig_path = path; return ret; } @@ -611,24 +613,25 @@ mext_check_coverage(struct inode *inode, ext4_lblk_t from, ext4_lblk_t count, { struct ext4_ext_path *path = NULL; struct ext4_extent *ext; + int ret = 0; ext4_lblk_t last = from + count; while (from < last) { *err = get_ext_path(inode, from, &path); if (*err) - return 0; + goto out; ext = path[ext_depth(inode)].p_ext; - if (!ext) { - ext4_ext_drop_refs(path); - return 0; - } - if (uninit != ext4_ext_is_uninitialized(ext)) { - ext4_ext_drop_refs(path); - return 0; - } + if (uninit != ext4_ext_is_uninitialized(ext)) + goto out; from += ext4_ext_get_actual_len(ext); ext4_ext_drop_refs(path); } - return 1; + ret = 1; +out: + if (path) { + ext4_ext_drop_refs(path); + kfree(path); + } + return ret; } /** -- cgit From 83cdadd8b0559c93728d065d23ca3485fa567e54 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Fri, 22 Feb 2013 13:32:56 -0500 Subject: xfs: fix potential infinite loop in xfs_iomap_prealloc_size() If freesp == 0, we could end up in an infinite loop while squashing the preallocation. Break the loop when we've killed the prealloc entirely. Signed-off-by: Brian Foster Reviewed-by: Dave Chinner Signed-off-by: Ben Myers (cherry picked from commit e78c420bfc2608bb5f9a0b9165b1071c1e31166a) --- fs/xfs/xfs_iomap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 912d83d8860a..b0b0f448e843 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -413,7 +413,7 @@ xfs_iomap_prealloc_size( * have a large file on a small filesystem and the above * lowspace thresholds are smaller than MAXEXTLEN. */ - while (alloc_blocks >= freesp) + while (alloc_blocks && alloc_blocks >= freesp) alloc_blocks >>= 4; } -- cgit From 3325beed46d8d14d873e94d89ea57ee900dec942 Mon Sep 17 00:00:00 2001 From: Mark Tinguely Date: Sun, 24 Feb 2013 13:04:37 -0600 Subject: xfs: fix xfs_iomap_eof_prealloc_initial_size type Fix the return type of xfs_iomap_eof_prealloc_initial_size() to xfs_fsblock_t to reflect the fact that the return value may be an unsigned 64 bits if XFS_BIG_BLKNOS is defined. Signed-off-by: Mark Tinguely Reviewed-by: Dave Chinner Signed-off-by: Ben Myers (cherry picked from commit e8108cedb1c5d1dc359690d18ca997e97a0061d2) --- fs/xfs/xfs_iomap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index b0b0f448e843..5a30dd899d2b 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -325,7 +325,7 @@ xfs_iomap_eof_want_preallocate( * rather than falling short due to things like stripe unit/width alignment of * real extents. */ -STATIC int +STATIC xfs_fsblock_t xfs_iomap_eof_prealloc_initial_size( struct xfs_mount *mp, struct xfs_inode *ip, -- cgit From e001873853d87674dd5b3cfa2851885023616695 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 12 Mar 2013 23:30:34 +1100 Subject: xfs: ensure we capture IO errors correctly Failed buffer readahead can leave the buffer in the cache marked with an error. Most callers that then issue a subsequent read on the buffer do not zero the b_error field out, and so we may incorectly detect an error during IO completion due to the stale error value left on the buffer. Avoid this problem by zeroing the error before IO submission. This ensures that the only IO errors that are detected those captured from are those captured from bio submission or completion. Signed-off-by: Dave Chinner Reviewed-by: Mark Tinguely Signed-off-by: Ben Myers (cherry picked from commit c163f9a1760229a95d04e37b332de7d5c1c225cd) --- fs/xfs/xfs_buf.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 4e8f0df82d02..8459b5d8cb71 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -1334,6 +1334,12 @@ _xfs_buf_ioapply( int size; int i; + /* + * Make sure we capture only current IO errors rather than stale errors + * left over from previous use of the buffer (e.g. failed readahead). + */ + bp->b_error = 0; + if (bp->b_flags & XBF_WRITE) { if (bp->b_flags & XBF_SYNCIO) rw = WRITE_SYNC; -- cgit From a517b608fa3d9b65930ef53ffe4a2f9800e10f7d Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 18 Mar 2013 10:49:07 -0400 Subject: nfsd: only unhash DRC entries that are in the hashtable It's not safe to call hlist_del() on a newly initialized hlist_node. That leads to a NULL pointer dereference. Only do that if the entry is hashed. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 62c1ee128aeb..18509bd6f587 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -102,7 +102,8 @@ nfsd_reply_cache_free_locked(struct svc_cacherep *rp) { if (rp->c_type == RC_REPLBUFF) kfree(rp->c_replvec.iov_base); - hlist_del(&rp->c_hash); + if (!hlist_unhashed(&rp->c_hash)) + hlist_del(&rp->c_hash); list_del(&rp->c_lru); --num_drc_entries; kmem_cache_free(drc_slab, rp); -- cgit From ac534ff2d5508bdff1358a55d88053da729ff46b Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 15 Mar 2013 09:16:29 -0400 Subject: nfsd: fix startup order in nfsd_reply_cache_init If we end up doing "goto out_nomem" in this function, we'll call nfsd_reply_cache_shutdown. That will attempt to walk the LRU list and free entries, but that list may not be initialized yet if the server is starting up for the first time. It's also possible for the shrinker to kick in before we've initialized the LRU list. Rearrange the initialization so that the LRU list_head and cache size are initialized before doing any of the allocations that might fail. Signed-off-by: Jeff Layton Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 18509bd6f587..ca05f6dc3544 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -119,6 +119,10 @@ nfsd_reply_cache_free(struct svc_cacherep *rp) int nfsd_reply_cache_init(void) { + INIT_LIST_HEAD(&lru_head); + max_drc_entries = nfsd_cache_size_limit(); + num_drc_entries = 0; + register_shrinker(&nfsd_reply_cache_shrinker); drc_slab = kmem_cache_create("nfsd_drc", sizeof(struct svc_cacherep), 0, 0, NULL); @@ -129,10 +133,6 @@ int nfsd_reply_cache_init(void) if (!cache_hash) goto out_nomem; - INIT_LIST_HEAD(&lru_head); - max_drc_entries = nfsd_cache_size_limit(); - num_drc_entries = 0; - return 0; out_nomem: printk(KERN_ERR "nfsd: failed to allocate reply cache\n"); -- cgit From 1ada47d9468fe3907f7f9e00179168f5e2f90803 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 20 Mar 2013 09:39:42 -0400 Subject: ext4: fix ext4_evict_inode() racing against workqueue processing code Commit 84c17543ab56 (ext4: move work from io_end to inode) triggered a regression when running xfstest #270 when the file system is mounted with dioread_nolock. The problem is that after ext4_evict_inode() calls ext4_ioend_wait(), this guarantees that last io_end structure has been freed, but it does not guarantee that the workqueue structure, which was moved into the inode by commit 84c17543ab56, is actually finished. Once ext4_flush_completed_IO() calls ext4_free_io_end() on CPU #1, this will allow ext4_ioend_wait() to return on CPU #2, at which point the evict_inode() codepath can race against the workqueue code on CPU #1 accessing EXT4_I(inode)->i_unwritten_work to find the next item of work to do. Fix this by calling cancel_work_sync() in ext4_ioend_wait(), which will be renamed ext4_ioend_shutdown(), since it is only used by ext4_evict_inode(). Also, move the call to ext4_ioend_shutdown() until after truncate_inode_pages() and filemap_write_and_wait() are called, to make sure all dirty pages have been written back and flushed from the page cache first. BUG: unable to handle kernel NULL pointer dereference at (null) IP: [] cwq_activate_delayed_work+0x3b/0x7e *pdpt = 0000000030bc3001 *pde = 0000000000000000 Oops: 0000 [#1] SMP DEBUG_PAGEALLOC Modules linked in: Pid: 6, comm: kworker/u:0 Not tainted 3.8.0-rc3-00013-g84c1754-dirty #91 Bochs Bochs EIP: 0060:[] EFLAGS: 00010046 CPU: 0 EIP is at cwq_activate_delayed_work+0x3b/0x7e EAX: 00000000 EBX: 00000000 ECX: f505fe54 EDX: 00000000 ESI: ed5b697c EDI: 00000006 EBP: f64b7e8c ESP: f64b7e84 DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068 CR0: 8005003b CR2: 00000000 CR3: 30bc2000 CR4: 000006f0 DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000 DR6: ffff0ff0 DR7: 00000400 Process kworker/u:0 (pid: 6, ti=f64b6000 task=f64b4160 task.ti=f64b6000) Stack: f505fe00 00000006 f64b7e9c c01de3d7 f6435540 00000003 f64b7efc c01def1d f6435540 00000002 00000000 0000008a c16d0808 c040a10b c16d07d8 c16d08b0 f505fe00 c16d0780 00000000 00000000 ee153df4 c1ce4a30 c17d0e30 00000000 Call Trace: [] cwq_dec_nr_in_flight+0x71/0xfb [] process_one_work+0x5d8/0x637 [] ? ext4_end_bio+0x300/0x300 [] worker_thread+0x249/0x3ef [] kthread+0xd8/0xeb [] ? manage_workers+0x4bb/0x4bb [] ? trace_hardirqs_on+0x27/0x37 [] ret_from_kernel_thread+0x1b/0x28 [] ? __init_kthread_worker+0x71/0x71 Code: 01 83 15 ac ff 6c c1 00 31 db 89 c6 8b 00 a8 04 74 12 89 c3 30 db 83 05 b0 ff 6c c1 01 83 15 b4 ff 6c c1 00 89 f0 e8 42 ff ff ff <8b> 13 89 f0 83 05 b8 ff 6c c1 6c c1 00 31 c9 83 EIP: [] cwq_activate_delayed_work+0x3b/0x7e SS:ESP 0068:f64b7e84 CR2: 0000000000000000 ---[ end trace a1923229da53d8a4 ]--- Signed-off-by: "Theodore Ts'o" Cc: Jan Kara --- fs/ext4/ext4.h | 2 +- fs/ext4/inode.c | 4 ++-- fs/ext4/page-io.c | 12 +++++++++++- 3 files changed, 14 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 167ff564bbfa..3b83cd604796 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -2617,7 +2617,7 @@ extern int ext4_move_extents(struct file *o_filp, struct file *d_filp, extern int __init ext4_init_pageio(void); extern void ext4_add_complete_io(ext4_io_end_t *io_end); extern void ext4_exit_pageio(void); -extern void ext4_ioend_wait(struct inode *); +extern void ext4_ioend_shutdown(struct inode *); extern void ext4_free_io_end(ext4_io_end_t *io); extern ext4_io_end_t *ext4_init_io_end(struct inode *inode, gfp_t flags); extern void ext4_end_io_work(struct work_struct *work); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 65bbc9339aca..ea5f24ffa60c 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -185,8 +185,6 @@ void ext4_evict_inode(struct inode *inode) trace_ext4_evict_inode(inode); - ext4_ioend_wait(inode); - if (inode->i_nlink) { /* * When journalling data dirty buffers are tracked only in the @@ -216,6 +214,7 @@ void ext4_evict_inode(struct inode *inode) filemap_write_and_wait(&inode->i_data); } truncate_inode_pages(&inode->i_data, 0); + ext4_ioend_shutdown(inode); goto no_delete; } @@ -225,6 +224,7 @@ void ext4_evict_inode(struct inode *inode) if (ext4_should_order_data(inode)) ext4_begin_ordered_truncate(inode, 0); truncate_inode_pages(&inode->i_data, 0); + ext4_ioend_shutdown(inode); if (is_bad_inode(inode)) goto no_delete; diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 809b31003ecc..047a6de04a0a 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -50,11 +50,21 @@ void ext4_exit_pageio(void) kmem_cache_destroy(io_page_cachep); } -void ext4_ioend_wait(struct inode *inode) +/* + * This function is called by ext4_evict_inode() to make sure there is + * no more pending I/O completion work left to do. + */ +void ext4_ioend_shutdown(struct inode *inode) { wait_queue_head_t *wq = ext4_ioend_wq(inode); wait_event(*wq, (atomic_read(&EXT4_I(inode)->i_ioend_count) == 0)); + /* + * We need to make sure the work structure is finished being + * used before we let the inode get destroyed. + */ + if (work_pending(&EXT4_I(inode)->i_unwritten_work)) + cancel_work_sync(&EXT4_I(inode)->i_unwritten_work); } static void put_io_page(struct ext4_io_page *io_page) -- cgit From 2b405bfa84063bfa35621d2d6879f52693c614b0 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 20 Mar 2013 09:42:11 -0400 Subject: ext4: fix data=journal fast mount/umount hang In data=journal mode, if we unmount the file system before a transaction has a chance to complete, when the journal inode is being evicted, we can end up calling into jbd2_log_wait_commit() for the last transaction, after the journalling machinery has been shut down. Arguably we should adjust ext4_should_journal_data() to return FALSE for the journal inode, but the only place it matters is ext4_evict_inode(), and so to save a bit of CPU time, and to make the patch much more obviously correct by inspection(tm), we'll fix it by explicitly not trying to waiting for a journal commit when we are evicting the journal inode, since it's guaranteed to never succeed in this case. This can be easily replicated via: mount -t ext4 -o data=journal /dev/vdb /vdb ; umount /vdb ------------[ cut here ]------------ WARNING: at /usr/projects/linux/ext4/fs/jbd2/journal.c:542 __jbd2_log_start_commit+0xba/0xcd() Hardware name: Bochs JBD2: bad log_start_commit: 3005630206 3005630206 0 0 Modules linked in: Pid: 2909, comm: umount Not tainted 3.8.0-rc3 #1020 Call Trace: [] warn_slowpath_common+0x68/0x7d [] ? __jbd2_log_start_commit+0xba/0xcd [] warn_slowpath_fmt+0x2b/0x2f [] __jbd2_log_start_commit+0xba/0xcd [] jbd2_log_start_commit+0x24/0x34 [] ext4_evict_inode+0x71/0x2e3 [] evict+0x94/0x135 [] iput+0x10a/0x110 [] jbd2_journal_destroy+0x190/0x1ce [] ? bit_waitqueue+0x50/0x50 [] ext4_put_super+0x52/0x294 [] generic_shutdown_super+0x48/0xb4 [] kill_block_super+0x22/0x60 [] deactivate_locked_super+0x22/0x49 [] deactivate_super+0x30/0x33 [] mntput_no_expire+0x107/0x10c [] sys_umount+0x2cf/0x2e0 [] sys_oldumount+0x12/0x14 [] syscall_call+0x7/0xb ---[ end trace 6a954cc790501c1f ]--- jbd2_log_wait_commit: error: j_commit_request=-1289337090, tid=0 Signed-off-by: "Theodore Ts'o" Reviewed-by: Jan Kara Cc: stable@vger.kernel.org --- fs/ext4/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index ea5f24ffa60c..85e41a2a39ad 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -205,7 +205,8 @@ void ext4_evict_inode(struct inode *inode) * don't use page cache. */ if (ext4_should_journal_data(inode) && - (S_ISLNK(inode->i_mode) || S_ISREG(inode->i_mode))) { + (S_ISLNK(inode->i_mode) || S_ISREG(inode->i_mode)) && + inode->i_ino != EXT4_JOURNAL_INO) { journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; tid_t commit_tid = EXT4_I(inode)->i_datasync_tid; -- cgit From cf4ab538f1516606d3ae730dce15d6f33d96b7e1 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 8 Mar 2013 12:56:37 -0500 Subject: NFSv4: Fix the string length returned by the idmapper Functions like nfs_map_uid_to_name() and nfs_map_gid_to_group() are expected to return a string without any terminating NUL character. Regression introduced by commit 57e62324e469e092ecc6c94a7a86fe4bd6ac5172 (NFS: Store the legacy idmapper result in the keyring). Reported-by: Dave Chiluk Signed-off-by: Trond Myklebust Cc: Bryan Schumaker Cc: stable@vger.kernel.org [>=3.4] --- fs/nfs/idmap.