oh sorry i thought we should open a ticket for every game my bad

Any suggestions about the issue though? @DSDV if you solve it let me know

System Information

• GPU: RTX 4080
• Video driver version: 4.6.0 NVIDIA 565.57.01
• SO: Arch Linux
• Kernel version: 6.12.1-arch1-1
• Link to full system information report as Gist: Hardware info
• Proton version: Proton experimental and ProtonGE 9.20

steam-2694490.log

Symptoms

As soon as the game starts, a sound is heard for a second, but then the sound is turned off. And the screen stays black since I start the game. After a while the game crashes without showing anything.

Reproduction

Start game

System Information

• GPU: NVIDIA GeForce RTX 4070
• Video driver version: 565.57.01
• Kernel version: 6.12.1
• Link to full system information report as Gist: https://gist.github.com/kittyandrew/8c2100d99b63982e54843a608b50e4f2
• Proton version: Proton experimental and GE-Proton9-20

steam-2694490.log

Symptoms

Game opens a window, screen stays black without any progress after 10 minutes.

Reproduction

Start game

Hello @kisak-valve, It has no effect, the same thing happens.

@kisak-valve unfortunately didn't help,
here is new log
steam-2694490.log

The issue seems to be related to wayland, the game launches just fine under X11

Edit: First launching the game in x11 and setting the renderer from vulkan to dx12 allows launching the game in wayland mode afterwards.

I had the same black screen issue, but am able to launch the game after editing game config to use DirectX12 instead of Vulkan.

Can be done by editing the following file:
~/.steam/steam/steamapps/compatdata/2694490/pfx/drive_c/users/steamuser/My Documents/My Games/Path of Exile 2/poe2_production_Config.ini

Changed renderer_type=Vulkan to renderer_type=DirectX12

~/.steam/steam/steamapps/compatdata/2694490/pfx/drive_c/users/steamuser/My Documents/My Games/Path of Exile 2/poe2_production_Config.ini

YES, thank you

Replying to https://github.com/ValveSoftware/Proton/issues/8296#issuecomment-2524200434

Works!!!, thank you

Replying to https://github.com/ValveSoftware/Proton/issues/8296#issuecomment-2524134603

There might be a wsi issue there as people reports Vulkan renderer of game works with X11 sessions on Nvidia while stays at black screen on Wayland sessions on Nvidia.

Thanks you for the reports and for taking steps to characterize the issue. For future reference, I've filed NV-internal bug 4999362 to track and investigate this issue.

From the client logs (logs folder next to the game executable), this looks like a Vulkan error not being handled by the game:
2024/12/07 09:35:44 8111398 db2b0f7b [INFO Client 340] [VULKAN] acquire: vk::Device::acquireNextImageKHR: ErrorOutOfDateKHR
This happens on launch when using Vulkan and after switching to Vulkan in-game (started as DX12) the same error appears sometimes when switching windows (other than that the Vulkan renderer works fine).
As per the Vulkan Spec, this should be handled by re-creating the swapchain

I can play just fine on but experience sudden performance degradation after about an hour where the framerate starts jumping up and down and the CPU frametime goes above 50ms if you move your character (stays low when you stop moving), only a restart seems to fix that.

I can play just fine on but experience sudden performance degradation after about an hour where the framerate starts jumping up and down and the CPU frametime goes above 50ms if you move your character (stays low when you stop moving), only a restart seems to fix that.

I had the same issue, enabling the steam overlay seems to have fixed this for me.

Thanks you for the reports and for taking steps to characterize the issue. For future reference, I've filed NV-internal bug 4999362 to track and investigate this issue.

Hello @cubanismo, I see that you are working on making Nvidia Linux drivers and that you have created a bug to fix this. If it helps, this same error has happened to me with BG3, Satisfactory and other games I don't remember. In these games, since the game was booting in Dirext12 and I was the one switching to Vulkan, I knew how to fix it.

The pattern I detected is the following: Wayland + Game under proton + Vulkan + Nvidia.

