From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-patches-return-213266-listarch-gcc-patches=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 11108 invoked by alias); 5 Mar 2008 16:27:15 -0000
Received: (qmail 11100 invoked by uid 22791); 5 Mar 2008 16:27:15 -0000
X-Spam-Check-By: sourceware.org
Received: from mx2.suse.de (HELO mx2.suse.de) (195.135.220.15)     by sourceware.org (qpsmtpd/0.31) with ESMTP; Wed, 05 Mar 2008 16:26:55 +0000
Received: from Relay1.suse.de (mail2.suse.de [195.135.221.8]) 	(using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) 	(No client certificate requested) 	by mx2.suse.de (Postfix) with ESMTP id C9D3D3994B; 	Wed,  5 Mar 2008 17:26:52 +0100 (CET)
Date: Wed, 05 Mar 2008 16:27:00 -0000
From: Richard Guenther <rguenther@suse.de>
To: Richard Kenner <kenner@vlsi1.ultra.nyu.edu>
Cc: gcc-patches@gcc.gnu.org, joseph@codesourcery.com, matz@suse.de
Subject: Re: [PATCH] Make alias_sets_conflict_p less conservative
In-Reply-To: <10803051624.AA07978@vlsi1.ultra.nyu.edu>
Message-ID: <Pine.LNX.4.64.0803051724530.4133@zhemvz.fhfr.qr>
References: <Pine.LNX.4.64.0803042155430.4133@zhemvz.fhfr.qr>                 <10803042355.AA00168@vlsi1.ultra.nyu.edu>                 <4aca3dc20803041620g70075ae3h40d22a4531ad7566@mail.gmail.com>                 <10803050136.AA01130@vlsi1.ultra.nyu.edu>                   <Pine.LNX.4.64.0803051057420.4133@zhemvz.fhfr.qr>                 <10803051132.AA03852@vlsi1.ultra.nyu.edu>                    <Pine.LNX.4.64.0803051241530.4133@zhemvz.fhfr.qr>                <10803051226.AA04718@vlsi1.ultra.nyu.edu>                <Pine.LNX.4.64.0803051336170.4133@zhemvz.fhfr.qr>                    <10803051256.AA05161@vlsi1.ultra.nyu.edu>                <Pine.LNX.4.64.0803051452510.4133@zhemvz.fhfr.qr>                 <10803051426.AA05981@vlsi1.ultra.nyu.edu>                    <Pine.LNX.4.64.0803051559450.4133@zhemvz.fhfr.qr>                <10803051515.AA06678@vlsi1.ultra.nyu.edu>                <Pine.LNX.4.64.0803051614490.4133@zhemvz.fhfr.qr>                 <Pine.LNX.4.64.0803051625050.20583@wotan.suse.de>                 <Pine.LNX.4.64.080305163    <Pine.LNX.4.64.0803051715100.4133@zhemvz.fhfr.qr>  <10803051624.AA07978@vlsi1.ultra.nyu.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-patches.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
X-SW-Source: 2008-03/txt/msg00336.txt.bz2

On Wed, 5 Mar 2008, Richard Kenner wrote:

> > Err, right.  But tree aliasing is computing a representation for TBAA,
> > using alias set conflict queries but _not_ looking at individual accesses.
> 
> That seems quite wrong to me.
> 
> > So, is alias_set_subset_of () wrong as currently implemented?  Does
> > it lack the ase->has_zero_child test in your opinion?  Otherwise I'll
> > try using that for these queries.
> 
> I don't know enough to answer those questions.

That is sad, because you knew enough to fix exactly the same code
in alias_sets_conflict_p.

> What I *do* know is that if you have x.y, with y addressable, any code
> that looks at the alias set of X *for any reason* is broken.

Well, you obviously do not have enough knowledge of the alias 
representation used in tree-ssa and I won't try to start to explain.
If you want, take the testcase in PR27799, build with
-O2 -fdump-tree-salias-vops-details and look at the 052t.salias dump.

Richard.