Install vulkan-radeon, delete amdvlk and try again.

Install vulkan-radeon, delete amdvlk and try again.

Thank you very much, this did indeed fix the issue.

Shall I close this issue now, or leave it open in case any other compatibility problems occur for other people?

Leave it open i guess, wouldn't hurt.

Sad that how AMDVLK is prompted as first option on Arch leads to a bad experience for most users.

Glad it helped.

Hi,

The game seems to always update one component upon launch:

Seems like some graphical libraries are being replaced, which also prompt to shader recompile - when this happens then I get this error:

If I retry multiple times I get lucky and then the game is smooth as butter.

I'm using Proton 6.3-8 on Ubuntu 20.04 on a 3080 Ti (nVidia drivers 495.46).

I did not see that D3D error window yet on my computer(RX 5700 XT) but the shader recompilation on every launching followed by its usual stuttering is here too.

I'm also getting the stutter on load mentioned a few posts up (the post with the video). Stutters fairly badly when first loading into town, then for a few seconds when entering a stage, as shown in the video. Stabilizes for the most part shortly after, but there will be an occasional stutter (~1 sec) during gameplay as well.

• AMD Ryzen 5 3400G
• NVIDIA GTX 1650 SUPER (470.86)
• 16GB RAM
• Pop OS 21.10

It may be irrelevant but a quick test led to a proton log with a large number of warnings like

warn:d3d12_pipeline_state_init_graphics: Ignoring stencilTestEnable due to lack of stencil aspect

steam-1446780.zip

More than 5000 lines like this and all I did was loading the game, walking a little in the village and closing it.

System Information

• GPU: GTX 2080ti

• Driver/LLVM version: nvidia 495.46

• Kernel version: 5.16.0-arch1-1

• Full system information report

• Proton version: 6.3.8-5

  Opted into Steam client beta?: [No]
  Have you checked for system updates?: [Yes]

Attaching the PROTON_LOG, although the issue is not with the game startup.
steam-1446780.zip

Here is a more detailed log which also captures the time period where the error occurs.
steam-1446780-extra.zip

Symptoms

I have difficulties with multiplayer. I can join hunts in progress, and others can join mine, but if I attempt to depart on a hunt from the gathering hub with others, I disconnect with a "Communication Error" and enter the hunt separately from other players.

Reproduction

Launch Monster hunter rise with proton (but have tried also with proton-experimental and proton-ge)
Join a lobby with friends (this works)
Start a new hunt, wait for everyone to ready up, then launch hunt.
One of the players gets a communication error. Both players enter the hunt, but in separate instances.

Brief update, the "Fatal D3D Error" is fairly common on Windows too, nothing special about Proton/wine seems like... Perhaps some dodgy CAPCOM code?

@blastermaster77 As a workaround to the stuttering, I've been deleting the steam shader cache for the game (in steamapps/shadercache/1446780) and shader.cache2 file from the game folder before starting the game. The game seems to be happier if it rebuilds its own cache each time.

I experience the loading screen stuttering as well, and have been wondering if that is somehow related to the multiplayer communication errors I'm seeing (if the game fails to communicate during those stutters, is that what causes the disconnect?).

I haven't yet tried deleting the shader caches manually yet, but I do know that proton and proton-experimental crash for me during the game intro (before reaching the main menu) if shader precaching is disabled in Steam settings.

I also get that crash when turning off precaching. And to clarify, it's not just loading screens that are stuttering for me. It seems that every shader that was compiled on the last run causes a massive stutter on the next run when it is used again.

@Keelhauled Hm, that seems to line up with my experience of the game recompiling its own cache every time I change runners and thus running great on that first boot, but then stuttering again on subsequent boots with the same runner. But I've seen a report on ProtonDB of someone getting a Denuvo lockout from changing runners too many times so that's not something I want to do too often.

EDIT: I tried just deleting shader.cache2 and that got rid of the stutters, so it doesn't seem that deleting Steam's cache is necessary. However, I've also noticed that on boots after shader.cache2 is deleted, while the game doesn't stutter, it does fade to a black screen for several seconds after a loading screen, essentially adding several seconds to load times. Without deleting shader.cache2, there's no black screen after loading screens, but the stuttering is present. So it seems like a choice between stuttering or longer loads.

The game plays perfectly fine, but I have experienced a strange graphical glitch at Lava Caverns. There are green and red glowing parts on walls in the caves where there are water falls. Has anybody else had this?

I just installed latest mesa-git (22.0.0) and this is how it looks like


:

This bug does not occur with Proton-Experimental

I did some more testing with the help of a gracious friend. It seems like deleting the shader cache does somewhat fix the multiplayer connection issues I am encountering. It seems like if I attempt to load a map with a multiplayer session, it will succeed with no issues only if, during the load, I get the "Compiling Shaders" bar. Otherwise a communication error occurs. Deleting the shader cache meant that I was able to depart on a hunt with my friend, and it succeeded, but a communication error occured when returning back to the lobby (I assume this is because I already had computed shaders for the town and so did not need to again). This also explains why joining a lobby works, there is no map transition when joining a lobby.

I can only guess why this happens, my best totally uninformed theory is that there's some kind of handshake failing during the loading screen hitches, which doesn't fail if shaders need to be computed (maybe the game communicates an expected delay in that case? pure speculation).

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise
• Steam AppID of the game: 1446780

System Information

• GPU: Amd radeon rx 580 8GB
• Driver/LLVM version: 4.6 Mesa 20.3.5
• Kernel version: 5.10.0-10-amd64
• Proton version: 6.3-8 and experimental fail similarly.

Symptoms

Update
After clearing shadercache, game runs correctly again. Leaving the error report as the error is still present (but probably unrelated).

Unhandled Exception:
System.TypeLoadException: Could not load type of field 'InstallerMessage.Form1+<CodecCheck>d__11:<codecQuary>5__2' (3) due to: Could not load file or assembly 'Windows.Foundation.UniversalApiContract, Version=8.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.
  at InstallerMessage.Form1..ctor () [0x0000c] in <a06ba4bb41904acfa9f23ab189e60bef>:0 
  at (wrapper remoting-invoke-with-check) InstallerMessage.Form1..ctor()
  at InstallerMessage.Program.Main () [0x0000b] in <a06ba4bb41904acfa9f23ab189e60bef>:0 
[ERROR] FATAL UNHANDLED EXCEPTION: System.TypeLoadException: Could not load type of field 'InstallerMessage.Form1+<CodecCheck>d__11:<codecQuary>5__2' (3) due to: Could not load file or assembly 'Windows.Foundation.UniversalApiContract, Version=8.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.
  at InstallerMessage.Form1..ctor () [0x0000c] in <a06ba4bb41904acfa9f23ab189e60bef>:0 
  at (wrapper remoting-invoke-with-check) InstallerMessage.Form1..ctor()
GameAction [AppID 1446780, ActionID 1] : LaunchApp changed task to RunningInstallScript with ""
GameAction [AppID 1446780, ActionID 1] : LaunchApp changed task to SynchronizingCloud with ""
GameAction [AppID 1446780, ActionID 1] : LaunchApp changed task to SynchronizingControllerConfig with ""
Opted-in Controller Mask for AppId 1446780: ffff
GameAction [AppID 1446780, ActionID 1] : LaunchApp changed task to ProcessingShaderCache with ""
Fossilize INFO: Autogroup scheduling is not enabled on this kernel. Will rely entirely on nice().
Fossilize INFO: Autogroup scheduling is not enabled on this kernel. Will rely entirely on nice().
Fossilize INFO: Autogroup scheduling is not enabled on this kernel. Will rely entirely on nice().
***** OUT OF MEMORY! attempted allocation size: 1698128704 ****
src/tier0/memstd.cpp (2838) : OUT OF MEMORY
src/tier0/memstd.cpp (2838) : OUT OF MEMORY
src/tier0/memstd.cpp (2838) : Fatal assert; application exiting
src/tier0/memstd.cpp (2838) : Fatal assert; application exiting
Installing breakpad exception handler for appid(steam)/version(1642451672)
assert_20220119202613_32.dmp[37423]: Uploading dump (out-of-process)
/tmp/dumps/assert_20220119202613_32.dmp
_ExitOnFatalAssert

Reproduction

Tried both steam stable and beta, both fail the same way.
Tried both proton 6.3-8 and proton experimental, both fail the same way.
Other games with proton 6.3-8 and proton experimental that ran before today run correctly.

Solution

In my case, the game worked again after removing the shader cache. The vulkan shader loading bar took less than normal and never went into the game but crashed steam instead, hopefully this helps others.

System Information

• GPU: AMD 6700xt
• Driver/LLVM version: Mesa 22.0.0-devel (git-f1f65e5bcf)
• Kernel version: 5.15.11
• Proton version: Experimental

Symptoms

Online play is not working properly: most of the time it will let you connect to another lobby once and play one hunt, after that the game will either crash or disconnect, either when returning to the lobby or when departing for another hunt. If it disconnects it won't be able to connect again (in game error or crash when joining).
There is also an element of randomness on whether it lets you connect even one time, it seems like if you try it as soon as you get in game it mostly works fine, but otherwise if you try after playing a while even in single player it can fail the first time too.

Reproduction

Speak with the "Courier" npc (see attached picture), select play online->find a lobby and join a game. If it lets you, play one mission then come back to the lobby. If it doesn't disconnect you or crash, join another mission on the same lobby to trigger the issue. I tried 5 times and I got disconnected or crashed 100% of the time. If disconnection happens, trying to join another lobby will result in an error. I stopped testing because I didn't want to disturb other players

Unrelated to the issue, but experimental bleeding edge stopped working with one of the last updates, it crashes once it gets tho the health disclaimer on boot so I can't try that either (that used to work, not sure about the online because I never tried with bleeding edge. I haven't played in a week so I'm not sure when it broke).
Below it's a log for the bleeding edge crash
steam-1446780.log

In terms of the "Fatal D3D Error", seems like it may be caused by enabling MANGOHUD (v 0.6.5) - apparently enabling it cause this issue to happen ~50% of the times the game is started, I've disabled it and so far never experienced the error once - will post further updates if this is the case.

