[BUG] With 64K page size (aarch64 with 64K page size config) and 4K btrfs block size, the following workload can easily lead to a corrupted read:
mkfs.btrfs -f -s 4k $dev > /dev/null mount -o compress $dev $mnt xfs_io -f -c "pwrite -S 0xff 0 64k" $mnt/base > /dev/null echo "correct result:" od -Ad -t x1 $mnt/base xfs_io -f -c "reflink $mnt/base 32k 0 32k" \ -c "reflink $mnt/base 0 32k 32k" \ -c "pwrite -S 0xff 60k 4k" $mnt/new > /dev/null echo "incorrect result:" od -Ad -t x1 $mnt/new umount $mnt
This shows the following result:
correct result: 0000000 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff * 0065536 incorrect result: 0000000 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff * 0032768 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 * 0061440 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff * 0065536
Notice the zero in the range [32K, 60K), which is incorrect.
[CAUSE] With extra trace printk, it shows the following events during od: (some unrelated info removed like CPU and context)
od-3457 btrfs_do_readpage: enter r/i=5/258 folio=0(65536) prev_em_start=0000000000000000
The "r/i" is indicating the root and inode number. In our case the file "new" is using ino 258 from fs tree (root 5).
Here notice the @prev_em_start pointer is NULL. This means the btrfs_do_readpage() is called from btrfs_read_folio(), not from btrfs_readahead().
od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=0 got em start=0 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=4096 got em start=0 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=8192 got em start=0 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=12288 got em start=0 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=16384 got em start=0 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=20480 got em start=0 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=24576 got em start=0 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=28672 got em start=0 len=32768
These above 32K blocks will be read from the first half of the compressed data extent.
od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=32768 got em start=32768 len=32768
Note here there is no btrfs_submit_compressed_read() call. Which is incorrect now. Although both extent maps at 0 and 32K are pointing to the same compressed data, their offsets are different thus can not be merged into the same read.
So this means the compressed data read merge check is doing something wrong.
od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=36864 got em start=32768 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=40960 got em start=32768 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=45056 got em start=32768 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=49152 got em start=32768 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=53248 got em start=32768 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=57344 got em start=32768 len=32768 od-3457 btrfs_do_readpage: r/i=5/258 folio=0(65536) cur=61440 skip uptodate od-3457 btrfs_submit_compressed_read: cb orig_bio: file off=0 len=61440
The function btrfs_submit_compressed_read() is only called at the end of folio read. The compressed bio will only have an extent map of range [0, 32K), but the original bio passed in is for the whole 64K folio.
This will cause the decompression part to only fill the first 32K, leaving the rest untouched (aka, filled with zero).
This incorrect compressed read merge leads to the above data corruption.
There were similar problems that happened in the past, commit 808f80b46790 ("Btrfs: update fix for read corruption of compressed and shared extents") is doing pretty much the same fix for readahead.
But that's back to 2015, where btrfs still only supports bs (block size) == ps (page size) cases. This means btrfs_do_readpage() only needs to handle a folio which contains exactly one block.
Only btrfs_readahead() can lead to a read covering multiple blocks. Thus only btrfs_readahead() passes a non-NULL @prev_em_start pointer.
With v5.15 kernel btrfs introduced bs < ps support. This breaks the above assumption that a folio can only contain one block.
Now btrfs_read_folio() can also read multiple blocks in one go. But btrfs_read_folio() doesn't pass a @prev_em_start pointer, thus the existing bio force submission check will never be triggered.
In theory, this can also happen for btrfs with large folios, but since large folio is still experimental, we don't need to bother it, thus only bs < ps support is affected for now.
[FIX] Instead of passing @prev_em_start to do the proper compressed extent check, introduce one new member, btrfs_bio_ctrl::last_em_start, so that the existing bio force submission logic will always be triggered.
CC: stable@vger.kernel.org # 5.15+ Reviewed-by: Filipe Manana fdmanana@suse.com Signed-off-by: Qu Wenruo wqu@suse.com Signed-off-by: David Sterba dsterba@suse.com --- Changelog: v4: - Minor grammar fixes - Update the subject to reflect the bug better
This is the merged version inside for-next branch. However for people who are not following our development branch but still want to verify the incoming fstests test case, we need this fix to be publicly avaiable. Newer minor changes will only be done inside the development branch.
v3: - Use a single btrfs_bio_ctrl::last_em_start Since the existing prev_em_start is using U64_MAX as the initial value to indicate no em hit yet, we can use the same logic, which saves another 8 bytes from btrfs_bio_ctrl.
- Update the reproducer Previously I failed to reproduce using a minimal workload. As regular read from od/md5sum always trigger readahead thus not hitting the @prev_em_start == NULL path.