As I didn't know, until now, where the error came from, I didn't know where to put the bug.

I add this as possible information to detect the bug.

I had the same issue, enabling the steam overlay seems to have fixed this for me.

That seems to have worked indeed, much appreciated!

Can anyone launch the game (using Nvidia drivers) with gamescope? I am trying to running it with these options

DXVK_HDR=1 gamescope -f -w 2560 -h 1440 --hdr-enabled -- env VKD3D_DISABLE_EXTENSIONS=VK_KHR_present_wait %command%

in order to get HDR going. The renderer has been set to DX12 and the game launched without gamescope fine but won't run at all with gamescope. Using KDE Plasma & under Arch with 1080 GTX card.

I have the game running pretty well. Unfortunately, any time a video is playing (including the cutscene at the end of Act 1), it completely kills all sound output on my system. Linux Mint 21.3 with Xanmod 6.11.10-x64v3 kernel.
FWIW, I'm using the NVIDIA 550 drivers installed through apt on X11 with Cinnamon. I'm also using Pipewire/Pulse for my sound system, although I suspect that isn't related to this issue.
Oh, I'm also running Proton Experimental. May try one of the GE flavors and see what that does.

Edit: gamemode has, in fact, not fixed it. I'm not 100% sure what's causing it, but having a YouTube video playing on my second monitor seems to be part of it.

Game works with nvidia wayland if you start it in windowed mode.
Game renderer vulkan.

You can then set fullscreen via KDE window settings and it will work fine

We determined this is an issue we've already resolved in pending driver releases. In the meantime, could those who are running into the issue verify it does not occur with our Vulkan developer beta drivers? Note the first drivers on this page are our regular releases. The beta drivers are further down under an appropriate heading.

The fix should already be present there. However, please be aware these drivers are not meant for general use and have not gone through our usual QA process. They're intended primarily for developers who need to develop code using the latest Vulkan API features, and for verifying pending bug fixes, like this one.

Replying to https://github.com/ValveSoftware/Proton/issues/8296#issuecomment-2530429943

Yeah, game works fine on latest 550.40.81 vulkan dev drivers.
No DLSS support though

Been playing the game for a bit on my system and comparing it to Windows.
The framerate is about almost cut in half whenever visual effects happen. The loading screens are also almost doubled or tripled in length and the game initially seems to be extremely slow at streaming in effects and certain assets while its almost instantaneous on Windows.

This has mainly been through having DirectX12 renderer selected, but having Vulkan selected seems to produce the exact same results. Just tried it and it's still just as bad.

Running "NVIDIA RTX 2070" card with the "565.77" drivers on "Wayland" with the "6.12.3-arch1-1" kernel.

I am unable to get remote play to function properly with the game. I am running it with proton. Can this be a proton issue or is it something else? I am trying to stream to a Steam Deck. Remote play functions perfectly with other games.

EndeavourOS 6.12.3-arch1-1 kernel
Nvidia GTX 1070 (565.77 drivers) Plasma 6.2.4 X11
Ryzen 7 7800X3D

I can play just fine on but experience sudden performance degradation after about an hour where the framerate starts jumping up and down and the CPU frametime goes above 50ms if you move your character (stays low when you stop moving), only a restart seems to fix that.

I had the same issue, enabling the steam overlay seems to have fixed this for me.

Thank you, strangely this fixed the issue for me as well.

The ingame clock is off by one hour depending on the set timezone. The same is not seen on Windows. Tested on Proton Experimental (20241206).

To enable the ingame clock: ESC -> Options -> Interface tab -> Show Clock

$ date
Mo 16. Dez 17:09:21 CET 2024
$ localectl status
System Locale: LANG=en_US.UTF-8
               LC_NUMERIC=de_DE.UTF-8
               LC_TIME=de_DE.UTF-8
               LC_MONETARY=de_DE.UTF-8
               LC_PAPER=de_DE.UTF-8
               LC_MEASUREMENT=de_DE.UTF-8

The ingame clock is off by one hour depending on the set timezone. The same is not seen on Windows. Tested on Proton Experimental (20241206).

unset TZ; %command%

I suggest re-attempting to use the Vulkan Shaders. From what I can tell there was a hotfix released 12/21 that resolved the compatibility.

