http://bugs.winehq.org/show_bug.cgi?id=19659
Scott Murray scott@aikendrum.org changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |stefan@codeweavers.com
--- Comment #9 from Scott Murray scott@aikendrum.org 2009-08-24 00:23:22 --- Hi folks - the likely candidate from the regression testing I did seems to be this patch:
commit 014c4bfc70a4d4e60f033d579d1be13a46f65170 Author: Stefan Dösinger stefan@codeweavers.com Date: Thu Apr 9 18:40:57 2009 +0200
wined3d: Save some memory in vertex buffers.
see attached bisect log (regression-eq2-wine1.1.18-1.1.19-09082401448) for further details. The previously attached backtraces are still valid. (thread 09 etc)
I've put ~8 hrs running the bisected wine after excluding this patch, and the seg fault hasn't reoccured. 2 hrs is the longest I've run without seeing the fault during all the other bisects
I've also checked out out wine-1.1.19, reverted this patch (014c4bfc70a4d4e60f033d579d1be13a46f65170) out, compiled and I'v put ~3 hrs on this code to double-check the results of the bisect. (I can't easily revert this patch on later versions of code, as later patches require the added functionality Stefan has built in with this one, so the compile fails).
I use this app quite a bit - happy to assist with any further testing, patch evaluating etc if I can.
Cheers,
Scott Murray