The attached code produces the subject error. gfortran -c bug_n.f90 bug_n.f90:45: internal compiler error: Segmentation fault Please submit a full bug report, with preprocessed source if appropriate. See <URL:http://gcc.gnu.org/bugs.html> for instructions. [sfilippo@localhost NEMO]$ gfortran -v Using built-in specs. Target: i686-pc-linux-gnu Configured with: ../gcc-4.2-20070124/configure --prefix=/usr/local/gcc42 --with-mpfr=/home/travel/GCC/BUILDS/mpfr --with-gmp-lib=/home/travel/GCC/BUILDS/gmp/lib/ --with-gmp=/home/travel/GCC/BUILDS/gmp Thread model: posix gcc version 4.2.0 20070124 (prerelease)
Created attachment 12999 [details] test case
I'm testing a patch for this at the moment. The problem is you have renamed SIZE to an interface. That is, this is leading to the segfault: interface size module procedure get_fld_size end interface While the standard may permit this construct, it may be prudent to not rename standard intrinsic procedures.
Subject: Bug 30683 Author: kargl Date: Sat Feb 3 01:01:06 2007 New Revision: 121531 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=121531 Log: 2007-02-02 Steven G. Kargl <kargl@gcc.gnu.org> PR fortran/30683 * resolve.c (resolve_generic_f): Check for non-NULL sym. * gfortran.dg/interface_10.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/interface_10.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/resolve.c trunk/gcc/testsuite/ChangeLog
Fixed on trunk.
(In reply to comment #2) > I'm testing a patch for this at the moment. The problem is you > have renamed SIZE to an interface. That is, this is leading to > the segfault: > > > interface size > module procedure get_fld_size > end interface > > While the standard may permit this construct, it may be prudent to > not rename standard intrinsic procedures. > Agreed. I take it you mean it IS standard conforming, right? Just to make sure. In any case, an ice is an ice....
Subject: Re: [4.2 only] internal compiler error: Segmentation fault On Sat, Feb 03, 2007 at 12:05:53PM -0000, sfilippone at uniroma2 dot it wrote: > > > I'm testing a patch for this at the moment. The problem is you > > have renamed SIZE to an interface. That is, this is leading to > > the segfault: > > > > > > interface size > > module procedure get_fld_size > > end interface > > > > While the standard may permit this construct, it may be prudent to > > not rename standard intrinsic procedures. > > > Agreed. > I take it you mean it IS standard conforming, right? I a quick scan of F2003, I could not find anything that say the construct above is nonconforming. So, I believe that it is permitted. > Just to make sure. In any case, an ice is an ice.... Yes, indeed. Fortunately, this one was easy to find and fix.
Subject: Bug 30683 Author: kargl Date: Sat Feb 3 18:59:53 2007 New Revision: 121548 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=121548 Log: 2007-02-02 Steven G. Kargl <kargl@gcc.gnu.org> PR fortran/30683 * resolve.c (resolve_generic_f): Check for non-NULL sym. * gfortran.dg/interface_10.f Added: branches/gcc-4_2-branch/gcc/testsuite/gfortran.dg/interface_10.f90 Modified: branches/gcc-4_2-branch/gcc/fortran/ChangeLog branches/gcc-4_2-branch/gcc/fortran/resolve.c branches/gcc-4_2-branch/gcc/testsuite/ChangeLog
Fixed on 4.2.