[Bug 47970] Legends of Runeterra crashes at launch

WineHQ Bugzilla wine-bugs at winehq.org
Mon Mar 29 16:35:19 CDT 2021


https://bugs.winehq.org/show_bug.cgi?id=47970

--- Comment #22 from Daniel Bomar <dbdaniel42 at gmail.com> ---
(In reply to Zebediah Figura from comment #18)
> (In reply to Daniel Bomar from comment #17)
> > I did some testing on this using the Wine sources from git and I can confirm
> > a regression.  With all staging patches applied:
> > 
> > Last known working version: v5.9 (game seems to fully work and I was able to
> > complete a game online)
> > First known bad version: v.5.10 (game does not launch)
> > 
> > I was going to do a more thorough bisect but the issue is the game seems to
> > depend on multiple staging patches in order to launch.  I tried compiling
> > with only ntdll-x86_64_SegDs, ntdll-WRITECOPY, and
> > user32-InternalGetWindowIcon but this configuration failed to launch the
> > game on v5.9.
> > 
> > The wiki page on regression testing states that when staging patches are
> > required, the next step is to narrow down which patches are required but I'm
> > a little stuck here as I tried the only 3 patches mentioned in bug reports
> > and it still does not work.  I'm only able to get it to work by applying all
> > patches.
> > 
> > If anyone knows which patches are required, please let me know and I can do
> > a better bisect.
> 
> One thing you can do is to reverse-bisect between wine-staging and upstream
> wine. Of course, that's not ideal in its own ways. It's a rather unfortunate
> case when a game depends on several wine-staging patch sets.
> 
> Another possible approach, that may be helpful enough here, is to try
> individual wine-staging rebase versions between 5.9 and 5.10 That doesn't
> constitute a full bisect, of course, but would at least narrow down to the
> commits for a particular day, which may be enough of a clue in itself to
> make a good guess.

Thanks for the advice.  I just uploaded the results of the bisect plus the
error message I was getting from the bad revisions.  I wasn't sure how to
reverse-bisect between two different git repositories so I started with the 2nd
method you mentioned.  That showed the first bad commit was a rebase so I used
that information to do a bisect on upstream Wine.  I had to do a bit of manual
fiddling with the patches to get the intermediate builds to compile but I think
my results are correct.

# first bad commit: [be0eb9c92eb7a4fcd9d0d48568c8ed5e8326ef0b] ntdll: Move the
thread startup code to the Unix library

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.



More information about the wine-bugs mailing list