steam-526870.zip

https://github.com/ValveSoftware/Proton/assets/46057382/2f435b1c-ae6d-4d73-9b7c-72c4a0128623

RADV_PERFTEST=gsfastlaunch2

Hello @alf1e, can you test if setting the game's launch options to RADV_PERFTEST=gsfastlaunch2 %command% has an effect? If that helps, then the graphical artifacting is a 760M specific issue which is fixed in mesa 24.0.2 and newer (https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27618).

That partially worked, However textures such as the conveyor belts are still missing and it will not let me move my mouse all the way left until I ALT+TAB out and back into the game. I will update Mesa now but please set me know if you have any idea how to fix the other issues.

@kisak-valve Nevermind. Got into an actual game and it was still flickering with artefacts.

Screencast from 2024-03-17 16-35-39.webm

@alf1e isn't there an UE VCRuntime library to install (on the Steam Deck it works and it gets installed when launching the game the first time but not on Steam desktop), I cannot get the game to run no matter the parameters (I'm using an AMD RX780M) on NixOS.

@alf1e isn't there an UE VCRuntime library to install (on the Steam Deck it works and it gets installed when launching the game the first time but not on Steam desktop), I cannot get the game to run no matter the parameters (I'm using an AMD RX780M) on NixOS.

Not sure. All I know is Fedora 40 had a new Mesa version and that fixed my issues.

Following https://github.com/ValveSoftware/Proton/blob/proton_9.0/docs/DEBUGGING.md, I attempted to attach a debugger to find out what the exact error from vkWaitForFences might be. I was not able to do so, unfortunately -- vkWaitForFences does not appear to be symbolized.

That said, from https://registry.khronos.org/vulkan/specs/1.3-extensions/man/html/vkWaitForFences.html:

On success, this command returns
- VK_SUCCESS
- VK_TIMEOUT

However, the generated code in loader_thunks.c is:

VkResult WINAPI vkWaitForFences(VkDevice device, uint32_t fenceCount, const VkFence *pFences, VkBool32 waitAll, uint64_t timeout)                                                                                                                                                     
{
    struct vkWaitForFences_params params;
    NTSTATUS status;
    params.device = device;
    params.fenceCount = fenceCount;
    params.pFences = pFences;    
    params.waitAll = waitAll;
    params.timeout = timeout;
    status = UNIX_CALL(vkWaitForFences, &params);      
    assert(!status && "vkWaitForFences");      
    return params.result;
}

Which treats any nonzero value as an error. In fact vk.xml specifies that either VK_SUCCESS or VK_TIMEOUT should count as a success for this function, but make_vulkan does not appear to make any use of this information. Since the physical value of VK_TIMEOUT is 2, not 0, the code raises a spurious assertion failure if the wait times out.

Previously to this driver update Linux only supported implicit sync, which would IIUC never have timed out.

I therefore believe this is a bug in winevulkan.dll, not in Satisfactory.

EDIT: ...not anymore. The result is returned out-of-band, so it's some other status check that's failing, and the return value from vkWaitForFences is just returned as-is. Right, well, I'm clearly out of my depth here then. I don't see anything relevant in the log.

It just means a segfault happened on the other side.

Makes... sense. The log file does seem to be mostly crash logs.

FWIW I'm unable to reproduce this on NVIDIA, one of my machines ran the game fine (555.54.02 driver, Kwin 6 beta, and Xwayland 24.1) and the other spins at the loading screen indefinitely as the game starts.

Since this is suspected to be related to explicit sync I wonder if waiting on the sync fds within vkWaitForFences is going wrong, causing the error to be returned. If so it might be dependent on the kernel, I know there are some fixes for syncobjs although those should all be in 6.8.x kernels. Is anyone else reproducing this (on NVIDIA) easily and if so do you notice if a newer/older kernel has any effect?

I'm also no longer able to reproduce. It works fine at the moment.

