diff options
Diffstat (limited to 'fs')
33 files changed, 790 insertions, 302 deletions
@@ -1021,6 +1021,7 @@ void aio_complete(struct kiocb *iocb, long res, long res2) /* everything turned out well, dispose of the aiocb. */ kiocb_free(iocb); + put_reqs_available(ctx, 1); /* * We have to order our ring_info tail store above and test @@ -1062,6 +1063,9 @@ static long aio_read_events_ring(struct kioctx *ctx, if (head == tail) goto out; + head %= ctx->nr_events; + tail %= ctx->nr_events; + while (ret < nr) { long avail; struct io_event *ev; @@ -1100,8 +1104,6 @@ static long aio_read_events_ring(struct kioctx *ctx, flush_dcache_page(ctx->ring_pages[0]); pr_debug("%li h%u t%u\n", ret, head, tail); - - put_reqs_available(ctx, ret); out: mutex_unlock(&ctx->ring_lock); diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index b7e2c1c1ef36..be91397f4e92 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -1259,11 +1259,19 @@ struct btrfs_block_group_cache { spinlock_t lock; u64 pinned; u64 reserved; + u64 delalloc_bytes; u64 bytes_super; u64 flags; u64 sectorsize; u64 cache_generation; + /* + * It is just used for the delayed data space allocation because + * only the data space allocation and the relative metadata update + * can be done cross the transaction. + */ + struct rw_semaphore data_rwsem; + /* for raid56, this is a full stripe, without parity */ unsigned long full_stripe_len; @@ -3316,7 +3324,7 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, struct btrfs_key *ins); int btrfs_reserve_extent(struct btrfs_root *root, u64 num_bytes, u64 min_alloc_size, u64 empty_size, u64 hint_byte, - struct btrfs_key *ins, int is_data); + struct btrfs_key *ins, int is_data, int delalloc); int btrfs_inc_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf, int full_backref, int no_quota); int btrfs_dec_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -3330,7 +3338,8 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, u64 bytenr, u64 num_bytes, u64 parent, u64 root_objectid, u64 owner, u64 offset, int no_quota); -int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len); +int btrfs_free_reserved_extent(struct btrfs_root *root, u64 start, u64 len, + int delalloc); int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, u64 start, u64 len); void btrfs_prepare_extent_commit(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index fafb3e53ecde..99c253918208 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -105,7 +105,8 @@ static int find_next_key(struct btrfs_path *path, int level, 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); + u64 num_bytes, int reserve, + int delalloc); static int block_rsv_use_bytes(struct btrfs_block_rsv *block_rsv, u64 num_bytes); int btrfs_pin_extent(struct btrfs_root *root, @@ -3260,7 +3261,8 @@ again: spin_lock(&block_group->lock); if (block_group->cached != BTRFS_CACHE_FINISHED || - !btrfs_test_opt(root, SPACE_CACHE)) { + !btrfs_test_opt(root, SPACE_CACHE) || + block_group->delalloc_bytes) { /* * don't bother trying to write stuff out _if_ * a) we're not cached, @@ -5613,6 +5615,7 @@ int btrfs_exclude_logged_extents(struct btrfs_root *log, * @cache: The cache we are manipulating * @num_bytes: The number of bytes in question * @reserve: One of the reservation enums + * @delalloc: The blocks are allocated for the delalloc write * * This is called by the allocator when it reserves space, or by somebody who is * freeing space that was never actually used on disk. For example if you @@ -5631,7 +5634,7 @@ int btrfs_exclude_logged_extents(struct btrfs_root *log, * succeeds. */ static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, - u64 num_bytes, int reserve) + u64 num_bytes, int reserve, int delalloc) { struct btrfs_space_info *space_info = cache->space_info; int ret = 0; @@ -5650,12 +5653,18 @@ static int btrfs_update_reserved_bytes(struct btrfs_block_group_cache *cache, num_bytes, 0); space_info->bytes_may_use -= num_bytes; } + + if (delalloc) + cache->delalloc_bytes += num_bytes; } } else { if (cache->ro) space_info->bytes_readonly += num_bytes; cache->reserved -= num_bytes; space_info->bytes_reserved -= num_bytes; + + if (delalloc) + cache->delalloc_bytes -= num_bytes; } spin_unlock(&cache->lock); spin_unlock(&space_info->lock); @@ -6206,7 +6215,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags)); btrfs_add_free_space(cache, buf->start, buf->len); - btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE); + btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE, 0); trace_btrfs_reserved_extent_free(root, buf->start, buf->len); pin = 0; } @@ -6365,6 +6374,70 @@ enum btrfs_loop_type { LOOP_NO_EMPTY_SIZE = 3, }; +static inline void +btrfs_lock_block_group(struct btrfs_block_group_cache *cache, + int delalloc) +{ + if (delalloc) + down_read(&cache->data_rwsem); +} + +static inline void +btrfs_grab_block_group(struct btrfs_block_group_cache *cache, + int delalloc) +{ + btrfs_get_block_group(cache); + if (delalloc) + down_read(&cache->data_rwsem); +} + +static struct btrfs_block_group_cache * +btrfs_lock_cluster(struct btrfs_block_group_cache *block_group, + struct btrfs_free_cluster *cluster, + int delalloc) +{ + struct btrfs_block_group_cache *used_bg; + bool locked = false; +again: + spin_lock(&cluster->refill_lock); + if (locked) { + if (used_bg == cluster->block_group) + return used_bg; + + up_read(&used_bg->data_rwsem); + btrfs_put_block_group(used_bg); + } + + used_bg = cluster->block_group; + if (!used_bg) + return NULL; + + if (used_bg == block_group) + return used_bg; + + btrfs_get_block_group(used_bg); + + if (!delalloc) + return used_bg; + + if (down_read_trylock(&used_bg->data_rwsem)) + return used_bg; + + spin_unlock(&cluster->refill_lock); + down_read(&used_bg->data_rwsem); + locked = true; + goto again; +} + +static inline void +btrfs_release_block_group(struct btrfs_block_group_cache *cache, + int delalloc) +{ + if (delalloc) + up_read(&cache->data_rwsem); + btrfs_put_block_group(cache); +} + /* * walks the btree of allocated extents and find a hole of a given size. * The key ins is changed to record the hole: @@ -6379,7 +6452,7 @@ enum btrfs_loop_type { static noinline int find_free_extent(struct btrfs_root *orig_root, u64 num_bytes, u64 empty_size, u64 hint_byte, struct btrfs_key *ins, - u64 flags) + u64 flags, int delalloc) { int ret = 0; struct btrfs_root *root = orig_root->fs_info->extent_root; @@ -6467,6 +6540,7 @@ static noinline int find_free_extent(struct btrfs_root *orig_root, up_read(&space_info->groups_sem); } else { index = get_block_group_index(block_group); + btrfs_lock_block_group(block_group, delalloc); goto have_block_group; } } else if (block_group) { @@ -6481,7 +6555,7 @@ search: u64 offset; int cached; - btrfs_get_block_group(block_group); + btrfs_grab_block_group(block_group, delalloc); search_start = block_group->key.objectid; /* @@ -6529,16 +6603,16 @@ have_block_group: * the refill lock keeps out other * people trying to start a new cluster */ - spin_lock(&last_ptr->refill_lock); - used_block_group = last_ptr->block_group; - if (used_block_group != block_group && - (!used_block_group || - used_block_group->ro || - !block_group_bits(used_block_group, flags))) + used_block_group = btrfs_lock_cluster(block_group, + last_ptr, + delalloc); + if (!used_block_group) goto refill_cluster; - if (used_block_group != block_group) - btrfs_get_block_group(used_block_group); + if (used_block_group != block_group && + (used_block_group->ro || + !block_group_bits(used_block_group, flags))) + goto release_cluster; offset = btrfs_alloc_from_cluster(used_block_group, last_ptr, @@ -6552,16 +6626,15 @@ have_block_group: used_block_group, search_start, num_bytes); if (used_block_group != block_group) { - btrfs_put_block_group(block_group); + btrfs_release_block_group(block_group, + delalloc); block_group = used_block_group; } goto checks; } WARN_ON(last_ptr->block_group != used_block_group); - if (used_block_group != block_group) - btrfs_put_block_group(used_block_group); -refill_cluster: +release_cluster: /* If we are on LOOP_NO_EMPTY_SIZE, we can't * set up a new clusters, so lets just skip it * and let the allocator find whatever block @@ -6578,8 +6651,10 @@ refill_cluster: * succeeding in the unclustered * allocation. */ if (loop >= LOOP_NO_EMPTY_SIZE && - last_ptr->block_group != block_group) { + used_block_group != block_group) { spin_unlock(&last_ptr->refill_lock); + btrfs_release_block_group(used_block_group, + delalloc); goto unclustered_alloc; } @@ -6589,6 +6664,10 @@ refill_cluster: */ btrfs_return_cluster_to_free_space(NULL, last_ptr); + if (used_block_group != block_group) + btrfs_release_block_group(used_block_group, + delalloc); +refill_cluster: if (loop >= LOOP_NO_EMPTY_SIZE) { spin_unlock(&last_ptr->refill_lock); goto unclustered_alloc; @@ -6696,7 +6775,7 @@ checks: BUG_ON(offset > search_start); ret = btrfs_update_reserved_bytes(block_group, num_bytes, - alloc_type); + alloc_type, delalloc); if (ret == -EAGAIN) { btrfs_add_free_space(block_group, offset, num_bytes); goto loop; @@ -6708,13 +6787,13 @@ checks: trace_btrfs_reserve_extent(orig_root, block_group, search_start, num_bytes); - btrfs_put_block_group(block_group); + btrfs_release_block_group(block_group, delalloc); break; loop: failed_cluster_refill = false; failed_alloc = false; BUG_ON(index != get_block_group_index(block_group)); - btrfs_put_block_group(block_group); + btrfs_release_block_group(block_group, delalloc); } up_read(&space_info->groups_sem); @@ -6827,7 +6906,7 @@ again: int btrfs_reserve_extent(struct btrfs_root *root, u64 num_bytes, u64 min_alloc_size, u64 empty_size, u64 hint_byte, - struct btrfs_key *ins, int is_data) + struct btrfs_key *ins, int is_data, int delalloc) { bool final_tried = false; u64 flags; @@ -6837,7 +6916,7 @@ int btrfs_reserve_extent(struct btrfs_root *root, again: WARN_ON(num_bytes < root->sectorsize); ret = find_free_extent(root, num_bytes, empty_size, hint_byte, ins, - flags); + flags, delalloc); if (ret == -ENOSPC) { if (!final_tried && ins->offset) { @@ -6862,7 +6941,8 @@ again: } static int __btrfs_free_reserved_extent(struct btrfs_root *root, - u64 start, u64 len, int pin) + u64 start, u64 len, + int pin, int delalloc) { struct btrfs_block_group_cache *cache; int ret = 0; @@ -6881,7 +6961,7 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root, pin_down_extent(root, cache, start, len, 1); else { btrfs_add_free_space(cache, start, len); - btrfs_update_reserved_bytes(cache, len, RESERVE_FREE); + btrfs_update_reserved_bytes(cache, len, RESERVE_FREE, delalloc); } btrfs_put_block_group(cache); @@ -6891,15 +6971,15 @@ static int __btrfs_free_reserved_extent(struct btrfs_root *root, } int btrfs_free_reserved_extent(struct btrfs_root *root, - u64 start, u64 len) + u64 start, u64 len, int delalloc) { - return __btrfs_free_reserved_extent(root, start, len, 0); + return __btrfs_free_reserved_extent(root, start, len, 0, delalloc); } int btrfs_free_and_pin_reserved_extent(struct btrfs_root *root, u64 start, u64 len) { - return __btrfs_free_reserved_extent(root, start, len, 1); + return __btrfs_free_reserved_extent(root, start, len, 1, 0); } static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans, @@ -7114,7 +7194,7 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, return -EINVAL; ret = btrfs_update_reserved_bytes(block_group, ins->offset, - RESERVE_ALLOC_NO_ACCOUNT); + RESERVE_ALLOC_NO_ACCOUNT, 0); BUG_ON(ret); /* logic error */ ret = alloc_reserved_file_extent(trans, root, 0, root_objectid, 0, owner, offset, ins, 1); @@ -7256,7 +7336,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, return ERR_CAST(block_rsv); ret = btrfs_reserve_extent(root, blocksize, blocksize, - empty_size, hint, &ins, 0); + empty_size, hint, &ins, 0, 0); if (ret) { unuse_block_rsv(root->fs_info, block_rsv, blocksize); return ERR_PTR(ret); @@ -8659,6 +8739,7 @@ btrfs_create_block_group_cache(struct btrfs_root *root, u64 start, u64 size) start); atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); + init_rwsem(&cache->data_rwsem); INIT_LIST_HEAD(&cache->list); INIT_LIST_HEAD(&cache->cluster_list); INIT_LIST_HEAD(&cache->new_bg_list); diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index 15ce5f2a2b62..ccc264e7bde1 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -158,7 +158,6 @@ struct extent_buffer { * to unlock */ wait_queue_head_t read_lock_wq; - wait_queue_head_t lock_wq; struct page *pages[INLINE_EXTENT_BUFFER_PAGES]; #ifdef CONFIG_BTRFS_DEBUG struct list_head leak_list; diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index 1874aee69c86..225302b39afb 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -75,6 +75,8 @@ void free_extent_map(struct extent_map *em) if (atomic_dec_and_test(&em->refs)) { WARN_ON(extent_map_in_tree(em)); WARN_ON(!list_empty(&em->list)); + if (test_bit(EXTENT_FLAG_FS_MAPPING, &em->flags)) + kfree(em->bdev); kmem_cache_free(extent_map_cache, em); } } diff --git a/fs/btrfs/extent_map.h b/fs/btrfs/extent_map.h index e7fd8a56a140..b2991fd8583e 100644 --- a/fs/btrfs/extent_map.h +++ b/fs/btrfs/extent_map.h @@ -15,6 +15,7 @@ #define EXTENT_FLAG_PREALLOC 3 /* pre-allocated extent */ #define EXTENT_FLAG_LOGGING 4 /* Logging this extent */ #define EXTENT_FLAG_FILLING 5 /* Filling in a preallocated extent */ +#define EXTENT_FLAG_FS_MAPPING 6 /* filesystem extent mapping type */ struct extent_map { struct rb_node rb_node; diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 372b05ff1943..2b0a627cb5f9 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -274,18 +274,32 @@ struct io_ctl { }; static int io_ctl_init(struct io_ctl *io_ctl, struct inode *inode, - struct btrfs_root *root) + struct btrfs_root *root, int write) { + int num_pages; + int check_crcs = 0; + + num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> + PAGE_CACHE_SHIFT; + + if (btrfs_ino(inode) != BTRFS_FREE_INO_OBJECTID) + check_crcs = 1; + + /* Make sure we can fit our crcs into the first page */ + if (write && check_crcs && + (num_pages * sizeof(u32)) >= PAGE_CACHE_SIZE) + return -ENOSPC; + memset(io_ctl, 0, sizeof(struct io_ctl)); - io_ctl->num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> - PAGE_CACHE_SHIFT; - io_ctl->pages = kzalloc(sizeof(struct page *) * io_ctl->num_pages, - GFP_NOFS); + + io_ctl->pages = kzalloc(sizeof(struct page *) * num_pages, GFP_NOFS); if (!io_ctl->pages) return -ENOMEM; + + io_ctl->num_pages = num_pages; io_ctl->root = root; - if (btrfs_ino(inode) != BTRFS_FREE_INO_OBJECTID) - io_ctl->check_crcs = 1; + io_ctl->check_crcs = check_crcs; + return 0; } @@ -666,6 +680,13 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, generation = btrfs_free_space_generation(leaf, header); btrfs_release_path(path); + if (!BTRFS_I(inode)->generation) { + btrfs_info(root->fs_info, + "The free space cache file (%llu) is invalid. skip it\n", + offset); + return 0; + } + if (BTRFS_I(inode)->generation != generation) { btrfs_err(root->fs_info, "free space inode generation (%llu) " @@ -677,7 +698,7 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, if (!num_entries) return 0; - ret = io_ctl_init(&io_ctl, inode, root); + ret = io_ctl_init(&io_ctl, inode, root, 0); if (ret) return ret; @@ -957,19 +978,18 @@ fail: } static noinline_for_stack int -add_ioctl_entries(struct btrfs_root *root, - struct inode *inode, - struct btrfs_block_group_cache *block_group, - struct io_ctl *io_ctl, - struct extent_state **cached_state, - struct list_head *bitmap_list, - int *entries) +write_pinned_extent_entries(struct btrfs_root *root, + struct btrfs_block_group_cache *block_group, + struct io_ctl *io_ctl, + int *entries) { u64 start, extent_start, extent_end, len; - struct list_head *pos, *n; struct extent_io_tree *unpin = NULL; int ret; + if (!block_group) + return 0; + /* * We want to add any pinned extents to our free space cache * so we don't leak the space @@ -979,23 +999,19 @@ add_ioctl_entries(struct btrfs_root *root, */ unpin = root->fs_info->pinned_extents; - if (block_group) - start = block_group->key.objectid; + start = block_group->key.objectid; - while (block_group && (start < block_group->key.objectid + - block_group->key.offset)) { + while (start < block_group->key.objectid + block_group->key.offset) { ret = find_first_extent_bit(unpin, start, &extent_start, &extent_end, EXTENT_DIRTY, NULL); - if (ret) { - ret = 0; - break; - } + if (ret) + return 0; /* This pinned extent is out of our range */ if (extent_start >= block_group->key.objectid + block_group->key.offset) - break; + return 0; extent_start = max(extent_start, start); extent_end = min(block_group->key.objectid + @@ -1005,11 +1021,20 @@ add_ioctl_entries(struct btrfs_root *root, *entries += 1; ret = io_ctl_add_entry(io_ctl, extent_start, len, NULL); if (ret) - goto out_nospc; + return -ENOSPC; start = extent_end; } + return 0; +} + +static noinline_for_stack int +write_bitmap_entries(struct io_ctl *io_ctl, struct list_head *bitmap_list) +{ + struct list_head *pos, *n; + int ret; + /* Write out the bitmaps */ list_for_each_safe(pos, n, bitmap_list) { struct btrfs_free_space *entry = @@ -1017,36 +1042,24 @@ add_ioctl_entries(struct btrfs_root *root, ret = io_ctl_add_bitmap(io_ctl, entry->bitmap); if (ret) - goto out_nospc; + return -ENOSPC; list_del_init(&entry->list); } - /* Zero out the rest of the pages just to make sure */ - io_ctl_zero_remaining_pages(io_ctl); - - ret = btrfs_dirty_pages(root, inode, io_ctl->pages, io_ctl->num_pages, - 0, i_size_read(inode), cached_state); - io_ctl_drop_pages(io_ctl); - unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, - i_size_read(inode) - 1, cached_state, GFP_NOFS); + return 0; +} - if (ret) - goto fail; +static int flush_dirty_cache(struct inode *inode) +{ + int ret; ret = btrfs_wait_ordered_range(inode, 0, (u64)-1); - if (ret) { + if (ret) clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, inode->i_size - 1, EXTENT_DIRTY | EXTENT_DELALLOC, 0, 