[Pixman] mmx build regression

[Matt Turner]

On Wed, Feb 22, 2012 at 2:27 AM, Jeremy Huddleston
<jeremyhu at freedesktop.org> wrote:
> I just got my tinderbox back up today and noticed this build regression in pixman.  I haven't looked into it yet, but git-log blames Matt ;)
>
> http://tinderbox.x.org/builds/2012-02-22-0001/logs/pixman/#build
>
> The warnings are harmless.  I wish the error was more verbose, but that's all I got.  I'll try bisecting when I get a chance, but I'm overloaded right now and wanted to report this since you might have an idea before I get some time.
>
> --Jeremy
>
> libtool: compile:  /opt/local/libexec/llvm-3.0/libexec/scan-build/ccc-analyzer -DHAVE_CONFIG_H -I. -I.. -I/Users/jeremy/src/freedesktop/jhbuild/build/include -I/Users/jeremy/src/freedesktop/jhbuild/external/build/include -mmmx -Winline -O0 -pipe -Wall -Wextra -Wno-sign-compare -Wno-unused-parameter -Wno-missing-field-initializers -Wformat=2 -fdiagnostics-show-category=name -Wall -fno-strict-aliasing -fvisibility=hidden -D_REENTRANT -c pixman-mmx.c  -fno-common -DPIC -o .libs/libpixman_mmx_la-pixman-mmx.o
>
> pixman-mmx.c:63:73: warning: unknown attribute '__artificial__' ignored [-Wattributes,Semantic Issue]
> extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
>                                                                        ^
> pixman-mmx.c:73:73: warning: unknown attribute '__artificial__' ignored [-Wattributes,Semantic Issue]
> extern __inline __m64 __attribute__((__gnu_inline__, __always_inline__, __artificial__))
>                                                                        ^
> fatal error: error in backend: Invalid operand for inline asm constraint 'K'!
> make[3]: *** [libpixman_mmx_la-pixman-mmx.lo] Error 1
> make[2]: *** [all] Error 2
> make[1]: *** [all-recursive] Error 1
> make: *** [all] Error 2
>
> scan-build: 2 bugs found.
> scan-build: Run 'scan-view /Users/jeremy/src/freedesktop/jhbuild/analyzer/yuffie/20120221-1900/pixman/2012-02-21-1' to examine bug reports.
>
>
>

This looks like an LLVM/clang bug.

The 'K' constraint is documented here
http://gcc.gnu.org/onlinedocs/gcc/Machine-Constraints.html#Machine-Constraints
as an 8-bit immediate.