This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

fix remap_save_expr buglet


Seen on tree-ssa branch, where we actually hit this case, then generate
bad code because of it.  I don't know if it affects mainline, exactly,
but I can't believe its right anywhere.


r~


        * tree-inline.c (remap_save_expr): Map new save_expr to identity
        rather than to error_mark_node.

Index: tree-inline.c
===================================================================
RCS file: /cvs/gcc/gcc/gcc/tree-inline.c,v
retrieving revision 1.26.2.46
diff -u -p -r1.26.2.46 tree-inline.c
--- tree-inline.c	22 Sep 2003 16:54:34 -0000	1.26.2.46
+++ tree-inline.c	23 Sep 2003 19:59:34 -0000
@@ -1812,8 +1812,7 @@ remap_save_expr (tree *tp, void *st_, tr
 			     (splay_tree_key) *tp,
 			     (splay_tree_value) t);
       /* Make sure we don't remap an already-remapped SAVE_EXPR.  */
-      splay_tree_insert (st, (splay_tree_key) t,
-			 (splay_tree_value) error_mark_node);
+      splay_tree_insert (st, (splay_tree_key) t, (splay_tree_value) t);
     }
   else
     /* We've already walked into this SAVE_EXPR, so we needn't do it


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]