Lines Matching refs:vec0
378 v8i16 vec0, vec1, vec2; in I422ToARGBRow_MSA() local
392 vec0, vec1, vec2); in I422ToARGBRow_MSA()
393 STOREARGB(vec0, vec1, vec2, alpha, dst_argb); in I422ToARGBRow_MSA()
409 v8i16 vec0, vec1, vec2; in I422ToRGBARow_MSA() local
423 vec0, vec1, vec2); in I422ToRGBARow_MSA()
424 STOREARGB(alpha, vec0, vec1, vec2, dst_argb); in I422ToRGBARow_MSA()
442 v8i16 vec0, vec1, vec2; in I422AlphaToARGBRow_MSA() local
458 vec0, vec1, vec2); in I422AlphaToARGBRow_MSA()
460 STOREARGB(vec0, vec1, vec2, src3, dst_argb); in I422AlphaToARGBRow_MSA()
478 v8i16 vec0, vec1, vec2, vec3, vec4, vec5; in I422ToRGB24Row_MSA() local
503 vec0, vec1, vec2); in I422ToRGB24Row_MSA()
506 reg0 = (v16u8)__msa_ilvev_b((v16i8)vec1, (v16i8)vec0); in I422ToRGB24Row_MSA()
531 v8i16 vec0, vec1, vec2; in I422ToRGB565Row_MSA() local
544 vec0, vec2, vec1); in I422ToRGB565Row_MSA()
545 vec0 = __msa_srai_h(vec0, 3); in I422ToRGB565Row_MSA()
550 vec0 |= vec1; in I422ToRGB565Row_MSA()
551 dst0 = (v16u8)(vec2 | vec0); in I422ToRGB565Row_MSA()
569 v8i16 vec0, vec1, vec2; in I422ToARGB4444Row_MSA() local
584 vec0, vec1, vec2); in I422ToARGB4444Row_MSA()
585 reg0 = (v8u16)__msa_srai_h(vec0, 4); in I422ToARGB4444Row_MSA()
609 v8i16 vec0, vec1, vec2; in I422ToARGB1555Row_MSA() local
624 vec0, vec1, vec2); in I422ToARGB1555Row_MSA()
625 reg0 = (v8u16)__msa_srai_h(vec0, 3); in I422ToARGB1555Row_MSA()
663 v16u8 vec0, vec1, dst0, dst1; in YUY2ToUVRow_MSA() local
672 vec0 = __msa_aver_u_b(src0, src2); in YUY2ToUVRow_MSA()
674 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in YUY2ToUVRow_MSA()
675 dst1 = (v16u8)__msa_pckod_b((v16i8)vec1, (v16i8)vec0); in YUY2ToUVRow_MSA()
728 v16u8 vec0, vec1, dst0, dst1; in UYVYToUVRow_MSA() local
737 vec0 = __msa_aver_u_b(src0, src2); in UYVYToUVRow_MSA()
739 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in UYVYToUVRow_MSA()
740 dst1 = (v16u8)__msa_pckod_b((v16i8)vec1, (v16i8)vec0); in UYVYToUVRow_MSA()
773 v16u8 src0, src1, src2, src3, vec0, vec1, vec2, vec3, dst0; in ARGBToYRow_MSA() local
786 vec0 = (v16u8)__msa_pckev_b((v16i8)src1, (v16i8)src0); in ARGBToYRow_MSA()
790 reg0 = (v8u16)__msa_ilvev_b(zero, (v16i8)vec0); in ARGBToYRow_MSA()
794 reg4 = (v8u16)__msa_ilvod_b(zero, (v16i8)vec0); in ARGBToYRow_MSA()
825 v16u8 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7, vec8, vec9; in ARGBToUVRow_MSA() local
844 vec0 = (v16u8)__msa_pckev_b((v16i8)src1, (v16i8)src0); in ARGBToUVRow_MSA()
852 vec8 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBToUVRow_MSA()
856 vec0 = (v16u8)__msa_pckod_b((v16i8)vec1, (v16i8)vec0); in ARGBToUVRow_MSA()
862 reg4 = __msa_hadd_u_h(vec0, vec0); in ARGBToUVRow_MSA()
872 vec0 = (v16u8)__msa_pckev_b((v16i8)src1, (v16i8)src0); in ARGBToUVRow_MSA()
