You are not logged in.
Which process should I attach gdb to?
Offline
The one that crashes?
gnome-shell
Online
And which results should I report back?
Offline
The backtrace - hopefully.
You may wanna read up https://wiki.archlinux.org/title/Debugg … _the_trace first.
Online
I read the instructions, but I'm not sure what I'm doing wrong here.
I run gdb /usr/bin/gnome-shell, and then execute "run --replace" in gdb. But this causes my DE to crash, so I can't write the backtraces to gdb.txt
I also tried running gdb in a separate tty, but I can't replicate it there, because "run --replace" yields "Failed to configure, unsupported session type" error.
Last edited by GeneralPoxter (2023-08-19 23:30:41)
Offline
Okay I followed the instructions on https://wiki.gnome.org/Projects/GnomeShell/Debugging, and I was able to repeat the crash twice and get screenlogs with backtraces for both.
The first one I did here (it's a little messy though since I didn't really know what I was doing): https://drive.google.com/file/d/1rIQqrm … sp=sharing
The second time is here (this one should be cleaner): https://drive.google.com/file/d/1WUKgln … sp=sharing
Last edited by GeneralPoxter (2023-08-20 00:21:05)
Offline
I filed the following bug report, containing a similar backtrace a while back:
https://gitlab.gnome.org/GNOME/mutter/-/issues/2862
Still occurring on every logout.
This crasher began for me long before the systemd 254 upgrade.
[2023-06-05T09:19:42-0400] [ALPM] upgraded systemd (253.4-1 -> 253.5-1)
[2023-06-20T10:19:42-0400] [ALPM] upgraded systemd (253.5-1 -> 253.5-2)
[2023-07-12T09:36:29-0400] [ALPM] upgraded systemd (253.5-2 -> 253.6-2)
[2023-07-22T10:49:19-0400] [ALPM] upgraded systemd (253.6-2 -> 253.7-1)
[2023-08-01T18:07:10-0400] [ALPM] upgraded systemd (253.7-1 -> 254-1)
[2023-08-10T11:27:13-0400] [ALPM] upgraded systemd (254-1 -> 254.1-1)
EDIT: I would not rely on the Alt-F2...'r' method to restart the shell in recent years.
Disabling all extensions, logging out, and logging in ensures a reproducible scenario.
Last edited by tekstryder (2023-08-20 01:47:43)
Offline
Ah yes, it does appear that my system segfaults whenever I logout as well.
Do you think it's also related to this recent issue: https://gitlab.gnome.org/GNOME/mutter/-/issues/2977 ?
Last edited by GeneralPoxter (2023-08-20 01:53:21)
Offline
559 Thread 1 (Thread 0x7ffff3150d80 (LWP 3863) "gnome-shell"):
560 #0 0x00007ffff6cfef37 in _XSend () at /usr/lib/libX11.so.6
561 #1 0x00007ffff6cf77d9 in XQueryExtension () at /usr/lib/libX11.so.6
562 #2 0x00007ffff65f6e17 in () at /usr/lib/libXext.so.6
563 #3 0x00007ffff65fd088 in XSyncDestroyFence () at /usr/lib/libXext.so.6
564 #4 0x00007ffff73061b5 in () at /usr/lib/libmutter-12.so.0
565 #5 0x00007ffff7306385 in () at /usr/lib/libmutter-12.so.0
570 #10 0x00007ffff690ff5e in () at /usr/lib/libffi.so.8
571 #11 0x00007ffff6912b73 in ffi_call () at /usr/lib/libffi.so.8
572 #12 0x00007ffff77cd8c9 in () at /usr/lib/libgjs.so.0
573 #13 0x00007ffff77ce299 in () at /usr/lib/libgjs.so.0
574 #14 0x00007ffff5aaf506 in () at /usr/lib/libmozjs-102.so
575 #15 0x00007ffff5aa05f5 in () at /usr/lib/libmozjs-102.so
576 #16 0x00007ffff6126fdb in JS_CallFunctionValue(JSContext*, JS::Handle<JSObject*>, JS::Handle<JS::Value>, JS::HandleValueArray const&, JS::MutableHandle<JS::Value>) () at /usr/lib/libmozjs-102.so
