https://bugs.winehq.org/show_bug.cgi?id=39180
Bug ID: 39180 Summary: Witcher 3 Wild hunt Wine 1.7.50 not working Product: Wine Version: 1.7.50 Hardware: x86 OS: Linux Status: UNCONFIRMED Severity: blocker Priority: P2 Component: directx-d3dx10 Assignee: wine-bugs@winehq.org Reporter: f00dl3a@gmail.com Distribution: ---
Created attachment 52217 --> https://bugs.winehq.org/attachment.cgi?id=52217 Backtrace
Witcher 3 Wild Hunt is still not working under WINE. Instead of a vcomp110.dll error (as in Bug 38613) - now it is just crashing with a "Page fault on write access to 0x00000000 in 64-bit code" message.
Back trace attached. Also Witcher3 error log is attached.
Ubuntu 15.10, Witcher 3 patched to 1.06
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #1 from Anthony f00dl3a@gmail.com --- Created attachment 52218 --> https://bugs.winehq.org/attachment.cgi?id=52218 Witcher 3 crash log
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #2 from Anthony f00dl3a@gmail.com --- Please note, this is the GoG version
I am including the GoG DLL,
wine witcher3.exe goggame-1207664643.dll
https://bugs.winehq.org/show_bug.cgi?id=39180
Józef Kucia joseph.kucia@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |joseph.kucia@gmail.com
--- Comment #3 from Józef Kucia joseph.kucia@gmail.com --- Could you attach a debug log with +d3d11 channel?
Also, the importance and component fields are wrong. The "Component" field should be set to directx-d3d and "Importance" to normal. See http://wiki.winehq.org/Bugs
https://bugs.winehq.org/show_bug.cgi?id=39180
Nikolay Sivov bunglehead@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|directx-d3dx10 |directx-d3d Severity|blocker |normal
https://bugs.winehq.org/show_bug.cgi?id=39180
fjfrackiewicz@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |fjfrackiewicz@gmail.com
--- Comment #4 from fjfrackiewicz@gmail.com --- Created attachment 55096 --> https://bugs.winehq.org/attachment.cgi?id=55096 Wine 1.9.14 backtrace
I tried starting the game with "WINEDEBUG=+d3d11 wine witcher3.exe" but the game crashes almost immediately when attempting to run it and generates no useful terminal output.
https://bugs.winehq.org/show_bug.cgi?id=39180
Kimmo Myllyvirta kimmo.myllyvirta@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |kimmo.myllyvirta@gmail.com
--- Comment #5 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- Created attachment 55113 --> https://bugs.winehq.org/attachment.cgi?id=55113 Witcher 3 +d3d11
You need to start it with GoG dll, as Anthony mentioned in comment #2. Same goes for the Steam version - the game must be started via Steam. "wine witcher3.exe" alone will not work.
I don't have clean prefix, I have Steam version, and I'm using the latest wine from wine-git, but; - The game starts up and splash screens show up ok. - The intro video works ok (except that the small "Skip" hint at the lower right corner is messed up) - The main menu looks ok with just minor glitches - When a new game is started, the videos work ok, and after the videos are finished the game hangs. Just a black screen, and I need to kill the game.
https://bugs.winehq.org/show_bug.cgi?id=39180
Jonathan Strander mblackwell1024@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mblackwell1024@gmail.com
--- Comment #6 from Jonathan Strander mblackwell1024@gmail.com --- I have the Steam version and definitely get the crash that Anthony reported initially in the latest Wine version.
Not having a clean prefix doesn't really help narrow things down. ;)
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #7 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- (In reply to Jonathan Strander from comment #6)
I have the Steam version and definitely get the crash that Anthony reported initially in the latest Wine version.
Not having a clean prefix doesn't really help narrow things down. ;)
Can you attach the terminal output? Or even better, the +d3d11 debug log.
