560 release feedback & discussion

The poor quality of Linux drivers since 555 and 560 has kept me in 550 since January to upgrade things, which has problems, but far less than either 555 or 560. I upgraded then to test KDE 6.0 and ideally Wayland, but found Wayland+Nvidia still a mess, and remains largely so with small steps forward in 555/560 to fix explicit sync issues, but break everything else in the process. Particularly gaming that I enjoy now with Proton.

Watching 555/560 in these threads has made me rethink my strategy in any desktop/gaming system I buy in the future ever including Nvidia again, as at this point over some 15 years of Nvidia Linux drivers personally, I just give up and will try AMD for some hope of actual Wayland use.

AMD may not be entirely better, but it simply canā€™t be this bad to rub against any and every Linux graphics feature that Nvidia has across 20 years of GPU history. Thereā€™s a reason Steam chose AMD over Nvidia for the Steam Deck, not to mention Playstation and Xboxā€¦

1 Like

Still cannot run apps that render via Vulkan on Wayland with 560 drivers. I would downgrade to 555 but its almost impossible. Error is:

wl_surface#67: error 0: DRM Syncobj surface object already created for surface 67
Protocol error 0 on object wl_surface

This isnā€™t even gaming/ML related. Just trying to use my computer. Please Nvidia, dedicate some resources to fixing this!

1 Like

amd is no better right now itā€™s a huge flux and a mess sure 560 has minor issues but most of them are solved in whatever driver comes next so whining about it instead of waiting helps no one here. just report bugs and move on.
560 for me has been the best driver nvidia has had in ages! and it will only get better from here.

4 Likes

Getting hilariously bad performances on Sleeping Dogs (1 FPS, literally)
dmesg is spammed with:

rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)

nvidia-bug-report.log.gz (1.7 MB)

RTX4070 (laptop) / dxvk 2.4.1 / ge-proton 9-16 / linux 6.11.4 / X (i3) / nvidia-open-dkms 560.35.03-16

Gave 560.35.03 a try again on my Razer laptop (Intel + RTX3070), but the GPU is still unable to get into D3cold when no external displays are connected, both in open and closed drivers mode, with or without GSP.

Running Ubuntu 24.10 with Gnome+Wayland.

Still staying on Nouveau for now to have a somewhat usable battery life.

nvidia-bug-report.log.gz (1.9 MB)

when ran through gamescope, DiabloIV systematically crashes ~40sec after launch.
dmesg spammed with:

[45966.947639] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45966.962293] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF35E10000-0xDF35E1FFFF!
[45966.962315] NVRM: rmapiMapWithSecInfo: Nv04Map: map failed; status: Found a duplicate entry in the requested btree [NV_ERR_INSERT_DUPLICATE_NAME] (0x00000019)
[45966.962553] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF36E70000-0xDF36E7FFFF!
[45966.962567] NVRM: rmapiMapWithSecInfo: Nv04Map: map failed; status: Found a duplicate entry in the requested btree [NV_ERR_INSERT_DUPLICATE_NAME] (0x00000019)
[45966.962765] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF36E20000-0xDF36E6FFFF!
[45966.962779] NVRM: rmapiMapWithSecInfo: Nv04Map: map failed; status: Found a duplicate entry in the requested btree [NV_ERR_INSERT_DUPLICATE_NAME] (0x00000019)
[45966.964090] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45966.964107] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4798
[45966.964111] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2274
[45966.964121] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4798
[45966.964123] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2274
[45966.964132] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:4798
[45966.964134] NVRM: nvAssertFailedNoLog: Assertion failed: NV_OK == status @ gpu_vaspace.c:2274
[45966.968974] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45968.254781] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45968.259478] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45968.261697] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45972.759013] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45973.155196] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45973.936317] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45974.269888] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45974.810194] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45977.003088] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45977.101851] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45977.126604] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45977.140335] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45977.148817] NVRM: rmapiUnmapWithSecInfo: Nv04Unmap: ummap failed; status: Requested object not found [NV_ERR_OBJECT_NOT_FOUND] (0x00000057)
[45977.163588] NVRM: intermapRegisterDmaMapping: Failed to insert new mapping node for range 0xDF35C50000-0xDF35E0FFFF!

