From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 16436 invoked by alias); 26 Feb 2008 19:04:02 -0000 Received: (qmail 16427 invoked by uid 22791); 26 Feb 2008 19:04:02 -0000 X-Spam-Check-By: sourceware.org Received: from mk-outboundfilter-3.mail.uk.tiscali.com (HELO mk-outboundfilter-3.mail.uk.tiscali.com) (212.74.114.23) by sourceware.org (qpsmtpd/0.31) with ESMTP; Tue, 26 Feb 2008 19:03:42 +0000 X-Trace: 49245379/mk-outboundfilter-1.mail.uk.tiscali.com/F2S/$NILDRAM-ACCEPTED/nildram-customers/81.6.217.218 X-SBRS: None X-RemoteIP: 81.6.217.218 X-IP-MAIL-FROM: richard@nildram.co.uk X-IP-BHB: Once X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: Ao8CAH7ww0dRBtna/2dsb2JhbACBVat/ X-IP-Direction: IN Received: from 81-6-217-218.gotadsl.co.uk (HELO firetop.home) ([81.6.217.218]) by smtp.f2s.tiscali.co.uk with ESMTP/TLS/AES256-SHA; 26 Feb 2008 19:03:39 +0000 Received: from richard by firetop.home with local (Exim 4.63) (envelope-from ) id 1JU55X-0006jH-Ee; Tue, 26 Feb 2008 19:03:39 +0000 To: Peter Bergner Mail-Followup-To: Peter Bergner ,gcc-patches@gcc.gnu.org, rsandifo@nildram.co.uk Cc: gcc-patches@gcc.gnu.org Subject: Re: [PATCH] PR35371 GCSE loses track of REG_POINTER attribute References: <20080225222624.GA26857@vervain.rchland.ibm.com> <87skzfzkvh.fsf@firetop.home> <1204052325.7014.2.camel@otta> From: Richard Sandiford Date: Tue, 26 Feb 2008 19:45:00 -0000 In-Reply-To: <1204052325.7014.2.camel@otta> (Peter Bergner's message of "Tue\, 26 Feb 2008 12\:58\:45 -0600") Message-ID: <87fxvfzgj8.fsf@firetop.home> User-Agent: Gnus/5.11 (Gnus v5.11) Emacs/22.1 (gnu/linux) 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: List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org X-SW-Source: 2008-02/txt/msg01310.txt.bz2 Peter Bergner writes: > On Tue, 2008-02-26 at 17:29 +0000, Richard Sandiford wrote: >> Minor suggestion, but maybe this could go in emit-rtl.c, under a more >> generic name? Given the performance impact of losing pointer info, >> it would be nice to have a defined API for creating a register that's >> like another. > > Having tracked a similar problem in another file, I was thinking along > the same lines. I'm bad at names though. Care to suggest a name for > the new function? Me too, which is why I copped out. ;) Oh well. How about gen_reg_rtx_like? Richard