bug#74224: [PATCH][scratch/igc] pgtk: fix crashing bug in atimer used for checking scaling

["Gerd Möllmann" <gerd.moellmann@gmail.com>]

Jeff Walsh <fejfighter@gmail.com> writes:

>> LGTM. Can you commit that yourself to the branch?
>
> I don't have a savannah account for emacs, but I have just registered and will chase up approvals.
>
> timers are cleaned up by cancel_atimer(), then they get put on a free list (`free_atimers` on line 132 in the snippet).
> It appears that we rely on the OS to cleanup at emacs shutdown.

Very good, thanks!

I'll commit that for you, then, and close this bug when done.

>
> On Wed, Nov 6, 2024 at 9:36 PM Gerd Möllmann <gerd.moellmann@gmail.com> wrote:
>
>  Jeff Walsh <fejfighter@gmail.com> writes:
>
>  >  #ifdef WINDOWSNT
>  >  #define raise(s) w32_raise(s)
>  > @@ -132,7 +133,13 @@ start_atimer (enum atimer_type type, struct timespec timestamp,
>  >        free_atimers = t->next;
>  >      }
>  >    else
>  > -    t = xmalloc (sizeof *t);
>  > +    {
>  > +#ifdef HAVE_MPS
>  > +      t = igc_xzalloc_ambig (sizeof *t);
>  > +#else
>  > +      t = xmalloc (sizeof *t);
>  > +#endif
>  > +    }
>  >  
>  >    /* Fill the atimer structure.  */
>  >    memset (t, 0, sizeof *t);
>
>  On second thought, and I don't know if it's relevant, do we need to
>  igc_xfree that?