http://bugs.winehq.org/show_bug.cgi?id=35721
Bug ID: 35721 Summary: Black and White: Unhandled page fault on read access to 0x00000138 at address 0x37f204 Product: Wine Version: 1.7.13 Hardware: x86 OS: Mac OS X Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs@winehq.org Reporter: linus@folkdatorn.se
Created attachment 47710 --> http://bugs.winehq.org/attachment.cgi?id=47710 Complete output from wine runblack.exe
When trying to run Black and White on wine 1.7.13 it crashes when starting with the following error:
wine: Unhandled page fault on read access to 0x00000138 at address 0x37f204 (thread 003e), starting debugger...
This does NOT happen when using wine 1.7.5.
My computer takes forever just to build one version of wine so I haven't had the time for regression testing yet. If someone could do it for me I would be very thankful.
Attached is the complete output from `wine runblack.exe`.
http://bugs.winehq.org/show_bug.cgi?id=35721
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression Summary|Black and White: Unhandled |Black and White: crashes on |page fault on read access |start |to 0x00000138 at address | |0x37f204 |
http://bugs.winehq.org/show_bug.cgi?id=35721
Christian Widmer shadow@umbrox.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |shadow@umbrox.de
--- Comment #1 from Christian Widmer shadow@umbrox.de --- The most recent git version (commit 2bb1059ad24942f0ea13d3e49042c8ea85c4cbd6 at the time of writing) does not crash for me. Although the game does not crash at the completion of the temple, the textures are still messed up.
http://bugs.winehq.org/show_bug.cgi?id=35721
George Khan a3663644@drdrb.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |a3663644@drdrb.net
--- Comment #2 from George Khan a3663644@drdrb.net --- If you have more than 1 core on your machine you can speed up compile time by
make -j4
where number is your code amount, in this case 4.
It speeds ups a lot.
Also you can disable test cases when doing bisect (don't remember how), but it is somewhere winehq guide.
https://bugs.winehq.org/show_bug.cgi?id=35721
hanska2@luukku.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |hanska2@luukku.com
--- Comment #3 from hanska2@luukku.com --- 1.7.26 got just released. Are you able to try that?
https://bugs.winehq.org/show_bug.cgi?id=35721
~J~ funtimeschaos@rocketmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |funtimeschaos@rocketmail.co | |m
--- Comment #4 from ~J~ funtimeschaos@rocketmail.com --- Having the same problem with wine 1.7.30 so still an issue
https://bugs.winehq.org/show_bug.cgi?id=35721
Sebastian Lackner sebastian@fds-team.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |sebastian@fds-team.de
--- Comment #5 from Sebastian Lackner sebastian@fds-team.de --- The information in this bug report is not very useful so far. I don't have the game unfortunately, so I cannot test myself.
First of all it would be very useful if someone could attach the full terminal output (the attachment in the description is truncated, most likely the crash dialog was not closed, which is necessary to get the full crash dump). Also make sure to have debug symbols installed if possible.
When its really a regression, it would also be very useful to know which commit between 1.7.5 - 1.7.13 exactly broke the game. Take a look here for more information how to do a regression test: http://wiki.winehq.org/RegressionTesting
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #6 from ~J~ funtimeschaos@rocketmail.com --- Created attachment 49935 --> https://bugs.winehq.org/attachment.cgi?id=49935 Output of runblack.exe with WINE 1.6.2
I just tested with wine 1.6.2 and still experiencing the same symptoms. I've attached the base output - what debug flags do you want to run this with, exactly?
https://bugs.winehq.org/show_bug.cgi?id=35721
roger@mailinator.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |roger@mailinator.com
--- Comment #7 from roger@mailinator.com --- Even 1.6.2 is called stable, I think bugs should be opened against latest development version. You should try (and use) latest, which is currently at 1.7.30.
The gab between stable and development version grows every 2 weeks, because that's how often comes new version of development version.
Can you retest with 1.7.30 or newer?
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #8 from ~J~ funtimeschaos@rocketmail.com --- Created attachment 49950 --> https://bugs.winehq.org/attachment.cgi?id=49950 wine 1.7.30-git output
Done. Here's the output I get with wine I compiled from git.
Note that I do not get a crash dialogue.
https://bugs.winehq.org/show_bug.cgi?id=35721
Ken Sharp imwellcushtymelike@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords|regression |
--- Comment #9 from Ken Sharp imwellcushtymelike@gmail.com --- Please try again in Wine 1.7.48.
https://bugs.winehq.org/show_bug.cgi?id=35721
winetest@luukku.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |winetest@luukku.com
--- Comment #10 from winetest@luukku.com --- (In reply to Ken Sharp from comment #9)
Please try again in Wine 1.7.48.
Can you test newer wine version? 1.7 is really outdated already.