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index dc0f98dfa717..c516da5873fd 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -726,9 +726,9 @@ out1: return ret; } -static int nfs_idmap_instantiate(struct key *key, struct key *authkey, char *data) +static int nfs_idmap_instantiate(struct key *key, struct key *authkey, char *data, size_t datalen) { - return key_instantiate_and_link(key, data, strlen(data) + 1, + return key_instantiate_and_link(key, data, datalen, id_resolver_cache->thread_keyring, authkey); } @@ -738,6 +738,7 @@ static int nfs_idmap_read_and_verify_message(struct idmap_msg *im, struct key *key, struct key *authkey) { char id_str[NFS_UINT_MAXLEN]; + size_t len; int ret = -ENOKEY; /* ret = -ENOKEY */ @@ -747,13 +748,15 @@ static int nfs_idmap_read_and_verify_message(struct idmap_msg *im, case IDMAP_CONV_NAMETOID: if (strcmp(upcall->im_name, im->im_name) != 0) break; - sprintf(id_str, "%d", im->im_id); - ret = nfs_idmap_instantiate(key, authkey, id_str); + /* Note: here we store the NUL terminator too */ + len = sprintf(id_str, "%d", im->im_id) + 1; + ret = nfs_idmap_instantiate(key, authkey, id_str, len); break; case IDMAP_CONV_IDTONAME: if (upcall->im_id != im->im_id) break; - ret = nfs_idmap_instantiate(key, authkey, im->im_name); + len = strlen(im->im_name); + ret = nfs_idmap_instantiate(key, authkey, im->im_name, len); break; default: ret = -EINVAL; -- cgit From 991f76f837bf22c5bb07261cfd86525a0a96650c Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Wed, 20 Mar 2013 23:25:24 +0800 Subject: sysfs: fix race between readdir and lseek While readdir() is running, lseek() may set filp->f_pos as zero, then may leave filp->private_data pointing to one sysfs_dirent object without holding its reference counter, so the sysfs_dirent object may be used after free in next readdir(). This patch holds inode->i_mutex to avoid the problem since the lock is always held in readdir path. Reported-by: Dave Jones Tested-by: Sasha Levin Cc: Signed-off-by: Ming Lei Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 2fbdff6be25c..c9e16608f486 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -1058,10 +1058,21 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) return 0; } +static loff_t sysfs_dir_llseek(struct file *file, loff_t offset, int whence) +{ + struct inode *inode = file_inode(file); + loff_t ret; + + mutex_lock(&inode->i_mutex); + ret = generic_file_llseek(file, offset, whence); + mutex_unlock(&inode->i_mutex); + + return ret; +} const struct file_operations sysfs_dir_operations = { .read = generic_read_dir, .readdir = sysfs_readdir, .release = sysfs_dir_release, - .llseek = generic_file_llseek, + .llseek = sysfs_dir_llseek, }; -- cgit From e5110f411d2ee35bf8d202ccca2e89c633060dca Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Wed, 20 Mar 2013 23:25:25 +0800 Subject: sysfs: handle failure path correctly for readdir() In case of 'if (filp->f_pos == 0 or 1)' of sysfs_readdir(), the failure from filldir() isn't handled, and the reference counter of the sysfs_dirent object pointed by filp->private_data will be released without clearing filp->private_data, so use after free bug will be triggered later. This patch returns immeadiately under the situation for fixing the bug, and it is reasonable to return from readdir() when filldir() fails. Reported-by: Dave Jones Tested-by: Sasha Levin Cc: Signed-off-by: Ming Lei Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index c9e16608f486..e14512678c9b 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -1020,6 +1020,8 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) ino = parent_sd->s_ino; if (filldir(dirent, ".", 1, filp->f_pos, ino, DT_DIR) == 0) filp->f_pos++; + else + return 0; } if (filp->f_pos == 1) { if (parent_sd->s_parent) @@ -1028,6 +1030,8 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) ino = parent_sd->s_ino; if (filldir(dirent, "..", 2, filp->f_pos, ino, DT_DIR) == 0) filp->f_pos++; + else + return 0; } mutex_lock(&sysfs_mutex); for (pos = sysfs_dir_pos(ns, parent_sd, filp->f_pos, pos); -- cgit From 4376c94618c26225e69e17b7c91169c45a90b292 Mon Sep 17 00:00:00 2001 From: fanchaoting Date: Thu, 21 Mar 2013 09:15:30 +0800 Subject: pnfs-block: removing DM device maybe cause oops when call dev_remove when pnfs block using device mapper,if umounting later,it maybe cause oops. we apply "1 + sizeof(bl_umount_request)" memory for msg->data, the memory maybe overflow when we do "memcpy(&dataptr [sizeof(bl_msg)], &bl_umount_request, sizeof(bl_umount_request))", because the size of bl_msg is more than 1 byte. Signed-off-by: fanchaoting Cc: stable@vger.kernel.org Signed-off-by: Trond Myklebust --- fs/nfs/blocklayout/blocklayoutdm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/blocklayout/blocklayoutdm.c b/fs/nfs/blocklayout/blocklayoutdm.c index 737d839bc17b..6fc7b5cae92b 100644 --- a/fs/nfs/blocklayout/blocklayoutdm.c +++ b/fs/nfs/blocklayout/blocklayoutdm.c @@ -55,7 +55,8 @@ static void dev_remove(struct net *net, dev_t dev) bl_pipe_msg.bl_wq = &nn->bl_wq; memset(msg, 0, sizeof(*msg)); - msg->data = kzalloc(1 + sizeof(bl_umount_request), GFP_NOFS); + msg->len = sizeof(bl_msg) + bl_msg.totallen; + msg->data = kzalloc(msg->len, GFP_NOFS); if (!msg->data) goto out; @@ -66,7 +67,6 @@ static void dev_remove(struct net *net, dev_t dev) memcpy(msg->data, &bl_msg, sizeof(bl_msg)); dataptr = (uint8_t *) msg->data; memcpy(&dataptr[sizeof(bl_msg)], &bl_umount_request, sizeof(bl_umount_request)); - msg->len = sizeof(bl_msg) + bl_msg.totallen; add_wait_queue(&nn->bl_wq, &wq); if (rpc_queue_upcall(nn->bl_device_pipe, msg) < 0) { -- cgit From a073dbff359f4741013ae4b8395f5364c5e00b48 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 20 Mar 2013 12:34:32 -0400 Subject: NFSv4.1: Fix a race in pNFS layoutcommit We need to clear the NFS_LSEG_LAYOUTCOMMIT bits atomically with the NFS_INO_LAYOUTCOMMIT bit, otherwise we may end up with situations where the two are out of sync. The first half of the problem is to ensure that pnfs_layoutcommit_inode clears the NFS_LSEG_LAYOUTCOMMIT bit through pnfs_list_write_lseg. We still need to keep the reference to those segments until the RPC call is finished, so in order to make it clear _where_ those references come from, we add a helper pnfs_list_write_lseg_done() that cleans up after pnfs_list_write_lseg. Signed-off-by: Trond Myklebust Acked-by: Benny Halevy Cc: stable@vger.kernel.org --- fs/nfs/nfs4proc.c | 14 -------------- fs/nfs/pnfs.c | 19 ++++++++++++++++++- 2 files changed, 18 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index b2671cb0f901..6ccdd4fd9b59 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -6416,22 +6416,8 @@ nfs4_layoutcommit_done(struct rpc_task *task, void *calldata) static void nfs4_layoutcommit_release(void *calldata) { struct nfs4_layoutcommit_data *data = calldata; - struct pnfs_layout_segment *lseg, *tmp; - unsigned long *bitlock = &NFS_I(data->args.inode)->flags; pnfs_cleanup_layoutcommit(data); - /* Matched by references in pnfs_set_layoutcommit */ - list_for_each_entry_safe(lseg, tmp, &data->lseg_list, pls_lc_list) { - list_del_init(&lseg->pls_lc_list); - if (test_and_clear_bit(NFS_LSEG_LAYOUTCOMMIT, - &lseg->pls_flags)) - pnfs_put_lseg(lseg); - } - - clear_bit_unlock(NFS_INO_LAYOUTCOMMITTING, bitlock); - smp_mb__after_clear_bit(); - wake_up_bit(bitlock, NFS_INO_LAYOUTCOMMITTING); - put_rpccred(data->cred); kfree(data); } diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 48ac5aad6258..3d900916fd41 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -1746,11 +1746,27 @@ static void pnfs_list_write_lseg(struct inode *inode, struct list_head *listp) list_for_each_entry(lseg, &NFS_I(inode)->layout->plh_segs, pls_list) { if (lseg->pls_range.iomode == IOMODE_RW && - test_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags)) + test_and_clear_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags)) list_add(&lseg->pls_lc_list, listp); } } +static void pnfs_list_write_lseg_done(struct inode *inode, struct list_head *listp) +{ + struct pnfs_layout_segment *lseg, *tmp; + unsigned long *bitlock = &NFS_I(inode)->flags; + + /* Matched by references in pnfs_set_layoutcommit */ + list_for_each_entry_safe(lseg, tmp, listp, pls_lc_list) { + list_del_init(&lseg->pls_lc_list); + pnfs_put_lseg(lseg); + } + + clear_bit_unlock(NFS_INO_LAYOUTCOMMITTING, bitlock); + smp_mb__after_clear_bit(); + wake_up_bit(bitlock, NFS_INO_LAYOUTCOMMITTING); +} + void pnfs_set_lo_fail(struct pnfs_layout_segment *lseg) { pnfs_layout_io_set_failed(lseg->pls_layout, lseg->pls_range.iomode); @@ -1795,6 +1811,7 @@ void pnfs_cleanup_layoutcommit(struct nfs4_layoutcommit_data *data) if (nfss->pnfs_curr_ld->cleanup_layoutcommit) nfss->pnfs_curr_ld->cleanup_layoutcommit(data); + pnfs_list_write_lseg_done(data->args.inode, &data->lseg_list); } /* -- cgit From 24956804349ca0eadcdde032d65e8c00b4214096 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 20 Mar 2013 13:03:00 -0400 Subject: NFSv4.1: Always clear the NFS_INO_LAYOUTCOMMIT in layoutreturn Note that clearing NFS_INO_LAYOUTCOMMIT is tricky, since it requires you to also clear the NFS_LSEG_LAYOUTCOMMIT bits from the layout segments. The only two sites that need to do this are the ones that call pnfs_return_layout() without first doing a layout commit. Signed-off-by: Trond Myklebust Acked-by: Benny Halevy Cc: stable@vger.kernel.org --- fs/nfs/nfs4filelayout.c | 1 - fs/nfs/pnfs.c | 35 +++++++++++++++++++++++++++-------- 2 files changed, 27 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4filelayout.c b/fs/nfs/nfs4filelayout.c index 49eeb044c109..4fb234d3aefb 100644 --- a/fs/nfs/nfs4filelayout.c +++ b/fs/nfs/nfs4filelayout.c @@ -129,7 +129,6 @@ static void filelayout_fenceme(struct inode *inode, struct pnfs_layout_hdr *lo) { if (!test_and_clear_bit(NFS_LAYOUT_RETURN, &lo->plh_flags)) return; - clear_bit(NFS_INO_LAYOUTCOMMIT, &NFS_I(inode)->flags); pnfs_return_layout(inode); } diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 3d900916fd41..5044142c1216 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -417,6 +417,16 @@ should_free_lseg(struct pnfs_layout_range *lseg_range, lo_seg_intersecting(lseg_range, recall_range); } +static bool pnfs_lseg_dec_and_remove_zero(struct pnfs_layout_segment *lseg, + struct list_head *tmp_list) +{ + if (!atomic_dec_and_test(&lseg->pls_refcount)) + return false; + pnfs_layout_remove_lseg(lseg->pls_layout, lseg); + list_add(&lseg->pls_list, tmp_list); + return true; +} + /* Returns 1 if lseg is removed from list, 0 otherwise */ static int mark_lseg_invalid(struct pnfs_layout_segment *lseg, struct list_head *tmp_list) @@ -430,11 +440,8 @@ static int mark_lseg_invalid(struct pnfs_layout_segment *lseg, */ dprintk("%s: lseg %p ref %d\n", __func__, lseg, atomic_read(&lseg->pls_refcount)); - if (atomic_dec_and_test(&lseg->pls_refcount)) { - pnfs_layout_remove_lseg(lseg->pls_layout, lseg); - list_add(&lseg->pls_list, tmp_list); + if (pnfs_lseg_dec_and_remove_zero(lseg, tmp_list)) rv = 1; - } } return rv; } @@ -777,6 +784,21 @@ send_layoutget(struct pnfs_layout_hdr *lo, return lseg; } +static void pnfs_clear_layoutcommit(struct inode *inode, + struct list_head *head) +{ + struct nfs_inode *nfsi = NFS_I(inode); + struct pnfs_layout_segment *lseg, *tmp; + + if (!test_and_clear_bit(NFS_INO_LAYOUTCOMMIT, &nfsi->flags)) + return; + list_for_each_entry_safe(lseg, tmp, &nfsi->layout->plh_segs, pls_list) { + if (!test_and_clear_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags)) + continue; + pnfs_lseg_dec_and_remove_zero(lseg, head); + } +} + /* * Initiates a LAYOUTRETURN(FILE), and removes the pnfs_layout_hdr * when the layout segment list is empty. @@ -808,6 +830,7 @@ _pnfs_return_layout(struct inode *ino) /* Reference matched in nfs4_layoutreturn_release */ pnfs_get_layout_hdr(lo); empty = list_empty(&lo->plh_segs); + pnfs_clear_layoutcommit(ino, &tmp_list); pnfs_mark_matching_lsegs_invalid(lo, &tmp_list, NULL); /* Don't send a LAYOUTRETURN if list was initially empty */ if (empty) { @@ -820,8 +843,6 @@ _pnfs_return_layout(struct inode *ino) spin_unlock(&ino->i_lock); pnfs_free_lseg_list(&tmp_list); - WARN_ON(test_bit(NFS_INO_LAYOUTCOMMIT, &nfsi->flags)); - lrp = kzalloc(sizeof(*lrp), GFP_KERNEL); if (unlikely(lrp == NULL)) { status = -ENOMEM; @@ -1458,7 +1479,6 @@ static void pnfs_ld_handle_write_error(struct nfs_write_data *data) dprintk("pnfs write error = %d\n", hdr->pnfs_error); if (NFS_SERVER(hdr->inode)->pnfs_curr_ld->flags & PNFS_LAYOUTRET_ON_ERROR) { - clear_bit(NFS_INO_LAYOUTCOMMIT, &NFS_I(hdr->inode)->flags); pnfs_return_layout(hdr->inode); } if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) @@ -1613,7 +1633,6 @@ static void pnfs_ld_handle_read_error(struct nfs_read_data *data) dprintk("pnfs read error = %d\n", hdr->pnfs_error); if (NFS_SERVER(hdr->inode)->pnfs_curr_ld->flags & PNFS_LAYOUTRET_ON_ERROR) { - clear_bit(NFS_INO_LAYOUTCOMMIT, &NFS_I(hdr->inode)->flags); pnfs_return_layout(hdr->inode); } if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags)) -- cgit From 240286725d854331422cb15957f8d9bf2741d4e3 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 20 Mar 2013 13:23:33 -0400 Subject: NFSv4.1: Add a helper pnfs_commit_and_return_layout In order to be able to safely return the layout in nfs4_proc_setattr, we need to block new uses of the layout, wait for all outstanding users of the layout to complete, commit the layout and then return it. This patch adds a helper in order to do all this safely. Signed-off-by: Trond Myklebust Cc: Boaz Harrosh --- fs/nfs/nfs4proc.c | 2 +- fs/nfs/pnfs.c | 27 +++++++++++++++++++++++++++ fs/nfs/pnfs.h | 6 ++++++ 3 files changed, 34 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 6ccdd4fd9b59..26431cf62ddb 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2632,7 +2632,7 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, int status; if (pnfs_ld_layoutret_on_setattr(inode)) - pnfs_return_layout(inode); + pnfs_commit_and_return_layout(inode); nfs_fattr_init(fattr); diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 5044142c1216..4bdffe0ba025 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -866,6 +866,33 @@ out: } EXPORT_SYMBOL_GPL(_pnfs_return_layout); +int +pnfs_commit_and_return_layout(struct inode *inode) +{ + struct pnfs_layout_hdr *lo; + int ret; + + spin_lock(&inode->i_lock); + lo = NFS_I(inode)->layout; + if (lo == NULL) { + spin_unlock(&inode->i_lock); + return 0; + } + pnfs_get_layout_hdr(lo); + /* Block new layoutgets and read/write to ds */ + lo->plh_block_lgets++; + spin_unlock(&inode->i_lock); + filemap_fdatawait(inode->i_mapping); + ret = pnfs_layoutcommit_inode(inode, true); + if (ret == 0) + ret = _pnfs_return_layout(inode); + spin_lock(&inode->i_lock); + lo->plh_block_lgets--; + spin_unlock(&inode->i_lock); + pnfs_put_layout_hdr(lo); + return ret; +} + bool pnfs_roc(struct inode *ino) { struct pnfs_layout_hdr *lo; diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index 94ba80417748..f5f8a470a647 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -219,6 +219,7 @@ void pnfs_set_layoutcommit(struct nfs_write_data *wdata); void pnfs_cleanup_layoutcommit(struct nfs4_layoutcommit_data *data); int pnfs_layoutcommit_inode(struct inode *inode, bool sync); int _pnfs_return_layout(struct inode *); +int pnfs_commit_and_return_layout(struct inode *); void pnfs_ld_write_done(struct nfs_write_data *); void pnfs_ld_read_done(struct nfs_read_data *); struct pnfs_layout_segment *pnfs_update_layout(struct inode *ino, @@ -407,6 +408,11 @@ static inline int pnfs_return_layout(struct inode *ino) return 0; } +static inline int pnfs_commit_and_return_layout(struct inode *inode) +{ + return 0; +} + static inline bool pnfs_ld_layoutret_on_setattr(struct inode *inode) { -- cgit From 06ae43f34bcc07a0b6be8bf78a1c895bcd12c839 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 20 Mar 2013 13:19:30 -0400 Subject: Don't bother with redoing rw_verify_area() from default_file_splice_from() default_file_splice_from() ends up calling vfs_write() (via very convoluted callchain). It's an overkill, since we already have done rw_verify_area() in the caller by the time we call vfs_write() we are under set_fs(KERNEL_DS), so access_ok() is also pointless. Add a new helper (__kernel_write()), use it instead of kernel_write() in there. Signed-off-by: Al Viro --- fs/internal.h | 5 +++++ fs/read_write.c | 25 +++++++++++++++++++++++++ fs/splice.c | 4 +++- 3 files changed, 33 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/internal.h b/fs/internal.h index 507141fceb99..4be78237d896 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -125,3 +125,8 @@ extern int invalidate_inodes(struct super_block *, bool); * dcache.c */ extern struct dentry *__d_alloc(struct super_block *, const struct qstr *); + +/* + * read_write.c + */ +extern ssize_t __kernel_write(struct file *, const char *, size_t, loff_t *); diff --git a/fs/read_write.c b/fs/read_write.c index a698eff457fb..f7b5a23b804b 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -17,6 +17,7 @@ #include #include #include "read_write.h" +#include "internal.h" #include #include @@ -417,6 +418,30 @@ ssize_t do_sync_write(struct file *filp, const char __user *buf, size_t len, lof EXPORT_SYMBOL(do_sync_write); +ssize_t __kernel_write(struct file *file, const char *buf, size_t count, loff_t *pos) +{ + mm_segment_t old_fs; + const char __user *p; + ssize_t ret; + + old_fs = get_fs(); + set_fs(get_ds()); + p = (__force const char __user *)buf; + if (count > MAX_RW_COUNT) + count = MAX_RW_COUNT; + if (file->f_op->write) + ret = file->f_op->write(file, p, count, pos); + else + ret = do_sync_write(file, p, count, pos); + set_fs(old_fs); + if (ret > 0) { + fsnotify_modify(file); + add_wchar(current, ret); + } + inc_syscw(current); + return ret; +} + ssize_t vfs_write(struct file *file, const char __user *buf, size_t count, loff_t *pos) { ssize_t ret; diff --git a/fs/splice.c b/fs/splice.c index 718bd0056384..29e394e49ddd 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -31,6 +31,7 @@ #include #include #include +#include "internal.h" /* * Attempt to steal a page from a pipe buffer. This should perhaps go into @@ -1048,9 +1049,10 @@ static int write_pipe_buf(struct pipe_inode_info *pipe, struct pipe_buffer *buf, { int ret; void *data; + loff_t tmp = sd->pos; data = buf->ops->map(pipe, buf, 0); - ret = kernel_write(sd->u.file, data + buf->offset, sd->len, sd->pos); + ret = __kernel_write(sd->u.file, data + buf->offset, sd->len, &tmp); buf->ops->unmap(pipe, buf, data); return ret; -- cgit From f853c616883a8de966873a1dab283f1369e275a1 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 11 Mar 2013 09:52:19 -0400 Subject: cifs: ignore everything in SPNEGO blob after mechTypes We've had several reports of people attempting to mount Windows 8 shares and getting failures with a return code of -EINVAL. The default sec= mode changed recently to sec=ntlmssp. With that, we expect and parse a SPNEGO blob from the server in the NEGOTIATE reply. The current decode_negTokenInit function first parses all of the mechTypes and then tries to parse the rest of the negTokenInit reply. The parser however currently expects a mechListMIC or nothing to follow the mechTypes, but Windows 8 puts a mechToken field there instead to carry some info for the new NegoEx stuff. In practice, we don't do anything with the fields after the mechTypes anyway so I don't see any real benefit in continuing to parse them. This patch just has the kernel ignore the fields after the mechTypes. We'll probably need to reinstate some of this if we ever want to support NegoEx. Reported-by: Jason Burgess Reported-by: Yan Li Signed-off-by: Jeff Layton Cc: Signed-off-by: Steve French --- fs/cifs/asn1.c | 53 +++++------------------------------------------------ 1 file changed, 5 insertions(+), 48 deletions(-) (limited to 'fs') diff --git a/fs/cifs/asn1.c b/fs/cifs/asn1.c index cfd1ce34e0bc..1d36db114772 100644 --- a/fs/cifs/asn1.c +++ b/fs/cifs/asn1.c @@ -614,53 +614,10 @@ decode_negTokenInit(unsigned char *security_blob, int length, } } - /* mechlistMIC */ - if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - /* Check if we have reached the end of the blob, but with - no mechListMic (e.g. NTLMSSP instead of KRB5) */ - if (ctx.error == ASN1_ERR_DEC_EMPTY) - goto decode_negtoken_exit; - cFYI(1, "Error decoding last part negTokenInit exit3"); - return 0; - } else if ((cls != ASN1_CTX) || (con != ASN1_CON)) { - /* tag = 3 indicating mechListMIC */ - cFYI(1, "Exit 4 cls = %d con = %d tag = %d end = %p (%d)", - cls, con, tag, end, *end); - return 0; - } - - /* sequence */ - if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - cFYI(1, "Error decoding last part negTokenInit exit5"); - return 0; - } else if ((cls != ASN1_UNI) || (con != ASN1_CON) - || (tag != ASN1_SEQ)) { - cFYI(1, "cls = %d con = %d tag = %d end = %p (%d)", - cls, con, tag, end, *end); - } - - /* sequence of */ - if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - cFYI(1, "Error decoding last part negTokenInit exit 7"); - return 0; - } else if ((cls != ASN1_CTX) || (con != ASN1_CON)) { - cFYI(1, "Exit 8 cls = %d con = %d tag = %d end = %p (%d)", - cls, con, tag, end, *end); - return 0; - } - - /* general string */ - if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { - cFYI(1, "Error decoding last part negTokenInit exit9"); - return 0; - } else if ((cls != ASN1_UNI) || (con != ASN1_PRI) - || (tag != ASN1_GENSTR)) { - cFYI(1, "Exit10 cls = %d con = %d tag = %d end = %p (%d)", - cls, con, tag, end, *end); - return 0; - } - cFYI(1, "Need to call asn1_octets_decode() function for %s", - ctx.pointer); /* is this UTF-8 or ASCII? */ -decode_negtoken_exit: + /* + * We currently ignore anything at the end of the SPNEGO blob after + * the mechTypes have been parsed, since none of that info is + * used at the moment. + */ return 1; } -- cgit From d763448286377b8a0e3f179372e9e292bef3c337 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Mon, 11 Mar 2013 09:20:00 +0000 Subject: Btrfs: update to use fs_state bit Now that we use bit operation to check fs_state, update btrfs_free_fs_root()'s checker, otherwise we get back to memory leak case. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 7d84651e850b..127b23e8323b 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3253,7 +3253,7 @@ void btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root) if (btrfs_root_refs(&root->root_item) == 0) synchronize_srcu(&fs_info->subvol_srcu); - if (fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) { + if (test_bit(BTRFS_FS_STATE_ERROR, &fs_info->fs_state)) { btrfs_free_log(NULL, root); btrfs_free_log_root_tree(NULL, fs_info); } -- cgit From 835d974fabfa9bff4d173ad03c054ac2f673263f Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 19 Mar 2013 12:13:25 -0400 Subject: Btrfs: handle a bogus chunk tree nicely If you restore a btrfs-image file system and try to mount that file system we'll panic. That's because btrfs-image restores and just makes one big chunk to envelope the whole disk, since they are really only meant to be messed with by our btrfs-progs. So fix up btrfs_rmap_block and the callers of it for mount so that we no longer panic but instead just return an error and fail to mount. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 35 ++++++++++++++++++++++++++++++----- fs/btrfs/volumes.c | 13 ++++++++++++- 2 files changed, 42 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 350b9b18140c..a8ff25aedca1 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -257,7 +257,8 @@ static int exclude_super_stripes(struct btrfs_root *root, cache->bytes_super += stripe_len; ret = add_excluded_extent(root, cache->key.objectid, stripe_len); - BUG_ON(ret); /* -ENOMEM */ + if (ret) + return ret; } for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) { @@ -265,13 +266,17 @@ static int exclude_super_stripes(struct btrfs_root *root, ret = btrfs_rmap_block(&root->fs_info->mapping_tree, cache->key.objectid, bytenr, 0, &logical, &nr, &stripe_len); - BUG_ON(ret); /* -ENOMEM */ + if (ret) + return ret; while (nr--) { cache->bytes_super += stripe_len; ret = add_excluded_extent(root, logical[nr], stripe_len); - BUG_ON(ret); /* -ENOMEM */ + if (ret) { + kfree(logical); + return ret; + } } kfree(logical); @@ -7964,7 +7969,17 @@ int btrfs_read_block_groups(struct btrfs_root *root) * info has super bytes accounted for, otherwise we'll think * we have more space than we actually do. */ - exclude_super_stripes(root, cache); + ret = exclude_super_stripes(root, cache); + if (ret) { + /* + * We may have excluded something, so call this just in + * case. + */ + free_excluded_extents(root, cache); + kfree(cache->free_space_ctl); + kfree(cache); + goto error; + } /* * check for two cases, either we are full, and therefore @@ -8106,7 +8121,17 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, cache->last_byte_to_unpin = (u64)-1; cache->cached = BTRFS_CACHE_FINISHED; - exclude_super_stripes(root, cache); + ret = exclude_super_stripes(root, cache); + if (ret) { + /* + * We may have excluded something, so call this just in + * case. + */ + free_excluded_extents(root, cache); + kfree(cache->free_space_ctl); + kfree(cache); + return ret; + } add_new_free_space(cache, root->fs_info, chunk_offset, chunk_offset + size); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 5989a92236f7..2854c824ab64 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -4935,7 +4935,18 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, em = lookup_extent_mapping(em_tree, chunk_start, 1); read_unlock(&em_tree->lock); - BUG_ON(!em || em->start != chunk_start); + if (!em) { + printk(KERN_ERR "btrfs: couldn't find em for chunk %Lu\n", + chunk_start); + return -EIO; + } + + if (em->start != chunk_start) { + printk(KERN_ERR "btrfs: bad chunk start, em=%Lu, wanted=%Lu\n", + em->start, chunk_start); + free_extent_map(em); + return -EIO; + } map = (struct map_lookup *)em->bdev; length = em->len; -- cgit From 6113077cd319e747875ec71227d2b5cb54e08c76 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Tue, 19 Mar 2013 10:57:14 +0000 Subject: Btrfs: fix missing qgroup reservation before fallocating Steps to reproduce: mkfs.btrfs mount btrfs quota enable btrfs sub create /subv btrfs qgroup limit 10M /subv fallocate --length 20M /subv/data For the above example, fallocating will return successfully which is not expected, we try to fix it by doing qgroup reservation before fallocating. Signed-off-by: Wang Shilong Reviewed-by: Miao Xie Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/file.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 7bdb47faa12e..1be25b92d63c 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -2142,6 +2142,7 @@ static long btrfs_fallocate(struct file *file, int mode, { struct inode *inode = file->f_path.dentry->d_inode; struct extent_state *cached_state = NULL; + struct btrfs_root *root = BTRFS_I(inode)->root; u64 cur_offset; u64 last_byte; u64 alloc_start; @@ -2169,6 +2170,11 @@ static long btrfs_fallocate(struct file *file, int mode, ret = btrfs_check_data_free_space(inode, alloc_end - alloc_start); if (ret) return ret; + if (root->fs_info->quota_enabled) { + ret = btrfs_qgroup_reserve(root, alloc_end - alloc_start); + if (ret) + goto out_reserve_fail; + } /* * wait for ordered IO before we have any locks. We'll loop again @@ -2272,6 +2278,9 @@ static long btrfs_fallocate(struct file *file, int mode, &cached_state, GFP_NOFS); out: mutex_unlock(&inode->i_mutex); + if (root->fs_info->quota_enabled) + btrfs_qgroup_free(root, alloc_end - alloc_start); +out_reserve_fail: /* Let go of our reservation. */ btrfs_free_reserved_data_space(inode, alloc_end - alloc_start); return ret; -- cgit From d9abbf1c3131b679379762700201ae69367f3f62 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Wed, 20 Mar 2013 13:49:48 +0000 Subject: Btrfs: fix locking on ROOT_REPLACE operations in tree mod log To resolve backrefs, ROOT_REPLACE operations in the tree mod log are required to be tied to at least one KEY_REMOVE_WHILE_FREEING operation. Therefore, those operations must be enclosed by tree_mod_log_write_lock() and tree_mod_log_write_unlock() calls. Those calls are private to the tree_mod_log_* functions, which means that removal of the elements of an old root node must be logged from tree_mod_log_insert_root. This partly reverts and corrects commit ba1bfbd5 (Btrfs: fix a tree mod logging issue for root replacement operations). This fixes the brand-new version of xfstest 276 as of commit cfe73f71. Cc: stable@vger.kernel.org Signed-off-by: Jan Schmidt Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 30 ++++++++++++++++++++---------- 1 file changed, 20 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index ecd25a1b4e51..ca9d8f1a3bb6 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -651,6 +651,8 @@ tree_mod_log_insert_root(struct btrfs_fs_info *fs_info, if (tree_mod_dont_log(fs_info, NULL)) return 0; + __tree_mod_log_free_eb(fs_info, old_root); + ret = tree_mod_alloc(fs_info, flags, &tm); if (ret < 0) goto out; @@ -736,7 +738,7 @@ tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq) static noinline void tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst, struct extent_buffer *src, unsigned long dst_offset, - unsigned long src_offset, int nr_items) + unsigned long src_offset, int nr_items, int log_removal) { int ret; int i; @@ -750,10 +752,12 @@ tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst, } for (i = 0; i < nr_items; i++) { - ret = tree_mod_log_insert_key_locked(fs_info, src, - i + src_offset, - MOD_LOG_KEY_REMOVE); - BUG_ON(ret < 0); + if (log_removal) { + ret = tree_mod_log_insert_key_locked(fs_info, src, + i + src_offset, + MOD_LOG_KEY_REMOVE); + BUG_ON(ret < 0); + } ret = tree_mod_log_insert_key_locked(fs_info, dst, i + dst_offset, MOD_LOG_KEY_ADD); @@ -927,7 +931,6 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans, ret = btrfs_dec_ref(trans, root, buf, 1, 1); BUG_ON(ret); /* -ENOMEM */ } - tree_mod_log_free_eb(root->fs_info, buf); clean_tree_block(trans, root, buf); *last_ref = 1; } @@ -1046,6 +1049,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans, btrfs_set_node_ptr_generation(parent, parent_slot, trans->transid); btrfs_mark_buffer_dirty(parent); + tree_mod_log_free_eb(root->fs_info, buf); btrfs_free_tree_block(trans, root, buf, parent_start, last_ref); } @@ -1750,7 +1754,6 @@ static noinline int balance_level(struct btrfs_trans_handle *trans, goto enospc; } - tree_mod_log_free_eb(root->fs_info, root->node); tree_mod_log_set_root_pointer(root, child); rcu_assign_pointer(root->node, child); @@ -2995,7 +2998,7 @@ static int push_node_left(struct btrfs_trans_handle *trans, push_items = min(src_nritems - 8, push_items); tree_mod_log_eb_copy(root->fs_info, dst, src, dst_nritems, 0, - push_items); + push_items, 1); copy_extent_buffer(dst, src, btrfs_node_key_ptr_offset(dst_nritems), btrfs_node_key_ptr_offset(0), @@ -3066,7 +3069,7 @@ static int balance_node_right(struct btrfs_trans_handle *trans, sizeof(struct btrfs_key_ptr)); tree_mod_log_eb_copy(root->fs_info, dst, src, 0, - src_nritems - push_items, push_items); + src_nritems - push_items, push_items, 1); copy_extent_buffer(dst, src, btrfs_node_key_ptr_offset(0), btrfs_node_key_ptr_offset(src_nritems - push_items), @@ -3218,12 +3221,18 @@ static noinline int split_node(struct btrfs_trans_handle *trans, int mid; int ret; u32 c_nritems; + int tree_mod_log_removal = 1; c = path->nodes[level]; WARN_ON(btrfs_header_generation(c) != trans->transid); if (c == root->node) { /* trying to split the root, lets make a new one */ ret = insert_new_root(trans, root, path, level + 1); + /* + * removal of root nodes has been logged by + * tree_mod_log_set_root_pointer due to locking + */ + tree_mod_log_removal = 0; if (ret) return ret; } else { @@ -3261,7 +3270,8 @@ static noinline int split_node(struct btrfs_trans_handle *trans, (unsigned long)btrfs_header_chunk_tree_uuid(split), BTRFS_UUID_SIZE); - tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid); + tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid, + tree_mod_log_removal); copy_extent_buffer(split, c, btrfs_node_key_ptr_offset(0), btrfs_node_key_ptr_offset(mid), -- cgit From 1dd05682b3ef6e70409e130bfd83e91770801589 Mon Sep 17 00:00:00 2001 From: Tsutomu Itoh Date: Thu, 21 Mar 2013 04:32:32 +0000 Subject: Btrfs: fix memory leak in btrfs_create_tree() We should free leaf and root before returning from the error handling code. Signed-off-by: Tsutomu Itoh Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 127b23e8323b..6d19a0a554aa 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1291,6 +1291,7 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans, 0, objectid, NULL, 0, 0, 0); if (IS_ERR(leaf)) { ret = PTR_ERR(leaf); + leaf = NULL; goto fail; } @@ -1334,11 +1335,16 @@ struct btrfs_root *btrfs_create_tree(struct btrfs_trans_handle *trans, btrfs_tree_unlock(leaf); + return root; + fail: - if (ret) - return ERR_PTR(ret); + if (leaf) { + btrfs_tree_unlock(leaf); + free_extent_buffer(leaf); + } + kfree(root); - return root; + return ERR_PTR(ret); } static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans, -- cgit From 51f0885e5415b4cc6535e9cdcc5145bfbc134353 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 22 Mar 2013 11:44:04 -0700 Subject: vfs,proc: guarantee unique inodes in /proc Dave Jones found another /proc issue with his Trinity tool: thanks to the namespace model, we can have multiple /proc dentries that point to the same inode, aliasing directories in /proc//net/ for example. This ends up being a total disaster, because it acts like hardlinked directories, and causes locking problems. We rely on the topological sort of the inodes pointed to by dentries, and if we have aliased directories, that odering becomes unreliable. In short: don't do this. Multiple dentries with the same (directory) inode is just a bad idea, and the namespace code should never have exposed things this way. But we're kind of stuck with it. This solves things by just always allocating a new inode during /proc dentry lookup, instead of using "iget_locked()" to look up existing inodes by superblock and number. That actually simplies the code a bit, at the cost of potentially doing more inode [de]allocations. That said, the inode lookup wasn't free either (and did a lot of locking of inodes), so it is probably not that noticeable. We could easily keep the old lookup model for non-directory entries, but rather than try to be excessively clever this just implements the minimal and simplest workaround for the problem. Reported-and-tested-by: Dave Jones Analyzed-by: Al Viro Cc: stable@vger.kernel.org Signed-off-by: Linus Torvalds --- fs/proc/inode.