https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90857
Jonathan Wakely <redi at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Target Milestone|--- |12.0
--- Comment #5 from Jonathan Wakely <redi at gcc dot gnu.org> ---
One of the MSVC devs agreed with my interpretation and suggested that their
impl (which has a check for the position==last case) is paranoid.
I think the check in comment 3 is probably worth adding (for GCC 12), as long
as we keep the debug assertion to diagnose it when checking is enabled. We
could even consider __glibcxx_assert(__pos != __last) so that it's diagnosed
more aggressively.