Lines Matching refs:q1
279 #define GET_NOTHEV(p1, p0, q0, q1, hev_thresh, not_hev) { \ argument
282 const __m128i t2 = MM_ABS(q1, q0); \
292 #define GET_BASE_DELTA(p1, p0, q0, q1, o) { \ argument
294 o = _mm_subs_epi8(p1, q1); /* p1 - q1 */ \
327 const __m128i* q1, int thresh, __m128i *mask) { in NeedsFilter() argument
328 __m128i t1 = MM_ABS(*p1, *q1); // abs(p1 - q1) in NeedsFilter()
347 const __m128i* q1, int thresh) { in DoFilter2() argument
351 const __m128i q1s = _mm_xor_si128(*q1, sign_bit); in DoFilter2()
353 NeedsFilter(p1, p0, q0, q1, thresh, &mask); in DoFilter2()
368 __m128i* q0, __m128i* q1, in DoFilter4() argument
375 GET_NOTHEV(*p1, *p0, *q0, *q1, hev_thresh, not_hev); in DoFilter4()
378 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter4()
380 t1 = _mm_subs_epi8(*p1, *q1); // p1 - q1 in DoFilter4()
406 *q1 = _mm_subs_epi8(*q1, t3); // q1 -= t3 in DoFilter4()
410 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter4()
415 __m128i* q0, __m128i* q1, __m128i *q2, in DoFilter6() argument
421 GET_NOTHEV(*p1, *p0, *q0, *q1, hev_thresh, not_hev); in DoFilter6()
424 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter6()
427 GET_BASE_DELTA(*p1, *p0, *q0, *q1, a); in DoFilter6()
459 UPDATE_2PIXELS(*p1, *q1, a1_lo, a1_hi); in DoFilter6()
464 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter6()
515 __m128i* q0, __m128i* q1) { in Load16x4() argument
533 Load8x4(r8, stride, p0, q1); in Load16x4()
543 *q0 = _mm_unpacklo_epi64(t2, *q1); in Load16x4()
544 *q1 = _mm_unpackhi_epi64(t2, *q1); in Load16x4()
558 __m128i* q0, __m128i* q1) { in Store16x4() argument
570 *q0 = _mm_unpacklo_epi8(t1, *q1); in Store16x4()
571 *q1 = _mm_unpackhi_epi8(t1, *q1); in Store16x4()
582 *p1 = _mm_unpacklo_epi16(t1, *q1); in Store16x4()
583 *q1 = _mm_unpackhi_epi16(t1, *q1); in Store16x4()
591 Store4x4(q1, r8, stride); in Store16x4()
602 __m128i q1 = _mm_loadu_si128((__m128i*)&p[stride]); in SimpleVFilter16SSE2() local
604 DoFilter2(&p1, &p0, &q0, &q1, thresh); in SimpleVFilter16SSE2()
612 __m128i p1, p0, q0, q1; in SimpleHFilter16SSE2() local
616 Load16x4(p, p + 8 * stride, stride, &p1, &p0, &q0, &q1); in SimpleHFilter16SSE2()
617 DoFilter2(&p1, &p0, &q0, &q1, thresh); in SimpleHFilter16SSE2()
618 Store16x4(p, p + 8 * stride, stride, &p1, &p0, &q0, &q1); in SimpleHFilter16SSE2()
677 #define COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask) { \ argument
682 NeedsFilter(&p1, &p0, &q0, &q1, thresh, &fl_yes); \
691 __m128i p2, p1, p0, q0, q1, q2; in VFilter16SSE2() local
698 LOAD_H_EDGES4(p, stride, q0, q1, q2, t1); in VFilter16SSE2()
699 MAX_DIFF2(t1, q2, q1, q0, mask); in VFilter16SSE2()
701 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter16SSE2()
702 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in VFilter16SSE2()
709 _mm_storeu_si128((__m128i*)&p[1 * stride], q1); in VFilter16SSE2()
716 __m128i p3, p2, p1, p0, q0, q1, q2, q3; in HFilter16SSE2() local
722 Load16x4(p, p + 8 * stride, stride, &q0, &q1, &q2, &q3); // q0, q1, q2, q3 in HFilter16SSE2()
723 MAX_DIFF2(q3, q2, q1, q0, mask); in HFilter16SSE2()
725 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter16SSE2()
726 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in HFilter16SSE2()
729 Store16x4(p, p + 8 * stride, stride, &q0, &q1, &q2, &q3); in HFilter16SSE2()
737 __m128i t1, t2, p1, p0, q0, q1; in VFilter16iSSE2() local
747 LOAD_H_EDGES4(p, stride, q0, q1, t1, t2); in VFilter16iSSE2()
748 MAX_DIFF2(t2, t1, q1, q0, mask); in VFilter16iSSE2()
750 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter16iSSE2()
751 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in VFilter16iSSE2()
757 _mm_storeu_si128((__m128i*)&p[1 * stride], q1); in VFilter16iSSE2()
766 __m128i t1, t2, p1, p0, q0, q1; in HFilter16iSSE2() local
774 Load16x4(b, b + 8 * stride, stride, &q0, &q1, &t1, &t2); // q0, q1, q2, q3 in HFilter16iSSE2()
775 MAX_DIFF2(t2, t1, q1, q0, mask); in HFilter16iSSE2()
777 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter16iSSE2()
778 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in HFilter16iSSE2()
781 Store16x4(b, b + 8 * stride, stride, &p1, &p0, &q0, &q1); in HFilter16iSSE2()
791 __m128i t1, p2, p1, p0, q0, q1, q2; in VFilter8SSE2() local
798 LOADUV_H_EDGES4(u, v, stride, q0, q1, q2, t1); in VFilter8SSE2()
799 MAX_DIFF2(t1, q2, q1, q0, mask); in VFilter8SSE2()
801 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter8SSE2()
802 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in VFilter8SSE2()
809 STOREUV(q1, u, v, 1 * stride); in VFilter8SSE2()
816 __m128i p3, p2, p1, p0, q0, q1, q2, q3; in HFilter8SSE2() local
823 Load16x4(u, v, stride, &q0, &q1, &q2, &q3); // q0, q1, q2, q3 in HFilter8SSE2()
824 MAX_DIFF2(q3, q2, q1, q0, mask); in HFilter8SSE2()
826 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter8SSE2()
827 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in HFilter8SSE2()
830 Store16x4(u, v, stride, &q0, &q1, &q2, &q3); in HFilter8SSE2()
836 __m128i t1, t2, p1, p0, q0, q1; in VFilter8iSSE2() local
846 LOADUV_H_EDGES4(u, v, stride, q0, q1, t1, t2); in VFilter8iSSE2()
847 MAX_DIFF2(t2, t1, q1, q0, mask); in VFilter8iSSE2()
849 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter8iSSE2()
850 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in VFilter8iSSE2()
856 STOREUV(q1, u, v, 1 * stride); in VFilter8iSSE2()
862 __m128i t1, t2, p1, p0, q0, q1; in HFilter8iSSE2() local
868 Load16x4(u, v, stride, &q0, &q1, &t1, &t2); // q0, q1, q2, q3 in HFilter8iSSE2()
869 MAX_DIFF2(t2, t1, q1, q0, mask); in HFilter8iSSE2()
871 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter8iSSE2()
872 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in HFilter8iSSE2()
876 Store16x4(u, v, stride, &p1, &p0, &q0, &q1); in HFilter8iSSE2()