Subject: Re: compiling -current on amd64...
To: None <>
From: Matthias Drochner <>
List: current-users
Date: 06/26/2006 14:08:31
This is a multipart MIME message.

Content-Type: text/plain; charset=us-ascii said:
> /u/NetBSD/src/gnu/dist/binutils/libiberty/concat.c: In function
> 'reconcat': /u/NetBSD/src/gnu/dist/binutils/libiberty/concat.c:205:
> warning: second parameter of 'va_start' not last named argument 

This happens to me too, on i386.
It seems to be caused by a recent gcc fix which does a bit more
than intended. The appended patch helps for me.

best regards

Content-Type: text/plain ; name="va.txt"; charset=us-ascii
Content-Description: va.txt
Content-Disposition: attachment; filename="va.txt"

Index: builtins.c
RCS file: /cvsroot/src/gnu/dist/gcc4/gcc/builtins.c,v
retrieving revision 1.4
diff -u -p -r1.4 builtins.c
--- builtins.c	26 Jun 2006 00:33:11 -0000	1.4
+++ builtins.c	26 Jun 2006 12:05:54 -0000
@@ -9799,12 +9799,14 @@ fold_builtin_next_arg (tree arglist)
 	arg = TREE_OPERAND (arg, 0);
       if (arg != last_parm)
+#if 0
 	  /* FIXME: Sometimes with the tree optimizers we can get the
 	     not the last argument even though the user used the last
 	     argument.  We just warn and set the arg to be the last
 	     argument so that we will get wrong-code because of
 	     it.  */
 	  warning (0, "second parameter of %<va_start%> not last named argument");
       /* We want to verify the second parameter just once before the tree
          optimizers are run and then avoid keeping it in the tree,