I retested with wine 1.9.16. - make a clean 64bit prefix - install Steam - install the game (42GiBs to download, it takes a while) - press Play - The game starts up ok as previously, and hangs at the same spot after a new game is started. Except now when it hangs, the rotating loading -indicator is visible, and some parts of the game (maybe, hard to say since the game is about to fade-in and everything is quite dark).
No winetricks needed, nothing changed from winecfg etc. Ubuntu 15.10, GTX 980 (352.63 driver).
https://bugs.winehq.org/show_bug.cgi?id=39180
Ivan Set s.i.v.892@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |s.i.v.892@gmail.com
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #8 from Jonathan Strander mblackwell1024@gmail.com --- Created attachment 55372 --> https://bugs.winehq.org/attachment.cgi?id=55372 Wine 1.9.16 Crash Log
Since the game never actually starts (just crashes) you don't get anything new from adding +d3d11
This log is with a new prefix that has only Steam and The Witcher 3 installed on it. No fiddling with any settings of any kind in Wine - just installed Steam, downloaded game, and pressed Play.
Ubuntu 15.10, GTX 970 (v367.35)
https://bugs.winehq.org/show_bug.cgi?id=39180
winetest@luukku.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |winetest@luukku.com
--- Comment #9 from winetest@luukku.com --- (In reply to Jonathan Strander from comment #8)
Created attachment 55372 [details] Wine 1.9.16 Crash Log
Since the game never actually starts (just crashes) you don't get anything new from adding +d3d11
This log is with a new prefix that has only Steam and The Witcher 3 installed on it. No fiddling with any settings of any kind in Wine - just installed Steam, downloaded game, and pressed Play.
Ubuntu 15.10, GTX 970 (v367.35)
I have no idea what this means in practice but doesnt sound good.
"fixme:virtual:create_view out of memory in virtual heap for 0x7f94aa6a0000-0x7f9caa6a0000"
Taken from the last log provided.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #10 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- (In reply to Jonathan Strander from comment #8)
Created attachment 55372 [details] Wine 1.9.16 Crash Log
Since the game never actually starts (just crashes) you don't get anything new from adding +d3d11
Could you try the hack from bug #39885 ?
There's no patch, but just open dlls/ntdll/virtual.c, and change this; #define VIRTUAL_HEAP_SIZE (sizeof(void*)*1024*1024) to eg. this; #define VIRTUAL_HEAP_SIZE (sizeof(void*)*1024*1024*2) compile, and try to run the game again.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #11 from Jonathan Strander mblackwell1024@gmail.com --- Unfortunately I don't have a good environment to build Wine with like I did years ago (and don't have much time to get it up and running).
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #12 from winetest@luukku.com --- (In reply to Jonathan Strander from comment #11)
Unfortunately I don't have a good environment to build Wine with like I did years ago (and don't have much time to get it up and running).
I doubt that matters in this case much it seems that the cause has been found already. I would personally test myself but I don't have this specific game in my gamelist (I do have around 200 steam games).
https://bugs.winehq.org/show_bug.cgi?id=39180
Shmerl shtetldik@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |shtetldik@gmail.com
--- Comment #13 from Shmerl shtetldik@gmail.com --- I can confirm, that applying a patch to dlls/ntdll/virtual.c (latest master git) to double the size of the virtual heap, fixes the crashing for the Witcher 3 (I'm testing the GOG version, with all latest patches applied, i.e. up until patch 1.30 (GOG version 2.0.0.51) including.
Now it doesn't crash, but I get this error message (and it quits after pressing OK):
GPU doesn't meet minimal requirements. Support for DirectX 11 is required.
My GPU is Nvidia GTX 680. Should I open a separate bug for this?
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #14 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- (In reply to Shmerl from comment #13)
Now it doesn't crash, but I get this error message (and it quits after pressing OK):
GPU doesn't meet minimal requirements. Support for DirectX 11 is required.
My GPU is Nvidia GTX 680. Should I open a separate bug for this?
