Hi,
I was wondering if there were any drawbacks to doing __meinvf32(0.0f) and clamping the value between 0.0 and 1.0f afterwards. I'm asking because since IF statements slows down the CLA processing time, it would be faster to just streamline the code and avoid testing for 0.0f by doing :
float32 divide = __meinvf32(0.0f); divide = divide * (2.0f - divide * 0.0f); divide = divide * (2.0f - divide * 0.0f); __mmaxf32( __mminf32( divide, 1.0f), 0.0f)
Let me know what you think!
Regards,