Re: ready for the GDB 17.2 corrective release?

[Tom de Vries <tdevries@suse.de>]

On 4/20/26 8:32 PM, Joel Brobecker wrote:
> Hi Tom,
> 
>> I'd like to nominate PR31449 [1] and PR33794 [2].  They're both crashes that
>> happen when querying the user about enabling debuginfod while enabling TUI.
>>
>> I've submitted low-risk patches for both PRs (
>> https://sourceware.org/pipermail/gdb-patches/2026-April/226551.html ).
>>
>> I'm not sure if the PRs are severe enough to block the release for: crashes
>> are serious, but the likelihood that such a crash will trigger is not high,
>> I think, which is probably why we have lived with it, probably since the
>> query was introduced in 2021.
>>
>> But the fact that low-risk patches are available might be reason enough to
>> consider including them.
> 
> Thanks for pointing those two PRs out. Let's start by marking them as
> targetting 17.2 for the time being, so we don't forget about them.
> 
> How soon do you anticipate you'll be able to have those changes pushed?

Well, they're waiting for review.

> There are really two things to consider:
> 
>    * Are the patches suitable for the 17 branch? If they are low risk
>      and prevent a crash, that seems like an easy "yes".
Yes, they are.

>    * Are the issues problematic-enough that we would want to wait extra
>      for them. The answer depends on how frequently we think people
>      are going to hit the issue.

Right.

> 
>      If this issue is a regression of release 17 compared to release 16,
>      this changes the perspective as well.
> 

It's not a regression AFAIK.

> I have no strong opinion either way, FWIW.
> 

Ack, thanks.
  -Tom