Yep, I’m using the Starfield script extender exe to launch it. I tried different Proton versions in my troubleshooting but didn’t think to try running with the original exe.
Could you try running with the original game exe and confirm that it resolves the issue? If it does if you could give me the name of the script extender executable file that would also be helpful.
Hey, as I have the same problem as him, and was opening Starfield with script extender, I tried loading a save without the script extender and still have this issue:
Running openSUSE Tumbleweed with KDE/Wayland, latest beta drivers 565.57.01, RTX 2060 SUPER 8GB
Seems fixed on NixOS with linux 6.11 and 565.57.01.
560 was not working well with 6.11 (was not letting SDDM boot properly) and 560 with 6.10 KDE crashed (probably because of HDR as you said).
Hi @edgarscirulis ,
This remains under investigation. The fix is not available on this 565 release.
- What version of Minecraft application are you using?
- Do you see a crash on switching the application to fullscreen or on exiting the application. On recent versions of Minecraft, for example, 1.21.2 we no longer see a crash on switching to full screen but do see a crash on exiting the game.
Hi @thesword53 ,
- Have you reported this issue before? Please share a reference bug number if you have received any.
- Can you please attach a NVIDIA bug report from your system.
- Can you please share your steps to reproduce this issue?
Hi @lurkingaround ,
If you are referring to the crash tracked at item 7 from 560 release feedback & discussion - #210 by abchauhan
- 4717891 - On 555, Unreal engine crashes after running for some time
This is under investigation. We hope to have this fixed on our next release.
Thank you
1 Like
Hi, @abchauhan
On newer Minecraft versions, the game crashes on exit, while on older versions like 1.12.2, it crashes when entering full-screen mode and on exit.
Disabling / Setting to __GL_THREADED_OPTIMIZATIONS=0 seems to fix the issue, so it may be related to this.
1 Like
Yeah, I can confirm, this is with the 560 driver but it’s the same issue you’re experiencing with 565.
Crash on exit with Minecraft 1.16.4, but same happens with 1.21.3
# A fatal error has been detected by the Java Runtime Environment:
#
# SIGBUS (0x7) at pc=0x000063ae4a8387cc, pid=119092, tid=119093
#
# JRE version: OpenJDK Runtime Environment (17.0.13+11) (build 17.0.13+11)
# Java VM: OpenJDK 64-Bit Server VM (17.0.13+11, mixed mode, sharing, tiered, compressed oops, compressed class ptrs, g1 gc, linux-amd64)
# Problematic frame:
# C [libnvidia-glcore.so.560.35.03+0xa387cc]
#
I installed 565.57.01 from RPMFusion testing repo on Fedora 41, and the "KFENCE: use-after-free" trace still occasionally appears on the logs:
out 30 09:51:37 fedoracosta kernel: ==================================================================
out 30 09:51:37 fedoracosta kernel: Hardware name: Acer Nitro AN515-44/Stonic_RNS, BIOS V1.04 02/04/2021
out 30 09:51:37 fedoracosta kernel: Tainted: [P]=PROPRIETARY_MODULE, [O]=OOT_MODULE, [E]=UNSIGNED_MODULE
out 30 09:51:37 fedoracosta kernel: CPU: 4 UID: 1000 PID: 2469 Comm: gnome-shell Tainted: P OE 6.11.5-300.fc41.x86_64 #1
out 30 09:51:37 fedoracosta kernel:
out 30 09:51:37 fedoracosta kernel: entry_SYSCALL_64_after_hwframe+0x76/0x7e
out 30 09:51:37 fedoracosta kernel: do_syscall_64+0x82/0x160
out 30 09:51:37 fedoracosta kernel: __x64_sys_ioctl+0x97/0xd0
out 30 09:51:37 fedoracosta kernel: nvidia_unlocked_ioctl+0x53b/0x8d0 [nvidia]
out 30 09:51:37 fedoracosta kernel: rm_ioctl+0x7f/0x400 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv000763rm+0x20f/0xe00 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv000605rm+0x4d/0x60 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv049403rm+0x41/0x70 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv049402rm+0xdd/0x190 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv051319rm+0x6c5/0x730 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv051321rm+0xb7/0x180 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv049512rm+0xb3/0xd5 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv018262rm+0x5a4/0x690 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv006119rm+0x27/0xc0 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv004343rm+0xd/0x20 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv039478rm+0x65/0x130 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv039514rm+0xc2/0x430 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv044974rm+0x67/0xd0 [nvidia]
