From patchwork Tue Oct 26 21:34:39 2010
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
X-Patchwork-Submitter: Jan Hubicka <hubicka@ucw.cz>
X-Patchwork-Id: 69294
Return-Path: 
 <gcc-patches-return-276470-incoming=patchwork.ozlabs.org@gcc.gnu.org>
X-Original-To: incoming@patchwork.ozlabs.org
Delivered-To: patchwork-incoming@bilbo.ozlabs.org
Received: from sourceware.org (server1.sourceware.org [209.132.180.131])
	by ozlabs.org (Postfix) with SMTP id 16E2BB70AA
	for <incoming@patchwork.ozlabs.org>;
	Wed, 27 Oct 2010 08:34:59 +1100 (EST)
Received: (qmail 15615 invoked by alias); 26 Oct 2010 21:34:48 -0000
Received: (qmail 15497 invoked by uid 22791); 26 Oct 2010 21:34:47 -0000
X-SWARE-Spam-Status: No, hits=-1.8 required=5.0	tests=AWL, BAYES_00,
	T_RP_MATCHES_RCVD
X-Spam-Check-By: sourceware.org
Received: from nikam-dmz.ms.mff.cuni.cz (HELO nikam.ms.mff.cuni.cz)
	(195.113.20.16) by sourceware.org (qpsmtpd/0.43rc1) with
	ESMTP; Tue, 26 Oct 2010 21:34:41 +0000
Received: by nikam.ms.mff.cuni.cz (Postfix, from userid 16202)	id
	7E1409ACA2E; Tue, 26 Oct 2010 23:34:39 +0200 (CEST)
Date: Tue, 26 Oct 2010 23:34:39 +0200
From: Jan Hubicka <hubicka@ucw.cz>
To: gcc-patches@gcc.gnu.org
Subject: More cgraph_remove_unreachable_nodes fixes
Message-ID: <20101026213439.GA3813@kam.mff.cuni.cz>
MIME-Version: 1.0
Content-Disposition: inline
User-Agent: Mutt/1.5.18 (2008-05-17)
Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-patches.gcc.gnu.org>
List-Unsubscribe: 
 <mailto:gcc-patches-unsubscribe-incoming=patchwork.ozlabs.org@gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-patches/>
List-Post: <mailto:gcc-patches@gcc.gnu.org>
List-Help: <mailto:gcc-patches-help@gcc.gnu.org>
Sender: gcc-patches-owner@gcc.gnu.org
Delivered-To: mailing list gcc-patches@gcc.gnu.org

Hi,
cgraph_remove_unreachable_nodes still remove masters of clones from othr
partitions that confuse sanity checking.  Fixed by the following patch that
also cleans up how things are queued. Everything that is referenced by live
part of cgraph is enqueued, everything that is really reachable gets reachable
flag.

Bootstrapped/regtested x86_64-linux, also lto-bootstrapped and comitted.
	* ipa.c (process_references): Enqueue all referenced nodes;
	mark as reachable only non-external nodes.
	(cgraph_remove_unreachable_nodes): All referenced nodes should
	be enqueued; remove bogues node->needed check.

Index: ipa.c
===================================================================
--- ipa.c	(revision 165985)
+++ ipa.c	(working copy)
@@ -170,12 +170,11 @@ process_references (struct ipa_ref_list
 	{
 	  struct cgraph_node *node = ipa_ref_node (ref);
 	  if (!node->reachable
+	      && node->analyzed
 	      && (!DECL_EXTERNAL (node->decl)
 	          || before_inlining_p))
-	    {
-	      node->reachable = true;
-	      enqueue_cgraph_node (node, first);
-	    }
+	    node->reachable = true;
+	  enqueue_cgraph_node (node, first);
 	}
       else
 	{
@@ -304,15 +303,15 @@ cgraph_remove_unreachable_nodes (bool be
 	  if (node->reachable)
 	    {
 	      for (e = node->callees; e; e = e->next_callee)
-		if (!e->callee->reachable
-		    && node->analyzed
-		    && (!e->inline_failed || !e->callee->analyzed
-			|| (!DECL_EXTERNAL (e->callee->decl))
-			|| before_inlining_p))
-		  {
+		{
+		  if (!e->callee->reachable
+		      && node->analyzed
+		      && (!e->inline_failed
+			  || !DECL_EXTERNAL (e->callee->decl)
+			  || before_inlining_p))
 		    e->callee->reachable = true;
-		    enqueue_cgraph_node (e->callee, &first);
-		  }
+		  enqueue_cgraph_node (e->callee, &first);
+		}
 	      process_references (&node->ref_list, &first, &first_varpool, before_inlining_p);
 	    }
 
@@ -416,7 +415,7 @@ cgraph_remove_unreachable_nodes (bool be
 	      found = true;
 
 	  /* If so, we need to keep node in the callgraph.  */
-	  if (found || node->needed)
+	  if (found)
 	    {
 	      if (node->analyzed)
 		{