575 release feedback & discussion

amrits · August 26, 2025, 7:18am

Thanks for the quick update.

amrits · August 26, 2025, 10:10am

I tried repro on below setup with multiple GPUs RTX 3090 and RTX 4080 but did not observe any flickering on home screen of game “Hades”
I can see that FPS always remains at high value unlike your observation at 0 while flickering.

Setup - ASRock TRX40 Taichi + Arch Linux + kernel 6.14.5-arch1-1 + Driver 575.64.03 + NVIDIA GeForce RTX 3090
ASRock TRX40 Taichi + Arch Linux + kernel 6.14.5-arch1-1 + Driver 575.64.03 + NVIDIA GeForce RTX 4080
Display 1 - AOC AG276QZD with resolution 2560x1440 and refresh rate as 240 Hz
Display 2 - Asus VG27AQML1A with resolution 1920x1080 and refresh rate as 60 Hz

Could you be able to test with our latest 580 release driver and also share graphics settings used while playing game.

dr.silens · August 26, 2025, 12:21pm

Reproducible in 580 driver also for me.
Video showcase. Video contain rapid flashing.
In video, we can see, when I check “Fullscreen” option in game, monitor starting flashing. I can somewhat workaround this issue by unchecking “Borderless and Fullscreen” in the same time. Issue probably related to direct scanout?
I can reproduce this issue on my second monitor ( HDMI ) also. Main monitor on DP.
I also made bug report in proton-ge github. Link to bug report here

BlueGoliath · August 27, 2025, 9:50pm

Hello?

BlueGoliath · August 29, 2025, 5:49am

Another day, another driver crash. Anything remotely demanding can just crash with zero dmesg errors.

hatami · August 31, 2025, 8:49am

Hi,

After upgrading to 575.64.03 I’m having trouble using OpenCL-OpenGL interoperability. A working sample on previous versions (I remember it was working fine on 565 ) is crashing with segmentation fault with following dmesg error message:

[  996.465700] traps: pbo_interop[43406] general protection fault ip:772625718209 sp:7ffe1db00b00 error:0 in libnvidia-opencl.so.575.64.03[7726253ba000+eb8000]

It crashes on clEnqueueReleaseGLObjects(queue, 1, &clPBO, 0, NULL, NULL);

and I have no clue why this happens…

Thanks for your time and considerations.

nvidia-bug-report

MelanieB · August 31, 2025, 12:54pm

Just for info:

I now run Windows 11 with Nvidia 576.28 driver longer than 13 days and nights without any flicker, screen freezes or black screens and that on the same computer I have problems with my Tuxedo OS or even CachyOS (with 575 or 580s driver or older ones) with even flicker, screen freezes and black screens - most time within 24 to 48 hours uptime what is really annoying…

I really like Linux much better but the repeated unpredictable crashes because of Nvidia drivers on Linux I don’t like at all.

chris.bainbridge1 · September 2, 2025, 11:31am

@abchauhan

Reporting BUG

Synopsis: Intermittent Xorg crash when opening laptop lid after suspend/resume

Brief Description: This is an intermittent bug that happens about once a week. I don’t have exact steps for reproduction, but it appears to happen when opening laptop lid after suspend/resume.

Steps to Reproduce:

Disable hybrid graphics in BIOS (so only nVidia GPU is used).
Boot Ubuntu 24.04.
Connect dock.
Suspend.
Unplug dock.
Open laptop lid (laptop will resume).
(intermittent) Xorg crashes:

(EE) NVIDIA(GPU-0): Failed to acquire modesetting permission.
(EE)
Fatal server error:
(EE) EnterVT failed for screen 0
(EE)
(EE) 
Please consult the The X.Org Foundation support
         at http://wiki.x.org
 for help.
(EE) Please also check the log file at "/home/chrb/.local/share/xorg/Xorg.1.log" for additional information.
(EE)
(WW) NVIDIA(0): Failed to set the display configuration
(WW) NVIDIA(0):  - Setting a mode on head 0 failed: Insufficient permissions
(WW) NVIDIA(0):  - Setting a mode on head 1 failed: Insufficient permissions  
(WW) NVIDIA(0):  - Setting a mode on head 2 failed: Insufficient permissions
(WW) NVIDIA(0):  - Setting a mode on head 3 failed: Insufficient permissions
(EE) Server terminated with error (1). Closing log file.

System Configuration: Ubuntu 24.04 on Lenovo Legion 7 16ACHg6 (Geforce RTX 3070) with Dell WD19 dock and three external monitors.

