http://bugs.winehq.org/show_bug.cgi?id=34603
Bug #: 34603 Summary: Trainsimulator 2014 crash at start/fails to start Product: Wine Version: 1.7.2 Platform: x86-64 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: salamanderrake@gmail.com Classification: Unclassified
Created attachment 46088 --> http://bugs.winehq.org/attachment.cgi?id=46088 steam supplied overrides via winetricks install
Train simulator fails to start/crashes. New steam install in a new overlay. First with overrides provided by steams install via winetricks, then with d3dx9 override. Also have output from WINEDEBUG="-all,d3d". Also debug output with d3d11 disabled in winecfg.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #1 from salamander purake salamanderrake@gmail.com 2013-09-26 19:18:18 CDT --- Created attachment 46089 --> http://bugs.winehq.org/attachment.cgi?id=46089 output with d3dx9 override from winetricks
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #2 from salamander purake salamanderrake@gmail.com 2013-09-26 19:20:41 CDT --- Created attachment 46090 --> http://bugs.winehq.org/attachment.cgi?id=46090 output with WINEDEBUG="-all,d3d"
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #3 from salamander purake salamanderrake@gmail.com 2013-09-26 19:21:39 CDT --- Created attachment 46091 --> http://bugs.winehq.org/attachment.cgi?id=46091 output with WINEDEBUG="-all,d3d" and d3d11 disabled in winecfg (archive)
http://bugs.winehq.org/show_bug.cgi?id=34603
joaopa jeremielapuree@yahoo.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jeremielapuree@yahoo.fr
--- Comment #4 from joaopa jeremielapuree@yahoo.fr 2013-09-26 23:18:51 CDT ---
err:x11settings:X11DRV_ChangeDisplaySettingsEx No matching mode found 1920x1200x32 @75! (XRandR 1.2)
At first, you need to fix that
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #5 from salamander purake salamanderrake@gmail.com 2013-09-27 07:46:57 CDT --- (In reply to comment #4)
err:x11settings:X11DRV_ChangeDisplaySettingsEx No matching mode found 1920x1200x32 @75! (XRandR 1.2)
At first, you need to fix that
I don't know how that is since my monitor is set to 1920x1200x32 @60.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #6 from salamander purake salamanderrake@gmail.com 2013-09-27 14:05:46 CDT --- I found the location to the config file, config.xml and config_fullscreen.xml. I had to go to the line <mGfxRefreshRate>75</mGfxRefreshRate> and change it to <mGfxRefreshRate>60</mGfxRefreshRate>. Still crashes but its a start.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #7 from salamander purake salamanderrake@gmail.com 2013-09-27 14:06:44 CDT --- Created attachment 46113 --> http://bugs.winehq.org/attachment.cgi?id=46113 output from terminal running steam/railworks. with the refresh rate fixed.
http://bugs.winehq.org/show_bug.cgi?id=34603
salamander purake salamanderrake@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #46089|0 |1 is obsolete| |
--- Comment #8 from salamander purake salamanderrake@gmail.com 2013-09-27 14:08:47 CDT --- Comment on attachment 46089 --> http://bugs.winehq.org/attachment.cgi?id=46089 output with d3dx9 override from winetricks
Obsoleted by "output from terminal running steam/railworks. with the refresh rate fixed."
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #9 from salamander purake salamanderrake@gmail.com 2013-09-27 15:36:52 CDT --- After upgrading to wine 1.7.3 I reinstalled everything and the overrides from steam. I also added these overrides to the list d3dx9, physx, dotnet40, and vcrun2010, and disabled d3d11 in winetricks. Got one dialog with a back trace rebuilt wine with debug symblos and got a message about .net missing file or executalbe and no more crash dialogs or backtraces. Please bump up to version 1.7.3. It has the same error message so I wouldn't need a new post because the wine version changed.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #10 from salamander purake salamanderrake@gmail.com 2013-09-27 15:37:21 CDT --- Created attachment 46115 --> http://bugs.winehq.org/attachment.cgi?id=46115 one and only back trace.
