From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964865AbWA0AbM (ORCPT ); Thu, 26 Jan 2006 19:31:12 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S964843AbWA0AbM (ORCPT ); Thu, 26 Jan 2006 19:31:12 -0500 Received: from hiauly1.hia.nrc.ca ([132.246.100.193]:31247 "EHLO hiauly1.hia.nrc.ca") by vger.kernel.org with ESMTP id S964797AbWA0AbI (ORCPT ); Thu, 26 Jan 2006 19:31:08 -0500 Message-Id: <200601270028.k0R0STBB021468@hiauly1.hia.nrc.ca> Subject: Re: [parisc-linux] Re: [PATCH 3/6] C-language equivalents of To: grundler@parisc-linux.org (Grant Grundler) Date: Thu, 26 Jan 2006 19:28:29 -0500 (EST) From: "John David Anglin" Cc: grundler@parisc-linux.org, mita@miraclelinux.com, linux-kernel@vger.kernel.org, ink@jurassic.park.msu.ru, spyro@f2s.com, dev-etrax@axis.com, dhowells@redhat.com, ysato@users.sourceforge.jp, torvalds@osdl.org, linux-ia64@vger.kernel.org, takata@linux-m32r.org, linux-m68k@vger.kernel.org, gerg@uclinux.org, linux-mips@linux-mips.org, parisc-linux@parisc-linux.org, linuxppc-dev@ozlabs.org, linux390@de.ibm.com, linuxsh-dev@lists.sourceforge.net, linuxsh-shmedia-dev@lists.sourceforge.net, sparclinux@vger.kernel.org, ultralinux@vger.kernel.org, uclinux-v850@lsi.nec.co.jp, ak@suse.de, chris@zankel.net In-Reply-To: <20060126230443.GC13632@colo.lackof.org> from "Grant Grundler" at Jan 26, 2006 04:04:43 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > Yeah, about the same for parisc. > > > Clearly the smallest of the lot with the smallest register pressure, > > being the best candidate out of the lot, whether we inline it or not. > > Agreed. But I expect parisc will have to continue using it's asm > sequence and ignore the generic version. AFAIK, the compiler isn't that > good with instruction nullification and I have other issues I'd > rather work on. I looked at the assembler code generated on parisc with 4.1.0 (prerelease). The toernig code is definitely inferior. The mita sequence is four instructions longer than the arm sequence, but it didn't have any branches. The arm sequence has four branches. Thus, it's not clear to me which would perform better in the real world. There were no nullified instructions generated for any of the sequences. However, neither is as good as the handcraft asm sequence currently being used. Dave -- J. David Anglin dave.anglin@nrc-cnrc.gc.ca National Research Council of Canada (613) 990-0752 (FAX: 952-6602)