421 #define ST_B8(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, pdst, stride) \  argument
424     ST_B4(RTYPE, in4, in5, in6, in7, (pdst) + 4 * stride, stride);         \
447 #define ST_H8(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, pdst, stride) \  argument
450     ST_H4(RTYPE, in4, in5, in6, in7, (pdst) + 4 * stride, stride);         \
614 #define AVER_UB4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
618     AVER_UB2(RTYPE, in4, in5, in6, in7, out2, out3)                         \
1130 #define ILVL_B4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1134     ILVL_B2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1187 #define ILVR_B4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1191     ILVR_B2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1198 #define ILVR_B8(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, in8, in9, in10, \  argument
1202     ILVR_B4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, out2,   \
1224 #define ILVR_H4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1228     ILVR_H2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1240 #define ILVR_W4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1244     ILVR_W2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1264 #define ILVR_D3(RTYPE, in0, in1, in2, in3, in4, in5, out0, out1, out2) \  argument
1267     out2 = (RTYPE)__msa_ilvr_d((v2i64)(in4), (v2i64)(in5));            \
1271 #define ILVR_D4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1275     ILVR_D2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1403 #define PCKEV_B4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1407     PCKEV_B2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1429 #define PCKEV_H4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1433     PCKEV_H2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1453 #define PCKEV_D4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1457     PCKEV_D2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1491 #define XORI_B7_128(RTYPE, in0, in1, in2, in3, in4, in5, in6) \  argument
1494     XORI_B3_128(RTYPE, in4, in5, in6);                        \
1507 #define AVE_SH4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1512     out2 = (RTYPE)__msa_ave_s_h((v8i16)in4, (v8i16)in5);                   \
1532 #define ADDS_SH4(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1536     ADDS_SH2(RTYPE, in4, in5, in6, in7, out2, out3);                        \
1666 #define MUL4(in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, out2, out3) \  argument
1669     MUL2(in4, in5, in6, in7, out2, out3);                                    \
1683 #define ADD4(in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, out2, out3) \  argument
1686     ADD2(in4, in5, in6, in7, out2, out3);                                    \
1700 #define SUB4(in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, out2, out3) \  argument
1704     out2 = in4 - in5;                                                        \
1776 #define BUTTERFLY_8(in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, out2, \  argument
1782     out3 = in3 + in4;                                                         \
1784     out4 = in3 - in4;                                                         \
1795 #define BUTTERFLY_16(in0, in1, in2, in3, in4, in5, in6, in7, in8, in9, in10,  \  argument
1804     out4 = in4 + in11;                                                        \
1812     out11 = in4 - in11;                                                       \
1824 #define TRANSPOSE8x8_UB(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0,   \  argument
1830     ILVR_B4_SB(in2, in0, in3, in1, in6, in4, in7, in5, tmp0_m, tmp1_m, tmp2_m, \
1847 #define TRANSPOSE16x8_UB_UB(in0, in1, in2, in3, in4, in5, in6, in7, in8, in9, \  argument
1856     ILVEV_D2_UB(in4, in12, in5, in13, out3, out2);                            \
1909 #define TRANSPOSE4X8_SH_SH(in0, in1, in2, in3, in4, in5, in6, in7, out0, out1, \  argument
1916     ILVR_H4_SH(in1, in0, in3, in2, in5, in4, in7, in6, tmp0_n, tmp1_n, tmp2_n, \
1952 #define TRANSPOSE8x8_H(RTYPE, in0, in1, in2, in3, in4, in5, in6, in7, out0, \  argument
1959     ILVR_H2_SH(in6, in4, in7, in5, s0_m, s1_m);                             \
1961     ILVL_H2_SH(in6, in4, in7, in5, s0_m, s1_m);                             \