570 release feedback & discussion

Graphics / Linux Linux
392

Hi @matt-schwartz ,

4924590 is still under investigation.

Hi @BlueGoliath ,

Framegen in The Finals is still broken and the VRAM issues still aren’t fixed.

Sorry, have you received any tracking or bug IDs for these issues? Thanks

1 Like
393

Any news about power management issues [BUG Report] Idle Power Draw is ASTRONOMICAL with RTX 3090 ? The topic has been opened in 2020 and the issue is still here.

394

Nope.

If you’re feeling generous, it would be nice to get a status update on:

This bug prevents my software from running on Blackwell GPUs.

1 Like
395

Regarding the VRAM issues, he means this, “invisible VRAM usage”, the numbers in nvidia-smi don’t add up, don’t think it’s tracked.
It’ss even more off when you use tools like nvtop.

2 Likes
396

Between a Plasma monitor widget, nvidia-smi, mangohud and nvtop I often get 4 different values for VRAM usage…

397

Is any build fix coming? Build misses a file that is critical to install driver from rpm

398

I have installed the 570.124.04 version of the driver on a fairly fresh install of Kubuntu 24.10 and after my computer goes to sleep and wakes up, anything that “triggers” VRR causes massive black screen flickering or even a black screen entirely(screen is on but is black) until you move the mouse.

This also happens when I move a fullscreen application(I tested with a youtube video in Microsoft Edge) to any of my monitors of the same model ( ASUS VG248QG )

It seems to go back to expected/normal behaviour after a reboot

System Information:
OS: Kubuntu 24.10 x86_64
Kernel: 6.11.0-19-generic
Resolution: 1920x1080
DE: Plasma 6.1.5
WM: kwin
CPU: AMD Ryzen 9 7950X3D (32) @ 5.759GHz
GPU: NVIDIA GeForce RTX 3080 Lite Hash Rate
Memory: 7308MiB / 31738MiB
Display Server: Wayland

nvidia-bug-report.log.gz (1.5 MB)

399

All 5 issues I’m currently tracking are still present with nVidia 570.133.07 Production Branch drivers.

Linking back to overall summary tracker:

Latest stack:

  • Arch Linux | Kernel 6.13.7
  • Gnome-shell | Mutter 48.0
  • Wayland (meson_options: xwayland=false, x11=false)
  • Gtk4 4.18.2
  • Mesa 25.0.2
  • vulkan-icd-loader 1.4.304
  • nVidia 570.133.07
400

Thanks @abchauhan! If you have some time, would you be able to check the status of 5089016 as well? So far I have not found any workarounds on the gamescope side yet.

401

You forgot:

402

Nope.

I’m only tracking issues I’ve reported myself or am directly affected by.

404

The second one is affecting all users. You can test it by running glxinfo, vulkaninfo, eglinfo and moving cursor or running glxgears on a Wayland sessions. You will see screen stutters and this especially noticeable at high refresh rate.

405

You’ll need to be far more specific with steps to reproduce.

glxinfo and glxgearsare exclusive to the X11 protocol. I’m running a pure Wayland environment without an X11 session nor Xwayland available.

vulkaninfo and eglinfo both produce textual summaries of those platforms’ integration. How would that test GPU memory issues?

vkcube and eglgears-wayland both run at 60fps here, capped by display refresh rate. Neither trigger any GPU memory issues observable with nvtop.

406

When you start any Vulkan/OpenGL/CUDA application, those programs will create GPU contexts and context creation allocates some GPU memory.
To reproduce, you can:

  • Run clinfo while running vkcube. You need to have clinfo and opencl-nvidia
  • Run vainfo while running vkcube. You need to have libva-utils and libva-nvidia-driver
  • Run eglinfo or vulkaninfo while running `vkcube but this will not be noticeable on a 60Hz display.

Here is a video to illustrate the issue:

407

Perhaps my 60fps cap is too low to reproduce this.

No visible disruption when running vainfo or vulkaninfo whilst cube spinning:

$ vkcube --wsi wayland --width 2560 --height 1440
Selected GPU 0: NVIDIA GeForce GTX 1050 Ti, type: DiscreteGpu

The issue you’re seeing could also be X11/Xwayland-related and I don’t have those available to test.