Edit - can confirm having MANGOHUD enabled triggers the "Fatal D3D Error", restarted the game multiple times without it and never got it - otherwise is quite frequent.

adding to my other message, today I played 2-3 hours with a build of proton-tkg based on wine-staging 95bf6698950d82622f7461111b3a199343ce9a21, didn't get disconnected even once

Is there a way to disable vulkan shader compilation just for monster hunter rise? I noticed that when I start a session on wayland the game compiles the shaders on its own and its smooth as butter, wen I go back to x11 and wayland for the second time the steam vulkan shader compilation appears and it gets stutter like the video i posted above. Its with nvidia drivers.

@blastermaster77 afaik there is no way to disable it per game

95bf6698950d82622f7461111b3a199343ce9a21

Could you please link to the proton version you used?

95bf6698950d82622f7461111b3a199343ce9a21

Could you please link to the proton version you used?

My comment was mostly directed at Valve devs, I don't know if it's ok to share unofficial builds here. You could compile it yourself with proton-tkg for now.

Compatibility Report

Name of the game with compatibility issues: Monster Hunter Rise
Steam AppID of the game: 1446780

System Information

GPU: GeForce RTX 2060 SUPER
Driver/LLVM version: Mesa 22.0.0-devel
Kernel version: 5.4.0-96
Proton version: Proton Experimental

Symptoms & Reproduction

I start the game with any Proton version, it crashes & gives me an error.

Greetings,

I run this game on Linux Lite & I get this error message. I updated everything & cannot find a solution.
The other bug reports didnt work for me or maybe I did something wrong.
Im relatively new to Linux & this is my first bug report, so yeah probably getting something wrong

steam-1446780.log
.

I have just noticed the following dmesg output after launching Rise

umip: MonsterHunterRi[34229] ip:14f533690 sp:93f740: SGDT instruction cannot be used by applications.
umip: MonsterHunterRi[34229] ip:14f533690 sp:93f740: For now, expensive software emulation returns the result.
umip: MonsterHunterRi[34229] ip:160386cb2 sp:93d4c8: SGDT instruction cannot be used by applications.
umip: MonsterHunterRi[34229] ip:160386cb2 sp:93d4c8: For now, expensive software emulation returns the result.
umip: MonsterHunterRi[34229] ip:15e660f48 sp:93d0c8: SGDT instruction cannot be used by applications.

Is it something to be concerned from a performance point of view?

adding to my other message, today I played 2-3 hours with a build of proton-tkg based on wine-staging 95bf6698950d82622f7461111b3a199343ce9a21, didn't get disconnected even once

Just tried this as well and got 4 successful connections out of 5 (the disconnect might've been a fluke, although I didn't have a single disconnect of that kind in the ~50 hours I played on Windows). There's still seems to be something very screwed with the shader caches, as several people here are reporting. I'll try deleting the shader.cache2 file - are you guys deleting it every time before you launch the game?

Also, is Steam performing 'first time setup' every time you start the game for anyone else as well?

@lptnmachine In my case I have to delete shader.cache2 and nvidiav1/GLCache/ folder before every launch for the stutters to not happen. I also skip steams own "processing vulkan shaders" popup because it doesn't seem to matter. And yeah first time setup happens very often, not sure if every time but almost.

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issuecomment-1024186800

I left the shader cache alone, it only stutters on load screens and maybe briefly as soon as I get in a hunt, otherwise works fine during gameplay.
The first time setup windows appears multiple (every? Didn't pay much attention) times for me too, but it's only for a split second

are you guys deleting it every time before you launch the game?

No. But the consequence is that the whole computer freezes when launching Rise. After some seconds things go back to normal. It is annoying but having to manually delete a file every time also is.

Also, is Steam performing 'first time setup' every time you start the game for anyone else as well?

Yes. But at least here it finishes fast so I do not bother with it as much as with the current performance in the flooded forest. What kind of fps are you guys getting there? I have a Ryzen 3700x and a Rx 5700 XT and the fps goes from a solid 60 down to 40 or 35 and sometimes 30. But as I do not have Windows in my PC I am not sure about what to expect in that map.

Edit: and my monitor resolution is 2560 x 1080(ultrawide)

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issuecomment-1024254180

That doesn't seem right, I'm on a 1440p monitor and a 6700xt, which I understand is a bit better but I get way better performance than that, for sure never below 60. Are you cpu or gpu limited when performance drops? What vulkan driver are you using?

Are you cpu or gpu limited when performance drops?

That is the weirdest part. None of them is saturated. Far from that. I think none of my cpu cores is getting beyond 30% usage. Similar thing in the output of radeontop. My gpu is essentially begging to receive something to do. I think there is some kind of thread synchronization bottleneck happening on my side. At least in perf output I see something like this

If that is normal or not I am not sure. At first sight it is almost like Proton spends most of its time acquiring locks to synchronize threads.

What vulkan driver are you using?

Mesa vulkan-radeon 21.3.4

I get the feeling you guys are having a different problem with shaders if it doesn't stutter when running around in the village for example.

@wwmm Are you sure it's just flooded forest that has bad fps? I solved my fps issues by setting WINEDLLOVERRIDES="dxgi=b" in start parameters like i've mentioned here.

if it doesn't stutter when running around in the village for example.

That happens to me only in the first time I walk in the village after loading the game. Later visits to the village are fine while the game is running.

Are you sure it's just flooded forest that has bad fps?

Later today I will pay attention to that. But the flooded forest is definitely the map that makes me think about performance loss. The whole map has bad performance here.

I solved my fps issues by setting WINEDLLOVERRIDES="dxgi=b" in start parameters like i've mentioned

I will try to do that later and see what happens.

That happens to me only in the first time I walk in the village after loading the game. Later visits to the village are fine while the game is running.

Could you test if those first time stutters go away after deleting shader.cache2 and steamapps/shadercache/1446780/nvidiav1/GLCache folder? I don't get any stutters during gameplay if I delete those first.

@wwmm next time you play take a look at your gpu clocks and see if it's an issue with the gpu not clocking up, if it isn't you could try forcing a different power profile

echo "manual" > /sys/class/drm/card0/device/power_dpm_force_performance_level
echo "1" > /sys/class/drm/card0/device/pp_power_profile_mode

this has solved some issues for me and other people (not related to mhr but worth a try) on NAVI since the default profile it's a bit to aggressive in clocking down and it resulted in bad frametimes
the card number could vary, and on the second power profile I choose 3D_FULL_SCREEN, which for you could be a different number so adjust accordingly. (more info here)

@Keelhauled the issue for me it's pretty random, so it's hard to tell if it's helping or not, (but I do get stutters also in the village, only right after the loading screen). I'm on AMD so it might just be a different behavior with the drivers. There is a solution in development that will remove the stutters completely for vkd3d-proton, but I don't know when it will be merged

Could you test if those first time stutters go away after deleting shader.cache2

Yes. They go away. As I have a Radeon GPU it does not make sense to delete the other folder here but deleting shader.cache2 fixes these stutters. The problem is that steam stays for much more time in the "Compiling Shaders" stage. And that also happens in the game. A lot of time waiting for the shaders to be recompiled.

Thanks everyone for the responses. I played and tested some more, and here's some things I think I figured out (keep in mind that all these are uneducated guesses at best lol):

• I built Proton with the commit mentioned by @EmmerichFrog
• On first run, the game goes through the cache building routine in-game, and then stutters when Proton converts the shaders
• On subsequent runs: if I delete shader.cache2 before I run the game, the game does the game's cache building in most (all?) loading screens for a few seconds. No stutters whatsoever after (except some short ones in loading screens sometimes), runs fine and only few connection issues (about one every hour, maybe. In Windows I never had any disconnects at all). If I don't delete shader.cache2 however, the game doesn't do the in-game shader building, but stutters as if Proton is converting shaders when entering a map that hasn't been entered since relaunching the game.
• I think these stutters in the loading screen really are what's dropping the connection. The internet of a friend of mine drops out sometimes, and in MH:R he gets disconnected immediately, while in something like Rocket League he'll just lag out for half a minute or so. In Windows, the loading screens run completely smooth - the game building the shaders in-game at the loading screens runs smooth in Linux, but when Proton converts shaders I get heavy stutters.
• So why are the shaders being constantly regenerated and reconverted? My guess is that for some reason after restarting the game Proton loses the connection between the shaders generated by the game (the ones in shader.cache2 I assume) and the converted version of these shaders in the Proton cache, so they get regenerated each time.

I haven't touched the steamapps/shadercache/1446780/nvidiav1/GLCache folder at all.
This is the result of just a couple of hours of 'testing' (mostly just playing because it really did work well most of the time) so might not be 100% accurate, but I really do think that there's connection between the shader (re)building and the disconnects.

I solved my fps issues by setting WINEDLLOVERRIDES="dxgi=b"

This makes no difference here.

next time you play take a look at your gpu clocks and see if it's an issue with the gpu not clocking up, if it isn't you could try forcing a different power profile

That also made no difference. I was already using profile_peak in power_dpm_force_performance_level. So the clocks were already fixed.

Just to show what is happening on my computer the image below is from the flooded forest. At its top right corner is the steam fps meter showing 39 fps

And that is the htop image from the same moment

If the cpu usage gets lower than that it will be almost on idle XD. And this is what radeontop shows

My gpu is also begging for work.

@lptnmachine with that proton build I get no issues online, regardless of lags during/after loading screens, so I'm not sure it's related, but since it works fine now for me I'd rather not touch anything else

@wwmm no idea then sorry, I see you have a xwayland process so I assume you're on wayland, have you tried running the game on x11? I don't think it will solve anything but ¯_(ツ)_/¯

I see you have a xwayland process so I assume you're on wayland, have you tried running the game on x11? I don't think it will solve anything but ¯(ツ)/¯

I tried that a few minutes ago. And there was no difference at all. What is good in a way because it means wayland is not losing to Xorg. But unfortunately in both of them the performance is not as good as it can be.

I have to take a careful look at my computer configuration. I must have something enabled somewhere that is influencing this somehow...

@wwmm

this is what I get roughly on the same spot, I am cpu limited too as you can see, so it might not be a gpu issue after all.
I get way better performance but I'm on a 5800x. I'm not sure if the delta should be this huge. Is the cpu clocking up fine?

this is what I get roughly on the same spot

85 fps??? Holy Moly XD. Something is really bad on my side.

Is the cpu clocking up fine?

