https://bugs.winehq.org/show_bug.cgi?id=35900
Bug ID: 35900 Summary: L.A. Noire 1.3 (Steam version) crashes during upload of WINED3DFMT_DXT1 volume texture Product: Wine Version: 1.7.15 Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: directx-d3d Assignee: wine-bugs@winehq.org Reporter: focht@gmx.net
Hello folks,
with bug 35112 fixed, it runs a bit farther - only to crash again.
Prerequisite: 'winetricks -q dotnet35sp1'
--- snip --- $ pwd /home/focht/wine-apps/wineprefix-steam/wineprefix/drive_c/Program Files/Steam
$ WINEDEBUG=+tid,+seh,+loaddll,+process,+msvcrt,+d3d,+d3d9,+d3d_texture,+d3d_surface wine ./steam.exe -applaunch 110800 >>~/Downloads/log.txt 2>&1 ... 0041:trace:d3d:device_update_volume device 0x31b52f00, src_volume 0x34dc7ea0, dst_volume 0x34dc7c78. 0041:trace:d3d_surface:wined3d_volume_map volume 0x34dc7ea0, map_desc 0x3328c4, box (nil), flags 0x10. 0041:trace:d3d_surface:wined3d_volume_map Base memory pointer 0x34e1c8e0. 0041:trace:d3d_surface:wined3d_volume_get_pitch Returning row pitch 128, slice pitch 2048. 0041:trace:d3d_surface:wined3d_volume_map No box supplied - all is ok 0041:trace:d3d_surface:wined3d_volume_map Returning memory 0x34e1c8e0, row pitch 128, slice pitch 2048. 0041:trace:d3d:context_acquire device 0x31b52f00, target (nil). 0041:trace:d3d:context_enter Entering context 0x31b804f8, level 1. 0041:trace:d3d_texture:wined3d_texture_bind texture 0x34dc7b10, context 0x31b804f8, srgb 0. 0041:trace:d3d:context_bind_texture glBindTexture call ok /home/focht/projects/wine/wine.repo/src/dlls/wined3d/context.c / 2065 0041:trace:d3d_surface:wined3d_volume_load_location Volume 0x34dc7c78, loading WINED3D_LOCATION_TEXTURE_RGB, have WINED3D_LOCATION_TEXTURE_RGB. 0041:trace:d3d_surface:wined3d_volume_load_location Location(s) already up to date. 0041:trace:d3d_surface:wined3d_volume_upload_data volume 0x34dc7c78, context 0x31b804f8, level 0, format WINED3DFMT_DXT1 (0x31545844). 0041:trace:seh:raise_exception code=c0000005 flags=0 addr=0xf5b6cde9 ip=f5b6cde9 tid=0041 0041:trace:seh:raise_exception info[0]=00000000 0041:trace:seh:raise_exception info[1]=34e40000 0041:trace:seh:raise_exception eax=00000000 ebx=00000000 ecx=00000004 edx=00000000 esi=34e40000 edi=f18a1728 0041:trace:seh:raise_exception ebp=00000038 esp=0033231c cs=0023 ds=002b es=002b fs=0063 gs=006b flags=00010202 0041:trace:seh:call_stack_handlers calling handler at 0x1048b64 code=c0000005 flags=0 0041:trace:seh:call_stack_handlers handler at 0x1048b64 returned 1 0041:trace:seh:call_stack_handlers calling handler at 0x1060f8d code=c0000005 flags=0 0041:trace:seh:call_stack_handlers handler at 0x1060f8d returned 1 0041:trace:seh:call_stack_handlers calling handler at 0x1068980 code=c0000005 flags=0 0041:trace:seh:call_stack_handlers handler at 0x1068980 returned 1 0041:trace:seh:call_stack_handlers calling handler at 0x10521c5 code=c0000005 flags=0 0041:trace:seh:call_stack_handlers handler at 0x10521c5 returned 1 0041:trace:seh:call_stack_handlers calling handler at 0xedc215 code=c0000005 flags=0 --- snip ---
This is with the Nvidia binary driver:
$ nvidia-smi
+------------------------------------------------------+ | NVIDIA-SMI 4.304... Driver Version: 304.119 | |-------------------------------+----------------------+----------------------+ | GPU Name | Bus-Id Disp. | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | |===============================+======================+======================| | 0 GeForce GT 425M | 0000:01:00.0 N/A | N/A | | N/A 61C N/A N/A / N/A | 13% 128MB / 1023MB | N/A Default | +-------------------------------+----------------------+----------------------+
Interestingly, when I disable DXTC support (commenting out the "GL_EXT_texture_compression_s3tc" from gl_extension_map[] in dlls/wined3d/directx.c -> http://source.winehq.org/git/wine.git/blob/62f7ba2555c475a65ad6bf143780b24fb...) it runs much farther: several splash/intros can be seen until it crashes again (probably due to disabling of DXTC support), after lots of:
