>+movh       [r0],       m0
>+movhps     [r0 + r1],  m0

>>change movh to movlps is better, movh+movhps is mixed float and integer path
Will movh+movhps cause any problem ? I thought movh will be faster.
In old CPU, the data across float and integer path need extra latency
_______________________________________________
x265-devel mailing list
[email protected]
https://mailman.videolan.org/listinfo/x265-devel

Reply via email to