https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89285
Jakub Jelinek <jakub at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|UNCONFIRMED |NEW
Last reconfirmed| |2019-02-11
CC| |dmalcolm at gcc dot gnu.org,
| |jakub at gcc dot gnu.org,
| |redi at gcc dot gnu.org
Target Milestone|--- |8.3
Summary|ICE after casting the this |[8 Regression] ICE after
|pointer in the constructor |casting the this pointer in
|in C++17 mode |the constructor in C++17
| |mode
Ever confirmed|0 |1
--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
This started to ICE with r259629 and stopped ICEing with r267272.
Not really sure if this is ice-on-invalid (and thus accepts-invalid) on the
trunk or just ice-on-valid on the branch, given the reinterpret_casts in the
constexpr body.