--- snip --- ... 0042:trace:d3d9:d3d9_device_CreateCubeTexture iface 0xfa0a540, edge_length 128, levels 8, usage 0, format 0x31545844, pool 0, texture 0x332d84, shared_handle (nil). 0042:trace:d3d_texture:wined3d_texture_create device 0xfd12f00, desc 0x332c68, level_count 8, surface_flags 0, parent 0x3524e438, parent_ops 0xf5f77e78, texture 0x3524e448. 0042:fixme:d3d:debug_d3dusage Unrecognized usage flag(s) 0x10000000 0042:trace:d3d_texture:wined3d_texture_init texture 0x3524e488, texture_ops 0xf5eab0b0, layer_count 6, level_count 8, resource_type WINED3D_RTYPE_CUBE_TEXTURE, format WINED3DFMT_DXT1, multisample_type 0, multisample_quality 0, usage 0, pool WINED3D_POOL_DEFAULT, width 128, height 128, depth 1, device 0xfd12f00, parent 0x3524e438, parent_ops 0xf5f77e78, resource_ops 0xf5eab0bc. 0042:warn:d3d_texture:wined3d_texture_init Failed to initialize resource, returning 0x8876086c 0042:warn:d3d_texture:cubetexture_init Failed to initialize texture, returning 0x8876086c 0042:warn:d3d_texture:wined3d_texture_create Failed to initialize texture, returning 0x8876086c. 0042:warn:d3d9:cubetexture_init Failed to create wined3d cube texture, hr 0x8876086c. 0042:warn:d3d9:d3d9_device_CreateCubeTexture Failed to initialize cube texture, hr 0x8876086c. 0042:trace:seh:raise_exception code=c0000005 flags=0 addr=0xd66b7e ip=00d66b7e tid=0042 0042:trace:seh:raise_exception info[0]=00000001 0042:trace:seh:raise_exception info[1]=00000014 0042:trace:seh:raise_exception eax=00000000 ebx=00000008 ecx=10998bf0 edx=00110064 esi=10998b70 edi=31545844 0042:trace:seh:raise_exception ebp=1169bd90 esp=00332db8 cs=0023 ds=002b es=002b fs=0063 gs=006b flags=00010246 0042:trace:seh:call_stack_handlers calling handler at 0x1048bb6 code=c0000005 flags=0 --- snip ---
There is also bug 29128 - IMHO messed up with various issues, unusable for problem analysis.
$ wine --version wine-1.7.15-133-g62f7ba2
Regards
https://bugs.winehq.org/show_bug.cgi?id=35900
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- URL| |http://store.steampowered.c | |om/app/110800/
https://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #1 from Anastasius Focht focht@gmx.net --- Created attachment 47936 --> https://bugs.winehq.org/attachment.cgi?id=47936 WINEDEBUG=+tid,+seh,+loaddll,+process,+msvcrt,+d3d,+d3d9,+d3d_texture,+d3d_surface
Hello folks,
attached is full trace log, bzip2 compressed
Uncompressed: 3.7 MB
Regards
https://bugs.winehq.org/show_bug.cgi?id=35900
Henri Verbeet hverbeet@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |stefan@codeweavers.com
--- Comment #2 from Henri Verbeet hverbeet@gmail.com --- We don't support volume DXT textures, and IIRC according to Stefan neither does NVIDIA on Windows. The interesting question is probably why the application thinks this is supported. We may e.g. simply need to fail volume texture creation with DXT formats.
https://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #3 from Stefan Dösinger stefan@codeweavers.com --- If I remember correctly GeForce8+ GPUs support it, but GeForce 7 and earlier do not. d3d9-class radeon GPUs have support for dxtn volume textures.
There's an equivalent GL extension called GL_NV_texture_compression_vtc, but it uses a different block layout than d3d, so using it isn't trivial.
https://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #4 from Stefan Dösinger stefan@codeweavers.com --- I just double-checked, my GeForce GTX 460 supports DXT1 to DXT5 on volume textures. Still it would be interesting to see how the game responds if we just fail the CreateVolumeTexture if the format is a DXTn format call.
http://bugs.winehq.org/show_bug.cgi?id=35900
Igor Zinovyev zinigor+winehq@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |zinigor+winehq@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=35900
Jarkko K jarkko_korpi@hotmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jarkko_korpi@hotmail.com
--- Comment #5 from Jarkko K jarkko_korpi@hotmail.com --- So what's next step?
https://support.rockstargames.com/hc/en-us/articles/200149466-L-A-Noire-PC-S...
According to that
also
Microsoft Visual C++ 2008 SP1 Redistributable Package (x86)
is needed.