Will send out a fstest case for it using the minimal reproducer. But it will still need a 16K/64K page sized system to reproduce.
Fix the problem by doing an block aligned write into the folio, so that the folio will be partially dirty and not go through the readahead path.
- Update the analyze This includes the trace events of the minimal reproducer, and mentioning of previous similar fixes and why they do not work for subpage cases.
- Update the CC tag Since it's only affecting bs < ps cases (for non-experimental builds), only need to fix kernels with subpage btrfs supports.
v2: - Only save extent_map::start/len to save memory for btrfs_bio_ctrl It's using on-stack memory which is very limited inside the kernel.
- Remove the commit message mentioning of clearing last saved em Since we're using em::start/len, there is no need to clear them. Either we hit the same em::start/len, meaning hitting the same extent map, or we hit a different em, which will have a different start/len. --- fs/btrfs/extent_io.c | 40 ++++++++++++++++++++++++++++++---------- 1 file changed, 30 insertions(+), 10 deletions(-)
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 426a6791a0b2..ca7174fa0240 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -131,6 +131,24 @@ struct btrfs_bio_ctrl { */ unsigned long submit_bitmap; struct readahead_control *ractl; + + /* + * The start offset of the last used extent map by a read operation. + * + * This is for proper compressed read merge. + * U64_MAX means we are starting the read and have made no progress yet. + * + * The current btrfs_bio_is_contig() only uses disk_bytenr as + * the condition to check if the read can be merged with previous + * bio, which is not correct. E.g. two file extents pointing to the + * same extent but with different offset. + * + * So here we need to do extra checks to only merge reads that are + * covered by the same extent map. + * Just extent_map::start will be enough, as they are unique + * inside the same inode. + */ + u64 last_em_start; };
/* @@ -965,7 +983,7 @@ static void btrfs_readahead_expand(struct readahead_control *ractl, * return 0 on success, otherwise return error */ static int btrfs_do_readpage(struct folio *folio, struct extent_map **em_cached, - struct btrfs_bio_ctrl *bio_ctrl, u64 *prev_em_start) + struct btrfs_bio_ctrl *bio_ctrl) { struct inode *inode = folio->mapping->host; struct btrfs_fs_info *fs_info = inode_to_fs_info(inode); @@ -1076,12 +1094,11 @@ static int btrfs_do_readpage(struct folio *folio, struct extent_map **em_cached, * non-optimal behavior (submitting 2 bios for the same extent). */ if (compress_type != BTRFS_COMPRESS_NONE && - prev_em_start && *prev_em_start != (u64)-1 && - *prev_em_start != em->start) + bio_ctrl->last_em_start != U64_MAX && + bio_ctrl->last_em_start != em->start) force_bio_submit = true;
- if (prev_em_start) - *prev_em_start = em->start; + bio_ctrl->last_em_start = em->start;
em_gen = em->generation; btrfs_free_extent_map(em); @@ -1296,12 +1313,15 @@ int btrfs_read_folio(struct file *file, struct folio *folio) const u64 start = folio_pos(folio); const u64 end = start + folio_size(folio) - 1; struct extent_state *cached_state = NULL; - struct btrfs_bio_ctrl bio_ctrl = { .opf = REQ_OP_READ }; + struct btrfs_bio_ctrl bio_ctrl = { + .opf = REQ_OP_READ, + .last_em_start = U64_MAX, + }; struct extent_map *em_cached = NULL; int ret;
lock_extents_for_read(inode, start, end, &cached_state); - ret = btrfs_do_readpage(folio, &em_cached, &bio_ctrl, NULL); + ret = btrfs_do_readpage(folio, &em_cached, &bio_ctrl); btrfs_unlock_extent(&inode->io_tree, start, end, &cached_state);
btrfs_free_extent_map(em_cached); @@ -2641,7 +2661,8 @@ void btrfs_readahead(struct readahead_control *rac) { struct btrfs_bio_ctrl bio_ctrl = { .opf = REQ_OP_READ | REQ_RAHEAD, - .ractl = rac + .ractl = rac, + .last_em_start = U64_MAX, }; struct folio *folio; struct btrfs_inode *inode = BTRFS_I(rac->mapping->host); @@ -2649,12 +2670,11 @@ void btrfs_readahead(struct readahead_control *rac) const u64 end = start + readahead_length(rac) - 1; struct extent_state *cached_state = NULL; struct extent_map *em_cached = NULL; - u64 prev_em_start = (u64)-1;
lock_extents_for_read(inode, start, end, &cached_state);
while ((folio = readahead_folio(rac)) != NULL) - btrfs_do_readpage(folio, &em_cached, &bio_ctrl, &prev_em_start); + btrfs_do_readpage(folio, &em_cached, &bio_ctrl);
btrfs_unlock_extent(&inode->io_tree, start, end, &cached_state);