Ticket #13021: macports-mplayer-mmx.diff

libavcodec/i386/dsputil_mmx.c

@@ -631 +631 @@
 }
 
 static inline void transpose4x4(uint8_t *dst, uint8_t *src, int dst_stride, int src_stride){
+#if 1
+  /* borrowed from: http://lists-archives.org/mplayer-dev-eng/14875-compiling-with-intel-c.html */
+  /* Ugly, but uses a minimal number of operands just to fit into Intel C limits */
+  asm volatile("movd (%1), %%mm0               \n\t"
+               "mov %3, %%eax                  \n\t"
+               "movd (%1, %%eax), %%mm1        \n\t"
+               "add %3, %%eax                  \n\t"
+               "movd (%1, %%eax), %%mm2        \n\t"
+               "add %3, %%eax                  \n\t"
+               "movd (%1, %%eax), %%mm3        \n\t"
+               "punpcklbw %%mm1, %%mm0         \n\t"
+               "punpcklbw %%mm3, %%mm2         \n\t"
+               "movq %%mm0, %%mm1              \n\t"
+               "punpcklwd %%mm2, %%mm0         \n\t"
+               "punpckhwd %%mm2, %%mm1         \n\t"
+               "movd %%mm0, (%0)               \n\t"
+               "punpckhdq %%mm0, %%mm0         \n\t"
+               "mov %2, %%eax                  \n\t"
+               "movd %%mm0, (%0, %%eax)        \n\t"
+               "add %2, %%eax                  \n\t"
+               "movd %%mm1, (%0, %%eax)        \n\t"
+               "punpckhdq %%mm1, %%mm1         \n\t"
+               "add %2, %%eax                  \n\t"
+               "movd %%mm1, (%0, %%eax)        \n\t"
+               :: "r" (dst), "r" (src), "r"(dst_stride), "r"(src_stride)
+               : "eax");
+#else
     asm volatile( //FIXME could save 1 instruction if done as 8x4 ...
         "movd  %4, %%mm0                \n\t"
         "movd  %5, %%mm1                \n\t"
@@ -657 +684 @@
            "m" (*(uint32_t*)(src + 2*src_stride)),
            "m" (*(uint32_t*)(src + 3*src_stride))
     );
+#endif
 }
 
 static void h263_h_loop_filter_mmx(uint8_t *src, int stride, int qscale){