I'm assuming you are using nvidia proprietary drivers. +d3d11,+dxgi debug log might help. You should see; trace:dxgi:dxgi_check_feature_level_support Choosing supported feature level D3D_FEATURE_LEVEL_11_0 (SM5).
You could also try setting the MaxVersionGL registry key, see; https://appdb.winehq.org/objectManager.php?sClass=version&iId=33252 and comment; Minimal requirements dialog box by Guillaume Charifi
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #15 from Shmerl shtetldik@gmail.com --- That didn't help. I set those registry entries, and enabled debug flags. Still, same error message.
Log:
err:winediag:wined3d_dll_init Setting maximum allowed wined3d GL version to 4.5. fixme:heap:RtlSetHeapInformation 0x29c0000 0 0x212fd60 4 stub trace:dxgi:CreateDXGIFactory riid {7b7166ec-21c7-44ae-b21a-c9ae321ae369}, factory 0x212f7a0 warn:dxgi:dxgi_factory_create Failed to initialize factory, hr 0x887a0004. fixme:msvcrt:__clean_type_info_names_internal (0x2421878) stub fixme:msvcrt:__clean_type_info_names_internal (0x2335768) stub
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #16 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- Created attachment 55545 --> https://bugs.winehq.org/attachment.cgi?id=55545 Screenshot, GTX680
I tried with latest wine from git, GTX 680 (367.44 driver), Ubuntu 16.04, clean prefix. It starts up without problems, screenshot attached. Are you sure you are using nvidia drivers (not mesa/nouveau), check with glxinfo, you should see; .... OpenGL vendor string: NVIDIA Corporation OpenGL renderer string: GeForce GTX 680/PCIe/SSE2 OpenGL core profile version string: 4.5.0 NVIDIA 367.44 OpenGL core profile shading language version string: 4.50 NVIDIA
if you are, and it still fails, then please attach WINEDEBUG=+d3d11,+dxgi,+d3d debug log, since this doesn't look good; warn:dxgi:dxgi_factory_create Failed to initialize factory, hr 0x887a0004.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #17 from Shmerl shtetldik@gmail.com --- Yes. Current driver: 367.44.
To clarify, it started happening after all latest patches (up until 1.30) are applied. GOG version without patches launches to the same menu screen OK. So there is obviously some change in those latest patches. Check if you are testing the very latest version of TW3.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #18 from Shmerl shtetldik@gmail.com --- log:
err:winediag:wined3d_dll_init Setting maximum allowed wined3d GL version to 4.5. trace:d3d:wined3d_dll_init Use 2048MiB = 0x80000000 bytes for emulated_textureram fixme:heap:RtlSetHeapInformation 0x29c0000 0 0x212fd60 4 stub trace:dxgi:CreateDXGIFactory riid {7b7166ec-21c7-44ae-b21a-c9ae321ae369}, factory 0x212f7a0 trace:d3d:wined3d_init Initializing adapters. trace:d3d:wined3d_adapter_init adapter 0x87060, ordinal 0. warn:d3d:wined3d_init Failed to initialize adapter. warn:d3d:wined3d_create Failed to initialize wined3d object, hr 0x80004005. warn:dxgi:dxgi_factory_create Failed to initialize factory, hr 0x887a0004. fixme:msvcrt:__clean_type_info_names_internal (0x2421878) stub fixme:msvcrt:__clean_type_info_names_internal (0x2335768) stub
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #19 from Henri Verbeet hverbeet@gmail.com --- (In reply to Shmerl from comment #18)
trace:d3d:wined3d_init Initializing adapters. trace:d3d:wined3d_adapter_init adapter 0x87060, ordinal 0. warn:d3d:wined3d_init Failed to initialize adapter.
If that's complete, that means it fails right when trying to retrieve the wgl_driver structure in wined3d_adapter_init(). Otherwise there would be a "Allocated LUID ..." message from wined3d_adapter_init().
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #20 from Shmerl shtetldik@gmail.com --- So it's a problem with application, or with my setup?
