diff options
Diffstat (limited to 'fs/buffer.c')
| -rw-r--r-- | fs/buffer.c | 457 |
1 files changed, 244 insertions, 213 deletions
diff --git a/fs/buffer.c b/fs/buffer.c index 4f73d23c2c46..8cf4a1dc481e 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -176,21 +176,11 @@ void end_buffer_write_sync(struct buffer_head *bh, int uptodate) } EXPORT_SYMBOL(end_buffer_write_sync); -/* - * Various filesystems appear to want __find_get_block to be non-blocking. - * But it's the page lock which protects the buffers. To get around this, - * we get exclusion from try_to_free_buffers with the blockdev mapping's - * i_private_lock. - * - * Hack idea: for the blockdev mapping, i_private_lock contention - * may be quite high. This code could TryLock the page, and if that - * succeeds, there is no need to take i_private_lock. - */ static struct buffer_head * -__find_get_block_slow(struct block_device *bdev, sector_t block) +__find_get_block_slow(struct block_device *bdev, sector_t block, bool atomic) { - struct inode *bd_inode = bdev->bd_inode; - struct address_space *bd_mapping = bd_inode->i_mapping; + struct address_space *bd_mapping = bdev->bd_mapping; + const int blkbits = bd_mapping->host->i_blkbits; struct buffer_head *ret = NULL; pgoff_t index; struct buffer_head *bh; @@ -199,15 +189,33 @@ __find_get_block_slow(struct block_device *bdev, sector_t block) int all_mapped = 1; static DEFINE_RATELIMIT_STATE(last_warned, HZ, 1); - index = ((loff_t)block << bd_inode->i_blkbits) / PAGE_SIZE; + index = ((loff_t)block << blkbits) / PAGE_SIZE; folio = __filemap_get_folio(bd_mapping, index, FGP_ACCESSED, 0); if (IS_ERR(folio)) goto out; - spin_lock(&bd_mapping->i_private_lock); + /* + * Folio lock protects the buffers. Callers that cannot block + * will fallback to serializing vs try_to_free_buffers() via + * the i_private_lock. + */ + if (atomic) + spin_lock(&bd_mapping->i_private_lock); + else + folio_lock(folio); + head = folio_buffers(folio); if (!head) goto out_unlock; + /* + * Upon a noref migration, the folio lock serializes here; + * otherwise bail. + */ + if (test_bit_acquire(BH_Migrate, &head->b_state)) { + WARN_ON(!atomic); + goto out_unlock; + } + bh = head; do { if (!buffer_mapped(bh)) @@ -233,10 +241,13 @@ __find_get_block_slow(struct block_device *bdev, sector_t block) (unsigned long long)block, (unsigned long long)bh->b_blocknr, bh->b_state, bh->b_size, bdev, - 1 << bd_inode->i_blkbits); + 1 << blkbits); } out_unlock: - spin_unlock(&bd_mapping->i_private_lock); + if (atomic) + spin_unlock(&bd_mapping->i_private_lock); + else + folio_unlock(folio); folio_put(folio); out: return ret; @@ -258,7 +269,6 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) } else { clear_buffer_uptodate(bh); buffer_io_error(bh, ", async page read"); - folio_set_error(folio); } /* @@ -287,7 +297,6 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate) still_busy: spin_unlock_irqrestore(&first->b_uptodate_lock, flags); - return; } struct postprocess_bh_ctx { @@ -391,7 +400,6 @@ static void end_buffer_async_write(struct buffer_head *bh, int uptodate) buffer_io_error(bh, ", lost async page write"); mark_buffer_write_io_error(bh); clear_buffer_uptodate(bh); - folio_set_error(folio); } first = folio_buffers(folio); @@ -413,7 +421,6 @@ static void end_buffer_async_write(struct buffer_head *bh, int uptodate) still_busy: spin_unlock_irqrestore(&first->b_uptodate_lock, flags); - return; } /* @@ -658,7 +665,9 @@ EXPORT_SYMBOL(generic_buffers_fsync); void write_boundary_block(struct block_device *bdev, sector_t bblock, unsigned blocksize) { - struct buffer_head *bh = __find_get_block(bdev, bblock + 1, blocksize); + struct buffer_head *bh; + + bh = __find_get_block_nonatomic(bdev, bblock + 1, blocksize); if (bh) { if (buffer_dirty(bh)) write_dirty_buffer(bh, 0); @@ -687,30 +696,37 @@ void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode) } EXPORT_SYMBOL(mark_buffer_dirty_inode); -/* - * Add