I'm using Proton-GE-9.21
20 Dec 2024 - I had to modify PoE2 rendering to use DirectX12
21 Dec 2024 - After hotfix, Vulkan Shaders was default again, but successfully loaded the game

My system reference:

• Fedora 40
• Proton-GE 9,21
• AMD 5950X
• Nvidia Quadro M2000
• Drivers nvidia-6.11.10-200

I suggest re-attempting to use the Vulkan Shaders. From what I can tell there was a hotfix released 12/21 that resolved the compatibility.

No luck, same behavior as pre hotfix.

there was a hotfix

There was no hotfix.

there was a hotfix

There was no hotfix.

Then please explain why:

• Steam required an update for PoE 2 this morning
• It didn't work yesterday
• It worked today

Replying to https://github.com/ValveSoftware/Proton/issues/8296#issuecomment-2558305125

Not every "update" is a "hotfix." The "updates" published on Steam are mostly just pre-cached shaders. The application version number did not increment. GGG did not announce a release (edit: linkify, this is where GGG puts patch notes). An actual Nvidia developer, on this thread, already indicated a driver issue.

Literally anything about your situation or your machine could be the reason for the difference between "it didn't work" and "it worked" for you because you did not qualify anything about your experience and because you provided no sources for your information at all. You might not have even been experiencing the same problem in the first place. You might not be using Wayland. Nobody knows.

Do not ask people to explain your own Russell's Teapot for you, and definitely please do not speak on behalf of potential solutions you are not the source of, or this thread will get real crowded real fast.

Is this thread appropriate to bring up potential AMD issues up in?

I'm experiencing shutdowns of my machine in this title and have been meticulously observing voltage, current, temperature and total draw of all components in an attempt to rule out potential hardware instability. I've yet to find anything indicative of hardware instability: I have limited RAM to default settings (36 timing, 4800MHz) and retrained it. I have disabled any board-specific "boost" via MSI, set strict limits on draw to Intel's manual spec for the 13600K (125W/181W), and brought the lite load down to a reasonable value (mode 5. I am indeed stable at mode 2, but either way this implies much less juice running through it and a much more even curve than the default). I've monitored temperatures and ensured that I never break 85C on my CPU.

My initial instinct was that my 7900XTX may be tripping my SF750 via transient, but I'm a little unconvinced: it still occurs at low resolutions and ultra perf scaling (some 160W constant draw).

I've found this occurs in both DX12 and Vulkan configs: confounding further.

I have an SF1000 around that I was planning on returning... but may slot in if necessary. I'm as yet unconvinced that's the case.

Looks like it may be related to the Mesa driver.

4:10:16 bazzite kernel: amdgpu 0000:03:00.0: [drm] REG_WAIT timeout 1us * 100 tries - dcn32_program_compbuf_size

The game often freezes when minimizing/switching the windows.

My specs:

OS: Ubuntu 24.10 x86_64 
Host: XPS 17 9700 
Kernel: 6.11.0-13-generic 
Resolution: 2560x1440 
DE: GNOME 47.0 
WM: Mutter 
Icons: ePapirus-Dark [GTK2/3] 
CPU: Intel i7-10750H (12) @ 2.600GHz 
GPU: Intel CometLake-H GT2 [UHD Graphics] 
GPU: NVIDIA GeForce GTX 1650 Ti Mobile 
Memory: 11758MiB / 31327MiB 

Driver version is 565.77. Tried 560, same issue. Before that I had the 550 driver, and there were no such freezes with it.

Using Vulkan renderer with X11 and windowed fullscreen. DirectX12 with Wayland/X11 doesn't work at all.

