MMX(2) accelerated yuv420->rgb24 routine, trivial as essentially as it just

swaps R and B inputs for the existing yuv420->bgr24 routine.

Originally committed as revision 30254 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

libswscale/x86/yuv2rgb_mmx.c

@@ -72,7 +72,8 @@ SwsFunc ff_yuv2rgb_init_mmx(SwsContext *c)
                 if (HAVE_7REGS) return yuva420_bgr32_MMX2;
                 break;
             } else return yuv420_bgr32_MMX2;
-        case PIX_FMT_BGR24:  return yuv420_rgb24_MMX2;
+        case PIX_FMT_RGB24:  return yuv420_rgb24_MMX2;
+        case PIX_FMT_BGR24:  return yuv420_bgr24_MMX2;
         case PIX_FMT_RGB565: return yuv420_rgb16_MMX2;
         case PIX_FMT_RGB555: return yuv420_rgb15_MMX2;
         }
@@ -89,7 +90,8 @@ SwsFunc ff_yuv2rgb_init_mmx(SwsContext *c)
                 if (HAVE_7REGS) return yuva420_bgr32_MMX;
                 break;
             } else return yuv420_bgr32_MMX;
-        case PIX_FMT_BGR24:  return yuv420_rgb24_MMX;
+        case PIX_FMT_RGB24:  return yuv420_rgb24_MMX;
+        case PIX_FMT_BGR24:  return yuv420_bgr24_MMX;
         case PIX_FMT_RGB565: return yuv420_rgb16_MMX;
         case PIX_FMT_RGB555: return yuv420_rgb15_MMX;
         }