From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 18493 invoked by alias); 22 Oct 2004 19:12:53 -0000 Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org Received: (qmail 18486 invoked from network); 22 Oct 2004 19:12:52 -0000 Received: from unknown (HELO mx1.redhat.com) (66.187.233.31) by sourceware.org with SMTP; 22 Oct 2004 19:12:52 -0000 Received: from int-mx1.corp.redhat.com (int-mx1.corp.redhat.com [172.16.52.254]) by mx1.redhat.com (8.12.11/8.12.11) with ESMTP id i9MJClZp019910; Fri, 22 Oct 2004 15:12:52 -0400 Received: from [172.16.83.146] (vpn83-146.boston.redhat.com [172.16.83.146]) by int-mx1.corp.redhat.com (8.11.6/8.11.6) with ESMTP id i9MJCkr00337; Fri, 22 Oct 2004 15:12:46 -0400 Subject: Re: Fix a tcb crash and a potential bug on mainline From: Jeffrey A Law Reply-To: law@redhat.com To: Steven Bosscher Cc: Andrew MacLeod , gcc-patches , Diego Novillo In-Reply-To: <200410201527.38061.stevenb@suse.de> References: <200410182303.52997.stevenb@suse.de> <1098227633.2915.166.camel@localhost.localdomain> <1098278085.5695.3886.camel@pain> <200410201527.38061.stevenb@suse.de> Content-Type: text/plain Organization: Red Hat, Inc Message-Id: <1098472364.2915.216.camel@localhost.localdomain> Mime-Version: 1.0 Date: Fri, 22 Oct 2004 19:15:00 -0000 Content-Transfer-Encoding: 7bit X-SW-Source: 2004-10/txt/msg01950.txt.bz2 On Wed, 2004-10-20 at 07:27, Steven Bosscher wrote: > > > 2 - The issue may also exist outside cleanup_control_flow, so why does > > remove_phi_arg_num have to remove the PHI node when is removes the last > > argument. > > That is what I believe to be the problem. Jeff thinks this is "the > right thing" to do. What I think is right is to ensure that we release the dead PHIs and their associated PHI_RESULTs. What Andrew made clearer in his message is that we would get a chance to collect them when we remove unreachable blocks since the only time we can have a PHI with no args is when we have unreachable blocks. I'm pretty sure we don't release PHIs and SSA_NAMEs in that code, but it shouldn't be terribly hard to add it in a way that is safe (since the unreachable block removal code is used by both trees and rtl). > > > > Why not allow the PHI node to exist with no arguments. > > That would work too, I guess. And I like that a lot better than removing the PHI and leaving in references to the result.