Can you maybe try this with amdvlk as well? GPU hangs are often vulkan driver bugs.
Same issue here with AMDVLK.
Jän 12 00:43:06 domi-pc net.lutris.Lutris.desktop[3841]: debug: Compiling graphics pipeline...
Jän 12 00:43:06 domi-pc net.lutris.Lutris.desktop[3841]: debug: vs : VS_1a586b378d63907f6c6cf4c8913ca281a5d37758
Jän 12 00:43:06 domi-pc net.lutris.Lutris.desktop[3841]: debug: fs : FS_b4806b860cdbaf2975cf4f8b12b95d0d000f379b
Jän 12 00:43:06 domi-pc net.lutris.Lutris.desktop[3841]: debug: attr 0 : location 0, binding 0, format VK_FORMAT_R32G32_SFLOAT, offset 0
Jän 12 00:43:06 domi-pc net.lutris.Lutris.desktop[3841]: debug: attr 1 : location 1, binding 0, format VK_FORMAT_R32G32_SFLOAT, offset 8
Jän 12 00:43:06 domi-pc net.lutris.Lutris.desktop[3841]: debug: binding 0 : binding 0, stride 20, rate 0, divisor 0
Jän 12 00:43:12 domi-pc NetworkManager[1313]: <info> [1673480592.9312] manager: NetworkManager state is now CONNECTED_SITE
Jän 12 00:43:12 domi-pc systemd[1]: Starting NetworkManager-dispatcher.service - Network Manager Script Dispatcher Service...
Jän 12 00:43:13 domi-pc NetworkManager[1313]: <info> [1673480593.2790] manager: NetworkManager state is now CONNECTED_GLOBAL
Jän 12 00:43:16 domi-pc kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, signaled seq=310421, emitted seq=310423
Jän 12 00:43:16 domi-pc kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process eldorado.exe pid 10431 thread eldorado.exe pid 10450
Jän 12 00:43:16 domi-pc kernel: amdgpu 0000:0c:00.0: amdgpu: GPU reset begin!
Jän 12 00:43:17 domi-pc kernel: amdgpu 0000:0c:00.0: [drm:amdgpu_ring_test_helper [amdgpu]] *ERROR* ring kiq_2.1.0 test failed (-110)
Jän 12 00:43:17 domi-pc kernel: [drm:gfx_v10_0_hw_fini [amdgpu]] *ERROR* KGQ disable failed
Jän 12 00:43:17 domi-pc kernel: [drm:gfx_v10_0_hw_fini [amdgpu]] *ERROR* failed to halt cp gfx
Jän 12 00:43:17 domi-pc kernel: [drm] free PSP TMR buffer
Jän 12 00:43:17 domi-pc kernel: amdgpu 0000:0c:00.0: amdgpu: MODE1 reset
Jän 12 00:43:17 domi-pc kernel: amdgpu 0000:0c:00.0: amdgpu: GPU mode1 reset
Jän 12 00:43:17 domi-pc kernel: amdgpu 0000:0c:00.0: amdgpu: GPU smu mode1 reset
Jän 12 00:43:17 domi-pc kernel: amdgpu 0000:0c:00.0: amdgpu: GPU reset succeeded, trying to resume
The apitrace replays just fine on the Nvidia driver and there are no notable validation errors. Please open a Mesa bug report.
Done, opened an issue here: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8067
@DomiStyle Even if the mesa issue is still open could i get you to try this again with dxvk 2.3 and latest driver?
@Blisto91 Tried with DXVK 2.3 and Mesa 23.1.6 but unfortunately GPU still hangs on start.
Issue appears to be fixed with Mesa 24.0.3 and later.
ge-proton7-35x1 2023-01DXVK_LOG_PATHx1 2023-01RADV_TRACE_FILEx1 2023-01
When starting Eldewrito 0.6.1 (aka Halo Online) with DXVK the game will freeze at the start screen for a few seconds before the X11 session crashes.
Sometimes (especially on Wayland) the game will start but performance will be abysmal and looking at the syslog reveals a bunch of soft resets every 20-30s.
The game launches fine with Gallium Nine and WineD3D, albeit with some graphical and performance issues.
I can provide download links for the game files since it's a hassle to patch them on Linux.
Software information
Eldewrito 0.6.1, default settings
System information
Apitrace file(s)
Apitrace with Gallium Nine:
https://mega.nz/file/vFoVAaQK#bx_dbeqZw8EbZLbuD065kWmi2lboYmnaQFxIZyByr58
Log files
I tried to get DXVK debug logs and RADV trace files but none are created even though
DXVK_LOG_PATHandRADV_TRACE_FILEare set. Just tell me if they (or something else) are needed and I will take another look.syslog for the amdgpu timeout:
halo_crash.txt