[Bug middle-end/63273] atomic operations lead to inefficient code

[dvyukov at google dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63273

--- Comment #4 from Dmitry Vyukov <dvyukov at google dot com> ---
For the record, this bug is the result of my attempt to use std atomic
operations in ThreadSanitizer runtime. We do a bunch of relaxed loads and
stores during processing of every memory access in the target program. Usage of
std atomics instead of hand-rolled atomics has decreased end-to-end performance
by 30%, as the result we had to reject usage of std atomics and keep own
implementation.