x86/vf_blend: Add SSE2 optimization for divide
4.5x faster than C float version with autovectorization 10 x faster than C int version 25 x faster than C float version without autovectorization
Showing
Please
register
or
sign in
to comment