[Bug c++/116095] satisfaction of atomic constraint depends on itself, when piping boost range into range-v3

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116095

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Mainly also see https://cplusplus.github.io/CWG/issues/2369.html .