880 vec8 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBToUVRow_MSA()
884 vec0 = (v16u8)__msa_pckod_b((v16i8)vec1, (v16i8)vec0); in ARGBToUVRow_MSA()
890 reg4 += __msa_hadd_u_h(vec0, vec0); in ARGBToUVRow_MSA()
986 v16u8 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in ARGBToRGB565Row_MSA() local
992 vec0 = (v16u8)__msa_srai_b((v16i8)src0, 3); in ARGBToRGB565Row_MSA()
1004 vec0 = __msa_binsli_b(vec0, vec1, 2); in ARGBToRGB565Row_MSA()
1008 vec0 = (v16u8)__msa_ilvev_b((v16i8)vec1, (v16i8)vec0); in ARGBToRGB565Row_MSA()
1010 dst0 = (v16u8)__msa_pckev_h((v8i16)vec4, (v8i16)vec0); in ARGBToRGB565Row_MSA()
1022 v16u8 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7, vec8, vec9; in ARGBToARGB1555Row_MSA() local
1028 vec0 = (v16u8)__msa_srai_b((v16i8)src0, 3); in ARGBToARGB1555Row_MSA()
1030 vec2 = (v16u8)__msa_srai_b((v16i8)vec0, 3); in ARGBToARGB1555Row_MSA()
1044 vec0 = __msa_binsli_b(vec0, vec1, 2); in ARGBToARGB1555Row_MSA()
1050 vec0 = (v16u8)__msa_ilvev_b((v16i8)vec1, (v16i8)vec0); in ARGBToARGB1555Row_MSA()
1052 dst0 = (v16u8)__msa_pckev_h((v8i16)vec1, (v8i16)vec0); in ARGBToARGB1555Row_MSA()
1064 v16u8 vec0, vec1; in ARGBToARGB4444Row_MSA() local
1071 vec0 = (v16u8)__msa_srai_b((v16i8)src0, 4); in ARGBToARGB4444Row_MSA()
1075 vec0 = __msa_binsli_b(vec0, src0, 3); in ARGBToARGB4444Row_MSA()
1077 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBToARGB4444Row_MSA()
1090 v8u16 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in ARGBToUV444Row_MSA() local
1112 vec0 = (v8u16)__msa_ilvr_b(zero, (v16i8)src0); in ARGBToUV444Row_MSA()
1118 vec10 = vec0 * const_18; in ARGBToUV444Row_MSA()
1124 vec0 *= const_112; in ARGBToUV444Row_MSA()
1134 vec0 += const_32896; in ARGBToUV444Row_MSA()
1138 vec0 -= vec2; in ARGBToUV444Row_MSA()
1142 vec0 = (v8u16)__msa_srai_h((v8i16)vec0, 8); in ARGBToUV444Row_MSA()
1146 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBToUV444Row_MSA()
1162 v8u16 vec0, vec1, vec2, vec3; in ARGBMultiplyRow_MSA() local
1169 vec0 = (v8u16)__msa_ilvr_b((v16i8)src0, (v16i8)src0); in ARGBMultiplyRow_MSA()
1173 reg0 = (v4u32)__msa_ilvr_h(zero, (v8i16)vec0); in ARGBMultiplyRow_MSA()
1174 reg1 = (v4u32)__msa_ilvl_h(zero, (v8i16)vec0); in ARGBMultiplyRow_MSA()
1185 vec0 = (v8u16)__msa_pckev_h((v8i16)reg1, (v8i16)reg0); in ARGBMultiplyRow_MSA()
1187 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBMultiplyRow_MSA()
1242 v8u16 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7, vec8, vec9; in ARGBAttenuateRow_MSA() local
1250 vec0 = (v8u16)__msa_ilvr_b((v16i8)src0, (v16i8)src0); in ARGBAttenuateRow_MSA()
1254 vec4 = (v8u16)__msa_fill_h(vec0[3]); in ARGBAttenuateRow_MSA()
1255 vec5 = (v8u16)__msa_fill_h(vec0[7]); in ARGBAttenuateRow_MSA()
