Re: xfs: XLOG_WAS_CONT_TRANS on a transaction with no buffered regions reads ri_buf[-1] - null deref Oops during mount log recovery

[Dave Chinner]

On Sat, Nov 07, 2026 at 10:00:00AM +0800, Guikun Yang wrote:
> Hi,
> 
> Mounting an XFS volume whose log record after the transaction header
> carries an operation header with the XLOG_WAS_CONT_TRANS flag panics
> the kernel during log recovery - on a stock kernel, no sanitizer
> configuration needed:
> 
>   BUG: unable to handle page fault for address: fffffffffffffff0
>   #PF: supervisor read access in kernel mode
>   Oops: Oops: 0000 [#1] SMP KASAN PTI
>   RIP: xlog_recover_add_to_cont_trans+0xf9/0x460
>    xlog_recovery_process_trans
>    xlog_recover_process_ophdr
>    xlog_recover_process_data
>    xlog_recover_process
>    xlog_do_recovery_pass
>   Kernel panic - not syncing: Fatal exception

Please stop with all these bug reports about things we -already know
are not validated- and -are already being fixed-.

https://lore.kernel.org/linux-xfs/20261013060204.3138690-1-dgc@xxxxxxxxxx/

Please run all your tests and analysis on the log verification
series that addresses all these problems, and find holes in that
series that still need to be plugged.

Do *not waste our time* reporting known problems that we are
actively fixing. Make yourself and your LLM tokens useful by
creating fstests reporducers for all the issues and then -testing
the proposed fixes- and giving feedback on the fixes that already
exist.

-Dave.
-- 
Dave Chinner
dgc@xxxxxxxxxx