[incremental] Patch: FYI: fix thinko in multi-hunk check

Tom Tromey tromey@redhat.com
Wed Dec 12 18:58:00 GMT 2007


I'm checking this in on the incremental compiler branch.

This fixes a bug in the failure check for multi-hunk bindings.
The code was checking the wrong variable for the wrong value.

Also, this fixes a bad comment I noticed in c-tree.h.

With this patch, plus the ones from yesterday, I can now successfully
build GNU idutils, which is a multi-directory project.  This shows a
performance problem compiling regex.c :(, so I'll probably look at
that next.

Tom

ChangeLog:
2007-12-12  Tom Tromey  <tromey@redhat.com>

	* c-parser.c (check_hunk_binding): Fix multi-hunk failure check.
	* c-tree.h (C_IS_RESERVED_WORD): Update documentation.

Index: c-tree.h
===================================================================
--- c-tree.h	(revision 130776)
+++ c-tree.h	(working copy)
@@ -58,8 +58,8 @@
 #define C_TYPE_INCOMPLETE_VARS(TYPE) TYPE_VFIELD (TYPE)
 
 /* In an IDENTIFIER_NODE, nonzero if this identifier is actually a
-   keyword.  C_RID_CODE (node) is then the RID_* value of the keyword,
-   and C_RID_YYCODE is the token number wanted by Yacc.  */
+   keyword.  C_RID_CODE (node) is then the RID_* value of the
+   keyword.  */
 #define C_IS_RESERVED_WORD(ID) TREE_LANG_FLAG_0 (ID)
 
 struct lang_type GTY(())
Index: c-parser.c
===================================================================
--- c-parser.c	(revision 130506)
+++ c-parser.c	(working copy)
@@ -1969,7 +1969,7 @@
       self_iter = self_iter->next;
     }
 
-  if (binding == NULL)
+  if (binding_iter)
     return true;
 
   info->binding = binding;



More information about the Gcc-patches mailing list