As far as I can see it is. At least setting the performance governor made no difference when compared to schedutils.

Although I did not fix my performance problem yet I've noticed some interesting things that are affecting performance. I did the previous tests with VKD3D_CONFIG=force_host_cached and now that I tried without it I am losing from 15 to 20 fps in places where I was having 60 fps without problems. Like the Village. And without it I see some stuttering there as well while walking around. I wonder if that was the reason I was seeing less stuttering there when compared to other people. In any case it is clear I need force_host_cached on my hardware. It makes a big positive difference on this game performance so I will keep using it.

Something else I noticed is that smart access memory is not playing nice with Rise. At least not with my GPU. After setting RADV_PERFTEST=nosam I gained from 7 to 9 fps in the same map location I showed in one of my posts. The one where I was only getting 39 fps.

@wwmm what happens with VKD3D_CONFIG=no_upload_hvv? It should disable bar within vkd3d. BTW there are been other people that had issues with re-bar on other games with some cpu/gpu combination heavily degrading performance so it might be worth it to try with it disabled in bios

what happens with VKD3D_CONFIG=no_upload_hvv?

It seems to have the same effect as nosam. There is some performance gain but nothing magical.

it might be worth it to try with it disabled in bios

And to my surprise that made a big difference. After disabling Above 4G decoding in my bios I started to see 60 fps almost everywhere in the flood forest :-). The place I showed in one of my images before is an exception now where I see the fps decreasing to around 54. "Suddenly" my performance issues are gone. Thanks!

But honestly I think it is really odd that even nosam and no_upload_hvv together do not have the same effects as disabling the resizable bar in the bios. And that is not all. Compare the perf output I have now

with the one I showed before. The function fast_rtlAcquireSRWLockShared that was the most called function by very far suffered a massive decrease in its number of calls. It is insane. That did not happen when setting nosam or no_upload_hvv.

I wonder what else could be using the resizable bar besides radv and vkd3d. Or could it be that one of them has a bug and is not fully disabling the resizable bar usage when we tell it to do so?

@wwmm I'm not sure about the details, but just having rebar enabled causes much slower memory access on some hw configuration, it's probably why amd has it restricted to 6xxx cards and 5xx mobo on windows

it's probably why amd has it restricted to 6xxx cards and 5xx mobo on windows

I see. I have my RX 5700 XT on a X470 motherboard. But I still think that the fast_rtlAcquireSRWLockShared situation in perf's output is a little weird.

I can confirm that using the launch option WINEDLLOVERRIDES=dxgi=b fixes the stutters with nvidia gpu and I get no disconnects form online game. The game now always compile the shaders at the beginning. The only downside is that you have to wait for the compilation of the shaders but Its way better that way until we get a proper fix.

Using WINEDLLOVERRIDES="dxgi=b" does not fix the shader stuttering for me. Also on NVIDIA GPU.

Using WINEDLLOVERRIDES="dxgi=b" does not fix the shader stuttering for me. Also on NVIDIA GPU.

yep it does not fix it darn!

The game stopped working on my desktop and just fails on start.
DMC5, which uses the same engine, works fine.

I've tried deleting the prefix, shadercache and reinstalling the game, and using different Proton versions, to no avail.

inxi -b (Nvidia GPU is disabled via kernel arguments)

System:
  Host: toolbox Kernel: 5.16.8-200.fc35.x86_64 x86_64 bits: 64 Desktop: GNOME
    Distro: Fedora release 35 (Thirty Five)
Machine:
  Type: Desktop Mobo: Micro-Star
    model: MPG X570 GAMING PRO CARBON WIFI (MS-7B93) v: 1.0
    serial: <superuser required> UEFI: American Megatrends LLC. v: 1.D0
    date: 07/05/2021
CPU:
  Info: 8-core AMD Ryzen 7 5800X [MT MCP] speed (MHz): avg: 2595
    min/max: 2200/4850
Graphics:
  Device-1: NVIDIA TU106 [GeForce RTX 2060 SUPER] driver: N/A
  Device-2: AMD Navi 21 [Radeon RX 6800/6800 XT / 6900 XT] driver: amdgpu
    v: kernel
  Device-3: Elgato Systems Game Capture HD60 S+ type: USB
    driver: hid-generic,snd-usb-audio,usbhid,uvcvideo
  Display: wayland server: X.Org 1.21.1.4 driver: loaded: nvidia
    unloaded: fbdev,modesetting,nouveau,vesa resolution: 1: 2560x1440~144Hz
    2: 2560x1440~60Hz 3: 1920x1080~60Hz
  Message: Unable to show advanced data. Required tool glxinfo missing.
Network:
  Device-1: Intel I211 Gigabit Network driver: igb
  Device-2: Intel Wi-Fi 6 AX200 driver: iwlwifi
Drives:
  Local Storage: total: 3.64 TiB used: 1.96 TiB (53.9%)
Info:
  Processes: 527 Uptime: 3h 26m Memory: 31.32 GiB used: 6.01 GiB (19.2%)
  Shell: Bash inxi: 3.3.12

steam-1446780.log
http://ix.io/3PoA

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issuecomment-1037243873

Did you delete your compatdata folder when switching proton versions? Like unofficial ones like GE?

@TheHooly if you're using mangohud try without that

@blastermaster77 I did delete the prefix every time I used a different Proton version, yes.
@EmmerichFrog Trying to run the game without MangoHUD seems to make no difference, http://ix.io/3Ppl

@blastermaster77 I did delete the prefix every time I used a different Proton version, yes. @EmmerichFrog Trying to run the game without MangoHUD seems to make no difference, http://ix.io/3Ppl

Maybe it's selecting the wrong card? Try forcing the amd ICD with VK_ICD_FILENAMES=/usr/share/vulkan/icd.d/radeon_icd.x86_64.json %command%

Adjust the path if it's not there on fedora

I only have 1 active card,
I tried adding your env var, but that doesn't make a difference, either.

@TheHooly comparing your logs with some of my old logs the only difference that seems relevant is

19426.155:0124:0128:fixme:ntdll:NtRaiseHardError 50000018 stub

right before the crash. But unfortunately the logs are not clear(at least to me) about why the game would be raising a "hard error" message. Did you try to take a look at your system logs? Maybe a command like sudo journalctl | grep -i steam will show something from past attempts that is interesting. Or running sudo journalctl -fb before launching the game so you can see what the system is printing to its logs while the game tries to start.

sudo journalctl | grep -i steam didn't reveal anything that looked interesting, but sudo journalctl -fb did: http://ix.io/3PvM

Can some one help me make a launch option script to be able to delete the shader.cache2 automatically when I launch the game? I delete it because that fixes the stutters and the online disconnects. Been trying but cat not make it to work. Any help will be apreciated.

UPDATE: nevermind fixed it I had to point to my /home/user/bin/script.sh; %command% I was missing the semicolon at the end of script.sh in the launch option now it works flawlessly!
this is the simple script:
#!/bin/sh
rm ~/.steam/steam/steamapps/common/MonsterHunterRise/shader.cache2

I put the script on the user folder /bin/<script.sh>
give premissions with chmod +x script.sh
and the launch option on steam : ~/bin/script.sh; %command%

hope this helps people having disconnect problems and the stutters with the game at the beginning. This will make the game compile shaders every time you launch the game, in my system it does it fast. You will notice that the game is in sync with the audio now.
video of compilation every time you launch it. The flickering of the video is a problem while recording with nvidia with obs, I forgot to deselect the allowfliping option on the nvidia-settings manager.

https://user-images.githubusercontent.com/9307070/153937160-016d2488-f16a-48aa-a0cb-2c09ce0f1465.mov

The latest update (3.9.0.0) hit some kind of regression for me with a config that worked before the update. During the last moments of the initial loading screen (with the spinning logo at the lower right corner) after shader compile the game drops to 1 fps and doesn't recover from there. The start menu and village (if you manage to get there) continue running with single digit fps.

Tested WINEDLLOVERRIDES="dxgi=b", deleting shader cache, previosly working Proton GE 7.x, latest GE 7-4, Proton Experiemental, with Steam runtime and native without success.

Running NVIDIA and Intel on Arch Linux, Xorg, KDE, no special kernel or anything else.

One issue I have and do not know why but on Manjaro KDE latest and running NVIDIA, Intel and Xorg. On whatever proton release (GE-7.9, Experimental or 7.x) when I use my keyboard and mouse everything is fine but then when I hook my controller on USB or Bluetooth using a weapon with effects like the BOW I get an insane FPS drop when the bow glows red and then when I stop it goes back to normal.
Here's a video to show you this weird behaviour...

https://user-images.githubusercontent.com/14289711/158038934-9f3de9e3-f75d-4ea3-9fcb-b88a5dcdddad.mp4

WINEDLLOVERRIDES="dxgi=b" Also fixed FPS for me. Went from 39 fps to 120 fps on 5600g + rx 6600. Using GE-7.20, since regular proton and experimental crashed for me. I removed my pfx on compatdata and all shader info before changing proton versions.

So I've been playing with no problems up to this point, until I started fighting Basarios. After shooting him once, the game froze one the frame of impact, music still playing and my controller was vibrating. After like, a minute, the game continued until I tried to shoot him again and the game froze. I was still able to alt+tab out and do other things though. (I did record this, but the video is too big for github)

EDIT (Forgot to link gist): https://gist.github.com/DigitalDevilSummoner/d7a227765539daee04f9fd1d98d2be93#file-gistfile1-txt

I don't have a log for this yet, might try again tonight, but the stall appears to correlate with controller vibration. At least when using a Dualshock 4, large motor activation seems to do this. When I disabled vibration entirely, I was able to complete a mission without a stall. With it on, the vibration would start, a few frames later the game would stall for 1-2 minutes, vibration would stop after ~15 seconds (as far as I know this is a built in controller feature so vibration doesn't go on indefinitely when not signaled), and then it'd get a few more frames through and vibration would kick back in, triggering the stall again. I didn't sit through it long enough to get past the event that was triggering the vibration as this probably would have taken 20-30 minutes.

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issuecomment-1171094232

This is actually the exact issue I was having and I also couldn't get a log, and now that i'm trying to, it doesn't happen.

