Re: BUG: kernel NULL pointer dereference, address: 000000000000002c - RIP: 0010:alloc_extent_buffer

[Qu Wenruo]

在 2024/5/30 23:26, Naresh Kamboju 写道:
> The following kernel BUG: and kernel crash noticed while running xfstests btfs
> filesystem testing on qemu-x86_64 with loop back.
>
> Steps to reproduce link provided.
>
> Test details:
> ----
>    Tests:  xfstests-btrfs  btrfs/232
>    SKIP_INSTALL='true'
>    TEST_DEV='/dev/loop0'
>    SCRATCH_DEV='/dev/loop1'
>    TEST_DIR='/mnt/test'
>    SCRATCH_DIR='/mnt/scratch'
>    FILESYSTEM='btrfs'
>    T_SIZE='5G'
>    S_SIZE='8G'

That's a pretty common setup, and unable to reproduce it here.

> Reported-by: Linux Kernel Functional Testing <lkft@xxxxxxxxxx>
>
> Test log:
> -------
>   <12>[ 6457.571628] run fstests btrfs/232 at 2024-05-29 16:31:29
>
> <6>[ 6464.685165] BTRFS: device fsid
> 50147eec-0761-4d75-8e77-df9c50ac385e devid 1 transid 6 /dev/loop1
> (7:1) scanned by mount (152729)
> <6>[ 6464.715051] BTRFS info (device loop1): first mount of filesystem
> 50147eec-0761-4d75-8e77-df9c50ac385e
> <6>[ 6464.719266] BTRFS info (device loop1): using crc32c
> (crc32c-generic) checksum algorithm
> <6>[ 6464.724996] BTRFS info (device loop1): using free-space-tree
> <6>[ 6464.789867] BTRFS info (device loop1): checking UUID tree
> <6>[ 6499.694309] BTRFS info (device loop1): qgroup scan completed
> (inconsistency flag cleared)
> <6>[ 6499.766172] BTRFS info (device loop1): qgroup scan completed
> (inconsistency flag cleared)
> <1>[ 6572.421678] BUG: kernel NULL pointer dereference, address:
> 000000000000002c
> <1>[ 6572.423036] #PF: supervisor read access in kernel mode
> <1>[ 6572.423070] #PF: error_code(0x0000) - not-present page
> <6>[ 6572.423143] PGD 0 P4D 0
> <4>[ 6572.424555] Oops: Oops: 0000 [#1] PREEMPT SMP PTI
> <4>[ 6572.424814] CPU: 0 PID: 152772 Comm: fsstress Not tainted
> 6.10.0-rc1-next-20240529 #1
> <4>[ 6572.424946] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009),
> BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014
> <4>[ 6572.425062] RIP: 0010:alloc_extent_buffer+0x253/0x820

Any line number and code context for it?

It may be a clue for the recent bug of various bad page status.

Thanks,
Qu
> --
> Linaro LKFT
> https://lkft.linaro.org