PATCH Reversion

[Mark Mitchell]

Nathan --

  As noted on the list, your patch to decl2.c somehow broke libjava.
I don't understand why from looking at the patch, which is why I
approved it in the first place. :-) Still, I can confirm that removing
it fixes the problem.  So, I've reverted this little portion of the
patch, which pretty well eviscerates the effect of the patch.  Please
track down what's going on as soon as possible.

  I only reverted this patch on the branch since wer're a little more
tolerant of problems on the mainline.

  Still, the rules have changed: now that libjava gets built by
default, we need to be testing that too whenever we make a C++
front-end change; just running `make check-g++' is no longer good
enough.  (Note that this is a policy change from historical practice.)

  Thanks,

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com

2001-05-01  Mark Mitchell  <mark@codesourcery.com>

	* decl2.c (do_using_directive): Revert previous patch.

Index: decl2.c
===================================================================
RCS file: /cvs/gcc/gcc/gcc/cp/decl2.c,v
retrieving revision 1.437.2.16
retrieving revision 1.437.2.17
diff -c -p -r1.437.2.16 -r1.437.2.17
*** decl2.c	2001/05/01 12:54:43	1.437.2.16
--- decl2.c	2001/05/01 23:41:03	1.437.2.17
*************** do_using_directive (namespace)
*** 5226,5233 ****
  {
    if (namespace == fake_std_node)
      return;
-   if (building_stmt_tree ())
-     add_stmt (build_stmt (USING_STMT, namespace));
    
    /* using namespace A::B::C; */
    if (TREE_CODE (namespace) == SCOPE_REF)
--- 5226,5231 ----