I quess the game has
http://www.gameshield.com/Product/Ironwrap
https://bugs.winehq.org/show_bug.cgi?id=35900
Rui Caridade rcaridade145@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |rcaridade145@gmail.com
--- Comment #6 from Rui Caridade rcaridade145@gmail.com --- (In reply to Stefan Dösinger from comment #4)
I just double-checked, my GeForce GTX 460 supports DXT1 to DXT5 on volume textures. Still it would be interesting to see how the game responds if we just fail the CreateVolumeTexture if the format is a DXTn format call.
I was trying to found information on that and i came across
http://www.ogre3d.org/forums/viewtopic.php?f=4&t=53723#p365595
Hope it helps
https://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #7 from Jarkko K jarkko_korpi@hotmail.com --- Hello Jarkko,
I take this rather offline ...
Although I appreciate any user contributing to Wine by triaging bugs it's sometimes 'overdone'.
This happend with some bugs now so I feel the need to put some things into perspective. I find it quite annoying that you interfere with bug discussion (especially moderated by me) by:
* adding redundant information (state obvious things again, already contained in previous comments) * doing just 'guesswork' which isn't backed hard facts/analysis * make it look like there is some 'important' information missing (no, it's not) * just 'ping' various bugs but didn't take the time to download and retest the actual thing by yourself => trying to moderate/push bugs
To be quite frank: I don't like this at all for my own bugs or bugs by others that _I_ moderate - and I moderate/resolve _many_ Wine bugs.
If a Wine developer or myself is on a bug there is _no_ need to further comment/join in unless explicitly asked for ("please retest xxx"). In that case there are _already_ the best guys one can get for working on Wine bugs.
So please refrain from commenting in bugs created by me or where I put analysis (comments) on. I don't bear a grude against you - but just do it the proper way - for everyone's benefit.
Thanks.
Regards,
A. Focht
https://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #8 from Anastasius Focht focht@gmx.net --- Hello Jarkko,
the email was intended to stay private. Bugzilla shouldn't be abused for personal things. Thanks.
Regards
http://bugs.winehq.org/show_bug.cgi?id=35900
Berillions berillions@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |berillions@gmail.com
--- Comment #10 from Berillions berillions@gmail.com --- Hi Anastasius,
When you said this : "Interestingly, when I disable DXTC support it runs much farther: several splash/intros can be seen until it crashes again (probably due to disabling of DXTC support) etc...
Which splash intro dis you see ?
I installed LA Noire into a fresh wineprefix + winetricks dotnet35.
My Nvidia driver binary :
+------------------------------------------------------+ | NVIDIA-SMI 334.21 Driver Version: 334.21 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | |===============================+======================+======================| | 0 GeForce GTX 770 Off | 0000:01:00.0 N/A | N/A | | 34% 26C N/A N/A / N/A | 84MiB / 2047MiB | N/A Default | +-------------------------------+----------------------+----------------------+
I had a crash about a missing dll "socialclub.dll" at the first launch. Install "Rockstar Game Social Club" setup from Rockstar website resolved this issue.
Now, when i launch the game from Steam, i can see the splash screen (attached at my message) and the game crash because of :
--- snip --- fixme:advapi:CreateProcessAsUserW 0x3cc L"C:\Program Files\Rockstar Games\Social Club\renderer.exe" L""C:\Program Files\Rockstar Games\Social Club\renderer.exe" --type=renderer --disable-logging --disable-desktop-notifications --disable-speech-input --enable-accelerated-compositing --lang=en-US --channel=88.0E727300.1072086138 /prefetch:3" (nil) (nil) 0 0x0100040c (nil) (null) 0xf3ee0fc 0xf3ee0e4 - semi-stub fixme:shell:URL_ParseUrl failed to parse L"mscorlib.resources" err:module:attach_process_dlls "KERNEL32.dll" failed to initialize, aborting err:module:LdrInitializeThunk Main exe initialization for L"C:\Program Files\Rockstar Games\Social Club\renderer.exe" failed, status c0000005 --- snip ---
Cheers, Max
http://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #11 from Berillions berillions@gmail.com --- Created attachment 47947 --> http://bugs.winehq.org/attachment.cgi?id=47947 Splash Screen crash
LA Noire splash screen crash with error log in the output console
http://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #12 from Anastasius Focht focht@gmx.net --- Hello Berillions,
--- quote --- Which splash intro dis you see ? --- quote ---
your screenshot shows the first splash.
There is another series of splash screens and short animated sequences following to 'cover up' for the time spent to upload textures, initializing shaders and the like.
The crash in 'renderer.exe' process entry point seen in your snippet is a secondary fault, caused by main executable unable to reach a certain point due to volume texture upload problem.
Basically the 'channel' parameter denotes an inter-process communication mechanism which is used to exchange data between launcher, main and renderer process.
