Bug 509359

Summary: Spectacle caused amdgpu to crash
Product: [Plasma] kwin Reporter: Greeniac <quarro>
Component: platform-drmAssignee: KWin default assignee <kwin-bugs-null>
Status: RESOLVED UPSTREAM    
Severity: crash CC: xaver.hugl
Priority: NOR    
Version First Reported In: 6.4.4   
Target Milestone: ---   
Platform: Fedora RPMs   
OS: Linux   
Latest Commit: Version Fixed/Implemented In:
Sentry Crash Report:

Description Greeniac 2025-09-10 20:16:22 UTC 
Been using my computer normally, but when I launched spectacle to take a screenshot, amdgpu crashed, and the only way to recover it was to do a force shutdown, and when it booted back again, I could not reproduce the issue, so I am not sure what truly caused it.

But with journalctl, I got the logs of the crash (line 131710 of the attatched log file):

Sep 10 22:38:01 fedora systemd[2004]: Starting app-org.kde.spectacle.service - Spectacle screenshot capture utility...
Sep 10 22:38:02 fedora systemd[2004]: Started app-org.kde.spectacle.service - Spectacle screenshot capture utility.
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu: GPU fault detected: 147 0x00024402
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu:  for process spectacle pid 253190 thread spectacle:cs0 pid 253194
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu:   VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x00000C00
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu:   VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x09044002
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu: VM fault (0x02, vmid 4, pasid 32830) at page 3072, write from 'TC5' (0x54433500) (68)
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu: GPU fault detected: 147 0x00004802
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu:  for process spectacle pid 253190 thread spectacle:cs0 pid 253194
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu:   VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x00000C00
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu:   VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x08048002
Sep 10 22:38:02 fedora kernel: amdgpu 0000:01:00.0: amdgpu: VM fault (0x02, vmid 4, pasid 32830) at page 3072, read from 'TC4' (0x54433400) (72)
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: Dumping IP State
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: Dumping IP State Completed
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: [drm] AMDGPU device coredump file has been created
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: ring gfx timeout, signaled seq=58859845, emitted seq=58859848
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: Process information: process spectacle pid 253190 thread spectacle:cs0 pid 253194
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: Starting gfx ring reset
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: Ring gfx reset failed
Sep 10 22:38:12 fedora kernel: amdgpu 0000:01:00.0: amdgpu: GPU reset begin!
Sep 10 22:38:13 fedora kernel: amdgpu: cp is busy, skip halt cp
Sep 10 22:38:13 fedora kernel: amdgpu: rlc is busy, skip halt rlc
Sep 10 22:38:13 fedora kernel: amdgpu 0000:01:00.0: amdgpu: PCI CONFIG reset
Sep 10 22:38:13 fedora kernel: amdgpu 0000:01:00.0: amdgpu: GPU reset succeeded, trying to resume
Sep 10 22:38:13 fedora kernel: [drm] PCIE GART of 256M enabled (table at 0x000000F402000000).
Sep 10 22:38:13 fedora kernel: [drm] VRAM is lost due to GPU reset!
Sep 10 22:38:13 fedora kernel: [drm] UVD and UVD ENC initialized successfully.
Sep 10 22:38:13 fedora kernel: [drm] VCE initialized successfully.
Sep 10 22:38:13 fedora kernel: amdgpu 0000:01:00.0: amdgpu: GPU reset(1) succeeded!
Sep 10 22:38:13 fedora kernel: amdgpu 0000:01:00.0: [drm] device wedged, but recovered through reset
Sep 10 22:38:13 fedora kwin_wayland[2305]: kwin_scene_opengl: 0x4: GL_CONTEXT_LOST in context lost
Sep 10 22:38:13 fedora kwin_wayland[2305]: kwin_scene_opengl: 0x4: GL_CONTEXT_LOST in context lost
Sep 10 22:38:14 fedora audit[253190]: ANOM_ABEND auid=1000 uid=1000 gid=1000 ses=3 subj=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 pid=253190 comm="spectacle:cs0" exe="/usr/bin/spectacle" sig=6 res=1
Sep 10 22:38:14 fedora spectacle[253190]: amdgpu: The CS has cancelled because the context is lost. This context is innocent.
Sep 10 22:38:14 fedora systemd-coredump[253294]: Process 253190 (spectacle) of user 1000 terminated abnormally with signal 6/ABRT, processing...
Sep 10 22:38:14 fedora audit: BPF prog-id=691 op=LOAD
Sep 10 22:38:14 fedora audit: BPF prog-id=692 op=LOAD
Sep 10 22:38:14 fedora audit: BPF prog-id=693 op=LOAD
Sep 10 22:38:14 fedora systemd[1]: Started systemd-coredump@13-253294-0.service - Process Core Dump (PID 253294/UID 0).
Sep 10 22:38:14 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@13-253294-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Sep 10 22:38:14 fedora audit: BPF prog-id=694 op=LOAD
Sep 10 22:38:14 fedora audit: BPF prog-id=695 op=LOAD
Sep 10 22:38:14 fedora audit: BPF prog-id=696 op=LOAD
Sep 10 22:38:14 fedora systemd[1]: Started drkonqi-coredump-processor@13-253294-0.service - Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Sep 10 22:38:14 fedora audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=drkonqi-coredump-processor@13-253294-0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Sep 10 22:38:14 fedora systemd-coredump[253295]: Removed old coredump core.tf_linux64.1000.c81a76ea03da473299f8305f6f7011a8.50595.1756641657000000.zst.
Sep 10 22:38:14 fedora abrt-dump-journal-core[1494]: Failed to obtain all required information from journald
Sep 10 22:38:14 fedora abrt-dump-journal-core[1494]: Failed to obtain all required information from journald
Sep 10 22:38:15 fedora systemd-coredump[253295]: Process 253190 (spectacle) of user 1000 dumped core.

STEPS TO REPRODUCE
could not reproduce



SOFTWARE/OS VERSIONS

Operating System: Fedora Linux 42
KDE Plasma Version: 6.4.4
KDE Frameworks Version: 6.17.0
Qt Version: 6.9.1
Kernel Version: 6.16.4-200.fc42.x86_64 (64-bit)
Graphics Platform: Wayland
Processors: 4 × Intel® Core™ i5-7600K CPU @ 3.80GHz
Memory: 24 GiB of RAM (23.4 GiB usable)
Graphics Processor: AMD Radeon Pro 580X
Manufacturer: Apple Inc.
Product Name: iMac18,3
System Version: 1.0
Comment 1 Greeniac 2025-09-10 20:18:33 UTC 
I decided not to attach the log file, since it's 34mb (uptime was 4 days), and I fear that it may include some information that can be traced as I saw some Mullvad logs. But if further information is needed, I will try my best to provide them.
Comment 2 Zamundaaa 2025-09-10 20:57:26 UTC 
This is a bug in the OpenGL driver. If you can't replicate it, it's probably very difficult to get useful information about it, but you can report it at https://gitlab.freedesktop.org/mesa/mesa/-/issues (especially if it does happen again)