Open kzdixon opened 1 year ago
Game launches and runs fine here, at least during the first mission.
~However, I'm having issues getting the game to use my Dualsense controller (Flatpak Steam). Tried fiddling a bit with Steam input but no luck so far.~ Fixed by using gamescope.
From Easy Anti-cheat window: "Launch Error" "Unexpected error. (#1)"
This system uses the same game install folder for 2 (two) users. However each user has its own game-running-software in the $HOME directories--Easy Anti-cheat; Proton; runtime_sniper/solder, etc.
Running Steam from console show no permission errors (anymore).
1633.789:0028:002c:warn:seh:check_bpf_jit_enable Could not open /proc/sys/net/core/bpf_jit_enable.
1635.257:0128:0170:err:kerberos:kerberos_LsaApInitializePackage no Kerberos support, expect problems
1635.262:0128:0170:err:ntlm:ntlm_LsaApInitializePackage no NTLM support, expect problems
I played the first 10-11 mission. But my game seems to freeze when I fight Balteus. Does anybody else got this issue?
Found my issue. It seems that resizable bar made the game freeze on this fight. This is not the first game causing issues with rezisable bar on my system. So I don't think it's Armored core related.
NVIDIA 3070 RTX AMD RYZEN 3700X NVIDIA 535.104.05
Game also crashes for me a few seconds into the Balteus fight in the Watchpoint level. This is a few hours in and 100% repeatable.
Aug 24 23:59:59 torrent kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, signaled seq=84250579, emitted seq=84250581
Aug 24 23:59:59 torrent kernel: [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process armoredcore6.ex pid 44352 thread armoredcore6.ex pid 44352
Aug 24 23:59:59 torrent kernel: amdgpu 0000:03:00.0: amdgpu: GPU reset begin!
Aug 24 23:59:59 torrent kernel: amdgpu 0000:03:00.0: amdgpu: BACO reset
Aug 25 00:00:01 torrent kernel: amdgpu 0000:03:00.0: amdgpu: GPU reset succeeded, trying to resume
Aug 25 00:00:01 torrent kernel: [drm] PCIE GART of 512M enabled (table at 0x00000081FEE00000).
Aug 25 00:00:01 torrent kernel: [drm] VRAM is lost due to GPU reset!
Aug 25 00:00:01 torrent kernel: [drm] PSP is resuming...
Aug 25 00:00:01 torrent
kernel: [drm] reserve 0x900000 from 0x81fd000000 for PSP TMR
GPU crashes then recovers but takes X with it in the process.
Guess I'll have to try disabling resizable BAR.
Edit: Disabling resizable BAR did not help.
@brandonegbert are you to provide a save game if they are sharable? 👀
Reproducing this is going to be a nightmare if we cannot share saves. I'm seriously struggling getting past the tutorial boss ...
Not having any problems with my 5900X+6700XT yet, besides lower performance than I would have expected.
But while experimenting with WINE_CPU_TOPOLOGY I managed to get the game to crash as it launches after the EAC popup.
WINE_CPU_TOPOLOGY=12:0,1,2,3,4,5,12,13,14,15,16,17
to contain the game to the first CCD crashes every time. steam-1888160.log
WINE_CPU_TOPOLOGY=12:0,1,2,3,4,5,6,7,8,9,10,11
for just physical threads of both CCDs works.
The game that lead to me finding out about WINE_CPU_TOPOLOGY, Mechwarrior Online, doesn't crash with the first option.
I got past the boss and grinding through the (trivial) missions ...
I'm stuck at "[Press Any Button]"
Disregard, I guess. Maybe I wasn't giving it enough time to compile shaders or something. It seems to work fine now.
My game was running completely fine for 4 hours, but that was until I got to the boss: "AA P07 Balteus". The game keeps crashing when he does a flamethrower spinning attack, it's the second move he does upon entering combat. I've crashed at the same point 6 times now, I've tried different Proton versions like 7.0-6, 8.0-3 and experimental.
Also getting the Watchpoint/Baltheus crash. Both on Proton 8 and bleeding-edge. The annoying part is it happens on the end boss of a lengthy level and you have to start it over every time it crashes. Will continue testing.
