Re: [mainline] powerpc/TM: Unexpected TM Bad Thing during core dump on POWER9 (7.1‑rc1)

[Venkat Rao Bagalkote <venkat88@linux.ibm.com>]

On 28/04/26 5:48 pm, Venkat Rao Bagalkote wrote:
> Greetings!!
>
> IBM CI has reported a kernel crash while running 
> selftests/powerpc/signal on a POWER9 pSeries system.
>
> I attempted to reproduce this issue manually, but was not successful 
> so far.

I’m now able to reproduce this issue reliably and consistently. The 
reproduction steps are simple:

Run selftests/powerpc/signal


Regards,

Venkat.

>
> Below are the details of the crash as reported by CI.
>
> Crash Details:
>
> [ 9798.880148] Unexpected TM Bad Thing exception at c00000000000dbac
>                (msr 0x8000000302a03031) tm_scratch=800000010280b033
> [ 9798.880160] Oops: Unrecoverable exception, sig: 6 [#1]
> [ 9798.880165] LE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=8192 NUMA pSeries
> [ 9798.880173] Modules linked in: nvram(E) rpadlpar_io(E) rpaphp(E) ...
> [ 9798.880233] CPU: 8 UID: 0 PID: 1039530 Comm: sigfuz
>                Tainted: G            E       7.1.0-rc1 #1 PREEMPT
> [ 9798.880245] Hardware name: IBM,8375-42A POWER9 (architected)
>                hv:phyp pSeries
> [ 9798.880251] NIP:  c00000000000dbac LR: 0000000010001e58
> [ 9798.880262] MSR:  8000000302a03031 <SF,VEC,VSX,FP,ME,IR,DR,LE,TM[SE]>
>
>
> Call Trace:
>
>
> NIP [c00000000000dbac] interrupt_return_srr_kernel+0x15c/0x18c
> Call Trace:
>  tm_reclaim_thread
>  flush_tmregs_to_thread
>  vsr_get
>  regset_get_alloc
>  fill_thread_core_info.isra.0
>  fill_note_info
>  elf_core_dump
>  coredump_write
>  do_coredump
>  vfs_coredump
>  get_signal
>  do_signal
>  do_notify_resume
>  interrupt_exit_user_prepare_main
>  interrupt_exit_user_prepare
>  interrupt_return_srr_user
>
>
> Please let me know if further details are required.
>
>
> If you happen to fix this issue, please add below tag.
>
> Reported-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
>
>
>
> Regards,
>
> Venkat.
>
>