http://bugs.winehq.org/show_bug.cgi?id=19392
Summary: Runes of magic : crash while entering world Product: Wine Version: 1.1.26 Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: anonymous.xzxz@yahoo.fr
Created an attachment (id=22473) --> (http://bugs.winehq.org/attachment.cgi?id=22473) Console output
I start the game, accept the EULA, select a server, select a character and clic 'enter world'. A screen with a blue progression bar is displayed. Then afetr a few seconds a crash occurs.
Console output attached.
wine version : wine-1.1.26-44-gc2ada64 Clean wineprefix except (else I hit bug #17818) : HKEY_CURRENT_USER/Software/Wine/Direct3D/UseGLSL="disabled"
http://bugs.winehq.org/show_bug.cgi?id=19392
Juan Lang juan_lang@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #22473|application/octet-stream |text/plain mime type| |
http://bugs.winehq.org/show_bug.cgi?id=19392
3vi1 winehq.org@eternaldusk.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |winehq.org@eternaldusk.com
--- Comment #1 from 3vi1 winehq.org@eternaldusk.com 2009-08-09 17:46:47 --- I can confirm this. Crash at same location, and I have matching console output:
err:seh:raise_exception Unhandled exception code c0000005 flags 0 addr 0x34cc2c
http://bugs.winehq.org/show_bug.cgi?id=19392
3vi1 winehq.org@eternaldusk.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |julliard@winehq.org
--- Comment #2 from 3vi1 winehq.org@eternaldusk.com 2009-08-09 21:39:17 --- Since it was working in earlier versions, I did a regression test. My bisect turns this up as the culprit:
75792a3e6994e124a777b8825ef9b2916e7fe333 is first bad commit commit 75792a3e6994e124a777b8825ef9b2916e7fe333 Author: Alexandre Julliard julliard@winehq.org Date: Fri Jul 10 14:46:55 2009 +0200
dbghelp: Grow all arrays by doubling the size.
:040000 040000 a1f845b3b8715beca6f753b124c76dc43820cda8 37d2df93816a6a1ffe831609d96b765a472a4a40 M dlls
Unfortunately, it does not look like this commit can be backed out in the current version:
evil@mars:~/src/wine$ git show 75792a3e6994e124a777b8825ef9b2916e7fe333 | patch -p1 -R patching file dlls/dbghelp/coff.c patching file dlls/dbghelp/dbghelp_private.h patching file dlls/dbghelp/minidump.c Hunk #3 FAILED at 253. Hunk #4 FAILED at 382. Hunk #5 succeeded at 873 (offset 6 lines). 2 out of 5 hunks FAILED -- saving rejects to file dlls/dbghelp/minidump.c.rej patching file dlls/dbghelp/module.c patching file dlls/dbghelp/msc.c patching file dlls/dbghelp/source.c patching file dlls/dbghelp/stabs.c patching file dlls/dbghelp/symbol.c
(compilation of course fails)
http://bugs.winehq.org/show_bug.cgi?id=19392
Vitaliy Margolen vitaliy@kievinfo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1
--- Comment #3 from Vitaliy Margolen vitaliy@kievinfo.com 2009-08-10 00:43:15 --- Confirming per comment #1.
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #4 from 3vi1 winehq.org@eternaldusk.com 2009-08-13 21:01:56 --- Created an attachment (id=23053) --> (http://bugs.winehq.org/attachment.cgi?id=23053) Hack to reduce minidump memory consumption.
It appears like the catalyst for the crash are the changes in dbghelp/minidump.c. The attached quick hack (tested w/1.1.27) reduces its memory consumption and allows RoM to start again.
I won't pretend to fully understand what's happening behind the scenes (yet), but these minidump functions eat a ton of memory when allowed to double.
http://bugs.winehq.org/show_bug.cgi?id=19392
Ken Sharp kennybobs@o2.co.uk changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download, regression URL| |http://www3.firstclass.com/ | |ClientDownloads/IntlClients | |/uk/FC81UKDownloadFiles/FC8 | |326UK.exe CC| |kennybobs@o2.co.uk Target Milestone|--- |1.2.0
--- Comment #5 from Ken Sharp kennybobs@o2.co.uk 2009-08-14 16:03:53 --- Also affects FirstClass client. Workaround is very simple, disable dbghelp in winecfg for this app. Technically (dogfood) I guess...