Saves are too big to upload here and don't compress at all, so here a link: https://drive.google.com/file/d/1HJ5uHyvlOEP7mtbbAofGBeee6ycounn6/view?usp=sharing Saves are found in (guessing the numbers folder might have different numbers): pfx/drive_c/users/steamuser/AppData/Roaming/ArmoredCore6/76561197967793517/
Looks like @frozen-sea got you a save game but if you need another I can get one uploaded tonight.
On a hunch, I started by trying it in offline mode without the anti-cheat. Was able to make multiple attempts lasting several minutes each (even got him to below 50% once!) at the boss in question without crashing, where as previously it would crash within like 30 seconds.
Launch option: eval $(echo "%command%" | sed "s/start_protected_game.exe/armoredcore6.exe/")
Adding log of that session for good measure. steam-1888160-nocrash.zip
Edit: I then un-did the workaround and can no longer reproduce it on Proton 8 with no launch options... Literally nothing else has changed, not even rebooted.
For anyone that has it working: are you on RDNA3 at all?
I'm having the same issue
Arch, kernel 6.4.12, nvidia 535.104.05, proton 8.0-3/bleeding-edge, Hybrid and dedicated graphics tested. System: Intel i5 12500H, RTX 3050 mobile and 16 GB of ram.
I've already tested a variety of graphics combinations, but with no luck.
Edit: I forgot to mention the issue only occurs at the start of the chapter 1 last boss as it was mentioned in another comment. But to add: I could reproduce the freeze as well on the Strider mission setting the graphics settings to low (from normal or high) for some reason.
I cannot reproduce any hang on RX 7600. Finally got to the boss after playing this all day <_<
I can reproduce it on 5700xt however.
Steam Overlay doesn't work. I had this problem also on Elden Ring. idk If that has something to do with that, but the Easy Anti Cheat slpash screen has the Steam Overlay.
Couldn't beat it but I got Balteus to about 25% left on the second phase without a crash of any sort. That super assault armor attack of it tanks my fps to 30 every time, though... Still 6700XT.
Tried @frozen-sea solution of disabling EAC, but unfortunately it doesn't do the trick for me - game still crashes at around boss's second salvo, never experienced a crash before this mission.
Using Nvidia RTX 1650 Mobile (535.98), with AMD Ryzen 5 4600H on Arch Linux 6.4.11-arch1-1.
Tried disabling EAC, tried some other suggestions including: reducing graphics fidelity and disabling auto-adjust but I also kept crashing on the first salvo of Balteus.
My only solution was to get AC6 on my Steam Deck and then beat the boss from there.
Now I have a 100% reproducible error on the mission immediately afterwards. Where 'restart from checkpoint' will cause my system to crash on the fight with 'The Cleaner.'
Finally root caused it. It's a game bug.
It crashes in a shader which samples a texture, but there's actually a uniform buffer there, which causes GPU crash:
============
Fault type: MISMATCH_DESCRIPTOR_TYPE
CBV_SRV_UAV heap cookie: 7
Shader hash and instruction: b8140dcf1aa2320a (1)
Accessed resource/view cookie: 9
Shader desired descriptor type: 1 (SAMPLED_IMAGE)
Found descriptor type in heap: 4 (UNIFORM_BUFFER)
Failed heap index: 10336
==========
0298:err:vkd3d_descriptor_debug_qa_check_entry: Num failed checks: 212
0298:err:vkd3d_descriptor_debug_qa_check_report_fault:
============
Fault type: MISMATCH_DESCRIPTOR_TYPE
CBV_SRV_UAV heap cookie: 7
Shader hash and instruction: b8140dcf1aa2320a (1)
Accessed resource/view cookie: 9
Shader desired descriptor type: 1 (SAMPLED_IMAGE)
Found descriptor type in heap: 4 (UNIFORM_BUFFER)
Failed heap index: 24823
==========
0298:err:vkd3d_descriptor_debug_qa_check_entry: Num failed checks: 212
0298:err:vkd3d_descriptor_debug_qa_check_report_fault:
============
Fault type: MISMATCH_DESCRIPTOR_TYPE
CBV_SRV_UAV heap cookie: 7
Shader hash and instruction: 8450c12e29bda693 (8)
Accessed resource/view cookie: 9
Shader desired descriptor type: 1 (SAMPLED_IMAGE)
Found descriptor type in heap: 4 (UNIFORM_BUFFER)
Failed heap index: 1521
==========
0298:err:vkd3d_descriptor_debug_qa_check_entry: Num failed checks: 905
0298:err:vkd3d_descriptor_debug_qa_check_report_fault:
============
Fault type: MISMATCH_DESCRIPTOR_TYPE
CBV_SRV_UAV heap cookie: 7
Shader hash and instruction: 8450c12e29bda693 (8)
Accessed resource/view cookie: 9
Shader desired descriptor type: 1 (SAMPLED_IMAGE)
Found descriptor type in heap: 4 (UNIFORM_BUFFER)
Failed heap index: 5745
==========
Using descriptor QA also masks the bug since invalid access will be turned into a NULL descriptor.
