bug#72158: 29.4; Core Dump

[Eli Zaretskii]

> From: David Wen Riccardi-Zhu <dwrz@dwrz.net>
> Date: Wed, 17 Jul 2024 13:05:42 +0000
> 
> 
> I'm getting frequent core dumps while using the lsp-mode package:
> 
> Jul 17 12:59:01 main.dwrz.net emacs[38432]: Backtrace:
> Jul 17 12:59:01 main.dwrz.net emacs[38432]: 
> /usr/bin/emacs(+0x19155b)[0x5cbc3bc7755b]
> Jul 17 12:59:01 main.dwrz.net emacs[38432]: 
> /usr/bin/emacs(+0x2495c)[0x5cbc3bb0a95c]
> Jul 17 12:59:01 main.dwrz.net emacs[38432]: 
> /usr/bin/emacs(+0x25767)[0x5cbc3bb0b767]
> Jul 17 12:59:01 main.dwrz.net emacs[38432]: 
> /usr/bin/emacs(+0x2e18e5)[0x5cbc3bdc78e5]
> Jul 17 12:59:01 main.dwrz.net emacs[38432]: 
> /usr/lib/libc.so.6(+0x3cae0)[0x710025ae8ae0]

Thanks, but these addresses can only be interpreted on your system.
Can you run Emacs under GDB, and when it crashes, produce a backtrace
using the "thread apply all bt" GDB command, and post the results
here?

We had already similar crash reports, on Fedora and Arch Linux, and in
both cases the problem went away when Emacs was rebuilt from sources.
So my guess is that something in the non-default compilation switches
or some other aspect of the configuration and build triggers these
segfaults.  If you can build Emacs from sources, and if that build
doesn't crash, then this is something to take up with the folks who
maintain the downstream distro you are using.