I can still reproduce the vkWaitForFences error message with the following setup:
Arch linux
Kernel 6.9.10
Nvidia driver 555.58.02
Proton 9.0, Proton experimental, Proton GE (did not try any others)
DE: Sway
Satisfactory EGS version via Heroic Launcher

Will try again with kernel 6.10 and nvidia 560 beta tomorrow and edit this message if I remember to

Edit: Same error message with kernel 6.10 and nvidia 560.28.03

I was able to reproduce a few days ago on Arch Linux, kernel 6.10.2-arch1.1, nvidia-dkms (proprietary) 555.58.02, Proton Experimental, KDE Plasma 6.1, Satisfactory running in Vulkan through Steam.

Update: I can get into the game on driver 560.35.03 with egl-wayland 1.1.16. :tada: Tested with Proton GE. I am no longer experiencing the vkWaitForFences issue.

Edit: vkWaitForFences just happened again, but I tried to move the window while the game was loading or something. I avoided doing that and got into the game fine.

just a 'me too' regarding vkWaitForFences.
nonvidia 560.35.03, 6.10.7-amd64, KDE Plasma, Wayland, any proton, same result.

1.0 got just released. Same problem on Proton Experimental:

as a side note, this seems to be related to having previously used the EXPERIMENTAL vulkan support.

deleting steamapps/compatdata/526870/pfx/drive_c/users/steamuser/Local Settings/Application Data/FactoryGame/Saved/Config/Windows/Engine.ini will restore graphics settings to default, and game will launch correctly in DX12 mode.

I also have the assertion failed on exit on my system but the game works (tested Dx11 and Vulkan)
Proton experimental, Kernel Debian 6.7.12-1~bpo12+1, NVidia 550.107.02.

tried ofcaah idea and detection is now ok

I still play with Vulkan rendering which proved to be far faster on No Man's Sky and Satisfactory here.

I still play with Vulkan rendering

I'm having same error if I switch to Vulkan; only DX12 works for me.

The vkWaitForFences issue seems to be tied to nvidia & explicit sync along with the game running using Vulkan (-vulkan) under Wayland. I stopped my game from crashing by passing the __NV_DISABLE_EXPLICIT_SYNC=1 environment variable. Alternatively, running the game in -dx12 mode also fixes the issue, but at the cost of lost performance.

Note that this needs a fairly recent version of egl-wayland to work (1.1.15+) as this functionality was added recently.

Edit: This workaround seems to have made the crashes less frequent, but they still seem to happen from time to time. DX12 is the only reliable option right now, which is not even close as performant. :/

[…]
Edit: This workaround seems to have made the crashes less frequent, but they still seem to happen from time to time. DX12 is the only reliable option right now, which is not even close as performant. :/

>50h playing now, only having to click "cancel" button after exiting the game, and I have no performance issues. Button vs performance. My choice is made until it's resolved :)

Edit on 22/09 :
Kernel 6.7 and NVidia 550.107.02 as NVidia asked to wait for driver fully compatible with newer kernels. Today's stable version is 550.120 which only fixes issues which doesn't affect my system.
PROTON_ENABLE_NVAPI is always 1 as I set it at login
The exception I get isn't a bad exit status with vkWaitForFences but vkDestroySampler. (a function might return 0 instead of 1 or the opposite to say "true" I hate numeric value for booleans :laughing: )

same issue here, i get assertion failed "!status" & "vkWaitForFences"" when i start the game.
proton: GE-Proton9-13, proton bleeding-edge | kernel: Gentoo 6.10.11 | nvidia: 560.35.03
explicit sync disabled and on Hyprland (wayland)

game works fine on dx11/dx12

Edit:
I switched to gnome from hyprland, nothing else was changed and now instead of crashing on startup it crashes when i exit. Atleast i can play now!

