[Bug target/59767] __atomic_load_n with __ATOMIC_SEQ_CST should result in a memory barrier
jamrial at gmail dot com
gcc-bugzilla@gcc.gnu.org
Thu Jul 2 16:35:00 GMT 2015
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=59767
James Almer <jamrial at gmail dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jamrial at gmail dot com
--- Comment #1 from James Almer <jamrial at gmail dot com> ---
This still happens with gcc 4.9 and gcc 5.
int foo(int *bar)
{
return *bar;
}
and
int foo(int *bar)
{
return __atomic_load_n(bar, __ATOMIC_SEQ_CST);
}
Generate the exact same assembly. The latter should add an mfence before the
mov instruction.
More information about the Gcc-bugs
mailing list