module m contains subroutine s character(:), allocatable :: a1 a1 = a1 return contains integer function f1() f1 = 0 end function f1 end subroutine s integer function f2() character(:), allocatable :: a2 f2 = 0 ! this line causes a seg fault! a2 = a2 end function f2 end module m ##################### The above do-nothing construction gives an ICE. Difficult for me to know why but it is probably something to do with the allocatable character scalars. As far as I know, it is legal Fortran. P.S. Thanks Paul, Dominique et al. for the recent push on the these things. gcc version 6.0.0 20160410 (experimental) (GCC) :: mingw build from equation.com running on x86_w64.
Ah, maybe I reduced this too far as a1 and a2 are obviously uninitialized, but still...
Ah, maybe I reduced this too far as a1 and a2 are obviously uninitialized... Is this better - it also produces an ICE? module m contains subroutine s(in) character(*), intent(in) :: in character(:), allocatable :: a a = in a = a return contains integer function f1() f1 = 0 end function f1 end subroutine s integer function f2(a) character(*), intent(in) :: a character(:), allocatable :: in in = a in = in f2 = 0 end function f2 end module m
The tests compile with 4.8.5 and 4.9.3, but give an ICE with 5.3 and trunk (6.0). The change occurred between revisions r219174 (2015-01-04, OK) and r219318 (2015-01-07, ICE); r219193 (pr47674)?
It seems the problem usually arises when there is reassignment in one line, eg character(:), allocatable: s, t s = s(2:) ! or even s = s whereas forcing the temporary t = s(2:) s = t seems to work. But it is not quite that simple as the above doesn't always fail. It requires other things to be happening too.
Author: pault Date: Mon Jun 13 07:48:25 2016 New Revision: 237358 URL: https://gcc.gnu.org/viewcvs?rev=237358&root=gcc&view=rev Log: 2016-06-13 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * frontend-passes.c (realloc_string_callback): Add a call to gfc_dep_compare_expr. 2016-06-13 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * gfortran.dg/pr70673.f90: New test. Added: trunk/gcc/testsuite/gfortran.dg/pr70673.f90 Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/frontend-passes.c trunk/gcc/testsuite/ChangeLog
Author: pault Date: Mon Jun 27 19:18:21 2016 New Revision: 237802 URL: https://gcc.gnu.org/viewcvs?rev=237802&root=gcc&view=rev Log: 2016-06-27 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * frontend-passes.c (realloc_string_callback): Add a call to gfc_dep_compare_expr. 2016-06-27 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * gfortran.dg/pr70673.f90: New test. Added: branches/gcc-6-branch/gcc/testsuite/gfortran.dg/pr70673.f90 Modified: branches/gcc-6-branch/gcc/fortran/ChangeLog branches/gcc-6-branch/gcc/fortran/frontend-passes.c branches/gcc-6-branch/gcc/testsuite/ChangeLog
Author: pault Date: Mon Jun 27 20:54:56 2016 New Revision: 237804 URL: https://gcc.gnu.org/viewcvs?rev=237804&root=gcc&view=rev Log: 2016-06-27 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * frontend-passes.c (realloc_string_callback): Add a call to gfc_dep_compare_expr. 2016-06-27 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * gfortran.dg/pr70673.f90: New test. Added: branches/gcc-5-branch/gcc/testsuite/gfortran.dg/pr70673.f90 Modified: branches/gcc-5-branch/gcc/fortran/frontend-passes.c
Author: pault Date: Mon Jun 27 20:57:13 2016 New Revision: 237805 URL: https://gcc.gnu.org/viewcvs?rev=237805&root=gcc&view=rev Log: 2016-06-27 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * frontend-passes.c (realloc_string_callback): Add a call to gfc_dep_compare_expr. 2016-06-27 Paul Thomas <pault@gcc.gnu.org> PR fortran/70673 * gfortran.dg/pr70673.f90: New test. Modified: branches/gcc-5-branch/gcc/fortran/ChangeLog branches/gcc-5-branch/gcc/testsuite/ChangeLog
Closing rather belatedly! Thanks for the report. Paul