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #21 from Henri Verbeet hverbeet@gmail.com --- (In reply to Shmerl from comment #20)
So it's a problem with application, or with my setup?
It looks like an issue with the setup, perhaps the (32-bit) OpenGL driver. Adding +wgl may give more information.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #22 from Shmerl shtetldik@gmail.com --- May be it's a problem with Wine that I built from source. I'll run some more tests to narrow it down.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #23 from Shmerl shtetldik@gmail.com --- You were right, I somehow messed up Wine when building it. I just reinstalled the game (without latest patches to avoid the crash), and used regular Wine (64-bit) and it launched OK, why my custom build one produced same DX11 error.
https://bugs.winehq.org/show_bug.cgi?id=39180
Fabian Maurer dark.shadow4@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dark.shadow4@web.de
https://bugs.winehq.org/show_bug.cgi?id=39180
Behem0th grantipak@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |grantipak@gmail.com
https://bugs.winehq.org/show_bug.cgi?id=39180
Guillaume Charifi guillaume.charifi@sfr.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |guillaume.charifi@sfr.fr
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #24 from Guillaume Charifi guillaume.charifi@sfr.fr --- I have a some patches locally that greatly improves the 2D interface (texts not broken nor rendered at wrong places, ...), and I started to work on stuff for missing 3D features (without much results for now). Please ask if you want to try them, maybe it can help into getting them committed...
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #25 from Shmerl shtetldik@gmail.com --- (In reply to Guillaume Charifi from comment #24)
I have a some patches locally that greatly improves the 2D interface (texts not broken nor rendered at wrong places, ...), and I started to work on stuff for missing 3D features (without much results for now). Please ask if you want to try them, maybe it can help into getting them committed...
Most of the DX11 errors that I see are stubs of functions missing from https://source.winehq.org/git/wine.git/history/HEAD:/dlls/d3d11/device.c
I.e. stuff like:
d3d11_immediate_context_CSSetConstantBuffers d3d11_immediate_context_CSSetShader d3d11_immediate_context_CSSetShaderResources d3d11_immediate_context_CSSetUnorderedAccessViews d3d11_immediate_context_Dispatch d3d11_immediate_context_DSSetConstantBuffers d3d11_immediate_context_HSSetConstantBuffers
Are you going to work on those?
I'm interested in testing this, if you'd post your patches somewhere. I just need to figure how to build Wine properly (probably will use a VM next time).
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #26 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- Created attachment 55736 --> https://bugs.winehq.org/attachment.cgi?id=55736 Deferred rendering hack patch
I'm also interested on those patches. Currently the game is blocked by unimplemented deferred rendering. The game calls stub CreateDeferredContext and tries to use NULL context. It catches the exception and hangs.
Here's a semi-hack patch that should get you in-game, although the game is still far from playable :)
https://bugs.winehq.org/show_bug.cgi?id=39180
wine-devel@hurrikhan.eu changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |wine-devel@hurrikhan.eu
--- Comment #27 from wine-devel@hurrikhan.eu --- Created attachment 55739 --> https://bugs.winehq.org/attachment.cgi?id=55739 compressed log, unpacks to 340MB
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #28 from wine-devel@hurrikhan.eu --- Here above is a log with +d3d11,+dxgi,+d3d ... sent faster than expected. I get the menu, the black screen + progress wheel and then it freezes, with the mouse still moving.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #29 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- Created attachment 55740 --> https://bugs.winehq.org/attachment.cgi?id=55740 Witcher 3 in-game
(In reply to wine-devel from comment #28)
Here above is a log with +d3d11,+dxgi,+d3d ... sent faster than expected. I get the menu, the black screen + progress wheel and then it freezes, with the mouse still moving.
You are hitting the fixme:d3d11:d3d11_device_CreateDeferredContext. You could try the patch I sent earlier, it fixes this freeze. But don't expect too much, see the screenshot.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #30 from wine-devel@hurrikhan.eu --- I applied the patch and indeed : I can see myself moving, hear the "witcher sense" activating, but the rendering is about the same as you sent (I was able to see the bath though). I presume it's because of all the shaders issues in the big log (is it ok to send something so big btw?)