How did you manage to start it at all? My easy anti cheat launcher just is stuck "Initializing" indefinitely. nvidia 3090 on ubuntu 22.04
https://discord.com/channels/481706387278397440/1144637082799710280/1144742374623416380
There's a handful of reports of similar crashes at the start of the Balteus fight on Windows in this thread: https://steamcommunity.com/app/1888160/discussions/0/3820795131605892308/ And this one: https://steamcommunity.com/app/1888160/discussions/1/3820795131608400290/
May not be isolated to Linux.
There's a handful of reports of similar crashes at the start of the Balteus fight on Windows in this thread: https://steamcommunity.com/app/1888160/discussions/0/3820795131605892308/ And this one: https://steamcommunity.com/app/1888160/discussions/1/3820795131608400290/
May not be isolated to Linux.
With this, the threads got cross-referenced. I wonder if there could be a proton workaround for the problem or if this can only be fixed by the developers. I really appreciate the finding, and I wonder how we could report this to Bandai to get this fixed asap.
Finally root caused it. It's a game bug.
It crashes in a shader which samples a texture, but there's actually a uniform buffer there, which causes GPU crash:
Create a new mesa issue and request a workaround: https://gitlab.freedesktop.org/mesa/mesa/-/issues
GiGurra did you try @frozen-sea launch option? That workaround gets the game to start on my machine: eval $(echo "%command%" | sed "s/start_protected_game.exe/armoredcore6.exe/")
Timon was having a similar EAC launch issue on windows: https://steamcommunity.com/app/1888160/discussions/0/3820795131605739862/?ctp=3#c3820795131606924426
Timons windows solution:
However, I read that wine by default runs programs with administrator previlages. Does someone know how to confirm administrator previlages for start_protected_game.exe and armoredcore6.exe running in Proton?
I was finally able to finish Balteus without crashing - in offline mode without the anti-cheat by using @frozen-sea launch option:
eval $(echo "%command%" | sed "s/start_protected_game.exe/armoredcore6.exe/")
Replying to https://github.com/ValveSoftware/Proton/issues/7045#issuecomment-1694143730
I did try running the armoredcore6.exe instead of the eac launcher, like was suggested here, but that only launches, turns into a black screen with game cursor, and then 1 second later closes.
I am currently still running into the issue of starting up the game I get a white screen for about 4 seconds with a curious that I can see and moves when it goes to black my cursor is gone and the game crashes anyone any advice I tried everything in these guides
https://www.ghostarrow.com/armored-core-6-crashing-not-launching-white-black-screen-fix https://updatecrazy.com/armored-core-6-wont-launch-black-screen-easy-fix/ https://digistatement.com/armored-core-6-ac6-fires-of-rubicon-black-screen-issue-is-there-any-fix-yet/ https://www.reddit.com/r/armoredcore/comments/160gcs1/black_screen_on_startup/
Replying to #7045 (comment)
I did try running the armoredcore6.exe instead of the eac launcher, like was suggested here, but that only launches, turns into a black screen with game cursor, and then 1 second later closes.
If I try to add extra commands to the eval
part e.g. gamemoderun mangohud %command%
- game crashes either with white or black screen. Removing gamemoderun
and using only mangohud %command%
does work for me, tho.
I was finally able to finish Balteus without crashing - in offline mode without the anti-cheat by using @frozen-sea launch option:
eval $(echo "%command%" | sed "s/start_protected_game.exe/armoredcore6.exe/")
@Cyber-Oto what is the windows equivalent for this, is it just running in administrator mode? if i try launching one in administrator mode, how can i be assured the other exe will as well, even if i set the compatibility setting to administrator mode permanently?