System Information

• GPU: 7900 XT
• Video driver version: Mesa 24.3.1-arch1.3
• Kernel version: 6.12.4-arch1-1
• Link to full system information report as Gist: https://gist.github.com/gelebflamefinger/e2854257d660f9cb3adff19d8c1db454
• Proton version: Proton Experimental

I confirm:

• [x] that I haven't found an existing compatibility report for this game.
• [x] that I have checked whether there are updates for my system available.

Symptoms

The in-game cursors have no filtering/anti-aliasing applied, leading to the cursors being very jaggy looking. This is not how the cursors look when playing on Windows. Using Wayland, present with both DX12 and Vulkan renderers.

Reproduction

Play the game, set the cursor to various sizes/types like high contrast small, observe aliasing.

The game runs at 50-60 FPS with choppy frame times after 20 mins of playtime in gamescope. Doesn't occur when running the game natively. EDIT: LD_PRELOAD="" fixes this.

MANGOHUD_CONFIGFILE=~/.config/MangoHud/MangoHud.conf gamemoderun gamescope --hdr-enabled --adaptive-sync --mangoapp -e -f -w 3840 -h 1600 -- %command%  --nologo

Proton-GE flatpak (com.valvesoftware.Steam.CompatibilityTool.Proton-GE)
7900XTX, 3700X

When running the game natively, alt-tab switches the game out of fullscreen or windowed mode. However I've been told this is an issue on Windows as well.

The game runs at 50-60 FPS with choppy frame times after 20 mins of playtime in gamescope. Doesn't occur when running the game natively. EDIT: LD_PRELOAD="" fixes this.

Out of interest, does Steam overlay work for you at all when running gamescope under that full flatpak environment? On my end the overlay toggle functionally works but its UI is invisible even though the FPS counter isn't. I'm wondering if it's then not somehow related to the previously mentioned stutters with disabled overlay. There's quite a few issue reports with gamescope on wayland (https://github.com/ValveSoftware/gamescope/issues/835), but I don't see them mentioning flatpak environment which is the only configuration where I experience it.

When running the game natively, alt-tab switches the game out of fullscreen or windowed mode. However I've been told this is an issue on Windows as well.

If you set windowed fullscreen and restart the game right after, the setting should stick and you'll avoid this behavior.

does Steam overlay work for you at all when running gamescope under that full flatpak environment

It does as long as I don't LD_PRELOAD="" in my launch options. If I leave it enabled (flatpak or not, I have both) the game stutters after 20/30 mins, so it's clearly some issue with Steam Overlay. Someone mentioned to me that it wasn't present before the "game recording" update.

If you set windowed fullscreen and restart the game right after, the setting should stick and you'll avoid this behavior

Thank you. I'll try this. I don't have this issue anymore though because I use gamescope.

Replying to https://github.com/ValveSoftware/Proton/issues/8296#issuecomment-2559523483

UPD: I managed to run the game with Wayland and DirectX12. Interestingly enough, it only runs under Proton 8.0-5. Every other version of Proton (including Experimental, Hotfix, GE, etc.) doesn't work, and the game crashes even before the logo is shown.

Wayland and DirectX12 eliminated the problem with freezes on alt-tabbing, though.

It does as long as I don't LD_PRELOAD="" in my launch options.

I have tried to appy this when running steam and no luck, get stutters after haln an hour. Did you also apply LD_PRELOAD to run steam (LD_PRELOAD="" steam) or did it somewhere else?

I have tried to appy this when running steam and no luck, get stutters after haln an hour. Did you also apply LD_PRELOAD to run steam (LD_PRELOAD="" steam) or did it somewhere else?

I added it to the game's launch options. Like LD_PRELOAD="" %command%

I added it to the game's launch options. Like LD_PRELOAD="" %command%

Can confirm this works for 7800XT, thank you so much for clarification

Replying to [#8296 (comment)](https://github.com/ValveSoftware/Proton/issues/8296#issuecomment-2559523483)

UPD: I managed to run the game with Wayland and DirectX12. Interestingly enough, it only runs under Proton 8.0-5. Every other version of Proton (including Experimental, Hotfix, GE, etc.) doesn't work, and the game crashes even before the logo is shown.

