Thank you.
I just saw your answer.
With 580s driver I have a GSP problem too. So I will try it.
For people who, like me, cannot run v580 due to its enormous amount of bugs, and prefer to stick with 575.64.05, attached is a patch for compiling the GPL modules of the latter against Linux 6.17.
Good news: after manually installing driver 580.82.09, the warning spam in dmesg is gone. Looks like that issue has been solved. đ
1 Like
Iâm sometimes seeing this error, especially when running some heavier games (that use advanced options like DLSS or RT):
[ 7196.695178] NVRM: nvGpuOpsReportFatalError: uvm encountered global fatal error 0x60, requiring os reboot to recover.
[ 7196.697695] NVRM: GPU at PCI:0000:01:00: GPU-94de4b2d-e5a4-fd60-a9e3-fd87115fbaef
[ 7196.697697] NVRM: Xid (PCI:0000:01:00): 154, GPU recovery action changed from 0x0 (None) to 0x2 (Node Reboot Required)
[ 7196.847221] NVRM: nvGpuOpsReportFatalError: uvm encountered global fatal error 0x60, requiring os reboot to recover.
But I donât see what isnât working: Everything still works as expected.
But thereâs one issue: Thereâs around 1.2 GB difference in nvidia-smi between VRAM usage reported and per-app VRAM usage summarized. Tho, Iâm not sure if this is the general âmemory leakâ that is still in the driver, or if this is directly related to the uvm error reported.
nvidia-bug-report.log.gz (675,8 KB)
I got a lot of follow-up errors (I didnât reboot yet):
[92124.519234] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E930000-0xDF1E93FFFF!
[92124.519447] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E800000-0xDF1E81FFFF!
[92124.519645] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1EAF0000-0xDF1EAFFFFF!
...
[92124.525775] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92124.525779] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92124.525784] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
...
[92124.596464] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E930000-0xDF1E93FFFF!
[92124.596671] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E800000-0xDF1E81FFFF!
[92124.596864] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1EAF0000-0xDF1EAFFFFF!
...
[92124.602731] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92124.602734] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92124.602739] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
...
[92124.792067] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E930000-0xDF1E93FFFF!
[92124.792287] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E800000-0xDF1E81FFFF!
[92124.792494] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1EAF0000-0xDF1EAFFFFF!
...
[92124.799095] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92124.799098] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92124.799104] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
...
[92124.813018] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1EA20000-0xDF1EAAFFFF!
[92124.814748] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92124.814751] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92124.942773] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E930000-0xDF1E93FFFF!
...
[92124.949481] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92124.949485] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92124.949490] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
...
[92124.949629] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92124.964695] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1EA20000-0xDF1EAAFFFF!
[92124.966136] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92124.966139] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92125.145773] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1E930000-0xDF1E93FFFF!
...
[92125.153301] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92125.153304] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
[92125.153310] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
...
[92125.168641] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF1EA20000-0xDF1EAAFFFF!
[92125.171587] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4838
[92125.171595] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2272
nvidia-bug-report.log.gz (713,0 KB)
Hello Chris, We have created a local setup but did not get this issue reproduced locally. Are you still able to reproduce this issue?
This issue is fixed in latest 580 driver, Can you please check and update status?
Hello Chris, We have created a local setup but did not get this issue reproduced locally. Are you still able to reproduce this issue?
Unfortunately I think this bug is hard to reliably reproduce. It was an intermittent fault that only happened when using the âdynamic graphicsâ setting in the BIOS. Iâm not sure of the exact conditions under which it occurred, apart from that it happened shortly after starting a custom Unreal engine app.
To workaround the problem, I switched to âdiscrete graphicsâ in the BIOS and havenât seen it since. I just tried enabling âdynamicâ graphics with 580.82.09 and had a play for a few minutes with full-screening parallel instances of the app, when I closed them down I got an error nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c67e:4:0:1258 but it didnât hard hang with 100% CPU like before. A more immediate problem is that the external monitors no longer work with âdynamic graphicsâ when the laptop is booted with the dock attached, but presumably this is the same/related problem Iâve already reported in the 580 thread.
Hello, Can you please try upgrading driver to latest 580 build and check if you get this issue there as well?
This issue also reproducible on âPeakâ steam link with Vulkan renderer. Same symptoms: zero fps reported, rapid flickering.
Hi, unfortunately the issue still exists with the current 580 driver. The freezing of both image and sound can be interrupted by tabbing out of the game and then back in. Iâve also noticed a similar problem recently with Age of Empires 4. I seem to recall that someone else reported a similar issue before, and the assumption was that some blocked state from the KDE desktop wasnât being released. Does that sound familiar to you?
Please confirm if you are still seeing gamescope crash issue with our latest released driver.
Also please share repro video for reference if issue still persists.