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #31 from Shmerl shtetldik@gmail.com --- (In reply to Kimmo Myllyvirta from comment #29)
You are hitting the fixme:d3d11:d3d11_device_CreateDeferredContext. You could try the patch I sent earlier, it fixes this freeze.
Will your patch be accepted upstream? Or may be in Wine staging?
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #32 from fjfrackiewicz@gmail.com --- (In reply to Shmerl from comment #31)
(In reply to Kimmo Myllyvirta from comment #29)
You are hitting the fixme:d3d11:d3d11_device_CreateDeferredContext. You could try the patch I sent earlier, it fixes this freeze.
Will your patch be accepted upstream? Or may be in Wine staging?
It's being reviewed by Henri Verbeet here:
http://source.winehq.org/patches/ (look for ID 126768 )
So I guess it just needs time before it can be accepted upstream.
https://bugs.winehq.org/show_bug.cgi?id=39180
Matteo Bruni matteo.mystral@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Ever confirmed|0 |1 Status|UNCONFIRMED |NEW
--- Comment #33 from Matteo Bruni matteo.mystral@gmail.com --- (In reply to fjfrackiewicz from comment #32)
It's being reviewed by Henri Verbeet here:
http://source.winehq.org/patches/ (look for ID 126768 )
Well, it's been reviewed: https://www.winehq.org/pipermail/wine-devel/2016-October/114904.html
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #34 from Shmerl shtetldik@gmail.com --- By the way, how can I request the increased virtual heap limit fix to be included in Wine staging? I don't think it was included yet.
https://bugs.winehq.org/show_bug.cgi?id=39180
winetest@luukku.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |41636
https://bugs.winehq.org/show_bug.cgi?id=39180
Alistair Leslie-Hughes leslie_alistair@hotmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |patch
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #35 from Shmerl shtetldik@gmail.com --- Please, add dependency on bug #39885.
https://bugs.winehq.org/show_bug.cgi?id=39180
Józef Kucia joseph.kucia@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Depends on| |39885
https://bugs.winehq.org/show_bug.cgi?id=39180
Simon Bolokanov sbolokanov@abv.bg changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |sbolokanov@abv.bg
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #36 from Simon Bolokanov sbolokanov@abv.bg --- That's how it looks on Wine 2.0-rc2 with memory workaround.
https://www.youtube.com/watch?v=7w8GDpmfNEk
https://bugs.winehq.org/show_bug.cgi?id=39180
Evan Goers megatog615@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |megatog615@gmail.com
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #37 from winetest@luukku.com --- (In reply to Shmerl from comment #34)
By the way, how can I request the increased virtual heap limit fix to be included in Wine staging? I don't think it was included yet.
Staging now has a patch for that.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #38 from Shmerl shtetldik@gmail.com --- Created attachment 56876 --> https://bugs.winehq.org/attachment.cgi?id=56876 Loading screen distroted.
Witcher 3 loading screen is distorted.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #39 from Shmerl shtetldik@gmail.com --- I installed the latest updated version from GOG (1.31), and tried with Wine staging 2.0rc5 which now includes the virtual heap fix.
The game launches, but starting screen looks pretty distorted. See attachement: https://bugs.winehq.org/attachment.cgi?id=56876
I'm using RX 480 with Mesa 13.0.2. Same happens with latest Mesa built from source. Does anyone experience same distortion with Nvidia blob? If it's Mesa specific, I can file a Mesa bug.