1274 reg0 *= (v4u32)__msa_ilvr_h(zero, (v8i16)vec0); in ARGBAttenuateRow_MSA()
1275 reg1 *= (v4u32)__msa_ilvl_h(zero, (v8i16)vec0); in ARGBAttenuateRow_MSA()
1290 vec0 = (v8u16)__msa_pckev_h((v8i16)reg1, (v8i16)reg0); in ARGBAttenuateRow_MSA()
1294 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBAttenuateRow_MSA()
1309 v16u8 src0, src1, dst0, vec0, vec1; in ARGBToRGB565DitherRow_MSA() local
1321 vec0 = (v16u8)__msa_pckev_b((v16i8)src1, (v16i8)src0); in ARGBToRGB565DitherRow_MSA()
1323 reg0 = (v8i16)__msa_ilvev_b(zero, (v16i8)vec0); in ARGBToRGB565DitherRow_MSA()
1325 reg2 = (v8i16)__msa_ilvod_b(zero, (v16i8)vec0); in ARGBToRGB565DitherRow_MSA()
1354 v16i8 vec0; in ARGBShuffleRow_MSA() local
1358 vec0 = (v16i8)__msa_fill_w(val); in ARGBShuffleRow_MSA()
1359 shuffler_vec += vec0; in ARGBShuffleRow_MSA()
1378 v8u16 vec0, vec1; in ARGBShadeRow_MSA() local
1388 vec0 = (v8u16)__msa_ilvr_b((v16i8)src0, (v16i8)src0); in ARGBShadeRow_MSA()
1390 reg0 = (v4u32)__msa_ilvr_h(zero, (v8i16)vec0); in ARGBShadeRow_MSA()
1391 reg1 = (v4u32)__msa_ilvl_h(zero, (v8i16)vec0); in ARGBShadeRow_MSA()
1402 vec0 = (v8u16)__msa_pckev_h((v8i16)reg1, (v8i16)reg0); in ARGBShadeRow_MSA()
1404 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBShadeRow_MSA()
1413 v16u8 src0, src1, vec0, vec1, dst0, dst1; in ARGBGrayRow_MSA() local
1421 vec0 = (v16u8)__msa_pckev_h((v8i16)src1, (v8i16)src0); in ARGBGrayRow_MSA()
1423 reg0 = __msa_dotp_u_h(vec0, const_0x4B0F); in ARGBGrayRow_MSA()
1426 vec0 = (v16u8)__msa_ilvev_b((v16i8)reg0, (v16i8)reg0); in ARGBGrayRow_MSA()
1427 vec1 = (v16u8)__msa_ilvod_b((v16i8)vec1, (v16i8)vec0); in ARGBGrayRow_MSA()
1428 dst0 = (v16u8)__msa_ilvr_b((v16i8)vec1, (v16i8)vec0); in ARGBGrayRow_MSA()
1429 dst1 = (v16u8)__msa_ilvl_b((v16i8)vec1, (v16i8)vec0); in ARGBGrayRow_MSA()
1438 v16u8 src0, src1, dst0, dst1, vec0, vec1, vec2, vec3, vec4, vec5; in ARGBSepiaRow_MSA() local
1451 vec0 = (v16u8)__msa_pckev_h((v8i16)src1, (v8i16)src0); in ARGBSepiaRow_MSA()
1454 reg0 = (v8u16)__msa_dotp_u_h(vec0, const_0x4411); in ARGBSepiaRow_MSA()
1455 reg1 = (v8u16)__msa_dotp_u_h(vec0, const_0x5816); in ARGBSepiaRow_MSA()
1456 reg2 = (v8u16)__msa_dotp_u_h(vec0, const_0x6218); in ARGBSepiaRow_MSA()
1465 vec0 = (v16u8)__msa_pckev_b((v16i8)reg0, (v16i8)reg0); in ARGBSepiaRow_MSA()
1468 vec4 = (v16u8)__msa_ilvr_b((v16i8)vec2, (v16i8)vec0); in ARGBSepiaRow_MSA()
1482 v8u16 vec0, vec1, vec2, vec3; in ARGB4444ToARGBRow_MSA() local
1488 vec0 = (v8u16)__msa_andi_b(src0, 0x0F); in ARGB4444ToARGBRow_MSA()
1492 vec0 |= (v8u16)__msa_slli_b((v16i8)vec0, 4); in ARGB4444ToARGBRow_MSA()
1496 dst0 = (v16u8)__msa_ilvr_b((v16i8)vec2, (v16i8)vec0); in ARGB4444ToARGBRow_MSA()
1497 dst1 = (v16u8)__msa_ilvl_b((v16i8)vec2, (v16i8)vec0); in ARGB4444ToARGBRow_MSA()