Linux Bug Report:

nvidia-bug-report.log.gz (742.6 KB)

dr.silens · September 3, 2025, 9:44am

Still reproducible on 580.82.07

tjwmmd2 · September 7, 2025, 11:07am

Thanks for improving the new 580.82.07 (early Sept 2025 release) driver so that it no longer crashes the kernel when bugged versions of Electron/Chromium are used. :)

I have encountered that bug in Element and have documented it in more depth and reported it to the upstream. There is probably just a small list of applications that used a “new but not latest” version of Electron which will need to update themselves.

github.com/flathub/im.riot.Riot

Wayland socket crashes desktop on dual GPU (iGPU + dGPU) systems where dGPU is primary output

opened 07:55PM - 03 Sep 25 UTC

Arcitec

# Edit: Solved. Element client is using a bugged version of Electron, which brea…ks on multi-GPU systems which contain non-NVIDIA GPUs. See https://github.com/flathub/im.riot.Riot/issues/542#issuecomment-3263682588 --- - Fedora 42 with GNOME 48.4 on Wayland. - NVIDIA 3090 with driver 575.64.05. - Latest Element Flatpak. - Kernel 6.16.3 All of a sudden, Element crashes on startup with a terminal message about being unable to create a "gbm" connection (the Wayland rendering protocol), but the bugged Element processes still stick around and freeze the computer (it takes several minutes for the desktop to recover, but sometimes it doesn't recover at all). Workaround: Use Flatseal and disable the "Wayland windowing system" socket for this Flatpak to force it to use X11 via Xwayland instead. If you aren't using advanced features like screensharing, you don't need it to be running via native Wayland anyway. I suspected that this issue was due to an outdated Electron version. But they use a modern one from mid-2025: - https://github.com/element-hq/element-desktop/blob/develop/package.json#L92 - https://www.electronjs.org/docs/latest/tutorial/electron-timelines So I checked what my latest OS update installed, which is the last thing I did before Element stopped working. That update had installed a new kernel and new `glibc`. So I downgraded Kernel to 6.15.10 and glibc from `2.41-11.fc42` back to `2.41-10.fc42` (Fedora builds them from Master code, so it was a potential reason). My NVIDIA driver + GNOME versions have not changed between when Element last worked a week ago and the problem now. Well, the results are: - Regardless of kernel or glibc version, Element crashes and breaks the desktop. Worth noting: - Other Electron apps such as Vesktop and Heroic Launcher works. - Brave browser works.

MelanieB · September 7, 2025, 11:57am

After I had an uptime of more than 15 days and 12 hours on Windows 11 until I restarted it myself again on the same machine I run Linux too I tried Linux since yesterday again and I had 2 screen freezes again today within 25 minutes.

nvidia v575 bug report_NVME0N1P2_2025-09-07_01.txt (35.1 KB)

nvidia-bug-report.log__NVME0N1P2_2025-09-07_01.gz (915.1 KB)

nvidia v575 bug report_NVME0N1P2_2025-09-07_02.txt (35.0 KB)

nvidia-bug-report.log__NVME0N1P2_2025-09-07_02.gz (932.6 KB)

Edit:
And because it was so wonderful, there was just one more black screen on Linux.
But I don’t feel like documenting that right now.
I’m back on Windows 11 for now.

vrachatte · September 8, 2025, 12:31pm

This issue has been root caused internally and fix will be available in upcoming release.

MelanieB · September 14, 2025, 12:58am

Black screen

after almost 52 hours uptime (longest uptime I’ve achieved last time on Linux)
while listening to a video on Youtube in Firefox (hardware acceleration disabled) besides I checked posts in https://forums.developer.nvidia.com

Sound still was going on with a short interrupt until I switched pc off via power button (pressed it ~5 sec).
Additional Kernel parameter: nvidia-drm.modeset=1

I have no idea what made the black screen. More than 1 1/2 minutes before nothing was logged (see attached bug report).

Before that long uptime there were more than 300 system updates (because I was longer on Windows 11 too) and KDE Plasma and Frameworks versions changed.