Wayland and DirectX12 eliminated the problem with freezes on alt-tabbing, though.

Here are two logs,

for Proton 8.0-5, which runs the game without issues:
steam-2694490-805.log

and for Proton 9.0-4, with which the game does not start:
steam-2694490-904.log

I believe the part of the log related to the error starts on line 3694 with:

wine: Unhandled page fault on read access to 0000000000000010 at address 00006FFFF766E892 (thread 0188), starting debugger...

And then some info about the error is on line 4085:

197329.258:0184:0188:warn:seh:dispatch_exception backtrace: --- Exception 0xc0000005.
197329.258:0184:0188:trace:seh:dispatch_exception code=c0000005 flags=0 addr=00006FFFF766E892 ip=6ffff766e892
197329.258:0184:0188:trace:seh:dispatch_exception  info[0]=0000000000000000
197329.259:0184:0188:trace:seh:dispatch_exception  info[1]=0000000000000010
197329.259:0184:0188:warn:seh:dispatch_exception EXCEPTION_ACCESS_VIOLATION exception (code=c0000005) raised

But I wasn't able to decipher that. There are no such lines in the Proton 8.0-5 log, though.

@kisak-valve Experimental indeed works. Thanks!

the mouse back and forward buttons are indistinguishable from horizontal scroll buttons, both are recognized as mouse 4 and 5 in-game.
this causes issues for me because the G502 X sends horizontal scroll events by tilting the scroll wheel, which causes scroll wheel presses to sometimes send horizontal scroll activating the wrong skill in game

I use LD_PRELOAD="" %command% with DX12 (Vulkan is extremely laggy, but works) and the game works completely fine.

There are only two issues:

• I can't disable triple buffering, the game would instantly re-enable it
• The game runs at max 60 fps even though I have a monitor that supports 120Hz (It's also definitely not a performance cap, since it applies even on lowest settings)

Not sure if any of these are a proton thing or not.
I only use Wayland, so not sure if there is a difference on x11.

One of the things that are asked of people to do, and I didn't was to verify the files. I read several posts on the forums talking about crashes on the same area, that didn't seem to have the same context, and some of them were fixed by verifying the files. It was indeed some kind of corruption. 14 files were affected.