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/proc/inode.c b/fs/proc/inode.c index a86aebc9ba7c..869116c2afbe 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -446,9 +446,10 @@ static const struct file_operations proc_reg_file_ops_no_compat = { struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) { - struct inode *inode = iget_locked(sb, de->low_ino); + struct inode *inode = new_inode_pseudo(sb); - if (inode && (inode->i_state & I_NEW)) { + if (inode) { + inode->i_ino = de->low_ino; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; PROC_I(inode)->pde = de; @@ -476,7 +477,6 @@ struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) inode->i_fop = de->proc_fops; } } - unlock_new_inode(inode); } else pde_put(de); return inode; -- cgit From e49dbbf3e770aa590a8a464ac4978a09027060b9 Mon Sep 17 00:00:00 2001 From: Kent Overstreet Date: Fri, 22 Mar 2013 11:18:24 -0700 Subject: nfsd: fix bad offset use vfs_writev() updates the offset argument - but the code then passes the offset to vfs_fsync_range(). Since offset now points to the offset after what was just written, this is probably not what was intended Introduced by face15025ffdf664de95e86ae831544154d26c9c "nfsd: use vfs_fsync_range(), not O_SYNC, for stable writes". Signed-off-by: Kent Overstreet Cc: Al Viro Cc: "Eric W. Biederman" Cc: stable@vger.kernel.org Reviewed-by: Zach Brown Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 2a7eb536de0b..2b2e2396a869 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1013,6 +1013,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, int host_err; int stable = *stablep; int use_wgather; + loff_t pos = offset; dentry = file->f_path.dentry; inode = dentry->d_inode; @@ -1025,7 +1026,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, /* Write the data. */ oldfs = get_fs(); set_fs(KERNEL_DS); - host_err = vfs_writev(file, (struct iovec __user *)vec, vlen, &offset); + host_err = vfs_writev(file, (struct iovec __user *)vec, vlen, &pos); set_fs(oldfs); if (host_err < 0) goto out_nfserr; -- cgit From 4adaa611020fa6ac65b0ac8db78276af4ec04e63 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 26 Mar 2013 13:07:00 -0400 Subject: Btrfs: fix race between mmap writes and compression Btrfs uses page_mkwrite to ensure stable pages during crc calculations and mmap workloads. We call clear_page_dirty_for_io before we do any crcs, and this forces any application with the file mapped to wait for the crc to finish before it is allowed to change the file. With compression on, the clear_page_dirty_for_io step is happening after we've compressed the pages. This means the applications might be changing the pages while we are compressing them, and some of those modifications might not hit the disk. This commit adds the clear_page_dirty_for_io before compression starts and makes sure to redirty the page if we have to fallback to uncompressed IO as well. Signed-off-by: Chris Mason Reported-by: Alexandre Oliva cc: stable@vger.kernel.org --- fs/btrfs/extent_io.c | 33 +++++++++++++++++++++++++++++++++ fs/btrfs/extent_io.h | 2 ++ fs/btrfs/inode.c | 14 ++++++++++++++ 3 files changed, 49 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index f173c5af6461..cdee391fc7bf 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -1257,6 +1257,39 @@ int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end) GFP_NOFS); } +int extent_range_clear_dirty_for_io(struct inode *inode, u64 start, u64 end) +{ + unsigned long index = start >> PAGE_CACHE_SHIFT; + unsigned long end_index = end >> PAGE_CACHE_SHIFT; + struct page *page; + + while (index <= end_index) { + page = find_get_page(inode->i_mapping, index); + BUG_ON(!page); /* Pages should be in the extent_io_tree */ + clear_page_dirty_for_io(page); + page_cache_release(page); + index++; + } + return 0; +} + +int extent_range_redirty_for_io(struct inode *inode, u64 start, u64 end) +{ + unsigned long index = start >> PAGE_CACHE_SHIFT; + unsigned long end_index = end >> PAGE_CACHE_SHIFT; + struct page *page; + + while (index <= end_index) { + page = find_get_page(inode->i_mapping, index); + BUG_ON(!page); /* Pages should be in the extent_io_tree */ + account_page_redirty(page); + __set_page_dirty_nobuffers(page); + page_cache_release(page); + index++; + } + return 0; +} + /* * helper function to set both pages and extents in the tree writeback */ diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index 6068a1985560..258c92156857 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -325,6 +325,8 @@ int map_private_extent_buffer(struct extent_buffer *eb, unsigned long offset, unsigned long *map_len); int extent_range_uptodate(struct extent_io_tree *tree, u64 start, u64 end); +int extent_range_clear_dirty_for_io(struct inode *inode, u64 start, u64 end); +int extent_range_redirty_for_io(struct inode *inode, u64 start, u64 end); int extent_clear_unlock_delalloc(struct inode *inode, struct extent_io_tree *tree, u64 start, u64 end, struct page *locked_page, diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 1f26888825e2..6a6e13c53086 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -353,6 +353,7 @@ static noinline int compress_file_range(struct inode *inode, int i; int will_compress; int compress_type = root->fs_info->compress_type; + int redirty = 0; /* if this is a small write inside eof, kick off a defrag */ if ((end - start + 1) < 16 * 1024 && @@ -415,6 +416,17 @@ again: if (BTRFS_I(inode)->force_compress) compress_type = BTRFS_I(inode)->force_compress; + /* + * we need to call clear_page_dirty_for_io on each + * page in the range. Otherwise applications with the file + * mmap'd can wander in and change the page contents while + * we are compressing them. + * + * If the compression fails for any reason, we set the pages + * dirty again later on. + */ + extent_range_clear_dirty_for_io(inode, start, end); + redirty = 1; ret = btrfs_compress_pages(compress_type, inode->i_mapping, start, total_compressed, pages, @@ -554,6 +566,8 @@ cleanup_and_bail_uncompressed: __set_page_dirty_nobuffers(locked_page); /* unlocked later on in the async handlers */ } + if (redirty) + extent_range_redirty_for_io(inode, start, end); add_async_extent(async_cow, start, end - start + 1, 0, NULL, 0, BTRFS_COMPRESS_NONE); *num_added += 1; -- cgit From 7ea600b5314529f9d1b9d6d3c41cb26fce6a7a4a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 26 Mar 2013 18:25:57 -0400 Subject: Nest rename_lock inside vfsmount_lock ... lest we get livelocks between path_is_under() and d_path() and friends. The thing is, wrt fairness lglocks are more similar to rwsems than to rwlocks; it is possible to have thread B spin on attempt to take lock shared while thread A is already holding it shared, if B is on lower-numbered CPU than A and there's a thread C spinning on attempt to take the same lock exclusive. As the result, we need consistent ordering between vfsmount_lock (lglock) and rename_lock (seq_lock), even though everything that takes both is going to take vfsmount_lock only shared. Spotted-by: Brad Spengler Cc: stable@vger.kernel.org Signed-off-by: Al Viro --- fs/dcache.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/dcache.c b/fs/dcache.c index fbfae008ba44..e8bc3420d63e 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -2542,7 +2542,6 @@ static int prepend_path(const struct path *path, bool slash = false; int error = 0; - br_read_lock(&vfsmount_lock); while (dentry != root->dentry || vfsmnt != root->mnt) { struct dentry * parent; @@ -2572,8 +2571,6 @@ static int prepend_path(const struct path *path, if (!error && !slash) error = prepend(buffer, buflen, "/", 1); -out: - br_read_unlock(&vfsmount_lock); return error; global_root: @@ -2590,7 +2587,7 @@ global_root: error = prepend(buffer, buflen, "/", 1); if (!error) error = is_mounted(vfsmnt) ? 1 : 2; - goto out; + return error; } /** @@ -2617,9 +2614,11 @@ char *__d_path(const struct path *path, int error; prepend(&res, &buflen, "\0", 1); + br_read_lock(&vfsmount_lock); write_seqlock(&rename_lock); error = prepend_path(path, root, &res, &buflen); write_sequnlock(&rename_lock); + br_read_unlock(&vfsmount_lock); if (error < 0) return ERR_PTR(error); @@ -2636,9 +2635,11 @@ char *d_absolute_path(const struct path *path, int error; prepend(&res, &buflen, "\0", 1); + br_read_lock(&vfsmount_lock); write_seqlock(&rename_lock); error = prepend_path(path, &root, &res, &buflen); write_sequnlock(&rename_lock); + br_read_unlock(&vfsmount_lock); if (error > 1) error = -EINVAL; @@ -2702,11 +2703,13 @@ char *d_path(const struct path *path, char *buf, int buflen) return path->dentry->d_op->d_dname(path->dentry, buf, buflen); get_fs_root(current->fs, &root); + br_read_lock(&vfsmount_lock); write_seqlock(&rename_lock); error = path_with_deleted(path, &root, &res, &buflen); + write_sequnlock(&rename_lock); + br_read_unlock(&vfsmount_lock); if (error < 0) res = ERR_PTR(error); - write_sequnlock(&rename_lock); path_put(&root); return res; } @@ -2830,6 +2833,7 @@ SYSCALL_DEFINE2(getcwd, char __user *, buf, unsigned long, size) get_fs_root_and_pwd(current->fs, &root, &pwd); error = -ENOENT; + br_read_lock(&vfsmount_lock); write_seqlock(&rename_lock); if (!d_unlinked(pwd.dentry)) { unsigned long len; @@ -2839,6 +2843,7 @@ SYSCALL_DEFINE2(getcwd, char __user *, buf, unsigned long, size) prepend(&cwd, &buflen, "\0", 1); error = prepend_path(&pwd, &root, &cwd, &buflen); write_sequnlock(&rename_lock); + br_read_unlock(&vfsmount_lock); if (error < 0) goto out; @@ -2859,6 +2864,7 @@ SYSCALL_DEFINE2(getcwd, char __user *, buf, unsigned long, size) } } else { write_sequnlock(&rename_lock); + br_read_unlock(&vfsmount_lock); } out: -- cgit From 3151527ee007b73a0ebd296010f1c0454a919c7d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 15 Mar 2013 01:45:51 -0700 Subject: userns: Don't allow creation if the user is chrooted Guarantee that the policy of which files may be access that is established by setting the root directory will not be violated by user namespaces by verifying that the root directory points to the root of the mount namespace at the time of user namespace creation. Changing the root is a privileged operation, and as a matter of policy it serves to limit unprivileged processes to files below the current root directory. For reasons of simplicity and comprehensibility the privilege to change the root directory is gated solely on the CAP_SYS_CHROOT capability in the user namespace. Therefore when creating a user namespace we must ensure that the policy of which files may be access can not be violated by changing the root directory. Anyone who runs a processes in a chroot and would like to use user namespace can setup the same view of filesystems with a mount namespace instead. With this result that this is not a practical limitation for using user namespaces. Cc: stable@vger.kernel.org Acked-by: Serge Hallyn Reported-by: Andy Lutomirski Signed-off-by: "Eric W. Biederman" --- fs/namespace.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 50ca17d3cb45..a3035223d421 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -2732,6 +2732,30 @@ bool our_mnt(struct vfsmount *mnt) return check_mnt(real_mount(mnt)); } +bool current_chrooted(void) +{ + /* Does the current process have a non-standard root */ + struct path ns_root; + struct path fs_root; + bool chrooted; + + /* Find the namespace root */ + ns_root.mnt = ¤t->nsproxy->mnt_ns->root->mnt; + ns_root.dentry = ns_root.mnt->mnt_root; + path_get(&ns_root); + while (d_mountpoint(ns_root.dentry) && follow_down_one(&ns_root)) + ; + + get_fs_root(current->fs, &fs_root); + + chrooted = !path_equal(&fs_root, &ns_root); + + path_put(&fs_root); + path_put(&ns_root); + + return chrooted; +} + static void *mntns_get(struct task_struct *task) { struct mnt_namespace *ns = NULL; -- cgit From 90563b198e4c6674c63672fae1923da467215f45 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 22 Mar 2013 03:10:15 -0700 Subject: vfs: Add a mount flag to lock read only bind mounts When a read-only bind mount is copied from mount namespace in a higher privileged user namespace to a mount namespace in a lesser privileged user namespace, it should not be possible to remove the the read-only restriction. Add a MNT_LOCK_READONLY mount flag to indicate that a mount must remain read-only. CC: stable@vger.kernel.org Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- fs/namespace.