[Bug 35605] New: 3Dmark 2000: Choosing "Run Default Benchmark" says "Benchmark Aborted!"
http://bugs.winehq.org/show_bug.cgi?id=35605
Bug ID: 35605 Summary: 3Dmark 2000: Choosing "Run Default Benchmark" says "Benchmark Aborted!" Product: Wine Version: 1.7.12 Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs@winehq.org Reporter: wylda@volny.cz Classification: Unclassified
Created attachment 47554 --> http://bugs.winehq.org/attachment.cgi?id=47554 Console +relay log from wine-1.7.12-141-gaa22a3c
As the title says, one can't run the "Default Benchark" because immediately message box pops-up saying "Benchmark Aborted!". At the attached relay log:
0024:Call user32.MessageBoxA(00010072,0055b274 "Benchmark Aborted!",0055b074 "3DMark Information",00000040) ret=0046ada3
Custom benchmark runs fine, so no problem with 3D enviroment. In case of re-test, be aware, that on the net, there are broken installers, which crash even on WinXP.
http://bugs.winehq.org/show_bug.cgi?id=35605
Wylda wylda@volny.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download URL| |http://www.computerbase.de/ | |downloads/software/benchmar | |ks/futuremark_3dmark2000/
--- Comment #1 from Wylda wylda@volny.cz ---
Adding correct URL for download.
$ sha1sum -b 3dmark2000v11.exe 7fa48d56464ef3ad03f746dffa851cef276e29ff *3dmark2000v11.exe
http://bugs.winehq.org/show_bug.cgi?id=35605
Wylda wylda@volny.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #47554|application/octet-stream |application/x-bzip2 mime type| |
https://bugs.winehq.org/show_bug.cgi?id=35605
hanska2@luukku.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |hanska2@luukku.com
--- Comment #2 from hanska2@luukku.com --- Interesting. I wanted to try this bug. Starting the demo gives me an error message that 3dmark needs at least 4Mb video memory.
So there is definitely something wrong with wine source code.
https://bugs.winehq.org/show_bug.cgi?id=35605
--- Comment #3 from Wylda wylda@volny.cz --- Still in wine-1.7.27.
https://bugs.winehq.org/show_bug.cgi?id=35605
Teras teras@luukku.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |teras@luukku.com
--- Comment #4 from Teras teras@luukku.com --- Still not working
wine-1.7.37-49-g75a2e85
https://bugs.winehq.org/show_bug.cgi?id=35605
Béla Gyebrószki gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |gyebro69@gmail.com
--- Comment #5 from Béla Gyebrószki gyebro69@gmail.com --- For me the default benchmark is aborting when running the application in the traditional full-screen mode, but it runs properly in virtual desktop mode.
If someone is up for some regression testing: Wine 1.5.15: the benchmark starts properly in full-screen mode Wine 1.5.20: reports "Benchmark aborted!"
Fedora 21 Nvidia binary drivers 340.76 XOrg 1.16.3 XFCE 4.11
3dmark2000v11.exe md5sum: eb025480a0ee5665bddcbda716c67f73
https://bugs.winehq.org/show_bug.cgi?id=35605
Wylda wylda@volny.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression CC| |julliard@winehq.org Regression SHA1| |a0f039a6049641cc2be8fa32f2a | |fa3c3f394a99b
--- Comment #6 from Wylda wylda@volny.cz --- Many thanks for great tip, GyB. It was broken in version wine-1.5.17-270-ga0f039a:
commit a0f039a6049641cc2be8fa32f2afa3c3f394a99b Author: Alexandre Julliard julliard@winehq.org Date: Thu Nov 22 11:57:15 2012 +0100
user32: Fix the wait event mask in PeekMessage.
https://bugs.winehq.org/show_bug.cgi?id=35605
Wylda wylda@volny.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|-unknown |user32 Version|1.7.12 |1.5.18
https://bugs.winehq.org/show_bug.cgi?id=35605
--- Comment #7 from Wylda wylda@volny.cz ---
commit a0f039a6049641cc2be8fa32f2afa3c3f394a99b
FWIW this commit should fix bug 32264.
https://bugs.winehq.org/show_bug.cgi?id=35605
--- Comment #8 from Alexandre Julliard julliard@winehq.org --- It looks like it's getting confused by an asynchronous take focus event.
https://bugs.winehq.org/show_bug.cgi?id=35605
super_man@post.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |focht@gmx.net, | |super_man@post.com
--- Comment #9 from super_man@post.com --- Works for me 1.8.rc4. Anyone else willing to confirm?
https://bugs.winehq.org/show_bug.cgi?id=35605
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |FIXED
--- Comment #10 from Anastasius Focht focht@gmx.net --- Hello folks,
running 'Default Benchmark' works for me too, resolving 'FIXED'.
It would be nice if people with some spare time could add the commit that fixed this by using reverse regression test or just educated guess + cherry pick ;-)
You might get results faster by limiting the bisect process to 'user32' since the regression was caused by a change to this module ('git bisect start -- dlls/user32').
Generally speaking this kind of janitorial task is useful and doesn't require a lot of knowledge.
