catastrophic error: **Internal compiler error: internal abort**

[hs-napa]

Hi,

Here is a simple test case that crashes the compiler:

[fortran]MODULE TEST_MOD

   PUBLIC :: TEST_TYPE, LISTNODE

   PRIVATE

   TYPE TEST_TYPE
      CHARACTER(LEN=:), ALLOCATABLE :: STR
   END TYPE


   TYPE LISTNODE
      TYPE(TEST_TYPE), POINTER :: ITEM
      CLASS(LISTNODE), POINTER :: NEXT_NODE
   END TYPE

   TYPE LIST
      CLASS(LISTNODE), POINTER :: FIRST_NODE
   CONTAINS
      PROCEDURE :: INSERT_FIRST
   END TYPE

CONTAINS

   SUBROUTINE INSERT_FIRST( THIS, ITEM )
      CLASS(LIST) :: THIS
      TYPE(TEST_TYPE) :: ITEM
      CLASS(LISTNODE), POINTER :: NEW
      NEW => NEW_NODE()
      NEW % ITEM = ITEM

      IF ( ASSOCIATED( THIS % FIRST_NODE ) ) THEN
         NEW % NEXT_NODE => THIS % FIRST_NODE
      END IF
      THIS % FIRST_NODE => NEW

   END SUBROUTINE



   FUNCTION NEW_NODE()
      CLASS(LISTNODE), POINTER :: NEW_NODE
      ALLOCATE( LISTNODE :: NEW_NODE )
      ALLOCATE( NEW_NODE % ITEM )
   END FUNCTION

END MODULE
[/fortran]

Compiling the test file gives the following output:

>ifort alloc_test.f90
Intel Visual Fortran Compiler Professional for applications running on IA-32, Version 11.1 Build 20091130 Package ID: w_cprof_p_11.1.054
Copyright (C) 1985-2009 Intel Corporation. All rights reserved.

alloc_test.f90(29): catastrophic error: **Internal compiler error: internal abort** Please report this error along with the circumstances in which it occurred in a Software Problem
Report. Note: File and line given may not be explicit cause of this error.
NEW => NEW_NODE()
------^
compilation aborted for alloc_test.f90 (code 3)

I can work around the problem by using a subroutine returning the newly allocated pointer as an INTENT(OUT) parameter.

BR,
-Heikki

[TimP]

No crash on compilers issued since July. Time to upgrade?

[Steven_L_Intel1]

Confirming Tim's report, this compiles fine with 11.1.067 (Update 7).

[hs-napa]

Thanks for your answers.

We are in a process of upgrading to newer version of the compiler. However, with more than a million lines of fortran in our codebase the testing takes a while before we can take a new compiler version into production. Our production team is busy with that. Currently I'm stuck with the 11.1.054 in my daily work and I can't do much about it.

Anyway, I will install update 7 for testing these bugs before submitting them to not report bugs that have already been fixed.

I find this forum very useful. Thank you guys, keep up the good work!

BR,
-Heikki