This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
[patch, fortran] PR35724 compile time segmentation fault for eoshift
- From: Jerry DeLisle <jvdelisle at verizon dot net>
- To: Fortran List <fortran at gcc dot gnu dot org>
- Cc: gcc-patches <gcc-patches at gcc dot gnu dot org>
- Date: Wed, 16 Apr 2008 21:27:29 -0700
- Subject: [patch, fortran] PR35724 compile time segmentation fault for eoshift
This patch fixes the same problem we had with cshift.
Regression tested on x86-64.
Will commit as obvious.
Jerry
2008-04-16 Jerry DeLisle <jvdelisle@gcc.gnu.org>
PR fortran/35724
* iresolve.c (gfc_resolve_eoshift): Check for NULL symtree in test for
optional argument attribute.
! { dg-do compile }
! pr35724 compile time segmentation fault for eoshift with negative third arg
subroutine ra0072(dda,lda,nf10,nf1,mf1,nf2)
real dda(10,10)
logical lda(10,10)
dda = eoshift(dda,(/mf1,nf1/),tws0r,nf3-nf1)
lda = cshift(lda,(/mf1,nf1/),nf3-nf1)
where (lda) dda = eoshift(dda,1,1.0,-mf1)
end subroutine
Index: iresolve.c
===================================================================
--- iresolve.c (revision 134368)
+++ iresolve.c (working copy)
@@ -738,7 +738,8 @@ gfc_resolve_eoshift (gfc_expr *f, gfc_ex
if (dim != NULL)
{
- if (dim->expr_type != EXPR_CONSTANT && dim->symtree->n.sym->attr.optional)
+ if (dim->expr_type != EXPR_CONSTANT && dim->symtree != NULL
+ && dim->symtree->n.sym->attr.optional)
{
/* Mark this for later setting the type in gfc_conv_missing_dummy. */
dim->representation.length = shift->ts.kind;