Also, Wine log shows the following:
fixme:d3d11:d3d11_immediate_context_OMSetBlendState Per-rendertarget blend not implemented. fixme:d3d11:d3d11_immediate_context_OMSetBlendState Color mask > 3 not implemented. fixme:d3d11:d3d11_immediate_context_HSSetConstantBuffers iface 0xacfa0, start_slot 1, buffer_count 1, buffers 0x5fbadcf8 stub! fixme:d3d11:d3d11_immediate_context_DSSetConstantBuffers iface 0xacfa0, start_slot 1, buffer_count 1, buffers 0x5fbadcf8 stub! fixme:dxgi:dxgi_swapchain_Present Unimplemented sync interval 1
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #40 from Fabian Maurer dark.shadow4@web.de --- Well, could you provide an apitrace? (https://github.com/apitrace/apitrace/blob/master/docs/USAGE.markdown) I could test it on Win10.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #41 from Shmerl shtetldik@gmail.com --- (In reply to Fabian Maurer from comment #40)
Well, could you provide an apitrace? (https://github.com/apitrace/apitrace/blob/master/docs/USAGE.markdown) I could test it on Win10.
How do you want me to trace it, like this?
apitrace trace --api gl ...?
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #42 from Fabian Maurer dark.shadow4@web.de --- Yeah, trace it in opengl mode. "apitrace trace wine XYZ.exe" is usually enough. Just be careful since it gets big very fast, also, if the problem appears in windowed mode, that would be preferable
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #43 from Shmerl shtetldik@gmail.com --- The problem is in fullscreen mode. I managed to produce a trace until the broken menu screen appears, and killed the process soon afterwards, to prevent the massive file size.
The resulted file is still big, so I can't attach it here. I uploaded it to: https://uploadfiles.io/caf8c
Going back to the original question. Can anyone please confirm that this also happens on Nvidia blob or not? Because if not, it can be a Mesa issue.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #44 from Guillaume Charifi guillaume.charifi@sfr.fr --- I'm not having this problem, Wine upstream, Mesa upstream.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #45 from Fabian Maurer dark.shadow4@web.de --- Created attachment 56880 --> https://bugs.winehq.org/attachment.cgi?id=56880 Broken rendering in the apitrace
Can confirm that the apitrace looks weird. Attached a screenshots, probably shouldn't be like that, right? Tested on win10, so it's a wine issue.
https://bugs.winehq.org/show_bug.cgi?id=39180
Filip Volejnik f.volejnik@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |f.volejnik@gmail.com
--- Comment #46 from Filip Volejnik f.volejnik@gmail.com --- I can confirm this issue happens with nvidia drivers as well. Same sort of distorsion on main menu.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #47 from Shmerl shtetldik@gmail.com --- Created attachment 56884 --> https://bugs.winehq.org/attachment.cgi?id=56884 Witcher 3 main menu distortion (no CSMT)
Ah, I think I found part of the problem! With CSMT enabled (wine-staging), a major distortion happens like in my previous screenshot. Without CSMT, it looks much better, only the text of the menu is messed up.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #48 from Shmerl shtetldik@gmail.com --- Some report, that without CSMT with Nvidia blob, starting menu looks OK. So I opened this Mesa bug: https://bugs.freedesktop.org/show_bug.cgi?id=99444
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #49 from Guillaume Charifi guillaume.charifi@sfr.fr --- No, it is *not* a Mesa bug, it's a missing feature in Wine. I implemented it locally and text displays correctly. By the way, I've implemented enough of tessellation to be able to run Unigine Heaven, and I'm now working on compute shaders. :)
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #50 from Shmerl shtetldik@gmail.com --- (In reply to Guillaume Charifi from comment #49)
No, it is *not* a Mesa bug, it's a missing feature in Wine. I implemented it locally and text displays correctly.
Thanks, for clarifying. I can close the Mesa bug since you confirm it's not Mesa specific.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #51 from Guillaume Charifi guillaume.charifi@sfr.fr --- Mh, sorry, I think I read too fast. It is not directly Mesa-related, because it works fine with nouveau, but indeed probably radeonsi-related. However, keep in mind that by using wine-staging, it can also be a side-effect of other patches in the repository. I'll send a trace so that you can check if there are still artifacts (in which case it would be radeonsi's fault).
