Intel ARCHITECTURE IA-32 - Manual (Page 236)

IA-32 Intel® Architecture Optimization
4-16
Move Byte Mask to Integer
The pmovmskb instruction returns a bit mask formed from the most
significant bits of each byte of its source operand. When used with the
64-bit MMX registers, this produces an 8-bit mask, zeroing out the
upper 24 bits in the destination register. When used with the 128-bit
XMM registers, it produces a 16-bit mask, zeroing out the upper 16 bits
in the destination register. The 64-bit version is shown in Figure 4-7 and
Example 4-10.
Example 4-9 Repeated pinsrw Instruction Code
; Input:
; edx pointer to structure containing source
; values at offsets: of +0, +10, +13, and +24
; immediate value: “1”
; Output:
; MMX register with new 16-bit value inserted
;
pxor mm0, mm0 ; Breaks dependedncy on previous value of mm0
mov eax, [edx]
pinsrw mm0, eax, 0
mov eax, [edx+10]
pinsrw mm0, eax, 1
mov eax, [edx+13]
pinsrw mm0, eax, 2
mov eax, [edx+24]
pinsrw mm0, eax, 3