[Issue 15007] core.atomic match C++11

[via Digitalmars-d-bugs]

https://issues.dlang.org/show_bug.cgi?id=15007

--- Comment #1 from z.p.gaal.devel at gmail.com ---
Also plese take a look at this forum post:
https://forum.dlang.org/thread/muisadtexibcuudzjtbx@forum.dlang.org

What is the state of fences and compiler optimization?
Is it guaranteed that, no instruction is moved increctly before/after a
fence/atomic operation ?

--