https://bugs.winehq.org/show_bug.cgi?id=52045
Bug ID: 52045 Summary: The Witcher 3: crashes when launching Product: Wine-staging Version: 6.21 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs@winehq.org Reporter: gireham910@erpipo.com CC: leslie_alistair@hotmail.com, z.figura12@gmail.com Distribution: ---
Created attachment 71056 --> https://bugs.winehq.org/attachment.cgi?id=71056 The output from the wine app, after the crash
Hello,
I noticed that on Fedora 35, with the latest wine-staging package (6.21) "The Witcher 3" fails to launch (I have attached the output that wine itself generated and asked me to upload, as well as the output from the terminal when I manually tried to launch the app using "wine witcher3.exe").
The program used to work on wine-staging 6.18, and stopped working immediately after updating to 6.21 (on Fedora there were no 6.19 or 6.20 packages). I don't have other windows apps installed.
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #1 from gireham910@erpipo.com --- Created attachment 71057 --> https://bugs.winehq.org/attachment.cgi?id=71057 This is the output from the terminal
https://bugs.winehq.org/show_bug.cgi?id=52045
Louis Lenders xerox.xerox2000x@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|The Witcher 3: crashes when |The Witcher 3: crashes when |launching |launching (regression) CC| |xerox.xerox2000x@gmail.com Keywords| |regression
--- Comment #2 from Louis Lenders xerox.xerox2000x@gmail.com --- Hi,
Any chance you could do a regressiontest? That would be the quickest way to pinpoint where the bug is I think
(It appears to be crashing inside dbghelp.dll. Maybe worth to test again with upcoming wine-6.22 as there seems quite some work going in for this particular dll)
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #3 from gireham910@erpipo.com --- Hi,
Sadly, from what I've read on the wiki, I don't think I have the skills to do this.
https://bugs.winehq.org/show_bug.cgi?id=52045
sporif@posteo.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |sporif@posteo.net
--- Comment #4 from sporif@posteo.net --- The bug is still present on latest wine-staging-git.
A bisect points to [b10ac7d8](https://github.com/wine-staging/wine-staging/commit/b10ac7d8acd88850d743f854...) as the cause.
It's also not reproducible on plain wine.
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #5 from sporif@posteo.net --- Ok, so the first offending wine commit is [690c2e51](https://source.winehq.org/git/wine.git/commit/690c2e51bfba7028b69523c6f334a5...) and building wine-staging 6.21 without dbghelp-Debug_Symbols allows witcher 3 to launch.
https://bugs.winehq.org/show_bug.cgi?id=52045
Morgan Wesström winehq@pp.dyndns.biz changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |winehq@pp.dyndns.biz
https://bugs.winehq.org/show_bug.cgi?id=52045
Gijs Vermeulen gijsvrm@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|The Witcher 3: crashes when |The Witcher 3 crashes on |launching (regression) |launch
--- Comment #6 from Gijs Vermeulen gijsvrm@gmail.com --- (In reply to sporif from comment #5) -snip-
So, just to confirm, this problem is only present with wine-staging and not upstream wine?
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #7 from sporif@posteo.net --- (In reply to Gijs Vermeulen from comment #6)
(In reply to sporif from comment #5) -snip-
So, just to confirm, this problem is only present with wine-staging and not upstream wine?
Yes
https://bugs.winehq.org/show_bug.cgi?id=52045
Eric Pouech eric.pouech@orange.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |eric.pouech@orange.fr
--- Comment #8 from Eric Pouech eric.pouech@orange.fr --- Created attachment 71091 --> https://bugs.winehq.org/attachment.cgi?id=71091 proposed patch
looking at the trace (in comment #1) from the bug report, there are several items that don't seem right:
1)
0198:fixme:dbghelp:elf_search_auxv can't find symbol in module
this cryptic one means that dbghelp cannot read in debuggee libwine.so.1 for getting back to ELF env...
2) in crash backtrace
1 0x00000003be5cf5cc symt_new_function+0xdc(module=00000000020D7150, compiland =0000000000000000, name="main", addr=0x7d001270, size=0x1e8, sig_type=0000000000 000000) [Z:\builddir\build\BUILD\wine-6.21\dlls\dbghelp\symbol.c:357] in dbghelp (0x00000003be5dc208)
compiland=0000000000000000 is bad :-( (the NULL compiland could stem from problem in 1) (BTW it's very likely related to bisect result in comment #5)
please test the attached patch that should at least prevent the crash
3) related to the patchset dbghelp-Debug_Symbols patchset (and why it creates issue only when applied), I need the both traces the dig further: WINEDEBUG=+dbghelp with the patchset applied and not applied
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #9 from sporif@posteo.net --- Created attachment 71101 --> https://bugs.winehq.org/attachment.cgi?id=71101 Output from app with wine-staging 6.22
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #10 from sporif@posteo.net --- Created attachment 71102 --> https://bugs.winehq.org/attachment.cgi?id=71102 Output from app with wine-staging 6.22 without dbghelp-Debug_Symbols patchset
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #11 from sporif@posteo.net --- (In reply to Eric Pouech from comment #8)
Created attachment 71091 [details] please test the attached patch that should at least prevent the crash
I tested the proposed patch on wine-staging 6.22 and it doesn't prevent the crash.
