Fix tgamma errno setting on domain error (bug 6809)

Joseph S. Myers joseph@codesourcery.com
Fri May 10 11:39:00 GMT 2013


On Fri, 10 May 2013, Ondrej Bilka wrote:

> > -	if(__builtin_expect(!__finite(y), 0)&&__finite(x)
> > +	if(__builtin_expect(!__finite(y), 0)
> > +	   && (__finite (x) || __isinf (x) < 0)
> A nit why expect here is only in part of expression but not whole
> expression?

If the return value isn't finite, we're already in the unlikely case and 
there isn't really much point optimizing the conditionals that come after 
that.  It makes sense for the first conditional to be the one that splits 
things into likely / unlikely cases, since it's the likely case (finite 
result) that's most important to optimize.

-- 
Joseph S. Myers
joseph@codesourcery.com



More information about the Libc-alpha mailing list