0, NULL, GFP_NOFS); - goto fail; - } - return 0; -fail: - return -1; - -out_nospc: - return -ENOSPC; + return ret; } static void noinline_for_stack @@ -1056,6 +1069,7 @@ cleanup_write_cache_enospc(struct inode *inode, struct list_head *bitmap_list) { struct list_head *pos, *n; + list_for_each_safe(pos, n, bitmap_list) { struct btrfs_free_space *entry = list_entry(pos, struct btrfs_free_space, list); @@ -1088,64 +1102,104 @@ static int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, { struct extent_state *cached_state = NULL; struct io_ctl io_ctl; - struct list_head bitmap_list; + LIST_HEAD(bitmap_list); int entries = 0; int bitmaps = 0; int ret; - int err = -1; - - INIT_LIST_HEAD(&bitmap_list); if (!i_size_read(inode)) return -1; - ret = io_ctl_init(&io_ctl, inode, root); + ret = io_ctl_init(&io_ctl, inode, root, 1); if (ret) return -1; + if (block_group && (block_group->flags & BTRFS_BLOCK_GROUP_DATA)) { + down_write(&block_group->data_rwsem); + spin_lock(&block_group->lock); + if (block_group->delalloc_bytes) { + block_group->disk_cache_state = BTRFS_DC_WRITTEN; + spin_unlock(&block_group->lock); + up_write(&block_group->data_rwsem); + BTRFS_I(inode)->generation = 0; + ret = 0; + goto out; + } + spin_unlock(&block_group->lock); + } + /* Lock all pages first so we can lock the extent safely. */ io_ctl_prepare_pages(&io_ctl, inode, 0); lock_extent_bits(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1, 0, &cached_state); - - /* Make sure we can fit our crcs into the first page */ - if (io_ctl.check_crcs && - (io_ctl.num_pages * sizeof(u32)) >= PAGE_CACHE_SIZE) - goto out_nospc; - io_ctl_set_generation(&io_ctl, trans->transid); + /* Write out the extent entries in the free space cache */ ret = write_cache_extent_entries(&io_ctl, ctl, block_group, &entries, &bitmaps, &bitmap_list); if (ret) goto out_nospc; - ret = add_ioctl_entries(root, inode, block_group, &io_ctl, - &cached_state, &bitmap_list, &entries); + /* + * Some spaces that are freed in the current transaction are pinned, + * they will be added into free space cache after the transaction is + * committed, we shouldn't lose them. + */ + ret = write_pinned_extent_entries(root, block_group, &io_ctl, &entries); + if (ret) + goto out_nospc; - if (ret == -ENOSPC) + /* At last, we write out all the bitmaps. */ + ret = write_bitmap_entries(&io_ctl, &bitmap_list); + if (ret) goto out_nospc; - else if (ret) + + /* Zero out the rest of the pages just to make sure */ + io_ctl_zero_remaining_pages(&io_ctl); + + /* Everything is written out, now we dirty the pages in the file. */ + ret = btrfs_dirty_pages(root, inode, io_ctl.pages, io_ctl.num_pages, + 0, i_size_read(inode), &cached_state); + if (ret) + goto out_nospc; + + if (block_group && (block_group->flags & BTRFS_BLOCK_GROUP_DATA)) + up_write(&block_group->data_rwsem); + /* + * Release the pages and unlock the extent, we will flush + * them out later + */ + io_ctl_drop_pages(&io_ctl); + + unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, + i_size_read(inode) - 1, &cached_state, GFP_NOFS); + + /* Flush the dirty pages in the cache file. */ + ret = flush_dirty_cache(inode); + if (ret) goto out; - err = update_cache_item(trans, root, inode, path, offset, + /* Update the cache item to tell everyone this cache file is valid. */ + ret = update_cache_item(trans, root, inode, path, offset, entries, bitmaps); - out: io_ctl_free(&io_ctl); - if (err) { + if (ret) { invalidate_inode_pages2(inode->i_mapping); BTRFS_I(inode)->generation = 0; } btrfs_update_inode(trans, root, inode); - return err; + return ret; out_nospc: - cleanup_write_cache_enospc(inode, &io_ctl, &cached_state, &bitmap_list); + + if (block_group && (block_group->flags & BTRFS_BLOCK_GROUP_DATA)) + up_write(&block_group->data_rwsem); + goto out; } @@ -1165,6 +1219,12 @@ int btrfs_write_out_cache(struct btrfs_root *root, spin_unlock(&block_group->lock); return 0; } + + if (block_group->delalloc_bytes) { + block_group->disk_cache_state = BTRFS_DC_WRITTEN; + spin_unlock(&block_group->lock); + return 0; + } spin_unlock(&block_group->lock); inode = lookup_free_space_inode(root, block_group, path); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 8925f66a1411..3668048e16f8 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -693,7 +693,7 @@ retry: ret = btrfs_reserve_extent(root, async_extent->compressed_size, async_extent->compressed_size, - 0, alloc_hint, &ins, 1); + 0, alloc_hint, &ins, 1, 1); if (ret) { int i; @@ -794,7 +794,7 @@ retry: out: return ret; out_free_reserve: - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); out_free: extent_clear_unlock_delalloc(inode, async_extent->start, async_extent->start + @@ -917,7 +917,7 @@ static noinline int cow_file_range(struct inode *inode, cur_alloc_size = disk_num_bytes; ret = btrfs_reserve_extent(root, cur_alloc_size, root->sectorsize, 0, alloc_hint, - &ins, 1); + &ins, 1, 1); if (ret < 0) goto out_unlock; @@ -995,7 +995,7 @@ out: return ret; out_reserve: - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); out_unlock: extent_clear_unlock_delalloc(inode, start, end, locked_page, EXTENT_LOCKED | EXTENT_DO_ACCOUNTING | @@ -2599,6 +2599,21 @@ out_kfree: return NULL; } +static void btrfs_release_delalloc_bytes(struct btrfs_root *root, + u64 start, u64 len) +{ + struct btrfs_block_group_cache *cache; + + cache = btrfs_lookup_block_group(root->fs_info, start); + ASSERT(cache); + + spin_lock(&cache->lock); + cache->delalloc_bytes -= len; + spin_unlock(&cache->lock); + + btrfs_put_block_group(cache); +} + /* as ordered data IO finishes, this gets called so we can finish * an ordered extent if the range of bytes in the file it covers are * fully written. @@ -2698,6 +2713,10 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent) logical_len, logical_len, compress_type, 0, 0, BTRFS_FILE_EXTENT_REG); + if (!ret) + btrfs_release_delalloc_bytes(root, + ordered_extent->start, + ordered_extent->disk_len); } unpin_extent_cache(&BTRFS_I(inode)->extent_tree, ordered_extent->file_offset, ordered_extent->len, @@ -2750,7 +2769,7 @@ out: !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); + ordered_extent->disk_len, 1); } @@ -6535,21 +6554,21 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode, alloc_hint = get_extent_allocation_hint(inode, start, len); ret = btrfs_reserve_extent(root, len, root->sectorsize, 0, - alloc_hint, &ins, 1); + alloc_hint, &ins, 1, 1); if (ret) return ERR_PTR(ret); em = create_pinned_em(inode, start, ins.offset, start, ins.objectid, ins.offset, ins.offset, ins.offset, 0); if (IS_ERR(em)) { - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); return em; } ret = btrfs_add_ordered_extent_dio(inode, start, ins.objectid, ins.offset, ins.offset, 0); if (ret) { - btrfs_free_reserved_extent(root, ins.objectid, ins.offset); + btrfs_free_reserved_extent(root, ins.objectid, ins.offset, 1); free_extent_map(em); return ERR_PTR(ret); } @@ -7437,7 +7456,7 @@ free_ordered: if (!test_bit(BTRFS_ORDERED_PREALLOC, &ordered->flags) && !test_bit(BTRFS_ORDERED_NOCOW, &ordered->flags)) btrfs_free_reserved_extent(root, ordered->start, - ordered->disk_len); + ordered->disk_len, 1); btrfs_put_ordered_extent(ordered); btrfs_put_ordered_extent(ordered); } @@ -8808,7 +8827,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode, cur_bytes = min(num_bytes, 256ULL * 1024 * 1024); cur_bytes = max(cur_bytes, min_size); ret = btrfs_reserve_extent(root, cur_bytes, min_size, 0, - *alloc_hint, &ins, 1); + *alloc_hint, &ins, 1, 0); if (ret) { if (own_trans) btrfs_end_transaction(trans, root); @@ -8822,7 +8841,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode, BTRFS_FILE_EXTENT_PREALLOC); if (ret) { btrfs_free_reserved_extent(root, ins.objectid, - ins.offset); + ins.offset, 0); btrfs_abort_transaction(trans, root, ret); if (own_trans) btrfs_end_transaction(trans, root); diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c index 01277b8f2373..5665d2149249 100644 --- a/fs/btrfs/locking.c +++ b/fs/btrfs/locking.c @@ -33,14 +33,14 @@ static void btrfs_assert_tree_read_locked(struct extent_buffer *eb); */ void btrfs_set_lock_blocking_rw(struct extent_buffer *eb, int rw) { - if (eb->lock_nested) { - read_lock(&eb->lock); - if (eb->lock_nested && current->pid == eb->lock_owner) { - read_unlock(&eb->lock); - return; - } - read_unlock(&eb->lock); - } + /* + * no lock is required. The lock owner may change if + * we have a read lock, but it won't change to or away + * from us. If we have the write lock, we are the owner + * and it'll never change. + */ + if (eb->lock_nested && current->pid == eb->lock_owner) + return; if (rw == BTRFS_WRITE_LOCK) { if (atomic_read(&eb->blocking_writers) == 0) { WARN_ON(atomic_read(&eb->spinning_writers) != 1); @@ -65,14 +65,15 @@ void btrfs_set_lock_blocking_rw(struct extent_buffer *eb, int rw) */ void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw) { - if (eb->lock_nested) { - read_lock(&eb->lock); - if (eb->lock_nested && current->pid == eb->lock_owner) { - read_unlock(&eb->lock); - return; - } - read_unlock(&eb->lock); - } + /* + * no lock is required. The lock owner may change if + * we have a read lock, but it won't change to or away + * from us. If we have the write lock, we are the owner + * and it'll never change. + */ + if (eb->lock_nested && current->pid == eb->lock_owner) + return; + if (rw == BTRFS_WRITE_LOCK_BLOCKING) { BUG_ON(atomic_read(&eb->blocking_writers) != 1); write_lock(&eb->lock); @@ -99,6 +100,9 @@ void btrfs_clear_lock_blocking_rw(struct extent_buffer *eb, int rw) void btrfs_tree_read_lock(struct extent_buffer *eb) { again: + BUG_ON(!atomic_read(&eb->blocking_writers) && + current->pid == eb->lock_owner); + read_lock(&eb->lock); if (atomic_read(&eb->blocking_writers) && current->pid == eb->lock_owner) { @@ -132,7 +136,9 @@ int btrfs_try_tree_read_lock(struct extent_buffer *eb) if (atomic_read(&eb->blocking_writers)) return 0; - read_lock(&eb->lock); + if (!read_trylock(&eb->lock)) + return 0; + if (atomic_read(&eb->blocking_writers)) { read_unlock(&eb->lock); return 0; @@ -151,7 +157,10 @@ int btrfs_try_tree_write_lock(struct extent_buffer *eb) if (atomic_read(&eb->blocking_writers) || atomic_read(&eb->blocking_readers)) return 0; - write_lock(&eb->lock); + + if (!write_trylock(&eb->lock)) + return 0; + if (atomic_read(&eb->blocking_writers) || atomic_read(&eb->blocking_readers)) { write_unlock(&eb->lock); @@ -168,14 +177,15 @@ int btrfs_try_tree_write_lock(struct extent_buffer *eb) */ void btrfs_tree_read_unlock(struct extent_buffer *eb) { - if (eb->lock_nested) { - read_lock(&eb->lock); - if (eb->lock_nested && current->pid == eb->lock_owner) { - eb->lock_nested = 0; - read_unlock(&eb->lock); - return; - } - read_unlock(&eb->lock); + /* + * if we're nested, we have the write lock. No new locking + * is needed as long as we are the lock owner. + * The write unlock will do a barrier for us, and the lock_nested + * field only matters to the lock owner. + */ + if (eb->lock_nested && current->pid == eb->lock_owner) { + eb->lock_nested = 0; + return; } btrfs_assert_tree_read_locked(eb); WARN_ON(atomic_read(&eb->spinning_readers) == 0); @@ -189,14 +199,15 @@ void btrfs_tree_read_unlock(struct extent_buffer *eb) */ void btrfs_tree_read_unlock_blocking(struct extent_buffer *eb) { - if (eb->lock_nested) { - read_lock(&eb->lock); - if (eb->lock_nested && current->pid == eb->lock_owner) { - eb->lock_nested = 0; - read_unlock(&eb->lock); - return; - } - read_unlock(&eb->lock); + /* + * if we're nested, we have the write lock. No new locking + * is needed as long as we are the lock owner. + * The write unlock will do a barrier for us, and the lock_nested + * field only matters to the lock owner. + */ + if (eb->lock_nested && current->pid == eb->lock_owner) { + eb->lock_nested = 0; + return; } btrfs_assert_tree_read_locked(eb); WARN_ON(atomic_read(&eb->blocking_readers) == 0); @@ -244,6 +255,7 @@ void btrfs_tree_unlock(struct extent_buffer *eb) BUG_ON(blockers > 1); btrfs_assert_tree_locked(eb); + eb->lock_owner = 0; atomic_dec(&eb->write_locks); if (blockers) { diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index ac80188eec88..b6d198f5181e 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -2725,11 +2725,8 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent); length = btrfs_dev_extent_length(l, dev_extent); - if (found_key.offset + length <= start) { - key.offset = found_key.offset + length; - btrfs_release_path(path); - continue; - } + if (found_key.offset + length <= start) + goto skip; chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent); chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent); @@ -2740,10 +2737,12 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, * the chunk from going away while we scrub it */ cache = btrfs_lookup_block_group(fs_info, chunk_offset); - if (!cache) { - ret = -ENOENT; - break; - } + + /* some chunks are removed but not committed to disk yet, + * continue scrubbing */ + if (!cache) + goto skip; + dev_replace->cursor_right = found_key.offset + length; dev_replace->cursor_left = found_key.offset; dev_replace->item_needs_writeback = 1; @@ -2802,7 +2801,7 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx, dev_replace->cursor_left = dev_replace->cursor_right; dev_replace->item_needs_writeback = 1; - +skip: key.offset = found_key.offset + length; btrfs_release_path(path); } diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index ffeed6d6326f..c83b24251e53 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2543,9 +2543,6 @@ static int btrfs_relocate_chunk(struct btrfs_root *root, remove_extent_mapping(em_tree, em); write_unlock(&em_tree->lock); - kfree(map); - em->bdev = NULL; - /* once for the tree */ free_extent_map(em); /* once for us */ @@ -4301,9 +4298,11 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, em = alloc_extent_map(); if (!em) { + kfree(map); ret = -ENOMEM; goto error; } + set_bit(EXTENT_FLAG_FS_MAPPING, &em->flags); em->bdev = (struct block_device *)map; em->start = start; em->len = num_bytes; @@ -4346,7 +4345,6 @@ error_del_extent: /* One for the tree reference */ free_extent_map(em); error: - kfree(map); kfree(devices_info); return ret; } @@ -4558,7 +4556,6 @@ void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree) write_unlock(&tree->map_tree.lock); if (!em) break; - kfree(em->bdev); /* once for us */ free_extent_map(em); /* once for the tree */ @@ -5362,6 +5359,15 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, return 0; } +static inline void btrfs_end_bbio(struct btrfs_bio *bbio, struct bio *bio, int err) +{ + if (likely(bbio->flags & BTRFS_BIO_ORIG_BIO_SUBMITTED)) + bio_endio_nodec(bio, err); + else + bio_endio(bio, err); + kfree(bbio); +} + static void btrfs_end_bio(struct bio *bio, int err) { struct btrfs_bio *bbio = bio->bi_private; @@ -5402,12 +5408,6 @@ static void btrfs_end_bio(struct bio *bio, int err) bio = bbio->orig_bio; } - /* - * We have original bio now. So increment bi_remaining to - * account for it in endio - */ - atomic_inc(&bio->bi_remaining); - bio->bi_private = bbio->private; bio->bi_end_io = bbio->end_io; btrfs_io_bio(bio)->mirror_num = bbio->mirror_num; @@ -5424,9 +5424,8 @@ static void btrfs_end_bio(struct bio *bio, int err) set_bit(BIO_UPTODATE, &bio->bi_flags); err = 0; } - kfree(bbio); - bio_endio(bio, err); + btrfs_end_bbio(bbio, bio, err); } else if (!is_orig_bio) { bio_put(bio); } @@ -5589,12 +5588,15 @@ static void bbio_error(struct btrfs_bio *bbio, struct bio *bio, u64 logical) { atomic_inc(&bbio->error); if (atomic_dec_and_test(&bbio->stripes_pending)) { + /* Shoud be the original bio. */ + WARN_ON(bio != bbio->orig_bio); + bio->bi_private = bbio->private; bio->bi_end_io = bbio->end_io; btrfs_io_bio(bio)->mirror_num = bbio->mirror_num; bio->bi_iter.bi_sector = logical >> 9; - kfree(bbio); - bio_endio(bio, -EIO); + + btrfs_end_bbio(bbio, bio, -EIO); } } @@ -5681,6 +5683,7 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, BUG_ON(!bio); /* -ENOMEM */ } else { bio = first_bio; + bbio->flags |= BTRFS_BIO_ORIG_BIO_SUBMITTED; } submit_stripe_bio(root, bbio, bio, @@ -5822,6 +5825,7 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key, return -ENOMEM; } + set_bit(EXTENT_FLAG_FS_MAPPING, &em->flags); em->bdev = (struct block_device *)map; em->start = logical; em->len = length; @@ -5846,7 +5850,6 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key, map->stripes[i].dev = btrfs_find_device(root->fs_info, devid, uuid, NULL); if (!map->stripes[i].dev && !btrfs_test_opt(root, DEGRADED)) { - kfree(map); free_extent_map(em); return -EIO; } @@ -5854,7 +5857,6 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key, map->stripes[i].dev = add_missing_dev(root, devid, uuid); if (!map->stripes[i].dev) { - kfree(map); free_extent_map(em); return -EIO; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 1a15bbeb65e2..2aaa00c47816 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -190,11 +190,14 @@ struct btrfs_bio_stripe { struct btrfs_bio; typedef void (btrfs_bio_end_io_t) (struct btrfs_bio *bio, int err); +#define BTRFS_BIO_ORIG_BIO_SUBMITTED 0x1 + struct btrfs_bio { atomic_t stripes_pending; struct btrfs_fs_info *fs_info; bio_end_io_t *end_io; struct bio *orig_bio; + unsigned long flags; void *private; atomic_t error; int max_errors; diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index 0227b45ef00a..15e9505aa35f 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -290,7 +290,8 @@ int cifsConvertToUTF16(__le16 *target, const char *source, int srclen, const struct nls_table *cp, int mapChars) { - int i, j, charlen; + int i, charlen; + int j = 0; char src_char; __le16 dst_char; wchar_t tmp; @@ -298,12 +299,11 @@ cifsConvertToUTF16(__le16 *target, const char *source, int srclen, if (!mapChars) return cifs_strtoUTF16(target, source, PATH_MAX, cp); - for (i = 0, j = 0; i < srclen; j++) { + for (i = 0; i < srclen; j++) { src_char = source[i]; charlen = 