[Bug 22523] New: Unreal Engine 1 Games fail at Startup due to wined3d regression in 1.1.41-1.1.42

http://bugs.winehq.org/show_bug.cgi?id=22523

Summary: Unreal Engine 1 Games fail at Startup due to wined3d
                    regression in 1.1.41-1.1.42
           Product: Wine
           Version: 1.1.42
          Platform: x86
        OS/Version: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: directx-d3d
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

Created an attachment (id=27604)
 --> (http://bugs.winehq.org/attachment.cgi?id=27604)
wine-1.1.42 output, and d3d trace

Unreal Engine 1 Games fail at Startup due to wined3d regression in
1.1.41-1.1.42

Due to a regression in wine during the 1.1.41 to 1.1.42 timeframe, Unreal
Engine 1 based games (tested with Unreal Gold, Unreal Tournament, and Deus Ex)
fail to startup due to a d3d error. The error dialog box produced by the
application says:

----
General protection fault!

History: UD3DRenderDevice::SetRes <- UD3DRenderDevice::Init <-
UWindowsViewport::TryRenderDevice <-
UWindowsViewport::OpenWindow <- UGameEngine::Init <- InitEngine
----

A regression test in dlls/wined3d shows the following to be the regression:

19b6f5ca75b43f61a366aa521eab9ddbfc4f99dd is first bad commit
commit 19b6f5ca75b43f61a366aa521eab9ddbfc4f99dd
Author: Stefan Dösinger 
Date:   Sun Mar 28 16:25:04 2010 +0200

wined3d: Control the onscreen depth stencil format in the swapchain.

This allows the swapchain to know what depth format its window contexts
have to see if the
    requested depth format is compatible or a FBO fallback is needed, and it
will be needed to
    set the onscreen format to the requested auto depth stencil format instead
of the
    let's-hope-it-fits D24_UNORM_S8_UINT format.

:040000 040000 2f6e9d3017c2e6715b6561d25cca4e947ee95000
44d70c59cce2f82aa3f2d8b6e19679588f81e1b8 M    dlls

http://source.winehq.org/git/wine.git/?a=commit;h=19b6f5ca75b43f61a366aa521e...

This is on a laptop running Slackware 13.0 (linux-2.6.30-5, xorg-server 1.6.3)
using intel-2.8.1 driver, and OffscreenRenderingMode=backbuffer. I'll probably
end up testing this on a similiar (albeit nvidia-195.36.15,
OffscreenRenderingMode=fbo) desktop.

I first noticed this problem while on wine-1.1.43, but attached is the plain
output, followed by a verbose d3d trace (WINEDEBUG="+d3d") of wine-1.1.42.