I was finally able to finish Balteus without crashing - in offline mode without the anti-cheat by using @frozen-sea launch option:
eval $(echo "%command%" | sed "s/start_protected_game.exe/armoredcore6.exe/")
@Cyber-Oto what is the windows equivalent for this, is it just running in administrator mode? if i try launching one in administrator mode, how can i be assured the other exe will as well, even if i set the compatibility setting to administrator mode permanently?
@aram909 What this command does is telling Steam to use armoredcore6.exe
instead of start_protected_game.exe
.
I couldn't really find any suitable one-liner for Windows, so even though there should be one, what I can advice is to just paste there your full path to the .exe, something like this.
"C:\Program Files (x86)\Steam\steamapps\common\ARMORED CORE VI FIRES OF RUBICON\Game\armoredcore6.exe" %command%
It will try to use that .exe
directly. I believe you must keep %command%
, or Steam will ignore your custom command altogether. Just make sure that path is adapted to your setup and points to armoredcore6.exe
on your machine.
Not sure about setting administrator mode, you can try to set if it doesn't work. A sign that it worked will be that game stars without EAC popup and uses offline mode.
Also, I should point out that it is not guaranteed to fix Balteus crash, unfortunately, since it worked for some and didn't for another, like me.
yeah EAC kept popping up
do I put it into the alternate launch options for steam, like so?
or directly into the terminal/cmd prompt?
@aram909 Yes, directly there in Launch options, make sure that you have set your actual path to the file and kept all double quotes with %command%
, i.e. "FULL_PATH_TO_EXE" %command%
otherwise it won't work. If all is correct and it still won't work, not sure if I can help further, since I use Linux without dual-boot to Windows to check.
%command%
Isn't a thing in Windows Steam. Unless they added it with the GUI update.
This seems to launch fine straight out of the box and works fine with keyboard and mouse but not with my ps4 controller or xbox controller when they work in all other games ok.
Symptoms
On starting the game, it crashes after briefly loading. I expect a pre-rendered video is the issue, but no versions of Proton I've tried including GE seem to get it working past the menu screen on my end.
Reproduction
* Open the game. * Click New Game or Continue. * It crashes.
It seems that Steam's preload may have been the root cause of my problems regarding EAC and the game not letting me play.
After verifying the files, the game patched ~35GB and downloaded ~800MB and EAC now seems to allow me to log in. I can also now actually get past the first loading screen.
is there any confirmed fix to the crashes on windows?
seems like a patch is necessary for some of the issues, unfortunately. based on earlier in the thread, its a shader compilation issue?
I repeated under a number of different settings configurations, probably 2 dozen in all, and i found that when i turned my settings to low, and at lower and lower resolutions, the crash occured later and later in the particular fight.
Saw a post on reddit here https://www.reddit.com/r/armoredcore/comments/16184hj/comment/jxzrcdu/?utm_source=share&utm_medium=web2x&context=3 recommending proton 7. I went from crashing consistently after about the second volley in the Balteus fight, to being able to do 5 attempts without a crash. Seems to have fixed, or at least improved the situation.
I lost on purpose to test it multiple times, clearly.
Saw a post on reddit here https://www.reddit.com/r/armoredcore/comments/16184hj/comment/jxzrcdu/?utm_source=share&utm_medium=web2x&context=3 recommending proton 7. I went from crashing consistently after about the second volley in the Balteus fight, to being able to do 5 attempts without a crash. Seems to have fixed, or at least improved the situation.
I lost on purpose to test it multiple times, clearly.
I've finished the fight! I had to lower the game resolution to 800x600 and settings to low to get decent FPS, but going back to Proton 7 made the fight not freeze/crash the game.
Agreed, swapped to proton 7 and was able to get through the fight
At least, it held up as long enough as my skill issue took to catch up to get through the boss.
Can confirm that using Proton 7.0-6 allowed me to kill the boss after several attempts. No crashes at all, but performance got a severe hit, especially in Sortie screen and when assembling AC.
Unfortunately the next mission is basically unplayable for me due to extremely low fps, regardless of graphics settings. So I'd say while 7.0-6 allowed to pass Balteus, further playthrough is out of the question like this.