out 30 09:51:37 fedoracosta kernel: nv_dma_release_sgt+0x49/0x70 [nvidia]
out 30 09:51:37 fedoracosta kernel: freed by task 2469 on cpu 4 at 808.510300s:
out 30 09:51:37 fedoracosta kernel:
out 30 09:51:37 fedoracosta kernel: entry_SYSCALL_64_after_hwframe+0x76/0x7e
out 30 09:51:37 fedoracosta kernel: do_syscall_64+0x82/0x160
out 30 09:51:37 fedoracosta kernel: __x64_sys_ioctl+0x97/0xd0
out 30 09:51:37 fedoracosta kernel: drm_ioctl+0x28b/0x540
out 30 09:51:37 fedoracosta kernel: drm_ioctl_kernel+0xb3/0x100
out 30 09:51:37 fedoracosta kernel: drm_gem_prime_fd_to_handle+0xea/0x220
out 30 09:51:37 fedoracosta kernel: drm_gem_prime_import_dev+0x96/0x180
out 30 09:51:37 fedoracosta kernel: nv_drm_gem_prime_import_sg_table+0x2d/0xb0 [nvidia_drm]
out 30 09:51:37 fedoracosta kernel: allocated by task 2469 on cpu 4 at 808.479455s:
out 30 09:51:37 fedoracosta kernel: kfence-#217: 0x00000000c2533f2c-0x00000000cf5865c4, size=384, cache=kmalloc-rnd-15-512
out 30 09:51:37 fedoracosta kernel:
out 30 09:51:37 fedoracosta kernel: entry_SYSCALL_64_after_hwframe+0x76/0x7e
out 30 09:51:37 fedoracosta kernel: do_syscall_64+0x82/0x160
out 30 09:51:37 fedoracosta kernel: __x64_sys_ioctl+0x97/0xd0
out 30 09:51:37 fedoracosta kernel: nvidia_unlocked_ioctl+0x53b/0x8d0 [nvidia]
out 30 09:51:37 fedoracosta kernel: rm_ioctl+0x7f/0x400 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv000763rm+0x20f/0xe00 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv000605rm+0x4d/0x60 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv049403rm+0x41/0x70 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv049402rm+0xdd/0x190 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv051319rm+0x6c5/0x730 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv051321rm+0xb7/0x180 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv049512rm+0xb3/0xd5 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv018262rm+0x5a4/0x690 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv006119rm+0x27/0xc0 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv004343rm+0xd/0x20 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv039478rm+0x65/0x130 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv039514rm+0xc2/0x430 [nvidia]
out 30 09:51:37 fedoracosta kernel: _nv044974rm+0x67/0xd0 [nvidia]
out 30 09:51:37 fedoracosta kernel: nv_dma_release_sgt+0x29/0x70 [nvidia]
out 30 09:51:37 fedoracosta kernel: Use-after-free read at 0x00000000a008055b (in kfence-#217):
out 30 09:51:37 fedoracosta kernel: BUG: KFENCE: use-after-free read in nv_dma_release_sgt+0x29/0x70 [nvidia]
out 30 09:51:37 fedoracosta kernel: ==================================================================
Other than that, seems pretty solid so far, but I don’t use it for gaming. Secondary monitor works on HDMI, but usually is a bit laggy (especially for text input on browser windows).
nvidia-bug-report.log.gz (461.7 KB)
1 Like
The 565 driver is working fine on the Acer Predator G9-793-71SJ, driver 550, 555 and 560 all failed with black screens of death or a blinking cursor during boot. Alot of issues with suspend (black screen of death). With 565 driver installed manually all the issues i had before are gone now. I am using Pop! OS 22.04 with kernel 6.9
[565.57.01] GTK4 apps background CPU usage with Vulkan renderer
As of gtk 4.16 the Vulkan rendering backend is used by default.
GTK4 apps are consuming unexpected CPU cycles while idle in the background when using the Vulkan backend.
This does not occur with the NGL renderer.
This issue is still occurring with the 565.57.01 beta.
