Lines Matching refs:T1
48 __m128i T0, T1, T2, T3;
126 T1 = _mm_unpackhi_epi64(transpose1_0, transpose1_1);
143 // c = MUL(T1, K2) - MUL(T3, K1) = MUL(T1, k2) - MUL(T3, k1) + T1 - T3
144 const __m128i c1 = _mm_mulhi_epi16(T1, k2);
146 const __m128i c3 = _mm_sub_epi16(T1, T3);
149 // d = MUL(T1, K1) + MUL(T3, K2) = MUL(T1, k1) + MUL(T3, k2) + T1 + T3
150 const __m128i d1 = _mm_mulhi_epi16(T1, k1);
152 const __m128i d3 = _mm_add_epi16(T1, T3);
188 T1 = _mm_unpackhi_epi64(transpose1_0, transpose1_1);
222 dst1 = _mm_add_epi16(dst1, T1);
335 *not_hev = _mm_cmpeq_epi8(t_max_h, zero); // not_hev <= t1 && not_hev <= t2
390 const __m128i t1 = MM_ABS(*p1, *q1); // abs(p1 - q1)
392 const __m128i t2 = _mm_and_si128(t1, kFE); // set lsb of each byte to zero
435 __m128i t1, t2, t3;
443 t1 = _mm_subs_epi8(*p1, *q1); // p1 - q1
444 t1 = _mm_andnot_si128(not_hev, t1); // hev(p1 - q1)
446 t1 = _mm_adds_epi8(t1, t2); // hev(p1 - q1) + 1 * (q0 - p0)
447 t1 = _mm_adds_epi8(t1, t2); // hev(p1 - q1) + 2 * (q0 - p0)
448 t1 = _mm_adds_epi8(t1, t2); // hev(p1 - q1) + 3 * (q0 - p0)
449 t1 = _mm_and_si128(t1, *mask); // mask filter values we don't care about
451 t2 = _mm_adds_epi8(t1, k3); // 3 * (q0 - p0) + hev(p1 - q1) + 3
452 t3 = _mm_adds_epi8(t1, k4); // 3 * (q0 - p0) + hev(p1 - q1) + 4
576 const __m128i t1 = *p1;
578 *p1 = _mm_unpacklo_epi64(t1, *p0);
579 *p0 = _mm_unpackhi_epi64(t1, *p0);
600 __m128i t1, p1_s, p0_s, q0_s, q1_s;
604 t1 = *p0;
605 p0_s = _mm_unpacklo_epi8(*p1, t1);
606 p1_s = _mm_unpackhi_epi8(*p1, t1);
610 t1 = *q0;
611 q0_s = _mm_unpacklo_epi8(t1, *q1);
612 q1_s = _mm_unpackhi_epi8(t1, *q1);
616 t1 = p0_s;
617 p0_s = _mm_unpacklo_epi16(t1, q0_s);
618 q0_s = _mm_unpackhi_epi16(t1, q0_s);
622 t1 = p1_s;
623 p1_s = _mm_unpacklo_epi16(t1, q1_s);
624 q1_s = _mm_unpackhi_epi16(t1, q1_s);
736 __m128i t1;
741 LOAD_H_EDGES4(p - 4 * stride, stride, t1, p2, p1, p0);
742 MAX_DIFF1(t1, p2, p1, p0, mask);
745 LOAD_H_EDGES4(p, stride, q0, q1, q2, t1);
746 MAX_DIFF2(t1, q2, q1, q0, mask);
845 __m128i t1, p2, p1, p0, q0, q1, q2;
848 LOADUV_H_EDGES4(u - 4 * stride, v - 4 * stride, stride, t1, p2, p1, p0);
849 MAX_DIFF1(t1, p2, p1, p0, mask);
852 LOADUV_H_EDGES4(u, v, stride, q0, q1, q2, t1);
853 MAX_DIFF2(t1, q2, q1, q0, mask);
890 __m128i t1, t2, p1, p0, q0, q1;
893 LOADUV_H_EDGES4(u, v, stride, t2, t1, p1, p0);
894 MAX_DIFF1(t2, t1, p1, p0, mask);
900 LOADUV_H_EDGES4(u, v, stride, q0, q1, t1, t2);
901 MAX_DIFF2(t2, t1, q1, q0, mask);
916 __m128i t1, t2, p1, p0, q0, q1;
917 Load16x4(u, v, stride, &t2, &t1, &p1, &p0); // p3, p2, p1, p0
918 MAX_DIFF1(t2, t1, p1, p0, mask);
922 Load16x4(u, v, stride, &q0, &q1, &t1, &t2); // q0, q1, q2, q3
923 MAX_DIFF2(t2, t1, q1, q0, mask);
