void ff_put_vc1_mspel_mc00_mmx(uint8_t *dst, const uint8_t *src,
int stride, int rnd)
{
- ff_put_pixels8_mmx(dst, src, stride, 8);
+ put_pixels8_mmx(dst, src, stride, 8);
}
- #endif /* HAVE_YASM */
- #if HAVE_INLINE_ASM
+#if CONFIG_DIRAC_DECODER
+#define DIRAC_PIXOP(OPNAME2, OPNAME, EXT)\
+void ff_ ## OPNAME2 ## _dirac_pixels8_ ## EXT(uint8_t *dst, const uint8_t *src[5], int stride, int h)\
+{\
+ if (h&3)\
+ ff_ ## OPNAME2 ## _dirac_pixels8_c(dst, src, stride, h);\
+ else\
+ OPNAME ## _pixels8_ ## EXT(dst, src[0], stride, h);\
+}\
+void ff_ ## OPNAME2 ## _dirac_pixels16_ ## EXT(uint8_t *dst, const uint8_t *src[5], int stride, int h)\
+{\
+ if (h&3)\
+ ff_ ## OPNAME2 ## _dirac_pixels16_c(dst, src, stride, h);\
+ else\
+ OPNAME ## _pixels16_ ## EXT(dst, src[0], stride, h);\
+}\
+void ff_ ## OPNAME2 ## _dirac_pixels32_ ## EXT(uint8_t *dst, const uint8_t *src[5], int stride, int h)\
+{\
+ if (h&3) {\
+ ff_ ## OPNAME2 ## _dirac_pixels32_c(dst, src, stride, h);\
+ } else {\
+ OPNAME ## _pixels16_ ## EXT(dst , src[0] , stride, h);\
+ OPNAME ## _pixels16_ ## EXT(dst+16, src[0]+16, stride, h);\
+ }\
+}
+
+#if HAVE_MMX_INLINE
+DIRAC_PIXOP(put, put, mmx)
+DIRAC_PIXOP(avg, avg, mmx)
+#endif
+
+#if HAVE_YASM
+DIRAC_PIXOP(avg, ff_avg, mmxext)
+
+void ff_put_dirac_pixels16_sse2(uint8_t *dst, const uint8_t *src[5], int stride, int h)
+{
+ if (h&3)
+ ff_put_dirac_pixels16_c(dst, src, stride, h);
+ else
+ ff_put_pixels16_sse2(dst, src[0], stride, h);
+}
+void ff_avg_dirac_pixels16_sse2(uint8_t *dst, const uint8_t *src[5], int stride, int h)
+{
+ if (h&3)
+ ff_avg_dirac_pixels16_c(dst, src, stride, h);
+ else
+ ff_avg_pixels16_sse2(dst, src[0], stride, h);
+}
+void ff_put_dirac_pixels32_sse2(uint8_t *dst, const uint8_t *src[5], int stride, int h)
+{
+ if (h&3) {
+ ff_put_dirac_pixels32_c(dst, src, stride, h);
+ } else {
+ ff_put_pixels16_sse2(dst , src[0] , stride, h);
+ ff_put_pixels16_sse2(dst+16, src[0]+16, stride, h);
+ }
+}
+void ff_avg_dirac_pixels32_sse2(uint8_t *dst, const uint8_t *src[5], int stride, int h)
+{
+ if (h&3) {
+ ff_avg_dirac_pixels32_c(dst, src, stride, h);
+ } else {
+ ff_avg_pixels16_sse2(dst , src[0] , stride, h);
+ ff_avg_pixels16_sse2(dst+16, src[0]+16, stride, h);
+ }
+}
+#endif
+#endif
+
+/* XXX: Those functions should be suppressed ASAP when all IDCTs are
+ * converted. */
+#if CONFIG_GPL
+static void ff_libmpeg2mmx_idct_put(uint8_t *dest, int line_size,
+ int16_t *block)
+{
+ ff_mmx_idct(block);
+ ff_put_pixels_clamped_mmx(block, dest, line_size);
+}
+
+static void ff_libmpeg2mmx_idct_add(uint8_t *dest, int line_size,
+ int16_t *block)
+{
+ ff_mmx_idct(block);
+ ff_add_pixels_clamped_mmx(block, dest, line_size);
+}
+
+static void ff_libmpeg2mmx2_idct_put(uint8_t *dest, int line_size,
+ int16_t *block)
+{
+ ff_mmxext_idct(block);
+ ff_put_pixels_clamped_mmx(block, dest, line_size);
+}
+
+static void ff_libmpeg2mmx2_idct_add(uint8_t *dest, int line_size,
+ int16_t *block)
+{
+ ff_mmxext_idct(block);
+ ff_add_pixels_clamped_mmx(block, dest, line_size);
+}
+#endif
+
static void vector_clipf_sse(float *dst, const float *src,
float min, float max, int len)
{