Lines Matching refs:v6
33 ld1 {v5.16b, v6.16b}, [x0], #32
42 shl v20.8h, v6.8h, #4 // 16 * src[40]
43 shl v21.8h, v6.8h, #2 // 4 * src[40]
62 mla v7.8h, v6.8h, v0.h[1] // t1 = 16 * src[8] + 15 * src[24] + 9 * src[40] + 4 * src[56]
71 mls v18.8h, v6.8h, v0.h[2] // -t4 = - 4 * src[8] + 9 * src[24] - 15 * src[40] + 16 * src[56]
76 neg v6.8h, v19.8h // +t3
81 ssra v16.8h, v6.8h, #1 // (t7 + t3) >> 1
89 srshr v6.8h, v17.8h, #2 // (t8 + t4 + 4) >> 3
95 trn2 v18.8h, v5.8h, v6.8h
105 trn1 v5.8h, v5.8h, v6.8h
108 trn1 v6.2d, v21.2d, v18.2d
115 shl v22.8h, v6.8h, #2 // 4 * src[8]
119 shl v2.8h, v6.8h, #4 // 16 * src[8]
127 mls v17.8h, v6.8h, v0.h[2] // - 15 * src[8] + 4 * src[24] + 16 * src[40]
130 mls v5.8h, v6.8h, v0.h[1] // - 9 * src[8] + 16 * src[24] - 4 * src[40]
131 shl v6.8h, v19.8h, #3 // 16/2 * src[16]
140 mla v6.8h, v1.8h, v0.h[0] // t3/2 = 16/2 * src[16] + 6/2 * src[48]
146 add v18.8h, v20.8h, v6.8h // t5/2 = t1/2 + t3/2
151 sub v0.8h, v20.8h, v6.8h // t8/2 = t1/2 - t3/2
154 sub v23.8h, v20.8h, v6.8h // t8/2 = t1/2 - t3/2
156 add v6.8h, v20.8h, v6.8h // t5/2 = t1/2 + t3/2
166 srsra v6.8h, v3.8h, #1 // (t5 - t1 + 1) >> 1
175 srshr v1.8h, v6.8h, #6 // (t5 - t1 + 65) >> 7
195 trn2 v6.4h, v1.4h, v3.4h
204 trn1 v19.2s, v6.2s, v3.2s
205 trn2 v3.2s, v6.2s, v3.2s
206 trn1 v6.2s, v7.2s, v4.2s
214 shl v17.4h, v6.4h, #4 // 16 * src[5]
220 shl v27.4h, v6.4h, #2 // 4 * src[5]
236 mla v21.4h, v6.4h, v0.h[1] // t1 = 16 * src[1] + 15 * src[3] + 9 * src[5] + 4 * src[7]
241 mls v23.4h, v6.4h, v0.h[2] // -t4 = - 4 * src[1] + 9 * src[3] - 15 * src[5] + 16 * src[7]
243 neg v6.4h, v21.4h // -t1
262 ssra v26.4h, v6.4h, #1 // (t5 - t1) >> 1
271 trn2 v6.8h, v1.8h, v2.8h
275 trn2 v4.4s, v6.4s, v2.4s
279 trn1 v2.4s, v6.4s, v2.4s
281 mul v6.8h, v7.8h, v0.h[6] // 17 * src[0]
285 add v0.8h, v6.8h, v1.8h // t1 = 17 * src[0] + 17 * src[16]
286 sub v1.8h, v6.8h, v1.8h // t2 = 17 * src[0] - 17 * src[16]
288 neg v6.8h, v4.8h // -t3/2
292 ssra v6.8h, v0.8h, #1 // (t1 - t3) >> 1
296 srshr v3.8h, v6.8h, #6 // (t1 - t3 + 64) >> 7
332 ld1 {v6.s}[0], [x0], x1
345 ld1 {v6.s}[1], [x0], x1
435 uaddw v2.8h, v2.8h, v6.8b
470 ld1 {v6.s}[0], [x0], x1
473 ld1 {v6.s}[1], [x0]
525 uaddw v1.8h, v1.8h, v6.8b
554 ld1 {v6.8b}, [x0], x1
567 uaddw v6.8h, v16.8h, v6.8b
576 sqxtun v1.8b, v6.8h
720 ld1 {v6.s}[0], [x0], x1 // P7
730 uxtl v3.8h, v6.8b // P7
733 uxtl v6.8h, v7.8b // P4
738 mls v5.4h, v6.4h, v0.h[1] // 2*P3-5*P4
740 sub v3.4h, v6.4h, v1.4h // P4-P5
741 mls v2.4h, v6.4h, v0.h[0] // 2*P1-5*P2+5*P3-2*P4
770 mls v6.4h, v0.4h, v3.4h // invert d depending on clip_sign & a0_sign, or zero it if they match, and accumulate into P4
772 sqxtun v0.8b, v6.8h
793 trn1 v6.8b, v1.8b, v2.8b
797 trn1 v4.4h, v6.4h, v2.4h // P1, P5
799 trn2 v2.4h, v6.4h, v2.4h // P3, P7
802 uxtl v6.8h, v7.8b // P2, P6
805 mls v3.8h, v6.8h, v0.h[1] // 2*P1-5*P2, 2*P5-5*P6
809 mov d6, v6.d[1] // P6
815 mls v2.4h, v6.4h, v0.h[0] // 2*P3-5*P4+5*P5-2*P6
817 abs v6.4h, v7.4h
821 sshr v6.4h, v6.4h, #1 // clip
824 cmeq v18.4h, v6.4h, #0 // test clip == 0
837 cmhs v5.4h, v0.4h, v6.4h
839 bsl v5.8b, v6.8b, v0.8b // FFMIN(d, clip)
865 ld1 {v6.8b}, [x3], x1 // P3
874 uxtl v20.8h, v6.8b // P3
877 ushll v6.8h, v6.8b, #1 // 2*P3
885 mls v6.8h, v7.8h, v0.h[1] // 2*P3-5*P4
891 mla v6.8h, v1.8h, v0.h[1] // 2*P3-5*P4+5*P5
893 mls v6.8h, v5.8h, v0.h[0] // 2*P3-5*P4+5*P5-2*P6
898 srshr v6.8h, v6.8h, #3
900 abs v20.8h, v6.8h // a0
901 sshr v6.8h, v6.8h, #8 // a0_sign
904 sub v4.8h, v4.8h, v6.8h // clip_sign - a0_sign
905 uqsub v6.8h, v20.8h, v18.8h // a0 >= a3 ? a0-a3 : 0 (a0 > a3 in all cases where filtering is enabled, so makes more sense to subtract this way round than the opposite and then taking the abs)
908 mul v0.8h, v6.8h, v0.h[1] // a0 >= a3 ? 5*(a0-a3) : 0
944 ld1 {v6.8b}, [x3], x1
952 trn1 v18.8b, v5.8b, v6.8b // P1[4], P1[5], P3[4]...