- related to the patchset dbghelp-Debug_Symbols patchset (and why it creates issue only when applied), I need the both traces the dig further:
WINEDEBUG=+dbghelp with the patchset applied and not applied
I've attached the output of these two scenarios.
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #12 from Eric Pouech eric.pouech@orange.fr --- thanks for posting the traces; I've got what I need
for the patch not fixing the crash, are you sure the crash still happen in the same place (symt_new_function) but not elsewhere (like pe_load_debug_info ?)
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #13 from sporif@posteo.net --- (In reply to Eric Pouech from comment #12)
thanks for posting the traces; I've got what I need
for the patch not fixing the crash, are you sure the crash still happen in the same place (symt_new_function) but not elsewhere (like pe_load_debug_info ?)
I just checked and the output with the patch is the same as stock.
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #14 from Eric Pouech eric.pouech@orange.fr --- Comment on attachment 71091 --> https://bugs.winehq.org/attachment.cgi?id=71091 proposed patch
that's very strange you get the same output: - as the patch really protects the crash (so at least it shouldn't crash at the same place) - some other testers gave positive feedback
so would you mind checking that the patch has really been applied, recompiled, reinstalled
TIA
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #15 from sporif@posteo.net --- (In reply to Eric Pouech from comment #14)
Comment on attachment 71091 [details] proposed patch
that's very strange you get the same output:
- as the patch really protects the crash (so at least it shouldn't crash at
the same place)
- some other testers gave positive feedback
so would you mind checking that the patch has really been applied, recompiled, reinstalled
TIA
Yes the patch really is applied, fully recompiled and installed and it gives the same output. I checked properly, only the symbol addresses change.
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #16 from Eric Pouech eric.pouech@orange.fr --- can you post the part of output similar to what has been provided in comment #1, the top two line of backtrace, like =>0 0x00000003be5ccf18 vector_add+0x8(v=0000000000000020, pool=00000000020D1960) [Z:\builddir\build\BUILD\wine-6.21\dlls\dbghelp\storage.c:171] in dbghelp (0x00000003be5dc208) 1 0x00000003be5cf5cc symt_new_function+0xdc(module=00000000020D0BD0, compiland=0000000000000000, name="main", addr=0x7d001270, size=0x1e8, sig_type=0000000000000000) [Z:\builddir\build\BUILD\wine-6.21\dlls\dbghelp\symbol.c:357] in dbghelp (0x00000003be5dc208) TIA
https://bugs.winehq.org/show_bug.cgi?id=52045
José Pekkarinen koalinux@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |koalinux@gmail.com
--- Comment #17 from José Pekkarinen koalinux@gmail.com --- Hi,
I've been playing for a short while using 6.22, suddenly Gentoo stabilized 6.23 and it broke. I'll attach my output.
Thanks!
José.
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #18 from José Pekkarinen koalinux@gmail.com --- Created attachment 71388 --> https://bugs.winehq.org/attachment.cgi?id=71388 Output when executing TW3 on wine 6.23
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #19 from José Pekkarinen koalinux@gmail.com --- Actually it works, I wasn't plugin my audio devices. Nevermind my comments.
Thanks!
José.
https://bugs.winehq.org/show_bug.cgi?id=52045
--- Comment #20 from sporif@posteo.net --- I don't see this bug anymore on wine-staging v7.0
Could've been fixed in 6.23 but I never tested that.
https://bugs.winehq.org/show_bug.cgi?id=52045
Eric Pouech eric.pouech@orange.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |0bbb161c41dd6dcb505ddef40d1 | |8d32065925341 Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED
--- Comment #21 from Eric Pouech eric.pouech@orange.fr --- closing then adding likely SHA1 for the fix in reply to #20: we don't explicitely track in which version it has been actually fixed, so the important point is that the issue no longer appears
https://bugs.winehq.org/show_bug.cgi?id=52045
Alistair Leslie-Hughes leslie_alistair@hotmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #22 from Alistair Leslie-Hughes leslie_alistair@hotmail.com --- Closing Fixed wine-staging bug.