1; switch (src_char) { case 0: - put_unaligned(0, &target[j]); goto ctoUTF16_out; case ':': dst_char = cpu_to_le16(UNI_COLON); @@ -350,6 +350,7 @@ cifsConvertToUTF16(__le16 *target, const char *source, int srclen, } ctoUTF16_out: + put_unaligned(0, &target[j]); /* Null terminate target unicode string */ return j; } diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 2c90d07c0b3a..888398067420 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -725,6 +725,19 @@ out_nls: goto out; } +static ssize_t +cifs_loose_read_iter(struct kiocb *iocb, struct iov_iter *iter) +{ + ssize_t rc; + struct inode *inode = file_inode(iocb->ki_filp); + + rc = cifs_revalidate_mapping(inode); + if (rc) + return rc; + + return generic_file_read_iter(iocb, iter); +} + static ssize_t cifs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct inode *inode = file_inode(iocb->ki_filp); @@ -881,7 +894,7 @@ const struct inode_operations cifs_symlink_inode_ops = { const struct file_operations cifs_file_ops = { .read = new_sync_read, .write = new_sync_write, - .read_iter = generic_file_read_iter, + .read_iter = cifs_loose_read_iter, .write_iter = cifs_file_write_iter, .open = cifs_open, .release = cifs_close, @@ -939,7 +952,7 @@ const struct file_operations cifs_file_direct_ops = { const struct file_operations cifs_file_nobrl_ops = { .read = new_sync_read, .write = new_sync_write, - .read_iter = generic_file_read_iter, + .read_iter = cifs_loose_read_iter, .write_iter = cifs_file_write_iter, .open = cifs_open, .release = cifs_close, diff --git a/fs/cifs/link.c b/fs/cifs/link.c index 264ece71bdb2..68559fd557fb 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -374,7 +374,7 @@ cifs_create_mf_symlink(unsigned int xid, struct cifs_tcon *tcon, oparms.cifs_sb = cifs_sb; oparms.desired_access = GENERIC_WRITE; oparms.create_options = create_options; - oparms.disposition = FILE_OPEN; + oparms.disposition = FILE_CREATE; oparms.path = path; oparms.fid = &fid; oparms.reconnect = false; diff --git a/fs/eventpoll.c b/fs/eventpoll.c index b73e0621ce9e..b10b48c2a7af 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -910,7 +910,7 @@ static const struct file_operations eventpoll_fops = { void eventpoll_release_file(struct file *file) { struct eventpoll *ep; - struct epitem *epi; + struct epitem *epi, *next; /* * We don't want to get "file->f_lock" because it is not @@ -926,7 +926,7 @@ void eventpoll_release_file(struct file *file) * Besides, ep_remove() acquires the lock, so we can't hold it here. */ mutex_lock(&epmutex); - list_for_each_entry_rcu(epi, &file->f_ep_links, fllink) { + list_for_each_entry_safe(epi, next, &file->f_ep_links, fllink) { ep = epi->ep; mutex_lock_nested(&ep->mtx, 0); ep_remove(ep, epi); diff --git a/fs/locks.c b/fs/locks.c index da57c9b7e844..717fbc404e6b 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -431,7 +431,7 @@ static int lease_init(struct file *filp, long type, struct file_lock *fl) if (assign_type(fl, type) != 0) return -EINVAL; - fl->fl_owner = (fl_owner_t)filp; + fl->fl_owner = (fl_owner_t)current->files; fl->fl_pid = current->tgid; fl->fl_file = filp; diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index c496f8a74639..9927913c97c2 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -147,6 +147,17 @@ int nfs_sync_mapping(struct address_space *mapping) return ret; } +static void nfs_set_cache_invalid(struct inode *inode, unsigned long flags) +{ + struct nfs_inode *nfsi = NFS_I(inode); + + if (inode->i_mapping->nrpages == 0) + flags &= ~NFS_INO_INVALID_DATA; + nfsi->cache_validity |= flags; + if (flags & NFS_INO_INVALID_DATA) + nfs_fscache_invalidate(inode); +} + /* * Invalidate the local caches */ @@ -162,17 +173,16 @@ static void nfs_zap_caches_locked(struct inode *inode) memset(NFS_I(inode)->cookieverf, 0, sizeof(NFS_I(inode)->cookieverf)); if (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)) { - nfs_fscache_invalidate(inode); - nfsi->cache_validity |= NFS_INO_INVALID_ATTR + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR | NFS_INO_INVALID_DATA | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL - | NFS_INO_REVAL_PAGECACHE; + | NFS_INO_REVAL_PAGECACHE); } else - nfsi->cache_validity |= NFS_INO_INVALID_ATTR + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL - | NFS_INO_REVAL_PAGECACHE; + | NFS_INO_REVAL_PAGECACHE); nfs_zap_label_cache_locked(nfsi); } @@ -187,8 +197,7 @@ void nfs_zap_mapping(struct inode *inode, struct address_space *mapping) { if (mapping->nrpages != 0) { spin_lock(&inode->i_lock); - NFS_I(inode)->cache_validity |= NFS_INO_INVALID_DATA; - nfs_fscache_invalidate(inode); + nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA); spin_unlock(&inode->i_lock); } } @@ -209,7 +218,7 @@ EXPORT_SYMBOL_GPL(nfs_zap_acl_cache); void nfs_invalidate_atime(struct inode *inode) { spin_lock(&inode->i_lock); - NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATIME; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATIME); spin_unlock(&inode->i_lock); } EXPORT_SYMBOL_GPL(nfs_invalidate_atime); @@ -369,7 +378,7 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr, st inode->i_mode = fattr->mode; if ((fattr->valid & NFS_ATTR_FATTR_MODE) == 0 && nfs_server_capable(inode, NFS_CAP_MODE)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); /* Why so? Because we want revalidate for devices/FIFOs, and * that's precisely what we have in nfs_file_inode_operations. */ @@ -415,36 +424,36 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr, st if (fattr->valid & NFS_ATTR_FATTR_ATIME) inode->i_atime = fattr->atime; else if (nfs_server_capable(inode, NFS_CAP_ATIME)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); if (fattr->valid & NFS_ATTR_FATTR_MTIME) inode->i_mtime = fattr->mtime; else if (nfs_server_capable(inode, NFS_CAP_MTIME)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); if (fattr->valid & NFS_ATTR_FATTR_CTIME) inode->i_ctime = fattr->ctime; else if (nfs_server_capable(inode, NFS_CAP_CTIME)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); if (fattr->valid & NFS_ATTR_FATTR_CHANGE) inode->i_version = fattr->change_attr; else if (nfs_server_capable(inode, NFS_CAP_CHANGE_ATTR)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); if (fattr->valid & NFS_ATTR_FATTR_SIZE) inode->i_size = nfs_size_to_loff_t(fattr->size); else - nfsi->cache_validity |= NFS_INO_INVALID_ATTR - | NFS_INO_REVAL_PAGECACHE; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR + | NFS_INO_REVAL_PAGECACHE); if (fattr->valid & NFS_ATTR_FATTR_NLINK) set_nlink(inode, fattr->nlink); else if (nfs_server_capable(inode, NFS_CAP_NLINK)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); if (fattr->valid & NFS_ATTR_FATTR_OWNER) inode->i_uid = fattr->uid; else if (nfs_server_capable(inode, NFS_CAP_OWNER)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); if (fattr->valid & NFS_ATTR_FATTR_GROUP) inode->i_gid = fattr->gid; else if (nfs_server_capable(inode, NFS_CAP_OWNER_GROUP)) - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ATTR); if (fattr->valid & NFS_ATTR_FATTR_BLOCKS_USED) inode->i_blocks = fattr->du.nfs2.blocks; if (fattr->valid & NFS_ATTR_FATTR_SPACE_USED) { @@ -550,6 +559,9 @@ static int nfs_vmtruncate(struct inode * inode, loff_t offset) spin_lock(&inode->i_lock); i_size_write(inode, offset); + /* Optimisation */ + if (offset == 0) + NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_DATA; spin_unlock(&inode->i_lock); truncate_pagecache(inode, offset); @@ -578,7 +590,8 @@ void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr) inode->i_uid = attr->ia_uid; if ((attr->ia_valid & ATTR_GID) != 0) inode->i_gid = attr->ia_gid; - NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ACCESS|NFS_INO_INVALID_ACL; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_ACCESS + | NFS_INO_INVALID_ACL); spin_unlock(&inode->i_lock); } if ((attr->ia_valid & ATTR_SIZE) != 0) { @@ -1101,7 +1114,7 @@ static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr && inode->i_version == fattr->pre_change_attr) { inode->i_version = fattr->change_attr; if (S_ISDIR(inode->i_mode)) - nfsi->cache_validity |= NFS_INO_INVALID_DATA; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA); ret |= NFS_INO_INVALID_ATTR; } /* If we have atomic WCC data, we may update some attributes */ @@ -1117,7 +1130,7 @@ static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr && timespec_equal(&inode->i_mtime, &fattr->pre_mtime)) { memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime)); if (S_ISDIR(inode->i_mode)) - nfsi->cache_validity |= NFS_INO_INVALID_DATA; + nfs_set_cache_invalid(inode, NFS_INO_INVALID_DATA); ret |= NFS_INO_INVALID_ATTR; } if ((fattr->valid & NFS_ATTR_FATTR_PRESIZE) @@ -1128,9 +1141,6 @@ static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr ret |= NFS_INO_INVALID_ATTR; } - if (nfsi->cache_validity & NFS_INO_INVALID_DATA) - nfs_fscache_invalidate(inode); - return ret; } @@ -1189,7 +1199,7 @@ static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fat