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #52 from Guillaume Charifi guillaume.charifi@sfr.fr --- Created attachment 56903 --> https://bugs.winehq.org/attachment.cgi?id=56903 The working trace
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #53 from winetest@luukku.com --- This thread has a patch for CreateDeferredContext and it seems to be that Sleeping dogs is currently blocked by that feature. The output was very limited.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #54 from Kimmo Myllyvirta kimmo.myllyvirta@gmail.com --- Yes, it needs it, with the stub implementation the sleeping dogs demo goes "much" further (it may need a lot more than simple stubs, it seems to use deferred contexts extensively).
On-topic; Since Witcher 1.31, the game doesn't call CreateDeferredContext at all anymore, at least not at the very beginning of a new game (Steam version). The main menu looks the same as before (GTX 980, 375.20, staging and vanilla). And yes, CSMT breaks it.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #55 from Shmerl shtetldik@gmail.com --- (In reply to Guillaume Charifi from comment #51)
Mh, sorry, I think I read too fast. It is not directly Mesa-related, because it works fine with nouveau, but indeed probably radeonsi-related. However, keep in mind that by using wine-staging, it can also be a side-effect of other patches in the repository. I'll send a trace so that you can check if there are still artifacts (in which case it would be radeonsi's fault).
I replayed your trace, and I can see distortions in the menu, which looks similar to what I get when I run it in Wine. So I'll reopen Mesa bug then.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #56 from Shmerl shtetldik@gmail.com --- (In reply to Guillaume Charifi from comment #52)
Created attachment 56903 [details] The working trace
Can you please answer the question here: https://bugs.freedesktop.org/show_bug.cgi?id=99444#c5
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #57 from Evan Goers megatog615@gmail.com --- This appears to be fixed on wine 2.0-rc6.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #58 from Shmerl shtetldik@gmail.com --- (In reply to Evan Goers from comment #57)
This appears to be fixed on wine 2.0-rc6.
What do you mean? It's still unplayable because of many missing DX11 features.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #59 from Austin English austinenglish@gmail.com --- (In reply to Shmerl from comment #58)
(In reply to Evan Goers from comment #57)
This appears to be fixed on wine 2.0-rc6.
What do you mean? It's still unplayable because of many missing DX11 features.
The original bug was about a crash. If it's not crashing, but there are other graphical bugs or other problems, those should be separate reports.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #60 from Shmerl shtetldik@gmail.com --- There can be separate reports, but this is a generic enough bug to make it include others. If you want - open another one and make this one depend on it.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #61 from Bruno Jesus 00cpxxx@gmail.com --- (In reply to Shmerl from comment #60)
There can be separate reports, but this is a generic enough bug to make it include others. If you want - open another one and make this one depend on it.
In general we try to work in one problem per bug avoiding what we call "meta bugs". That simplifies bug fixing by reducing the scope of the problem (divide and conquer).
This bug for example has been used for the original crash, for wine build problems, for heap problems and for long discussions into different patches about missing features.
The longer the bug is opened with such generic title the longer it will remain open with growing discussions about new/different issues (which makes it less useful and harder to understand when reading the entire thread). I don't have the game nor the required knowledge to split this bug so I'll leave as is for now.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #62 from Shmerl shtetldik@gmail.com --- (In reply to Bruno Jesus from comment #61)
In general we try to work in one problem per bug avoiding what we call "meta bugs". That simplifies bug fixing by reducing the scope of the problem (divide and conquer).
I agree that it's better to have a separate bug for each focused issue, but it still helps to have a meta bug for that will encompass them all. It simply helps keeping track of things. Mozilla do it for Firefox for instance. I.e. open an encompassing general bug, and then open multiple individual bugs which affect it and mark the main one as dependent on the individual ones.