a page to the dirty page list. - * - * It is a sad fact of life that this function is called from several places - * deeply under spinlocking. It may not sleep. - * - * If the page has buffers, the uptodate buffers are set dirty, to preserve - * dirty-state coherency between the page and the buffers. It the page does - * not have buffers then when they are later attached they will all be set - * dirty. - * - * The buffers are dirtied before the page is dirtied. There's a small race - * window in which a writepage caller may see the page cleanness but not the - * buffer dirtiness. That's fine. If this code were to set the page dirty - * before the buffers, a concurrent writepage caller could clear the page dirty - * bit, see a bunch of clean buffers and we'd end up with dirty buffers/clean - * page on the dirty page list. - * - * We use i_private_lock to lock against try_to_free_buffers while using the - * page's buffer list. Also use this to protect against clean buffers being - * added to the page after it was set dirty. - * - * FIXME: may need to call ->reservepage here as well. That's rather up to the - * address_space though. +/** + * block_dirty_folio - Mark a folio as dirty. + * @mapping: The address space containing this folio. + * @folio: The folio to mark dirty. + * + * Filesystems which use buffer_heads can use this function as their + * ->dirty_folio implementation. Some filesystems need to do a little + * work before calling this function. Filesystems which do not use + * buffer_heads should call filemap_dirty_folio() instead. + * + * If the folio has buffers, the uptodate buffers are set dirty, to + * preserve dirty-state coherency between the folio and the buffers. + * Buffers added to a dirty folio are created dirty. + * + * The buffers are dirtied before the folio is dirtied. There's a small + * race window in which writeback may see the folio cleanness but not the + * buffer dirtiness. That's fine. If this code were to set the folio + * dirty before the buffers, writeback could clear the folio dirty flag, + * see a bunch of clean buffers and we'd end up with dirty buffers/clean + * folio on the dirty folio list. + * + * We use i_private_lock to lock against try_to_free_buffers() while + * using the folio's buffer list. This also prevents clean buffers + * being added to the folio after it was set dirty. + * + * Context: May only be called from process context. Does not sleep. + * Caller must ensure that @folio cannot be truncated during this call, + * typically by holding the folio lock or having a page in the folio + * mapped and holding the page table lock. + * + * Return: True if the folio was dirtied; false if it was already dirtied. */ bool block_dirty_folio(struct address_space *mapping, struct folio *folio) { @@ -731,15 +747,12 @@ bool block_dirty_folio(struct address_space *mapping, struct folio *folio) * Lock out page's memcg migration to keep PageDirty * synchronized with per-memcg dirty page counters. */ - folio_memcg_lock(folio); newly_dirty = !folio_test_set_dirty(folio); spin_unlock(&mapping->i_private_lock); if (newly_dirty) __folio_mark_dirty(folio, mapping, 1); - folio_memcg_unlock(folio); - if (newly_dirty) __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); @@ -769,12 +782,11 @@ EXPORT_SYMBOL(block_dirty_folio); static int fsync_buffers_list(spinlock_t *lock, struct list_head *list) { struct buffer_head *bh; - struct list_head tmp; struct address_space *mapping; int err = 0, err2; struct blk_plug plug; + LIST_HEAD(tmp); - INIT_LIST_HEAD(&tmp); blk_start_plug(&plug); spin_lock(lock); @@ -851,8 +863,7 @@ static int fsync_buffers_list(spinlock_t *lock, struct list_head *list) * done a sync(). Just drop the buffers from the inode list. * * NOTE: we take the inode's blockdev's mapping's i_private_lock. Which - * assumes that all the buffers are against the blockdev. Not