This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
[Patch, fortran] PR29392 -Segfault on data character initialization
- From: Paul Thomas <paulthomas2 at wanadoo dot fr>
- To: patch <gcc-patches at gcc dot gnu dot org>, "'fortran at gcc dot gnu dot org'" <fortran at gcc dot gnu dot org>, François-Xavier Coudert <fxcoudert at gmail dot com>
- Date: Mon, 16 Oct 2006 17:23:31 +0200
- Subject: [Patch, fortran] PR29392 -Segfault on data character initialization
:ADDPATCH fortran:
The fix here was simple... literally. The expression for the substring
start and end needed copying, so that it doesn't get frozen for one
value of the iterator, and simplifying. The testcase is that of the
reporter.
Regtested on PIV/Cygwin_NT - OK for trunk?
Paul
Index: gcc/fortran/ChangeLog
===================================================================
*** gcc/fortran/ChangeLog (révision 117568)
--- gcc/fortran/ChangeLog (copie de travail)
***************
*** 1,3 ****
--- 1,10 ----
+ 2006-10-16 Paul Thomas <pault@gcc.gnu.org>
+
+ PR fortran/29392
+ * data.c (create_character_intializer): Copy and simplify
+ the expressions for the start and end of a sub-string
+ reference.
+
2006-10-08 Erik Edelmann <edelmann@gcc.gnu.org>
Paul Thomas <pault@gcc.gnu.org>
Index: gcc/fortran/data.c
===================================================================
*** gcc/fortran/data.c (révision 117568)
--- gcc/fortran/data.c (copie de travail)
*************** create_character_intializer (gfc_expr *
*** 167,179 ****
if (ref)
{
gcc_assert (ref->type == REF_SUBSTRING);
/* Only set a substring of the destination. Fortran substring bounds
are one-based [start, end], we want zero based [start, end). */
! gfc_extract_int (ref->u.ss.start, &start);
start--;
! gfc_extract_int (ref->u.ss.end, &end);
}
else
{
--- 167,192 ----
if (ref)
{
+ gfc_expr *start_expr, *end_expr;
+
gcc_assert (ref->type == REF_SUBSTRING);
/* Only set a substring of the destination. Fortran substring bounds
are one-based [start, end], we want zero based [start, end). */
! start_expr = gfc_copy_expr (ref->u.ss.start);
! end_expr = gfc_copy_expr (ref->u.ss.end);
!
! if ((gfc_simplify_expr (start_expr, 1) == FAILURE)
! || (gfc_simplify_expr (end_expr, 1)) == FAILURE)
! {
! gfc_error ("failure to simplify substring reference in DATA"
! "statement at %L", &ref->u.ss.start->where);
! return NULL;
! }
!
! gfc_extract_int (start_expr, &start);
start--;
! gfc_extract_int (end_expr, &end);
}
else
{
Index: gcc/testsuite/ChangeLog
===================================================================
*** gcc/testsuite/ChangeLog (révision 117568)
--- gcc/testsuite/ChangeLog (copie de travail)
***************
*** 1,3 ****
--- 1,8 ----
+ 2006-10-16 Paul Thomas <pault@gcc.gnu.org>
+
+ PR fortran/29392
+ * gfortran.dg/data_char_3.f90: New test.
+
2006-10-08 Erik Edelmann <edelmann@gcc.gnu.org>
Paul Thomas <pault@gcc.gnu.org>
Index: gcc/testsuite/gfortran.dg/data_char_3.f90
===================================================================
*** gcc/testsuite/gfortran.dg/data_char_3.f90 (révision 0)
--- gcc/testsuite/gfortran.dg/data_char_3.f90 (révision 0)
***************
*** 0 ****
--- 1,11 ----
+ ! { dg-do run }
+ ! { dg-options "-O2" }
+ ! Tests the fix PR29392, in which the iterator valued substring
+ ! reference would cause a segfault.
+ !
+ ! Contributed by Francois-Xavier Coudert <fxcoudert@gcc.gnu.org>
+ !
+ character(LEN=2) :: a(2)
+ data ((a(I)(k:k),I=1,2),k=1,2) /2*'a',2*'z'/
+ IF (ANY(a.NE."az")) CALL ABORT()
+ END