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index a3035223d421..8505b5ece5de 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1713,6 +1713,9 @@ static int change_mount_flags(struct vfsmount *mnt, int ms_flags) if (readonly_request == __mnt_is_readonly(mnt)) return 0; + if (mnt->mnt_flags & MNT_LOCK_READONLY) + return -EPERM; + if (readonly_request) error = mnt_make_readonly(real_mount(mnt)); else -- cgit From 132c94e31b8bca8ea921f9f96a57d684fa4ae0a9 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 22 Mar 2013 04:08:05 -0700 Subject: vfs: Carefully propogate mounts across user namespaces As a matter of policy MNT_READONLY should not be changable if the original mounter had more privileges than creator of the mount namespace. Add the flag CL_UNPRIVILEGED to note when we are copying a mount from a mount namespace that requires more privileges to a mount namespace that requires fewer privileges. When the CL_UNPRIVILEGED flag is set cause clone_mnt to set MNT_NO_REMOUNT if any of the mnt flags that should never be changed are set. This protects both mount propagation and the initial creation of a less privileged mount namespace. Cc: stable@vger.kernel.org Acked-by: Serge Hallyn Reported-by: Andy Lutomirski Signed-off-by: "Eric W. Biederman" --- fs/namespace.c | 6 +++++- fs/pnode.c | 6 ++++++ fs/pnode.h | 1 + 3 files changed, 12 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 8505b5ece5de..968d4c5eae03 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -798,6 +798,10 @@ static struct mount *clone_mnt(struct mount *old, struct dentry *root, } mnt->mnt.mnt_flags = old->mnt.mnt_flags & ~MNT_WRITE_HOLD; + /* Don't allow unprivileged users to change mount flags */ + if ((flag & CL_UNPRIVILEGED) && (mnt->mnt.mnt_flags & MNT_READONLY)) + mnt->mnt.mnt_flags |= MNT_LOCK_READONLY; + atomic_inc(&sb->s_active); mnt->mnt.mnt_sb = sb; mnt->mnt.mnt_root = dget(root); @@ -2342,7 +2346,7 @@ static struct mnt_namespace *dup_mnt_ns(struct mnt_namespace *mnt_ns, /* First pass: copy the tree topology */ copy_flags = CL_COPY_ALL | CL_EXPIRE; if (user_ns != mnt_ns->user_ns) - copy_flags |= CL_SHARED_TO_SLAVE; + copy_flags |= CL_SHARED_TO_SLAVE | CL_UNPRIVILEGED; new = copy_tree(old, old->mnt.mnt_root, copy_flags); if (IS_ERR(new)) { up_write(&namespace_sem); diff --git a/fs/pnode.c b/fs/pnode.c index 3e000a51ac0d..8b29d2164da6 100644 --- a/fs/pnode.c +++ b/fs/pnode.c @@ -9,6 +9,7 @@ #include #include #include +#include #include "internal.h" #include "pnode.h" @@ -220,6 +221,7 @@ static struct mount *get_source(struct mount *dest, int propagate_mnt(struct mount *dest_mnt, struct dentry *dest_dentry, struct mount *source_mnt, struct list_head *tree_list) { + struct user_namespace *user_ns = current->nsproxy->mnt_ns->user_ns; struct mount *m, *child; int ret = 0; struct mount *prev_dest_mnt = dest_mnt; @@ -237,6 +239,10 @@ int propagate_mnt(struct mount *dest_mnt, struct dentry *dest_dentry, source = get_source(m, prev_dest_mnt, prev_src_mnt, &type); + /* Notice when we are propagating across user namespaces */ + if (m->mnt_ns->user_ns != user_ns) + type |= CL_UNPRIVILEGED; + child = copy_tree(source, source->mnt.mnt_root, type); if (IS_ERR(child)) { ret = PTR_ERR(child); diff --git a/fs/pnode.h b/fs/pnode.h index 19b853a3445c..a0493d5ebfbf 100644 --- a/fs/pnode.h +++ b/fs/pnode.h @@ -23,6 +23,7 @@ #define CL_MAKE_SHARED 0x08 #define CL_PRIVATE 0x10 #define CL_SHARED_TO_SLAVE 0x20 +#define CL_UNPRIVILEGED 0x40 static inline void set_mnt_shared(struct mount *mnt) { -- cgit From 87a8ebd637dafc255070f503909a053cf0d98d3f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 24 Mar 2013 14:28:27 -0700 Subject: userns: Restrict when proc and sysfs can be mounted Only allow unprivileged mounts of proc and sysfs if they are already mounted when the user namespace is created. proc and sysfs are interesting because they have content that is per namespace, and so fresh mounts are needed when new namespaces are created while at the same time proc and sysfs have content that is shared between every instance. Respect the policy of who may see the shared content of proc and sysfs by only allowing new mounts if there was an existing mount at the time the user namespace was created. In practice there are only two interesting cases: proc and sysfs are mounted at their usual places, proc and sysfs are not mounted at all (some form of mount namespace jail). Cc: stable@vger.kernel.org Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- fs/namespace.c | 21 +++++++++++++++++++++ fs/proc/root.c | 4 ++++ fs/sysfs/mount.c | 4 ++++ 3 files changed, 29 insertions(+) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 968d4c5eae03..d581e45c0a9f 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -2763,6 +2763,27 @@ bool current_chrooted(void) return chrooted; } +void update_mnt_policy(struct user_namespace *userns) +{ + struct mnt_namespace *ns = current->nsproxy->mnt_ns; + struct mount *mnt; + + down_read(&namespace_sem); + list_for_each_entry(mnt, &ns->list, mnt_list) { + switch (mnt->mnt.mnt_sb->s_magic) { + case SYSFS_MAGIC: + userns->may_mount_sysfs = true; + break; + case PROC_SUPER_MAGIC: + userns->may_mount_proc = true; + break; + } + if (userns->may_mount_sysfs && userns->may_mount_proc) + break; + } + up_read(&namespace_sem); +} + static void *mntns_get(struct task_struct *task) { struct mnt_namespace *ns = NULL; diff --git a/fs/proc/root.c b/fs/proc/root.c index c6e9fac26bac..9c7fab1d23f0 100644 --- a/fs/proc/root.c +++ b/fs/proc/root.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include #include @@ -108,6 +109,9 @@ static struct dentry *proc_mount(struct file_system_type *fs_type, } else { ns = task_active_pid_ns(current); options = data; + + if (!current_user_ns()->may_mount_proc) + return ERR_PTR(-EPERM); } sb = sget(fs_type, proc_test_super, proc_set_super, flags, ns); diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c index 8d924b5ec733..afd83273e6ce 100644 --- a/fs/sysfs/mount.c +++ b/fs/sysfs/mount.c @@ -19,6 +19,7 @@ #include #include #include +#include #include "sysfs.h" @@ -111,6 +112,9 @@ static struct dentry *sysfs_mount(struct file_system_type *fs_type, struct super_block *sb; int error; + if (!(flags & MS_KERNMOUNT) && !current_user_ns()->may_mount_sysfs) + return ERR_PTR(-EPERM); + info = kzalloc(sizeof(*info), GFP_KERNEL); if (!info) return ERR_PTR(-ENOMEM); -- cgit From 3e84f48edfd33b2e209a117c11fb9ce637cc9b67 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 27 Mar 2013 15:20:30 +0000 Subject: vfs/splice: Fix missed checks in new __kernel_write() helper Commit 06ae43f34bcc ("Don't bother with redoing rw_verify_area() from default_file_splice_from()") lost the checks to test existence of the write/aio_write methods. My apologies ;-/ Eventually, we want that in fs/splice.c side of things (no point repeating it for every buffer, after all), but for now this is the obvious minimal fix. Reported-by: Dave Jones Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- fs/read_write.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/read_write.c b/fs/read_write.c index f7b5a23b804b..e6ddc8dceb96 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -424,6 +424,9 @@ ssize_t __kernel_write(struct file *file, const char *buf, size_t count, loff_t const char __user *p; ssize_t ret; + if (!file->f_op || (!file->f_op->write && !file->f_op->aio_write)) + return -EINVAL; + old_fs = get_fs(); set_fs(get_ds()); p = (__force const char __user *)buf; -- cgit From adaa4b8e4d47eeb114513c2f7a172929154b94bd Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Thu, 21 Mar 2013 14:30:23 +0000 Subject: Btrfs: fix EIO from btrfs send in is_extent_unchanged for punched holes When you take a snapshot, punch a hole where there has been data, then take another snapshot and try to send an incremental stream, btrfs send would give you EIO. That is because is_extent_unchanged had no support for holes being punched. With this patch, instead of returning EIO we just return 0 (== the extent is not unchanged) and we're good. Signed-off-by: Jan Schmidt Cc: Alexander Block Signed-off-by: Josef Bacik --- fs/btrfs/send.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index 68da757615ae..ed897dc11356 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -3945,12 +3945,10 @@ static int is_extent_unchanged(struct send_ctx *sctx, found_key.type != key.type) { key.offset += right_len; break; - } else { - if (found_key.offset != key.offset + right_len) { - /* Should really not happen */ - ret = -EIO; - goto out; - } + } + if (found_key.offset != key.offset + right_len) { + ret = 0; + goto out; } key = found_key; } -- cgit From f4881bc7a83eff263789dd524b7c269d138d4af5 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 25 Mar 2013 16:03:35 -0400 Subject: Btrfs: fix space leak when we fail to reserve metadata space Dave reported a warning when running xfstest 275. We have been leaking delalloc metadata space when our reservations fail. This is because we were improperly calculating how much space to free for our checksum reservations. The problem is we would sometimes free up space that had already been freed in another thread and we would end up with negative usage for the delalloc space. This patch fixes the problem by calculating how much space the other threads would have already freed, and then calculate how much space we need to free had we not done the reservation at all, and then freeing any excess space. This makes xfstests 275 no longer have leaked space. Thanks Cc: stable@vger.kernel.org Reported-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 47 +++++++++++++++++++++++++++++++++++++++++------ 1 file changed, 41 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a8ff25aedca1..a22b5cc921ad 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4815,14 +4815,49 @@ out_fail: * If the inodes csum_bytes is the same as the original * csum_bytes then we know we haven't raced with any free()ers * so we can just reduce our inodes csum bytes and carry on. - * Otherwise we have to do the normal free thing to account for - * the case that the free side didn't free up its reserve - * because of this outstanding reservation. */ - if (BTRFS_I(inode)->csum_bytes == csum_bytes) + if (BTRFS_I(inode)->csum_bytes == csum_bytes) { calc_csum_metadata_size(inode, num_bytes, 0); - else - to_free = calc_csum_metadata_size(inode, num_bytes, 0); + } else { + u64 orig_csum_bytes = BTRFS_I(inode)->csum_bytes; + u64 bytes; + + /* + * This is tricky, but first we need to figure out how much we + * free'd from any free-ers that occured during this + * reservation, so we reset ->csum_bytes to the csum_bytes + * before we dropped our lock, and then call the free for the + * number of bytes that were freed while we were trying our + * reservation. + */ + bytes = csum_bytes - BTRFS_I(inode)->csum_bytes; + BTRFS_I(inode)->csum_bytes = csum_bytes; + to_free = calc_csum_metadata_size(inode, bytes, 0); + + + /* + * Now we need to see how much we would have freed had we not + * been making this reservation and our ->csum_bytes were not + * artificially inflated. + */ + BTRFS_I(inode)->csum_bytes = csum_bytes - num_bytes; + bytes = csum_bytes - orig_csum_bytes; + bytes = calc_csum_metadata_size(inode, bytes, 0); + + /* + * Now reset ->csum_bytes to what it should be. If bytes is + * more than to_free then we would have free'd more space had we + * not had an artificially high ->csum_bytes, so we need to free + * the remainder. If bytes is the same or less then we don't + * need to do anything, the other free-ers did the correct + * thing. + */ + BTRFS_I(inode)->csum_bytes = orig_csum_bytes - num_bytes; + if (bytes > to_free) + to_free = bytes - to_free; + else + to_free = 0; + } spin_unlock(&BTRFS_I(inode)->lock); if (dropped) to_free += btrfs_calc_trans_metadata_size(root, dropped); -- cgit From 6e137ed3f30574f314733d4b7a86ea6523232b14 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 26 Mar 2013 15:26:55 -0400 Subject: Btrfs: fix space accounting for unlink and rename We are way over-reserving for unlink and rename. Rename is just some random huge number and unlink accounts for tree log operations that don't actually happen during unlink, not to mention the tree log doesn't take from the trans block rsv anyway so it's completely useless. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 6a6e13c53086..8cab424c75f8 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3693,11 +3693,9 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, * 1 for the dir item * 1 for the dir index * 1 for the inode ref - * 1 for the inode ref in the tree log - * 2 for the dir entries in the log * 1 for the inode */ - trans = btrfs_start_transaction(root, 8); + trans = btrfs_start_transaction(root, 5); if (!IS_ERR(trans) || PTR_ERR(trans) != -ENOSPC) return trans; @@ -8141,7 +8139,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, * inodes. So 5 * 2 is 10, plus 1 for the new link, so 11 total items * should cover the worst case number of items we'll modify. */ - trans = btrfs_start_transaction(root, 20); + trans = btrfs_start_transaction(root, 11); if (IS_ERR(trans)) { ret = PTR_ERR(trans); goto out_notrans; -- cgit From db1d607d3ca5cbb283cbb17d648cd7e8dc67cc7b Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 26 Mar 2013 15:29:11 -0400 Subject: Btrfs: hold the ordered operations mutex when waiting on ordered extents We need to hold the ordered_operations mutex while waiting on ordered extents since we splice and run the ordered extents list. We need to make sure anybody else who wants to wait on ordered extents does actually wait for them to be completed. This will keep us from bailing out of flushing in case somebody is already waiting on ordered extents to complete. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ordered-data.