true - * for reiserfs. + * assumes that all the buffers are against the blockdev. */ void invalidate_inode_buffers(struct inode *inode) { @@ -953,12 +964,9 @@ no_grow: } EXPORT_SYMBOL_GPL(folio_alloc_buffers); -struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size, - bool retry) +struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size) { gfp_t gfp = GFP_NOFS | __GFP_ACCOUNT; - if (retry) - gfp |= __GFP_NOFAIL; return folio_alloc_buffers(page_folio(page), size, gfp); } @@ -1034,12 +1042,12 @@ static sector_t folio_init_buffers(struct folio *folio, static bool grow_dev_folio(struct block_device *bdev, sector_t block, pgoff_t index, unsigned size, gfp_t gfp) { - struct inode *inode = bdev->bd_inode; + struct address_space *mapping = bdev->bd_mapping; struct folio *folio; struct buffer_head *bh; sector_t end_block = 0; - folio = __filemap_get_folio(inode->i_mapping, index, + folio = __filemap_get_folio(mapping, index, FGP_LOCK | FGP_ACCESSED | FGP_CREAT, gfp); if (IS_ERR(folio)) return false; @@ -1073,10 +1081,10 @@ static bool grow_dev_folio(struct block_device *bdev, sector_t block, * lock to be atomic wrt __find_get_block(), which does not * run under the folio lock. */ - spin_lock(&inode->i_mapping->i_private_lock); + spin_lock(&mapping->i_private_lock); link_dev_buffers(folio, bh); end_block = folio_init_buffers(folio, bdev, size); - spin_unlock(&inode->i_mapping->i_private_lock); + spin_unlock(&mapping->i_private_lock); unlock: folio_unlock(folio); folio_put(folio); @@ -1112,6 +1120,8 @@ static struct buffer_head * __getblk_slow(struct block_device *bdev, sector_t block, unsigned size, gfp_t gfp) { + bool blocking = gfpflags_allow_blocking(gfp); + /* Size must be multiple of hard sectorsize */ if (unlikely(size & (bdev_logical_block_size(bdev)-1) || (size < 512 || size > PAGE_SIZE))) { @@ -1127,12 +1137,15 @@ __getblk_slow(struct block_device *bdev, sector_t block, for (;;) { struct buffer_head *bh; - bh = __find_get_block(bdev, block, size); - if (bh) - return bh; - if (!grow_buffers(bdev, block, size, gfp)) return NULL; + + if (blocking) + bh = __find_get_block_nonatomic(bdev, block, size); + else + bh = __find_get_block(bdev, block, size); + if (bh) + return bh; } } @@ -1193,13 +1206,11 @@ void mark_buffer_dirty(struct buffer_head *bh) struct folio *folio = bh->b_folio; struct address_space *mapping = NULL; - folio_memcg_lock(folio); if (!folio_test_set_dirty(folio)) { mapping = folio->mapping; if (mapping) __folio_mark_dirty(folio, mapping, 0); } - folio_memcg_unlock(folio); if (mapping) __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); } @@ -1212,33 +1223,33 @@ void mark_buffer_write_io_error(struct buffer_head *bh) /* FIXME: do we need to set this in both places? */ if (bh->b_folio && bh->b_folio->mapping) mapping_set_error(bh->b_folio->mapping, -EIO); - if (bh->b_assoc_map) { + if (bh->b_assoc_map) mapping_set_error(bh->b_assoc_map, -EIO); - errseq_set(&bh->b_assoc_map->host->i_sb->s_wb_err, -EIO); - } } EXPORT_SYMBOL(mark_buffer_write_io_error); -/* - * Decrement a buffer_head's reference count. If all buffers against a page - * have zero reference count, are clean and unlocked, and if the page is clean - * and unlocked then try_to_free_buffers() may strip the buffers from the page - * in preparation for freeing it (sometimes, rarely, buffers are removed from - * a page but it ends up not being freed, and buffers may later be reattached). +/** + * __brelse - Release a buffer. + * @bh: The buffer to release. + * + * This variant of brelse() can be called if @bh is guaranteed to not be NULL. */ -void __brelse(struct buffer_head * buf) +void __brelse(struct buffer_head *bh) { - if (atomic_read(&buf->b_count)) { - put_bh(buf); + if (atomic_read(&bh->b_count)) { + put_bh(bh); return; } WARN(1, KERN_ERR "VFS: brelse: Trying to free free buffer\n"); } EXPORT_SYMBOL(__brelse); -/* - * bforget() is like brelse(), except