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #6 from 3vi1 winehq.org@eternaldusk.com 2009-08-14 20:23:38 --- I figured out what's going on and how to fix it properly. I'll upload the patch later tonight. Going to go watch some DragonBallZ with my son first. :)
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #7 from 3vi1 winehq.org@eternaldusk.com 2009-08-14 23:14:43 --- Created an attachment (id=23066) --> (http://bugs.winehq.org/attachment.cgi?id=23066) Proposed fix.
Attached is my proposed fix.
add_module and add_memory_block are doubling their memory allocations every single time they are called. They should only be doubling the allocation when the memory they last allocated is full.
http://bugs.winehq.org/show_bug.cgi?id=19392
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |patch CC| |dank@kegel.com
--- Comment #8 from Dan Kegel dank@kegel.com 2009-08-15 05:06:04 --- Thanks. URL of the submitted patch in the archives is http://www.winehq.org/pipermail/wine-patches/2009-August/077161.html
Next time when you post a patch please also mention the bug number in the patch description.
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #9 from 3vi1 winehq.org@eternaldusk.com 2009-08-15 08:54:43 --- (In reply to comment #8)
Thanks. URL of the submitted patch in the archives is http://www.winehq.org/pipermail/wine-patches/2009-August/077161.html
Next time when you post a patch please also mention the bug number in the patch description.
Will do... but please disregard that patch. :P
I woke up this morning with a weird feeling that I had made an error in the add_memory_block logic... and sure enough, it appears I have.
I'll attach the patch with the correction as soon as I make the change and test.
Sorry for the false start.
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #10 from 3vi1 winehq.org@eternaldusk.com 2009-08-15 09:25:07 --- Created an attachment (id=23071) --> (http://bugs.winehq.org/attachment.cgi?id=23071) Proposed fix, pass #2.
Okay, the bug is fixed and it tests good. Patch attached.
In my previous attempt, add_memory block would have been constantly growing by 32 allocations, and not by doing reallocs. :\ I flipped the check on the if statement to match the way it's done in add_module, so that the logic is easier to follow.
Dan: Do I need to send this to wine-patches to, like I did with my previous attempt, or do you devs usually pick up these smaller bug-fixes directly from here in bugzilla?
Thanks again for your assistance.
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #11 from Ken Sharp kennybobs@o2.co.uk 2009-08-15 09:26:56 --- Thanks for the patch. You will need to submit to wine-patches. Patches aren't picked up from Bugzilla.
http://bugs.winehq.org/show_bug.cgi?id=19392
3vi1 winehq.org@eternaldusk.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #23066|Proposed fix. |Proposed fix (broken - description| |don't use) Attachment #23066|0 |1 is obsolete| |
--- Comment #11 from Ken Sharp kennybobs@o2.co.uk 2009-08-15 09:26:56 --- Thanks for the patch. You will need to submit to wine-patches. Patches aren't picked up from Bugzilla.
http://bugs.winehq.org/show_bug.cgi?id=19392
3vi1 winehq.org@eternaldusk.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #23053|0 |1 is obsolete| |
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #12 from 3vi1 winehq.org@eternaldusk.com 2009-08-15 09:28:53 --- (In reply to comment #11)
Thanks for the patch. You will need to submit to wine-patches. Patches aren't picked up from Bugzilla.
Okay - will do. Thanks Ken!
http://bugs.winehq.org/show_bug.cgi?id=19392
Evil wine@eternaldusk.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |wine@eternaldusk.com
--- Comment #13 from Evil wine@eternaldusk.com 2009-08-16 11:31:28 --- The archive url for the final (crosses fingers) patch is: http://www.winehq.org/pipermail/wine-patches/2009-August/077179.html
http://bugs.winehq.org/show_bug.cgi?id=19392
--- Comment #14 from 3vi1 winehq.org@eternaldusk.com 2009-08-21 12:17:56 --- The status for this can be changed to RESOLVED. The patch is included in today's 1.1.28.
http://bugs.winehq.org/show_bug.cgi?id=19392
Ken Sharp kennybobs@o2.co.uk changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #15 from Ken Sharp kennybobs@o2.co.uk 2009-08-21 13:00:51 --- Fixed.
http://bugs.winehq.org/show_bug.cgi?id=19392
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #16 from Alexandre Julliard julliard@winehq.org 2009-09-02 14:28:48 --- Closing bugs fixed in 1.1.29.
http://bugs.winehq.org/show_bug.cgi?id=19392
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |43a35958f698871ff01573b6103 | |0d80d0097e008 CC| |focht@gmx.net Component|-unknown |dbghelp
-
wine-bugs＠winehq.org