debugger

[garraleta_fortran]

DEBUG issues continue with the new FORTRAN 2025.1.0

Am I doing something wrong?

[andrew_4619]

Program test
  USE DATOS
  IMPLICIT NONE
  TYPE(VECTOR) :: IMAG
  
  IMAG%X=100
  IMAG%Y=200
  CALL SUB0(IMAG)
  CALL SUB4(IMAG)
  CALL SUB5(IMAG)
  CALL SUB1(IMAG)
  CALL SUB2(IMAG)
  CALL SUB3(IMAG)
END program test

Firstly  I agree there are is something wrong here. I had a play with your project in 2025.1 and see exactly what you report.  I had a further play, I named the main but other then getting rib of _UNNAMED_MAIN_$$ that is not significant. Using proper Fortran (Imag%X rather than imag.x) also has no bearing.  I seems to me that the use of external subroutines in separate source files is part of the trigger for the problem. The compiler is questioning the derived type "Uniqueness", it should not because  of the module definition.

I added sub4 and sub5 to the sub0 source file and that throws up a new strange variations of type(datos:vector). If the subs are in module(s) the problem goes away. I am guess adding explicit interfaces directly in the main program would also make it go away. I think intel need to look at this again.

It is not a problem I would personally ever see as external subroutines do not exist in my codebase.

[Xiaoping_D_Intel]

I can reproduce the error and have opened a bug report for it.