953 trn2 v5.8b, v5.8b, v6.8b // P2[4], P2[5], P4[4]...
954 trn1 v6.4h, v16.4h, v2.4h // P1[0], P1[1], P1[2], P1[3], P5[0]...
964 trn1 v7.2s, v6.2s, v3.2s // P1
966 trn2 v3.2s, v6.2s, v3.2s // P5
967 trn2 v6.2s, v19.2s, v16.2s // P6
975 uxtl v6.8h, v6.8b // P6
980 mls v19.8h, v6.8h, v0.h[1] // 2*P5-5*P6
995 mls v5.8h, v6.8h, v0.h[0] // 2*P3-5*P4+5*P5-2*P6
996 cmeq v6.8h, v1.8h, #0 // test clip == 0
1009 orr v4.16b, v6.16b, v4.16b // test clip == 0 || a0 >= pq
1050 ld1 {v6.16b}, [x3], x1 // P3
1065 uxtl v23.8h, v6.8b // P3[0..7]
1068 ushll v4.8h, v6.8b, #1 // 2*P3[0..7]
1071 uxtl2 v26.8h, v6.16b // P3[8..15]
1074 ushll2 v6.8h, v6.16b, #1 // 2*P3[8..15]
1086 mls v6.8h, v18.8h, v0.h[1] // 2*P3[8..15]-5*P4[8..15]
1103 mla v6.8h, v1.8h, v0.h[1] // 2*P3[8..15]-5*P4[8..15]+5*P5[8..15]
1108 mls v6.8h, v5.8h, v0.h[0] // 2*P3[8..15]-5*P4[8..15]+5*P5[8..15]-2*P6[8..15]
1114 srshr v6.8h, v6.8h, #3
1120 abs v19.8h, v6.8h // a0[8..15]
1124 sshr v6.8h, v6.8h, #8 // a0_sign[8..15]
1131 sub v6.8h, v7.8h, v6.8h // clip_sign[8..15] - a0_sign[8..15]
1156 mls v18.8h, v0.8h, v6.8h // invert d[8..15] depending on clip_sign[8..15] & a0_sign[8..15], or zero it if they match, and accumulate into P4[8..15]
1159 mla v1.8h, v0.8h, v6.8h // invert d[8..15] depending on clip_sign[8..15] & a0_sign[8..15], or zero it if they match, and accumulate into P5[8..15]
1185 ld1 {v6.8b}, [x3], x1
1195 trn1 v20.8b, v5.8b, v6.8b // P1[4], P1[5], P3[4]...
1197 trn2 v5.8b, v5.8b, v6.8b // P2[4], P2[5], P4[4]...
1198 ld1 {v6.8b}, [x3], x1
1211 trn1 v28.8b, v6.8b, v23.8b // P1[12], P1[13], P3[12]...
1212 trn2 v6.8b, v6.8b, v23.8b // P2[12], P2[13], P4[12]...
1224 trn1 v18.4h, v6.4h, v17.4h // P2[12], P2[13], P2[14], P2[15], P6[12]...
1235 trn2 v4.4h, v6.4h, v17.4h // P4[12], P4[13], P4[14], P4[15], P8[12]...
1237 ushll v6.8h, v19.8b, #1 // 2*P5[0..7]
1250 mls v6.8h, v20.8h, v0.h[1] // 2*P5[0..7]-5*P6[0..7]
1266 mla v6.8h, v16.8h, v0.h[1] // 2*P5[0..7]-5*P6[0..7]+5*P7[0..7]
1283 mls v6.8h, v1.8h, v0.h[0] // 2*P5[0..7]-5*P6[0..7]+5*P7[0..7]-2*P8[0..7]
1290 srshr v2.8h, v6.8h, #3
1292 srshr v6.8h, v24.8h, #3
1298 abs v6.8h, v6.8h // a1[8..15]
1303 cmhs v24.8h, v6.8h, v17.8h // test a1[8..15] >= a2[8.15]
1308 bsl v24.16b, v17.16b, v6.16b // a3[8..15]
1310 sshr v6.8h, v7.8h, #8 // a0_sign[0..7]
1318 sub v6.8h, v21.8h, v6.8h // clip_sign[0..7] - a0_sign[0..7]
1342 mls v27.8h, v2.8h, v6.8h // invert d[0..7] depending on clip_sign[0..7] & a0_sign[0..7], or zero it if they match, and accumulate into P4
1343 mla v16.8h, v2.8h, v6.8h // invert d[0..7] depending on clip_sign[0..7] & a0_sign[0..7], or zero it if they match, and accumulate into P5