GNU bug report logs -
#61960
30.0.50; Unexec build reliably crashes during loadup
Previous Next
Reported by: Eli Zaretskii <eliz <at> gnu.org>
Date: Sat, 4 Mar 2023 14:56:02 UTC
Severity: normal
Found in version 30.0.50
Done: Stefan Kangas <stefankangas <at> gmail.com>
Bug is archived. No further changes may be made.
Full log
Message #23 received at 61960 <at> debbugs.gnu.org (full text, mbox):
On Sun, 2023-03-05 at 00:56 +0300, Konstantin Kharlamov wrote:
> On Sat, 2023-03-04 at 21:45 +0000, Andrea Corallo wrote:
> > Konstantin Kharlamov <hi-angel <at> yandex.ru> writes:
> >
> > > Oh, I am sorry, I posted the configuration line from the wrong emacs
> > > build.
> > > It's
> > > supposed to be:
> > >
> > > --prefix=/usr --sysconfdir=/etc --libexecdir=/usr/lib --localstatedir=/var
> > > -
> > > -
> > > mandir=/usr/share/man --with-gameuser=:games --with-modules --without-
> > > libotf
> > > --
> > > without-m17n-flt --without-gconf --with-native-compilation=yes --with-
> > > xinput2 --
> > > with-x-toolkit=gtk3 --without-xaw3d --with-sound=no --with-tree-sitter --
> > > with-
> > > unexec --without-gpm --without-compress-install 'CFLAGS=-O0 -g3'
> >
> > Hi Konstantin,
> >
> > maybe the crash you see is not related but native-compilation is not
> > supposed to work with unexec builds.
> >
> > I think we should really add a configure time error for this. Eli could
> > this change go to emacs-29?
>
> emacs-29 haven't got the BLOCK_ALIGN change, so is unaffected.
Ah, sorry, I failed to parse your text correctly, because I'm in context of the
debugging session :) Yeah, if native compilation isn't supposed to work with
unexec(), it might be a good idea to disable that, sure.
> I should note though that I'm not the reporter :)
>
> ---------------
>
> Regarding my current findings: apparently the `unexec` has always been broken.
> I built it with sanitizer and found out that the variable `bss_size_growth`
> when doing the dump has too big size. The only difference between "before" and
> "after" the BLOCK_ALIGN change is that the difference "after" became quite
> large. It was just 440 bytes before, and became 31494584 bytes after.
>
> However, when built with sanitizer, sanitizer catches the problem in both
> cases, so there's that.
This bug report was last modified 94 days ago.
Previous Next
GNU bug tracking system
Copyright (C) 1999 Darren O. Benham,
1997,2003 nCipher Corporation Ltd,
1994-97 Ian Jackson.