http://bugs.winehq.org/show_bug.cgi?id=29464
Bug #: 29464 Summary: Diablo II Unhandled exception: Access_violation (c0000005) Product: Wine Version: 1.3.29 Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: vityokster@gmail.com Classification: Unclassified
Created attachment 38144 --> http://bugs.winehq.org/attachment.cgi?id=38144 Bzipped logfile
I try to run Diablo II Lord Of Destruction 1.13d with wine on ~x86 Gentoo Linux.
When run Diablo in console I see:
fixme:advapi:SetSecurityInfo stub fixme:d3d_caps:wined3d_guess_card No card selector available for GL vendor 0x4 and card vendor 0000 (using GL_RENDERER "Gallium 0.4 on i915 (chipset: 945GM)"). fixme:win:EnumDisplayDevicesW ((null),0,0x33e9b4,0x00000000), stub! err:ntdll:RtlDeleteResource Deleting active MRSW lock (0x1126ec), expect failure err:ntdll:RtlDeleteResource Deleting active MRSW lock (0x1121cc), expect failure
On wine version 1.3.28 and earlier Diablo runs without this error.
I try recognize which commit breaks wine using git bisect starting from 1.3.28 - good and 1.3.29 - bad.
After some tests git says:
b9fcf71c714e8587def75742f94759d7eca84ad2 is the first bad commit commit b9fcf71c714e8587def75742f94759d7eca84ad2 Author: Henri Verbeet hverbeet@codeweavers.com Date: Thu Sep 15 20:01:51 2011 +0200
ddraw: Use real flips.
:040000 040000 dde17c19243dd91a3830083879701b13c2a4476f 81d5b536d9f76c42bcdac7b3b3fb41814ebc6216 M dlls
I make log-file by using this command:
WINEDEBUG=+reg wine .wine/dosdevices/c:/Program\ Files/Diablo\ II/Diablo\ II.exe 2>DiabloIILOD_20111228_202000.log
Archived log in attachment.
http://bugs.winehq.org/show_bug.cgi?id=29464
Wylda wylda@volny.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |wylda@volny.cz
--- Comment #1 from Wylda wylda@volny.cz 2011-12-28 18:13:38 CST ---
Can't confirm, works perfectly here. I played D2 LoD even on Battle.net under: * wine-1.3.29, wine-1.3.35-230-g277361d * Debian Squeeze 32bit * nVidia GT240 v275.09.07
Please: 1. Test under 1.3.35 or latter wine version 2. Try to revert the commit to confirm yours regression test result 3. The way you run the game and debug channel look suspicious, so:
cd ~/.wine/drive_c/Program\ Files/Diablo\ II wine Diablo\ II.exe &> plain_console_log.txt
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #2 from Viktor Yu. Kovalskiy vityokster@gmail.com 2011-12-29 08:26:49 CST --- Created attachment 38161 --> http://bugs.winehq.org/attachment.cgi?id=38161 Log of running wine-1.3.35
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #3 from Viktor Yu. Kovalskiy vityokster@gmail.com 2011-12-29 08:27:52 CST --- I test wine-1.3.35 and current git wine commit:277361d7be206e1ff7634998f0d750f7c6d66498 Diablo II LOD crashes with pop-up: Unhandled exception: Access_violation (c0000005)
Also I test:
revision 1aeee59bd8b6172c848c8029e1a55ddac1f93c44 (commit before problem commit)
Diablo II.exe running.
revision b9fcf71c714e8587def75742f94759d7eca84ad2 (problem commit)
Diablo II.exe crashes.
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #4 from Viktor Yu. Kovalskiy vityokster@gmail.com 2011-12-29 08:28:35 CST --- Created attachment 38162 --> http://bugs.winehq.org/attachment.cgi?id=38162 Log of running wine (problem commit)
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #5 from Viktor Yu. Kovalskiy vityokster@gmail.com 2011-12-29 08:34:36 CST --- Created attachment 38163 --> http://bugs.winehq.org/attachment.cgi?id=38163 Log of running wine (before problem commit)
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #6 from Viktor Yu. Kovalskiy vityokster@gmail.com 2011-12-29 08:35:24 CST --- Created attachment 38164 --> http://bugs.winehq.org/attachment.cgi?id=38164 Log of running wine (current git version)
http://bugs.winehq.org/show_bug.cgi?id=29464
Wylda wylda@volny.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression CC| |hverbeet@gmail.com Component|-unknown |directx-d3d Regression SHA1| |b9fcf71c714e8587def75742f94 | |759d7eca84ad2
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #7 from Henri Verbeet hverbeet@gmail.com 2011-12-30 01:21:03 CST --- Does this happen with the demo for you? From the log, you seem to be using the Gallium i915 driver. Is that on purpose? As far as I'm aware that driver is somewhat experimental, the regular driver supported by Intel is the "classic" i915 driver. Does it work any better with the software (LIBGL_ALWAYS_SOFTWARE=1) renderer? Also, note that the exception address 0x6f9b8ac8 points to d2sound, which doesn't look like it would be related to 3d rendering.
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #8 from Viktor Yu. Kovalskiy vityokster@gmail.com 2011-12-30 06:05:03 CST --- After switching from Gallium to Classic Diablo II works fine.
I suppose If Gallium is used wine can't recognize video card.
For example, in log:
No card selector available for GL vendor 4 and card vendor 0000
And this situation starts from b9fcf71c714e8587def75742f94759d7eca84ad2 commit.
http://bugs.winehq.org/show_bug.cgi?id=29464
--- Comment #9 from Henri Verbeet hverbeet@gmail.com 2011-12-31 01:22:02 CST --- Created attachment 38196 --> http://bugs.winehq.org/attachment.cgi?id=38196 Detect i915g
(In reply to comment #8)
After switching from Gallium to Classic Diablo II works fine.
I suppose If Gallium is used wine can't recognize video card.
For example, in log:
No card selector available for GL vendor 4 and card vendor 0000
And this situation starts from b9fcf71c714e8587def75742f94759d7eca84ad2 commit.
Well no, http://bugs.winehq.org/attachment.cgi?id=38163 has that line as well. The attached patch is probably enough to recognize your card with the Gallium driver, but I don't think it will make the crash go away.
The main difference that commit makes is that it causes framebuffer blits to the frontbuffer to happen where it previously didn't. Perhaps those are special enough that it makes a difference. If you're interested in that kind of thing, the piglit fbo-sys-blit test may be a quick test you can try. Either way, unless the attached patch makes a difference, and given that it works with the classic driver, this is probably a bug in the driver. Please file a bug at freedesktop.org against the i915g driver instead.
http://bugs.winehq.org/show_bug.cgi?id=29464
Henri Verbeet hverbeet@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |INVALID
--- Comment #10 from Henri Verbeet hverbeet@gmail.com 2012-01-05 05:54:19 CST --- Resolving INVALID. For what it's worth, most Intel cards should now be recognized since 1a3aabf6f2122bcfadb8e87178b3f2e204337da9 and 51d54a909c43bfbd0d8509b27865e6f89a227994.
http://bugs.winehq.org/show_bug.cgi?id=29464
Henri Verbeet hverbeet@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #11 from Henri Verbeet hverbeet@gmail.com 2012-01-20 09:09:39 CST --- Closing.
-
wine-bugs＠winehq.org