Hello, please let me know if this isn't helpful or if I should submit my own separate issue but I've been encountering the same issues seemingly as this one. I tried all of the possible fixes I could find like various launch options and many of the things KingOfStrangeness mentioned or linked here. I get the same sort of error readout and it happens very randomly, I can sometimes go 6+ hours without encountering this problem at all, other times I'll crash 3 times in one hour. Here's my dmesg log for the amdgpu parts:

[ 3956.836281] amdgpu 0000:06:00.0: [gfxhub] page fault (src_id:0 ring:24 vmid:7 pasid:392)
[ 3956.836286] amdgpu 0000:06:00.0: Process DD2.exe pid 56325 thread vkd3d_queue pid 56461
[ 3956.836289] amdgpu 0000:06:00.0: in page starting at address 0x00008000e4200000 from client 0x1b (UTCL2)
[ 3956.836292] amdgpu 0000:06:00.0: GCVM_L2_PROTECTION_FAULT_STATUS:0x00701430
[ 3956.836293] amdgpu 0000:06:00.0: Faulty UTCL2 client ID: SQC (data) (0xa)
[ 3956.836295] amdgpu 0000:06:00.0: MORE_FAULTS: 0x0
[ 3956.836296] amdgpu 0000:06:00.0: WALKER_ERROR: 0x0
[ 3956.836298] amdgpu 0000:06:00.0: PERMISSION_FAULTS: 0x3
[ 3956.836299] amdgpu 0000:06:00.0: MAPPING_ERROR: 0x0
[ 3956.836300] amdgpu 0000:06:00.0: RW: 0x0
[ 3958.846953] amdgpu 0000:06:00.0: Dumping IP State
[ 3958.851241] amdgpu 0000:06:00.0: Dumping IP State Completed
[ 3958.851368] amdgpu 0000:06:00.0: [drm] AMDGPU device coredump file has been created
[ 3958.851370] amdgpu 0000:06:00.0: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
[ 3958.851373] amdgpu 0000:06:00.0: ring gfx_0.0.0 timeout, signaled seq=1319039, emitted seq=1319042
[ 3958.851375] amdgpu 0000:06:00.0: Process DD2.exe pid 56325 thread vkd3d_queue pid 56461
[ 3958.851378] amdgpu 0000:06:00.0: Starting gfx_0.0.0 ring reset
[ 3958.852542] amdgpu 0000:06:00.0: Ring gfx_0.0.0 reset succeeded
[ 3958.852544] amdgpu 0000:06:00.0: [drm] device wedged, but no recovery needed

editing to add specs:

Hardware / software
GPU: AMD Radeon RX 6600 XT
OS: CachyOS
Desktop: KDE Plasma
CPU: Ryzen 5 5600X
RAM: 16 GB
Game: Dragon's Dogma 2
Game version: Update 3.2
Proton: tried many versions, mostly using GE proton 11.7, all versions I've tried have experienced this same crash though.

Just let me know if there's anything else I should test or if I should make my own post etc. From what I gather this has been an issue since monster hunter wilds though and doesn't seem to have any sort of permanent fix yet?

Please upload a Proton log and which graphics settings you're using when these hangs happen. If you've noticed any patterns like particular in-game locations or activities that trigger it, or if it's just random, that's useful as well.

Thanks. I'll provide the Proton log and the graphics settings I'm using.

The hangs are not tied to one specific action as far as I can tell. They appear to be somewhat random, although they've become much more frequent after reaching Battahl. When the hang occurs, both monitors lose signal and the GPU appears to reset/crash rather than the game simply closing normally.

The relevant kernel error I've captured is:

amdgpu: ring gfx_0.0.0 timeout
[gfxhub] Page fault observed
Faulty page starting at address: 0x0000000000000000
process_name: DD2.exe
thread: vkd3d_queue

The kernel then attempts a ring reset, which fails, followed by a MODE1 GPU reset. VRAM is subsequently reported as lost.

I'm running an RX 9070 XT 16 GB with RADV on CachyOS.

My graphics settings are set to the highest available settings at 1080p, targeting 60 FPS. I am not using any upscaling or ray tracing. If there are any specific graphics settings you'd like me to provide beyond that, I can list them.

As for reproducibility, I haven't found a reliable sequence that triggers it. Normally it happens after a few hours of gameplay, if I'm lucky. However, after trying to reach Battahl via both the coast route and the caves, it has started happening much more frequently, to the point where I currently can't progress. During these attempts, it usually happens within 5–10 minutes, if I'm lucky.

It has happened during combat, while simply wandering around, and while talking to pawns, so at the moment I'd describe it as random rather than being tied to a particular activity.

steam-2054970.log

Additional information: I have now reproduced a very similar GPU hang in Elden Ring as well, so this does not appear to be exclusive to Dragon's Dogma 2.

The Elden Ring crash produced the following kernel messages:

amdgpu: ring gfx_0.0.0 timeout
Process eldenring.exe pid ... thread vkd3d_queue
Starting gfx_0.0.0 ring reset
Ring gfx_0.0.0 reset failed
GPU reset begin
...
MODE1 reset
GPU reset succeeded
VRAM is lost due to GPU reset

After the reset, amdgpu also reported GFXHUB page faults involving kwin_wayland.

This makes me suspect there may be a broader issue somewhere in the RX 9070 XT / RADV / amdgpu / vkd3d-proton stack rather than this being specific to Dragon's Dogma 2. I don't want to assume which component is responsible yet, but the fact that both games hit the gfx ring timeout through vkd3d_queue seems relevant.

The Elden Ring crash occurred during normal gameplay as well.

Pasted markdown(7).md