From patchwork Wed Aug  4 21:08:34 2010
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
X-Patchwork-Submitter: Bernd Schmidt <bernds@codesourcery.com>
X-Patchwork-Id: 60896
Return-Path: 
 <gcc-patches-return-269825-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 11794B6EE9
	for <incoming@patchwork.ozlabs.org>;
	Thu,  5 Aug 2010 07:08:53 +1000 (EST)
Received: (qmail 4972 invoked by alias); 4 Aug 2010 21:08:51 -0000
Received: (qmail 4956 invoked by uid 22791); 4 Aug 2010 21:08:51 -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 mail.codesourcery.com (HELO mail.codesourcery.com)
	(38.113.113.100) by sourceware.org (qpsmtpd/0.43rc1) with
	ESMTP; Wed, 04 Aug 2010 21:08:46 +0000
Received: (qmail 32007 invoked from network); 4 Aug 2010 21:08:45 -0000
Received: from unknown (HELO ?84.152.241.194?) (bernds@127.0.0.2) by
	mail.codesourcery.com with ESMTPA; 4 Aug 2010 21:08:45 -0000
Message-ID: <4C59D6D2.9040509@codesourcery.com>
Date: Wed, 04 Aug 2010 23:08:34 +0200
From: Bernd Schmidt <bernds@codesourcery.com>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US;
	rv:1.9.2.7) Gecko/20100724 Thunderbird/3.1.1
MIME-Version: 1.0
To: GCC Patches <gcc-patches@gcc.gnu.org>
Subject: Word DCE DEBUG_INSN problems
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

Because the old byte_dce and byte_lr code was disabled, it was never
updated to ignore DEBUG_INSNs, which caused regressions for some folks.
 I committed the following as obvious after bootstrap and test on
i686-linux.


Bernd

Index: ChangeLog
===================================================================
--- ChangeLog	(revision 162880)
+++ ChangeLog	(working copy)
@@ -1,3 +1,9 @@
+2010-08-04  Bernd Schmidt  <bernds@codesourcery.com>
+
+	PR rtl-optimization/45162
+	* df-problems.c (df_word_lr_bb_local_compute): Ignore DEBUG_INSNs.
+	* dce.c (word_dce_process_block): Likewise.
+
 2010-08-04  Steve Ellcey  <sje@cup.hp.com>
 
 	PR target/44583
Index: dce.c
===================================================================
--- dce.c	(revision 162821)
+++ dce.c	(working copy)
@@ -795,7 +795,7 @@ word_dce_process_block (basic_block bb, 
   bitmap_copy (local_live, DF_WORD_LR_OUT (bb));
 
   FOR_BB_INSNS_REVERSE (bb, insn)
-    if (INSN_P (insn))
+    if (NONDEBUG_INSN_P (insn))
       {
 	bool any_changed;
 	/* No matter if the instruction is needed or not, we remove
Index: df-problems.c
===================================================================
--- df-problems.c	(revision 162821)
+++ df-problems.c	(working copy)
@@ -2472,7 +2472,7 @@ df_word_lr_bb_local_compute (unsigned in
     {
       unsigned int uid = INSN_UID (insn);
 
-      if (!INSN_P (insn))
+      if (!NONDEBUG_INSN_P (insn))
 	continue;
       for (def_rec = DF_INSN_UID_DEFS (uid); *def_rec; def_rec++)
 	{