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #11 from Linus Unnebäck linus@folkdatorn.se --- Created attachment 61161 --> https://bugs.winehq.org/attachment.cgi?id=61161 Backtrace from Wine 3.6
This is still happening on Wine 3.6, see the attached backtrace.
I'd be happy to do more debugging and/or provide more information.
It seems like it only happens on patched versions of Black and White:
1.00 - Works 1.20 - Crashes 1.41 - Crashes
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #12 from Christian Widmer shadow@umbrox.de --- The game crashed neither with version 1.0 nor with version 1.2 on my end. I might add, that I used a 32-bit prefix in both cases with the Windows version set to Windows 7 in at least the 1.2 case (I do not remember what was set in the 1.0 one).
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #13 from Linus Unnebäck linus@folkdatorn.se ---
The game crashed neither with version 1.0 nor with version 1.2 on my end. I might add, that I used a 32-bit prefix in both cases with the Windows version set to Windows 7 in at least the 1.2 case (I do not remember what was set in the 1.0 one).
I tried installing, patching and running the game in a WINEARCH=win32 prefix, but got the same crash :(
I'm on macOS 10.13.4 btw, which system are you running?
In an effort to try and track down the original regression I used Wineskin to try some different versions of wine. This was all tested with Black & White 1, patched to 1.20 using the official patch.
1.7.5 - Works 1.7.6 - Works 1.7.7 - Works 1.7.8 - Works 1.7.9 - Works 1.7.10 - Works 1.7.11 - Works 1.7.12 - Crashes
I'm currently trying to build wine from source to track down the commit, but I cannot get the building to work :(
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #14 from Linus Unnebäck linus@folkdatorn.se --- Okay, so it seems like 1.0 is also crashing for me when installed using the CD-ROM Setup.exe.
Also, I cannot get the RIP-version from myabandonware.com to work either, which is the one I tested with in the first place that worked. It doesn't crash, but instead hangs with a black screen.
I've been at this all day and it seems like I'm worse of then when I started :D
(RIP-version: https://www.myabandonware.com/game/black-white-a33)
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #15 from Linus Unnebäck linus@folkdatorn.se --- Okay, just figured out a big piece of the puzzle. There is something in wine, that at least on macOS affects Black & White. But it doesn't affect the game itself, it affects the installer.
If I run the installer in Wineskin with wine 1.7.5, then copy the entire wine-prefix out and use it with wine 3.6 from Homebrew, then I can successfully run the game!
I'm going to try and investigate the difference between the two prefixes, but my guess is that it has something do to with the registry. I tried to debug the issue using +relay and it seems to crash after reading something from the registry... I'm not 100% about this though :)
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #16 from Linus Unnebäck linus@folkdatorn.se --- Created attachment 61176 --> https://bugs.winehq.org/attachment.cgi?id=61176 Backtrace v1.41 with Wine 3.6
Okay, the last post before bed.
I just found out that I've made a horrible mistake! Apparently, Black & White need you to execute runblack.exe from the directory it's installed to. Otherwise, it crashes with this very error.
So much of my testing on macOS might, unfortunately, be invalid...
Now version 1.20 seems to be working in Wine 3.6 on macOS 10.13.4.
That being said, I'm still experiencing a segfault with version 1.41, which prevents me from using that patch.
The error occurs slightly later than the other one though, as soon as the loading screen has drawn the first frame.
Maybe I should open another bug for that though?
Reading my first comment here, it seems like that was a problem that I can no longer reproduce. That was a read access to 0x00000138, and today I've been debugging the 0x00000004 one, which is caused by being in the wrong directory. (I guess that this would happen on Windows as well, although I haven't verified that)
The problem with 1.41 is a read access to 0x00000008:
wine: Unhandled page fault on read access to 0x00000008 at address 0x56e1fe (thread 0009), starting debugger...
Should we open a new bug, and close this as fixed?
https://bugs.winehq.org/show_bug.cgi?id=35721
--- Comment #17 from Christian Widmer shadow@umbrox.de --- Well, that is quite embarrassing. I actually confused this bug with bug 15984 to which I commented just a few days ago. When I then got the email notification for this bug here, I just read "crash" and "Black and White" and immediately thought of the other bug. I should have read more carefully.
Because I am actually using Gentoo Linux x86_64 and this bug seems to be specific to Mac OS X my comments are most likely irrelevant. I am sorry for the noise.
I am happy that you seem to have managed to resolve your issue in the meantime, though.
https://bugs.winehq.org/show_bug.cgi?id=35721
tokktokk fdsfgs@krutt.org changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |fdsfgs@krutt.org
https://bugs.winehq.org/show_bug.cgi?id=35721
mrdeathjr28@yahoo.es changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mrdeathjr28@yahoo.es