invalid |= NFS_INO_INVALID_ATIME; if (invalid != 0) - nfsi->cache_validity |= invalid; + nfs_set_cache_invalid(inode, invalid); nfsi->read_cache_jiffies = fattr->time_start; return 0; @@ -1402,13 +1412,11 @@ EXPORT_SYMBOL_GPL(nfs_refresh_inode); static int nfs_post_op_update_inode_locked(struct inode *inode, struct nfs_fattr *fattr) { - struct nfs_inode *nfsi = NFS_I(inode); + unsigned long invalid = NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; - nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_REVAL_PAGECACHE; - if (S_ISDIR(inode->i_mode)) { - nfsi->cache_validity |= NFS_INO_INVALID_DATA; - nfs_fscache_invalidate(inode); - } + if (S_ISDIR(inode->i_mode)) + invalid |= NFS_INO_INVALID_DATA; + nfs_set_cache_invalid(inode, invalid); if ((fattr->valid & NFS_ATTR_FATTR) == 0) return 0; return nfs_refresh_inode_locked(inode, fattr); @@ -1601,6 +1609,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) if ((nfsi->npages == 0) || new_isize > cur_isize) { i_size_write(inode, new_isize); invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA; + invalid &= ~NFS_INO_REVAL_PAGECACHE; } dprintk("NFS: isize change on server for file %s/%ld " "(%Ld to %Ld)\n", @@ -1702,10 +1711,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) invalid &= ~NFS_INO_INVALID_DATA; if (!NFS_PROTO(inode)->have_delegation(inode, FMODE_READ) || (save_cache_validity & NFS_INO_REVAL_FORCED)) - nfsi->cache_validity |= invalid; - - if (invalid & NFS_INO_INVALID_DATA) - nfs_fscache_invalidate(inode); + nfs_set_cache_invalid(inode, invalid); return 0; out_err: diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index f63cb87cd730..ba2affa51941 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -230,7 +230,7 @@ int nfs_atomic_open(struct inode *, struct dentry *, struct file *, extern struct file_system_type nfs4_fs_type; /* nfs4namespace.c */ -struct rpc_clnt *nfs4_create_sec_client(struct rpc_clnt *, struct inode *, struct qstr *); +struct rpc_clnt *nfs4_negotiate_security(struct rpc_clnt *, struct inode *, struct qstr *); struct vfsmount *nfs4_submount(struct nfs_server *, struct dentry *, struct nfs_fh *, struct nfs_fattr *); int nfs4_replace_transport(struct nfs_server *server, diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c index 3d5dbf80d46a..3d83cb1fdc70 100644 --- a/fs/nfs/nfs4namespace.c +++ b/fs/nfs/nfs4namespace.c @@ -139,16 +139,22 @@ static size_t nfs_parse_server_name(char *string, size_t len, * @server: NFS server struct * @flavors: List of security tuples returned by SECINFO procedure * - * Return the pseudoflavor of the first security mechanism in - * "flavors" that is locally supported. Return RPC_AUTH_UNIX if - * no matching flavor is found in the array. The "flavors" array + * Return an rpc client that uses the first security mechanism in + * "flavors" that is locally supported. The "flavors" array * is searched in the order returned from the server, per RFC 3530 - * recommendation. + * recommendation and each flavor is checked for membership in the + * sec= mount option list if it exists. + * + * Return -EPERM if no matching flavor is found in the array. + * + * Please call rpc_shutdown_client() when you are done with this rpc client. + * */ -static rpc_authflavor_t nfs_find_best_sec(struct nfs_server *server, +static struct rpc_clnt *nfs_find_best_sec(struct rpc_clnt *clnt, + struct nfs_server *server, struct nfs4_secinfo_flavors *flavors) { - rpc_authflavor_t pseudoflavor; + rpc_authflavor_t pflavor; struct nfs4_secinfo4 *secinfo; unsigned int i; @@ -159,62 +165,73 @@ static rpc_authflavor_t nfs_find_best_sec(struct nfs_server *server, case RPC_AUTH_NULL: case RPC_AUTH_UNIX: case RPC_AUTH_GSS: - pseudoflavor = rpcauth_get_pseudoflavor(secinfo->flavor, + pflavor = rpcauth_get_pseudoflavor(secinfo->flavor, &secinfo->flavor_info); - /* make sure pseudoflavor matches sec= mount opt */ - if (pseudoflavor != RPC_AUTH_MAXFLAVOR && - nfs_auth_info_match(&server->auth_info, - pseudoflavor)) - return pseudoflavor; - break; + /* does the pseudoflavor match a sec= mount opt? */ + if (pflavor != RPC_AUTH_MAXFLAVOR && + nfs_auth_info_match(&server->auth_info, pflavor)) { + struct rpc_clnt *new; + struct rpc_cred *cred; + + /* Cloning creates an rpc_auth for the flavor */ + new = rpc_clone_client_set_auth(clnt, pflavor); + if (IS_ERR(new)) + continue; + /** + * Check that the user actually can use the + * flavor. This is mostly for RPC_AUTH_GSS + * where cr_init obtains a gss context + */ + cred = rpcauth_lookupcred(new->cl_auth, 0); + if (IS_ERR(cred)) { + rpc_shutdown_client(new); + continue; + } + put_rpccred(cred); + return new; + } } } - - /* if there were any sec= options then nothing matched */ - if (server->auth_info.flavor_len > 0) - return -EPERM; - - return RPC_AUTH_UNIX; + return ERR_PTR(-EPERM); } -static rpc_authflavor_t nfs4_negotiate_security(struct inode *inode, struct qstr *name) +/** + * nfs4_negotiate_security - in response to an NFS4ERR_WRONGSEC on lookup, + * return an rpc_clnt that uses the best available security flavor with + * respect to the secinfo flavor list and the sec= mount options. + * + * @clnt: RPC client to clone + * @inode: directory inode + * @name: lookup name + * + * Please call rpc_shutdown_client() when you are done with this rpc client. + */ +struct rpc_clnt * +nfs4_negotiate_security(struct rpc_clnt *clnt, struct inode *inode, + struct qstr *name) { struct page *page; struct nfs4_secinfo_flavors *flavors; - rpc_authflavor_t flavor; + struct rpc_clnt *new; int err; page = alloc_page(GFP_KERNEL); if (!page) - return -ENOMEM; + return ERR_PTR(-ENOMEM); + flavors = page_address(page); err = nfs4_proc_secinfo(inode, name, flavors); if (err < 0) { - flavor = err; + new = ERR_PTR(err); goto out; } - flavor = nfs_find_best_sec(NFS_SERVER(inode), flavors); + new = nfs_find_best_sec(clnt, NFS_SERVER(inode), flavors); out: put_page(page); - return flavor; -} - -/* - * Please call rpc_shutdown_client() when you are done with this client. - */ -struct rpc_clnt *nfs4_create_sec_client(struct rpc_clnt *clnt, struct inode *inode, - struct qstr *name) -{ - rpc_authflavor_t flavor; - - flavor = nfs4_negotiate_security(inode, name); - if ((int)flavor < 0) - return ERR_PTR((int)flavor); - - return rpc_clone_client_set_auth(clnt, flavor); + return new; } static struct vfsmount *try_location(struct nfs_clone_mount *mountdata, @@ -397,11 +414,6 @@ struct vfsmount *nfs4_submount(struct nfs_server *server, struct dentry *dentry, if (client->cl_auth->au_flavor != flavor) flavor = client->cl_auth->au_flavor; - else { - rpc_authflavor_t new = nfs4_negotiate_security(dir, name); - if ((int)new >= 0) - flavor = new; - } mnt = nfs_do_submount(dentry, fh, fattr, flavor); out: rpc_shutdown_client(client); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 285ad5334018..4bf3d97cc5a0 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3247,7 +3247,7 @@ static int nfs4_proc_lookup_common(struct rpc_clnt **clnt, struct inode *dir, err = -EPERM; if (client != *clnt) goto out; - client = nfs4_create_sec_client(client, dir, name); + client = nfs4_negotiate_security(client, dir, name); if (IS_ERR(client)) return PTR_ERR(client); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 3ee5af4e738e..98ff061ccaf3 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -934,12 +934,14 @@ static bool nfs_write_pageuptodate(struct page *page, struct inode *inode) if (nfs_have_delegated_attributes(inode)) goto out; - if (nfsi->cache_validity & (NFS_INO_INVALID_DATA|NFS_INO_REVAL_PAGECACHE)) + if (nfsi->cache_validity & NFS_INO_REVAL_PAGECACHE) return false; smp_rmb(); if (test_bit(NFS_INO_INVALIDATING, &nfsi->flags)) return false; out: + if (nfsi->cache_validity & NFS_INO_INVALID_DATA) + return false; return PageUptodate(page) != 0; } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index c0d45cec9958..2204e1fe5725 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -41,6 +41,7 @@ #include <linux/ratelimit.h> #include <linux/sunrpc/svcauth_gss.h> #include <linux/sunrpc/addr.h> +#include <linux/hash.h> #include "xdr4.h" #include "xdr4cb.h" #include "vfs.h" @@ -364,6 +365,79 @@ static struct nfs4_ol_stateid * nfs4_alloc_stateid(struct nfs4_client *clp) return openlockstateid(nfs4_alloc_stid(clp, stateid_slab)); } +/* + * When we recall a delegation, we should be careful not to hand it + * out again straight away. + * To ensure this we keep a pair of bloom filters ('new' and 'old') + * in which the filehandles of recalled delegations are "stored". + * If a filehandle appear in either filter, a delegation is blocked. + * When a delegation is recalled, the filehandle is stored in the "new" + * filter. + * Every 30 seconds we swap the filters and clear the "new" one, + * unless both are empty of course. + * + * Each filter is 256 bits. We hash the filehandle to 32bit and use the + * low 3 bytes as hash-table indices. + * + * 'state_lock', which is always held when block_delegations() is