1511 v8u16 vec0, vec1, vec2, vec3, vec4, vec5; in ARGB1555ToARGBRow_MSA() local
1519 vec0 = src0 & const_0x1F; in ARGB1555ToARGBRow_MSA()
1531 reg0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGB1555ToARGBRow_MSA()
1560 v8u16 src0, src1, vec0, vec1, vec2, vec3, vec4, vec5; in RGB565ToARGBRow_MSA() local
1571 vec0 = src0 & const_0x1F; in RGB565ToARGBRow_MSA()
1577 reg0 = (v8u16)__msa_slli_h((v8i16)vec0, 3); in RGB565ToARGBRow_MSA()
1583 reg0 |= (v8u16)__msa_srli_h((v8i16)vec0, 2); in RGB565ToARGBRow_MSA()
1608 v16u8 vec0, vec1, vec2; in RGB24ToARGBRow_MSA() local
1617 vec0 = (v16u8)__msa_sldi_b((v16i8)src1, (v16i8)src0, 12); in RGB24ToARGBRow_MSA()
1621 dst1 = (v16u8)__msa_vshf_b(shuffler, (v16i8)alpha, (v16i8)vec0); in RGB24ToARGBRow_MSA()
1633 v16u8 vec0, vec1, vec2; in RAWToARGBRow_MSA() local
1642 vec0 = (v16u8)__msa_sldi_b((v16i8)src1, (v16i8)src0, 12); in RAWToARGBRow_MSA()
1646 dst1 = (v16u8)__msa_vshf_b(mask, (v16i8)alpha, (v16i8)vec0); in RAWToARGBRow_MSA()
1659 v8u16 src0, src1, vec0, vec1, vec2, vec3, vec4, vec5; in ARGB1555ToYRow_MSA() local
1671 vec0 = src0 & const_0x1F; in ARGB1555ToYRow_MSA()
1681 reg0 = (v8u16)__msa_slli_h((v8i16)vec0, 3); in ARGB1555ToYRow_MSA()
1683 reg0 |= (v8u16)__msa_srai_h((v8i16)vec0, 2); in ARGB1555ToYRow_MSA()
1716 v8u16 src0, src1, vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in RGB565ToYRow_MSA() local
1730 vec0 = src0 & const_0x1F; in RGB565ToYRow_MSA()
1736 reg0 = (v8u16)__msa_slli_h((v8i16)vec0, 3); in RGB565ToYRow_MSA()
1742 reg0 |= (v8u16)__msa_srli_h((v8i16)vec0, 2); in RGB565ToYRow_MSA()
1748 vec0 = (v8u16)__msa_ilvr_h((v8i16)reg1, (v8i16)reg0); in RGB565ToYRow_MSA()
1756 res0 = __msa_dotp_u_w(vec0, (v8u16)const_0x810019); in RGB565ToYRow_MSA()
1768 vec0 = (v8u16)__msa_pckev_h((v8i16)res1, (v8i16)res0); in RGB565ToYRow_MSA()
1770 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in RGB565ToYRow_MSA()
1780 v8u16 vec0, vec1, vec2, vec3; in RGB24ToYRow_MSA() local
1799 vec0 = (v8u16)__msa_pckev_h((v8i16)reg1, (v8i16)reg0); in RGB24ToYRow_MSA()
1803 vec0 = __msa_dotp_u_h((v16u8)vec0, (v16u8)const_0x8119); in RGB24ToYRow_MSA()
1805 vec0 = __msa_dpadd_u_h(vec0, (v16u8)vec2, (v16u8)const_0x42); in RGB24ToYRow_MSA()
1807 vec0 += const_0x1080; in RGB24ToYRow_MSA()
1809 vec0 = (v8u16)__msa_srai_h((v8i16)vec0, 8); in RGB24ToYRow_MSA()
1811 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in RGB24ToYRow_MSA()
1821 v8u16 vec0, vec1, vec2, vec3; in RAWToYRow_MSA() local
1840 vec0 = (v8u16)__msa_pckev_h((v8i16)reg1, (v8i16)reg0); in RAWToYRow_MSA()
1844 vec0 = __msa_dotp_u_h((v16u8)vec0, (v16u8)const_0x8142); in RAWToYRow_MSA()
1846 vec0 = __msa_dpadd_u_h(vec0, (v16u8)vec2, (v16u8)const_0x19); in RAWToYRow_MSA()
