From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965289AbWHOIKi (ORCPT ); Tue, 15 Aug 2006 04:10:38 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S965291AbWHOIKi (ORCPT ); Tue, 15 Aug 2006 04:10:38 -0400 Received: from cantor.suse.de ([195.135.220.2]:19870 "EHLO mx1.suse.de") by vger.kernel.org with ESMTP id S965289AbWHOIKh (ORCPT ); Tue, 15 Aug 2006 04:10:37 -0400 To: David Howells Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org# Subject: Re: [RHEL5 PATCH 1/4] Provide fallback full 64-bit divide/modulus ops for gcc References: <20060814211504.27190.10491.stgit@warthog.cambridge.redhat.com> <20060814211507.27190.61876.stgit@warthog.cambridge.redhat.com> From: Andi Kleen Date: 15 Aug 2006 10:10:32 +0200 In-Reply-To: <20060814211507.27190.61876.stgit@warthog.cambridge.redhat.com> Message-ID: User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org David Howells writes: > Provide simple, reasonably quick full 64-bit divide and modulus ops for gcc to > call behind the scenes as: > > __udivmoddi4 > __udivdi3 > __umoddi3 At least Linus' traditional argument against this is that it's better to open code these (do_div) so that it's clear to the coder that they are really costly. -Andi