[Bug 11305] New: Assaulttech1: Stack overflow while loading

http://bugs.winehq.org/show_bug.cgi?id=11305

Summary: Assaulttech1: Stack overflow while loading
           Product: Wine
           Version: 0.9.49.
          Platform: PC
               URL: http://www.benzinestudios.com/storage/demo/AT1a220ins.ex
                    e |
                    http://www.benzinestudios.com/storage/demo/AT1nofasttext
                    .zip
        OS/Version: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: directx-d3d
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

While 0.9.48 works fine, the game freezes while loading on 0.9.49 after
displaying the intro pics. I ran a regression test:

f6637cae983323eb60d989e5d144b15f1fee2846 is first bad commit
commit f6637cae983323eb60d989e5d144b15f1fee2846
Author: Stefan Dösinger 
Date:   Sat Oct 13 12:54:10 2007 +0200

wined3d: Move texture loading to LoadLocation.

:040000 040000 e71a95ad75ae7e126fb8ad65260c3deacf80ec14
caf23730cc67f4ee80653a799cf67cfbe84956e3 M      dlls

Steps to reproduce:
- Install the game using the First URL and use the binary provided by the
second URL to start it.
- Select a resolution. Fonts may corrupted, just click on the upper left hex to
select. This is only needed once.
- Intro pics will get displayed. Click mouse1 to skip, there are 3 of them.
- The game will now sit there sucking CPU time for about a minute or so. After
that there will be a message that the engine started. Just wait for it, and
don't alt+tab
- In case of 0.9.48: You will be provided with a dialog to change graphic
settings. This is the sign of success
- In case of 0.9.49: The game will freeze with the following console output:
err:seh:setup_exception stack overflow 4 bytes in thread 0009 eip 7ef84e81 esp
00230ffc stack 0x231000-0x340000

I'm sorry that I'm not able to test 0.9.53 (the current one), but that version
will not even get to the point of displaying the intro pics, it will freeze on
the spot without any suspicious console output.
I'm still investigating that as I'm unsure whether this is a seperate bug. If
someone could write a patch for me that disables the bad commit mentioned
above, I could try applying it to current sources and see what happens.