protonscr

[RDNA4/gfx12] Page faults in Marvel Rivals causing crash/freeze (RX 9060 XT)

vkd3dopen
HansKristian-Work/vkd3d-proton#2933 · opened 2026-04-11 by dudleybarbosa · updated 2026-04-11 · 2 comments · github
Ddudleybarbosa 2026-04-11 github

I'm having a lot of crashs and freezes in Marvel Rivals. Sometimes, even the REISUB can't unfreeze me, and i hold the power button.
The IAs said my GPU is too recent, and that's why I'm in trouble.
(After a lot of tries,) Claud sugested me to use the
sudo journalctl -f | grep -i "amdgpu|reset|hang|timeout"
, and that's the log.
(Claud saig it could be useful to solve the issue, that's why I'm here.)
Log:

dudu@HAL9000:~$ sudo journalctl -f | grep -i "amdgpu|reset|hang|timeout"
[sudo] senha para dudu:
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000080c10232e000 from client 10
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00600C30
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CPG (0x6)
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x0
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x3
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:48:55 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ee34000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00600031
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CB (0x0)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x3
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ede5000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ee43000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015edf5000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ee24000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00600031
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CB (0x0)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x3
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015edf5000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00600031
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CB (0x0)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x3
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ee34000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00600031
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CB (0x0)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x3
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ee04000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00640051
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CB (0x0)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x5
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ee04000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00640051
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CB (0x0)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x5
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:6 pasid:32807)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Process GameThread pid 60881 thread vkd3d_queue pid 61477
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x000000015ee33000 from client 10
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00640051
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CB (0x0)
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x5
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x0
abr 11 15:55:52 HAL9000 kernel: amdgpu 0000:03:00.0: amdgpu: RW: 0x1
^C
dudu@HAL9000:~$
.

Software information

Marvel Rivals (Steam App ID: 2767030)
Command: DISABLE_FOSSILIZE=1 gamemoderun FSR4_UPGRADE=1 %command%
Sorry. I wasn't using the protonlog. I'll use it.

System information

  • GPU: AMD RX 9060 XT 16GB (device ID 0x7590)
  • Driver: amdgpu, Mesa 25.2.8
  • Kernel: 6.17.0-20-generic
  • VKD3D-Proton version: I didn't especificated one.
  • OS: Ubuntu Studio 24.04.4
Mmbriar 2026-04-11 github

First of all you should probably update your Mesa version. 25.2.x is pretty old and it is quite likely that this is a driver bug that is already fixed in newer versions. You can use kisak ppa on ubuntu.

Ddudleybarbosa 2026-04-11 github

Update: Issue Resolved & Performance Boost! But I'm still crashing. :(
Context & Performance:
After upgrading to Mesa 26.0.4~kisak1 on Ubuntu Studio, I am happy to report that the RX 9060 XT is performing significantly better than before. The frame delivery is smoother, and the new driver architecture (GFX12) feels much more optimized for high-end gaming.

Desktop Restoration:
I initially faced some UI issues where icons and dependencies were removed during the full-upgrade. This wasn't a bug in Mesa itself, but a package conflict. With some AI-assisted troubleshooting, I reinstalled ubuntustudio-desktop and the icon themes. The system is now 100% stable and looking great. Gemini helped a lot in this fase.

The Unreal Engine 5 "Access Violation" Fix:
I encountered a crash in Marvel Rivals when returning to the character selection screen. Below is the log for reference:

Plaintext
Fatal error!
Unhandled Exception: EXCEPTION_ACCESS_VIOLATION reading address 0xffffffffffffffff
Crash in runnable thread Foreground Worker #0
The Solution:
This was a memory mapping issue (typical for UE5 on Linux). Increasing the vm.max_map_count solved it instantly.

Steps to optimize and fix:

Shader Cache Reset: rm -rf ~/.cache/mesa_shader_cache (essential after a driver jump).

Memory Limit Increase:

Bash
echo 'vm.max_map_count=2147483642' | sudo tee -a /etc/sysctl.conf && sudo sysctl -p
Final Verdict:
The performance gain with Mesa 26 is worth the initial setup tweaks. Everything is running flawlessly now. Thanks to the community and AI collaboration for the support!

Addendum: Hard Freeze Encountered

I experienced one "Hard Freeze" during gameplay where the character kept moving, audio looped, and the system became completely unresponsive (even Magic SysRq/REISUB failed). This required a hard power reset. Since it didn't leave a trace in journalctl, it might be a rare sync issue between the new Mesa 26 and Kernel 6.17 on GFX12 hardware. I will monitor if this recurs after the vm.max_map_count fix.

[Yes, I needed Gemini to help me to elaborate the text. You know. I'm not a native english speaker]

Launch lines