GNU bug report logs - #75846
31.0.50; feature/igc pgtk build crash on Linux/wayland (kde plasma6)

Previous Next

Full log

Message #23 received at 75846 <at> debbugs.gnu.org (full text, mbox):

From: Pip Cet <pipcet <at> protonmail.com>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: execvy <at> gmail.com, 75846 <at> debbugs.gnu.org
Subject: Re: bug#75846: 31.0.50;
 feature/igc pgtk build crash on Linux/wayland (kde plasma6)
Date: Sun, 26 Jan 2025 10:08:55 +0000

"Eli Zaretskii" <eliz <at> gnu.org> writes:

>> Cc: 75846 <at> debbugs.gnu.org
>> Date: Sun, 26 Jan 2025 03:02:35 +0000
>> From:  Pip Cet via "Bug reports for GNU Emacs,
>>  the Swiss army knife of text editors" <bug-gnu-emacs <at> gnu.org>
>>
>> > Configured using:
>> >  'configure 'CFLAGS=-O3 -march=native'
>>
>> I tried configuring like that, and my CFLAGS didn't include
>> -fno-omit-frame-pointer.
>
> I wonder if we want to support -O3 compilations at all, let alone

I'm not sure what you are suggesting.  We don't enable -O3 by default or
handle it in any special way.  I don't see how we could do so without
losing valuable debug information for both us and the GCC project.

> those with -march=native.

I don't know whether -march=native has been fixed to produce the
compiler options necessary for reproducing the build in the object files
or elsewhere.  If it has, we should include such information in
report-emacs-bug, at the risk of revealing more about our user's social
status than they expect.

> GCC is likely to produce buggy code with these options.

Not my experience.  It's a good way to get an early heads-up on proposed
GCC optimizations which break our (undocumented) conservative GC
assumptions.

My proposal is not to change anything on feature/igc.

Merely because that is relevant to that question, which I fear will have
to be discussed in detail in other forums:

I've currently checked out the commit before 67e60210577 and haven't
even decided whether I'll even *pull* feature/igc changes starting with
67e60210577, nevermind whether I want to *push* any further changes to
feature/igc after that.

Depending on that decision of mine, you might want to ignore anything I
recommend should or should not happen on that branch; that wouldn't
change much, but it would save some time.

Pip

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.