--- snip --- -=[ ProtectionID v0.6.6.7 DECEMBER]=- (c) 2003-2015 CDKiLLER & TippeX Build 24/12/14-22:48:13 Ready... Scanning -> Z:\home\focht\Downloads\3dmark2000v11.exe File Type : 32-Bit Exe (Subsystem : Win GUI / 2), Size : 20007579 (01314A9Bh) Byte(s) Compilation TimeStamp : 0x38DFA3F6 -> Mon 27th Mar 2000 18:09:58 (GMT) [TimeStamp] 0x38DFA3F6 -> Mon 27th Mar 2000 18:09:58 (GMT) | PE Header | - | Offset: 0x000000F0 | VA: 0x004000F0 | - -> File has 19843739 (012ECA9Bh) bytes of appended data starting at offset 028000h [File Heuristics] -> Flag #1 : 00000000000000000000000000000100 (0x00000004) [Entrypoint Section Entropy] : 6.59 (section #0) ".text " | Size : 0x105CA (67018) byte(s) [DllCharacteristics] -> Flag : (0x0000) -> NONE [SectionCount] 4 (0x4) | ImageSize 0x2A000 (172032) byte(s) [VersionInfo] Company Name : MadOnion.com [VersionInfo] Product Name : 3DMark2000 [VersionInfo] Product Version : Build 340 [VersionInfo] File Version : Build 340 [VersionInfo] Original FileName : stub32i.exe [VersionInfo] Internal Name : stub32i.exe [VersionInfo] Legal Copyrights : Copyright (c) 1998-2000 MadOnion.com [-= Installer =-] InstallShield v5.53.168.0 Setup Module [CompilerDetect] -> Visual C++ 6.0 - Scan Took : 0.172 Second(s) [0000000ACh (172) tick(s)] [558 of 573 scan(s) done]
Scanning -> C:\Program Files\MadOnion.com\3DMark2000\3DMark2000.exe File Type : 32-Bit Exe (Subsystem : Win GUI / 2), Size : 1552384 (017B000h) Byte(s) Compilation TimeStamp : 0x394DCAF4 -> Mon 19th Jun 2000 07:25:40 (GMT) [TimeStamp] 0x394DCAF4 -> Mon 19th Jun 2000 07:25:40 (GMT) | PE Header | - | Offset: 0x00000100 | VA: 0x00400100 | - [File Heuristics] -> Flag #1 : 00000000000000000000000000000000 (0x00000000) [Entrypoint Section Entropy] : 6.54 (section #0) ".text " | Size : 0x11EA8B (1174155) byte(s) [DllCharacteristics] -> Flag : (0x0000) -> NONE [SectionCount] 4 (0x4) | ImageSize 0x183000 (1585152) byte(s) [VersionInfo] Company Name : 6FileDescription [VersionInfo] Product Name : 3DMark Application [VersionInfo] Product Version : 1. 0. 0. 1 [VersionInfo] File Description : 3DMark [VersionInfo] File Version : 1. 0. 0. 1 [VersionInfo] Original FileName : 3DMark2000.exe [VersionInfo] Internal Name : 3DMark [VersionInfo] Legal Trademarks : FOriginalFilename [VersionInfo] Legal Copyrights : Copyright (C) 1998-1999 MadOnion.com [CdKeySerial] found "Registration Key" @ VA: 0x001551D8 / Offset: 0x001551D8 [CompilerDetect] -> Visual C++ 6.0 [!] File appears to have no protection or is using an unknown protection - Scan Took : 0.461 Second(s) [0000001CDh (461) tick(s)] [499 of 573 scan(s) done] --- snip ---
$ sha1sum 3dmark2000v11.exe 7fa48d56464ef3ad03f746dffa851cef276e29ff 3dmark2000v11.exe
$ du -sh 3dmark2000v11.exe 20M 3dmark2000v11.exe
$ wine --version wine-1.8-rc4
Regards
https://bugs.winehq.org/show_bug.cgi?id=35605
Wylda wylda@volny.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |bfb845f9fccb2ff5bff0b0ba323 | |8fec7f9f7b710
--- Comment #11 from Wylda wylda@volny.cz --- (In reply to Anastasius Focht from comment #10)
It would be nice if people with some spare time could add the commit that fixed this by using reverse regression test...
You're welcome. Fixed by wine-1.7.54-152-gbfb845f:
commit bfb845f9fccb2ff5bff0b0ba3238fec7f9f7b710 Author: Sebastian Lackner sebastian@fds-team.de Date: Thu Nov 5 15:57:29 2015 +0100
dinput: Ensure X11 input events are forwarded to wineserver even without message loop.
This commit basically reverts b22ff8018aca7c365e505f1db7732f7050ae259b.
Windows does not do that, but our current implementation of winex11 requires periodic event polling to forward events to the wineserver.
You might get results faster by limiting the bisect process to 'user32' since the regression was caused by a change to this module ('git bisect start -- dlls/user32').
I guess, that this wouldn't work this time ;)
Thanks Sebastian for fixing this bug!
https://bugs.winehq.org/show_bug.cgi?id=35605
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #12 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 1.9.0.
-
wine-bugs＠winehq.org