After the latest game update the game did not launch on Proton 9.0-3. (But not sure if the issue is proton or the game...) Only error appearing in log is, but not sure if relevant:

Adding process 213085 for gameID 526870
Adding process 213086 for gameID 526870
Adding process 213087 for gameID 526870
Adding process 213239 for gameID 526870
Traceback (most recent call last):
  File "/mnt/HDD2/Pelit/Steam/steamapps/common/Proton 9.0 (Beta)/proton", line 228, in copyfile_reflink
    bytes_to_copy -= copy_file_range(src.fileno(), dst.fileno(), bytes_to_copy)
OSError: [Errno 18] Invalid cross-device link

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/mnt/HDD2/Pelit/Steam/steamapps/common/Proton 9.0 (Beta)/proton", line 1601, in <module>
    g_session.init_session(sys.argv[1] != "runinprefix")
  File "/mnt/HDD2/Pelit/Steam/steamapps/common/Proton 9.0 (Beta)/proton", line 1508, in init_session
    g_compatdata.setup_prefix()
  File "/mnt/HDD2/Pelit/Steam/steamapps/common/Proton 9.0 (Beta)/proton", line 976, in setup_prefix
    try_copy(g_proton.lib64_dir + "wine/nvapi/nvapi64.dll", "drive_c/windows/system32",
  File "/mnt/HDD2/Pelit/Steam/steamapps/common/Proton 9.0 (Beta)/proton", line 161, in try_copy
    copyfile(src, dst)
  File "/mnt/HDD2/Pelit/Steam/steamapps/common/Proton 9.0 (Beta)/proton", line 234, in copyfile_reflink
    shutil.copyfile(srcname, dstname)
  File "/usr/lib/python3.9/shutil.py", line 275, in copyfile
    _fastcopy_sendfile(fsrc, fdst)
  File "/usr/lib/python3.9/shutil.py", line 172, in _fastcopy_sendfile
    raise err
  File "/usr/lib/python3.9/shutil.py", line 152, in _fastcopy_sendfile
    sent = os.sendfile(outfd, infd, offset, blocksize)
OSError: [Errno 5] Input/output error: '/mnt/HDD2/Pelit/Steam/steamapps/common/Proton 9.0 (Beta)/files/lib64/wine/nvapi/nvapi64.dll' -> '/mnt/SSD2/Pelit/Steam/steamapps/compatdata/526870/pfx/drive_c/windows/system32/nvapi64.dll'
Removing process 213239 for gameID 526870
Removing process 213087 for gameID 526870
Removing process 213086 for gameID 526870
Removing process 213085 for gameID 526870

Got it to launch as it used to by switching to Proton Experimental though. The above errors do not display then.

I'm having the same issues (on a AMD GPU) but only if I use Vulkan, sometimes it forces dx11 and works, other times some changing some settings to worse and better fixes it, then dx12 causes random VRAM problem like issues when AMD FSR is enabled, when its off it works perfectly, dx11 has the worst performance but doesn't have issues. Also the issues on Vulkan aren't from proton/wine, a different machine running Windows 11 has the same issue only on Vulkan.

i solved the problem, the issue was in keyboard layout - satisfactory used non-standart keyboard layout. In final, i just swapped default keyboard layout to eng, and all worked.

Since the 1.0 update, I have been getting infrequent long (5+ seconds) hangs when running the game. Nothing specific seems to trigger it. I've only seen them happen with DX12 rendering (the default), though I haven't tested too extensively with Vulkan or DX11 because those have pretty severe visual quality issues (something really nasty looking with shadow maps, especially apparent up on the pink forest / red bamboo plateau).

I looked through my logs to see if anything was printing during those hitches, and saw this:

