https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88713

--- Comment #33 from Marc Glisse <glisse at gcc dot gnu.org> ---
(In reply to Chris Elrod from comment #32)
> (In reply to Marc Glisse from comment #31)
> > What we need to understand is why gcc doesn't try to generate rsqrt

Without -mavx512er, we do not have an expander for rsqrtv16sf2, and without
that I don't know how the machinery can guess how to use rsqrt (there are
probably ways).

> The approximate sqrt, and then approximate reciprocal approximations were
> slower on my computer than just vsqrt followed by div.

We can probably split that into the speed of sqrt vs its approximation and
inverse (div) vs its approximation. At least one of them seems to be a
pessimization on that platform.

Reply via email to