Do you think it's a sensible approach?
https://bugs.winehq.org/show_bug.cgi?id=39180
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |FIXED Status|NEW |RESOLVED
--- Comment #63 from Austin English austinenglish@gmail.com --- (In reply to Shmerl from comment #62)
(In reply to Bruno Jesus from comment #61)
In general we try to work in one problem per bug avoiding what we call "meta bugs". That simplifies bug fixing by reducing the scope of the problem (divide and conquer).
I agree that it's better to have a separate bug for each focused issue, but it still helps to have a meta bug for that will encompass them all. It simply helps keeping track of things. Mozilla do it for Firefox for instance. I.e. open an encompassing general bug, and then open multiple individual bugs which affect it and mark the main one as dependent on the individual ones.
Do you think it's a sensible approach?
That is not Wine's policy. Of course Mozilla is entitled to their own.
The original crash has been fixed. New issues belong in separate reports.
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #64 from Shmerl shtetldik@gmail.com --- OK, I opened a new bug for the current issue (bug #42292).
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #65 from Simon Bolokanov sbolokanov@abv.bg --- Created attachment 57012 --> https://bugs.winehq.org/attachment.cgi?id=57012 terminal-output-for-crash
Weird, is it just me? It's still crashing.
Wine 2.0-rc6 f1d812d4109cfd501532e0ce809a62e8 setup_the_witcher_3_wild_hunt_2.0.0.52.exe
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #66 from Shmerl shtetldik@gmail.com --- (In reply to Simon Bolokanov from comment #65)
Weird, is it just me? It's still crashing.
Wine 2.0-rc6
It's still crashing with regular Wine. The virtual heap fix which prevents this crash is so far only in Wine staging, as indicated in bug #39885.
https://bugs.winehq.org/show_bug.cgi?id=39180
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #67 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 2.1.
https://bugs.winehq.org/show_bug.cgi?id=39180
Anthony f00dl3a@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|1.7.50 |2.2 Status|CLOSED |REOPENED Resolution|FIXED |--- Hardware|x86 |x86-64
--- Comment #68 from Anthony f00dl3a@gmail.com --- Still exists in Wine Staging v2.2
Can't even launch because it says I don't have a "DX11" graphics card. I applied the d3d.reg patch but no luck. Using a GeForce GTX 960 card which oddly enough was bundled with Witcher 3 so I would hope it has DX11 support!
https://bugs.winehq.org/show_bug.cgi?id=39180
--- Comment #69 from Shmerl shtetldik@gmail.com --- (In reply to Anthony from comment #68)
Can't even launch because it says I don't have a "DX11" graphics card. I applied the d3d.reg patch but no luck. Using a GeForce GTX 960 card which oddly enough was bundled with Witcher 3 so I would hope it has DX11 support!
Wine developers decided to close this bug, because it's not specific enough. For the current list of bugs, see:
* https://bugs.winehq.org/show_bug.cgi?id=42292 * https://bugs.winehq.org/show_bug.cgi?id=42414 * https://bugs.winehq.org/show_bug.cgi?id=42509
https://bugs.winehq.org/show_bug.cgi?id=39180
Bruno Jesus 00cpxxx@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |FIXED Status|REOPENED |RESOLVED
--- Comment #70 from Bruno Jesus 00cpxxx@gmail.com --- Resolving again.
https://bugs.winehq.org/show_bug.cgi?id=39180
Bruno Jesus 00cpxxx@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|2.2 |1.7.50 Status|RESOLVED |CLOSED
--- Comment #71 from Bruno Jesus 00cpxxx@gmail.com --- Closing again. Anthony, for new issues please open new bugs if they are not already reported as Shmerl said in comment 69.
https://bugs.winehq.org/show_bug.cgi?id=39180 Bug 39180 depends on bug 39885, which changed state.
Bug 39885 Summary: Cannot allocate more than 32gb of memory https://bugs.winehq.org/show_bug.cgi?id=39885
What |Removed |Added ---------------------------------------------------------------------------- Status|STAGED |RESOLVED Resolution|--- |FIXED