http://bugs.winehq.org/show_bug.cgi?id=34603
salamander purake salamanderrake@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|1.7.2 |1.7.3
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #11 from Austin English austinenglish@gmail.com 2013-09-27 20:34:43 CDT --- fixme:winsock:convert_aiflag_w2u Unhandled windows AI_xxx flags 8 ... Backtrace: =>0 0x7bc5232d in ntdll (+0x4232d) (0x003375a8) 1 0x7bc5251c in ntdll (+0x4251b) (0x00337638) 2 0x7bc5328a RtlFreeHeap+0xc9() in ntdll (0x003376a8) 3 0xf698d762 FreeAddrInfoW+0x41() in ws2_32 (0x003376e8)
could you install debug symbols and get another backtrace please? That winsock fixme looks a bit suspicious, a +winsock,+tid trace may help
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #12 from salamander purake salamanderrake@gmail.com 2013-09-27 21:00:17 CDT --- I can do +winsock,+tid but I can't get a backtrace anymore for some reason.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #13 from salamander purake salamanderrake@gmail.com 2013-09-27 21:04:19 CDT --- Created attachment 46118 --> http://bugs.winehq.org/attachment.cgi?id=46118 (archive) +winsock,+tid output
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #14 from Bruno Jesus 00cpxxx@gmail.com 2013-09-28 00:37:23 CDT --- Created attachment 46120 --> http://bugs.winehq.org/attachment.cgi?id=46120 add AI_NUMERICSERV to getaddrinfo
The application is freeing memory from an invalid call to getaddrinfo. Please try the attached patch, if it works the problem is certainly in winsock.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #15 from salamander purake salamanderrake@gmail.com 2013-09-28 02:12:10 CDT --- Created attachment 46121 --> http://bugs.winehq.org/attachment.cgi?id=46121 (archive) +winsock,+tid post patch output
Still fails, also note it failed to compile on my system for 64bit, not major just wanted to give you a heads up. Might be my system or setup, will check better in the morning.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #16 from Hans Leidekker hans@meelstraat.net 2013-09-28 05:12:40 CDT --- getaddrinfo/GetAddrInfoW should probably set res to NULL on all error paths.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #17 from Bruno Jesus 00cpxxx@gmail.com 2013-09-28 10:18:45 CDT --- The AI_xxx missing line is no longer present so I thought the function should have succeeded.
Is the backtrace the same as before?
It may sound silly but if the crash is the same as before you could comment the code from WS_freeaddrinfo and FreeAddrInfoW just to check if it gets any further.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #18 from salamander purake salamanderrake@gmail.com 2013-09-28 10:28:57 CDT --- (In reply to comment #17)
The AI_xxx missing line is no longer present so I thought the function should have succeeded.
Is the backtrace the same as before?
It may sound silly but if the crash is the same as before you could comment the code from WS_freeaddrinfo and FreeAddrInfoW just to check if it gets any further.
You mean everything involving them?
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #19 from salamander purake salamanderrake@gmail.com 2013-09-28 10:33:21 CDT --- Also I still can't get a backtrace. It just does not show up anymore. Not crash dialog either.
http://bugs.winehq.org/show_bug.cgi?id=34603
Bruno Jesus 00cpxxx@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #46120|0 |1 is obsolete| |
--- Comment #20 from Bruno Jesus 00cpxxx@gmail.com 2013-09-28 10:40:38 CDT --- Created attachment 46128 --> http://bugs.winehq.org/attachment.cgi?id=46128 add AI_NUMERICSERV and remove freeaddrinfo
Something like this patch.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #21 from salamander purake salamanderrake@gmail.com 2013-09-28 10:49:43 CDT --- Created attachment 46129 --> http://bugs.winehq.org/attachment.cgi?id=46129 (archive) with the patch to disalbe *freeaddrinfo*
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #22 from salamander purake salamanderrake@gmail.com 2013-09-28 10:50:30 CDT --- Still fails to load, look at attachment 46129 ^
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #23 from Hans Leidekker hans@meelstraat.net 2013-09-28 10:51:24 CDT --- Created attachment 46130 --> http://bugs.winehq.org/attachment.cgi?id=46130 patch
Does this fix it?
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #24 from salamander purake salamanderrake@gmail.com 2013-09-28 10:58:03 CDT --- (In reply to comment #22)
Still fails to load, look at attachment 46129 [details] ^
This was made with my patch( for the freeaddrinfo) and your previous patch, did not see your new patch or Hans Leidekker's patch, doing Hans Leidekker's patch first then yours.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #25 from salamander purake salamanderrake@gmail.com 2013-09-28 11:00:26 CDT --- I do get a rundll32.exe error on failure to inilize some .net thing, would that cause the issues I am having.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #26 from salamander purake salamanderrake@gmail.com 2013-09-28 11:01:44 CDT --- Created attachment 46131 --> http://bugs.winehq.org/attachment.cgi?id=46131 (archive) Hans Leidekker patch fails also
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #27 from salamander purake salamanderrake@gmail.com 2013-09-28 11:11:43 CDT --- Created attachment 46132 --> http://bugs.winehq.org/attachment.cgi?id=46132 (archive) with add AI_NUMERICSERV and remove freeaddrinfo patch
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #28 from Bruno Jesus 00cpxxx@gmail.com 2013-09-28 14:08:09 CDT --- I guess the remaining option is to get a +relay log to find the last call before the crash (without any patches).
