https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125391
Bug ID: 125391
Summary: [16 Regression] ICE on derived type (with final) array
assignment inside polymorphic class
Product: gcc
Version: 16.1.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: fortran
Assignee: unassigned at gcc dot gnu.org
Reporter: federico.perini at gmail dot com
Target Milestone: ---
Created attachment 64502
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=64502&action=edit
minimal example reproducing the issue.
Verify on Compiler Explorer at: https://godbolt.org/z/63errv1sP
I found this bug after MSYS2 upgraded to gfortran 16.1, but the same is still
present on trunk (see Compiler Explorer).
> in gimplify_var_or_parm_decl, at gimplify.cc:3426 (16.1)
> in gimplify_var_or_parm_decl, at gimplify.cc:3266 (trunk, 26/5/20)
Happens when:
- a derived type `inner` has `final :: ` procedure, even empty;
- wrapper type `outer` contains `type(inner), allocatable :: array(:)`, may or
may not have a `final ::`
- polymorphic procedure has `class(outer), intent(inout)`, not necessarily
type-bound
- inside the procedure, assignment happens: `this%array = [constructor]` or
`this%array = that%array`
The bug does not appear on gfortran-15 or previous versions. I've tried since
gfortran-7 approximately.
Thanks for the great work on gfortran!
Best regards,
Federico
