On 05/18/2013 06:11 PM, Kostya Shishkov wrote: > On Sat, May 18, 2013 at 03:40:54PM +0200, Luca Barbato wrote: >> On recent hardware and compiler using a plain division is faster. >> --- >> >> Some numbers using gcc-4.8 in -Og mode. > > It would be more interesting to test on RISCy architectures that lack division > instruction. Got MIPS or PowerPC?
ppc32 + gcc4.7 isn't happy, ppc64 _should_ be happy since it has a proper div (and playing with addresses has an additional cost), couldn't test on it right now. I'm about to push j-b patch about that assert. lu _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
