GNU bug report logs - #76004
31.0.50; feature/igc crashed on Linux, wayland (with full gdb debug info)

Previous Next

Package: emacs;

Reported by: Eval EXEC <execvy <at> gmail.com>

Date: Sun, 2 Feb 2025 04:58:01 UTC

Severity: normal

Found in version 31.0.50

Done: Pip Cet <pipcet <at> protonmail.com>

Bug is archived. No further changes may be made.

Full log


View this message in rfc822 format

From: Pip Cet <pipcet <at> protonmail.com>
To: Eval EXEC <execvy <at> gmail.com>
Cc: 76004 <at> debbugs.gnu.org
Subject: bug#76004: 31.0.50; feature/igc crashed on Linux, wayland (with full gdb debug info)
Date: Sun, 02 Feb 2025 09:02:15 +0000
"Eval EXEC" <execvy <at> gmail.com> writes:

> Hello,
>
> I'm helping to test feature/igc branch, and I got a crash, with full gdb
> debug info.
>
> I build this emacs executable with:
> ```bash
> set -ex
>
> make extraclean
>
> BRANCH_NAME=$(git branch --show-current | sed 's/\//_/g')
> COMMIT_ID=$(git rev-parse HEAD)
> BUILD_DIR=${BRANCH_NAME}-${COMMIT_ID}-gdb
> INSTALL_PREFIX=$(realpath ../emacs-build/${BUILD_DIR})
>
> ./autogen.sh
> ./configure  CFLAGS='-g -ggdb -march=native' \
>   --prefix=${INSTALL_PREFIX} \
>   --with-imagemagick  \
>   --with-mps=yes \
>   --with-xwidgets \
>   --with-modules \
>   --with-cairo --with-cairo-xcb \
>   --without-compress-install \
>   --with-mailutils\
>   --enable-link-time-optimization \
>   --with-tree-sitter --with-xinput2 \
>   --with-dbus \
>   --with-native-compilation=aot \
>   --with-file-notification=inotify\
>   && make -j20 install
> ```
>
> What should I do to help you to investigate this crash?

Could you also print the address of lispsym:

p &lispsym

as well as provide the mapped memory ranges; if you have a corefile,

objdump -h <core>

from the shell prompt will work.  If you don't, run "gcore" from gdb to
generate one.

I'm very confused about

         binding = XIL(0x7f6c2ae81440)

Lisp symbols aren't represented as pointers to their structs, there's an
offset in there, by the address of &lispsym (so the 0 word corresponds
to &lispsym[0] rather than a NULL pointer).

On my system, I've never seen the address of &lispsym be so small that
binding would remain in the valid 47-bit address range once we add the
lispsym offset.  This would explain the segfault (though not why GDB
makes inaccurate claims).

Pip





This bug report was last modified 85 days ago.

Previous Next


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