Having an issue with audio playback that I can't reliably reproduce. Sometimes the audio, specifically the music will cut out for 2-3 second intervals. The only correlation I have to go on is I'm using PulseAudio instead of PipeWire and I was playing the audio of a video in the background when I noticed this happening. I just found it weird that it wasn't all game audio cutting out but just the vocals/music. Haven't played natively on Windows so can't say if this is an actual bug with the game itself or not. Considered it being data streaming issue but I wasn't performing any intensive read/write operations in the background and I'm using an SSD. Have verified files.
gist
steam-1446780.log

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise
• Steam AppID of the game: 1446780

System Information

• GPU: GTX 1080
• Driver/LLVM version: nvidia 515.57
• Kernel version: 5.18.8-arch1-1
• Link to full system information report as Gist: https://gist.github.com/ConnorDY/8b46a1d37b8079d4f5281e8524fc13eb
• Proton version: Experimental & 7.0-3

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

After the Sunbreak expansion update (not sure if you need the DLC installed or not, but I do have it installed), The Frost Islands map causes the game to crash shortly after loading into it. You are able to walk around for a bit, sometimes for a few minutes, but it always freezes. My entire system freezes for a bit when this happens. Afterwards, the game remains frozen however I can still hear the game audio. This requires me to restart the game.

I haven't ran into this issue anywhere else in the game. I was able to play this map just fine before the Sunbreak update without any issues.

Reproduction

1. Start the game and load into the Frost Islands map. This can be via an Expedition Tour or any other quest.
2. Attempt to move around the map. The game should freeze within a few minutes, though sometimes it'll take a bit longer.

Proton log file

steam-1446780.log

Commenting as to give my proton log dump after experiencing the same crash as ConnorDY. Entering the Frost Island causes game to freeze. Time to freeze is variable.

System Info
Proton Log

I experience the same issue and found another report on protondb.
We all share the same graphics card; a GTX 1080.
System Info

I couldn't get mangohud to show ram, vram, and swap usage, but could the game/vkd3d be eating up vram, ram, and swap? I do know that out of my 8GB of ram and 4GB of swap, ram was ~95% and swap was ~50% at time of one of my crashes.

A quick look through of a successful hunt, on Sandy Plains, vkd3d doesn't seem to write data to system memory. More specifically this error:
015c:warn:vkd3d_allocate_device_memory: Memory allocation failed, falling back to system memory.

The successful hunt also did not seem to throw these errors as well:
0168:warn:d3d12_pipeline_state_init_graphics: Ignoring stencilTestEnable due to lack of stencil aspect.
016c:warn:d3d12_pipeline_state_init_graphics: Unused input element 0-9

Thoughts?

No Crash Proton Log
EDIT Added in that the successful hunt was in Sandy Plains, and not Frost Islands.

@MrQuestionMarks I monitored my VRAM, RAM, GPU, and CPU usage while playing. I'm not getting anywhere close to maxing out any of those when I'm freezing on Frost Islands.

I set all graphics options to low (except render scale) and it seems to work fine now for the three hunts in the Frozen Islands so far.
I was also not close to exhausting VRAM or RAM before.

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise: Sunbreak
• Steam AppID of the game: 1446780

System Information

• GPU: NVIDIA GeForce RTX 3080
• Driver/LLVM version: 4.6.0 NVIDIA 515.57
• Kernel version: 5.15.50-1-MANJARO
• Link to full system information report as Gist:
• Proton version: Experimental and GE 7-24

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.

Attached Log

Symptoms

Game blackscreens and freezes at a random chance at the end of any cutscene, including: Selecting character, Eating food, new location and new monster cutscenes, and end of hunt cutscenes.

The

There is no consistency

Reproduction

The fastest way to reproduce this is to launch the game, select your character, and watching the eating cutscene.

Adding to @ThatBigBirb 's report above, and I think I can give more context since I'm encountering the same issue.
Looking in the kernel logs turned up this tidbit:

NVRM: Xid (PCI:0000:09:00): 31, pid=42769, name=MonsterHunterRi, Ch 00000031, intr 00000000. MMU Fault: ENGINE HOST2 HUBCLIENT_ESC faulted @ 0x0_94f3f000. Fault is of type FAULT_PTE ACCESS_TYPE_VIRT_WRITE

I am also using an RTX 3080Ti, and this only started happening after a fairly recent graphics driver update.

Kernel version: 5.18.11-200.fc36 (Fedora 36)
Driver version: Nvidia 515.57-1
Proton version: Experimental and 7.0-3

I crash immediately whenever I attempt to start the 7* urgent quest to hunt Narwa.
Log:
https://drive.google.com/file/d/1Tuxo_JZhPAv0aj1FvaF4mHpYPyQ3TTDF/view?usp=sharing
Using:
Native Steam
Pop OS 22.04 LTS
RTX 2060 notebook GPU
nvidia driver 515.48.07
Proton experimental
Note: Using external display with internal display disabled is required in order to run the game on this setup. The Intel iGPU should not be involved at any point as the output is directly from the 2060.

Launch options:
PROTON_LOG=1 PROTON_ENABLE_NVAPI=1 PROTON_HIDE_NVIDIA_GPU=0 __NV_PRIME_RENDER_OFFLOAD=1 __GLX_VENDOR_LIBRARY_NAME=nvidia %command%

@almightyfoon I had a similar issue with my laptop. In order to stop this I had to disable my internal monitor,

I also use these options PROTON_LOG=1 PROTON_ENABLE_NVAPI=1 PROTON_HIDE_NVIDIA_GPU=0 __NV_PRIME_RENDER_OFFLOAD=1 __GLX_VENDOR_LIBRARY_NAME=nvidia %command%

EDIT: This seems to be finally fixed as of NVIDIA 515.76.

MHRise refuses to load vkd3d-proton.cache files on drivers 515.57 and 515.65.01 on a 3090.

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise
• Steam AppID of the game: 1446780

System Information

• GPU: NVIDIA GeForce RTX 3090
• Driver/LLVM version: nvidia 510.73.05 (working); nvidia 515.57/515.65.01 (broken)
• Kernel version: 5.18.15-gentoo-dist
• Link to full system information report as Gist: https://gist.github.com/kzdixon/180193f9a5eed363cdb05525162db5a7
• Proton version: 7.0/Proton Experimental (Bleeding-Edge)/GE-7.28 (same behavior across all three)

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.

Nothing is printed from adding PROTON_LOG=1 regarding the crash on attempting to load vkd3d-proton.cache files. The game simply freezes.

Symptoms

When launching the game on newer Nvidia drivers (515.57 or 515.65.01) on a 3090 after a first boot (fresh shadercache, no cache files in root dir) the game will freeze upon attempting to load the shader cache, assumedly for the RE Logo Intro movie and beyond. dinput8 modding will never load successfully when using these drivers either.
Downgrading my nvidia-drivers to the stable Gentoo version of 510.73.05 completely remedies this problem and the behavior is as it used to be before my two installs (Gentoo and Arch) pulled down the 515.57 update.

Reproduction

Install Nvidia driver 515.57 or greater while using a 3090 on Gentoo or Arch.
Launch the game fresh (this will work). Close the game.
Launch the game a second time, it will freeze after a bit of loading at the splash screen and never recover until stopped at the Steam UI, this also causes a slight lock-up of the system as it closes.
To avoid a freeze, remove any vkd3d-proton.cache files in the MonsterHunterRise directory and it will launch (assuming there is no attempt to load dinput8.dll as well).

If needed, mangohud and gamemoderun are being used as well, but these work perfectly when running drivers before 515.57, as was my experience for 200 hours of gameplay on Proton.

Gist: https://gist.github.com/YoungMaestro/9bdf3bcb1a725e1e87b141d68be11017
(couldnt find appid.log)

It didn't have any effect

Anything I can try to fix it?
edit:
Running this command allowed me to use the graphics card and put the game full screen although now is a complete black screen and I can interact with the game (hear intro music and Navigate through the start menu)
WINEDLLOVERRIDES="dxgi=b" DRI_PRIME=1 %command%

The newest Proton - Experimental (build id: 10039633) crash Monster Hunter Rise.
steam-1446780.log
Previous Experimental version and Proton Next (7.0-5) is working.

@Billli11 Could you provide some more information? Please copy your system information from Steam (Steam -> Help -> System Information) and put it in a gist, then include a link to the gist in this issue report. Could you also give more details about when the game crashes?

system info

the game crash right after shader compilation and intro movies,then launch CrashReport.exe.

Tried remove and recreated prefix and still crash.

Tried 3 other game. Those are working.(include DX11 and DX12 title)

Getting some vertex explosions with the new nvidia driver 525.60.11-1 on a 2070 using arch linux. Downgraded to 520.56.06-12 and everything is fine again. The lucent nargacuga hunt was a consistent way to get the bug to show. Anyone else having the same?

@Billli11 I'm using very similar hardware to you, and also running Arch (but with kwin not wlroots and I'm not on mesa-git) and I do not see this crash. I noticed that you have Steam runtime disabled - could you see if enabling it helps?

@Billli11 I'm using very similar hardware to you, and also running Arch (but with kwin not wlroots and I'm not on mesa-git) and I do not see this crash. I noticed that you have Steam runtime disabled - could you see if enabling it helps?

Think for the suggestion but it's also crash when start steam with steam-runtime and not steam-native.

I also tried the following:

• PROTON_NO_FSYNC=1
  crash
• PROTON_NO_ESYNC=1 PROTON_NO_FSYNC=1
  crash
• mesa and lib32-mesa 22.2.3-1 packages from arch repo
  crash
• Linux kernel 6.0.11.arch1-1 from arch repo
  crash
• Launch game in gamescope
  crash
• Disable Steam Pre-Caching
  crash
• Disable steam overlay
  crash
• Use gamepadui (Steam Deck UI) overlay
  crash
• Using i3 as DE instead of sway (X11 instead of wayland)
  crash
• Switching proton experimental to beta branch (expermental-7.0-20221115 - pinned version) (build id: 9941395)
  work

The strange thing is that all the other games that I played with proton experimental have been running as usual. I only encounter problem with this game.

@Billli11 is there anything suspicious in the output of sudo journalctl -b | grep -i steam after the game crashes? Since the launch day I have seen random crashes in this game. When they happen usually relaunching the game 3 or 4 times in a row is enough to "make it work". It is almost like it has to invalidate some kind of bad data in the shader cache or something similar and repeated attempts to launch somehow make it fix itself for a while.

