[LAD] Floating point Denormals: C++ and Waf

Jens M Andreasen jens.andreasen at comhem.se
Sat Aug 4 08:49:33 UTC 2012


On Fri, 2012-08-03 at 15:50 +0100, Harry van Haaren wrote:

> 
> Thanks, it seems that this is the easiest cross-CPU / architecture
> etc 

The problem is unique for Intel, in which case you can use:

-msse -mfpmath=sse -ffast-math





More information about the Linux-audio-dev mailing list