408

The issue you’re seeing could also be X11/Xwayland-related and I don’t have those available to test.

I ran vkcube --wsi wayland in Wayland plasma session. I don’t use X11.

410

nvidia-modeset system freeze since kernel-13-x, driver 86
Fedora 42 on EliteBook 8760w laptop with Geforce M3000M Maxwell graphics:

Last tested with kmod-nvidia-570.133.07-1.fc42.x86_64 from rpmfusion, which builds but locks up at boot.
As soon as I start X, plug in HDMI or turn on nvidia_modesetting=1, total system lockup. Black screen. No keyboard.
Requires hard reset. I hear the server version of nvidia driver works. As does text mode, nouveau.
nvidia-bug-report.log.gz (665.7 KB)

Trying datacenter-driver Downloads | NVIDIA Developer next.

411

I had a reproducer lately:

Playing Star Wars Outlaws pretty reliably “leaks” VRAM in the driver which nvidia-smi doesn’t show. What usually follows are errors like this:

Mär 21 02:20:32 jupiter kernel: NVRM: Xid (PCI:0000:01:00): 69, pid=7566, name=chrome, Class Error: ChId 00b0, Class 0000902d, Offset 0000023c, Data 00000000, ErrorCode 00000004
Mär 21 02:20:33 jupiter kernel: NVRM: Xid (PCI:0000:01:00): 69, pid=82105, name=chrome, Class Error: ChId 00b0, Class 0000902d, Offset 0000023c, Data 00000000, ErrorCode 00000004
Mär 21 02:20:34 jupiter kernel: NVRM: Xid (PCI:0000:01:00): 69, pid=82137, name=chrome, Class Error: ChId 00b0, Class 0000902d, Offset 0000023c, Data 00000000, ErrorCode 00000004
Mär 21 02:20:34 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object
Mär 21 02:20:34 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object
Mär 21 02:20:34 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object
Mär 21 02:20:34 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object
Mär 21 02:20:35 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object
Mär 21 02:20:35 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object
Mär 21 02:20:35 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object
Mär 21 02:20:35 jupiter kernel: [drm:nv_drm_gem_alloc_nvkms_memory_ioctl [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NVKMS memory for GEM object

After quitting the game and restarting all the processes using VRAM according to nvidia-smi, there were still around 500 MB missing. DXVK games seem to not trigger the missing memory behavior. Maybe because DXVK games tend to do a lot less Vulkan memory allocations due to the chunk allocator?

Anyway, after the NVKMS errors in dmesg, I usually also see errors about the kernel having page fault errors and I have to reboot. So the driver probably leaks not only VRAM but also somehow damages other kernel memory structures:

Mär 22 03:22:56 jupiter kernel: Huh VM_FAULT_OOM leaked out to the #PF handler. Retrying PF
Mär 22 03:22:56 jupiter kernel: Huh VM_FAULT_OOM leaked out to the #PF handler. Retrying PF
Mär 22 03:22:56 jupiter kernel: Huh VM_FAULT_OOM leaked out to the #PF handler. Retrying PF

I am not able to reproduce the latter without the NVIDIA driver or on system without NVIDIA hardware but otherwise very similar configuration. It happens with both DXVK and vkd3d games.

This has become better and worse with the latest 570 driver series: While the desktop no longer renders black windows randomly or completely crashes, I can trigger this problem within 24 hours usually.

The problem is worse using the kernel open driver, so I’m currently going with the closed driver.

1 Like
412

This is still happening in 570.133.07.

Mar 22 14:53:29 zzzzzz kernel: NVRM: GPU at PCI:0000:08:00: GPU-738d510d-abfd-7290-16db-025b3eb23693
Mar 22 14:53:29 zzzzzz kernel: NVRM: Xid (PCI:0000:08:00): 109, pid=6586, name=SOTTR.exe, Ch 00000050, errorString CTX SWITCH TIMEOUT, Info 0x5ec05a

nvidia-bug-report.log.gz (1.6 MB)

1 Like
413

[Monster Hunter wilds Vertex Explosions and performance issues - Graphics / Linux / Linux - NVIDIA Developer Forums](https://forums.developer.nvidia.com/t/monster-hunter-wilds-vertex-explosions-and-performance-issues reported another game with vertex issues and if you all have this game and have vertex issues please report it