577 #17 0x00007ffff77c2f69 in () at /usr/lib/libgjs.so.0
578 #18 0x00007ffff77fb7dc in () at /usr/lib/libgjs.so.0
579 #19 0x00007ffff7f4a6c0 in g_closure_invoke () at /usr/lib/libgobject-2.0.so.0
580 #20 0x00007ffff7f78938 in () at /usr/lib/libgobject-2.0.so.0
581 #21 0x00007ffff7f694cc in g_signal_emit_valist () at /usr/lib/libgobject-2.0.so.0
582 #22 0x00007ffff7f69d34 in g_signal_emit () at /usr/lib/libgobject-2.0.so.0
583 #23 0x00007ffff72dcb7f in () at /usr/lib/libmutter-12.so.0
584 #24 0x00007ffff7ad2644 in () at /usr/lib/libgio-2.0.so.0
585 #25 0x00007ffff7ad66ad in () at /usr/lib/libgio-2.0.so.0
586 #26 0x00007ffff7a769da in () at /usr/lib/libgio-2.0.so.0
587 #27 0x00007ffff7a6cd17 in () at /usr/lib/libgio-2.0.so.0
588 #28 0x00007ffff7ad2644 in () at /usr/lib/libgio-2.0.so.0
589 #29 0x00007ffff7ad66ad in () at /usr/lib/libgio-2.0.so.0
590 #30 0x00007ffff7a6d240 in () at /usr/lib/libgio-2.0.so.0
591 #31 0x00007ffff7a9f71f in () at /usr/lib/libgio-2.0.so.0
592 #32 0x00007ffff7ad2644 in () at /usr/lib/libgio-2.0.so.0
593 #33 0x00007ffff7ad66ad in () at /usr/lib/libgio-2.0.so.0
594 #34 0x00007ffff7aa5a65 in () at /usr/lib/libgio-2.0.so.0
595 #35 0x00007ffff7aa5bed in () at /usr/lib/libgio-2.0.so.0
596 #36 0x00007ffff7939a31 in g_main_context_dispatch () at /usr/lib/libglib-2.0.so.0
597 #37 0x00007ffff7996cc9 in () at /usr/lib/libglib-2.0.so.0
598 #38 0x00007ffff7938fef in g_main_loop_run () at /usr/lib/libglib-2.0.so.0
599 #39 0x00007ffff72cddcb in meta_context_run_main_loop () at /usr/lib/libmutter-12.so.0
600 #40 0x0000555555557526 in ()
601 #41 0x00007ffff6e27cd0 in () at /usr/lib/libc.so.6
602 #42 0x00007ffff6e27d8a in __libc_start_main () at /usr/lib/libc.so.6
603 #43 0x0000555555557835 in ()
https://gitlab.gnome.org/GNOME/mutter/-/issues/2977 goes down in libffi but tekstryder's https://gitlab.gnome.org/GNOME/mutter/-/issues/2862 is yours.
Edit: though the context or 2977 fits and the trace segment is similar. The stack might be somewhat corrupted at this point, but tekstryder apparently can pin it to to libx11 update.
Edit #2: you're both on nvidia - can you test and reproduce this w/ nouveau (and the modesetting driver)?
Last edited by seth (2023-08-20 08:14:17)
Online
Looks like there may be a fix for this crash issue around the corner.
@GeneralPoxter's bug report got duped to mine.
Mine was then duped to:
https://gitlab.gnome.org/GNOME/mutter/-/issues/2852
And an MR is pending:
https://gitlab.gnome.org/GNOME/mutter/- … uests/3202
Offline
MR !3202 didn't make it into 44.4, so I built mutter 44.4 patched with only this single MR.
I can confirm this resolves the crash at logout/restart.
Could be a while before this lands in Arch due to the timing of the 45 release and 44.5 maintenance releases will likely coincide. Hopefully 44.5 gets packaged while 45 sits in unstable and/or testing repos.
Offline
Ah yes, it does appear that my system segfaults whenever I logout as well.
Do you think it's also related to this recent issue: https://gitlab.gnome.org/GNOME/mutter/-/issues/2977 ?
@GeneralPoxter.. mutter 44.5 landed with the logout crasher fix:
https://gitlab.gnome.org/GNOME/mutter/- … dd76df1a4f
If that resolves your issue, could you go ahead and mark this thread [SOLVED]?
Offline
Yep, it works now without segfault
Offline