Same for me : Black screen or Glitches after a resume from Hibernation (Ubuntu 24.10)
I have this problem for a long time now (i mean last version of Ubuntu and lot a NVIDIA driver version).
Also, trying lot of settings but resume doesnā€™t work.

A lot of people have hibernate problem, so i thing, that some distro Linux set up a team to work with NVIDIA to solve these problems once and for all.

Iā€™m having a severe performance issue with Vulkan apps running in XWayland. The frame rate is extremely low with this combination. All other combinations (Wayland+Vulkan, Wayland+OpenGL, XWayland+OpenGL) work great with very high performance. This performance issue is regardless of the VkPresentModeKHR. The framerate gets worse with window size (see video).

To reproduce, compile GitHub - krh/vkcube: Spinning Vulkan Cube, and run once with ./vkcube -m wayland to observe everything works great. Then run again with ./vkcube -m xcb, maximize the window, and see the extremely bad framerate.

I donā€™t know if itā€™s related, but simply running glxgears and maximizing the window is also horrible for me. Yet, glxgears uses OpenGL instead of Vulkan.

Iā€™m on 560.35.03 (with GTX1660Ti). I validated that the test program (vkcube) is running on the GPU through nvidia-smi. Also mangohud confirms this:

Hey @courteauxmartijn Im afraid to say its Hyprland issue , Hyprland has so many issues with nvidia, poorly implemented protocols etc. Iā€™d suggest try gnome and kwin, if you prefer window managers use sway, although its unsupported on nvidia but their implementation is way better.

1 Like

565.57.01 | BETA is out now Driver Details | NVIDIA

1 Like

just tested 565, gamescope mouse cursor being cut off is fixed, however wayland native osu!lazer still crashes under vulkan renderer, and gamescope minecraft 1.8.9 still has quite intense flickering

can we please get 565 open kernel modules soon on the github

Usually they only commit once they hit a final release tag. I donā€™t think beta code will be released this time either.

Open modules are available as of 19 hours ago: Release 565.57.01 Ā· NVIDIA/open-gpu-kernel-modules Ā· GitHub

2 Likes

Not sure if the Unreal Engine Editor problem is ever going to be fixed in 565, for some reason the 4060Ti wasnā€™t compatible with 535 where UE editor was functional so forced to update to 560 and this happens almost immediately after opening a project from 5.1 and migrating it to 5.4.

Operating System: Arch Linux
KDE Plasma Version: 6.2.2
KDE Frameworks Version: 6.7.0
Qt Version: 6.8.0
Kernel Version: 6.11.5-arch1-1 (64-bit)
Graphics Platform: Wayland
Processors: 12 Ɨ 12th Gen IntelĀ® Coreā„¢ i5-12400F
Memory: 62.6 GiB of RAM
Graphics Processor: NVIDIA GeForce RTX 4060 Ti/PCIe/SSE2
Manufacturer: ASUS