@Billli11 Wow, thank you for testing all those variables! Those were definitely the major suspects I had after looking at your dist output :D. I have a couple more suggestions in addition to @wwmm ...

1. Could you try adding "VKD3D_CONFIG=swapchain_legacy %command%" to the launch options?
2. If (1 doesn't fix the problem, we very conveniently have experimental-7.0-20221115 pinned as a branch right now. Could you verify that version still works and doesn't crash?
3. If that (2 is a yes, then could you try dropping the vkd3d-proton files from that build (or from 7.0-5) into experimental-7.0-20221129? I think the easiest way is to copy the contents of Proton Experimental/files/lib/wine/vkd3d-proton and files/lib64/wine/vkd3d-proton when you are on the experimental-7.0-20221115 branch and drop them back in place after switching to 1129. Then you can flip-flop experimental branches again or save the original 1129 files before messing around to get back to the original shipped version.

After switching proton experimental from pinned version back to main branch, the game now working now.

So may be there are some corrupted files when proton updated to build id: 10039633 and got fixed by re-downloading the affected files?

I recall having verified integrity of proton's file when the problem happen but it all validated.

I would like to test swapping vkd3d but I can't reproduce the problem and currently locked out by denuvo.

@wwmm Noting weird show up in journalctl. And I tried remove the game's shader.cache2, vkd3d-proton.cache and steam's shadercache.

So it's not likely to be shader compilation problem.

@Billli11 Oh shoot - denuvo :(

So may be there are some corrupted files when proton updated to build id: 10039633 and got fixed by re-downloading the affected files?
That is definitely possible. "verify integrity of files" unfortunately doesn't work properly with Proton Experimental specifically right now because the system doesn't treat nested symlinks properly. So switching branches to trigger a reinstall is a better bet.

It sounds like you are no longer seeing the problem with "stock" experimental-7.0-20221129? Just verifying that I understood you correctly and the problem is fixed :)

@Billli11 Oh shoot - denuvo :(

So may be there are some corrupted files when proton updated to build id: 10039633 and got fixed by re-downloading the affected files?
That is definitely possible. "verify integrity of files" unfortunately doesn't work properly with Proton Experimental specifically right now because the system doesn't treat nested symlinks properly. So switching branches to trigger a reinstall is a better bet.

It sounds like you are no longer seeing the problem with "stock" experimental-7.0-20221129? Just verifying that I understood you correctly and the problem is fixed :)

Yes. Both pinned version and "stock" version of experimental are working before i got locked out.

Yes. Both pinned version and "stock" version of experimental are working before i got locked out.

Awesome :D In that case (in case it wasn't clear), seems unnecessary to do more testing. Luckily the ban is only 24 hr not longer 😅

Getting some vertex explosions with the new nvidia driver 525.60.11-1 on a 2070 using arch linux. Downgraded to 520.56.06-12 and everything is fine again. The lucent nargacuga hunt was a consistent way to get the bug to show. Anyone else having the same?

I also have same problem with 525 driver, my system is nobara on a 3080.

I also have same problem with 525 driver, my system is nobara on a 3080.

Forgot to mention it in my post but I'm also using flatpak steam. What about you?

I also have same problem with 525 driver, my system is nobara on a 3080.

Forgot to mention it in my post but I'm also using flatpak steam. What about you?

mine is from rpmfusion, so I guess steam version has nothing to do with this issue

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise
• Steam AppID of the game: 1446780

System Information

• GPU: GeForce RTX 2070 SUPER
• Driver/LLVM version: 525.60.11 Mesa 22.2.3-1
• Kernel version: 6.0.12-arch1-1
• Proton version: 7.0-5, Experimental, Hotfix, GE 7-42

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

Game crashes when trying to join or create a lobby. Here is the log: steam-1446780.log

Reproduction

Create or join a lobby

Additional Comments

I'm not sure this is a problem with Monster Hunter Rise or with steam. The core dump from coredumpctl dump steam gives

          PID: 33277 (steam)
           UID: 1000 ([REDACTED])
           GID: 1000 ([REDACTED])
        Signal: 6 (ABRT)
     Timestamp: Sat 2022-12-17 10:29:28 CET (1min 22s ago)
  Command Line: /home/[REDACTIED]/.local/share/Steam/ubuntu12_32/steam
    Executable: /home/[REDACTED]/.local/share/Steam/ubuntu12_32/steam
 Control Group: /user.slice/user-1000.slice/session-1.scope
          Unit: session-1.scope
         Slice: user-1000.slice
       Session: 1
     Owner UID: 1000 ([REDACTED])
       Boot ID: 3c24d012f18f4bcb855d9c58253cf3f7
    Machine ID: 457fe70eb0ae401798aad8134b283d11
      Hostname: [REDACTED]
       Storage: /var/lib/systemd/coredump/core.steam.1000.3c24d012f18f4bcb855d9c58253cf3f7.33277.1671269368000000.zst (present)
  Size on Disk: 60.0M
       Message: Process 33277 (steam) of user 1000 dumped core.
                
                Stack trace of thread 33391:
                #0  0x00000000f7ed0549 __kernel_vsyscall (linux-gate.so.1 + 0x549)
                #1  0x00000000f7a89667 n/a (libc.so.6 + 0x89667)
                #2  0x00000000f7a35677 raise (libc.so.6 + 0x35677)
                #3  0x00000000f7a1e126 abort (libc.so.6 + 0x1e126)
                #4  0x00000000edbcf235 n/a (/home/[REDACTED]/.local/share/Steam/ubuntu12_32/vgui2_s.so + 0x15235)
                ELF object binary architecture: Intel 80386

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise
• Steam AppID of the game: 1446780

System Information

• OS: Nobara 36 - Official
• GPU: ASUS ROG Strix GTX 1080-A8G
• Driver/LLVM version: 525.60.11
• Kernel version: 525.60.11
• Link to full system information report as Gist: Gist link
• Proton version: Proton Experimental [bleeding-edge], Proton 7.05 and ProtonGE 42.

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

Missing textures / flickering and black boxes while running game above 60 fps limit cap. Doesn't matter the quality settings

Reproduction

Set FPS cap above 60 (mainly 120+)

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issuecomment-1356144832

I'm experiencing the same issue, same core dump output. I have yet to try running with the logging cranked to max, but that's my planned next step.

@YoungMaestro, @kisak-valve

Anything I can try to fix it? edit: Running this command allowed me to use the graphics card and put the game full screen although now is a complete black screen and I can interact with the game (hear intro music and Navigate through the start menu) WINEDLLOVERRIDES="dxgi=b" DRI_PRIME=1 %command%

I had exact same issue : #6342

The launch parameters given by Maestro (WINEDLLOVERRIDES="dxgi=b" DRI_PRIME=1 %command%) fixed the issue forr me.

My specs are:
Nvidia RTX 2060
AMD Ryzen 3600
16 Gb RAM
Running on Nobara Linux

The real issue is this one : when I try to connect to a lobby or create mine, the game and Steam both crash. (I tried both Steam and Steam flatpak, as well as regular Proton, Proton GE 43, Proton experimental and Proton Hotfix)
First, the game freezes when it tries to find / create the lobby, then it closes and Steam crashes with it. Anyone else experiencing this issue?

This is the same issue that I have, and it appears one other person has. I personally haven't made any progress on this at all since I posted my report, and I am at a loss at what to try next, more or less. It doesn't seem like anything can be changed on my end to fix this issue, but I would love suggestions.

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise
• Steam AppID of the game: 1446780

System Information

• GPU: RX 7900 XTX
• Driver/LLVM version: 4.6 Mesa 22.3.3; LLVM 15.0.6-r1
• Kernel version: 6.1.4-zen2-znver3-o3-pds
• Link to full system information report as Gist: https://gist.github.com/kzdixon/89c2e8be835357675ede275c054d9dc5
• Proton version: Proton Experimental and/or Proton 7.0

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

Monster Hunter Rise never fully compiles shaders on RX 7900 XTX.

Reproduction

1. Launch Monster Hunter Rise
2. Watch as it "compiles" shaders
3. See that it never finishes even if left open for hours

Tested on all versions of Proton I usually use: Proton 7.0-5/6, Proton Experimental (Bleeding-Edge) and GE-Proton7-43.

Update: This is fixed a recent mesa merge (see https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/20941). Should be released in the near future I assume.

I believe the fix has been merged in and released as of mesa 22.3.5.

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issuecomment-1380390130

Very likely this issue wasn't specific to MHR, and the game would merely crash due to the excessive amount of X11 clients spawned due to a bug in the Steam client beta, see here: https://github.com/ValveSoftware/steam-for-linux/issues/9094. This issue has been fixed and the game launches fine now.

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issuecomment-1356271126

I have also been experiencing this exact issue.

System
OS: Linux Mint 21.1 x86_64
Kernel: 5.15.0-58-generic
GPU: Nvidia RTX 3080 TI
GPU Driver: 525.78.01-0ubuntu0.22.04.1

The issue appears to be related to the Nvidia driver; rolling back to 515.86.01 resolves the problem. I first saw this issue on 525.60.11.

System Information

• GPU: AMD RX 7900 XTX
• Driver/LLVM version: Mesa 23.1.0_devel.166448.ebc4893947c.932463d268438ce945b21718552d92ab, LLVM 15.0.7
• Kernel version: 6.1.10-zen
• Link to full system information report as Gist: https://gist.github.com/sclu1034/3a6880b91e42cfb4041fa107c878e43a
• Proton version: experimental & GE-Proton-7-49

steam-1446780.log.gz

Symptoms

The game window never goes into fullscreen. When trying to make it fullscreen via the window manager, it will go back to windowed immediately. It also re-positions itself to 0, 0 when moved.
At the same time, the game's display options are set to fullscreen.
Attempting to change any game option related to the display (e.g. Output Display, Screen Mode, Resolution, Display Frequency) is not possible. When trying to select another value for one of those options, it'll jump back immediately. Moving the window between screens does make it snap to that screen's 0, 0, and the Output Display option changes. But it changes to "4. Display", when there are only three screens physically and only two of them active.

Additionally, the Display Frequency option does not recognize that one display's maximum refresh rate is higher than the other. It only goes up to 144Hz for both of them, but one is capable of 165Hz.

