This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patches] Re: SSE min/max take2 - documentation
>
>
> On Tue, 27 Feb 2001, Jan Hubicka wrote:
> |> I'd say that the tree construct for 'max' should have exactly the same
> |> semantics as the (x > y ? x : y) construct in C. On IEEE compliant machines,
> |> this means we have all the proper behaviour regarding signed zeroes and
> |> NaN's. On other machines, or on -ffast-math compilations, we have additional
> |> freedom.
> |PowerPC is having min/max pattern using >= and <= comparators, unfortunately.
>
> We can use the identity (x > y ? x : y) == (-y <= -x ? y : x)
> to uniformly handle all cases on all platforms. This identity
> holds in presence of signed zeroes, infinities and NaN's.
hmm, thats nice, but how this helps to expanders?
If we define min expander to generate three negations and max on target
platform, I don't think we will win the code quality....
Honza
>
> -Geert