#define load_16way(src, x0, x1, x2, x3, x4, x5, x6, x7) \
vmovdqu (0*32)(src), x0; \
vmovdqu (1*32)(src), x1; \
vmovdqu (2*32)(src), x2; \
vmovdqu (3*32)(src), x3; \
vmovdqu (4*32)(src), x4; \
vmovdqu (5*32)(src), x5; \
vmovdqu (6*32)(src), x6; \
vmovdqu (7*32)(src), x7;
#define store_16way(dst, x0, x1, x2, x3, x4, x5, x6, x7) \
vmovdqu x0, (0*32)(dst); \
vmovdqu x1, (1*32)(dst); \
vmovdqu x2, (2*32)(dst); \
vmovdqu x3, (3*32)(dst); \
vmovdqu x4, (4*32)(dst); \
vmovdqu x5, (5*32)(dst); \
vmovdqu x6, (6*32)(dst); \
vmovdqu x7, (7*32)(dst);
#define store_cbc_16way(src, dst, x0, x1, x2, x3, x4, x5, x6, x7, t0) \
vpxor t0, t0, t0; \
vinserti128 $1, (src), t0, t0; \
vpxor t0, x0, x0; \
vpxor (0*32+16)(src), x1, x1; \
vpxor (1*32+16)(src), x2, x2; \
vpxor (2*32+16)(src), x3, x3; \
vpxor (3*32+16)(src), x4, x4; \
vpxor (4*32+16)(src), x5, x5; \
vpxor (5*32+16)(src), x6, x6; \
vpxor (6*32+16)(src), x7, x7; \
store_16way(dst, x0, x1, x2, x3, x4, x5, x6, x7);