[Bug fortran/45092] New: internal compiler error regression bug in the latest trunk build of the gfortran compiler

["clerman at fuse dot net" <gcc-bugzilla at gcc dot gnu dot org>]

Hello,

  I am e-mailing to gcc-bugzilla@gcc.gnu.org file bug8.tar.gz. Please unpack it
and then invoke the shell script bug8.sh (the name of the compiler I'm
symbolically linked to is gfortran-trunk). The following message appears
compiling file functiontypeM.f90 (see file functiontypesM.xyz): 

norm@oxford:~/design/gfortran/bug8$ more functiontypesM.xyz
functiontypesM.f90:206.65:

    & [(0.0_DOUB_PREC, iPars = 1_TinyInteger, NUMBER_OF_TERM_PARS)], [" ", "
"])
                                                                 1
Warning: Possible change of value in conversion from INTEGER(4) to INTEGER(1)
at (1)
functiontypesM.f90:383:0: internal compiler error: in
output_constructor_regular_field, at varasm.c:5027
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

  This bug is a regression; versions 4.4 and 4.5 successfully compile the
routine.

  Please note that I question the warning message issued for line 206: All the
entities involved are of the same kind, so, I assume, no conversion should be
done.

  Thank you for your attention.

Yours truly,

Norman S. Clerman
consultant


-- 
           Summary: internal compiler error regression bug in the latest
                    trunk build of the gfortran compiler
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: blocker
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: clerman at fuse dot net


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45092