1848 vec0 += const_0x1080; in RAWToYRow_MSA()
1850 vec0 = (v8u16)__msa_srai_h((v8i16)vec0, 8); in RAWToYRow_MSA()
1852 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in RAWToYRow_MSA()
1869 v8u16 vec0, vec1, vec2, vec3, vec4, vec5, vec6; in ARGB1555ToUVRow_MSA() local
1884 vec0 = src0 & const_0x1F; in ARGB1555ToUVRow_MSA()
1886 vec0 += src2 & const_0x1F; in ARGB1555ToUVRow_MSA()
1888 vec0 = (v8u16)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGB1555ToUVRow_MSA()
1907 vec0 = __msa_hadd_u_h((v16u8)vec0, (v16u8)vec0); in ARGB1555ToUVRow_MSA()
1910 vec6 = (v8u16)__msa_slli_h((v8i16)vec0, 1); in ARGB1555ToUVRow_MSA()
1911 vec6 |= (v8u16)__msa_srai_h((v8i16)vec0, 6); in ARGB1555ToUVRow_MSA()
1912 vec0 = (v8u16)__msa_slli_h((v8i16)vec2, 1); in ARGB1555ToUVRow_MSA()
1913 vec0 |= (v8u16)__msa_srai_h((v8i16)vec2, 6); in ARGB1555ToUVRow_MSA()
1917 reg1 = vec0 * const_0x4A; in ARGB1555ToUVRow_MSA()
1919 reg3 = vec0 * const_0x5E; in ARGB1555ToUVRow_MSA()
1950 v8u16 vec0, vec1, vec2, vec3, vec4, vec5; in RGB565ToUVRow_MSA() local
1966 vec0 = src0 & const_0x1F; in RGB565ToUVRow_MSA()
1968 vec0 += src2 & const_0x1F; in RGB565ToUVRow_MSA()
1970 vec0 = (v8u16)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in RGB565ToUVRow_MSA()
1989 vec0 = __msa_hadd_u_h((v16u8)vec0, (v16u8)vec0); in RGB565ToUVRow_MSA()
1992 vec3 = (v8u16)__msa_slli_h((v8i16)vec0, 1); in RGB565ToUVRow_MSA()
1993 vec3 |= (v8u16)__msa_srai_h((v8i16)vec0, 6); in RGB565ToUVRow_MSA()
2031 v8u16 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in RGB24ToUVRow_MSA() local
2064 vec0 = (v8u16)__msa_ilvr_b((v16i8)src4, (v16i8)src0); in RGB24ToUVRow_MSA()
2072 vec0 = (v8u16)__msa_hadd_u_h((v16u8)vec0, (v16u8)vec0); in RGB24ToUVRow_MSA()
2080 reg0 = (v8i16)__msa_pckev_d((v2i64)vec1, (v2i64)vec0); in RGB24ToUVRow_MSA()
2084 reg0 += (v8i16)__msa_pckod_d((v2i64)vec1, (v2i64)vec0); in RGB24ToUVRow_MSA()
2096 vec0 = (v8u16)__msa_pckev_h((v8i16)vec5, (v8i16)vec4); in RGB24ToUVRow_MSA()
2099 vec3 = vec0 * const_0x70; in RGB24ToUVRow_MSA()
2104 vec0 *= const_0x12; in RGB24ToUVRow_MSA()
2108 reg3 = __msa_subv_h((v8i16)const_0x8080, (v8i16)vec0); in RGB24ToUVRow_MSA()
2136 v8u16 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in RAWToUVRow_MSA() local
2169 vec0 = (v8u16)__msa_ilvr_b((v16i8)src4, (v16i8)src0); in RAWToUVRow_MSA()
2177 vec0 = (v8u16)__msa_hadd_u_h((v16u8)vec0, (v16u8)vec0); in RAWToUVRow_MSA()
2185 reg0 = (v8i16)__msa_pckev_d((v2i64)vec1, (v2i64)vec0); in RAWToUVRow_MSA()
2189 reg0 += (v8i16)__msa_pckod_d((v2i64)vec1, (v2i64)vec0); in RAWToUVRow_MSA()
2201 vec0 = (v8u16)__msa_pckod_h((v8i16)vec5, (v8i16)vec4); in RAWToUVRow_MSA()
2204 vec3 = vec0 * const_0x70; in RAWToUVRow_MSA()
2209 vec0 *= const_0x12; in RAWToUVRow_MSA()