called, + * is used to manage concurrent access. Testing does not need the lock + * except when swapping the two filters. + */ +static struct bloom_pair { + int entries, old_entries; + time_t swap_time; + int new; /* index into 'set' */ + DECLARE_BITMAP(set[2], 256); +} blocked_delegations; + +static int delegation_blocked(struct knfsd_fh *fh) +{ + u32 hash; + struct bloom_pair *bd = &blocked_delegations; + + if (bd->entries == 0) + return 0; + if (seconds_since_boot() - bd->swap_time > 30) { + spin_lock(&state_lock); + if (seconds_since_boot() - bd->swap_time > 30) { + bd->entries -= bd->old_entries; + bd->old_entries = bd->entries; + memset(bd->set[bd->new], 0, + sizeof(bd->set[0])); + bd->new = 1-bd->new; + bd->swap_time = seconds_since_boot(); + } + spin_unlock(&state_lock); + } + hash = arch_fast_hash(&fh->fh_base, fh->fh_size, 0); + if (test_bit(hash&255, bd->set[0]) && + test_bit((hash>>8)&255, bd->set[0]) && + test_bit((hash>>16)&255, bd->set[0])) + return 1; + + if (test_bit(hash&255, bd->set[1]) && + test_bit((hash>>8)&255, bd->set[1]) && + test_bit((hash>>16)&255, bd->set[1])) + return 1; + + return 0; +} + +static void block_delegations(struct knfsd_fh *fh) +{ + u32 hash; + struct bloom_pair *bd = &blocked_delegations; + + hash = arch_fast_hash(&fh->fh_base, fh->fh_size, 0); + + __set_bit(hash&255, bd->set[bd->new]); + __set_bit((hash>>8)&255, bd->set[bd->new]); + __set_bit((hash>>16)&255, bd->set[bd->new]); + if (bd->entries == 0) + bd->swap_time = seconds_since_boot(); + bd->entries += 1; +} + static struct nfs4_delegation * alloc_init_deleg(struct nfs4_client *clp, struct nfs4_ol_stateid *stp, struct svc_fh *current_fh) { @@ -372,6 +446,8 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_ol_stateid *stp, struct sv dprintk("NFSD alloc_init_deleg\n"); if (num_delegations > max_delegations) return NULL; + if (delegation_blocked(¤t_fh->fh_handle)) + return NULL; dp = delegstateid(nfs4_alloc_stid(clp, deleg_slab)); if (dp == NULL) return dp; @@ -2770,6 +2846,8 @@ static void nfsd_break_one_deleg(struct nfs4_delegation *dp) /* Only place dl_time is set; protected by i_lock: */ dp->dl_time = get_seconds(); + block_delegations(&dp->dl_fh); + nfsd4_cb_recall(dp); } diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 2d305a121f37..83baf2bfe9e9 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -2687,6 +2687,7 @@ nfsd4_encode_dirent(void *ccdv, const char *name, int namlen, nfserr = nfserr_toosmall; goto fail; case nfserr_noent: + xdr_truncate_encode(xdr, start_offset); goto skip_entry; default: /* diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h index a106b3f2b22a..fae17c640df3 100644 --- a/fs/ocfs2/dlm/dlmcommon.h +++ b/fs/ocfs2/dlm/dlmcommon.h @@ -331,6 +331,7 @@ struct dlm_lock_resource u16 state; char lvb[DLM_LVB_LEN]; unsigned int inflight_locks; + unsigned int inflight_assert_workers; unsigned long refmap[BITS_TO_LONGS(O2NM_MAX_NODES)]; }; @@ -910,6 +911,9 @@ void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm, void dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res); +void __dlm_lockres_grab_inflight_worker(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res); + void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock); void dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock); void __dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock); diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 3087a21d32f9..82abf0cc9a12 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -581,6 +581,7 @@ static void dlm_init_lockres(struct dlm_ctxt *dlm, atomic_set(&res->asts_reserved, 0); res->migration_pending = 0; res->inflight_locks = 0; + res->inflight_assert_workers = 0; res->dlm = dlm; @@ -683,6 +684,43 @@ void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm, wake_up(&res->wq); } +void __dlm_lockres_grab_inflight_worker(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res) +{ + assert_spin_locked(&res->spinlock); + res->inflight_assert_workers++; + mlog(0, "%s:%.*s: inflight assert worker++: now %u\n", + dlm->name, res->lockname.len, res->lockname.name, + res->inflight_assert_workers); +} + +static void dlm_lockres_grab_inflight_worker(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res) +{ + spin_lock(&res->spinlock); + __dlm_lockres_grab_inflight_worker(dlm, res); + spin_unlock(&res->spinlock); +} + +static void __dlm_lockres_drop_inflight_worker(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res) +{ + assert_spin_locked(&res->spinlock); + BUG_ON(res->inflight_assert_workers == 0); + res->inflight_assert_workers--; + mlog(0, "%s:%.*s: inflight assert worker--: now %u\n", + dlm->name, res->lockname.len, res->lockname.name, + res->inflight_assert_workers); +} + +static void dlm_lockres_drop_inflight_worker(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res) +{ + spin_lock(&res->spinlock); + __dlm_lockres_drop_inflight_worker(dlm, res); + spin_unlock(&res->spinlock); +} + /* * lookup a lock resource by name. * may already exist in the hashtable. @@ -1603,7 +1641,8 @@ send_response: mlog(ML_ERROR, "failed to dispatch assert master work\n"); response = DLM_MASTER_RESP_ERROR; dlm_lockres_put(res); - } + } else + dlm_lockres_grab_inflight_worker(dlm, res); } else { if (res) dlm_lockres_put(res); @@ -2118,6 +2157,8 @@ static void dlm_assert_master_worker(struct dlm_work_item *item, void *data) dlm_lockres_release_ast(dlm, res); put: + dlm_lockres_drop_inflight_worker(dlm, res); + dlm_lockres_put(res); mlog(0, "finished with dlm_assert_master_worker\n"); @@ -3088,11 +3129,15 @@ static int dlm_add_migration_mle(struct dlm_ctxt *dlm, /* remove it so that only one mle will be found */ __dlm_unlink_mle(dlm, tmp); __dlm_mle_detach_hb_events(dlm, tmp); - ret = DLM_MIGRATE_RESPONSE_MASTERY_REF; - mlog(0, "%s:%.*s: master=%u, newmaster=%u, " - "telling master to get ref for cleared out mle " - "during migration\n", dlm->name, namelen, name, - master, new_master); + if (tmp->type == DLM_MLE_MASTER) { + ret = DLM_MIGRATE_RESPONSE_MASTERY_REF; + mlog(0, "%s:%.*s: master=%u, newmaster=%u, " + "telling master to get ref " + "for cleared out mle during " + "migration\n", dlm->name, + namelen, name, master, + new_master); + } } spin_unlock(&tmp->spinlock); } diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index 5de019437ea5..45067faf5695 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -1708,7 +1708,8 @@ int dlm_master_requery_handler(struct o2net_msg *msg, u32 len, void *data, mlog_errno(-ENOMEM); /* retry!? */ BUG(); - } + } else + __dlm_lockres_grab_inflight_worker(dlm, res); } else /* put.. incase we are not the master */ dlm_lockres_put(res); spin_unlock(&res->spinlock); diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c index 9db869de829d..69aac6f088ad 100644 --- a/fs/ocfs2/dlm/dlmthread.c +++ b/fs/ocfs2/dlm/dlmthread.c @@ -259,12 +259,15 @@ static void dlm_run_purge_list(struct dlm_ctxt *dlm, * refs on it. */ unused = __dlm_lockres_unused(lockres); if (!unused || - (lockres->state & DLM_LOCK_RES_MIGRATING)) { + (lockres->state & DLM_LOCK_RES_MIGRATING) || + (lockres->inflight_assert_workers != 0)) { mlog(0, "%s: res %.*s is in use or being remastered, " - "used %d, state %d\n", dlm->name, - lockres->lockname.len, lockres->lockname.name, - !unused, lockres->state); - list_move_tail(&dlm->purge_list, &lockres->purge); + "used %d, state %d, assert master workers %u\n", + dlm->name, lockres->lockname.len, + lockres->lockname.name, + !unused, lockres->state, + lockres->inflight_assert_workers); + list_move_tail(&lockres->purge, &dlm->purge_list); spin_unlock(&lockres->spinlock); continue; } diff --git a/fs/ocfs2/dlm/dlmunlock.c b/fs/ocfs2/dlm/dlmunlock.c index 5698b52cf5c9..2e3c9dbab68c 100644 --- a/fs/ocfs2/dlm/dlmunlock.c +++ b/fs/ocfs2/dlm/dlmunlock.c @@ -191,7 +191,9 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm, DLM_UNLOCK_CLEAR_CONVERT_TYPE); } else if (status == DLM_RECOVERING || status == DLM_MIGRATING || - status == DLM_FORWARD) { + status == DLM_FORWARD || + status == DLM_NOLOCKMGR + ) { /* must clear the actions because this unlock * is about to be retried. cannot free or do * any list manipulation. */ @@ -200,7 +202,8 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm, res->lockname.name, status==DLM_RECOVERING?"recovering": (status==DLM_MIGRATING?"migrating": - "forward")); + (status == DLM_FORWARD ? "forward" : + "nolockmanager"))); actions = 0; } if (flags & LKM_CANCEL) @@ -364,7 +367,10 @@ static enum dlm_status dlm_send_remote_unlock_request(struct dlm_ctxt *dlm, * updated state to the recovery master. this thread * just needs to finish out the operation and call * the unlockast. */ - ret = DLM_NORMAL; + if (dlm_is_node_dead(dlm, owner)) + ret = DLM_NORMAL; + else + ret = DLM_NOLOCKMGR; } else { /* something bad. this will BUG in ocfs2 */ ret = dlm_err_to_dlm_status(tmpret); @@ -638,7 +644,9 @@ retry: if (status == DLM_RECOVERING || status == DLM_MIGRATING || - status == DLM_FORWARD) { + status == DLM_FORWARD || + status == DLM_NOLOCKMGR) { + /* We want to go away for a tiny bit to allow recovery * / migration to complete on this resource. I don't * know of any wait queue we could sleep on as this @@ -650,7 +658,7 @@ retry: msleep(50); mlog(0, "retrying unlock due to pending recovery/" - "migration/in-progress\n"); + "migration/in-progress/reconnect\n"); goto retry; } diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 2060fc398445..8add6f1030d7 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -205,6 +205,21 @@ static struct inode *ocfs2_get_init_inode(struct inode *dir, umode_t mode) return inode; } +static void ocfs2_cleanup_add_entry_failure(struct ocfs2_super *osb, + struct dentry *dentry, struct inode *inode) +{ + struct ocfs2_dentry_lock *dl = dentry->d_fsdata; + + ocfs2_simple_drop_lockres(osb, &dl->dl_lockres); + ocfs2_lock_res_free(&dl->dl_lockres); + BUG_ON(dl->dl_count != 1); + spin_lock(&dentry_attach_lock); + dentry->d_fsdata = NULL; + spin_unlock(&dentry_attach_lock); + kfree(dl); + iput(inode); +} + static int ocfs2_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, @@ -231,6 +246,7 @@ static int ocfs2_mknod(struct inode *dir, sigset_t oldset; int did_block_signals = 0; struct posix_acl *default_acl = NULL, *acl = NULL; + struct ocfs2_dentry_lock *dl = NULL; trace_ocfs2_mknod(dir, dentry, dentry->d_name.len, dentry->d_name.name, (unsigned long long)OCFS2_I(dir)->ip_blkno, @@ -423,6 +439,8 @@ static int ocfs2_mknod(struct inode *dir, goto leave; } + dl = dentry->d_fsdata; + status = ocfs2_add_entry(handle, dentry, inode, OCFS2_I(inode)->ip_blkno, parent_fe_bh, &lookup); @@ -469,6 +487,9 @@ leave: * ocfs2_delete_inode will mutex_lock again. */ if ((status < 0) && inode) { + if (dl) + ocfs2_cleanup_add_entry_failure(osb, dentry, inode); + OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SKIP_ORPHAN_DIR; clear_nlink(inode); iput(inode); @@ -991,6 +1012,65 @@ leave: return status; } +static int ocfs2_check_if_ancestor(struct ocfs2_super *osb, + u64 src_inode_no, u64 dest_inode_no) +{ + int ret = 0, i = 0; + u64 parent_inode_no = 0; + u64 child_inode_no = src_inode_no; + struct inode *child_inode; + +#define MAX_LOOKUP_TIMES 32 + while (1) { + child_inode = ocfs2_iget(osb, child_inode_no, 0, 0); + if (IS_ERR(child_inode)) { + ret = PTR_ERR(child_inode); + break; + } + + ret = ocfs2_inode_lock(child_inode, NULL, 0); + if (ret < 0) { + iput(child_inode); + if (ret != -ENOENT) + mlog_errno(ret); + break; + } + + ret = ocfs2_lookup_ino_from_name(child_inode, "..", 2, + &parent_inode_no); + ocfs2_inode_unlock(child_inode, 0); + iput(child_inode); + if (ret < 0) { + ret = -ENOENT; + break; + } + + if (parent_inode_no == dest_inode_no) { + ret = 1; + break; + } + + if (parent_inode_no == osb->root_inode->i_ino) { + ret = 0; + break; + } + + child_inode_no = parent_inode_no; + + if (++i >= MAX_LOOKUP_TIMES) { + mlog(ML_NOTICE, "max lookup times reached, filesystem " + "may have nested directories, " + "src inode: %llu, dest inode: %llu.\n", + (unsigned long long)src_inode_no, + (unsigned long long)dest_inode_no); + ret = 0; + break; + } + } + + return ret; +} + /* * The only place this should be used is rename! * if they have the same id, then the 1st one is the only one locked. @@ -1002,6 +1082,7 @@ static int ocfs2_double_lock(struct ocfs2_super *osb, struct inode *inode2) { int status; + int inode1_is_ancestor, inode2_is_ancestor; struct ocfs2_inode_info *oi1 = OCFS2_I(inode1); struct ocfs2_inode_info *oi2 = OCFS2_I(inode2); struct buffer_head **tmpbh; @@ -1015,9 +1096,26 @@ static int ocfs2_double_lock(struct ocfs2_super *osb, if (*bh2) *bh2 = NULL; - /* we always want to lock the one with the lower lockid first. */ + /* we always want to lock the one with the lower lockid first. + * and if they are nested, we lock ancestor first */ if (oi1->ip_blkno != oi2->ip_blkno) { - if (oi1->ip_blkno < oi2->ip_blkno) { + inode1_is_ancestor = ocfs2_check_if_ancestor(osb, oi2->ip_blkno, + oi1->ip_blkno); + if (inode1_is_ancestor < 0) { + status = inode1_is_ancestor; + goto bail; + } + + inode2_is_ancestor = ocfs2_check_if_ancestor(osb, oi1->ip_blkno, + oi2->ip_blkno); + if (inode2_is_ancestor < 0) { + status = inode2_is_ancestor; + goto bail; + } + + if ((inode1_is_ancestor == 1) || + (oi1->ip_blkno < oi2->ip_blkno && + inode2_is_ancestor == 0)) { /* switch id1 and id2 around */ tmpbh = bh2; bh2 = bh1; @@ -1098,6 +1196,7 @@ static int ocfs2_rename(struct inode *old_dir, struct ocfs2_dir_lookup_result old_entry_lookup = { NULL, }; struct ocfs2_dir_lookup_result orphan_insert = { NULL, }; struct ocfs2_dir_lookup_result target_insert = { NULL, }; + bool should_add_orphan = false; /* At some point it might be nice to break this function up a * bit. */ @@ -1134,6 +1233,21 @@ static int ocfs2_rename(struct inode *old_dir, goto bail; } rename_lock = 1; + + /* here we cannot guarantee the inodes haven't just been + * changed, so check if they are nested again */ + status = ocfs2_check_if_ancestor(osb, new_dir->i_ino, + old_inode->i_ino); + if (status < 0) { + mlog_errno(status); + goto bail; + } else if (status == 1) { + status = -EPERM; + trace_ocfs2_rename_not_permitted( + (unsigned long long)old_inode->i_ino, + (unsigned long long)new_dir->i_ino); + goto bail; + } } /* if old and new are the same, this'll just do one lock. */ @@ -1304,6 +1418,7 @@ static int ocfs2_rename(struct inode *old_dir, mlog_errno(status); goto bail; } + should_add_orphan = true; } } else { BUG_ON(new_dentry->d_parent->d_inode != new_dir); @@ -1348,17 +1463,6 @@ static int ocfs2_rename(struct inode *old_dir, goto bail; } - if (S_ISDIR(new_inode->i_mode) || - (ocfs2_read_links_count(newfe) == 1)) { - status = ocfs2_orphan_add(osb, handle, new_inode, - newfe_bh, orphan_name, - &orphan_insert, orphan_dir); - if (status < 0) { - mlog_errno(status); - goto bail; - } - } - /* change the dirent to point to the correct inode */ status = ocfs2_update_entry(new_dir, handle, &target_lookup_res, old_inode); @@ -1373,6 +1477,15 @@ static int ocfs2_rename(struct inode *old_dir, else ocfs2_add_links_count(newfe, -1); ocfs2_journal_dirty(handle, newfe_bh); + if (should_add_orphan) { + status = ocfs2_orphan_add(osb, handle, new_inode, + newfe_bh, orphan_name, + &orphan_insert, orphan_dir); + if (status < 0) { + mlog_errno(status); + goto bail; + } + } } else { /* if the name was not found in new_dir, add it now */ status = ocfs2_add_entry(handle, new_dentry, old_inode, @@ -1642,6 +1755,7 @@ static int ocfs2_symlink(struct inode *dir, struct ocfs2_dir_lookup_result lookup = { NULL, }; sigset_t oldset; int did_block_signals = 0; + struct ocfs2_dentry_lock *dl = NULL; trace_ocfs2_symlink_begin(dir, dentry, symname, dentry->d_name.len, dentry->d_name.name); @@ -1830,6 +1944,8 @@ static int ocfs2_symlink(struct inode *dir, goto bail; } + dl = dentry->d_fsdata; + status = ocfs2_add_entry(handle, dentry, inode, le64_to_cpu(fe->i_blkno), parent_fe_bh, &lookup); @@ -1864,6 +1980,9 @@ bail: if (xattr_ac) ocfs2_free_alloc_context(xattr_ac); if ((status < 0) && inode) { + if (dl) + ocfs2_cleanup_add_entry_failure(osb, dentry, inode); + OCFS2_I(inode)->ip_flags |= OCFS2_INODE_SKIP_ORPHAN_DIR; clear_nlink(inode); iput(inode); diff --git a/fs/ocfs2/ocfs2_trace.h b/fs/ocfs2/ocfs2_trace.h index 1b60c62aa9d6..6cb019b7c6a8 100644 --- a/fs/ocfs2/ocfs2_trace.h +++ b/fs/ocfs2/ocfs2_trace.h @@ -2292,6 +2292,8 @@ TRACE_EVENT(ocfs2_rename, __entry->new_len, __get_str(new_name)) ); +DEFINE_OCFS2_ULL_ULL_EVENT(ocfs2_rename_not_permitted); + TRACE_EVENT(ocfs2_rename_target_exists, TP_PROTO(int new_len, const char *new_name), TP_ARGS(new_len, new_name), diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index 714e53b9cc66..636aab69ead5 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -4288,9 +4288,16 @@ static int ocfs2_reflink(struct dentry *old_dentry, struct inode *dir, goto out; } + error = ocfs2_rw_lock(inode, 1); + if (error) { + mlog_errno(error); + goto out; + } + error = ocfs2_inode_lock(inode, &old_bh, 1); if (error) { mlog_errno(error); + ocfs2_rw_unlock(inode, 1); goto out; } @@ -4302,6 +4309,7 @@ static int ocfs2_reflink(struct dentry *old_dentry, struct inode *dir, up_write(&OCFS2_I(inode)->ip_xattr_sem); ocfs2_inode_unlock(inode, 1); + ocfs2_rw_unlock(inode, 1); brelse(old_bh); if (error) { diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index c7a89cea5c5d..ddb662b32447 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1925,15 +1925,11 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err) ocfs2_shutdown_local_alloc(osb); + ocfs2_truncate_log_shutdown(osb); + /* This will disable recovery and flush any recovery work. */ ocfs2_recovery_exit(osb); - /* - * During dismount, when it recovers another node it will call - * ocfs2_recover_orphans and queue delayed work osb_truncate_log_wq. - */ - ocfs2_truncate_log_shutdown(osb); - ocfs2_journal_shutdown(osb); ocfs2_sync_blockdev(sb); |