Skip to content

Commit

Permalink
btrfs: grab correct extent map for subpage compressed extent read
Browse files Browse the repository at this point in the history
[BUG]
When subpage compressed read write support is enabled, btrfs/038 always
fails with EIO.

A simplified script can easily trigger the problem:

  mkfs.btrfs -f -s 4k $dev
  mount $dev $mnt -o compress=lzo

  xfs_io -f -c "truncate 118811" $mnt/foo
  xfs_io -c "pwrite -S 0x0d -b 39987 92267 39987" $mnt/foo > /dev/null

  sync
  btrfs subvolume snapshot -r $mnt $mnt/mysnap1

  xfs_io -c "pwrite -S 0x3e -b 80000 200000 80000" $mnt/foo > /dev/null
  sync

  xfs_io -c "pwrite -S 0xdc -b 10000 250000 10000" $mnt/foo > /dev/null
  xfs_io -c "pwrite -S 0xff -b 10000 300000 10000" $mnt/foo > /dev/null

  sync
  btrfs subvolume snapshot -r $mnt $mnt/mysnap2

  cat $mnt/mysnap2/foo
  # Above cat will fail due to EIO

[CAUSE]
The problem is in btrfs_submit_compressed_read().

When it tries to grab the extent map of the read range, it uses the
following call:

	em = lookup_extent_mapping(em_tree,
  				   page_offset(bio_first_page_all(bio)),
				   fs_info->sectorsize);

The problem is in the page_offset(bio_first_page_all(bio)) part.

The offending inode has the following file extent layout

        item 10 key (257 EXTENT_DATA 131072) itemoff 15639 itemsize 53
                generation 8 type 1 (regular)
                extent data disk byte 13680640 nr 4096
                extent data offset 0 nr 4096 ram 4096
                extent compression 0 (none)
        item 11 key (257 EXTENT_DATA 135168) itemoff 15586 itemsize 53
                generation 8 type 1 (regular)
                extent data disk byte 0 nr 0
        item 12 key (257 EXTENT_DATA 196608) itemoff 15533 itemsize 53
                generation 8 type 1 (regular)
                extent data disk byte 13676544 nr 4096
                extent data offset 0 nr 53248 ram 86016
                extent compression 2 (lzo)

And the bio passed in has the following parameters:

page_offset(bio_first_page_all(bio))	= 131072
bio_first_bvec_all(bio)->bv_offset	= 65536

If we use page_offset(bio_first_page_all(bio) without adding bv_offset,
we will get an extent map for file offset 131072, not 196608.

This means we read uncompressed data from disk, and later decompression
will definitely fail.

[FIX]
Take bv_offset into consideration when trying to grab an extent map.

And add an ASSERT() to ensure we're really getting a compressed extent.

Thankfully this won't affect anything but subpage, thus we only need to
ensure this patch get merged before we enabled basic subpage support.

Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
  • Loading branch information
adam900710 authored and kdave committed Aug 23, 2021
1 parent ca62e85 commit 557023e
Showing 1 changed file with 6 additions and 3 deletions.
9 changes: 6 additions & 3 deletions fs/btrfs/compression.c
Original file line number Diff line number Diff line change
Expand Up @@ -682,6 +682,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
struct page *page;
struct bio *comp_bio;
u64 cur_disk_byte = bio->bi_iter.bi_sector << 9;
u64 file_offset;
u64 em_len;
u64 em_start;
struct extent_map *em;
Expand All @@ -691,15 +692,17 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,

em_tree = &BTRFS_I(inode)->extent_tree;

file_offset = bio_first_bvec_all(bio)->bv_offset +
page_offset(bio_first_page_all(bio));

/* we need the actual starting offset of this extent in the file */
read_lock(&em_tree->lock);
em = lookup_extent_mapping(em_tree,
page_offset(bio_first_page_all(bio)),
fs_info->sectorsize);
em = lookup_extent_mapping(em_tree, file_offset, fs_info->sectorsize);
read_unlock(&em_tree->lock);
if (!em)
return BLK_STS_IOERR;

ASSERT(em->compress_type != BTRFS_COMPRESS_NONE);
compressed_len = em->block_len;
cb = kmalloc(compressed_bio_size(fs_info, compressed_len), GFP_NOFS);
if (!cb)
Expand Down

0 comments on commit 557023e

Please sign in to comment.