2213 reg3 = __msa_subv_h((v8i16)const_0x8080, (v8i16)vec0); in RAWToUVRow_MSA()
2238 v8i16 vec0, vec1, vec2; in NV12ToARGBRow_MSA() local
2255 vec0, vec1, vec2); in NV12ToARGBRow_MSA()
2256 res0 = (v16u8)__msa_ilvev_b((v16i8)vec2, (v16i8)vec0); in NV12ToARGBRow_MSA()
2275 v8i16 vec0, vec1, vec2; in NV12ToRGB565Row_MSA() local
2291 vec0, vec1, vec2); in NV12ToRGB565Row_MSA()
2292 vec0 = vec0 >> 3; in NV12ToRGB565Row_MSA()
2295 dst0 = (v16u8)(vec0 | vec1 | vec2); in NV12ToRGB565Row_MSA()
2311 v8i16 vec0, vec1, vec2; in NV21ToARGBRow_MSA() local
2330 vec0, vec1, vec2); in NV21ToARGBRow_MSA()
2331 res0 = (v16u8)__msa_ilvev_b((v16i8)vec2, (v16i8)vec0); in NV21ToARGBRow_MSA()
2347 v16u8 src0, src1, vec0, dst0, dst1, dst2, dst3; in SobelRow_MSA() local
2358 vec0 = __msa_adds_u_b(src0, src1); in SobelRow_MSA()
2359 dst0 = (v16u8)__msa_vshf_b(mask0, (v16i8)alpha, (v16i8)vec0); in SobelRow_MSA()
2360 dst1 = (v16u8)__msa_vshf_b(mask1, (v16i8)alpha, (v16i8)vec0); in SobelRow_MSA()
2361 dst2 = (v16u8)__msa_vshf_b(mask2, (v16i8)alpha, (v16i8)vec0); in SobelRow_MSA()
2362 dst3 = (v16u8)__msa_vshf_b(mask3, (v16i8)alpha, (v16i8)vec0); in SobelRow_MSA()
2396 v16u8 src0, src1, vec0, vec1, vec2; in SobelXYRow_MSA() local
2403 vec0 = __msa_adds_u_b(src0, src1); in SobelXYRow_MSA()
2406 reg0 = (v16u8)__msa_ilvr_b((v16i8)alpha, (v16i8)vec0); in SobelXYRow_MSA()
2407 reg1 = (v16u8)__msa_ilvl_b((v16i8)alpha, (v16i8)vec0); in SobelXYRow_MSA()
2508 v16u8 vec0, vec1, vec2, vec3; in ARGBToUVJRow_MSA() local
2537 vec0 = __msa_aver_u_b(src4, src6); in ARGBToUVJRow_MSA()
2557 ARGBTOUV(vec0, vec1, vec2, vec3, const_0x6B14, const_0x7F, const_0x2B54, in ARGBToUVJRow_MSA()
2577 v16u8 dst0, dst1, vec0, vec1, vec2, vec3; in BGRAToUVRow_MSA() local
2589 READ_ARGB(s, t, vec0, vec1, vec2, vec3); in BGRAToUVRow_MSA()
2590 ARGBTOUV(vec0, vec1, vec2, vec3, const_0x125E, const_0x7000, const_0x264A, in BGRAToUVRow_MSA()
2644 v16u8 dst0, dst1, vec0, vec1, vec2, vec3; in RGBAToUVRow_MSA() local
2656 READ_ARGB(s, t, vec0, vec1, vec2, vec3); in RGBAToUVRow_MSA()
2657 ARGBTOUV(vec0, vec1, vec2, vec3, const_0x125E, const_0x7000, const_0x264A, in RGBAToUVRow_MSA()
2677 v8u16 vec0, vec1, vec2; in I444ToARGBRow_MSA() local
2688 vec0 = (v8u16)__msa_ilvr_b((v16i8)src0, (v16i8)src0); in I444ToARGBRow_MSA()
2689 reg0 = (v4i32)__msa_ilvr_h((v8i16)zero, (v8i16)vec0); in I444ToARGBRow_MSA()
2690 reg1 = (v4i32)__msa_ilvl_h((v8i16)zero, (v8i16)vec0); in I444ToARGBRow_MSA()
2701 vec0 = (v8u16)__msa_ilvr_b((v16i8)zero, (v16i8)src1); in I444ToARGBRow_MSA()
2703 reg6 = (v4i32)__msa_ilvr_h((v8i16)zero, (v8i16)vec0); in I444ToARGBRow_MSA()
2704 reg7 = (v4i32)__msa_ilvl_h((v8i16)zero, (v8i16)vec0); in I444ToARGBRow_MSA()