Lastly, the game does not use the hardware potential at all. It sits at around 30 - 32 FPS and the GPU at ~30% utilization. And MangoHUD shows 0 FPS.

@sclu1034
I have the same GPU which works fine (though I am on nobara, mesa 22.3). Maybe you can try to disable integrated graphics in BIOS

@sclu1034 I have the same GPU which works fine (though I am on nobara, mesa 22.3). Maybe you can try to disable integrated graphics in BIOS

That did help, thanks.
Guess I'll have to make due without the iGPU for the time being.

System Information

• GPU: AMD RX 7900 XT
• Driver/LLVM version: Mesa 23.0.0-rc4, LLVM 15.0.7
• Kernel version: 6.1.11-gentoo (with linux-hardened)
• Link to full system information report as Gist: 0xC0ncord/6cc3c6680d0043fd380215e85cd68165
• Proton version: 7.0.6, experimental, GE-Proton-7-49

Note that I've had to add an LD_PRELOAD entry for libstdc++ or else the Steam client will crash on startup due to the presence of libc++ on my system (see https://github.com/ValveSoftware/steam-for-linux/issues/8923#issuecomment-1356656896). I've manually set LD_PRELOAD='' in the game's launch options to ensure this wasn't causing any issues.

Symptoms

Since update 14.0.0 to the game was released a few days ago, I have been completely unable to launch it. Steam will launch the game, and after about 10 seconds, the game crashes with no errors or warnings visible. No game window ever appears (unless it's Denuvo telling me I've reached the activation limit for the day). Prior to this update, everything was perfectly fine. No amount of switching Proton versions, clearing the shader cache, or reinstalling the game has helped thus far. I've also tried running the game under the Flatpak version of Steam without success.

steam-1446780.log

After the recent Proton Experimental version update from 7 to 8, my game crashed at the beginning during the intro videos.

Deleting the prefix fixed the issue.

After finishing a ingame quest, black squares artifacts appear prior to the Quest Cleared logo. Always. This seems to happen after Sunbreak Title Update 5 [15.0.0.0 game version). I tried both the 7-XX and 8-XX GE-Proton branches. I have been playing on Linux since day one and this is my first issue (anything else works fine). I'm not sure if this is the best place to report it though. Anyway, let me thank you for your hard work.

Not sure if this is the right avenue to report this issue, but recently I picked up a Steam Deck and I'm able to reliably reproduce a game freeze. (This issue isn't related to the problems I described above.)

If I connect an external controller to the Deck over USB-C and launch Monster Hunter Rise, the game freezes whenever the controller is supposed to rumble (e.g. when a monster roars). This happens even after disconnecting the external controller and using the Deck's built-in gamepad. A reboot of the Deck without the external controller connected is what is needed to resolve it.

@kisak-valve Turns out in order to reproduce the freeze, the external controller must be disconnected after connecting it at least once. Just to make sure we're on the same page, here is what I did to reproduce it:

1. Log into the Steam Deck.
2. Connect an external controller to the Deck. In my case I have a USB hub connected over USB-C with a Steam controller dongle connected.
3. Disconnect the Deck from the USB-C hub (and thus the Steam controller).
4. Start Monster Hunter Rise.
5. Do anything in-game that causes the controller to rumble and the game will freeze.

Also note that I was able to reproduce this on Proton 7.0-6 and Proton 8.0-2. Here's the log. Thanks!

steam-1446780.log.gz

Replying to https://github.com/ValveSoftware/Proton/issues/5495#issue-1103134357

My issue is very similar. It compiles shaders, then freezes on the health warning screen. except i have an RTX3060, not an AMD GPU

Compatibility Report

• Name of the game with compatibility issues: Monster Hunter Rise
• Steam AppID of the game: 1446780

System Information

• GPU: GTX 1050
• Driver/LLVM version: NVIDIA 530.41.03
• Kernel version: Linux 6.3.2-arch1-1
• Link to full system information report as Gist:Here
• Proton version: Proton Experimental, GE-Proton8-3, Proton 7.0-6

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.

steam-1446780.log

Symptoms

Game status is running, but sometimes show blank window and then goes invisible. My mouse and gamepad still interact with the game. Once in a while, the game runs as usual, usually after i play another game like DOTA2 or MH World. But most of the time, even after playing another game, the same symptom occur. I try to go with the GE version, the same symptoms occur, or just the game keep blinking (more like STROBING like strobe light effect). I try to run steam back into stable version, the same symptom still occur. The game runs well before on 7.0-6 on gnome43, after updating to gnome44, this start to occur. When I try to run it on steam --debug from terminal, I can see the "Fossilize" unable to handle the pipeline. There is a lot more memory address spewed on the error, but one of the Fossilize WARN is like this :

Fossilize WARN: Pipeline library handle 0x00007fe021bc33e0 is not registered.

It has either not been recorded, or it failed to be recorded earlier (which is expected if application uses an extension that is not recognized by Fossilize).

Reproduction

Start the game, or Start the game after playing another game.

Seems like this game has an issue with the shaders having to be built every time.
And sometimes during said every-boot shader compilation, it crashes with a Fatal D3D Error (6).

The crashing isn't very consistent; it happens maybe once every other boot.
This is with the game installed fresh from Steam.

Proton Log here (Experimental).
steam-1446780.log

EDIT: Both shaders not saving/shaders compiling on every boot and the crashing seems to have been resolved by NVIDIA 535.86.05 (was using 535.54.03 when this problem was observed)

There's an issue with flickering reflections or glare on the water in the Frost Islands map. The issue is dependent on which antialiasing setting you're using, specifically so that the flickering is present if TAA is not enabled.

I'm using Proton Experimental.

I recorded a short demonstration of what this looks like: https://gofile.io/d/8u32x1

System:
  Kernel: 6.4.4-200.fc38.x86_64 arch: x86_64 bits: 64 Desktop: GNOME v: 44.3
    Distro: Fedora release 38 (Thirty Eight)
Machine:
  Type: Desktop Mobo: ASRock model: X570 Taichi serial: <superuser required>
    UEFI: American Megatrends v: P5.01 date: 01/18/2023
Battery:
  ID-1: hidpp_battery_0 charge: 88% condition: N/A
CPU:
  Info: 8-core AMD Ryzen 7 5800X3D [MT MCP] speed (MHz): avg: 1807
    min/max: 550/4550
Graphics:
  Device-1: AMD Navi 21 [Radeon RX 6800/6800 XT / 6900 XT] driver: amdgpu
    v: kernel
  Device-2: Realtek Streamplify CAM driver: snd-usb-audio,uvcvideo type: USB
  Display: wayland server: X.Org v: 22.1.9 with: Xwayland v: 22.1.9
    compositor: gnome-shell driver: X: loaded: amdgpu
    unloaded: fbdev,modesetting,radeon,vesa dri: radeonsi gpu: amdgpu
    resolution: 2560x1440~240Hz
  API: OpenGL v: 4.6 Mesa 23.1.4 renderer: AMD Radeon RX 6800 XT (navi21
    LLVM 16.0.6 DRM 3.52 6.4.4-200.fc38.x86_64)
Network:
  Device-1: Intel Wi-Fi 6 AX200 driver: iwlwifi
  Device-2: Intel I211 Gigabit Network driver: igb
Drives:
  Local Storage: total: 1.82 TiB used: 1.09 TiB (60.1%)
Info:
  Processes: 517 Uptime: 42m Memory: available: 31.25 GiB
  used: 5.02 GiB (16.1%) Shell: Zsh inxi: 3.3.27

steam-1446780.log

I've attached the proton log, but it looks like the Steam Runtime Diagnostics is a no-go. Selecting Steam Runtime Diagnostics from the help menu opens the appropriate window as it should, but after a few seconds it causes my desktop session, or something, to crash so badly I need to hit the reset button. Otherwise all I can see is a black screen, and I can't even get into a virtual terminal. I tried it twice now with the same result.

Edit: looks like the Steam Runtime thing crashing with AMD cards is a known issue: https://gitlab.freedesktop.org/drm/amd/-/issues/2384#note_1769931

I've also got some flickering that appears to be a regression in Proton experimental (switching to Proton 8.0-3 fixes it).

log: steam-1446780.log

runtime diagnostics
system information

screenshots:

@Kaedras Could you see if this regression is fixed on the experimental bleeding-edge branch? We are hoping that this is the same regression impacting some other games, and the fix was just pushed live very recently :)

@alasky17 I've just tested it, and it seems like the issue has been fixed.

steam-1446780.log

@mattipulkkinen Thank you for the report.
This should now behave as native Windows on latest Proton Experimental. There will still be a flicker as that is just the game, but it should now not appear a lot worse with AMD + radv.

Hi!
I'm having a pretty big issue with Rise under Proton with my Manjaro. The game completely flickers black a few times every second.
Additionally, the resolution, window mode and display settings can't be changed - they immediately switch back to the original resolution (not shown in the video, since it would have taken too long to get to the menu)
The game's unplayable in this state, sadly.

https://github.com/ValveSoftware/Proton/assets/13204623/2a89bc59-bff1-4e86-b3b0-6c835a329bc4

steam-1446780.log
runtimeDisagnostics.txt

Compatibility Report

Name of the game with compatibility issues: Monster Hunter Rise

Steam AppID of the game: 1446780

System Information

• GPU: AMD Radeon RX 480
• Driver/LLVM version: Mesa 23.2.1-arch1.2 / vulkan-radeon 1:23.2.1-2 / LLVM 16.0.6
• Kernel version: Linux 6.5.7-arch1-1
• Proton version: Proton Experimental, Proton Hotfix, Proton 8.0-4, Proton 7.0-6, Proton 6.3-8

Link to full system information report (proton log and steam system information) error-logs.zip

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

I have the same problem mentioned by @NTrixner, here is a formal report.

The screen flickers whenever the game starts in full screen mode. I am unable to use the in-game settings to change resolution or switch to windowed mode. At around minute 1:52 I pressed Alt+enter to force the game out of full screen.
You can also change the resolution with the .ini file, but by pressing Alt+enter the changes are overwritten and goes back to a window with monitor resolution.

When the game is not flickering, the frame rate hovers at around 10 - 17 FPS. Back when the game first launched, it ran at +60 FPS.

