Hello,
Time after time, I have problem that my screen (display) became black (no-display - just like it is turned off), and sometimes freezes. I installed 16.0 from updater in Leap 15.6. Since I mostly work in browsers, I noticed this black-screen and freeze in web-browsers - more in Chrome than in Mozilla and also on YouTUBE, but this about “browser-statistic” could be just my impression.
During that interruption, sound doesn’t dissapear.
If I posted in wrong category, please move my topic to right one
If you turn off browser hardware acceleration, do you still experience the same issue?
For Chrome: Go to chrome://settings/system and turn off “Use hardware acceleration when available”. For firefox it can be disabled via about:config (accept the warning it prompts you with), find layers.acceleration.disabled, set it to “true”, then restart firefox.
I was working on Mozilla for many hours to test - everything were fine. But, suddenly, this started to happens. Maybe it is because of Youtube (many flash-videos per one page), but… I am not in this part of “computer-stupp”
The message “Desktop effects were restarted due to a graphics reset” indicates that the GPU driver reset itself and KWin recovered. Since this affects the entire desktop and not just the browser, this suggests an amdgpu / Mesa issue rather than a browser hardware acceleration problem.
Since the audio continues, it is not a complete desktop environment crash as such, just graphics related.
The next time it happens, examine sudo dmesg | grep -i amdgpu
or sudo journalctl -k | grep -i amdgpu
Look for “GPU reset”, “gfx timeout”, “failed” type messages.
BTW, is this a desktop PC? Just wondering if you have a possible marginal/failing power supply unit?
@Malcolmlewis suggested the 6.19 kernels were to include significant improvements in graphics support, so installing the 6.19 kernel built for 16.0 may be worth a try, given the flakiness of recent kernel support for GCN#1, which your Oland is.
What is the reasoning behind that suspicion? The docs say that Mesa 25.3 and higher should be fine. And definitely refrain from bringing up the usage of non-distro repos. What you do with your systems is your business, but should not be propagated to others.
Slowroll does have Mesa 25.3 still, which I have on an Oland with longterm 6.12.67 kernel. Other than kernel command lines, dmesg and journal look modestly different from OP’s (1st line of each pair of lines following is OP’s; matching lines are omitted):
dmesg:
] Command line: BOOT_IMAGE=/boot/vmlinuz-6.12.0-160000.25-default root=UUID=950fca0f-94e8-472f-b494-9b60730573c9 splash=silent resume=/dev/disk/by-uuid/514d3993-f4f2-40f5-a0af-a185edf2509a preempt=full mitigations=auto quiet security=apparmor radeon.si_support=0 amdgpu.si_support=1 rd.driver.blacklist=radeon
] Command line: root=LABEL=w71p19sslo ipv6.disable=1 net.ifnames=0 radeon.si_support=0 amdgpu.si_support=1 noresume consoleblank=0 mitigations=off
] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-6.12.0-160000.25-default root=UUID=950fca0f-94e8-472f-b494-9b60730573c9 splash=silent resume=/dev/disk/by-uuid/514d3993-f4f2-40f5-a0af-a185edf2509a preempt=full mitigations=auto quiet security=apparmor radeon.si_support=0 amdgpu.si_support=1 rd.driver.blacklist=radeon
] Kernel command line: root=LABEL=w71p19sslo ipv6.disable=1 net.ifnames=0 radeon.si_support=0 amdgpu.si_support=1 noresume consoleblank=0 mitigations=off
NA
] amdgpu 0000:01:00.0: No more image in the PCI ROM
] amdgpu: ATOM BIOS: 113-AD57900-200
] amdgpu: ATOM BIOS: 113-C5530300-105
NA
] amdgpu 0000:01:00.0: amdgpu: PCIE atomic ops is not supported
] [drm] amdgpu: 3941M of GTT memory ready.
] [drm] amdgpu: 3971M of GTT memory ready.
] amdgpu 0000:01:00.0: amdgpu: Using BOCO for runtime pm
] amdgpu 0000:01:00.0: amdgpu: Runtime PM not available
journal:
11:51 Command line: BOOT_IMAGE=/boot/vmlinuz-6.12.0-160000.25-default root=UUID=950fca0f-94e8-472f-b494-9b60730573c9 splash=silent resume=/dev/disk/by-uuid/514d3993-f4f2-40f5-a0af-a185edf2509a preempt=full mitigations=auto quiet security=apparmor radeon.si_support=0 amdgpu.si_support=1 rd.driver.blacklist=radeon
17:33 Command line: root=LABEL=w71p19sslo ipv6.disable=1 net.ifnames=0 radeon.si_support=0 amdgpu.si_support=1 noresume consoleblank=0 mitigations=off
11:51 Kernel command line: BOOT_IMAGE=/boot/vmlinuz-6.12.0-160000.25-default root=UUID=950fca0f-94e8-472f-b494-9b60730573c9 splash=silent resume=/dev/disk/by-uuid/514d3993-f4f2-40f5-a0af-a185edf2509a preempt=full mitigations=auto quiet security=apparmor radeon.si_support=0 amdgpu.si_support=1 rd.driver.blacklist=radeon
17:33 Kernel command line: root=LABEL=w71p19sslo ipv6.disable=1 net.ifnames=0 radeon.si_support=0 amdgpu.si_support=1 noresume consoleblank=0 mitigations=off
NA
18:12 amdgpu 0000:01:00.0: No more image in the PCI ROM
11:55 amdgpu: ATOM BIOS: 113-AD57900-200
18:12 amdgpu: ATOM BIOS: 113-C5530300-105
NA
18:12 amdgpu 0000:01:00.0: amdgpu: PCIE atomic ops is not supported
11:55 [drm] amdgpu: 3941M of GTT memory ready.
18:12 [drm] amdgpu: 3971M of GTT memory ready.
11:55 amdgpu 0000:01:00.0: amdgpu: Using BOCO for runtime pm
18:12 amdgpu 0000:01:00.0: amdgpu: Runtime PM not available