it discards any - * potentially dirty data. +/** + * __bforget - Discard any dirty data in a buffer. + * @bh: The buffer to forget. + * + * This variant of bforget() can be called if @bh is guaranteed to not + * be NULL. */ void __bforget(struct buffer_head *bh) { @@ -1389,16 +1400,18 @@ lookup_bh_lru(struct block_device *bdev, sector_t block, unsigned size) /* * Perform a pagecache lookup for the matching buffer. If it's there, refresh * it in the LRU and mark it as accessed. If it is not present then return - * NULL + * NULL. Atomic context callers may also return NULL if the buffer is being + * migrated; similarly the page is not marked accessed either. */ -struct buffer_head * -__find_get_block(struct block_device *bdev, sector_t block, unsigned size) +static struct buffer_head * +find_get_block_common(struct block_device *bdev, sector_t block, + unsigned size, bool atomic) { struct buffer_head *bh = lookup_bh_lru(bdev, block, size); if (bh == NULL) { /* __find_get_block_slow will mark the page accessed */ - bh = __find_get_block_slow(bdev, block); + bh = __find_get_block_slow(bdev, block, atomic); if (bh) bh_lru_install(bh); } else @@ -1406,8 +1419,23 @@ __find_get_block(struct block_device *bdev, sector_t block, unsigned size) return bh; } + +struct buffer_head * +__find_get_block(struct block_device *bdev, sector_t block, unsigned size) +{ + return find_get_block_common(bdev, block, size, true); +} EXPORT_SYMBOL(__find_get_block); +/* same as __find_get_block() but allows sleeping contexts */ +struct buffer_head * +__find_get_block_nonatomic(struct block_device *bdev, sector_t block, + unsigned size) +{ + return find_get_block_common(bdev, block, size, false); +} +EXPORT_SYMBOL(__find_get_block_nonatomic); + /** * bdev_getblk - Get a buffer_head in a block device's buffer cache. * @bdev: The block device. @@ -1415,12 +1443,22 @@ EXPORT_SYMBOL(__find_get_block); * @size: The size of buffer_heads for this @bdev. * @gfp: The memory allocation flags to use. * + * The returned buffer head has its reference count incremented, but is + * not locked. The caller should call brelse() when it has finished + * with the buffer. The buffer may not be uptodate. If needed, the + * caller can bring it uptodate either by reading it or overwriting it. + * * Return: The buffer head, or NULL if memory could not be allocated. */ struct buffer_head *bdev_getblk(struct block_device *bdev, sector_t block, unsigned size, gfp_t gfp) { - struct buffer_head *bh = __find_get_block(bdev, block, size); + struct buffer_head *bh; + + if (gfpflags_allow_blocking(gfp)) + bh = __find_get_block_nonatomic(bdev, block, size); + else + bh = __find_get_block(bdev, block, size); might_alloc(gfp); if (bh) @@ -1446,24 +1484,33 @@ void __breadahead(struct block_device *bdev, sector_t block, unsigned size) EXPORT_SYMBOL(__breadahead); /** - * __bread_gfp() - reads a specified block and returns the bh - * @bdev: the block_device to read from - * @block: number of block - * @size: size (in bytes) to read - * @gfp: page allocation flag - * - * Reads a specified block, and returns buffer head that contains it. - * The page cache can be allocated from non-movable area - * not to prevent page migration if you set gfp to zero. - * It returns NULL if the block was unreadable. + * __bread_gfp() - Read a block. + * @bdev: The block device to read from. + * @block: Block number in units of block size. + * @size: The block size of this device in bytes. + * @gfp: Not page allocation flags; see below. + * + * You are not expected to call this function. You should use one of + * sb_bread(), sb_bread_unmovable() or __bread(). + * + * Read a specified block, and return the buffer head that refers to it. + * If @gfp is 0, the memory will be allocated using the block device's + * default GFP flags. If @gfp is __GFP_MOVABLE, the memory may be + * allocated from a movable area. Do not pass