[2024.10.08-02.04.06:298][202]LogGame: Warning: FGSchematic::GetIsPlayerSpecific: class was nullpeter.
[2024.10.08-02.04.18:527][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x000000010E1A1500) on the [0x0000000026C30B00, 3D] queue has not completed after 5.000000 seconds.
[2024.10.08-02.04.18:657][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x000000010E1AF000) on the [0x0000000026C30B00, 3D] queue has not completed after 5.129178 seconds.
[2024.10.08-02.04.18:657][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000E1896000) on the [0x0000000026C30B00, 3D] queue has not completed after 5.128201 seconds.
[2024.10.08-02.04.18:657][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000E189F400) on the [0x0000000026C30B00, 3D] queue has not completed after 5.128437 seconds.
[2024.10.08-02.04.18:658][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000FA2EF100) on the [0x0000000026C30B00, 3D] queue has not completed after 5.128858 seconds.
[2024.10.08-02.04.18:658][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000001188EF300) on the [0x0000000026C30B00, 3D] queue has not completed after 5.129422 seconds.
[2024.10.08-02.04.18:659][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x000000010E1ADA00) on the [0x0000000026C30B00, 3D] queue has not completed after 5.129432 seconds.
[2024.10.08-02.04.18:660][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000BF22B300) on the [0x0000000026C30B00, 3D] queue has not completed after 5.130143 seconds.
[2024.10.08-02.04.18:664][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x0000000117DF8B00) on the [0x0000000026C30B00, 3D] queue has not completed after 5.133832 seconds.
[2024.10.08-02.04.18:664][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000E32E0900) on the [0x0000000026C30B00, 3D] queue has not completed after 5.134202 seconds.
[2024.10.08-02.04.18:668][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x0000000119091100) on the [0x0000000026C30B00, 3D] queue has not completed after 5.137531 seconds.
[2024.10.08-02.04.18:669][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000E8704C00) on the [0x0000000026C30B00, 3D] queue has not completed after 5.138395 seconds.
[2024.10.08-02.04.18:669][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000BF22AA00) on the [0x0000000026C30B00, 3D] queue has not completed after 5.137579 seconds.
[2024.10.08-02.04.18:671][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000C2E0A300) on the [0x0000000026C30B00, 3D] queue has not completed after 5.138341 seconds.
[2024.10.08-02.04.18:671][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000BA1E1800) on the [0x0000000026C30B00, 3D] queue has not completed after 5.138559 seconds.
[2024.10.08-02.04.18:684][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000BF22C000) on the [0x0000000026C30B00, 3D] queue has not completed after 5.150494 seconds.
[2024.10.08-02.04.18:687][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000001190A8F00) on the [0x0000000026C30B00, 3D] queue has not completed after 5.152743 seconds.
[2024.10.08-02.04.18:687][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000000B539F900) on the [0x0000000026C30B00, 3D] queue has not completed after 5.152804 seconds.
[2024.10.08-02.04.18:688][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x00000001190CD400) on the [0x0000000026C30B00, 3D] queue has not completed after 5.152833 seconds.
[2024.10.08-02.04.18:688][576]LogD3D12RHI: Warning: GPU timeout: A payload (0x0000000118C1CD00) on the [0x0000000026C30B00, 3D] queue has not completed after 5.153178 seconds.
[2024.10.08-02.04.19:449][617]LogGame: Error: UFGOutlineComponent::ShowOutline - Trying to copy Custom Primitive Data from invalid index. (Index: 0, Num: 0)

One thing I've noticed about any given one of these hangs is that alt-tabbing away from the game and then back always gets the game back on track.

Not sure if this is a Proton issue or a bug for Coffee Stain, but figured I'd post it here first to see if anyone else is seeing it. I wish I had more concrete repro steps for these hangs, but all I have is "play the game for a bit". It might have something to do with objects de-spawning? Eg firing projectiles, creatures dying and despawning, etc... but I can't say for sure.

Specs:
kernel 6.11
Nvidia GTX 1070 8GB with 560.35.03 drivers
GNOME desktop + Wayland

deleting steamapps/compatdata/526870/pfx/drive_c/users/steamuser/Local Settings/Application Data/FactoryGame/Saved/Config/Windows/Engine.ini will restore graphics settings to default

I believe I got this error as a result of upgrading to Ubuntu 24.10 yesterday as I have never seen it before and got it consistently after the upgrade

Fortunately deleting this file fixed this issue for me and the game started right up again.

Sorry for not being more helpful. An Ubuntu upgrade pretty much just throws everything up in the air.

Upon exiting Satisfactory "1.0" i get an error (both on this official Proton variant as well as the GE variant.)

This wasn't an issue in Satisfactory update8

The game works pretty much flawlessly for me, except that same error message upon closing the game.

After upgrading from Fedora 39 to Fedora 41 issue with mouse cursor jumping from input fields and when grabbing some controls that should be dragged, started happening. Proton 9.0-3. GNOME 47, Wayland.

