http://bugs.winehq.org/show_bug.cgi?id=35492
Bug ID: 35492 Summary: Bad Direct3d Performance in Starcraft: Brood War Product: Wine Version: 1.7.11 Hardware: x86 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs@winehq.org Reporter: muleyryan@fastmail.fm Classification: Unclassified
For me, with WINEDEBUG=WARN+ALL, I get a screen flooded with
warn:d3d_perf:wined3d_surface_map Mapping a dynamic surface without WINED3D_MAP_DISCARD. warn:d3d_surface:wined3d_surface_blt D3D not initialized, using fallback
I was to perform a full regression, and then I discovered it takes 3+ hours to make wine from source,
Hopefully this is something that can be solved without having to spend 30+ hours idling.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #1 from Henri Verbeet hverbeet@gmail.com --- (In reply to comment #0)
warn:d3d_surface:wined3d_surface_blt D3D not initialized, using fallback
It sounds like you either don't have a working (32-bit) GL setup, or explicitly disabled it with a registry setting.
http://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #2 from Ken Sharp imwellcushtymelike@gmail.com --- (In reply to comment #0)
I was to perform a full regression, and then I discovered it takes 3+ hours to make wine from source,
It takes ten minutes here. Use distcc and ccache. A regression test rarely takes more than an hour on very old regressions.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #3 from Austin English austinenglish@gmail.com --- (In reply to comment #2)
(In reply to comment #0)
I was to perform a full regression, and then I discovered it takes 3+ hours to make wine from source,
It takes ten minutes here. Use distcc and ccache. A regression test rarely takes more than an hour on very old regressions.
Well that greatly depends on your hardware..but that's not really a topic for bugzilla :)
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #4 from Bob Johnson muleyryan@fastmail.fm --- Ok, I think we can close this:
Upon deeper inspection of my Xorg.0.log, I discovered the following:
[ 2864.954] (II) SAVAGE(0): Direct rendering enabled [ 2864.954] (==) RandR enabled [ 2864.984] (II) AIGLX: Screen 0 is not DRI2 capable [ 2864.984] (EE) AIGLX: reverting to software rendering [ 2865.028] (II) AIGLX: Loaded and initialized swrast
So it would seem I'm having an Xorg problem, and it's not Wine.
Henri was onto it. Goede man.
Side note, if anyone knows how I can fix this problem, I'm still trying to fix it.
It takes ten minutes here
Thank you for those kind words, I'll tell that to my cpu.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #5 from Bob Johnson muleyryan@fastmail.fm --- Solved my problem, and yeah, invalid bug.
https://bugs.winehq.org/show_bug.cgi?id=35492
Bruno Jesus 00cpxxx@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution|--- |INVALID
--- Comment #6 from Bruno Jesus 00cpxxx@gmail.com --- (In reply to comment #5)
Solved my problem, and yeah, invalid bug.
Resolving invalid.
https://bugs.winehq.org/show_bug.cgi?id=35492
Bob Johnson muleyryan@fastmail.fm changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |UNCONFIRMED Resolution|INVALID |---
--- Comment #7 from Bob Johnson muleyryan@fastmail.fm --- Reopening. I fixed my DRI setup, and the problem remains.
When running with "WINEDEBUG=warn+all" console is filled with the same message: "warn:d3d_perf:wined3d_surface_blt Converted surface, using CPU blit.
Tried regressing to the commit before 3c023a5db5538a7622d4518e8ce24d36e1c459b1, as per bug 34547, and it made no difference.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #8 from Henri Verbeet hverbeet@gmail.com --- Could you attach a WINEDEBUG="+d3d" log?
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #9 from Bob Johnson muleyryan@fastmail.fm --- There's a great amount of output. What's the simplest way to get a tidy log from the massive output?
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #10 from Bruno Jesus 00cpxxx@gmail.com --- (In reply to comment #9)
There's a great amount of output. What's the simplest way to get a tidy log from the massive output?
Try compressing with bzip2 -9 or rzip -9.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #11 from Bob Johnson muleyryan@fastmail.fm --- Created attachment 47496 --> https://bugs.winehq.org/attachment.cgi?id=47496 'WINEDEBUG=+d3d' log
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #12 from Bruno Jesus 00cpxxx@gmail.com ---
trace:d3d:wined3d_adapter_init_gl_caps GL_VERSION: "1.2 Mesa 7.11.2".
I had several performance issues with mesa versions before 9 in my intel graphics hardware.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #13 from Bob Johnson muleyryan@fastmail.fm --- Thats the version in ARCH's community repo. Let me ask the maintainer what's up,
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #14 from Bob Johnson muleyryan@fastmail.fm --- I think that version of Mesa is required, because I just tried to build 10.0.3 with the --disable-shared-dricore option, and it was unrecognized.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #15 from Bob Johnson muleyryan@fastmail.fm --- "The death sentence to legacy Mesa drivers was carried out over the night. All DRI1 drivers have been removed from the Mesa tree along with support for BeOS and other old code, in an effort to remove drivers that receive little maintenance and are just causing a greater burden in re-factoring core Mesa code for the modern drivers. This means though that the Linux desktop loses its support for hardware like VIA Unichrome and 3dfx Voodoo graphics cards." - 27 August 2011
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #16 from Bruno Jesus 00cpxxx@gmail.com --- You could try the GDI renderer, read more at http://appdb.winehq.org/objectManager.php?sClass=version&iId=149 (topic Cure for slowness)
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #17 from Bob Johnson muleyryan@fastmail.fm --- GDI runs slightly faster, but still too slow to play.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #18 from Henri Verbeet hverbeet@gmail.com --- Yeah, that's probably the last version of Mesa that supports that hardware.
The d3d_perf warning is caused by the lack of support for ARB_fragment_program for converting paletted surfaces to RGBA. Unfortunately I don't know of an easy way to speed this up with the extensions supported by your GPU / driver.
https://bugs.winehq.org/show_bug.cgi?id=35492
--- Comment #19 from Austin English austinenglish@gmail.com --- This is your friendly reminder that there has been no bug activity for over a year. Is this still an issue in current (1.7.37 or newer) wine? If so, please attach the terminal output in 1.7.37 (see http://wiki.winehq.org/FAQ#get_log).
https://bugs.winehq.org/show_bug.cgi?id=35492
Ken Sharp imwellcushtymelike@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |Abandoned?, performance
https://bugs.winehq.org/show_bug.cgi?id=35492
Ken Sharp imwellcushtymelike@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |ABANDONED Status|UNCONFIRMED |RESOLVED
--- Comment #20 from Ken Sharp imwellcushtymelike@gmail.com --- Abandoned.
https://bugs.winehq.org/show_bug.cgi?id=35492
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #21 from Austin English austinenglish@gmail.com --- Closing.
-
wine-bugs＠winehq.org