Regards
http://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #13 from Berillions berillions@gmail.com --- We are at the same point actully. As Stefan said, we must to test the game if we fail the CreateVolumeTexture if the format is a DXTn format call. But how do this ?
http://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #14 from Anastasius Focht focht@gmx.net --- Created attachment 47953 --> http://bugs.winehq.org/attachment.cgi?id=47953 fail volume texture creation with DXTn formats
Hello folks,
I created a small patch which exactly does what Henri suggested.
--- quote --- We may e.g. simply need to fail volume texture creation with DXT formats. --- quote ---
and it fixes the issue - the game runs farther and is now at 'Neon sign' screen :)
Unfortunately it's stuck there with the infamous 'Synchronizing' message (another bug). You can find many problem reports from Windows 7/8 users:
"LA Noire synchronizing message stuck on Windows 7/8" "Stuck at Synchronizing Screen" "Synchronizing FOREVER :: L.A. Noire" "Freezing on Neon Sign/Title Screen during synchronization"
I already updated Social Club garbage to v1.1.0.6
The Game displays 'v1.3 (Build:2617) - DX9' at the bottom left screen.
Anyway, it's definitely progress and maybe something Henry/Stefan could properly integrate.
Regards
http://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #15 from Berillions berillions@gmail.com --- Hi AF,
I tried your patch. I pass the "LA Noire" splash screen but after it, i have a crash with a LA Noire window which appears told me : "AD9J: Impossible to create the d3d resources. Please reboot your computer."
I don't know which screen do you have. Possible to link it or attach it ? The same than me ?
Thanks
PS : I install dotnet35 with winetricks before to install Steam/LA Noire. But when i launch the game, it wants to install dotnet35. Possible to disable this installation ?
http://bugs.winehq.org/show_bug.cgi?id=35900
--- Comment #16 from Anastasius Focht focht@gmx.net --- Hello Berillions
--- quote --- "AD9J: Impossible to create the d3d resources. Please reboot your computer." --- quote ---
it seems I still had '-disablecrashhandler' in the launcher options which just ignores errors and continues (see http://bugs.winehq.org/show_bug.cgi?id=35112#c0 for explanation).
I run the game with: '-windowed -ignorefocus'
Currently, only post-mortem analysis is possible but I think it's linked to failing volume texture creation call.
There is a reference to 'graphicsdata/dirt.dds' string in callstack. The caller code seems to check a D3D call for 0x8876017C (DDER_OUTOFVIDEOMEMORY) result and throws code 'AD9Jb' = 'Unable to create d3d9 texture resource - out of memory'. After that it tests for D3D_OK - all other cases (fall-through) are handled with code 'AD9J' = 'Failed to create d3d resource. Please restart your computer'.
Windows users also reported this error code in various forums.
There is some official support entry which has rather generic problem solutions here ('get better hardware' and 'reinstall Windows' are obviously missing *g*):
https://support.rockstargames.com/hc/en-us/articles/200149816--Failed-to-cre...
So it seems just failing volume texture creation is not the way to go since using '-disablecrashhandler' is only a workaround and might cause real issues later.
Regards
https://bugs.winehq.org/show_bug.cgi?id=35900
Christopher Larson kergoth@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |kergoth@gmail.com
https://bugs.winehq.org/show_bug.cgi?id=35900
sacrediou vmerlet@april.org changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |vmerlet@april.org
https://bugs.winehq.org/show_bug.cgi?id=35900
Stefan Dösinger stefandoesinger@gmx.at changed:
What |Removed |Added ---------------------------------------------------------------------------- CC|stefan@codeweavers.com |stefandoesinger@gmx.at
https://bugs.winehq.org/show_bug.cgi?id=35900
Sergey Isakov isakov-sl@bk.ru changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |isakov-sl@bk.ru
--- Comment #17 from Sergey Isakov isakov-sl@bk.ru --- Volume DXTn textures are disabled during resolving the bug 38178. Dupe?
https://bugs.winehq.org/show_bug.cgi?id=35900
joaopa jeremielapuree@yahoo.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jeremielapuree@yahoo.fr
--- Comment #18 from joaopa jeremielapuree@yahoo.fr --- What about this bug in current wine(3.20)?
https://bugs.winehq.org/show_bug.cgi?id=35900
Matteo Bruni matteo.mystral@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |DUPLICATE
--- Comment #19 from Matteo Bruni matteo.mystral@gmail.com --- (In reply to joaopa from comment #18)
What about this bug in current wine(3.20)?
It should be fixed, actually retesting this would be nice though.
*** This bug has been marked as a duplicate of bug 39253 ***
https://bugs.winehq.org/show_bug.cgi?id=35900
Alistair Leslie-Hughes leslie_alistair@hotmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #20 from Alistair Leslie-Hughes leslie_alistair@hotmail.com --- Closing Duplicate.