Re: [PATCH v3 2/5] [PATCH 2/5] gdb: Improve SIGSEGV diagnostics for POE faults

Luis <[email protected]> Sat, 25 Jul 2026 08:39:22 +0100
Newsgroups gmane.comp.gdb.patches
Message-ID <[email protected]>
On 23/07/2026 10:03, Srinath Parvathaneni wrote:
>  >
>  >>>> +  /* For POE SEGSEGV, show additional information.  */
>  >>>
>  >>> Is this really supposed to be POE SEGSEGV or is it a typo?
>  >>>
>  >>
>  >> For now, in my knowledge, AARCH64_SEGV_PKUERR is only used for POE.
>  >> However, it will also be used soon for another upcoming feature.
>  >> I would recommend to either simply eliminate the comment, or make it 
> more clear like
>  >>
>  >> /* For memory permission faults, e.g. due to a permission overlay
>  >>     restriction when feature POE is enabled.  */
>  >>
>  >> Matthieu
>  >
>  >Sorry, I may not have been clear, but I was referring specifically to
>  >the SEGSEGV. Should it be SIGSEGV instead?
> 
> Yes Luis, that's a typo. It should be SIGSEGV instead of SEGSEGV, thanks for
> spotting it, I will update in the re-spin.
> 

Got it. Thanks.

> Regarding the comment itself, while SIGSEGV already implies a memory fault,
> AARCH64_SEGV_PKUERR is currently only generated for Permission Overlay
> violations. I think something like following would be more accurate 
> description.
> 
> /* Additional information for SIGSEGV caused by a permission overlay 
> violation.  */

Yes, I think that reads a bit better.