Versions before that updates:

KDE Plasma Version: 6.3.5
KDE Frameworks Version: 6.15.0

Recent system:

Operating System: TUXEDO OS
KDE Plasma Version: 6.4.4
KDE Frameworks Version: 6.17.0
Qt Version: 6.8.2
Kernel Version: 6.14.0-110029-tuxedo (64-bit)
Graphics Platform: X11

nvidia-bug-report.log_NVME0N1P2_2025-09-14_01.gz (906.5 KB)

nvidia v575 bug report_NVME0N1P2_2025-09-14_01.txt (21.1 KB)

Edit:

Sep 14 02:23:59 ANDROMEDA kernel: NVRM: GPU at PCI:0000:01:00: GPU-607aa084-1cf9-f43c-65b1-4be9d9db92f3
Sep 14 02:23:59 ANDROMEDA kernel: NVRM: Xid (PCI:0000:01:00): 158, pid=2425, name=nvidia-smi, - timeout error waiting for NV_UFLUSH_FB_FLUSH = 0x2 cnt=12887610
Sep 14 02:23:59 ANDROMEDA kernel: NVRM: Xid (PCI:0000:01:00): 154, GPU recovery action changed from 0x0 (None) to 0x1 (GPU Reset Required)
Sep 14 02:24:03 ANDROMEDA kernel: NVRM: Xid (PCI:0000:01:00): 158, - timeout error waiting for NV_UFLUSH_FB_FLUSH = 0x3 cnt=12140852
Sep 14 02:24:07 ANDROMEDA kernel: NVRM: Xid (PCI:0000:01:00): 158, - timeout error waiting for NV_UFLUSH_FB_FLUSH = 0x3 cnt=12197572
Sep 14 02:24:08 ANDROMEDA kernel: NVRM: Xid (PCI:0000:01:00): 120, GSP task panic: unknown error (0) @ pc:0x1001c16, aux:0x0, partition:2#0, task:1

Does it mean, that nvidia-smi caused it? What can I do?

marq · September 15, 2025, 8:59pm

Have you tried with HDMI instead? 575 was (or more precisely IS) a crashfest with DP here.

MelanieB · September 15, 2025, 9:54pm

Thank you, I will try it. Whole time I had DP only.

amrits · September 16, 2025, 10:38am

Hi @JyrkiVesterinen

Would you mind trying with our latest 580 released driver and share test results.

amrits · September 16, 2025, 12:07pm

Hi @dr.silens

Just wanted to update that I am able to duplicate flickering issue on Hades steam title.

Engineering team will have a look into repro setup and share further feedback once it is available.

MelanieB · September 27, 2025, 12:45am

Black Screen while watching video on YouTube in Firefox Browser

could initialize a reboot with Strg+Alt+Del and Enter but it ultimately got stuck somewhere at shutdown
switched pc off by pressing power button (~5 sec)
GPU <> Monitor connection: 8K HDMI cable 3 meter
Additional Kernel parameter: nvidia-drm.modeset=1

nvidia v575 bug report_NVME0N1P2_2025-09-27_01.txt (45.8 KB)

nvidia-bug-report.log_NVME0N1P2_2025-09-27_01.gz (914.6 KB)