in a complete set of + * GFP flags. + * + * The returned buffer head has its refcount increased. The caller should + * call brelse() when it has finished with the buffer. + * + * Context: May sleep waiting for I/O. + * Return: NULL if the block was unreadable. */ -struct buffer_head * -__bread_gfp(struct block_device *bdev, sector_t block, - unsigned size, gfp_t gfp) +struct buffer_head *__bread_gfp(struct block_device *bdev, sector_t block, + unsigned size, gfp_t gfp) { struct buffer_head *bh; - gfp |= mapping_gfp_constraint(bdev->bd_inode->i_mapping, ~__GFP_FS); + gfp |= mapping_gfp_constraint(bdev->bd_mapping, ~__GFP_FS); /* * Prefer looping in the allocator rather than here, at least that @@ -1567,8 +1614,8 @@ static void discard_buffer(struct buffer_head * bh) bh->b_bdev = NULL; b_state = READ_ONCE(bh->b_state); do { - } while (!try_cmpxchg(&bh->b_state, &b_state, - b_state & ~BUFFER_FLAGS_DISCARD)); + } while (!try_cmpxchg_relaxed(&bh->b_state, &b_state, + b_state & ~BUFFER_FLAGS_DISCARD)); unlock_buffer(bh); } @@ -1632,7 +1679,7 @@ void block_invalidate_folio(struct folio *folio, size_t offset, size_t length) if (length == folio_size(folio)) filemap_release_folio(folio, 0); out: - return; + folio_clear_mappedtodisk(folio); } EXPORT_SYMBOL(block_invalidate_folio); @@ -1696,16 +1743,16 @@ EXPORT_SYMBOL(create_empty_buffers); */ void clean_bdev_aliases(struct block_device *bdev, sector_t block, sector_t len) { - struct inode *bd_inode = bdev->bd_inode; - struct address_space *bd_mapping = bd_inode->i_mapping; + struct address_space *bd_mapping = bdev->bd_mapping; + const int blkbits = bd_mapping->host->i_blkbits; struct folio_batch fbatch; - pgoff_t index = ((loff_t)block << bd_inode->i_blkbits) / PAGE_SIZE; + pgoff_t index = ((loff_t)block << blkbits) / PAGE_SIZE; pgoff_t end; int i, count; struct buffer_head *bh; struct buffer_head *head; - end = ((loff_t)(block + len - 1) << bd_inode->i_blkbits) / PAGE_SIZE; + end = ((loff_t)(block + len - 1) << blkbits) / PAGE_SIZE; folio_batch_init(&fbatch); while (filemap_get_folios(bd_mapping, &index, end, &fbatch)) { count = folio_batch_count(&fbatch); @@ -1937,7 +1984,6 @@ recover: clear_buffer_dirty(bh); } } while ((bh = bh->b_this_page) != head); - folio_set_error(folio); BUG_ON(folio_test_writeback(folio)); mapping_set_error(folio->mapping, err); folio_start_writeback(folio); @@ -2148,15 +2194,14 @@ int __block_write_begin_int(struct folio *folio, loff_t pos, unsigned len, return err; } -int __block_write_begin(struct page *page, loff_t pos, unsigned len, +int __block_write_begin(struct folio *folio, loff_t pos, unsigned len, get_block_t *get_block) { - return __block_write_begin_int(page_folio(page), pos, len, get_block, - NULL); + return __block_write_begin_int(folio, pos, len, get_block, NULL); } EXPORT_SYMBOL(__block_write_begin); -static void __block_commit_write(struct folio *folio, size_t from, size_t to) +void block_commit_write(struct folio *folio, size_t from, size_t to) { size_t block_start, block_end; bool partial = false; @@ -2164,6 +2209,8 @@ static void __block_commit_write(struct folio *folio, size_t from, size_t to) struct buffer_head *bh, *head; bh = head = folio_buffers(folio); + if (!bh) + return; blocksize = bh->b_size; block_start = 0; @@ -2192,6 +2239,7 @@ static void __block_commit_write(struct folio *folio, size_t from, size_t to) if (!partial) folio_mark_uptodate(folio); } +EXPORT_SYMBOL(block_commit_write); /* * block_write_begin takes care of the basic task of block allocation and @@ -2200,33 +2248,33 @@ static void __block_commit_write(struct folio *folio, size_t from, size_t to) * The filesystem needs to handle block truncation upon failure. */ int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len, - struct page **pagep, get_block_t *get_block) + struct folio **foliop, get_block_t *get_block) { pgoff_t index = pos >> PAGE_SHIFT; - struct page *page; + struct folio *folio; int status; - page = grab_cache_page_write_begin(mapping, index); - if (!page) - return -ENOMEM; + folio = __filemap_get_folio(mapping, index, FGP_WRITEBEGIN, + mapping_gfp_mask(mapping)); + if (IS_ERR(folio)) + return PTR_ERR(folio); - status = __block_write_begin(page, pos, len, get_block); + status = __block_write_begin_int(folio, pos, len, get_block, NULL); if (unlikely(status)) { - unlock_page(page); - put_page(page); - page = NULL; + folio_unlock(folio); + folio_put(folio); + folio = NULL; } - *pagep = page; + *foliop = folio; return status; } EXPORT_SYMBOL(block_write_begin); int block_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct folio *folio, void *fsdata) { - struct folio *folio = page_folio(page); size_t start = pos - folio_pos(folio); if (unlikely(copied < len)) { @@ -2250,7 +2298,7 @@ int block_write_end(struct file *file, struct address_space *mapping, flush_dcache_folio(folio); /* This could be a short (even 0-length) commit */ - __block_commit_write(folio, start, start + copied); + block_commit_write(folio, start, start + copied); return copied; } @@ -2258,19 +2306,19 @@ EXPORT_SYMBOL(block_write_end); int generic_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct folio *folio, void *fsdata) { struct inode *inode = mapping->host; loff_t old_size = inode->i_size; bool i_size_changed = false; - copied = block_write_end(file, mapping, pos, len, copied, page, fsdata); + copied = block_write_end(file, mapping, pos, len, copied, folio, fsdata); /* * No need to use i_size_read() here, the i_size cannot change under us * because we hold i_rwsem. * - * But it's important to update i_size while still holding page lock: + * But it's important to update i_size while still holding folio lock: * page writeout could otherwise come in and zero beyond i_size. */ if (pos + copied > inode->i_size) { @@ -2278,8 +2326,8 @@ int generic_write_end(struct file *file, struct address_space *mapping, i_size_changed = true; } - unlock_page(page); - put_page(page); + folio_unlock(folio); + folio_put(folio); if (old_size < pos) pagecache_isize_extended(inode, old_size, pos); @@ -2349,9 +2397,8 @@ int block_read_full_folio(struct folio *folio, get_block_t *get_block) { struct inode *inode = folio->mapping->host; sector_t iblock, lblock; - struct buffer_head *bh, *head, *arr[MAX_BUF_PER_PAGE]; + struct buffer_head *bh, *head, *prev = NULL; size_t blocksize; - int nr, i; int fully_mapped = 1; bool page_error = false; loff_t limit = i_size_read(inode); @@ -2360,16 +2407,12 @@ int block_read_full_folio(struct folio *folio, get_block_t *get_block) if (IS_ENABLED(CONFIG_FS_VERITY) && IS_VERITY(inode)) limit = inode->i_sb->s_maxbytes; - VM_BUG_ON_FOLIO(folio_test_large(folio), folio); - head = folio_create_buffers(folio, inode, 0); blocksize = head->b_size; iblock = div_u64(folio_pos(folio), blocksize); lblock = div_u64(limit + blocksize - 1, blocksize); bh = head; - nr = 0; - i = 0; do { if (buffer_uptodate(bh)) @@ -2382,13 +2425,11 @@ int block_read_full_folio(struct folio *folio, get_block_t *get_block) if (iblock < lblock) { WARN_ON(bh->b_size != blocksize); err = get_block(inode, iblock, bh, 0); - if (err) { - folio_set_error(folio); + if (err) page_error = true; - } } if (!buffer_mapped(bh)) { - folio_zero_range(folio, i * blocksize, + folio_zero_range(folio, bh_offset(bh), blocksize); if (!err) set_buffer_uptodate(bh); @@ -2401,40 +2442,33 @@ int block_read_full_folio(struct folio *folio, get_block_t *get_block) if (buffer_uptodate(bh)) continue; } - arr[nr++] = bh; - } while (i++, iblock++, (bh = bh->b_this_page) != head); - - if (fully_mapped) - folio_set_mappedtodisk(folio); - - if (!nr) { - /* - * All buffers are uptodate or get_block() returned an - * error when trying to map them - we can finish the read. - */ - folio_end_read(folio, !page_error); - return 0; - } - /* Stage two: lock the buffers */ - for (i = 