On 26/05/14 09:44, Niels Möller wrote: > Luca Barbato <[email protected]> writes: > >> Which test file? Early micro optimization might lead to global slow >> down. > > Don't know. Anyway, replacing a function call by a float multiplication > by constant seems like a pretty safe micro-optimization to me.
This one looks surely safe and probably with -O3 some compilers might even optimize it that way. >> 1% is not small btw. > > I have some difficulty believing that number, but I haven't made any > benchmarks myself. That's why I wanted a sample to see how's possible. =) lu _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