2722 vec0 = (v8u16)__msa_pckev_h((v8i16)reg1, (v8i16)reg0); in I444ToARGBRow_MSA()
2725 vec0 = (v8u16)__msa_ilvev_b((v16i8)vec1, (v16i8)vec0); in I444ToARGBRow_MSA()
2727 dst0 = (v16u8)__msa_ilvr_h((v8i16)vec1, (v8i16)vec0); in I444ToARGBRow_MSA()
2728 dst1 = (v16u8)__msa_ilvl_h((v8i16)vec1, (v8i16)vec0); in I444ToARGBRow_MSA()
2740 v8i16 vec0, vec1; in I400ToARGBRow_MSA() local
2750 vec0 = (v8i16)__msa_ilvr_b((v16i8)src0, (v16i8)src0); in I400ToARGBRow_MSA()
2752 reg0 = (v4i32)__msa_ilvr_h(zero, vec0); in I400ToARGBRow_MSA()
2753 reg1 = (v4i32)__msa_ilvl_h(zero, vec0); in I400ToARGBRow_MSA()
2764 vec0 = (v8i16)__msa_pckev_h((v8i16)reg1, (v8i16)reg0); in I400ToARGBRow_MSA()
2766 vec0 += vec_ygb; in I400ToARGBRow_MSA()
2768 vec0 = __msa_srai_h(vec0, 6); in I400ToARGBRow_MSA()
2770 vec0 = __msa_maxi_s_h(vec0, 0); in I400ToARGBRow_MSA()
2772 vec0 = __msa_min_s_h(max, vec0); in I400ToARGBRow_MSA()
2774 res0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in I400ToARGBRow_MSA()
2791 v16u8 src0, vec0, vec1, vec2, vec3, dst0, dst1, dst2, dst3; in J400ToARGBRow_MSA() local
2796 vec0 = (v16u8)__msa_ilvr_b((v16i8)src0, (v16i8)src0); in J400ToARGBRow_MSA()
2800 dst0 = (v16u8)__msa_ilvr_b((v16i8)vec2, (v16i8)vec0); in J400ToARGBRow_MSA()
2801 dst1 = (v16u8)__msa_ilvl_b((v16i8)vec2, (v16i8)vec0); in J400ToARGBRow_MSA()
2816 v8i16 vec0, vec1, vec2; in YUY2ToARGBRow_MSA() local
2831 vec0, vec1, vec2); in YUY2ToARGBRow_MSA()
2832 STOREARGB(vec0, vec1, vec2, alpha, dst_argb); in YUY2ToARGBRow_MSA()
2844 v8i16 vec0, vec1, vec2; in UYVYToARGBRow_MSA() local
2859 vec0, vec1, vec2); in UYVYToARGBRow_MSA()
2860 STOREARGB(vec0, vec1, vec2, alpha, dst_argb); in UYVYToARGBRow_MSA()
2878 v8u16 vec0, vec1, vec2, vec3, y_frac; in InterpolateRow_MSA() local
2909 vec0 = (v8u16)__msa_ilvr_b((v16i8)src2, (v16i8)src0); in InterpolateRow_MSA()
2913 vec0 = (v8u16)__msa_dotp_u_h((v16u8)vec0, (v16u8)y_frac); in InterpolateRow_MSA()
2917 vec0 = (v8u16)__msa_srari_h((v8i16)vec0, 8); in InterpolateRow_MSA()
2921 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in InterpolateRow_MSA()
2988 v16u8 src0, src1, src2, src3, vec0, vec1, dst0; in ARGBExtractAlphaRow_MSA() local
2995 vec0 = (v16u8)__msa_pckod_b((v16i8)src1, (v16i8)src0); in ARGBExtractAlphaRow_MSA()
2997 dst0 = (v16u8)__msa_pckod_b((v16i8)vec1, (v16i8)vec0); in ARGBExtractAlphaRow_MSA()
3010 v8u16 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in ARGBBlendRow_MSA() local
3022 vec0 = (v8u16)__msa_ilvr_b(zero, (v16i8)src0); in ARGBBlendRow_MSA()
3030 vec8 = (v8u16)__msa_fill_h(vec0[3]); in ARGBBlendRow_MSA()
3031 vec9 = (v8u16)__msa_fill_h(vec0[7]); in ARGBBlendRow_MSA()
3054 vec0 += vec8; in ARGBBlendRow_MSA()
3058 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBBlendRow_MSA()
