- vec_u8 srcR1 = vec_ld(-2, src);
- vec_u8 srcR2 = vec_ld(14, src);
-
- switch (align) {
- default: {
- srcM2 = vec_perm(srcR1, srcR2, permM2);
- srcM1 = vec_perm(srcR1, srcR2, permM1);
- srcP0 = vec_perm(srcR1, srcR2, permP0);
- srcP1 = vec_perm(srcR1, srcR2, permP1);
- srcP2 = vec_perm(srcR1, srcR2, permP2);
- srcP3 = vec_perm(srcR1, srcR2, permP3);
- } break;
- case 11: {
- srcM2 = vec_perm(srcR1, srcR2, permM2);
- srcM1 = vec_perm(srcR1, srcR2, permM1);
- srcP0 = vec_perm(srcR1, srcR2, permP0);
- srcP1 = vec_perm(srcR1, srcR2, permP1);
- srcP2 = vec_perm(srcR1, srcR2, permP2);
- srcP3 = srcR2;
- } break;
- case 12: {
- vec_u8 srcR3 = vec_ld(30, src);
- srcM2 = vec_perm(srcR1, srcR2, permM2);
- srcM1 = vec_perm(srcR1, srcR2, permM1);
- srcP0 = vec_perm(srcR1, srcR2, permP0);
- srcP1 = vec_perm(srcR1, srcR2, permP1);
- srcP2 = srcR2;
- srcP3 = vec_perm(srcR2, srcR3, permP3);
- } break;
- case 13: {
- vec_u8 srcR3 = vec_ld(30, src);
- srcM2 = vec_perm(srcR1, srcR2, permM2);
- srcM1 = vec_perm(srcR1, srcR2, permM1);
- srcP0 = vec_perm(srcR1, srcR2, permP0);
- srcP1 = srcR2;
- srcP2 = vec_perm(srcR2, srcR3, permP2);
- srcP3 = vec_perm(srcR2, srcR3, permP3);
- } break;
- case 14: {
- vec_u8 srcR3 = vec_ld(30, src);
- srcM2 = vec_perm(srcR1, srcR2, permM2);
- srcM1 = vec_perm(srcR1, srcR2, permM1);
- srcP0 = srcR2;
- srcP1 = vec_perm(srcR2, srcR3, permP1);
- srcP2 = vec_perm(srcR2, srcR3, permP2);
- srcP3 = vec_perm(srcR2, srcR3, permP3);
- } break;
- case 15: {
- vec_u8 srcR3 = vec_ld(30, src);
- srcM2 = vec_perm(srcR1, srcR2, permM2);
- srcM1 = srcR2;
- srcP0 = vec_perm(srcR2, srcR3, permP0);
- srcP1 = vec_perm(srcR2, srcR3, permP1);
- srcP2 = vec_perm(srcR2, srcR3, permP2);
- srcP3 = vec_perm(srcR2, srcR3, permP3);
- } break;
- }
-
- srcP0A = (vec_s16) vec_mergeh(zero_u8v, srcP0);
- srcP0B = (vec_s16) vec_mergel(zero_u8v, srcP0);
- srcP1A = (vec_s16) vec_mergeh(zero_u8v, srcP1);
- srcP1B = (vec_s16) vec_mergel(zero_u8v, srcP1);
-
- srcP2A = (vec_s16) vec_mergeh(zero_u8v, srcP2);
- srcP2B = (vec_s16) vec_mergel(zero_u8v, srcP2);
- srcP3A = (vec_s16) vec_mergeh(zero_u8v, srcP3);
- srcP3B = (vec_s16) vec_mergel(zero_u8v, srcP3);
-
- srcM1A = (vec_s16) vec_mergeh(zero_u8v, srcM1);
- srcM1B = (vec_s16) vec_mergel(zero_u8v, srcM1);
- srcM2A = (vec_s16) vec_mergeh(zero_u8v, srcM2);
- srcM2B = (vec_s16) vec_mergel(zero_u8v, srcM2);
+ load_alignment(src, align, permM2, permM1, permP0, permP1, permP2, permP3);
+
+ srcP0A = (vec_s16) VEC_MERGEH(zero_u8v, srcP0);
+ srcP0B = (vec_s16) VEC_MERGEL(zero_u8v, srcP0);
+ srcP1A = (vec_s16) VEC_MERGEH(zero_u8v, srcP1);
+ srcP1B = (vec_s16) VEC_MERGEL(zero_u8v, srcP1);
+
+ srcP2A = (vec_s16) VEC_MERGEH(zero_u8v, srcP2);
+ srcP2B = (vec_s16) VEC_MERGEL(zero_u8v, srcP2);
+ srcP3A = (vec_s16) VEC_MERGEH(zero_u8v, srcP3);
+ srcP3B = (vec_s16) VEC_MERGEL(zero_u8v, srcP3);
+
+ srcM1A = (vec_s16) VEC_MERGEH(zero_u8v, srcM1);
+ srcM1B = (vec_s16) VEC_MERGEL(zero_u8v, srcM1);
+ srcM2A = (vec_s16) VEC_MERGEH(zero_u8v, srcM2);
+ srcM2B = (vec_s16) VEC_MERGEL(zero_u8v, srcM2);