Re: [BUG] xfs: undersized log triggers assertion in xfs_log_mount()

[铮]

Hi Carlos,

Thanks for the clarification. I am running syzkaller in isolated QEMU
guests for kernel testing, rather than using this configuration on a
production system.

I understand now that the fatal assertion is expected with
CONFIG_XFS_ASSERT_FATAL enabled. My intention was to check whether
this particular assertion on malformed filesystem input indicated
an issue worth investigating.

Your explanation answers my question. Thanks for your time.

Best regards,
Zheng Zheng

On Wed, Oct 28, 2026 at 1:32 AM Carlos Maiolino <cem@xxxxxxxxxx> wrote:
>
> On Wed, Oct 28, 2026 at 01:00:54AM +0800, 铮 wrote:
> > Hi Carlos,
> >
> > Executing the stored syzkaller testcases on Linux 7.3.0 ultimately triggers
> > an assertion in xfs_log_mount():
> >
> >   XFS (loop0): AAIEEE! Log failed size checks. Abort!
> >   XFS: Assertion failed: 0, file: fs/xfs/xfs_log.c, line: 597
> >   kernel BUG at fs/xfs/xfs_message.c:102!
> >   ...
> >   xfs_log_mount+0x4e3/0x590
> >   xfs_mountfs+0x10b9/0x2270
> >   xfs_fs_fill_super+0x15cc/0x20b0
> >   ...
> >   Kernel panic - not syncing: Fatal exception
> >
> > The log size check reports 516 blocks, below the required minimum of
> > 636 blocks.
> >
> > Is this an expected assertion with my XFS debug configuration, or should
> > this malformed image be rejected without triggering a kernel BUG?
>
> You have bug_on_assert set on your system, so yes, this is intended. If
> you don't want to panic() the machine there, you need to disable it on
> sysfs, but...
>
> > Questions
> > ---------
> > - Is triggering a fatal assertion for an undersized V5 log intentional
> >   with CONFIG_XFS_DEBUG=y and CONFIG_XFS_ASSERT_FATAL=y, or should this
> >   case return -EINVAL without triggering a BUG?
>
> The config you have enable explicitly says asserts are fatal. I'm not
> sure if I misunderstood you or you misunderstood the configuration
> option.
>
>
> > - If a diagnostic dump would help, what state should be printed before the
> >   assertion so the malformed image/log geometry can be identified?
>
> You are running a kernel configured with debug configuration that are
> meant to crash the system when some issue is found. Those are
> configuration settings used for development and debug purposes, not to
> use on a production or semi-production system.
>
> Cheers,
> Carlos