The issue is definitely related to remote play.

I'm running Arch Linux with NVIDIA 1070 GPU and was trying to stream to my M2 MacBook Pro.

Game ran fine on Arch but after attempting to stream to the MacBook, the game wouldn't launch anymore.

I run it from Steam, Proton 9.0-3 and gamemoderun %command%

After attempting to stream, I had to run this to make it work on Arch again.
rm -rf ~/.steam/steam/steamapps/compatdata/526870/

Same as other here, game works fine but I get vkDestroySampler error on exit

Satisfactory - mouse over text input box teleports mouse cursor

I'm having the same issue using pop_os up to date as of 2024-11-23, using a rx-6900xt and the kernel's reporting as Linux 6.9.3-76060903-generic (x86_64)

My work-around is that if I plug in an xbox controller I can use the right joystick to move the cursor over the few text boxes I need and I can use the mouse to click from there.

Satisfactory - mouse over text input box teleports mouse cursor

I have the same issue with Wayland, but not with XOrg.

Does anyone have workaround suggestions?

@Ruedii Not seeing that here with gnome/plasma wayland with proton 9. At least with the text input fields that you can find in the main menu from the new game and join game submenus, as well as the one in the in-game's builder menu. Does it only happen with specific text input fields? with what desktop environment?

I'm seeing an issue where playing from a Steam Deck in the official Dock causes audio to always come from the deck's speakers, disregarding the "use external sound" Steam system setting. The only way I was able to work around this was by booting into desktop mode, launching Satisfactory, then setting its OS-level application-specific audio output to be the external source (whose name might vary, depending on what you're plugging the dock into? can't tell). From that point on, including after I rebooted back into the regular Deck UI, it used the correct output. I didn't see any options in-game to set the audio output (and that doesn't seem like a usual feature) - so I'm not sure exactly whether the root of the problem was in the game, in Proton, or somewhere else.

There have been updates to Satisfactory since I last experienced the
issue. I will have to give it another go next time I am experimenting with
such stuff.

For replication's sake: My Wayland environment is KWin/Wayland with
Plasma. I have a mouse, and no touchscreen. I do not have GNOME
compatibility installed, but do have basic GTK+ compatibility installed.

It may be environment specific.

Notably KWin/Wayland uses a scaler compositor on full screen programs, and
passes through relative inputs such as mice in raw format. This may be
related.

Message ID: @.***>

Satisfactory - mouse over text input box teleports mouse cursor

Does anyone have workaround suggestions?

@Ruedii I don't know why this works but setting the launch command as such: gamemoderun gamescope --fullscreen --force-grab-cursor --backend wayland -H 2160 -h 2160 -- %command% seems to fix the teleporting mouse cursor for me.

EDIT: Change the resolution to whatever you're using (2160 for me above).

I'm having an issue with the map in satisfactory. Supposedly you're able to scroll the map around by clicking and dragging with right click. I've never been able to get this to work, though I haven't tried if it's possible on Windows. Does anyone else have this working?

I'm having an issue with the map in satisfactory. Supposedly you're able to scroll the map around by clicking and dragging with right click. I've never been able to get this to work, though I haven't tried if it's possible on Windows. Does anyone else have this working?

Never been a problem for me. Though I can't remember if it is right or left click - it's all in my muscle memory.

KWin/Wayland
Left Alt key doesn't work as pressable key in game. It works as modifier, but trying to use it to "hold to sprint" doesn't work. Even binding it to action to open build menu or inventory, on press it doesn't work. right alt key and ctrl keys works

Having issues with satisfactory on DX12 after the experimental 1.2 update.

Game hangs on loading screen and has to be force closed. You can hear the menu music behind the loading screen.

Upon running through terminal the proton output just keeps repeating

err:winevulkan:signaller_worker wait timed out with non-empty poll list.
endlessly.

System info
Arch on kernel 6.19.10
DE: Plasma 6.6.3
Mesa: 26.0.3
Nvidia driver: nvidia-open-dkms 595.58.03
Proton: Tried with various: experimental, 9.0, 10.0, GE 10-34.

