[Bug target/124042] [16 regression] Miscompare of 482.sphinx3 on aarch64

[pheeck at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124042

--- Comment #2 from Filip Kastl <pheeck at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #1)
> Is there a way to try an additional flag, -mearly-ra=none ?

Sure, I'll test that later this week.  You mean that I should try to bisect
with -mearly-ra=none or that I should just try if I still get a miscompare with
the latest trunk GCC and -mearly-ra=none?

Also, it doesn't reproduce with -O3.