For additional info, see initial report for the 560.35.03 driver:
Steps to reproduce
- Launch a GTK4 app (e.g. GTK4-demo, Nautilus, gnome-text-editor) using the Vulkan backend
- Observe the CPU usage of a GTK4 window’s thread
Current behavior
CPU usage constantly hovering between 2-3%, even when the window is ‘idle’ in the background
Expected outcome
Minimal CPU usage when app is idle, similar to what’s seen when using the ngl renderer
Version information
- Arch Linux | Kernel 6.11.5
- Gnome-shell | Mutter 47.1
- Wayland (meson_options: xwayland=false)
- Gtk4 4.16.3
- Mesa 24.2.5
nVidia 565.57.01- vulkan-icd-loader 1.3.295
Additional information
Here’s where things are spinning:
$ sudo strace -v -fp 9082
strace: Process 9082 attached with 11 threads
[pid 9094] openat(AT_FDCWD, "/proc/stat", O_RDONLY) = 42
[pid 9094] read(42, "cpu 34964 19 18870 9203976 5872"..., 299) = 299
[pid 9094] close(42) = 0
[pid 9094] clock_nanosleep(CLOCK_REALTIME, 0, {tv_sec=0, tv_nsec=5000000}, <unfinished ...>
[pid 9100] <... restart_syscall resumed>) = -1 ETIMEDOUT (Connection timed out)
[pid 9100] futex(0x6322a993e0c0, FUTEX_WAKE_PRIVATE, 1) = 0
[pid 9100] futex(0x7cab2c000f58, FUTEX_WAIT_BITSET_PRIVATE|FUTEX_CLOCK_REALTIME, 0, {tv_sec=1729865521, tv_nsec=735347000}, FUTEX_BITSET_MATCH_ANY <unfinished ...>
[pid 9094] <... clock_nanosleep resumed>0x7cab613ffb80) = 0
[pid 9094] openat(AT_FDCWD, "/proc/stat", O_RDONLY) = 42
[pid 9094] read(42, "cpu 34965 19 18870 9203988 5872"..., 299) = 299
[pid 9094] close(42) = 0
[pid 9094] clock_nanosleep(CLOCK_REALTIME, 0, {tv_sec=0, tv_nsec=5000000}, 0x7cab613ffb80) = 0
[pid 9094] openat(AT_FDCWD, "/proc/stat", O_RDONLY) = 42
[pid 9094] read(42, "cpu 34965 19 18870 9204001 5872"..., 299) = 299
[pid 9094] close(42) = 0
[pid 9094] clock_nanosleep(CLOCK_REALTIME, 0, {tv_sec=0, tv_nsec=5000000}, <unfinished ...>
[pid 9100] <... futex resumed>) = -1 ETIMEDOUT (Connection timed out)
[pid 9100] futex(0x6322a993e0c0, FUTEX_WAKE_PRIVATE, 1) = 0
[pid 9100] futex(0x7cab2c000f58, FUTEX_WAIT_BITSET_PRIVATE|FUTEX_CLOCK_REALTIME, 0, {tv_sec=1729865521, tv_nsec=745883000}, FUTEX_BITSET_MATCH_ANY <unfinished ...>
[pid 9094] <... clock_nanosleep resumed>0x7cab613ffb80) = 0
[pid 9094] openat(AT_FDCWD, "/proc/stat", O_RDONLY) = 42
[pid 9094] read(42, "cpu 34965 19 18870 9204014 5872"..., 299) = 299
[pid 9094] close(42) = 0
[pid 9094] clock_nanosleep(CLOCK_REALTIME, 0, {tv_sec=0, tv_nsec=5000000}, 0x7cab613ffb80) = 0
[pid 9094] openat(AT_FDCWD, "/proc/stat", O_RDONLY) = 42
[pid 9094] read(42, "cpu 34965 19 18872 9204025 5872"..., 299) = 299
[pid 9094] close(42) = 0
[pid 9094] clock_nanosleep(CLOCK_REALTIME, 0, {tv_sec=0, tv_nsec=5000000}, <unfinished ...>
[pid 9100] <... futex resumed>) = -1 ETIMEDOUT (Connection timed out)
nvidia-bug-report-565.57.01.log (16.3 MB)
1 Like
Many 565 packages are missing in theofficial nvidia debian12 repo: Index of /compute/cuda/repos/debian12/x86_64 /
2 Likes
565 is the first driver I have used that as made Alan Wake 2 totally playable. Huge frame time spikes when turning, going to mind place, switch to cut scenes are 95% gone. Performance as improved in general, possibly like 10 or so fps, but no science here yet. Can’t comment on vs Windows as don’t run it on main machine anymore. VRAM seemed to get freed and didn’t stack to 10GB and stay there.