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #29 from Hans Leidekker hans@meelstraat.net 2013-09-28 15:49:40 CDT --- (In reply to comment #28)
I guess the remaining option is to get a +relay log to find the last call before the crash (without any patches).
Yes, maybe add +seh,+tid in case it's a multi-threaded app.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #30 from salamander purake salamanderrake@gmail.com 2013-09-28 16:49:48 CDT --- Had to store the file on dropbox.
https://dl.dropboxusercontent.com/u/71062164/gentoo/relay-tid-seh.txt.tar.bz...
the text its self is 2.9 gigs unpacked.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #31 from Bruno Jesus 00cpxxx@gmail.com 2013-09-29 23:09:20 CDT --- Created attachment 46144 --> http://bugs.winehq.org/attachment.cgi?id=46144 relay extraction
This is what I could get from the relay log, the function GetAddrInfoW does not return. From what I could understand when it's converting the results to return to the caller the last memory allocation fails for unknown reasons.
It looks like wine was compiled with optimizations, that removed some calls from the log so part of my research is guessing.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #32 from salamander purake salamanderrake@gmail.com 2013-09-30 01:58:11 CDT --- I re-emerged wine without the custom-cflags use flag and recreated the debug out put, this time remembering to put in +tid and +seh along with +relay, the one drawback is now the unpacked archive is 4.5 gigs.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #33 from Hans Leidekker hans@meelstraat.net 2013-09-30 02:47:36 CDT --- Smells like heap corruption. Can you get a warn+heap,+winsock,+seh,+tid trace with 46120 and 46130 applied?
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #34 from salamander purake salamanderrake@gmail.com 2013-09-30 08:18:30 CDT --- Sorry the link took longer to get ready this time. https://dl.dropboxusercontent.com/u/71062164/gentoo/relay-tid-seh_no-custom-...
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #35 from salamander purake salamanderrake@gmail.com 2013-09-30 08:19:11 CDT --- Ill get the warn+heap,+winsock,+seh,+tid in a bit
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #36 from salamander purake salamanderrake@gmail.com 2013-09-30 08:28:46 CDT --- (In reply to comment #33)
Smells like heap corruption. Can you get a warn+heap,+winsock,+seh,+tid trace with 46120 and 46130 applied?
You will not believe this but with warn+heap,+winsock,+seh,+tid the game actually starts. Did not play it, I didn't want to corrupt the debug out put. Will be posting this debug output it a sec, will also try to run with war+heap only.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #37 from salamander purake salamanderrake@gmail.com 2013-09-30 08:30:00 CDT --- Created attachment 46152 --> http://bugs.winehq.org/attachment.cgi?id=46152 (archive) +winsock,warn+heap,+tid,+seh output with running game.
Here is the output of the game starting with those debug options.
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #38 from salamander purake salamanderrake@gmail.com 2013-09-30 08:32:21 CDT --- Forgot to mention in case its not clear, can not run without warn+heap
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #39 from salamander purake salamanderrake@gmail.com 2013-09-30 08:48:44 CDT --- Created attachment 46153 --> http://bugs.winehq.org/attachment.cgi?id=46153 this is run with -all,warn+heap only
I actually played it, it woks as good as expected minus some text bug. It did crash once but I started it up again and started playing, it crashed when I pushed esc trying to bypass the intro video. It only give the warn heap message once when steam, not the game, starts up.
http://bugs.winehq.org/show_bug.cgi?id=34603
spaaam42+wine@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |spaaam42+wine@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=34603
--- Comment #40 from salamander purake salamanderrake@gmail.com 2013-10-05 19:07:46 CDT --- I want to make this clearer, when I said I played it I mean I was able to play it with these debug symbols enabled. -all,warn+heap. (In reply to comment #39)
Created attachment 46153 [details] this is run with -all,warn+heap only
I actually played it, it woks as good as expected minus some text bug. It did crash once but I started it up again and started playing, it crashed when I pushed esc trying to bypass the intro video. It only give the warn heap message once when steam, not the game, starts up.
http://bugs.winehq.org/show_bug.cgi?id=34603
Doug Rintoul courtjestr@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |courtjestr@gmail.com
https://bugs.winehq.org/show_bug.cgi?id=34603
roger@mailinator.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |roger@mailinator.com
--- Comment #41 from roger@mailinator.com --- Any update on this?
https://bugs.winehq.org/show_bug.cgi?id=34603
salamander purake salamanderrake@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
--- Comment #42 from salamander purake salamanderrake@gmail.com --- Due to the version permanently changing from 2014 to 2015 and in wine 1.7.35 it runs perfectly I am marking it as resolved/fixed. It should probably be marked invalid but I don't want to give people the wrong impression, and sorry for leaving it in this state for this long.
https://bugs.winehq.org/show_bug.cgi?id=34603
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #43 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 1.7.36.
-
wine-bugs＠winehq.org