Справочник Пользователя для AMD 250
Appendix E
SSE and SSE2 Optimizations
359
Software Optimization Guide for AMD64 Processors
25112
Rev. 3.06
September 2005
E.3
Reuse of Dead Registers
Optimization
When it is necessary to save the contents of a register that is in FPS format to another unused (or
dead) register, where the previous contents of the dead register are unknown and could be a denormal,
then use
dead) register, where the previous contents of the dead register are unknown and could be a denormal,
then use
movaps xmm1, xmm2
instead of
movss xmm1, xmm2
.
Application
This optimization applies to:
•
32-bit software
•
64-bit software
Rationale
The
movss xmm1, xmm2
instruction takes additional time to execute if the previous contents of
XMM1 are a denormal.