https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97770
--- Comment #20 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to H.J. Lu from comment #19)
> (In reply to rguent...@suse.de from comment #18)
> >
> > That's going to work I guess but it will pessimize general optimization
> > which no longer know this is computing popcount (not sure if the old
> > version exposed that fact).
>
> Do we have testcase for that?
Sth like
int
foo (unsigned long long a)
{
return _mm_popcnt_u64 (a) == 0;
}
basically anything that triggers a simplification of POPCOUNT as implemented
in match.pd (there are quite some). But no, I don't think we have any
of those using the x86 intrinsics in the testsuite.
