[Bug c++/125064] gcc-17 regression: canonical types differ for identical types ‘a’ and ‘a’

[ppalka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125064

Patrick Palka <ppalka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ppalka at gcc dot gnu.org

--- Comment #1 from Patrick Palka <ppalka at gcc dot gnu.org> ---
Could you provide the original preprocessed source and does compiling it also
emit the "warning: ‘trivial_abi’ cannot be applied" warning?  The original ICE
from gimplify_expr seems interesting own its own.