[Bug fortran/78611] -march=native makes code 3x slower

[pepalogik at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=78611

--- Comment #4 from Jan Lachnitt <pepalogik at seznam dot cz> ---
Small test case with -march=core-avx-i:
real    0m1.300s
user    0m1.296s
sys     0m0.000s

I.e., reproduced.