Intel ARCHITECTURE IA-32 - Manual (Page 278)

IA-32 Intel® Architecture Optimization
5-16
// Start deswizzling here
movaps xmm7, xmm4 // xmm7= a1 a2 a3 a4
movhlps xmm7, xmm3 // xmm7= b3 b4 a3 a4
movaps xmm6, xmm2 // xmm6= g1 g2 g3 g4
movlhps xmm3, xmm4 // xmm3= b1 b2 a1 a2
movhlps xmm2, xmm1 // xmm2= r3 r4 g3 g4
movlhps xmm1, xmm6 // xmm1= r1 r2 g1 g2
movaps xmm6, xmm2 // xmm6= r3 r4 g3 g4
movaps xmm5, xmm1 // xmm5= r1 r2 g1 g2
shufps xmm2, xmm7, 0xDD // xmm2= r4 g4 b4 a4
shufps xmm1, xmm3, 0x88 // xmm4= r1 g1 b1 a1
shufps xmm5, xmm3, 0x88 // xmm5= r2 g2 b2 a2
shufps xmm6, xmm7, 0xDD // xmm6= r3 g3 b3 a3
movaps [edx], xmm4 // v1 = r1 g1 b1 a1
movaps [edx+16], xmm5 // v2 = r2 g2 b2 a2
movaps [edx+32], xmm6 // v3 = r3 g3 b3 a3
movaps [edx+48], xmm2 // v4 = r4 g4 b4 a4
// DESWIZZLING ENDS HERE
}
}
Example 5-7 Deswizzling Data 64-bit Integer SIMD Data
void mmx_deswizzle(IVertex_soa *in, IVertex_aos *out)
{
__asm {
mov ebx, in
mov edx, out
movq mm0, [ebx] // mm0= u1 u2
}
continued
Example 5-6 Deswizzling Data Using the movlhps and shuffle
Instructions (continued)