c99 complex multiplication and division

Paolo Carlini pcarlini@suse.de
Sat Feb 12 17:39:00 GMT 2005


Richard Henderson wrote:

>Now we must discuss when we want this new code enabled...
>  
>
Thanks for your work.

In my humble opinion, we should *definitely* enable it by default: the 
C++ library needs it (in order to consistently implement complex), 
fortran needs it, not to mention C, of course. Personally, I would 
consider it an important step forward for serious scientific computations.

In case of -ffast-math, I don't care much, frankly, probably the current 
behavior is the right one.

I'm sure that Joseph and Gaby, at least, have something to say.

Paolo.



More information about the Gcc-patches mailing list