3076 v8i16 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in ARGBQuantizeRow_MSA() local
3090 vec0 = (v8i16)__msa_ilvr_b(zero, (v16i8)src0); in ARGBQuantizeRow_MSA()
3098 tmp0 = (v4i32)__msa_ilvr_h((v8i16)zero, (v8i16)vec0); in ARGBQuantizeRow_MSA()
3099 tmp1 = (v4i32)__msa_ilvl_h((v8i16)zero, (v8i16)vec0); in ARGBQuantizeRow_MSA()
3146 vec0 = (v8i16)__msa_pckev_h((v8i16)tmp1, (v8i16)tmp0); in ARGBQuantizeRow_MSA()
3154 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in ARGBQuantizeRow_MSA()
3182 v8i16 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7, vec8, vec9; in ARGBColorMatrixRow_MSA() local
3190 vec0 = (v8i16)__msa_ilvr_b(zero, src0); in ARGBColorMatrixRow_MSA()
3208 vec10 = vec2 * vec0; in ARGBColorMatrixRow_MSA()
3210 vec12 = vec6 * vec0; in ARGBColorMatrixRow_MSA()
3216 vec14 = vec3 * vec0; in ARGBColorMatrixRow_MSA()
3218 vec16 = vec7 * vec0; in ARGBColorMatrixRow_MSA()
3236 vec2 = vec4 * vec0; in ARGBColorMatrixRow_MSA()
3238 vec3 = vec8 * vec0; in ARGBColorMatrixRow_MSA()
3244 vec4 = vec5 * vec0; in ARGBColorMatrixRow_MSA()
3246 vec5 = vec9 * vec0; in ARGBColorMatrixRow_MSA()
3354 v8i16 vec0, vec1, vec2, vec3, vec4, vec5; in SobelXRow_MSA() local
3368 vec0 = (v8i16)__msa_vshf_b(mask0, (v16i8)src1, (v16i8)src0); in SobelXRow_MSA()
3374 vec0 = (v8i16)__msa_hsub_u_h((v16u8)vec0, (v16u8)vec0); in SobelXRow_MSA()
3380 vec0 += vec2; in SobelXRow_MSA()
3384 vec0 += vec4; in SobelXRow_MSA()
3386 vec0 = __msa_add_a_h(zero, vec0); in SobelXRow_MSA()
3388 vec0 = __msa_maxi_s_h(vec0, 0); in SobelXRow_MSA()
3390 vec0 = __msa_min_s_h(max, vec0); in SobelXRow_MSA()
3392 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in SobelXRow_MSA()
3407 v8i16 vec0, vec1, vec2, vec3, vec4, vec5, vec6; in SobelYRow_MSA() local
3414 vec0 = (v8i16)__msa_ilvr_b((v16i8)zero, (v16i8)src0); in SobelYRow_MSA()
3418 vec0 -= vec2; in SobelYRow_MSA()
3422 vec2 = (v8i16)__msa_sldi_b((v16i8)vec1, (v16i8)vec0, 2); in SobelYRow_MSA()
3424 vec4 = (v8i16)__msa_sldi_b((v16i8)vec1, (v16i8)vec0, 4); in SobelYRow_MSA()
3426 vec0 += vec2; in SobelYRow_MSA()
3430 vec0 += vec4; in SobelYRow_MSA()
3432 vec0 = __msa_add_a_h(zero, vec0); in SobelYRow_MSA()
3434 vec0 = __msa_maxi_s_h(vec0, 0); in SobelYRow_MSA()
3436 vec0 = __msa_min_s_h(max, vec0); in SobelYRow_MSA()
3438 dst0 = (v16u8)__msa_pckev_b((v16i8)vec1, (v16i8)vec0); in SobelYRow_MSA()
3452 v4u32 vec0, vec1, vec2, vec3, vec4, vec5, vec6, vec7; in HalfFloatRow_MSA() local
3464 vec0 = (v4u32)__msa_ilvr_h(zero, (v8i16)src0); in HalfFloatRow_MSA()
3472 fvec0 = __msa_ffint_u_w(vec0); in HalfFloatRow_MSA()
3488 vec0 = ((v4u32)fvec0) >> 13; in HalfFloatRow_MSA()
3496 dst0 = (v8u16)__msa_pckev_h((v8i16)vec1, (v8i16)vec0); in HalfFloatRow_MSA()