22883.150:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
22883.158:02e4:02e8:trace:seh:register_module_exception_directory count 11, max_count 512.
22883.158:02e4:02e8:trace:loaddll:build_module Loaded L"C:\\windows\\system32\\imm32.dll" at 00006FFFFF460000: builtin
22883.761:02dc:02e0:fixme:dbghelp:elf_search_auxv can't find symbol in module
22883.874:0128:0160:trace:seh:call_handler handler at 00006FFFFFF82A00 returned 1
Unhandled exception: 0xe0000001 in 64-bit code (0x006fffffc1cf97).
22883.897:02dc:02e0:fixme:dbghelp:elf_search_auxv can't find symbol in module
Register dump:
 rip:00006fffffc1cf97 rsp:0000000006d2f150 rbp:0000000006d2f3c0 eflags:00000206 (   - --  I   - -P- )
 rax:0000000006d2f170 rbx:0000000000000016 rcx:0000000006d2f170 rdx:0000000000000000
 rsi:00000001400e77c0 rdi:0000000143bd7598  r8:0000000000000000  r9:0000000000000000 r10:c6c1864950d77370
 r11:0000000006d2f288 r12:0000000000000910 r13:0000000000000000 r14:0000000000000001 r15:0000000000040112
Stack dump:
0x00000006d2f150:  0000000006d2f170 0000000000000000
0x00000006d2f160:  0000000006d2f150 00006ffffff60f9a
0x00000006d2f170:  00000000e0000001 0000000000000000
0x00000006d2f180:  00006fffffc1cf97 0000000000000000
0x00000006d2f190:  0000000000000000 0000000000000910
0x00000006d2f1a0:  0000000006d2f3c0 00000001424e3e36
0x00000006d2f1b0:  0000000000000000 0001000100000000
0x00000006d2f1c0:  0000008000000007 0001000106d2f378
0x00000006d2f1d0:  0000000000000000 00006ffffff60f9a
0x00000006d2f1e0:  0000000000040112 00000001400e77d5
0x00000006d2f1f0:  0000000000000016 0000000143bd7598
0x00000006d2f200:  0000000000000001 0000000000000910
Backtrace:
=>0 0x006fffffc1cf97 in kernelbase (+0xcf97) (0x00000006d2f3c0)
  1 0x000001424e3d17 in pathofexilesteam (+0x24e3d17) (0x00000006d2f3c0)
  2 0x000001424da340 in pathofexilesteam (+0x24da340) (0x00000006d2f3c0)
  3 0x00000140d0de5e in pathofexilesteam (+0xd0de5e) (0x00000006d2f3c0)
  4 0x00000140d0df87 in pathofexilesteam (+0xd0df87) (0x00000006d2f550)
  5 0x00000140d0e290 in pathofexilesteam (+0xd0e290) (0x00000000a05eef)
  6 0x00000140cfee99 in pathofexilesteam (+0xcfee99) (0000000000000000)
  7 0x00000140d01eb7 in pathofexilesteam (+0xd01eb7) (0x00000006d2f700)
  8 0x00000140f71e41 in pathofexilesteam (+0xf71e41) (0x00000006d2f809)
  9 0x00000140f71fdd in pathofexilesteam (+0xf71fdd) (0x00000006d2f8f9)
  10 0x00000140f736d6 in pathofexilesteam (+0xf736d6) (0x00000981304a50)
  11 0x00000140f7b650 in pathofexilesteam (+0xf7b650) (0x00000981304a50)
  12 0x00000140ce64c0 in pathofexilesteam (+0xce64c0) (0x00000981304a50)
  13 0x00000140ce3dd1 in pathofexilesteam (+0xce3dd1) (0x00000006d2ffa0)
  14 0x000001401dd636 in pathofexilesteam (+0x1dd636) (0x00000006d2ffa0)
  15 0x000001401dcf71 in pathofexilesteam (+0x1dcf71) (0x00000006d2ffa0)
  16 0x00000140ce68b9 in pathofexilesteam (+0xce68b9) (0x00000006d2ffa0)
  17 0x006fffffed48bd in kernel32 (+0x148bd) (0x00000006d2ffa0)
  18 0x006ffffff4f4cf in ntdll (+0xf4cf) (0000000000000000)
0x006fffffc1cf97 kernelbase+0xcf97: add $0xc8, %rsp

I am getting crashes, somewhat randomly but regularly, can't even load one character in Clearfell villiage area as it always crashes.

Arch, nvidia-open-dkms 570.133.07-2, 6.14.2-2-cachyos, on bleeding edge Proton Experimental

