Re: [Mingw-cross-env-list] [call for review UPDATED] Qt 4.7-beta1

[Mark Brand]

> >   I posted yet another update. Explanation below.
> >
> > http://www.xs4all.nl/~brandm/mingw-cross-env/qt-1-win32.patch
> > http://www.xs4all.nl/~brandm/mingw-cross-env/qt.mk
> ...

> I'm still often confused by host/target and qt introduces a new
> xplatform that I haven't fully understood.

Describe the platform we are building Qt for (MinGW aka "win32-g++"):
 -arch  CFG_ARCH
-xplatform XPLATFORM
XQMAKESPEC
-host-big-endian or -host-little-endian switch CFG_HOST_ENDIAN

Describe the platform we are building Qt on (some form of Unix):
-host-arch CFG_HOST_ARCH
-platform PLATFORM
QMAKESPEC
-host-little-endian or -host-big-endian switch CFG_ENDIAN

It's of course not quite that simple, but it's a good start.

> I think I can see the
> intent of the patch, unfortunately it fails fairly early (log
> attached). I fiddled with using -host-arch and CFG_HOST_ARCH; that
> failed a lot later in the build, but was just a shot in the dark.

Now I see that the -arch value intended for CFG_ARCH also ends up in 
CFG_MAC_ARCHS, which causes the error.  I have added another patch to 
the configure script to keep "windows" out of CFG_MAC_ARCHS.

I've updated the files at the URLs above.

> The same error occurs in the 4.6 patch as soon as the cross-build starts.

If this solves the problem, let me know and I'll do the same thing for 
Qt 4.6.2.

> FWIW, there's no simple, cross-platform, way to determine the
> `bitness` of an OSX system. Configure scripts are easily mislead by
> tools such as uname, which is why we use gcc4.0 for mingw-cross-env.

I hope the configure script does the right thing on the Mac now that we 
don't specify -host-little-endian. If it goes wrong, could you try 
putting that option back in? You could also try "-host-arch i386".

-Mark