CPU: Ryzen 7 9800x3D
GPU: NVidia RTX 5090
RAM: 64GB DDR5

Hello @Kat299, with Proton 10 or Experimental, please add PROTON_LOG=1 %command% to the game's launch options and attach the generated $HOME/steam-$APPID.log to this issue report as a file. (Proton logs compress well if needed.)

steam-526870.zip
Here you go.

Somehow I can technically play if I don't interact with this error message. But I don't trust the stability. (This is on a 9070xt and not an nvidia gpu)

okay, touch wood, so far, this has resolved

%command% -dx12 -gpuCrashDebugging though an odd thing I'm seeing is it still shows the "Processing Vulkan Shaders" prompt despite it being set to not launch through vulkan

The last time it did completely freeze, I ran journalctl -b -1 > journaltcl.log and here's the results

journalctl.log

Replying to https://github.com/ValveSoftware/Proton/issues/7580#issuecomment-4163733863

This might be related to a driver issue affecting newer Nvidia cards on the 595 drivers.

I encountered very similar behavior in Windrose (another UE5 game): the game would freeze on the loading screen while audio/menu music continued in the background. People in that community seemed to be mentioning the 595 drivers being the culprit.

I was able to work around it by using HansKristian-Work/pyroveil with these launch arguments:

PYROVEIL=1 PYROVEIL_CONFIG=/home/USERNAME/pyroveil/hacks/roadcraft-nvidia-570-stable/pyroveil.json %command%

(Adjust path according to your own installation.)

I was also using Proton CachyOS with the VKD3D heap argument enabled, although I'm not sure whether that part was actually required for the fix.

I get the following error after splash screen when running Satisfactory 1.2 on Proton hotfix branch.

journalctl.log

Hello @kisak-valve, sorry for that, I was just trying to tweak settings randomly. I reverted it without forcing a compatibility tool to let Steam choose what's most appropriate.

Here is the logfile, hope it helps. May be worth mentionning I installed Steam with flatpak.
steam-526870.log

Let me know if I can help further.

To clarify, when running the game, I briefly see: a satisfactory splash screen, a huge full-white window and finally a full-white window called "Factory Crash Reporter".

Yep, I use NVIDIA proprietary driver. I also changed version recently when I failed to run satisfactory, that was probably a mistake. :/ I can wait the update cycle if needed.

Outside of flatpak, nvidia install looks ok to me, but I may miss something:

xavier@xavier-mint:~$ nvidia-smi
Tue Jun  2 22:15:19 2026       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 580.159.03             Driver Version: 580.159.03     CUDA Version: 13.0     |
+-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 3070        Off |   00000000:2D:00.0  On |                  N/A |
|  0%   55C    P3             40W /  240W |     630MiB /   8192MiB |     11%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A            1787      G   /usr/lib/xorg/Xorg                      357MiB |
|    0   N/A  N/A            3592      G   cinnamon                                 31MiB |
|    0   N/A  N/A           16035      G   /usr/share/librewolf/librewolf          179MiB |
+-----------------------------------------------------------------------------------------+
xavier@xavier-mint:~$ glxinfo -B
name of display: :0
display: :0  screen: 0
direct rendering: Yes
Memory info (GL_NVX_gpu_memory_info):
    Dedicated video memory: 8192 MB
    Total available memory: 8192 MB
    Currently available dedicated video memory: 7204 MB
OpenGL vendor string: NVIDIA Corporation
OpenGL renderer string: NVIDIA GeForce RTX 3070/PCIe/SSE2
OpenGL core profile version string: 4.6.0 NVIDIA 580.159.03
OpenGL core profile shading language version string: 4.60 NVIDIA
OpenGL core profile context flags: (none)
OpenGL core profile profile mask: core profile

OpenGL version string: 4.6.0 NVIDIA 580.159.03
OpenGL shading language version string: 4.60 NVIDIA
OpenGL context flags: (none)
OpenGL profile mask: (none)