Can confirm that using Proton 7.0-6 allowed me to kill the boss after several attempts. No crashes at all, but performance got a severe hit, especially in Sortie screen and when assembling AC.
Unfortunately the next mission is basically unplayable for me due to extremely low fps, regardless of graphics settings. So I'd say while 7.0-6 allowed to pass Balteus, further playthrough is out of the question like this.
I went back to proton 8.0-3 to continue playing and was able to complete the next mission and arena fights with no problem. People were reporting some weird red flickering during the boss fight, so the shader problem might cause freeze/crash on some GPUs while others only have the flicker effect.
Replying to https://github.com/ValveSoftware/Proton/issues/7045#issuecomment-1694789530
Checked further and yes, all arena fights and several next missions completed without a hitch on GE Proton 8-13 into the beginning of chapter 3. As a sidenote, using Low settings causes crashes even in usual fights, playing on High does not. I guess different shaders are being used depending on the quality, so some as broken, some are not.
Proton 7.0-6 also crashes for me in the same way on Balteus.
I will be looking into a potential workaround.
Getting a strange hang (on Proton 7.0, 8.0, Experimental, and GE 8-13) after some time playing the game (sometimes 30-40 minutes, sometimes 5-15). Specs are AMD Ryzen 5900X, 32GB 3200MHz, AMD 7900 XTX (specifically a Merc 310) Running on RebornOS/Arch.
My screens freeze, and if I'm lucky I can get out into a tty upon which they unfreeze again. Can't get back into KDE (on Wayland) though, even if I kill the game and related processes :/
I can get any logs that could help, I just need to know what those would be, and where they might be located.
This is cut from the end of dmesg
prior to shutting down via tty.
[ 6902.299753] amdgpu 0000:0c:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:3 pasid:32831, for process armoredcore6.ex pid 122266 thread armoredcore6.ex pid 122266)
[ 6902.299761] amdgpu 0000:0c:00.0: amdgpu: in page starting at address 0x000080000843c000 from client 10
[ 6902.299764] amdgpu 0000:0c:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00301430
[ 6902.299766] amdgpu 0000:0c:00.0: amdgpu: Faulty UTCL2 client ID: SQC (data) (0xa)
[ 6902.299768] amdgpu 0000:0c:00.0: amdgpu: MORE_FAULTS: 0x0
[ 6902.299770] amdgpu 0000:0c:00.0: amdgpu: WALKER_ERROR: 0x0
[ 6902.299771] amdgpu 0000:0c:00.0: amdgpu: PERMISSION_FAULTS: 0x3
[ 6902.299773] amdgpu 0000:0c:00.0: amdgpu: MAPPING_ERROR: 0x0
[ 6902.299775] amdgpu 0000:0c:00.0: amdgpu: RW: 0x0
[ 6912.734675] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx_0.0.0 timeout, but soft recovered
I'm using echo "mangohud %command%" | sed 's/start_protected_game.exe/armoredcore6.exe/' | sh
as the custom launch settings to run the game without EAC. I've also had gamemoderun
and DXVK_ASYNC
in there, but no difference (both to performance and crashyness).
quick edit: I know this has a non-zero chance of not being relevant here, but I couldn't get into the freedesktop gitlab (because it's timing out, 504) mesa to look around there to find somewhere more relevant
2023-08-29 update: I got it to work using the amdvlk
drivers instead of vulkan-radeon
. No crashes so far, and only a few issues towards the start of my session.
The game will no longer start for me. I just get a white screen and then I am booted out. It used to work just fine.
Replying to https://github.com/ValveSoftware/Proton/issues/7045#issuecomment-1694514368
Verifying the files did not work for me. I suspect it's a patch related thing.
Compatibility Report
System Information
I confirm:
steam-1888160.log
Symptoms
On starting the game, it crashes after briefly loading. I expect a pre-rendered video is the issue, but no versions of Proton I've tried including GE seem to get it working past the menu screen on my end.
Reproduction
Have tried every permutation possible to try and get it working on my system:
mesa
with and without LTOmangohud gamemoderun
Notably:
taskset
is used at allUPDATE: It seems that Steam's preload may have been the root cause of my problems regarding EAC and the game not letting me play.
After verifying the files, the game patched ~35GB and downloaded ~800MB and EAC now seems to allow me to log in. I can also now actually get past the first loading screen.