LoginId:5dd1433c01064299ae5582f2547adf36-000003e8

Caught signal 11 Segmentation fault

libnvidia-glcore.so.560.35.03!UnknownFunction(0xa10857)
libnvidia-glcore.so.560.35.03!UnknownFunction(0xe279e8)
libnvidia-glcore.so.560.35.03!UnknownFunction(0xe0d37f)
libnvidia-glcore.so.560.35.03!UnknownFunction(0xd4b78d)
libUnrealEditor-VulkanRHI.so!VulkanRHI::FFenceManager::WaitForFence(VulkanRHI::FFence*, unsigned long long) [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/VulkanRHI/Private/VulkanMemory.cpp:4690]
libUnrealEditor-VulkanRHI.so!FVulkanCommandBufferManager::WaitForCmdBuffer(FVulkanCmdBuffer*, float) [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/VulkanRHI/Private/VulkanCommandBuffer.cpp:502]
libUnrealEditor-VulkanRHI.so!FVulkanViewport::WaitForFrameEventCompletion() [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/VulkanRHI/Private/VulkanViewport.cpp:278]
libUnrealEditor-VulkanRHI.so!FVulkanViewport::Present(FVulkanCommandListContext*, FVulkanCmdBuffer*, FVulkanQueue*, FVulkanQueue*, bool) [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/VulkanRHI/Private/VulkanViewport.cpp:1048]
libUnrealEditor-VulkanRHI.so!FVulkanCommandListContext::RHIEndDrawingViewport(FRHIViewport*, bool, bool) [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/VulkanRHI/Private/VulkanRHI.cpp:1184]
libUnrealEditor-RHI.so!FRHICommand<FRHICommandEndDrawingViewport, FRHICommandEndDrawingViewportString2188>::ExecuteAndDestruct(FRHICommandListBase&, FRHICommandListDebugContext&) [/mnt/horde/++UE5/Sync/Engine/Source/Runtime/RHI/Public/RHICommandList.h:1295]
libUnrealEditor-RHI.so!FRHICommandListBase::Execute(TRHIPipelineArray<IRHIComputeContext*>&, FRHICommandListBase::FPersistentState::FGPUStats*) [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/RHI/Private/RHICommandList.cpp:477]
libUnrealEditor-RHI.so!UE::Core::Private::Function::TFunctionRefCaller<FRHICommandListImmediate::ExecuteAndReset(bool)::$_0, void ()>::Call(void*) [/mnt/horde/++UE5/Sync/Engine/Source/Runtime/Core/Public/Templates/Function.h:405]
libUnrealEditor-Core.so!TGraphTask<TFunctionGraphTaskImpl<void (), (ESubsequentsMode::Type)0> >::ExecuteTask(TArray<FBaseGraphTask*, TSizedDefaultAllocator<32> >&, ENamedThreads::Type, bool) [/mnt/horde/++UE5/Sync/Engine/Source/Runtime/Core/Public/Async/TaskGraphInterfaces.h:1235]
libUnrealEditor-Core.so!FNamedTaskThread::ProcessTasksNamedThread(int, bool) [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/Core/Private/Async/TaskGraph.cpp:760]
libUnrealEditor-Core.so!FNamedTaskThread::ProcessTasksUntilQuit(int) [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/Core/Private/Async/TaskGraph.cpp:650]
libUnrealEditor-RenderCore.so!FRHIThread::Run() [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/RenderCore/Private/RenderingThread.cpp:330]
libUnrealEditor-Core.so!FRunnableThreadPThread::Run() [/mnt/horde/++UE5/Sync/Engine/Source/./Runtime/Core/Private/HAL/PThreadRunnableThread.cpp:25]
libUnrealEditor-Core.so!FRunnableThreadPThread::_ThreadProc(void*) [/mnt/horde/++UE5/Sync/Engine/Source/Runtime/Core/Private/HAL/PThreadRunnableThread.h:187]
libc.so.6!UnknownFunction(0x9439c)
libc.so.6!UnknownFunction(0x11949b)

Does anyone know why NVENC no longer works in 560 driver? In older drivers and older GPUs the GPU load was always at least 40% while video conversion with NVENC (Handbrake). But now with this driver the GPU load is 1-2% and the whole load goes to the CPU which means NVENC is no longer working. The Intel CPU does NOT have an integrated GPU, like some older generations.
Why is that and how do I make NVENC work in this driver? The exact version is 560.35.03-18 from Arch repo and the GPU is RTX 3060.

Edit: after nearly 12 hours of testing and experimenting, I found out that NVENC is actually working and Handbrake is using it. The output file has the expected size and quality, so NVENC is working. But for whatever reason the GPU isnā€™t being loaded at all and conky indicates the most of the loads goes to the CPU. IDK, maybe nowadays GPUs are too powerful and with large VRAM sizes to even indicate any load. Whatever the cause of this misleading display, a handbrake encoding log clearly indicates itā€™s using NVENC for the conversion.

1 Like

Confirming NVBug #4840658 fixed with 565.57.01 Beta driver.

This was also affecting discrete GPUs running Gnome Wayland sessions, despite the specifics of the bug title.

1 Like

So There is a temp fix to flipping of the Wayward driver - So if you revert to the standard X driver you can eliminate some the problems in the meantime; Until Nvidia makes it priority. In the GUI logon screen there is an icon in lower right hand corner. You can select the display driver from that. Good Luck.

When can Linux users expect a solution to the Wayland Driver issue? At present the driver still presents problems to a majority of users and we are forced to revert back to standard X Display. Many of us believe you have no intentions of resolving the problem.

1 Like