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index dc08d77b717e..005c45db699e 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -557,6 +557,7 @@ void btrfs_wait_ordered_extents(struct btrfs_root *root, int delay_iput) INIT_LIST_HEAD(&splice); INIT_LIST_HEAD(&works); + mutex_lock(&root->fs_info->ordered_operations_mutex); spin_lock(&root->fs_info->ordered_extent_lock); list_splice_init(&root->fs_info->ordered_extents, &splice); while (!list_empty(&splice)) { @@ -600,6 +601,7 @@ void btrfs_wait_ordered_extents(struct btrfs_root *root, int delay_iput) cond_resched(); } + mutex_unlock(&root->fs_info->ordered_operations_mutex); } /* -- cgit From fdf30d1c1b386e1b73116cc7e0fb14e962b763b0 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 26 Mar 2013 15:31:45 -0400 Subject: Btrfs: limit the global reserve to 512mb A user reported a problem where he was getting early ENOSPC with hundreds of gigs of free data space and 6 gigs of free metadata space. This is because the global block reserve was taking up the entire free metadata space. This is ridiculous, we have infrastructure in place to throttle if we start using too much of the global reserve, so instead of letting it get this huge just limit it to 512mb so that users can still get work done. This allowed the user to complete his rsync without issues. Thanks Cc: stable@vger.kernel.org Reported-and-tested-by: Stefan Priebe Signed-off-by: Josef Bacik --- fs/btrfs/extent-tree.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a22b5cc921ad..0d8478700d78 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -4460,7 +4460,7 @@ static void update_global_block_rsv(struct btrfs_fs_info *fs_info) spin_lock(&sinfo->lock); spin_lock(&block_rsv->lock); - block_rsv->size = num_bytes; + block_rsv->size = min_t(u64, num_bytes, 512 * 1024 * 1024); num_bytes = sinfo->bytes_used + sinfo->bytes_pinned + sinfo->bytes_reserved + sinfo->bytes_readonly + -- cgit From a7975026ff9ddf91ba190ae2b71699dd156395e3 Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Mon, 25 Mar 2013 11:08:23 +0000 Subject: Btrfs: fix double free in the btrfs_qgroup_account_ref() The function btrfs_find_all_roots is responsible to allocate memory for 'roots' and free it if errors happen,so the caller should not free it again since the work has been done. Besides,'tmp' is allocated after the function btrfs_find_all_roots, so we can return directly if btrfs_find_all_roots() fails. Signed-off-by: Wang Shilong Reviewed-by: Miao Xie Reviewed-by: Jan Schmidt Signed-off-by: Josef Bacik --- fs/btrfs/qgroup.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 5471e47d6559..b44124dd2370 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -1153,7 +1153,7 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans, ret = btrfs_find_all_roots(trans, fs_info, node->bytenr, sgn > 0 ? node->seq - 1 : node->seq, &roots); if (ret < 0) - goto out; + return ret; spin_lock(&fs_info->qgroup_lock); quota_root = fs_info->quota_root; @@ -1275,7 +1275,6 @@ int btrfs_qgroup_account_ref(struct btrfs_trans_handle *trans, ret = 0; unlock: spin_unlock(&fs_info->qgroup_lock); -out: ulist_free(roots); ulist_free(tmp); -- cgit From 39847c4d3d91f487f9ab3d083ee5d0f8419f105c Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 28 Mar 2013 08:08:20 +0000 Subject: Btrfs: fix wrong reservation of csums We reserve the space for csums only when we write data into a file, in the other cases, such as tree log, log replay, we don't do reservation, so we can use the reservation of the transaction handle just for the former. And for the latter, we should use the tree's own reservation. But the function - btrfs_csum_file_blocks() didn't differentiate between these two types of the cases, fix it. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/file-item.c | 2 -- fs/btrfs/inode.c | 2 ++ 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index ec160202be3e..b7e529d2860f 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -728,7 +728,6 @@ int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans, return -ENOMEM; sector_sum = sums->sums; - trans->adding_csums = 1; again: next_offset = (u64)-1; found_next = 0; @@ -899,7 +898,6 @@ next_sector: goto again; } out: - trans->adding_csums = 0; btrfs_free_path(path); return ret; diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 8cab424c75f8..b88381582dab 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1757,8 +1757,10 @@ static noinline int add_pending_csums(struct btrfs_trans_handle *trans, struct btrfs_ordered_sum *sum; list_for_each_entry(sum, list, list) { + trans->adding_csums = 1; btrfs_csum_file_blocks(trans, BTRFS_I(inode)->root->fs_info->csum_root, sum); + trans->adding_csums = 0; } return 0; } -- cgit From 82d130ff390be67d980d8b6f39e921c0b1d8d8e0 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Thu, 28 Mar 2013 08:12:15 +0000 Subject: Btrfs: fix wrong return value of btrfs_lookup_csum() If we don't find the expected csum item, but find a csum item which is adjacent to the specified extent, we should return -EFBIG, or we should return -ENOENT. But btrfs_lookup_csum() return -EFBIG even the csum item is not adjacent to the specified extent. Fix it. Signed-off-by: Miao Xie Signed-off-by: Josef Bacik --- fs/btrfs/file-item.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index b7e529d2860f..c4628a201cb3 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -118,9 +118,11 @@ struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans, csums_in_item = btrfs_item_size_nr(leaf, path->slots[0]); csums_in_item /= csum_size; - if (csum_offset >= csums_in_item) { + if (csum_offset == csums_in_item) { ret = -EFBIG; goto fail; + } else if (csum_offset > csums_in_item) { + goto fail; } } item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_csum_item); -- cgit From d8fe29e9dea8d7d61fd140d8779326856478fc62 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 29 Mar 2013 08:09:34 -0600 Subject: Btrfs: don't drop path when printing out tree errors in scrub A user reported a panic where we were panicing somewhere in tree_backref_for_extent from scrub_print_warning. He only captured the trace but looking at scrub_print_warning we drop the path right before we mess with the extent buffer to print out a bunch of stuff, which isn't right. So fix this by dropping the path after we use the eb if we need to. Thanks, Cc: stable@vger.kernel.org Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/scrub.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index 53c3501fa4ca..85e072b956d5 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -542,7 +542,6 @@ static void scrub_print_warning(const char *errstr, struct scrub_block *sblock) eb = path->nodes[0]; ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item); item_size = btrfs_item_size_nr(eb, path->slots[0]); - btrfs_release_path(path); if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { do { @@ -558,7 +557,9 @@ static void scrub_print_warning(const char *errstr, struct scrub_block *sblock) ret < 0 ? -1 : ref_level, ret < 0 ? -1 : ref_root); } while (ret != 1); + btrfs_release_path(path); } else { + btrfs_release_path(path); swarn.path = path; swarn.dev = dev; iterate_extent_inodes(fs_info, found_key.objectid, -- cgit From d68772b7c83f4b518be15ae96f4827c8ed02f684 Mon Sep 17 00:00:00 2001 From: Matt Fleming Date: Fri, 8 Feb 2013 16:27:24 +0000 Subject: efivarfs: Move to fs/efivarfs Now that efivarfs uses the efivar API, move it out of efivars.c and into fs/efivarfs where it belongs. This move will eventually allow us to enable the efivarfs code without having to also enable CONFIG_EFI_VARS built, and vice versa. Furthermore, things like, mount -t efivarfs none /sys/firmware/efi/efivars will now work if efivarfs is built as a module without requiring the use of MODULE_ALIAS(), which would have been necessary when the efivarfs code was part of efivars.c. Cc: Matthew Garrett Cc: Jeremy Kerr Reviewed-by: Tom Gundersen Tested-by: Tom Gundersen Signed-off-by: Matt Fleming --- fs/Kconfig | 1 + fs/Makefile | 1 + fs/efivarfs/Kconfig | 12 +++ fs/efivarfs/Makefile | 7 ++ fs/efivarfs/file.c | 111 ++++++++++++++++++++ fs/efivarfs/inode.c | 173 ++++++++++++++++++++++++++++++++ fs/efivarfs/internal.h | 22 ++++ fs/efivarfs/super.c | 267 +++++++++++++++++++++++++++++++++++++++++++++++++ 8 files changed, 594 insertions(+) create mode 100644 fs/efivarfs/Kconfig create mode 100644 fs/efivarfs/Makefile create mode 100644 fs/efivarfs/file.c create mode 100644 fs/efivarfs/inode.c create mode 100644 fs/efivarfs/internal.h create mode 100644 fs/efivarfs/super.c (limited to 'fs') diff --git a/fs/Kconfig b/fs/Kconfig index 780725a463b1..c229f828eb01 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -211,6 +211,7 @@ source "fs/sysv/Kconfig" source "fs/ufs/Kconfig" source "fs/exofs/Kconfig" source "fs/f2fs/Kconfig" +source "fs/efivarfs/Kconfig" endif # MISC_FILESYSTEMS diff --git a/fs/Makefile b/fs/Makefile index 9d53192236fc..0fde6a369a72 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -127,3 +127,4 @@ obj-$(CONFIG_F2FS_FS) += f2fs/ obj-y += exofs/ # Multiple modules obj-$(CONFIG_CEPH_FS) += ceph/ obj-$(CONFIG_PSTORE) += pstore/ +obj-$(CONFIG_EFIVAR_FS) += efivarfs/ diff --git a/fs/efivarfs/Kconfig b/fs/efivarfs/Kconfig new file mode 100644 index 000000000000..1fb2b7f849f3 --- /dev/null +++ b/fs/efivarfs/Kconfig @@ -0,0 +1,12 @@ +config EFIVAR_FS + tristate "EFI Variable filesystem" + depends on EFI_VARS + help + efivarfs is a replacement filesystem for the old EFI + variable support via sysfs, as it doesn't suffer from the + same 1024-byte variable size limit. + + To compile this file system support as a module, choose M + here. The module will be called efivarfs. + + If unsure, say N. diff --git a/fs/efivarfs/Makefile b/fs/efivarfs/Makefile new file mode 100644 index 000000000000..955d478177d5 --- /dev/null +++ b/fs/efivarfs/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the efivarfs filesystem +# + +obj-$(CONFIG_EFIVAR_FS) += efivarfs.o + +efivarfs-objs := inode.o file.o super.o diff --git a/fs/efivarfs/file.c b/fs/efivarfs/file.c new file mode 100644 index 000000000000..aeb0368dace2 --- /dev/null +++ b/fs/efivarfs/file.c @@ -0,0 +1,111 @@ +/* + * Copyright (C) 2012 Red Hat, Inc. + * Copyright (C) 2012 Jeremy Kerr + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include + +#include "internal.h" + +static int efivarfs_file_open(struct inode *inode, struct file *file) +{ + file->private_data = inode->i_private; + return 0; +} + +static ssize_t efivarfs_file_write(struct file *file, + const char __user *userbuf, size_t count, loff_t *ppos) +{ + struct efivar_entry *var = file->private_data; + void *data; + u32 attributes; + struct inode *inode = file->f_mapping->host; + unsigned long datasize = count - sizeof(attributes); + ssize_t bytes = 0; + bool set = false; + + if (count < sizeof(attributes)) + return -EINVAL; + + if (copy_from_user(&attributes, userbuf, sizeof(attributes))) + return -EFAULT; + + if (attributes & ~(EFI_VARIABLE_MASK)) + return -EINVAL; + + data = kmalloc(datasize, GFP_KERNEL); + if (!data) + return -ENOMEM; + + if (copy_from_user(data, userbuf + sizeof(attributes), datasize)) { + bytes = -EFAULT; + goto out; + } + + bytes = efivar_entry_set_get_size(var, attributes, &datasize, + data, &set); + if (!set && bytes) + goto out; + + if (bytes == -ENOENT) { + drop_nlink(inode); + d_delete(file->f_dentry); + dput(file->f_dentry); + } else { + mutex_lock(&inode->i_mutex); + i_size_write(inode, datasize + sizeof(attributes)); + mutex_unlock(&inode->i_mutex); + } + + bytes = count; + +out: + kfree(data); + + return bytes; +} + +static ssize_t efivarfs_file_read(struct file *file, char __user *userbuf, + size_t count, loff_t *ppos) +{ + struct efivar_entry *var = file->private_data; + unsigned long datasize = 0; + u32 attributes; + void *data; + ssize_t size = 0; + int err; + + err = efivar_entry_size(var, &datasize); + if (err) + return err; + + data = kmalloc(datasize + sizeof(attributes), GFP_KERNEL); + + if (!data) + return -ENOMEM; + + size = efivar_entry_get(var, &attributes, &datasize, + data + sizeof(attributes)); + if (size) + goto out_free; + + memcpy(data, &attributes, sizeof(attributes)); + size = simple_read_from_buffer(userbuf, count, ppos, + data, datasize + sizeof(attributes)); +out_free: + kfree(data); + + return size; +} + +const struct file_operations efivarfs_file_operations = { + .open = efivarfs_file_open, + .read = efivarfs_file_read, + .write = efivarfs_file_write, + .llseek = no_llseek, +}; diff --git a/fs/efivarfs/inode.c b/fs/efivarfs/inode.c new file mode 100644 index 000000000000..640e289d522e --- /dev/null +++ b/fs/efivarfs/inode.c @@ -0,0 +1,173 @@ +/* + * Copyright (C) 2012 Red Hat, Inc. + * Copyright (C) 2012 Jeremy Kerr + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include +#include + +#include "internal.h" + +struct inode *efivarfs_get_inode(struct super_block *sb, + const struct inode *dir, int mode, dev_t dev) +{ + struct inode *inode = new_inode(sb); + + if (inode) { + inode->i_ino = get_next_ino(); + inode->i_mode = mode; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + switch (mode & S_IFMT) { + case S_IFREG: + inode->i_fop = &efivarfs_file_operations; + break; + case S_IFDIR: + inode->i_op = &efivarfs_dir_inode_operations; + inode->i_fop = &simple_dir_operations; + inc_nlink(inode); + break; + } + } + return inode; +} + +/* + * Return true if 'str' is a valid efivarfs filename of the form, + * + * VariableName-12345678-1234-1234-1234-1234567891bc + */ +bool efivarfs_valid_name(const char *str, int len) +{ + static const char dashes[EFI_VARIABLE_GUID_LEN] = { + [8] = 1, [13] = 1, [18] = 1, [23] = 1 + }; + const char *s = str + len - EFI_VARIABLE_GUID_LEN; + int i; + + /* + * We need a GUID, plus at least one letter for the variable name, + * plus the '-' separator + */ + if (len < EFI_VARIABLE_GUID_LEN + 2) + return false; + + /* GUID must be preceded by a '-' */ + if (*(s - 1) != '-') + return false; + + /* + * Validate that 's' is of the correct format, e.g. + * + * 12345678-1234-1234-1234-123456789abc + */ + for (i = 0; i < EFI_VARIABLE_GUID_LEN; i++) { + if (dashes[i]) { + if (*s++ != '-') + return false; + } else { + if (!isxdigit(*s++)) + return false; + } + } + + return true; +} + +static void efivarfs_hex_to_guid(const char *str, efi_guid_t *guid) +{ + guid->b[0] = hex_to_bin(str[6]) << 4 | hex_to_bin(str[7]); + guid->b[1] = hex_to_bin(str[4]) << 4 | hex_to_bin(str[5]); + guid->b[2] = hex_to_bin(str[2]) << 4 | hex_to_bin(str[3]); + guid->b[3] = hex_to_bin(str[0]) << 4 | hex_to_bin(str[1]); + guid->b[4] = hex_to_bin(str[11]) << 4 | hex_to_bin(str[12]); + guid->b[5] = hex_to_bin(str[9]) << 4 | hex_to_bin(str[10]); + guid->b[6] = hex_to_bin(str[16]) << 4 | hex_to_bin(str[17]); + guid->b[7] = hex_to_bin(str[14]) << 4 | hex_to_bin(str[15]); + guid->b[8] = hex_to_bin(str[19]) << 4 | hex_to_bin(str[20]); + guid->b[9] = hex_to_bin(str[21]) << 4 | hex_to_bin(str[22]); + guid->b[10] = hex_to_bin(str[24]) << 4 | hex_to_bin(str[25]); + guid->b[11] = hex_to_bin(str[26]) << 4 | hex_to_bin(str[27]); + guid->b[12] = hex_to_bin(str[28]) << 4 | hex_to_bin(str[29]); + guid->b[13] = hex_to_bin(str[30]) << 4 | hex_to_bin(str[31]); + guid->b[14] = hex_to_bin(str[32]) << 4 | hex_to_bin(str[33]); + guid->b[15] = hex_to_bin(str[34]) << 4 | hex_to_bin(str[35]); +} + +static int efivarfs_create(struct inode *dir, struct dentry *dentry, + umode_t mode, bool excl) +{ + struct inode *inode; + struct efivar_entry *var; + int namelen, i = 0, err = 0; + + if (!efivarfs_valid_name(dentry->d_name.name, dentry->d_name.len)) + return -EINVAL; + + inode = efivarfs_get_inode(dir->i_sb, dir, mode, 0); + if (!inode) + return -ENOMEM; + + var = kzalloc(sizeof(struct efivar_entry), GFP_KERNEL); + if (!var) { + err = -ENOMEM; + goto out; + } + + /* length of the variable name itself: remove GUID and separator */ + namelen = dentry->d_name.len - EFI_VARIABLE_GUID_LEN - 1; + + efivarfs_hex_to_guid(dentry->d_name.name + namelen + 1, + &var->var.VendorGuid); + + for (i = 0; i < namelen; i++) + var->var.VariableName[i] = dentry->d_name.name[i]; + + var->var.VariableName[i] = '\0'; + + inode->i_private = var; + + efivar_entry_add(var, &efivarfs_list); + d_instantiate(dentry, inode); + dget(dentry); +out: + if (err) { + kfree(var); + iput(inode); + } + return err; +} + +static int efivarfs_unlink(struct inode *dir, struct dentry *dentry) +{ + struct efivar_entry *var = dentry->d_inode->i_private; + + if (efivar_entry_delete(var)) + return -EINVAL; + + drop_nlink(dentry->d_inode); + dput(dentry); + return 0; +}; + +/* + * Handle negative dentry. + */ +static struct dentry *efivarfs_lookup(struct inode *dir, struct dentry *dentry, + unsigned int flags) +{ + if (dentry->d_name.len > NAME_MAX) + return ERR_PTR(-ENAMETOOLONG); + d_add(dentry, NULL); + return NULL; +} + +const struct inode_operations efivarfs_dir_inode_operations = { + .lookup = efivarfs_lookup, + .unlink = efivarfs_unlink, + .create = efivarfs_create, +}; diff --git a/fs/efivarfs/internal.h b/fs/efivarfs/internal.h new file mode 100644 index 000000000000..b5ff16addb7c --- /dev/null +++ b/fs/efivarfs/internal.h @@ -0,0 +1,22 @@ +/* + * Copyright (C) 2012 Red Hat, Inc. + * Copyright (C) 2012 Jeremy Kerr + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#ifndef EFIVAR_FS_INTERNAL_H +#define EFIVAR_FS_INTERNAL_H + +#include + +extern const struct file_operations efivarfs_file_operations; +extern const struct inode_operations efivarfs_dir_inode_operations; +extern bool efivarfs_valid_name(const char *str, int len); +extern struct inode *efivarfs_get_inode(struct super_block *sb, + const struct inode *dir, int mode, dev_t dev); + +extern struct list_head efivarfs_list; + +#endif /* EFIVAR_FS_INTERNAL_H */ diff --git a/fs/efivarfs/super.c b/fs/efivarfs/super.c new file mode 100644 index 000000000000..34c48f1fcdbc --- /dev/null +++ b/fs/efivarfs/super.c @@ -0,0 +1,267 @@ +/* + * Copyright (C) 2012 Red Hat, Inc. + * Copyright (C) 2012 Jeremy Kerr + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include +#include +#include +#include + +#include "internal.h" + +LIST_HEAD(efivarfs_list); + +static void efivarfs_evict_inode(struct inode *inode) +{ + clear_inode(inode); +} + +static const struct super_operations efivarfs_ops = { + .statfs = simple_statfs, + .drop_inode = generic_delete_inode, + .evict_inode = efivarfs_evict_inode, + .show_options = generic_show_options, +}; + +static struct super_block *efivarfs_sb; + +/* + * Compare two efivarfs file names. + * + * An efivarfs filename is composed of two parts, + * + * 1. A case-sensitive variable name + * 2. A case-insensitive GUID + * + * So we need to perform a case-sensitive match on part 1 and a + * case-insensitive match on part 2. + */ +static int efivarfs_d_compare(const struct dentry *parent, const struct inode *pinode, + const struct dentry *dentry, const struct inode *inode, + unsigned int len, const char *str, + const struct qstr *name) +{ + int guid = len - EFI_VARIABLE_GUID_LEN; + + if (name->len != len) + return 1; + + /* Case-sensitive compare for the variable name */ + if (memcmp(str, name->name, guid)) + return 1; + + /* Case-insensitive compare for the GUID */ + return strncasecmp(name->name + guid, str + guid, EFI_VARIABLE_GUID_LEN); +} + +static int efivarfs_d_hash(const struct dentry *dentry, + const struct inode *inode, struct qstr *qstr) +{ + unsigned long hash = init_name_hash(); + const unsigned char *s = qstr->name; + unsigned int len = qstr->len; + + if (!efivarfs_valid_name(s, len)) + return -EINVAL; + + while (len-- > EFI_VARIABLE_GUID_LEN) + hash = partial_name_hash(*s++, hash); + + /* GUID is case-insensitive. */ + while (len--) + hash = partial_name_hash(tolower(*s++), hash); + + qstr->hash = end_name_hash(hash); + return 0; +} + +/* + * Retaining negative dentries for an in-memory filesystem just wastes + * memory and lookup time: arrange for them to be deleted immediately. + */ +static int efivarfs_delete_dentry(const struct dentry *dentry) +{ + return 1; +} + +static struct dentry_operations efivarfs_d_ops = { + .d_compare = efivarfs_d_compare, + .d_hash = efivarfs_d_hash, + .d_delete = efivarfs_delete_dentry, +}; + +static struct dentry *efivarfs_alloc_dentry(struct dentry *parent, char *name) +{ + struct dentry *d; + struct qstr q; + int err; + + q.name = name; + q.len = strlen(name); + + err = efivarfs_d_hash(NULL, NULL, &q); + if (err) + return ERR_PTR(err); + + d = d_alloc(parent, &q); + if (d) + return d; + + return ERR_PTR(-ENOMEM); +} + +static int efivarfs_callback(efi_char16_t *name16, efi_guid_t vendor, + unsigned long name_size, void *data) +{ + struct super_block *sb = (struct super_block *)data; + struct efivar_entry *entry; + struct inode *inode = NULL; + struct dentry *dentry, *root = sb->s_root; + unsigned long size = 0; + char *name; + int len, i; + int err = -ENOMEM; + + entry = kmalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) + return err; + + memcpy(entry->var.VariableName, name16, name_size); + memcpy(&(entry->var.VendorGuid), &vendor, sizeof(efi_guid_t)); + + len = utf16_strlen(entry->var.VariableName); + + /* name, plus '-', plus GUID, plus NUL*/ + name = kmalloc(len + 1 + EFI_VARIABLE_GUID_LEN + 1, GFP_KERNEL); + if (!name) + goto fail; + + for (i = 0; i < len; i++) + name[i] = entry->var.VariableName[i] & 0xFF; + + name[len] = '-'; + + efi_guid_unparse(&entry->var.VendorGuid, name + len + 1); + + name[len + EFI_VARIABLE_GUID_LEN+1] = '\0'; + + inode = efivarfs_get_inode(sb, root->d_inode, S_IFREG | 0644, 0); + if (!inode) + goto fail_name; + + dentry = efivarfs_alloc_dentry(root, name); + if (IS_ERR(dentry)) { + err = PTR_ERR(dentry); + goto fail_inode; + } + + /* copied by the above to local storage in the dentry. */ + kfree(name); + + efivar_entry_size(entry, &size); + efivar_entry_add(entry, &efivarfs_list); + + mutex_lock(&inode->i_mutex); + inode->i_private = entry; + i_size_write(inode, size + sizeof(entry->var.Attributes)); + mutex_unlock(&inode->i_mutex); + d_add(dentry, inode); + + return 0; + +fail_inode: + iput(inode); +fail_name: + kfree(name); +fail: + kfree(entry); + return err; +} + +static int efivarfs_destroy(struct efivar_entry *entry, void *data) +{ + efivar_entry_remove(entry); + kfree(entry); + return 0; +} + +static int efivarfs_fill_super(struct super_block *sb, void *data, int silent) +{ + struct inode *inode = NULL; + struct dentry *root; + int err; + + efivarfs_sb = sb; + + sb->s_maxbytes = MAX_LFS_FILESIZE; + sb->s_blocksize = PAGE_CACHE_SIZE; + sb->s_blocksize_bits = PAGE_CACHE_SHIFT; + sb->s_magic = EFIVARFS_MAGIC; + sb->s_op = &efivarfs_ops; + sb->s_d_op = &efivarfs_d_ops; + sb->s_time_gran = 1; + + inode = efivarfs_get_inode(sb, NULL, S_IFDIR | 0755, 0); + if (!inode) + return -ENOMEM; + inode->i_op = &efivarfs_dir_inode_operations; + + root = d_make_root(inode); + sb->s_root = root; + if (!root) + return -ENOMEM; + + INIT_LIST_HEAD(&efivarfs_list); + + err = efivar_init(efivarfs_callback, (void *)sb, false, + true, &efivarfs_list); + if (err) + __efivar_entry_iter(efivarfs_destroy, &efivarfs_list, NULL, NULL); + + return err; +} + +static struct dentry *efivarfs_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *data) +{ + return mount_single(fs_type, flags, data, efivarfs_fill_super); +} + +static void efivarfs_kill_sb(struct super_block *sb) +{ + kill_litter_super(sb); + efivarfs_sb = NULL; + + /* Remove all entries and destroy */ + __efivar_entry_iter(efivarfs_destroy, &efivarfs_list, NULL, NULL); +} + +static struct file_system_type efivarfs_type = { + .name = "efivarfs", + .mount = efivarfs_mount, + .kill_sb = efivarfs_kill_sb, +}; + +static __init int efivarfs_init(void) +{ + if (!efi_enabled(EFI_RUNTIME_SERVICES)) + return 0; + + if (!efivars_kobject()) + return 0; + + return register_filesystem(&efivarfs_type); +} + +MODULE_AUTHOR("Matthew Garrett, Jeremy Kerr"); +MODULE_DESCRIPTION("EFI Variable Filesystem"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS_FS("efivarfs"); + +module_init(efivarfs_init); -- cgit From a9499fa7cd3fd4824a7202d00c766b269fa3bda6 Mon Sep 17 00:00:00 2001 From: Tom Gundersen Date: Fri, 8 Feb 2013 15:37:06 +0000 Subject: efi: split efisubsystem from efivars This registers /sys/firmware/efi/{,systab,efivars/} whenever EFI is enabled and the system is booted with EFI. This allows *) userspace to check for the existence of /sys/firmware/efi as a way to determine whether or it is running on an EFI system. *) 'mount -t efivarfs none /sys/firmware/efi/efivars' without manually loading any modules. [ Also, move the efivar API into vars.c and unconditionally compile it. This allows us to move efivars.c, which now only contains the sysfs variable code, into the firmware/efi directory. Note that the efivars.c filename is kept to maintain backwards compatability with the old efivars.ko module. With this patch it is now possible for efivarfs to be built without CONFIG_EFI_VARS - Matt ] Cc: Seiji Aguchi Cc: Tony Luck Cc: Mike Waychison Cc: Kay Sievers Cc: Jeremy Kerr Cc: Matthew Garrett Cc: Chun-Yi Lee Cc: Andy Whitcroft Cc: Tobias Powalowski Signed-off-by: Tom Gundersen Signed-off-by: Matt Fleming --- fs/efivarfs/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/efivarfs/Kconfig b/fs/efivarfs/Kconfig index 1fb2b7f849f3..367bbb10c543 100644 --- a/fs/efivarfs/Kconfig +++ b/fs/efivarfs/Kconfig @@ -1,6 +1,6 @@ config EFIVAR_FS tristate "EFI Variable filesystem" - depends on EFI_VARS + depends on EFI help efivarfs is a replacement filesystem for the old EFI variable support via sysfs, as it doesn't suffer from the -- cgit