1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.868:0128:012c:warn:vkd3d-proton:d3d12_command_list_QueryInterface: {5b2662fb-eb28-4aec-819e-1c1b4de060f6} not implemented, returning E_NOINTERFACE.
1054.872:0128:01b0:trace:seh:handle_syscall_fault code=c0000005 flags=0 addr=0x7835946de720 ip=7835946de720 tid=01b0
1054.872:0128:01b0:trace:seh:handle_syscall_fault  info[0]=0000000000000008
1054.872:0128:01b0:trace:seh:handle_syscall_fault  info[1]=00007835946de720
1054.872:0128:01b0:trace:seh:handle_syscall_fault  rax=0000000000000000 rbx=00007836b4d86f23 rcx=0000000000000024 rdx=0000783594b16b00
1054.872:0128:01b0:trace:seh:handle_syscall_fault  rsi=0000000000000024 rdi=0000000101f0b778 rbp=000078359620d980 rsp=0000000101f0b7c0
1054.872:0128:01b0:trace:seh:handle_syscall_fault   r8=0000000000000000  r9=00007835956c92c0 r10=00000000ffffffff r11=00007835969f6058
1054.872:0128:01b0:trace:seh:handle_syscall_fault  r12=0000000101f0ba70 r13=0000000000000000 r14=00007835949cb7c0 r15=00007835949cb7c0
1054.872:0128:01b0:warn:seh:handle_syscall_fault backtrace: --- Exception 0xc0000005 at 0x7835946de720:.
1054.872:0128:01b0:warn:seh:dump_syscall_fault backtrace: returning to user mode ip=00006ffffd16a21e ret=c0000005
1054.872:0128:01b0:err:msvcrt:_wassert (L"!status && \"vkCreateGraphicsPipelines\"",L"../src-wine/dlls/winevulkan/loader_thunks.c",3095)
1054.872:0128:01b0:trace:seh:raise (22)
1054.875:0128:01b0:info:nvapi64:<-NvAPI_DRS_DestroySession: OK
1054.875:0128:01b0:info:nvapi64:<-NvAPI_D3D12_DestroyCubinComputeShader: OK
1054.880:00e8:0410:warn:threadname:NtSetInformationThread Thread renamed to L"wine_threadpool_worker"
1055.366:0030:0414:warn:threadname:NtSetInformationThread Thread renamed to L"wine_threadpool_worker"
1055.366:0030:0418:warn:threadname:NtSetInformationThread Thread renamed to L"wine_threadpool_worker"
1055.366:0030:041c:warn:threadname:NtSetInformationThread Thread renamed to L"wine_threadpool_worker"
1055.366:0030:0420:warn:threadname:NtSetInformationThread Thread renamed to L"wine_threadpool_worker"
pid 7268 != 7267, skipping destruction (fork without exec?)

Edit: I guess I am in the same boat as @braiam, tried various different Proton versions and it all eventually crashes, usually when changing maps especially going back to Clearfell Encampment.

Here are last 1000 lines of the latest crash:
https://termbin.com/dgo4

Excerpt before crash

3933.752:03a8:03ac:trace:seh:register_module_exception_directory count 11, max_count 512.
3933.752:03a8:03ac:trace:loaddll:build_module Loaded L"C:\\windows\\system32\\imm32.dll" at 00006FFFFF460000: builtin
3934.330:03a0:03a4:fixme:dbghelp:elf_search_auxv can't find symbol in module
3934.356:0184:020c:trace:seh:call_handler handler at 00006FFFFFF829E0 returned 1
Unhandled exception: page fault on read access to 0xfffffffffffffff8 in 64-bit code (0x006ffffff595cb).

Edit2: Eventually I got a stable session, at first it seemed that removing gamemoderun helps but even after adding it back I didn't get any more crashes, Proton Experimental (bleeding-edge, build 18119139)

Performance are VERY poor with last PoE 2 update. heavy frame drops most of time 1-10 FPS!

Edit:
I tired Proton-GE, Proton-CachyOS, Proton-EM all are the same, in windows(same machine) its very stable and playable.

And here is a vid:

https://github.com/user-attachments/assets/cbacd39c-a7d9-4f58-8231-70807b43e177

i just cuts the vid, it keep drop frames and 1-10FPS

Performance are VERY poor with last PoE 2 update. heavy frame drops most of time 1-10 FPS!

Please add your system information and used Proton version.
I do not see poor performance on my personal AMDGPU systems and Proton Experimental.

Performance ok here on an AMD RX 9060 XT 16 GB card but HDR is greyed even though I have launched the game with gamescope etc.

DXVK_HDR=1 gamescope -f -w 2560 -h 1440 --hdr-enabled --force-grab-cursor --adaptive-sync -- env %command%

using Proton experimental.

