http://bugs.winehq.org/show_bug.cgi?id=26754
Summary: Regedit crashes on start when compiled with clang from trunk Product: Wine Version: 1.3.17 Platform: x86 OS/Version: Linux Status: NEW Keywords: download Severity: minor Priority: P2 Component: comctl32 AssignedTo: wine-bugs@winehq.org ReportedBy: austinenglish@gmail.com
Created an attachment (id=34075) --> (http://bugs.winehq.org/attachment.cgi?id=34075) backtrace
I hit a clang bug recently, that required upgrading to trunk from 2.9 (or going down to 2.8): http://llvm.org/bugs/show_bug.cgi?id=9676
that's been fixed. Compiling wine from trunk works, but running it crashes in wineboot. Backtrace indicates comctl32, and indeed, compiling just comctl32 with gcc gets regedit working.
Wineboot still shows a crash, but doesn't seem fatal. No backtrace though..
Comctl32/regedit backtrace attached.
http://bugs.winehq.org/show_bug.cgi?id=26754
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |source
--- Comment #1 from Austin English austinenglish@gmail.com 2011-04-12 18:55:36 CDT --- Interestingly, +relay also works around it. I'll attach a +seh,+tid,+imagelist log instead.
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #2 from Austin English austinenglish@gmail.com 2011-04-12 18:55:49 CDT --- Created an attachment (id=34076) --> (http://bugs.winehq.org/attachment.cgi?id=34076) +seh,+tid,+imagelist
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #3 from Austin English austinenglish@gmail.com 2011-04-12 19:00:13 CDT --- $ svn info Path: . URL: http://llvm.org/svn/llvm-project/cfe/trunk Repository Root: http://llvm.org/svn/llvm-project Repository UUID: 91177308-0d34-0410-b5e6-96231b3b80d8 Revision: 129385 Node Kind: directory Schedule: normal Last Changed Author: kremenek Last Changed Rev: 129366 Last Changed Date: 2011-04-12 12:21:33 -0500 (Tue, 12 Apr 2011)
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #4 from Austin English austinenglish@gmail.com 2011-04-12 19:06:20 CDT --- Also, the comctl32 rebar/toolbar tests show the same crash when running make test.
http://bugs.winehq.org/show_bug.cgi?id=26754
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dank@kegel.com
--- Comment #5 from Dan Kegel dank@kegel.com 2011-04-12 19:32:34 CDT --- Does dialing back the optimizer (compiling with -O1) help?
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #6 from Austin English austinenglish@gmail.com 2011-04-12 20:07:49 CDT --- (In reply to comment #5)
Does dialing back the optimizer (compiling with -O1) help?
-O1 fails to build, -O0 still has the bug. Same for bug 26755 (the wineboot/cabinet crash).
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #7 from Nikolay Sivov bunglehead@gmail.com 2011-04-30 01:11:27 CDT --- Does it really fail on ZeroMemory or it just corrupts stack?
ZeroMemory() should be dropped here anyway, cause we already explicitly initialize 11 fields, and only 6 are left to be zeroed with ZeroMemory().
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #8 from Austin English austinenglish@gmail.com 2011-05-01 16:50:22 CDT --- (In reply to comment #7)
Does it really fail on ZeroMemory or it just corrupts stack?
ZeroMemory() should be dropped here anyway, cause we already explicitly initialize 11 fields, and only 6 are left to be zeroed with ZeroMemory().
Well, initially tried valgrinding it, which crashes valgrind terribly (apparently it doesn't like recent clang either). I filed https://bugs.kde.org/show_bug.cgi?id=272189 for that.
Removing the ZeroMemory() call and initialized manually avoids the crash, I'll attach the patch I used.
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #9 from Austin English austinenglish@gmail.com 2011-05-01 16:50:51 CDT --- Created an attachment (id=34423) --> (http://bugs.winehq.org/attachment.cgi?id=34423) initialize manually
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #10 from Nikolay Sivov bunglehead@gmail.com 2011-05-02 00:36:59 CDT --- Patch could be sent anyway, but that doesn't explain a crash. Could be a compiler problem?
http://bugs.winehq.org/show_bug.cgi?id=26754
--- Comment #11 from Austin English austinenglish@gmail.com 2011-05-02 11:24:14 CDT --- (In reply to comment #10)
Patch could be sent anyway, but that doesn't explain a crash. Could be a compiler problem?
http://llvm.org/bugs/show_bug.cgi?id=9830
http://bugs.winehq.org/show_bug.cgi?id=26754
Nikolay Sivov bunglehead@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|comctl32 |-unknown
--- Comment #12 from Nikolay Sivov bunglehead@gmail.com 2011-05-07 01:45:29 CDT --- (In reply to comment #11)
(In reply to comment #10)
Patch could be sent anyway, but that doesn't explain a crash. Could be a compiler problem?
It looks confirmed on llvm.org side, so this is invalid. Let's keep it open until they fix compiler bug.
http://bugs.winehq.org/show_bug.cgi?id=26754
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |4adfb787f4e8c36a37ce1d53a7e | |6df16d03ecd8a Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #13 from Austin English austinenglish@gmail.com 2012-12-17 17:06:12 CST --- Fixed by http://source.winehq.org/git/wine.git/commit/4adfb787f4e8c36a37ce1d53a7e6df1...
http://bugs.winehq.org/show_bug.cgi?id=26754
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #14 from Alexandre Julliard julliard@winehq.org 2012-12-21 13:27:38 CST --- Closing bugs fixed in 1.5.20.
-
wine-bugs＠winehq.org