Bug 2247634 - Wayland dumps core on wake on Ryzen Framework 13
Summary: Wayland dumps core on wake on Ryzen Framework 13
Keywords:
Status: NEW
Alias: None
Product: Fedora
Classification: Fedora
Component: kwin
Version: 39
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Rex Dieter
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2023-11-02 13:26 UTC by Jed
Modified: 2023-11-03 02:43 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed:
Type: Bug
Embargoed:


Attachments (Terms of Use)
Core dump file (14.00 MB, application/octet-stream)
2023-11-03 02:43 UTC, Jed
no flags Details

Description Jed 2023-11-02 13:26:26 UTC
Description of problem:

I kept noticing that after my Ryzen Framework 13 would be asleep and then I’d log in, all my programs would be closed. I thought I was going crazy, but then I checked journalctl and I see that I’m experiencing core dumps on wake

...
Nov 01 17:33:34 fedora kactivitymanagerd[71673]: The Wayland connection broke. Did the Wayland compositor die?
Nov 01 17:33:34 fedora systemd[1]: systemd-coredump: Deactivated successfully.
Nov 01 17:33:33 fedora systemd-coredump[76013]: [🡕] Process 71458 (kwin_wayland) of user  1000 dumped core.
  Module libpipewire-module-session-manager.so from rpm pipewire-0.3.80-1.fc39.
  Module libpipewire-module-metadata.so from rpm pipewire-0.3.80-1.fc39.
  Module libpipewire-module-adapter.so from rpm pipewire-0.3.80-1.fc39.
  Module libpipewire-module-client-device.so from rpm pipewire-0.3.80-1.fc39.x86_64
  ...
  Stack trace of thread 71458:
  #0  0x00007f3444fbea99 _ZN4KWin12DrmGammaRampC2EPNS_7DrmCrtcERKSt10shared_ptrINS_19ColorTransformationEE (libkwin.so.5 + 0x3bea99)
  ...
Nov 01 17:33:33 fedora audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=systemd-coredump@1-76012-0 comm="systemd" exe="/usr/lib/systemd/systemd" host>
Nov 01 17:33:33 fedora systemd[1]: Started systemd-coredump - Process Core Dump (PID 76012/UID 0).
Nov 01 17:33:33 fedora audit: BPF prog-id=147 op=LOAD
Nov 01 17:33:33 fedora audit: BPF prog-id=146 op=LOAD
Nov 01 17:33:33 fedora audit: BPF prog-id=145 op=LOAD
Nov 01 17:33:33 fedora kernel: Code: 55 4c 8d 6f 08 41 54 53 48 89 d3 48 83 ec 78 48 89 bd 70 ff ff ff 48 89 b5 68 ff ff ff 64 48 8b 04 25 28 00 00 00 48 89 45 c8 <48> 8b 46 20 48 89 07 48 89 f7 e8 b8 d1 fd ff >
Nov 01 17:33:33 fedora kernel: kwin_wayland[71458]: segfault at 20 ip 00007f3444fbea99 sp 00007fffeaaec3d0 error 4 in libkwin.so.5.27.8[7f3444d70000+357000] likely on CPU 12 (core 6, socket 0)
Nov 01 17:33:33 fedora audit[71458]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=2 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=71458 comm="kwin_wayland" exe="/usr/bin/kwin_wayland" sig=11 res=1
...

Version-Release number of selected component (if applicable):

I’m on Fedora 39 KDE spin (Plasma v. 5.27.8, KDE Frameworks v. 5.111.0, kernel 6.5.8-300.fc39.x86_6). I’m also running with the amdgpu.sg_display=0 added as a kernel param to fix the white screen flashing issues.

How reproducible: Very


Steps to Reproduce:
1. Open a bunch of programs
2. Allow computer to go into sleep mode
3. Wake and re-log in

Actual results:

All programs are closed. Journalctl shows: (kwin_wayland) of user  1000 dumped core.

Expected results:
All programs are still running, no core dump

Comment 1 Olivier Fourdan 2023-11-02 17:31:49 UTC
Wayland is just a protocol specification in XML and the low-level C library that deals with the protocol.

This issue here is a bug not with the protocol itself but with either your compositor or desktop environment's implementation of the Wayland protocol and surrounding functionality.

In other words, that's a KDE issue, not Wayland, moving to kwin instead.

Comment 2 Jed 2023-11-03 02:43:43 UTC
Created attachment 1996867 [details]
Core dump file

Adding core dump file for analysis


Note You need to log in before you can comment on or make changes to this bug.