Did have one hard lock up and restart on that game, but didn’t see that again.
HDR on Samsung QN90B now enables and sets colour space DCI-P3. Looks good in Alan Wake 2 but not in Cyberpunk. Need to test and look into that more to feedback. Still cannot enable it in Horizon Zero Dawn without getting a black screen.
TTY is working now on kernel 6.11.5-200. Was not before, but I didn’t properly delve into that.
My feeling is ther is some desktop lag with GSP on, but I need to explore this more.
1 Like
This driver has a regression with the support for 2nd monitor in DVI-to-HDMI on Linux/wayland/gnome.
This was a problem in an earlier driver version but fixed. It’s now unusable so I have to downgrade to 560.35
I just tested this with the original game exe and it does resolve the issue. I made a save and reloaded it with SFSE turned back on and the corruption returned.
Script extender exe: sfse_loader.exe
Normal exe: Starfield.exe
One possible wrinkle is that I use a mod loader (Limo, but Mod Organizer 2 does this too) to overwrite Starfield.exe with a renamed sfse_loader.exe to ensure that Steam loads SFSE directly.
For the debian packages nvidia-smi is not included in the current 565. I had also weird problems that some symlinks werent missing. apt reinstall did fix it. But I am not sure why that issue occured.
also nvidia-settings is not found by kde for some reason. im not sure what that cause is. But I was able to run it with /usr/lib/nvidia/current/nvidia-settings. Seems like nvidia-vaapi is working and also after some test no regression found yet.
3 Likes
I hadn’t reported the issue before as the drivers prior 565 don’t support dma-buf mmap’ing.
nvidia-bug-report.log.gz (1.6 MB)
To reproduce the issue:
- Install
virglrenderer>= 1.0.0. Packagevirglrendereron Arch Linux and Ubuntu - Install Virtio Vulkan ICD (Venus) >= 24.2.0. Package
vulkan-virtioon Arch Linux,mesa-vulkan-driverson Ubuntu - Install Vulkan tools. Package
vulkan-toolson Arch Linux and Ubuntu - Run
virgl_test_server --venusin a terminal - Run
vulkaninfoin another terminal. A GPU device named “Virtio-GPU Venus (NVIDIA …)” should be listed in addition of Nvidia GPU. - Run
vkcube --gpu_number <id of virtio-gpu>
Edit: I am using KDE on Wayland
Dying Light 2 crashes when enabling anything that has RayTracing.
Apparently it’s been an issue since the release of 560 driver and 565 didn’t fix it. 550 doesn’t have this issue according to people in Proton Github Issues
i don’t think it’s a vram issue , i have 4090 , some games like throne and liberty got 20% performance increase still behind windows but that’s a good performance boost , some games like God of war Ragnorak and stormgate and space marine 2 had 0 performance increase and still behind windows by 30-35% , stormgate has a directx11 render path and it’s on par with windows when using dx11 , i think nvidia should priortize the dx12 performance issue, as many games don’t have directx11 anymore .
Another feedback. With both and 555 and 560 sleep didn’t work well in my case. nvidia-suspend was enabled, but I was getting black screen usually after trying to wake up machine. 565 seems to be resolved the problem completely, machine wakes up without any problem.
2 Likes
Is problem with suspend-then-hibernate resolved with this release? Does somebody tried it? Still I’m not able to use rtcwake to suspend notebook for 20 seconds when NVidia dGPU is on.
[ 763.905469] NVRM: GPU 0000:01:00.0: PreserveVideoMemoryAllocations module parameter is set. System Power Management attempted without driver procfs suspend interface. Please refer to the 'Configuring Power Management Support' section in the driver README.
[ 763.905483] nvidia 0000:01:00.0: PM: pci_pm_suspend(): nv_pmops_suspend [nvidia] returns -5
[ 763.905898] nvidia 0000:01:00.0: PM: dpm_run_callback(): pci_pm_suspend returns -5
[ 763.905907] nvidia 0000:01:00.0: PM: failed to suspend async: error -5
[ 763.906250] PM: Some devices failed to suspend, or early wake event detected