gfortran 4.1.2: "internal compiler error: Segmentation fault"

[Rob]

--- Bernhard Fischer <rep.nop@aon.at> wrote:

> On Fri, Sep 29, 2006 at 03:49:37PM -0700, Rob wrote:
> >
> >Hi,
> >
> >I have compile GCC 4.1.2 (prerelease) on an Alpha
> >system with CentOS 4 (clone of Redhat Enterprise).
> >
> >The required file to reproduce this problem are
> >here: http://surfion.snu.ac.kr/~lahaye/gcc/
> 
> Please reduce this to something that doesn't look
> like parts of vasp ;)
> I compiled (and ran) a current vasp.4.6 with
> gfortran a while ago and it did work fine.
> 
> I can confirm that, however, it currently fails for
> this reduced (and not vasp related) testcase:
> 
> $ cat no.f90 
>         MODULE MOD1
>         CONTAINS
>         SUBROUTINE SUB1(arg)
>         IMPLICIT NONE
>         CHARACTER (LEN=*), OPTIONAL :: arg
>         IF (PRESENT(arg)) WRITE(0,*) 'arg was ', arg
>         STOP
>         END SUBROUTINE SUB1
>         SUBROUTINE SUB2
>         CALL SUB1
>         END SUBROUTINE SUB2
>         END MODULE
> $ gfortran-4.2-HEAD -c no.f90
> no.f90: In function 'sub2':
> no.f90:3: internal compiler error: Segmentation
> fault

Yes, indeed I can now confirm this too.
Moreover, if you replace
  CALL SUB1
by
  CALL SUB1("")
the segfault goes away.
This way I can temporarily bypass this problem
with gfortran.

Also, this is not in the trunk, but in the
GCC 1.4.2 svn branch. In my case:

 $  gfortran --version
 GNU Fortran 95 (GCC) 4.1.2 20061001 (prerelease)

Regards,
Rob.


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com