thesword53 · September 27, 2025, 1:22pm

Sep 27 02:02:35 ANDROMEDA kernel: NVRM: GPU at PCI:0000:01:00: GPU-607aa084-1cf9-f43c-65b1-4be9d9db92f3
Sep 27 02:02:35 ANDROMEDA kernel: NVRM: Xid (PCI:0000:01:00): 158, pid=2904, name=openrgb, - timeout error waiting for NV_UFLUSH_FB_FLUSH = 0x2 cnt=12688200
Sep 27 02:02:36 ANDROMEDA kernel: NVRM: Xid (PCI:0000:01:00): 120, pid=2904, name=openrgb, GSP task panic: unknown error (0) @ pc:0x1001c16, aux:0x0, partition:2#0, task:1
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     Reported by libos partition:2#4 kernel v3.1 [0] @ ts:1758931355
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     RISC-V CSR State:
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         sstatus:0x0000000200000020  sscratch:0xffffffffa3013030     sie:0x0000000000000220  sip:0x0000000000000000
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         sepc:0x0000000001001c16     stval:0x0000000000000000  scause:0x0000000000000002
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     RISC-V GPR State:
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         ra:0x0000000001003b18   sp:0x0000000000307b10   gp:0x0000000000000000   tp:0x0000000000000000
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         a0:0x1868fabfb3880d40   a1:0x0000000000307aa8   a2:0x0000000000002000   a3:0x0000000004000008
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         a4:0x0000000000600000   a5:0x00000000000006fc   a6:0x0000000000000004   a7:0x0000000004003180
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         s0:0x0000000000307b20   s1:0x0000000004002900   s2:0x0000000004001000   s3:0x0000000000000000
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         s4:0x0000000004002000   s5:0x0000000004003170   s6:0x000000000000002a   s7:0x0000000004003168
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         s8:0x0000000000000023   s9:0x000000000000002f  s10:0x0000000004003148  s11:0x0000000000000000
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         t0:0x0000000000000024   t1:0x0000000000000004   t2:0x0000000000000000   t3:0x0000000004001000
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         t4:0x0000000000000001   t5:0x0000000001007548   t6:0xfffffffffffffff6
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     Stack Trace:
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x0000000001001c16
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x0000000001003b18
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x0000000001006cac
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x0000000001006c14
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     PC Trace:
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x0000000001001c16  0x0000000001003b14  0xffffffff93009d24  0x0000000001003b10  0x0000000001002b14
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x000000000100151a  0x000000000100117e  0xffffffff93002d08  0x0000000001001180  0x0000000001001510
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x0000000001002af6  0x00000000010029d8  0x00000000010025fe  0x0000000001002a16  0x0000000001002ae4
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x000000000100151a  0x000000000100117e  0xffffffff93002d08  0x0000000001001180
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     Local I/O Register State:
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x01281500:0x00000000   0x01281504:0xbadf5108   0x01281508:0x0084811c   0x01281510:0x00000000
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:         0x01281528:0x01049042   0x0128152c:0x01000040   0x01281530:0x00000000   0x01280008:0x00000040
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     ------------[ end crash report ]------------
Sep 27 02:02:36 ANDROMEDA kernel: NVRM: GPU0 GSP RPC buffer contains function 4128 (GSP_POST_NOCAT_RECORD) and data 0x0000000000000005 0x00000000010addbe.
Sep 27 02:02:36 ANDROMEDA kernel: NVRM: GPU0 RPC history (CPU -> GSP):
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     entry function                   data0              data1              ts_start           ts_end             duration actively_polling
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:      0    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000030 0x00063fbd1e6a3a4b 0x0000000000000000          y
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -1    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000038 0x00063fbd1e6851a0 0x00063fbd1e685528    904us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -2    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000030 0x00063fbd1e684c60 0x00063fbd1e685190   1328us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -3    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000038 0x00063fbd1e684814 0x00063fbd1e684c4f   1083us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -4    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000030 0x00063fbd1e68431e 0x00063fbd1e684801   1251us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -5    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000038 0x00063fbd1e683f25 0x00063fbd1e68430e   1001us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -6    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000030 0x00063fbd1e683b42 0x00063fbd1e683f14    978us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -7    76   GSP_RM_CONTROL        0x00000000402c0105 0x0000000000000038 0x00063fbd1e683785 0x00063fbd1e683b0b    902us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM: GPU0 RPC event history (CPU <- GSP):
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     entry function                   data0              data1              ts_start           ts_end             duration during_incomplete_rpc
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:      0    4128 GSP_POST_NOCAT_RECORD 0x0000000000000005 0x00000000010addbe 0x00063fbd1ea746e8 0x00063fbd1ea746e9      1us y
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -1    4130 RECOVERY_ACTION       0x0000000000000000 0x0000000000000000 0x00063fbd1ea746c8 0x00063fbd1ea746ca      2us y
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -2    4102 OS_ERROR_LOG          0x0000000000000000 0x0000000000000000 0x00063fbd1ea746ac 0x00063fbd1ea746c7     27us y
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -3    4128 GSP_POST_NOCAT_RECORD 0x0000000000000005 0x000002a5010a535c 0x00063fbd1ea7462a 0x00063fbd1ea7462d      3us y
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -4    4128 GSP_POST_NOCAT_RECORD 0x0000000000000005 0x0000000001280cf0 0x00063fbd1e69eb0e 0x00063fbd1e69eb10      2us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -5    4128 GSP_POST_NOCAT_RECORD 0x0000000000000005 0x00000000011d2610 0x00063fbd1b6f15de 0x00063fbd1b6f15de
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -6    4128 GSP_POST_NOCAT_RECORD 0x0000000000000005 0x0000000001280cf0 0x00063fbd1b6efa52 0x00063fbd1b6efa54      2us
Sep 27 02:02:36 ANDROMEDA kernel: NVRM:     -7    4128 GSP_POST_NOCAT_RECORD 0x0000000000000005 0x00000000011d2610 0x00063fbd18a4aa4e 0x00063fbd18a4aa4f      1us

Since you have GSP task panics, as a workaround, you can disable GSP by setting nvidia.NVreg_EnableGpuFirmware=0 in kernel parameters and using the proprietary kernel module.

hurikhan77 · September 28, 2025, 12:29am

After some longer uptime (5 days), I can no longer start games. The following messages will be logged to dmesg:

