bug#74966: 31.0.50; Crash report (using igc on macOS)

[Gerd Möllmann]

Pip Cet <pipcet@protonmail.com> writes:

> Gerd Möllmann <gerd.moellmann@gmail.com> writes:
>
>> The same happens with master and --enable-checking, so it's not igc.
>
> So... something's clobbering our subrs?  Any chance we can set a
> watchpoint and find out who it is, or whether the initialization is
> already using bad data?

I have no idea what could be happening ATM. An interesting details might
be that ns-win is in the .pdmp, judging from loadup.el.

I've stopped there for today because it was already a bit late for me. I
think I'll try tomorrow first if this depends on native compilation, or
if it also happens without. Just to be sure.

>> /Users/gerd/emacs/savannah/master % src/emacs -Q
>> 2024-12-19 16:20:25.163 emacs[83702:3766262] +[IMKClient subclass]: chose 
>> IMKClient_Modern
>> 2024-12-19 16:20:25.163 emacs[83702:3766262] +[IMKInputSession subclass]: 
>> chose IMKInputSession_Modern
>>
>> ./lisp.h:2000: Emacs fatal error: assertion failed: 0 <= idx && idx < 
>> gc_asize (array)
>> Fatal error 6: Aborted
>
> I assume that's in the same place?

I haven't checked, just assumed that if the same action leads to the
same result, it's pretty likely to have the same cause.

I'll let you know when I find something out tomorrow.