OpenGL ES profile version string: OpenGL ES 3.2 NVIDIA 580.159.03
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20

xavier@xavier-mint:~$ flatpak list | grep -i nvidia
nvidia-580-126-09	org.freedesktop.Platform.GL.nvidia-580-126-09		1.4	flathub	system
nvidia-580-159-03	org.freedesktop.Platform.GL.nvidia-580-159-03		1.4	flathub	system
nvidia-580-126-09	org.freedesktop.Platform.GL32.nvidia-580-126-09		1.4	flathub	system
nvidia-580-142	org.freedesktop.Platform.GL32.nvidia-580-142		1.4	flathub	system
nvidia-580-159-03	org.freedesktop.Platform.GL32.nvidia-580-159-03		1.4	flathub	system
Nvidia VAAPI driver	org.freedesktop.Platform.VAAPI.nvidia		25.08	flathub	system
i386	org.freedesktop.Platform.VAAPI.nvidia.i386		25.08	flathub	system

If relevant: when runningflatpak run com.valvesoftware.Steam

I see this:

Running query: 1 - GpuTopology
Response: gpu_topology {
  gpus {
    id: 1
    name: "NVIDIA GeForce RTX 3070"
    vram_size_bytes: 8847884288
    driver_id: k_EGpuDriverId_NvidiaProprietary
    driver_version_major: 580
    driver_version_minor: 159
    driver_version_patch: 3
    luid: 0
  }
  gpus {
    id: 2
    name: "NVIDIA GeForce RTX 3070"
    vram_size_bytes: 8847884288
    driver_id: k_EGpuDriverId_NvidiaProprietary
    driver_version_major: 580
    driver_version_minor: 159
    driver_version_patch: 3
    luid: 0
  }
  gpus {
    id: 3
    name: "NVIDIA GeForce RTX 3070"
    vram_size_bytes: 8847884288
    driver_id: k_EGpuDriverId_NvidiaProprietary
    driver_version_major: 580
    driver_version_minor: 159
    driver_version_patch: 3
    luid: 0
  }
  gpus {
    id: 4
    name: "llvmpipe (LLVM 21.1.8, 256 bits)"
    vram_size_bytes: 3221225472
    driver_id: k_EGpuDriverId_MesaLLVMPipe
    driver_version_major: 26
    driver_version_minor: 0
    driver_version_patch: 6
    luid: 0
  }
  default_gpu_id: 1
}

Fixed my problem by installing Steam from apt instead of flatpak.
Wiped my saves in the process as it seems that Satisfactory saves are not synced with Steam Cloud tho. Unlucky.

confirming the same problem since version 1.2, seems fixed by reverting to Proton 8. Which is old.

No it hasn't. Vulkan has been working fine between update 8 and v1.1
Proton 9 to 10 has been fine for me.

have this on mint with 7900xtx, was never an issue prior to 1.2

same issue! dx12 works, vulkan doesn't

I'm on AMD, not NVIDIA

For anyone getting the Unreal crash at start since the 1.2, 'ive got it working simply by uninstalling the game completely and deleting all the wine prefix that ever existed for the game. It then start properly (and prompt you to install a windows libs you can say yes).

Works on proton experimental and latest release of GE at the time. You might lose you saves if steam cloud is not enabled I can't confirm as I use steam cloud

For anyone getting the Unreal crash at start since the 1.2, 'ive got it working simply by uninstalling the game completely and deleting all the wine prefix that ever existed for the game. It then start properly (and prompt you to install a windows libs you can say yes).

Works on proton experimental and latest release of GE at the time. You might lose you saves if steam cloud is not enabled I can't confirm as I use steam cloud

Brilliant way to troll me into redownloading like 15gb or whatever the game is.

It does NOT work.

FWIW: Noticed current proton experimental causes the game to crash very frequently, whereas GE-Proton11-1 works fine for hours on end on my system. (Fedora 44, AMD 7900XTX)

no crashes here with dx12, proton experimental, rx 9070, mesa 26.1.4, linux 7.1.3. It's always a good idea to bring proton logs when getting crashes, and it'd also be good to check sudo dmesg to see if the gpu itself is crashing alongside the game

no crashes here with dx12, proton experimental, rx 9070, mesa 26.1.4, linux 7.1.3. It's always a good idea to bring proton logs when getting crashes, and it'd also be good to check sudo dmesg to see if the gpu itself is crashing alongside the game

With 1.2 there was no issue with DX12. Only vulkan.