Here is the video proof (seizure warning)

monster hunter flicker compressed.webm

After some time the game crashes and the Capcom crash report tool tries to do something but fails with error code: 0x2000000

Already tried the following launch options without success:

• WINE_FULLSCREEN_FSR=1
• WINE_FULLSCREEN_FSR_STRENGTH=1
• WINEDLLOVERRIDES="dinput8.dll=n,b"
• WINEDLLOVERRIDES="dinput8.dll=b"
• RADV_PERFTEST=nosam

Also, when the game starts, steam throws the following exception. Not sure if this is normal, but haven't seen it in other games.

Oct 21 13:58:51 Zordon-PC steam.desktop[7132]: fsync: up and running.
Oct 21 13:58:51 Zordon-PC steam.desktop[7132]: wine: RLIMIT_NICE is <= 20, unable to use setpriority safely
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]: Unhandled Exception:
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]: System.TypeLoadException: Could not load type of field 'InstallerMessage.Form1+<CodecCheck>d__11:<codecQuary>5__2' (3) due to: Could not load file or assembly 'Windows.Foundation.UniversalApiContract, Version=14.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]:   at InstallerMessage.Form1..ctor () [0x00063] in <fce48153e782462bbf37770256af3859>:0
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]:   at (wrapper remoting-invoke-with-check) InstallerMessage.Form1..ctor()
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]:   at InstallerMessage.Program.Main () [0x0000b] in <fce48153e782462bbf37770256af3859>:0
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]: [ERROR] FATAL UNHANDLED EXCEPTION: System.TypeLoadException: Could not load type of field 'InstallerMessage.Form1+<CodecCheck>d__11:<codecQuary>5__2' (3) due to: Could not load file or assembly 'Windows.Foundation.UniversalApiContract, Version=14.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]:   at InstallerMessage.Form1..ctor () [0x00063] in <fce48153e782462bbf37770256af3859>:0
Oct 21 13:58:53 Zordon-PC steam.desktop[7245]:   at (wrapper remoting-invoke-with-check) InstallerMessage.Form1..ctor()
Oct 21 13:58:54 Zordon-PC steam.desktop[2632]: /bin/sh\0-c\0mangohud /home/kenshinn1/.local/share/Steam/ubuntu12_32/reaper SteamLaunch AppId=1446780 -- /home/kenshinn1/.local/share/Steam/ubuntu12_32/steam-launch-wrapper -- '/home/kenshinn1/Extra2/SteamLibrary/steamapps/common/SteamLinuxRuntime_sniper'/_v2-entry-point --verb=waitforexitandrun -- '/home/kenshinn1/Extra2/SteamLibrary/steamapps/common/Proton - Experimental'/proton waitforexitandrun  '/home/kenshinn1/Extra2/SteamLibrary/steamapps/common/MonsterHunterRise/MonsterHunterRise.exe'\0
Oct 21 13:58:54 Zordon-PC steam.desktop[7270]: ERROR: ld.so: object '/home/kenshinn1/.local/share/Steam/ubuntu12_32/gameoverlayrenderer.so' from LD_PRELOAD cannot be preloaded (wrong ELF class: ELFCLASS32): ignored.
Oct 21 13:58:54 Zordon-PC steam.desktop[7270]: ERROR: ld.so: object '/home/kenshinn1/.local/share/Steam/ubuntu12_32/gameoverlayrenderer.so' from LD_PRELOAD cannot be preloaded (wrong ELF class: ELFCLASS32): ignored.
Oct 21 13:58:54 Zordon-PC steam.desktop[7270]: ERROR: ld.so: object '/home/kenshinn1/.local/share/Steam/ubuntu12_32/gameoverlayrenderer.so' from LD_PRELOAD cannot be preloaded (wrong ELF class: ELFCLASS32): ignored.
Oct 21 13:58:54 Zordon-PC steam.desktop[7270]: ERROR: ld.so: object '/home/kenshinn1/.local/share/Steam/ubuntu12_64/gameoverlayrenderer.so' from LD_PRELOAD cannot be preloaded (wrong ELF class: ELFCLASS64): ignored.
Oct 21 13:58:54 Zordon-PC steam.desktop[7270]: ERROR: ld.so: object '/usr/$LIB/mangohud/libMangoHud_opengl.so' from LD_PRELOAD cannot be preloaded (cannot open shared object file): ignored.
Oct 21 13:58:54 Zordon-PC steam.desktop[7271]: ERROR: ld.so: object '/home/kenshinn1/.local/share/Steam/ubuntu12_32/gameoverlayrenderer.so' from LD_PRELOAD cannot be preloaded (wrong ELF class: ELFCLASS32): ignored.
Oct 21 13:58:54 Zordon-PC steam.desktop[7271]: ERROR: ld.so: object '/home/kenshinn1/.local/share/Steam/ubuntu12_32/gameoverlayrenderer.so' from LD_PRELOAD cannot be preloaded (wrong ELF class: ELFCLASS32): ignored.
Oct 21 13:58:54 Zordon-PC steam.desktop[7272]: ERROR: ld.so: object '/home/kenshinn1/.local/share/Steam/ubuntu12_32/gameoverlayrenderer.so' from LD_PRELOAD cannot be preloaded (wrong ELF class: ELFCLASS32): ignored.
Oct 21 13:58:55 Zordon-PC steam.desktop[7394]: fsync: up and running.
Oct 21 13:58:55 Zordon-PC steam.desktop[7394]: wine: RLIMIT_NICE is <= 20, unable to use setpriority safely
Oct 21 13:58:57 Zordon-PC steam.desktop[7392]: Setting breakpad minidump AppID = 1446780

Reproduction

Start the game in full screen mode, when is not in full screen mode it has very low FPS.

Good day
@kisak-valve,

I have both installed, but I have set the environmental variable AMD_VULKAN_ICD to RADV (https://wiki.archlinux.org/title/Vulkan).
It shouldn't be using AMDVLK.

As you can see on the following picture, RPCS3 describes the GPU with RADV. If i switch the variable to AMDVLK the description changes.

Also tried running the game with AMDVLK with the same result. Do you think this could caused by AMDVLK?

Regards

Wow, thanks to your comment I managed to fix it.
It was AMDVLK, decided to uninstall it and that solved all my problems.

It seems that driver has issues with with Monster Hunter rise and the env variable doesn't switch the drivers.

Capcom updated the game today which completely broke steam deck support

Just going to corroborate this with a log since @LunaRosalie neglected to provide one: steam-1446780.log

Of particular note at the end of the file, possibly(?):

55308.256:012c:0130:trace:loaddll:build_module Loaded L"C:\\Program Files (x86)\\Steam\\gameoverlayrenderer64.dll" at 00000000034C0000: native
55308.256:012c:0130:fixme:steamclient:manual_convert_Set_SteamAPI_CCheckCallbackRegisteredInProcess_func_156 not implemented!
SteamInternal_SetMinidumpSteamID:  Caching Steam ID:  76561198090576201 [API loaded no]
55308.260:012c:0130:err:seh:NtRaiseException Unhandled exception code c0000409 flags 1 addr 0x1514421a8

Just crashes without showing a window

GPU: RTX 2060 Super
Driver: nvidia-dkms 545.29.06-1
steam-1446780-proton-experimental.log
steam-1446780-proton-8_0_4.log

Having the same issue as the above two users, game ran fine yesterday, now won't launch.
Was on Proton 8-0-4, tried Experimental, 7, and GE_8-27 with no success.

Running Steam in the console and looking at the output, this is the only thing that looks like it could be a relevant error.

System.TypeLoadException: Could not load type of field 'InstallerMessage.Form1+<CodecCheck>d__11:<codecQuary>5__2' (3) due to: Could not load file or assembly 'Windows.Foundation.UniversalApiContract, Version=14.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.
  at InstallerMessage.Form1..ctor () [0x00063] in <fce48153e782462bbf37770256af3859>:0 
  at (wrapper remoting-invoke-with-check) InstallerMessage.Form1..ctor()
  at InstallerMessage.Program.Main () [0x0000b] in <fce48153e782462bbf37770256af3859>:0 
[ERROR] FATAL UNHANDLED EXCEPTION: System.TypeLoadException: Could not load type of field 'InstallerMessage.Form1+<CodecCheck>d__11:<codecQuary>5__2' (3) due to: Could not load file or assembly 'Windows.Foundation.UniversalApiContract, Version=14.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.
  at InstallerMessage.Form1..ctor () [0x00063] in <fce48153e782462bbf37770256af3859>:0 
  at (wrapper remoting-invoke-with-check) InstallerMessage.Form1..ctor()
/bin/sh\0-c\0/home/ikj/.steam/debian-installation/ubuntu12_32/reaper SteamLaunch AppId=1446780 -- /home/ikj/.steam/debian-installation/ubuntu12_32/steam-launch-wrapper -- '/home/ikj/.steam/debian-installation/steamapps/common/SteamLinuxRuntime_sniper'/_v2-entry-point --verb=waitforexitandrun -- '/home/ikj/.steam/debian-installation/compatibilitytools.d/GE-Proton8-27'/proton waitforexitandrun  '/home/ikj/Games/steam/steamapps/common/MonsterHunterRise/MonsterHunterRise.exe'\0
chdir "/home/ikj/Games/steam/steamapps/common/MonsterHunterRise"

I do note it says <codecQuary>, but I can only find reference in Microsoft's docs to <codecQuery>. Typo?

The reference to codecQuary is likely internal code.

The true issue may be

Could not load file or assembly 'Windows.Foundation.UniversalApiContract, Version=14.0.0.0,
Culture=neutral, PublicKeyToken=null' or one of its dependencies.

A quick sleuth around the 'net seems to indicate this (possibly missing?) library is related to the Microsoft UWP toolkit.

We pushed a new build to Proton Hotfix that fixes the crash on launch. It wasn't really a Proton bug, but we were able to work around it easily :)

Limited play, but confirmed. The game is functional again.

Played some online and 1 hunt offline and it works

Working again, done a few hunts and working as it was before.

Since Capcom's patch, Proton 8.0-5 works fine again. With hotfix it takes me about three attempts starting the game for it to progress past the health warning screen on steam deck. With 8.0-5 it works every time. I recommend changing the default version back to 8.0-5.

