[Bug tree-optimization/123207] [16 Regression] Miscompare of 526.blender_r with -O3 on x86_64 since r16-6133-gca8d97f8cb8d28

[pheeck at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123207

--- Comment #8 from Filip Kastl <pheeck at gcc dot gnu.org> ---
(In reply to Sam James from comment #7)
> (In reply to Filip Kastl from comment #2)
> > (In reply to Richard Biener from comment #1)
> > > Does dropping -ffast-math work?
> > 
> > Just tried it (on Zen5).  -Ofast -fno-fast-math -march=native -flto still
> > results in a miscompare.
> 
> Just to double check: is this with Victor's fix (r16-6306-g57acdcf3fb3ec8)?
> Without checking, it'd be wrong-code.

I reproduced the miscompare on r16-6326-g9278f25eba9403.  So with r16-6306.