https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66713
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Keywords| |missed-optimization
Severity|normal |enhancement
--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
This is due to the cmp being done as a pointer to __atomic_compare_exchange_8
function insternally.
cmp_ = a_2(D);
__p2.4_5 = (long unsigned int) b_3(D);
__atomic_compare_exchange_8 (&MEM[(struct __atomic_base *)&aptr]._M_p, &cmp_,
__p2.4_5, 0, 4, 2);
It is not going to hurt performance most of the time.
