Lines Matching refs:q1
281 #define GET_NOTHEV(p1, p0, q0, q1, hev_thresh, not_hev) { \ argument
284 const __m128i t_2 = MM_ABS(q1, q0); \
294 #define GET_BASE_DELTA(p1, p0, q0, q1, o) { \ argument
296 o = _mm_subs_epi8(p1, q1); /* p1 - q1 */ \
329 const __m128i* q1, int thresh, __m128i *mask) { in NeedsFilter() argument
330 __m128i t1 = MM_ABS(*p1, *q1); // abs(p1 - q1) in NeedsFilter()
349 const __m128i* q1, int thresh) { in DoFilter2() argument
353 const __m128i q1s = _mm_xor_si128(*q1, sign_bit); in DoFilter2()
355 NeedsFilter(p1, p0, q0, q1, thresh, &mask); in DoFilter2()
370 __m128i* q0, __m128i* q1, in DoFilter4() argument
377 GET_NOTHEV(*p1, *p0, *q0, *q1, hev_thresh, not_hev); in DoFilter4()
380 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter4()
382 t1 = _mm_subs_epi8(*p1, *q1); // p1 - q1 in DoFilter4()
408 *q1 = _mm_subs_epi8(*q1, t3); // q1 -= t3 in DoFilter4()
412 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter4()
417 __m128i* q0, __m128i* q1, __m128i *q2, in DoFilter6() argument
423 GET_NOTHEV(*p1, *p0, *q0, *q1, hev_thresh, not_hev); in DoFilter6()
426 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter6()
429 GET_BASE_DELTA(*p1, *p0, *q0, *q1, a); in DoFilter6()
461 UPDATE_2PIXELS(*p1, *q1, a1_lo, a1_hi); in DoFilter6()
466 FLIP_SIGN_BIT4(*p1, *p0, *q0, *q1); in DoFilter6()
517 __m128i* q0, __m128i* q1) { in Load16x4() argument
535 Load8x4(r8, stride, p0, q1); in Load16x4()
545 *q0 = _mm_unpacklo_epi64(t2, *q1); in Load16x4()
546 *q1 = _mm_unpackhi_epi64(t2, *q1); in Load16x4()
560 __m128i* q0, __m128i* q1) { in Store16x4() argument
572 *q0 = _mm_unpacklo_epi8(t1, *q1); in Store16x4()
573 *q1 = _mm_unpackhi_epi8(t1, *q1); in Store16x4()
584 *p1 = _mm_unpacklo_epi16(t1, *q1); in Store16x4()
585 *q1 = _mm_unpackhi_epi16(t1, *q1); in Store16x4()
593 Store4x4(q1, r8, stride); in Store16x4()
604 __m128i q1 = _mm_loadu_si128((__m128i*)&p[stride]); in SimpleVFilter16SSE2() local
606 DoFilter2(&p1, &p0, &q0, &q1, thresh); in SimpleVFilter16SSE2()
614 __m128i p1, p0, q0, q1; in SimpleHFilter16SSE2() local
618 Load16x4(p, p + 8 * stride, stride, &p1, &p0, &q0, &q1); in SimpleHFilter16SSE2()
619 DoFilter2(&p1, &p0, &q0, &q1, thresh); in SimpleHFilter16SSE2()
620 Store16x4(p, p + 8 * stride, stride, &p1, &p0, &q0, &q1); in SimpleHFilter16SSE2()
679 #define COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask) { \ argument
684 NeedsFilter(&p1, &p0, &q0, &q1, thresh, &fl_yes); \
693 __m128i p2, p1, p0, q0, q1, q2; in VFilter16SSE2() local
700 LOAD_H_EDGES4(p, stride, q0, q1, q2, t1); in VFilter16SSE2()
701 MAX_DIFF2(t1, q2, q1, q0, mask); in VFilter16SSE2()
703 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter16SSE2()
704 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in VFilter16SSE2()
711 _mm_storeu_si128((__m128i*)&p[1 * stride], q1); in VFilter16SSE2()
718 __m128i p3, p2, p1, p0, q0, q1, q2, q3; in HFilter16SSE2() local
