| /hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp9/common/x86/ |
| H A D | vp9_loopfilter_intrin_sse2.c | 1467 static INLINE void transpose8x16(unsigned char *in0, unsigned char *in1, argument
1473 x0 = _mm_loadl_epi64((__m128i *)in0);
1475 x1 = _mm_loadl_epi64((__m128i *)(in0 + in_p));
1477 x2 = _mm_loadl_epi64((__m128i *)(in0 + 2 * in_p));
1479 x3 = _mm_loadl_epi64((__m128i *)(in0 + 3*in_p));
1481 x4 = _mm_loadl_epi64((__m128i *)(in0 + 4*in_p));
1483 x5 = _mm_loadl_epi64((__m128i *)(in0 + 5*in_p));
1485 x6 = _mm_loadl_epi64((__m128i *)(in0 + 6*in_p));
1487 x7 = _mm_loadl_epi64((__m128i *)(in0 + 7*in_p));
|
| H A D | vp9_idct_intrin_sse2.c | 333 #define TRANSPOSE_8X8(in0, in1, in2, in3, in4, in5, in6, in7, \
336 const __m128i tr0_0 = _mm_unpacklo_epi16(in0, in1); \
338 const __m128i tr0_2 = _mm_unpackhi_epi16(in0, in1); \
383 #define TRANSPOSE_8X4(in0, in1, in2, in3, out0, out1) \
385 const __m128i tr0_0 = _mm_unpacklo_epi16(in0, in1); \
388 in0 = _mm_unpacklo_epi32(tr0_0, tr0_1); /* i1 i0 */ \
392 #define TRANSPOSE_8X8_10(in0, in1, in2, in3, out0, out1) \
394 const __m128i tr0_0 = _mm_unpacklo_epi16(in0, in1); \
458 #define IDCT8(in0, in1, in2, in3, in4, in5, in6, in7, \
475 const __m128i lo_04 = _mm_unpacklo_epi16(in0, in
553 __m128i in0, in1, in2, in3, in4, in5, in6, in7; local
688 __m128i in0, in1, in2, in3, in4, in5, in6, in7; local
723 __m128i in0, in1, in2, in3, in4, in5, in6, in7; local
1012 __m128i in0, in1, in2, in3, in4, in5, in6, in7; local
2407 idct16_sse2(__m128i *in0, __m128i *in1) argument
2413 iadst16_sse2(__m128i *in0, __m128i *in1) argument
2497 __m128i in0[16], in1[16]; local
[all...] |
| /hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp9/encoder/x86/ |
| H A D | vp9_dct_avx2.c | 35 __m128i in0, in1, in2, in3; local
38 in0 = _mm_loadl_epi64((const __m128i *)(input + 0 * stride));
43 in0 = _mm_slli_epi16(in0, 4);
55 __m128i mask = _mm_cmpeq_epi16(in0, k__nonzero_bias_a);
56 in0 = _mm_add_epi16(in0, mask);
57 in0 = _mm_add_epi16(in0, k__nonzero_bias_b);
63 const __m128i r0 = _mm_add_epi16(in0, in
295 __m128i in0 = _mm_load_si128((const __m128i *)(input + 0 * stride)); local
827 __m128i in0, in1, in2, in3, in4, in5, in6, in7; local
1695 load_buffer_16x16_avx2(const int16_t* input, __m128i *in0, __m128i *in1, int stride) argument
1707 write_buffer_16x16_avx2(int16_t *output, __m128i *in0, __m128i *in1, int stride) argument
2535 fdct16_avx2(__m128i *in0, __m128i *in1) argument
2541 fadst16_avx2(__m128i *in0, __m128i *in1) argument
2549 __m128i in0[16], in1[16]; local
[all...] |
| H A D | vp9_dct_sse2.c | 70 __m128i in0, in1; local
74 in0 = _mm_loadl_epi64((const __m128i *)(input + 0 * stride));
78 in0 = _mm_unpacklo_epi64(in0, _mm_loadl_epi64((const __m128i *)
80 // in0 = [i0 i1 i2 i3 iC iD iE iF]
85 in0 = _mm_slli_epi16(in0, 4);
97 __m128i mask = _mm_cmpeq_epi16(in0, k__nonzero_bias_a);
98 in0 = _mm_add_epi16(in0, mas
396 __m128i in0 = _mm_load_si128((const __m128i *)(input + 0 * stride)); local
928 __m128i in0, in1, in2, in3, in4, in5, in6, in7; local
1796 load_buffer_16x16(const int16_t* input, __m128i *in0, __m128i *in1, int stride) argument
1808 write_buffer_16x16(int16_t *output, __m128i *in0, __m128i *in1, int stride) argument
2636 fdct16_sse2(__m128i *in0, __m128i *in1) argument
2642 fadst16_sse2(__m128i *in0, __m128i *in1) argument
2650 __m128i in0[16], in1[16]; local
[all...] |