@Sanaki Thank you for the report. It turns out that there was an intermittent bug that we started hitting that has been around forever and probably causing other random crashes that we could never track down. Luckily the failure was consistent enough with MHR that we were able to figure out the root cause and fix it :D The fix is in current hotfix and bleeding-edge. If you see that crash again with a Proton version newer than current hotfix or bleeding-edge, please ping me as currently I'm working under the assumption that it has been fixed. Note - the fix is not in 8.0-5 or experimental-8.0-20240111.

Crashing on launch with this error:

Wine C++ Runtime Library
Assertion failed!

Program: Z:\home\pez\.local\share\Steam\steamapps\common\MonsterHunterRise\MonsterHunterRise.exe
File: ../src-wine/dlls/winevulkan/loader_thunks.c
Line: 2934

Expression: !status && "vkCreateGraphicsPipelines""

Press OK to exit the program, or Cancel to start the Wine debugger

Sometimes I reach in game, go out on a quest and then the game crashes at the loading screen instead. Sometimes I reach "compiling shaders" which seems to hang forever. The log increases rapidly in size (8gb+!) so I can't upload those logs. This log is from crashing at launch:
steam-1446780.log

Hello @clankfish, can you check if https://gitlab.freedesktop.org/drm/amd/-/issues/3343 is relevant to your system? A quick test would be to reboot into an older kernel and see how the game behaves.

Hi @kisak-valve, enabling 4g decoding and resizeable bar in UEFI settings seems to have completely fixed this issue. Thank you very much! I was starting to worry that my new gpu was faulty or something lol phew :))

I (finally) got it to work too.

Just had to remove vulkan-amdgpu-pro, so it JUST uses vulkan-radeon.

as of recently i cant get MHR to work with Proton 8, 9, or Experimental as it just closes once it reaches the Health warning. Proton 8.0-5 just crashes at launch, 9.0-2 reaches health warning as does experimental.

It works fine with Proton GE 9-7 and worked fine with regular proton as of a few weeks ago at least which was when i got my new system and was trying the game out.

There doesnt appear to be much different in the proton logs between them, it just closes.

Crash at the health warning means the anticheat triggered. I'm afraid I can't tell you what would lead to that happening though. If you use mods, try updating REFramework.

Crash at the health warning means the anticheat triggered. I'm afraid I can't tell you what would lead to that happening though. If you use mods, try updating REFramework.

dont use mods, tried with and without RE framework. Really have no clue whats happening as it use to work fine with plain jane proton but there isnt anything in the logs to give me a direction as to whats going on with proton + the new DRM. I didnt play for a few weeks and now it just stopped working, I suppose at least GE works so i can play still.

EDIT: After i said this i tried a hail mary and remove DRI_PRIME option and that fixed it, apparently forcing a specific GPU caused it

heavy flickering on debian trixie with amd gpu
tried with multiple proton versions, even got the source code and compiled it myself, nothing seems to fix it

https://github.com/user-attachments/assets/7d32dd35-1ef0-4e23-bbe2-e29ddf89af14

System info:

Computer Information:

Manufacturer: Gigabyte Technology Co., Ltd.
Model: H410M H V2
Form Factor: Desktop
No Touch Input Detected
Processor Information:
CPU Vendor: GenuineIntel
CPU Brand: Intel(R) Core(TM) i5-10400 CPU @ 2.90GHz
CPU Family: 0x6
CPU Model: 0xa5
CPU Stepping: 0x3
CPU Type: 0x0
Speed: 4300 MHz
12 logical processors
6 physical processors
Hyper-threading: Supported
FCMOV: Supported
SSE2: Supported
SSE3: Supported
SSSE3: Supported
SSE4a: Unsupported
SSE41: Supported
SSE42: Supported
AES: Supported
AVX: Supported
AVX2: Supported
AVX512F: Unsupported
AVX512PF: Unsupported
AVX512ER: Unsupported
AVX512CD: Unsupported
AVX512VNNI: Unsupported
SHA: Unsupported
CMPXCHG16B: Supported
LAHF/SAHF: Supported
PrefetchW: Unsupported
BMI1: Supported
BMI2: Supported
F16C: Supported
FMA: Supported

Operating System:

Debian GNU/Linux 13 (trixie) (64 bit)
Kernel Name: Linux
Kernel Version: 6.12.43+deb13-amd64
X Server Vendor: The X.Org Foundation
X Server Release: 12101016
X Window Manager: KWin
Steam Runtime Version: steam-runtime_1.0.20250723.149234
Client Information:
Version: 1759461205
Browser GPU Acceleration Status: Enabled
Browser Canvas: Enabled
Browser Canvas out-of-process rasterization: Enabled
Browser Direct Rendering Display Compositor: Disabled
Browser Compositing: Enabled
Browser Multiple Raster Threads: Enabled
Browser OpenGL: Enabled
Browser Rasterization: Enabled
Browser Raw Draw: Disabled
Browser Skia Graphite: Disabled
Browser Video Decode: Enabled
Browser Video Encode: Disabled
Browser Vulkan: Disabled
Browser WebGL: Enabled
Browser WebGL2: Enabled
Browser WebGPU: Disabled
Browser WebNN: Disabled

Video Card:

Driver: AMD AMD Radeon RX 6600 (radeonsi, navi23, LLVM 19.1.7, DRM 3.61, 6.12.43+deb13-amd64)
Driver Version: 4.6 (Compatibility Profile) Mesa 25.0.7-2
Desktop Color Depth: 24 bits per pixel
Monitor Refresh Rate: 74 Hz
VendorID: 0x1002
DeviceID: 0x73ff
Revision Not Detected
Number of Monitors: 1
Number of Logical Video Cards: 1
Primary Display Resolution: 1920 x 1080
Desktop Resolution: 1920 x 1080
Primary Display Size: 23.54" x 13.23" (26.97" diag), 59.8cm x 33.6cm (68.5cm diag)
Primary VRAM: 8192 MB

Sound card:

Audio device: Realtek ALC897

Memory:

RAM: 15904 Mb

Is anyone else unable to enable HDR in this game?
My launch arguments are:
PROTON_ENABLE_WAYLAND=1 PROTON_ENABLE_HDR=1 ENABLE_HDR_WSI=1 %command%
This config works for me in other games like Overwatch but not in this game. Is this a bug?

Operating System: Arch Linux
KDE Plasma Version: 6.5.2
KDE Frameworks Version: 6.19.0
Qt Version: 6.10.0
Kernel Version: 6.17.7-5-cachyos (64-bit)
Graphics Platform: Wayland
Processors: 24 × 13th Gen Intel® Core™ i7-13700K
Memory: 64 GiB of RAM (62.5 GiB usable)
Graphics Processor: NVIDIA GeForce RTX 3090 Ti
Display: Philips PHL 34M2C8600 (HDR capable!)

@joni999 Which proton version? proton no not have wine wayland.
And PROTON_ENABLE_WAYLAND=1 and PROTON_ENABLE_HDR=1 are for other proton fork like ge-proton or proton-em.

Nvidia gpu also need This. (Which ENABLE_HDR_WSI=1 enable).

Sorry, you're right. I'm on GE-Proton10-13 so this is the wrong place to report this.

Thank you for the hint with the vk-hdr-layer for Nvidia.
For some reason I thought this was obsolete but obviously it's not because it works perfectly now with it installed.
The HDR switch is now toggleable and the game looks great.
I apologize for pinging everyone and thank you again!

When using Proton Experimental, quitting Rise has the process continue to run with a thread at 100%. I haven't let this go on longer than a minute or two before forcing it to stop through the library button. Proton 10.0-4 quits cleanly as expected.

Logs:
steam-1446780-experimental.log
steam-1446780-proton-10.log

@DonKatsu How specifically are you closing MH:R? Are you playing for a long time first, or can you get the game to hang just by launching and exiting again? Are you using an in-game exit (if so, which one) or alt+f4? Could you also please copy your system information from Steam (Steam -> Help -> System Information and Steam -> Help -> Steam Runtime Diagnostics) and put each in a gist, then include a link to the gists in this issue report?

Those logs were from starting the game, then selecting Quit from the main menu. But experimental and 10.0-4 behave the same after playing for say, an hour, then selecting Quit Game from the start menu.
To be exact, the game window does close out, but the process remains with a thread stuck at 100% utilization.
System Information
Steam Runtime Diagnostics

EDIT: Looks like it's because of reframework. Experimental quits properly with it disabled by removing its dinput8.dll.
Should I report it there, then? But given that it's not an issue with Proton 10.0-4, there's something going on with the Wine side...
EDIT2: Hmm, grabbed the latest reframework nightly and Expetimental quit properly... The one that isn't is the latest stable (v1.5.9.1).

Original report: https://gitlab.freedesktop.org/mesa/mesa/-/work_items/14999

Monster Hunter Rise crashes (segfault) when I'm connected to my company's VPN or behind its proxied network.

If I'm on the VPN or behind the corporate network (which requires the use of proxies), if I click PLAY on the Steam interface, the game will launch a screen with an image and a bar that says COMPILING SHADERS. After the bar is full (all shaders compiled), the game then crashes: it closes, and shows the Capcom "crash report" little window. If I disconnect from the VPN and try to do the same thing, the game will reach the main menu and work fine.

If I am in the lab nework inside the company, when I have the proxy stuff configured through the usual environment variables (http_proxy, https_proxy, no_proxy, socks_proxy, etc), the same crash happens. If I use some sort of Kernel-level solution to transparently handle proxies (it's a combination of a program called redsocks and iptables), then the game works fine.

I have tested this against multiple Intel machines: TGL, MTL, LNL, BMG, PTL. They all behave the same.

If I gdb -p the game, I see it gets SIGSEGV, but the backtrace is not very helpful.

Thread 37 "MonsterHunterRi" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x1024ff6c0 (LWP 545566)]
0x000000014350b751 in ?? ()
(gdb) bt
#0  0x000000014350b751 in ?? ()
#1  0x00000000243d2160 in ?? ()
#2  0x0000000006b6dc30 in ?? ()
#3  0x0000000000000000 in ?? ()

See the attached logs.

monster-hunter-rise-logs.tar.gz

There appears to be an issue with REFramework on aarch64. Game immediately crashes on Proton 11 ARM with the same exact installation on x86 REFramework Proton 11. Will try to gather logs later