@Mershl @afettouhi updated my comment

@CorrM did you switch to DirectX12 in graphic settings? vulkan works worse for me (but I am on AMD GPU build)

@CorrM did you switch to DirectX12 in graphic settings?

when i try to do that i get a blackscreen

I have Linux Mint 22.2 Zara fresh installed (Kernel 6.14.0-29)

Secure Boot and Nvidia Drivers 580.65.06 are installed fine

(Lutris with Battle.net and Diablo 4 works fine with Proton-GE 10-9)

Now i have fresh installed Steam and Downloaded Path of Exile 2. When i start Path of Exile 2 i ran into Black Screen (Mousecursor from POE2 is shown), but Game crashes.

Any Ideas?

steam-2694490.log

:Update: Works now, i have changed in Steam -> Steam Play force Proton 10 for Playing this Game and it works now.

Same symptom here, with NVIDIA 580.82.07 / openSUSE Tumbleweed 20250916 / KDE Plasma:

Wayland: Works fine With Proton 9 stable, Proton 10 Beta and Proton Experimental / Proton Bleeding Edge
X11: Proton 9 stable and Proton 10 Beta work fine, but Proton Experimental / Proton Bleeding Edge cause black screen on startup.

EDIT: Starting the game in windowed mode makes it work with X11 + Proton Experimental / Proton Bleeding Edge, so it might be related to #8812

@addohm Your issue (which, by the way, is present only when using the Vulkan renderer) is caused by dxvk-nvapi always reporting Vulkan Reflex as supported after https://github.com/jp7677/dxvk-nvapi/pull/308 to work around some games having a purple/pink tint when DLSS Frame Generation was enabled (like DOOM: The Dark Ages, as seen here). However, because PoE2 now also sees VkReflex as supported even if it isn't, it's trying to use Reflex's builtin FPS limiter which simply does nothing.

While we figure out a reasonable solution for this conflict so both games work as closely to "as expected" out of the box, PoE2's FPS limiter currently not working can be fixed with some manual tweaking using either of two workarounds below:

• Install and enable dxvk-nvapi's Vulkan Reflex layer for this game by launching it with DXVK_NVAPI_VKREFLEX=1 %command% (see linked document for installation instructions). This will fix both frame limiter and hopefully cause Reflex to actually reduce latency, but requires installing a new Vulkan layer on the host Linux system.
• Launch the game with DXVK_NVAPI_DISABLE_ENTRYPOINTS=NvAPI_Vulkan_InitLowLatencyDevice %command%. This pretty much reverts the change we applied to fix DOOM TDA, causing Reflex to be reported as not implemented and the game to fall back to other, functional means of framerate limiting. While easier to apply because it doesn't rely on having the layer, it will prevent the game from benefitting from Reflex's latency reduction.

Anyone else having very long verification times after patches?

The game is on a separate fast nvme drive which is not overheating (though smart data does export a 2nd temp sensor that always reports 82˚C). The CPU is mostly idling (load average: 4.16, 4.05, 4.02).
Steam shows disk usage jumping to ~280MB/s initially then it just keeps dropping, then jumps to ~25MB/s occasionally then decreasing down to ~15MB/s over next few minutes, on repeat. The last patch (last sunday I believe) took about an hour to finish.

CachyOS 7.0.10-2-cachyos
i9 13900kf
AMD RX 7900XTX
64G DDR5
Disk in question is KINGSTON FURY SFYRDK2000G 2TB
Steam Beta 1780352834

Update:
Same behavior on 6.18.33-2-cachyos-lts

Update2:
Booted windows and updated the firmware, then run a btrfs full balance in case this was because of charge bleed and the remaining verification finished in few minutes - misbehaving drive I guess.

There is an issue with Proton Experimental and Exiled Exchange 2 (3rd party in-game tool for POE2).
The tool works (as far as I know) by reading the clipboard and as far as I understand it can not read the clipboard when using Proton Experimental. With Proton 10.0-4 it works perfectly fine.