https://gcc.gnu.org/bugzilla/show_bug.cgi?id=89646
Bug ID: 89646
Summary: Spurious actual argument might interfere warning
Product: gcc
Version: 9.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: fortran
Assignee: unassigned at gcc dot gnu.org
Reporter: ian_harvey at bigpond dot com
Target Milestone: ---
The following:
MODULE m
IMPLICIT NONE
TYPE :: t
END TYPE t
CONTAINS
SUBROUTINE s
! To reproduce, both actual arguments must be TARGET,
! both arguments must be of derived type.
TYPE(t), TARGET :: a(5)
TYPE(t), TARGET :: b(5)
CALL move(a, b)
END SUBROUTINE s
! To reproduce, called procedure must be elemental.
ELEMENTAL SUBROUTINE move(from, to)
TYPE(t), INTENT(INOUT) :: from
TYPE(t), INTENT(OUT) :: to
END SUBROUTINE move
END MODULE m
when compiled with recent trunk (r26545) gives the following spurious warnings:
argument-interfereb.f90:12:14-17:
12 | CALL move(a, b)
| 1 2
Warning: INTENT(INOUT) actual argument at (1) might interfere with actual
argument at (2).
argument-interfereb.f90:12:14-17:
12 | CALL move(a, b)
| 2 1
Warning: INTENT(OUT) actual argument at (1) might interfere with actual
argument at (2).
I suspect that the compiler is trying to warn me about potential aliasing given
the TARGET attribute on the actual arguments, but there is no such aliasing in
this example, and the details required to trigger the warning (arguments must
be derived type, procedure must be elemental) are too specific for this warning
to be useful or intended.
