diff options
| author |   Matthew Wilcox (Oracle) <willy@infradead.org> | 2023-09-07 09:40:07 -0400 |
|---|---|---|
| committer |   Theodore Ts'o <tytso@mit.edu> | 2023-09-07 15:16:49 -0400 |
| commit | f94cf2206b066bd6d761d3347fd35f77b828c376 (patch) | |
| tree | 4ea354425663f47fb1d3145675bea1b914b5d530 /include/linux/buffer_head.h | |
| parent | ext4: fix slab-use-after-free in ext4_es_insert_extent() (diff) | |
| download | wireguard-linux-f94cf2206b066bd6d761d3347fd35f77b828c376.tar.xz wireguard-linux-f94cf2206b066bd6d761d3347fd35f77b828c376.zip | |
buffer: Make bh_offset() work for compound pages
If the buffer pointed to by the buffer_head is part of a compound page,
bh_offset() assumes that b_page is the precise page that contains
the data. A recent change to jbd2 inadvertently violated that assumption.
By using page_size(), we support both b_page being set to the head page
(as page_size() will return the size of the entire folio) and the precise
page (as it will return PAGE_SIZE for a tail page).
Fixes: 8147c4c4546f ("jbd2: use a folio in jbd2_journal_write_metadata_buffer()")
Reported-by: Zorro Lang <zlang@kernel.org>
Tested-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Diffstat (limited to 'include/linux/buffer_head.h')
| -rw-r--r-- | include/linux/buffer_head.h | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h index 6cb3e9af78c9..4ba242073adc 100644 --- a/include/linux/buffer_head.h +++ b/include/linux/buffer_head.h @@ -173,7 +173,10 @@ static __always_inline int buffer_uptodate(const struct buffer_head *bh) return test_bit_acquire(BH_Uptodate, &bh->b_state); } -#define bh_offset(bh) ((unsigned long)(bh)->b_data & ~PAGE_MASK) +static inline unsigned long bh_offset(const struct buffer_head *bh) +{ + return (unsigned long)(bh)->b_data & (page_size(bh->b_page) - 1); +} /* If we *know* page->private refers to buffer_heads */ #define page_buffers(page) \ |