>+movd        m0,    [r2]
>+pinsrw      m0,    [r2 + r4],    2
>+movd        m1,    [r2 + 2 * r4]
pinsrw is expensive instruction, it generate 2 uops
at here, the width is 2, so I suggest use general register future (movzx+sub)
_______________________________________________
x265-devel mailing list
[email protected]
https://mailman.videolan.org/listinfo/x265-devel

Reply via email to