On Sun, Jan 30, 2011 at 02:29:13PM +0100, Philipp Überbacher wrote:
i like -ffast-math... this whole thread is about float not being exact.
sure its worse with -ffast-math ... but if you dont assume correctness,
and do appropriate workarounds in case where you need correct results,
its the way to go.
Linux-audio-user mailing list