Re: [PATCH v2] btrfs: lzo: reject inline extents without compressed data

在 2026/7/15 18:35, David Lee 写道:
> inline Btrfs LZO payload shorter than two LZO headers.
> 
> Fix the validation or lifetime rule at the vulnerable boundary so malformed
> or racing input cannot reach the faulting path.
> 
> A four-byte extent produced the following KASAN report on Linux 7.2-rc2:
> 
>    BUG: KASAN: slab-out-of-bounds in lzo_decompress+0x57d/0x700
>    Read of size 4 at addr ffff888006f2e644 by task btrfs_lzo_inlin/77
> 
>    Call Trace:
>     <TASK>
>     dump_stack_lvl+0x5b/0x70
>     print_report+0xd1/0x610
>     ? _raw_spin_lock_irqsave+0x78/0xc0
>     ? kasan_complete_mode_report_info+0x2a/0x200
>     kasan_report+0xe0/0x110
>     ? lzo_decompress+0x57d/0x700
>     ? lzo_decompress+0x57d/0x700
>     __asan_report_load_n_noabort+0x13/0x20
>     lzo_decompress+0x57d/0x700
>     ? lzo_decompress_bio+0x1e80/0x1e80
>     btrfs_decompress+0x140/0x1c0
>     uncompress_inline+0x147/0x1b0
>     btrfs_get_extent+0xb23/0x10a0
>     ? btrfs_rename2+0xb0/0xb0
>     ? _raw_spin_lock+0x72/0xb0
>     ? _raw_read_lock_irq+0x50/0x50
>     ? __kasan_check_read+0x15/0x20
>     ? btrfs_folio_test_uptodate+0xc8/0x140
>     btrfs_do_readpage.constprop.0+0x538/0x1ac0
>     ? btrfs_lookup_ordered_extent+0x2c0/0x2c0
>     ? xa_load+0xb0/0x1e0
>     btrfs_readahead+0x32f/0x5f0
>     ? btrfs_writepages+0x180/0x180
>     ? lruvec_stat_mod_folio+0x240/0x330
>     ? write_one_eb+0xb30/0xb30
>     read_pages+0x16f/0x850
>     ? __kasan_check_read+0x15/0x20
>     ? __folio_batch_add_and_move+0x134/0x1c0
>     ? ractl_alloc_folio.constprop.0+0x90/0x90
>     ? filemap_add_folio+0x1cd/0x4f0
>     ? __folio_alloc_noprof+0x69/0xa0
>     page_cache_ra_unbounded+0x296/0x490
>     do_page_cache_ra+0xd9/0x130
>     page_cache_sync_ra+0x3ee/0x6f0
>     filemap_get_pages+0x306/0x15c0
>     ? is_bpf_text_address+0x14/0x30
>     ? kernel_text_address+0x89/0xf0
>     ? filemap_add_folio+0x4f0/0x4f0
>     ? putname+0x78/0xc0
>     filemap_read+0x329/0xd00
>     ? kasan_save_track+0x18/0x40
>     ? __kasan_slab_free+0x44/0x60
>     ? putname+0x78/0xc0
>     ? entry_SYSCALL_64_after_hwframe+0x4b/0x53
>     ? filemap_get_pages+0x15c0/0x15c0
>     btrfs_file_read_iter+0x1f8/0x2b0
>     vfs_read+0x4ef/0x720
>     ? putname+0x78/0xc0
>     ? __kasan_slab_free+0x56/0x60
>     ? kernel_read+0x180/0x180
>     ? putname+0x78/0xc0
>     ? __kasan_check_read+0x15/0x20
>     ksys_read+0xf8/0x1d0
>     ? vfs_write+0xa30/0xa30
>     ? __x64_sys_open+0x180/0x180
>     __x64_sys_read+0x71/0xb0
>     ? __kasan_check_read+0x15/0x20
>     x64_sys_call+0x1ab0/0x1b70
>     do_syscall_64+0x61/0x470
>     entry_SYSCALL_64_after_hwframe+0x4b/0x53
>     </TASK>
> 
> Link: https://lore.kernel.org/all/[email protected]/
> Fixes: a6fa6fae40ec ("btrfs: Add lzo compression support")
> Assisted-by: Codex:gpt-5.5
> Signed-off-by: David Lee <[email protected]>

Since the patch is fixing a real problem, I have pushed it to for-next, 
but with commit message fully reworked.

Thanks,
Qu