Adam Jackson írta:
On Mon, 2008-06-16 at 16:34 +0200, Zoltan Boszormenyi wrote:
> The rpm build machinery accepted that I want an i386 binary
> but the Mesa configure scripts didn't:
>
> ...
> gcc -c -I../../include -I../../src/mesa -I../../src/mesa/main
> -I../../src/mesa/glapi -I../../src/mesa/math -I../../src/mesa/tnl
> -I../../src/mesa/shader -I../../src/mesa/shader/grammar
> -I../../src/mesa/shader/slang -I../../src/mesa/swrast
> -I../../src/mesa/swrast_setup -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2
> -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m32
> -march=i386 -mtune=generic -fasynchronous-unwind-tables
> -fvisibility=hidden -Os -Wall -Wmissing-prototypes -std=c99 -ffast-math
> -fno-strict-aliasing -fPIC -DUSE_X86_64_ASM -D_POSIX_SOURCE
> -D_POSIX_C_SOURCE=199309L -D_BSD_SOURCE -D_SVID_SOURCE -D_GNU_SOURCE
> -DPTHREADS -DHAVE_POSIX_MEMALIGN x86-64/xform4.S -o x86-64/xform4.o
> ...
>
> Notice the "-m32 -march=i386" (that was added by RPM) and
> "-DUSE_X86_64_ASM" that's coming from Mesa.
>
Then Mesa's configure script is picking the wrong target script. Mesa
bug, not spec bug.
- ajax
I was able to fix it by modifying the spec file and one Makefile in Mesa:
https://bugzilla.redhat.com/show_bug.cgi?id=451668