ext4: fix delalloc i_disksize early update issue From: Mingming Cao Ext4_da_write_end() used walk_page_buffers() with a callback function of ext4_bh_unmapped_or_delay() to check if it extended the file size without allocating any blocks (since in this case i_disksize needs to be updated). However, this is didn't work proprely because the buffer head has not been marked dirty yet --- this is done later in block_commit_write() --- which caused ext4_bh_unmapped_or_delay() to always return false. In addition, walk_page_buffers() checks all of the buffer heads covering the page, and the only buffer_head that should be checked is the one covering the end of the write. Otherwise, given a 1k blocksize filesystem and a 4k page size, the buffer head covering the first 1k stripe of the file could be unmapped (because it was a sparse file), and the second or third buffer_head covering that page could be mapped, and using walk_page_buffers() would fail in this case since it would stop at the first unmapped buffer_head and return true. The core problem is that walk_page_buffers() was intended to do work in a callback function, and a non-zero return value indicated a failure, which termined the walk of the buffer heads covering the page. It was not intended to be used with a boolean function, such as ext4_bh_unmapped_or_delay(). Add addtional fix from Aneesh to protect i_disksize update rave with truncate. Signed-off-by: Mingming Cao Signed-off-by: Aneesh Kumar K.V Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 63 +++++++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 47 insertions(+), 16 deletions(-) Index: linux-2.6.26-rc9/fs/ext4/inode.c =================================================================== --- linux-2.6.26-rc9.orig/fs/ext4/inode.c 2008-07-11 16:05:14.000000000 -0700 +++ linux-2.6.26-rc9/fs/ext4/inode.c 2008-07-11 16:05:15.000000000 -0700 @@ -2292,6 +2292,29 @@ out: return ret; } +/* + * Check if we should update i_disksize + * when write to the end of file but not require block allocation + */ +static int ext4_da_should_update_i_disksize(struct page *page, + unsigned long offset) +{ + struct buffer_head *bh; + struct inode *inode = page->mapping->host; + unsigned int idx; + int i; + + bh = page_buffers(page); + idx = offset >> inode->i_blkbits; + + for (i=0; i < idx; i++) + bh = bh->b_this_page; + + if (!buffer_mapped(bh) || (buffer_delay(bh))) + return 0; + return 1; +} + static int ext4_da_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, @@ -2301,6 +2324,10 @@ static int ext4_da_write_end(struct file int ret = 0, ret2; handle_t *handle = ext4_journal_current_handle(); loff_t new_i_size; + unsigned long start, end; + + start = pos & (PAGE_CACHE_SIZE - 1); + end = start + copied -1; /* * generic_write_end() will run mark_inode_dirty() if i_size @@ -2309,18 +2336,23 @@ static int ext4_da_write_end(struct file */ new_i_size = pos + copied; - if (new_i_size > EXT4_I(inode)->i_disksize) - if (!walk_page_buffers(NULL, page_buffers(page), - 0, len, NULL, ext4_bh_unmapped_or_delay)){ - /* - * Updating i_disksize when extending file without - * needing block allocation - */ - if (ext4_should_order_data(inode)) - ret = ext4_jbd2_file_inode(handle, inode); + if (new_i_size > EXT4_I(inode)->i_disksize) { + if (ext4_da_should_update_i_disksize(page, end)) { + down_write(&EXT4_I(inode)->i_data_sem); + if (new_i_size > EXT4_I(inode)->i_disksize) { + /* + * Updating i_disksize when extending file + * without needing block allocation + */ + if (ext4_should_order_data(inode)) + ret = ext4_jbd2_file_inode(handle, + inode); - EXT4_I(inode)->i_disksize = new_i_size; + EXT4_I(inode)->i_disksize = new_i_size; + } + up_write(&EXT4_I(inode)->i_data_sem); } + } ret2 = generic_write_end(file, mapping, pos, len, copied, page, fsdata); copied = ret2; @@ -3388,6 +3420,11 @@ void ext4_truncate(struct inode *inode) goto out_stop; /* + * From here we block out all ext4_get_block() callers who want to + * modify the block allocation tree. + */ + down_write(&ei->i_data_sem); + /* * The orphan list entry will now protect us from any crash which * occurs before the truncate completes, so it is now safe to propagate * the new, shorter inode size (held for now in i_size) into the @@ -3396,12 +3433,6 @@ void ext4_truncate(struct inode *inode) */ ei->i_disksize = inode->i_size; - /* - * From here we block out all ext4_get_block() callers who want to - * modify the block allocation tree. - */ - down_write(&ei->i_data_sem); - if (n == 1) { /* direct blocks */ ext4_free_data(handle, inode, NULL, i_data+offsets[0], i_data + EXT4_NDIR_BLOCKS);