https://bugs.winehq.org/show_bug.cgi?id=45946
Bug ID: 45946 Summary: Regression in commit 580ea44bc65472c0304d74b7e873acfb7f680b85 Product: Wine Version: 3.17 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: directx-d3d Assignee: wine-bugs@winehq.org Reporter: zerooneo@01101001.net Distribution: ---
Created attachment 62487 --> https://bugs.winehq.org/attachment.cgi?id=62487 A screenshot showing several objects temporarily missing
A last-minute commit was included in release 3.17, numbered 580ea44bc65472c0304d74b7e873acfb7f680b85, and titled "wined3d: Use query buffer objects for occlusion queries."
This commit seems to negatively affect the game "The Witness". When I play the game with Wine compiled from this commit or a later one and go out into an open area, the framerate drops considerably, and landscape features and buildings that are at a distance start very rapidly flashing in and out of existence. Areas that are close to the player or indoors are unaffected.
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #1 from David Oftedal zerooneo@01101001.net --- Created attachment 62488 --> https://bugs.winehq.org/attachment.cgi?id=62488 A screenshot showing differences from the other screenshot.
https://bugs.winehq.org/show_bug.cgi?id=45946
David Oftedal zerooneo@01101001.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Regression SHA1| |580ea44bc65472c0304d74b7e87 | |3acfb7f680b85
https://bugs.winehq.org/show_bug.cgi?id=45946
David Oftedal zerooneo@01101001.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |awesie@gmail.com, | |zerooneo@01101001.net
https://bugs.winehq.org/show_bug.cgi?id=45946
David Oftedal zerooneo@01101001.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|Regression in commit |Regression in commit |580ea44bc65472c0304d74b7e87 |580ea44bc65472c0304d74b7e87 |3acfb7f680b85 |3acfb7f680b85: Landscape in | |"The Witness" flashes | |wildly
https://bugs.winehq.org/show_bug.cgi?id=45946
Gijs Vermeulen gijsvrm@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression
--- Comment #2 from Gijs Vermeulen gijsvrm@gmail.com --- Bug 45932 has the same regression sha1, can you try the patch attached there?
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #3 from Andrew Wesie awesie@gmail.com --- Those patches are probably ineffective. What gpu and driver are you using?
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #4 from David Oftedal zerooneo@01101001.net --- The model of the card according to the manufacturer spec is a Radeon 7670M: Type : AMD Radeon™ HD 7670M Graphics (2GB dedicated memory)
Various driver info:
Glxinfo: Device: AMD TURKS (DRM 2.50.0 / 4.18.0-8-generic, LLVM 7.0.0) (0x6840) OpenGL renderer string: AMD TURKS (DRM 2.50.0 / 4.18.0-8-generic, LLVM 7.0.0)
lsmod: radeon 1462272 52
syslog Oct 7 16:41:27 lightning kernel: [ 2.784659] [drm] Initialized radeon 2.50.0 20080528 for 0000:01:00.0 on minor 0
Xorg.0.log: [ 30.590] (II) RADEON: Driver for ATI/AMD Radeon chipsets: ATI Radeon X1950, ATI Radeon X1900, (etc. etc. etc.), [ 30.738] (II) RADEON(0): glamor X acceleration enabled on AMD TURKS (DRM 2.50.0 / 4.18.0-8-generic, LLVM 7.0.0)
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #5 from David Oftedal zerooneo@01101001.net --- As Andrew Wesie suggested, the patch posted at Bug 45932 does not affect this particular case.
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #6 from Andrew Wesie awesie@gmail.com --- Thanks for the information. Could you attach the output of glxinfo as a text file to this bug?
https://bugs.winehq.org/show_bug.cgi?id=45946
zzzzzyzz@hacari.org changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |zzzzzyzz@hacari.org
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #7 from David Oftedal zerooneo@01101001.net --- Created attachment 62503 --> https://bugs.winehq.org/attachment.cgi?id=62503 Output of glxinfo
Sure thing!
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #8 from Andrew Wesie awesie@gmail.com --- I was able to reproduce with an AMD HD 5750 that I had around. I spent the better part of the day trying to figure out the root cause to no avail.
I am convinced that it is a bug in the gallium r600 driver since calling glGetQueryObjectui64v multiple times in a row with the same arguments puts different values into the query buffer object. Also the gallium r600 driver does not pass the piglit qbo conherency tests, for what it is worth.
I'll probably poke at it some more since it is going to annoy me.
tl;dr probably a different issue than bug 45932, should not affect radeonsi users.
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #9 from Andrew Wesie awesie@gmail.com --- I can happily report that I found the bug in Mesa. I will file a bug report upstream at some point, but honestly right now I never want to see the Mesa source code again in my life.
The bug only affects r600 driver + query buffer objects.
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #10 from David Oftedal zerooneo@01101001.net --- Oh, wow! That makes at least two Mesa bugs uncovered with Wine and this game. Let's hope Mesa will work perfectly from here on out. Thanks for researching it so thoroughly.
https://bugs.winehq.org/show_bug.cgi?id=45946
Matteo Bruni matteo.mystral@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- See Also| |https://bugs.freedesktop.or | |g/show_bug.cgi?id=108311 Resolution|--- |NOTOURBUG Status|UNCONFIRMED |RESOLVED Summary|Regression in commit |Landscape in "The Witness" |580ea44bc65472c0304d74b7e87 |flashes wildly on r600g |3acfb7f680b85: Landscape in | |"The Witness" flashes | |wildly |
--- Comment #11 from Matteo Bruni matteo.mystral@gmail.com --- (In reply to Andrew Wesie from comment #9)
I can happily report that I found the bug in Mesa. I will file a bug report upstream at some point, but honestly right now I never want to see the Mesa source code again in my life.
Is it worse than those anti-cheating systems? :D
The bug only affects r600 driver + query buffer objects.
Resolving as NOTOURBUG, thank you for your work. I added the Mesa bug to the "See Also" section.
https://bugs.winehq.org/show_bug.cgi?id=45946
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #12 from Austin English austinenglish@gmail.com --- Closing.
https://bugs.winehq.org/show_bug.cgi?id=45946
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|CLOSED |RESOLVED
--- Comment #13 from Austin English austinenglish@gmail.com --- This was inadvertently caught up in my unclosed bugs filter. NOTOURBUG should only be closed when fixed upstream.
Setting back to RESOLVED NOTOURBUG.
Sorry for the spam.
https://bugs.winehq.org/show_bug.cgi?id=45946
--- Comment #14 from Andrew Wesie awesie@gmail.com --- Fix has been committed upstream (mesa) with 2ddd44d941648d49dc0d917e03a579baec3590d9.
https://bugs.winehq.org/show_bug.cgi?id=45946
Matteo Bruni matteo.mystral@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #15 from Matteo Bruni matteo.mystral@gmail.com --- (In reply to Andrew Wesie from comment #14)
Fix has been committed upstream (mesa) with 2ddd44d941648d49dc0d917e03a579baec3590d9.
Thank you, closing fixed upstream bug.
-
wine-bugs＠winehq.org