[455362.959137] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1353
[455362.959141] NVRM: sysmemConstruct_IMPL: *** Cannot allocate sysmem through fb heap
[455362.959153] NVRM: nvAssertOkFailedNoLog: Assertion failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from pRmApi->Alloc(pRmApi, device->session->handle, isSystemMemory ? device->handle : device->subhandle, &physHandle, isSystemMemory ? NV01_MEMORY_SYSTEM : NV01_MEMORY_LOCAL_USER, &memAllocParams, sizeof(memAllocParams)) @ nv_gpu_ops.c:4647
[455363.181997] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1353
[455363.182001] NVRM: sysmemConstruct_IMPL: *** Cannot allocate sysmem through fb heap
[455363.182079] NVRM: nvAssertOkFailedNoLog: Assertion failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from pRmApi->Alloc(pRmApi, device->session->handle, isSystemMemory ? device->handle : device->subhandle, &physHandle, isSystemMemory ? NV01_MEMORY_SYSTEM : NV01_MEMORY_LOCAL_USER, &memAllocParams, sizeof(memAllocParams)) @ nv_gpu_ops.c:4647
[455363.472908] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1353
[455363.472912] NVRM: sysmemConstruct_IMPL: *** Cannot allocate sysmem through fb heap
[455363.473180] NVRM: nvAssertOkFailedNoLog: Assertion failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from pRmApi->Alloc(pRmApi, device->session->handle, isSystemMemory ? device->handle : device->subhandle, &physHandle, isSystemMemory ? NV01_MEMORY_SYSTEM : NV01_MEMORY_LOCAL_USER, &memAllocParams, sizeof(memAllocParams)) @ nv_gpu_ops.c:4647
[455397.672545] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1353
[455397.672550] NVRM: sysmemConstruct_IMPL: *** Cannot allocate sysmem through fb heap
[455397.672618] NVRM: nvAssertOkFailedNoLog: Assertion failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from pRmApi->Alloc(pRmApi, device->session->handle, isSystemMemory ? device->handle : device->subhandle, &physHandle, isSystemMemory ? NV01_MEMORY_SYSTEM : NV01_MEMORY_LOCAL_USER, &memAllocParams, sizeof(memAllocParams)) @ nv_gpu_ops.c:4647
[455397.838114] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1353
[455397.838120] NVRM: sysmemConstruct_IMPL: *** Cannot allocate sysmem through fb heap
[455397.838138] NVRM: nvAssertOkFailedNoLog: Assertion failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from pRmApi->Alloc(pRmApi, device->session->handle, isSystemMemory ? device->handle : device->subhandle, &physHandle, isSystemMemory ? NV01_MEMORY_SYSTEM : NV01_MEMORY_LOCAL_USER, &memAllocParams, sizeof(memAllocParams)) @ nv_gpu_ops.c:4647

I can see that similar messages have been logged before but it doesn’t affect desktop usage:

[375849.320487] NVRM: nvCheckOkFailedNoLog: Check failed: Out of memory [NV_ERR_NO_MEMORY] (0x00000051) returned from _memdescAllocInternal(pMemDesc) @ mem_desc.c:1353
[375849.320503] NVRM: faultbufCtrlCmdMmuFaultBufferRegisterNonReplayBuf_IMPL: Error allocating client shadow fault buffer for non-replayable faults

A reboot will probably fix this so I suspect there’s some memory leak affecting the client shadow fault buffer?

If I try to use the 580er driver, launching games will fail with the same error message from the game, I don’t remember if I saw similar dmesg messages then. But I think it may be a different problem with 580 which I already reported in the other thread.

nvidia-bug-report.log.gz (716,2 KB)

Topic		Replies	Views
570 release feedback & discussion Linux	659	53030	March 5, 2026
555 release feedback & discussion Linux	277	47800	February 3, 2025
565 release feedback & discussion Linux	448	40288	November 10, 2025
595 release feedback & discussion Linux	317	18347	April 13, 2026
560 release feedback & discussion Linux	565	51789	November 18, 2025
590 release feedback & discussion Linux	451	35892	April 1, 2026
Nvidia, please get it together with external monitors on Wayland Linux wayland , linux , linux-driver	127	25140	April 9, 2026
550.78 release feedback & discussion thread Linux	22	5180	June 26, 2024
External monitor freezes when using dedicated GPU Linux	303	37263	March 25, 2026
Screen went black after instalation of driver 570+ on linux Drivers - Linux, Windows, MacOS linux-driver	10	5458	June 12, 2025

575 release feedback & discussion

Related topics