| /hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp8/common/arm/neon/ |
| H A D | dequant_idct_neon.c | 24 int16x8_t q1, q2, q3, q4, q5, q6; local
35 q4 = vld1q_s16(input);
55 q2 = vreinterpretq_s16_u16(vmulq_u16(vreinterpretq_u16_s16(q4),
64 q4 = vqdmulhq_n_s16(q2, cospi8sqrt2minus1);
67 q4 = vshrq_n_s16(q4, 1);
70 q4 = vqaddq_s16(q4, q2);
72 d10 = vqsub_s16(vget_low_s16(q3), vget_high_s16(q4));
73 d11 = vqadd_s16(vget_high_s16(q3), vget_low_s16(q4));
[all...] |
| /hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp9/common/mips/dspr2/ |
| H A D | vp9_loopfilter_masks_dspr2.h | 286 uint32_t q3, uint32_t q4,
301 /* flat |= (abs(q4 - q0) > thresh) */
302 "subu_s.qb %[c], %[q4], %[q0] \n\t"
303 "subu_s.qb %[r_k], %[q0], %[q4] \n\t"
362 [q2] "r" (q2), [q3] "r" (q3), [q4] "r" (q4),
282 vp9_flatmask5(uint32_t p4, uint32_t p3, uint32_t p2, uint32_t p1, uint32_t p0, uint32_t q0, uint32_t q1, uint32_t q2, uint32_t q3, uint32_t q4, uint32_t *flat2) argument
|
| H A D | vp9_loopfilter_filters_dspr2.h | 537 const uint32_t q4 = *oq4, q5 = *oq5, q6 = *oq6, q7 = *oq7; local
545 /* addition of p6,p5,p4,p3,p2,p1,p0,q0,q1,q2,q3,q4,q5,q6
557 "addu.ph %[add_p6toq6], %[add_p6toq6], %[q4] \n\t"
566 [q4] "r" (q4), [q5] "r" (q5), [q6] "r" (q6),
580 "subu.ph %[res_op6], %[res_op6], %[q4] \n\t"
594 "subu.ph %[res_op5], %[res_op5], %[q4] \n\t"
606 "subu.ph %[res_op4], %[res_op4], %[q4] \n\t"
616 "subu.ph %[res_op3], %[res_op3], %[q4] \n\t"
622 p0 + q0 + q1 + q2 + q3 + q4,
[all...] |
| H A D | vp9_mblpf_horiz_loopfilter_dspr2.c | 36 uint32_t p7, p6, p5, p4, p3, p2, p1, p0, q0, q1, q2, q3, q4, q5, q6, q7; local
102 "lw %[q4], (%[sq4]) \n\t"
108 [q7] "=&r" (q7), [q6] "=&r" (q6), [q5] "=&r" (q5), [q4] "=&r" (q4)
117 vp9_flatmask5(p7, p6, p5, p4, p0, q0, q4, q5, q6, q7, &flat2);
176 "sw %[q4], (%[sq4]) \n\t"
183 : [q6] "r" (q6), [q5] "r" (q5), [q4] "r" (q4), [q3] "r" (q3),
|
| H A D | vp9_mblpf_vert_loopfilter_dspr2.c | 34 uint32_t p7, p6, p5, p4, p3, p2, p1, p0, q0, q1, q2, q3, q4, q5, q6, q7; local
91 "lw %[q4], +4(%[s4]) \n\t"
95 [q5] "=&r" (q5), [q4] "=&r" (q4)
213 /* transpose q4, q5, q6, q7
219 q4 q4_0 q4_1 q4_2 q4_3
226 q4 q4_0 q5_0 q26_0 q7_0
231 "precrq.qb.ph %[prim3], %[q5], %[q4] \n\t"
232 "precr.qb.ph %[prim4], %[q5], %[q4] \n\t"
235 "precr.qb.ph %[q4],
[all...] |
| /hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp9/common/ |
| H A D | vp9_loopfilter_filters.c | 57 uint8_t q3, uint8_t q4) {
60 mask |= (abs(q4 - q0) > thresh) * -1;
246 q4 = *oq4, q5 = *oq5, q6 = *oq6, q7 = *oq7; local
258 q0 + q1 + q2 + q3 + q4, 4);
260 q0 + q1 + q2 + q3 + q4 + q5, 4);
262 q0 + q1 + q2 + q3 + q4 + q5 + q6, 4);
264 q0 * 2 + q1 + q2 + q3 + q4 + q5 + q6 + q7, 4);
266 q0 + q1 * 2 + q2 + q3 + q4 + q5 + q6 + q7 * 2, 4);
268 q0 + q1 + q2 * 2 + q3 + q4 + q5 + q6 + q7 * 3, 4);
270 q0 + q1 + q2 + q3 * 2 + q4
52 flat_mask5(uint8_t thresh, uint8_t p4, uint8_t p3, uint8_t p2, uint8_t p1, uint8_t p0, uint8_t q0, uint8_t q1, uint8_t q2, uint8_t q3, uint8_t q4) argument
[all...] |
| /hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp9/common/x86/ |
| H A D | vp9_loopfilter_intrin_avx2.c | 402 __m128i p4, p3, p2, p1, p0, q0, q1, q2, q3, q4; local
421 q4 = _mm_loadu_si128((__m128i *) (s + 4 * p));
532 _mm_or_si128(_mm_subs_epu8(q4, q0), _mm_subs_epu8(q0, q4)));
588 q256_4 = _mm256_cvtepu8_epi16(q4);
919 q4 = _mm_andnot_si128(flat2, q4);
921 q4 = _mm_or_si128(flat2_q4, q4);
922 _mm_storeu_si128((__m128i *) (s + 4 * p), q4);
[all...] |
| H A D | vp9_loopfilter_intrin_sse2.c | 388 __m128i p4, p3, p2, p1, p0, q0, q1, q2, q3, q4; local
401 q4 = _mm_loadu_si128((__m128i *)(s + 4 * p));
408 _mm_store_si128((__m128i *)&aq[4 * 16], q4);
524 _mm_or_si128(_mm_subs_epu8(q4, q0),
525 _mm_subs_epu8(q0, q4)));
600 q4 = _mm_unpacklo_epi8(_mm_loadl_epi64((__m128i *)(&aq[4 * 16] + off)),
664 c = _mm_add_epi16(q4, c);
712 c = _mm_add_epi16(_mm_sub_epi16(c, _mm_add_epi16(p3, q3)), q4);
718 c = _mm_add_epi16(_mm_sub_epi16(c, _mm_add_epi16(p2, q4)), q5);
|
| /hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp9/encoder/x86/ |
| H A D | vp9_dct_avx2.c | 325 const __m128i q4 = _mm_sub_epi16(in3, in4); local
394 const __m128i x0 = _mm_add_epi16(q4, r0);
395 const __m128i x1 = _mm_sub_epi16(q4, r0);
1236 const __m128i q4 = _mm_sub_epi16(input3, input4); local
1307 const __m128i x0 = _mm_add_epi16(q4, r0);
1308 const __m128i x1 = _mm_sub_epi16(q4, r0);
|
| H A D | vp9_dct_sse2.c | 426 const __m128i q4 = _mm_sub_epi16(in3, in4); local
495 const __m128i x0 = _mm_add_epi16(q4, r0);
496 const __m128i x1 = _mm_sub_epi16(q4, r0);
1337 const __m128i q4 = _mm_sub_epi16(input3, input4); local
1408 const __m128i x0 = _mm_add_epi16(q4, r0);
1409 const __m128i x1 = _mm_sub_epi16(q4, r0);
|