For anyone getting the Unreal crash at start since the 1.2, 'ive got it working simply by uninstalling the game completely and deleting all the wine prefix that ever existed for the game. It then start properly (and prompt you to install a windows libs you can say yes).
Works on proton experimental and latest release of GE at the time. You might lose you saves if steam cloud is not enabled I can't confirm as I use steam cloud

Brilliant way to troll me into redownloading like 15gb or whatever the game is.

It does NOT work.

Idk what to tell you, I tried 6 differents proton version, some GE version, vulkan, dx12, the only thing that ended up working was re-installing the game 🤷 You might try the latest GE proton, that's what I'm currently running.

I thought this ticket was only vulkan, oops. Latest ge works fine on dx12. It's vulkan that's completely broken, last proton that works with vulkan is 8.

As mentioned before, game works perfect with GE-Proton11-1, but with proton experimental (experimental-11.0-20260713-unstripped), it regularly crashes after a while (can take quite some time before it crashes). Seems to be a corrupted pointer:

4054.365:0158:01d0:trace:seh:dispatch_exception code=c0000005 (EXCEPTION_ACCESS_VIOLATION) flags=0 addr=00006FFFFA183DDA
4054.365:0158:01d0:trace:seh:dispatch_exception  info[0]=0000000000000000
4054.365:0158:01d0:trace:seh:dispatch_exception  info[1]=FFFFFFFFFFFFFFFF
4054.365:0158:01d0:trace:seh:dispatch_exception rip=00006ffffa183dda rsp=000000001c7ec6f0 rbp=0000000000000000 eflags=00010202
4054.365:0158:01d0:trace:seh:dispatch_exception rax=0000000000000000 rbx=0000000000000000 rcx=0000000000000000 rdx=000000001c7f0020
4054.365:0158:01d0:trace:seh:dispatch_exception rsi=8b4820ec83485340 rdi=0000000000000020  r8=0000000000000020  r9=0000000000000000
4054.365:0158:01d0:trace:seh:dispatch_exception r10=00006ffffa070000 r11=00006ffffa01ddf1 r12=0000000000000002 r13=0000000000cce080
4054.365:0158:01d0:trace:seh:dispatch_exception r14=0000000000000000 r15=000000001c7f0000 mxcsr=00001fbf cs=0033 ss=002b
4054.365:0158:01d0:trace:seh:call_vectored_handlers calling handler at 00006FFFE29F4910 code=c0000005 flags=0
4054.365:0158:01d0:trace:seh:call_vectored_handlers handler at 00006FFFE29F4910 returned 0
4054.365:0158:01d0:warn:seh:virtual_unwind backtrace: 00006FFFFA183DDA: L"FactoryGameSteam-Core-Win64-Shipping.dll" + 0000000000113DDA.
4054.365:0158:01d0:trace:unwind:RtlVirtualUnwind2 type 1 base 6ffffa070000 rip 6ffffa183dda rva 113dda rsp 1c7ec6f0

RSI pointer seems corrupted, causing the crash:

180113dd7  ff4a18             dec     dword [rdx+0x18]
180113dda  488b06             mov     rax, qword [rsi]   <<<< CRASH
180113ddd  48894210           mov     qword [rdx+0x10], rax
180113de1  4885db             test    rbx, rbx
180113de4  7479               je      0x180113e5f

Looks like something is trashing memory when using proton experimental.
Can play for hours on end with Proton11-1, so must be something in experimental causing it.

steam-crash-proton-experimental.zip

I thought this ticket was only vulkan, oops. Latest ge works fine on dx12. It's vulkan that's completely broken, last proton that works with vulkan is 8.

I have tried ge 11.1, it does not work. Game freezes on startup.

Btw is there any hope of this getting fixed anytime soon?

Any progress on this?