https://bugs.winehq.org/show_bug.cgi?id=41052
Bug ID: 41052 Summary: Failure to create D3D9 device on i915 Product: Wine Version: 1.9.14 Hardware: x86 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: directx-d3d Assignee: wine-bugs@winehq.org Reporter: b7.10110111@gmail.com Distribution: ---
Created attachment 55224 --> https://bugs.winehq.org/attachment.cgi?id=55224 Test case
I'm testing the attached source (compilable with wineg++ or mingw32-g++) on Intel Atom N550, which is i915-class GPU. I get the following output on it:
$ wine --version; ./d3d9-init-test.exe wine-1.9.14 fixme:win:EnumDisplayDevicesW ((null),0,0x32f858,0x00000000), stub! Error: Failed to create Direct3D device: invalid call
wine-1.5.17-7-g67886fed, despite multiple fixme and err prints, does at least create the device (i.e. the test prints "Init OK").
wine-1.7.33-84-gfecbc88 already fails.
https://bugs.winehq.org/show_bug.cgi?id=41052
Ruslan Kabatsayev b7.10110111@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression
https://bugs.winehq.org/show_bug.cgi?id=41052
Józef Kucia joseph.kucia@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |joseph.kucia@gmail.com
--- Comment #1 from Józef Kucia joseph.kucia@gmail.com --- Could you please attach "WINEDEBUG=+d3d" log?
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #2 from Ruslan Kabatsayev b7.10110111@gmail.com --- Created attachment 55225 --> https://bugs.winehq.org/attachment.cgi?id=55225 WINEDEBUG=+d3d output
Here it is.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #3 from Matteo Bruni matteo.mystral@gmail.com --- It looks like i915 doesn't support BGRA FBO attachments :/
Setting the "OffscreenRenderingMode" registry key to "backbuffer" might be a workaround, although that's going to break other stuff. FWIW I think "recent" Wine now fails just because it checks for renderability of the swapchain surfaces and Wine ~1.5.17 would fail to render properly later on anyway.
I have some wined3d changes planned (for OpenGL ES, which also has very limited BGRA formats support) which might help with i915 too. Don't hold your breath though...
You might also try with a newer Mesa. I think there have been a few i915 patches somewhat recently, although I don't recall if they had anything to do with this particular issue.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #4 from Matteo Bruni matteo.mystral@gmail.com --- ...except that BGRA FBO attachments seem to work just fine in the trace from bug 41055. Looking at the diff between the traces I think the driver might not support BGRA sRGB attachments specifically.
Does commenting out the "GL_EXT_texture_sRGB_decode" entry in gl_extension_map[] (wined3d/directx.c) make any difference? Otherwise it should be possible to disable the GL extension via the MESA_EXTENSION_OVERRIDE environment variable (so something like MESA_EXTENSION_OVERRIDE="-GL_EXT_texture_sRGB_decode" wine ...), that should have the same effect as the code change I suggested.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #5 from Bruno Jesus 00cpxxx@gmail.com --- Can you please attach the .exe or .exe.so compiled?
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #6 from Ruslan Kabatsayev b7.10110111@gmail.com --- Created attachment 55233 --> https://bugs.winehq.org/attachment.cgi?id=55233 Test case compiled with wineg++
Here it is.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #7 from Ruslan Kabatsayev b7.10110111@gmail.com --- (In reply to Matteo Bruni from comment #4)
Does commenting out the "GL_EXT_texture_sRGB_decode" entry in gl_extension_map[] (wined3d/directx.c) make any difference? Otherwise it should be possible to disable the GL extension via the MESA_EXTENSION_OVERRIDE environment variable (so something like MESA_EXTENSION_OVERRIDE="-GL_EXT_texture_sRGB_decode" wine ...), that should have the same effect as the code change I suggested.
Using MESA_EXTENSION_OVERRIDE does result in "Init OK" while without it I get the error.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #8 from Matteo Bruni matteo.mystral@gmail.com --- Created attachment 55330 --> https://bugs.winehq.org/attachment.cgi?id=55330 quirk patch
Can you try this patch and see if it helps?
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #9 from Ruslan Kabatsayev b7.10110111@gmail.com --- Yes, it does work.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #10 from Matteo Bruni matteo.mystral@gmail.com --- Created attachment 55929 --> https://bugs.winehq.org/attachment.cgi?id=55929 Alternative patch
Can you please test this new patch instead? I think I like it better than the previous one.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #11 from Ruslan Kabatsayev b7.10110111@gmail.com --- This one works too.
https://bugs.winehq.org/show_bug.cgi?id=41052
--- Comment #12 from Ruslan Kabatsayev b7.10110111@gmail.com --- Still present in 2.0-rc2-15-g21dd046.
https://bugs.winehq.org/show_bug.cgi?id=41052
André H. nerv@dawncrow.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |nerv@dawncrow.de
--- Comment #13 from André H. nerv@dawncrow.de --- Any updates on this Matteo?
https://bugs.winehq.org/show_bug.cgi?id=41052
Matteo Bruni matteo.mystral@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |FIXED Fixed by SHA1| |9fc3444d6ed8ad0e124a77d2faf | |1e2cc6639f40f Status|UNCONFIRMED |RESOLVED
--- Comment #14 from Matteo Bruni matteo.mystral@gmail.com --- This should be fixed by 9fc3444d6ed8ad0e124a77d2faf1e2cc6639f40f and 69006a7cc6c43256501f3d6c0a370ced2b2c53fa (which essentially are a split-up version of the latest attached patch.)
https://bugs.winehq.org/show_bug.cgi?id=41052
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #15 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 2.1.
https://bugs.winehq.org/show_bug.cgi?id=41052
Matteo Bruni matteo.mystral@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |shatteredlites@gmail.com
--- Comment #16 from Matteo Bruni matteo.mystral@gmail.com --- *** Bug 34871 has been marked as a duplicate of this bug. ***
https://bugs.winehq.org/show_bug.cgi?id=41052
Michael Stefaniuc mstefani@redhat.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Target Milestone|--- |2.0.x
https://bugs.winehq.org/show_bug.cgi?id=41052
Michael Stefaniuc mstefani@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Target Milestone|2.0.x |---
--- Comment #17 from Michael Stefaniuc mstefani@winehq.org --- Remove 2.0.x milestone.
-
wine-bugs＠winehq.org