0; i < nr; i++) { - bh = arr[i]; lock_buffer(bh); + if (buffer_uptodate(bh)) { + unlock_buffer(bh); + continue; + } + mark_buffer_async_read(bh); - } + if (prev) + submit_bh(REQ_OP_READ, prev); + prev = bh; + } while (iblock++, (bh = bh->b_this_page) != head); + + if (fully_mapped) + folio_set_mappedtodisk(folio); /* - * Stage 3: start the IO. Check for uptodateness - * inside the buffer lock in case another process reading - * the underlying blockdev brought it uptodate (the sct fix). + * All buffers are uptodate or get_block() returned an error + * when trying to map them - we must finish the read because + * end_buffer_async_read() will never be called on any buffer + * in this folio. */ - for (i = 0; i < nr; i++) { - bh = arr[i]; - if (buffer_uptodate(bh)) - end_buffer_async_read(bh, 1); - else - submit_bh(REQ_OP_READ, bh); - } + if (prev) + submit_bh(REQ_OP_READ, prev); + else + folio_end_read(folio, !page_error); + return 0; } EXPORT_SYMBOL(block_read_full_folio); @@ -2447,7 +2481,7 @@ int generic_cont_expand_simple(struct inode *inode, loff_t size) { struct address_space *mapping = inode->i_mapping; const struct address_space_operations *aops = mapping->a_ops; - struct page *page; + struct folio *folio; void *fsdata = NULL; int err; @@ -2455,11 +2489,11 @@ int generic_cont_expand_simple(struct inode *inode, loff_t size) if (err) goto out; - err = aops->write_begin(NULL, mapping, size, 0, &page, &fsdata); + err = aops->write_begin(NULL, mapping, size, 0, &folio, &fsdata); if (err) goto out; - err = aops->write_end(NULL, mapping, size, 0, 0, page, fsdata); + err = aops->write_end(NULL, mapping, size, 0, 0, folio, fsdata); BUG_ON(err > 0); out: @@ -2473,7 +2507,7 @@ static int cont_expand_zero(struct file *file, struct address_space *mapping, struct inode *inode = mapping->host; const struct address_space_operations *aops = mapping->a_ops; unsigned int blocksize = i_blocksize(inode); - struct page *page; + struct folio *folio; void *fsdata = NULL; pgoff_t index, curidx; loff_t curpos; @@ -2492,12 +2526,12 @@ static int cont_expand_zero(struct file *file, struct address_space *mapping, len = PAGE_SIZE - zerofrom; err = aops->write_begin(file, mapping, curpos, len, - &page, &fsdata); + &folio, &fsdata); if (err) goto out; - zero_user(page, zerofrom, len); + folio_zero_range(folio, offset_in_folio(folio, curpos), len); err = aops->write_end(file, mapping, curpos, len, len, - page, fsdata); + folio, fsdata); if (err < 0) goto out; BUG_ON(err != len); @@ -2525,12 +2559,12 @@ static int cont_expand_zero(struct file *file, struct address_space *mapping, len = offset - zerofrom; err = aops->write_begin(file, mapping, curpos, len, - &page, &fsdata); + &folio, &fsdata); if (err) goto out; - zero_user(page, zerofrom, len); + folio_zero_range(folio, offset_in_folio(folio, curpos), len); err = aops->write_end(file, mapping, curpos, len, len, - page, fsdata); + folio, fsdata); if (err < 0) goto out; BUG_ON(err != len); @@ -2546,7 +2580,7 @@ out: */ int cont_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, - struct page **pagep, void **fsdata, + struct folio **foliop, void **fsdata, get_block_t *get_block, loff_t *bytes) { struct inode *inode = mapping->host; @@ -2564,17 +2598,10 @@ int cont_write_begin(struct file *file, struct address_space *mapping, (*bytes)++; } - return block_write_begin(mapping, pos, len, pagep, get_block); + return block_write_begin(mapping, pos, len, foliop, get_block); } EXPORT_SYMBOL(cont_write_begin); -void block_commit_write(struct page *page, unsigned from, unsigned to) -{ - struct folio *folio = page_folio(page); - __block_commit_write(folio, from, to); -} -EXPORT_SYMBOL(block_commit_write); - /* * block_page_mkwrite() is not allowed to change the file size as it gets * called from a page fault handler when a page is first dirtied. Hence we must @@ -2620,7 +2647,7 @@ int block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, if (unlikely(ret)) goto out_unlock; - __block_commit_write(folio, 0, end); + block_commit_write(folio, 0, end); folio_mark_dirty(folio); folio_wait_stable(folio); @@ -2703,7 +2730,7 @@ unlock: EXPORT_SYMBOL(block_truncate_page); /* - * The generic ->writepage function for buffer-backed address_spaces + * The generic write folio function for buffer-backed address_spaces */ int block_write_full_folio(struct folio *folio, struct writeback_control *wbc, void *get_block) @@ -2723,7 +2750,7 @@ int block_write_full_folio(struct folio *folio, struct writeback_control *wbc, /* * The folio straddles i_size. It must be zeroed out on each and every - * writepage invocation because it may be mmapped. "A file is mapped + * writeback invocation because it may be mmapped. "A file is mapped * in multiples of the page size. For a file that is not a multiple of * the page size, the remaining memory is zeroed when mapped, and * writes to that region are not written out to the file." @@ -2788,7 +2815,7 @@ static void submit_bh_wbc(blk_opf_t opf, struct buffer_head *bh, bio->bi_iter.bi_sector = bh->b_blocknr * (bh->b_size >> 9); bio->bi_write_hint = write_hint; - __bio_add_page(bio, bh->b_page, bh->b_size, bh_offset(bh)); + bio_add_folio_nofail(bio, bh->b_folio, bh->b_size, bh_offset(bh)); bio->bi_end_io = end_bio_bh_io_sync; bio->bi_private = bh; @@ -2798,7 +2825,7 @@ static void submit_bh_wbc(blk_opf_t opf, struct buffer_head *bh, if (wbc) { wbc_init_bio(wbc, bio); - wbc_account_cgroup_owner(wbc, bh->b_page, bh->b_size); + wbc_account_cgroup_owner(wbc, bh->b_folio, bh->b_size); } submit_bio(bio); @@ -2861,26 +2888,6 @@ int sync_dirty_buffer(struct buffer_head *bh) } EXPORT_SYMBOL(sync_dirty_buffer); -/* - * try_to_free_buffers() checks if all the buffers on this particular folio - * are unused, and releases them if so. - * - * Exclusion against try_to_free_buffers may be obtained by either - * locking the folio or by holding its mapping's i_private_lock. - * - * If the folio is dirty but all the buffers are clean then we need to - * be sure to mark the folio clean as well. This is because the folio - * may be against a block device, and a later reattachment of buffers - * to a dirty folio will set *all* buffers dirty. Which would corrupt - * filesystem data on the same device. - * - * The same applies to regular filesystem folios: if all the buffers are - * clean then we set the folio clean and proceed. To do that, we require - * total exclusion from block_dirty_folio(). That is obtained with - * i_private_lock. - * - * try_to_free_buffers() is non-blocking. - */ static inline int buffer_busy(struct buffer_head *bh) { return atomic_read(&bh->b_count) | @@ -2914,6 +2921,30 @@ failed: return false; } +/** + * try_to_free_buffers - Release buffers attached to this folio. + * @folio: The folio. + * + * If any buffers are in use (dirty, under writeback, elevated refcount), + * no buffers will be freed. + * + * If the folio is dirty but all the buffers are clean then we need to + * be sure to mark the folio clean as well. This is because the folio + * may be against a block device, and a later reattachment of buffers + * to a dirty folio will set *all* buffers dirty. Which would corrupt + * filesystem data on the same device. + * + * The same applies to regular filesystem folios: if all the buffers are + * clean then we set the folio clean and proceed. To do that, we require + * total exclusion from block_dirty_folio(). That is obtained with + * i_private_lock. + * + * Exclusion against try_to_free_buffers may be obtained by either + * locking the folio or by holding its mapping's i_private_lock. + * + * Context: Process context. @folio must be locked. Will not sleep. + * Return: true if all buffers attached to this folio were freed. + */ bool try_to_free_buffers(struct folio *folio) { struct address_space * const mapping = folio->mapping; |