724 Load16x4(p, p + 8 * stride, stride, &q0, &q1, &q2, &q3); // q0, q1, q2, q3 in HFilter16SSE2()
725 MAX_DIFF2(q3, q2, q1, q0, mask); in HFilter16SSE2()
727 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter16SSE2()
728 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in HFilter16SSE2()
731 Store16x4(p, p + 8 * stride, stride, &q0, &q1, &q2, &q3); in HFilter16SSE2()
739 __m128i t1, t2, p1, p0, q0, q1; in VFilter16iSSE2() local
749 LOAD_H_EDGES4(p, stride, q0, q1, t1, t2); in VFilter16iSSE2()
750 MAX_DIFF2(t2, t1, q1, q0, mask); in VFilter16iSSE2()
752 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter16iSSE2()
753 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in VFilter16iSSE2()
759 _mm_storeu_si128((__m128i*)&p[1 * stride], q1); in VFilter16iSSE2()
768 __m128i t1, t2, p1, p0, q0, q1; in HFilter16iSSE2() local
776 Load16x4(b, b + 8 * stride, stride, &q0, &q1, &t1, &t2); // q0, q1, q2, q3 in HFilter16iSSE2()
777 MAX_DIFF2(t2, t1, q1, q0, mask); in HFilter16iSSE2()
779 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter16iSSE2()
780 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in HFilter16iSSE2()
783 Store16x4(b, b + 8 * stride, stride, &p1, &p0, &q0, &q1); in HFilter16iSSE2()
793 __m128i t1, p2, p1, p0, q0, q1, q2; in VFilter8SSE2() local
800 LOADUV_H_EDGES4(u, v, stride, q0, q1, q2, t1); in VFilter8SSE2()
801 MAX_DIFF2(t1, q2, q1, q0, mask); in VFilter8SSE2()
803 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter8SSE2()
804 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in VFilter8SSE2()
811 STOREUV(q1, u, v, 1 * stride); in VFilter8SSE2()
818 __m128i p3, p2, p1, p0, q0, q1, q2, q3; in HFilter8SSE2() local
825 Load16x4(u, v, stride, &q0, &q1, &q2, &q3); // q0, q1, q2, q3 in HFilter8SSE2()
826 MAX_DIFF2(q3, q2, q1, q0, mask); in HFilter8SSE2()
828 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter8SSE2()
829 DoFilter6(&p2, &p1, &p0, &q0, &q1, &q2, &mask, hev_thresh); in HFilter8SSE2()
832 Store16x4(u, v, stride, &q0, &q1, &q2, &q3); in HFilter8SSE2()
838 __m128i t1, t2, p1, p0, q0, q1; in VFilter8iSSE2() local
848 LOADUV_H_EDGES4(u, v, stride, q0, q1, t1, t2); in VFilter8iSSE2()
849 MAX_DIFF2(t2, t1, q1, q0, mask); in VFilter8iSSE2()
851 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in VFilter8iSSE2()
852 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in VFilter8iSSE2()
858 STOREUV(q1, u, v, 1 * stride); in VFilter8iSSE2()
864 __m128i t1, t2, p1, p0, q0, q1; in HFilter8iSSE2() local
870 Load16x4(u, v, stride, &q0, &q1, &t1, &t2); // q0, q1, q2, q3 in HFilter8iSSE2()
871 MAX_DIFF2(t2, t1, q1, q0, mask); in HFilter8iSSE2()
873 COMPLEX_FL_MASK(p1, p0, q0, q1, thresh, ithresh, mask); in HFilter8iSSE2()
874 DoFilter4(&p1, &p0, &q0, &q1, &mask, hev_thresh); in HFilter8iSSE2()
878 Store16x4(u, v, stride, &p1, &p0, &q0, &q1); in HFilter8iSSE2()