From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1030894AbWKOTEk (ORCPT ); Wed, 15 Nov 2006 14:04:40 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1030893AbWKOTEk (ORCPT ); Wed, 15 Nov 2006 14:04:40 -0500 Received: from mx2.mail.elte.hu ([157.181.151.9]:37086 "EHLO mx2.mail.elte.hu") by vger.kernel.org with ESMTP id S1030895AbWKOTEj (ORCPT ); Wed, 15 Nov 2006 14:04:39 -0500 Date: Wed, 15 Nov 2006 20:03:08 +0100 From: Ingo Molnar To: Jeremy Fitzhardinge Cc: Arjan van de Ven , akpm@osdl.org, ak@suse.de, linux-kernel@vger.kernel.org, Michael.Fetterman@cl.cam.ac.uk, Ian Campbell Subject: Re: i386 PDA patches use of %gs Message-ID: <20061115190308.GA9303@elte.hu> References: <1158046540.2992.5.camel@laptopd505.fenrus.org> <45075829.701@goop.org> <20060913095942.GA10075@elte.hu> <45082F1C.8000003@goop.org> <20061115182613.GA2227@elte.hu> <455B5ED8.5090005@goop.org> <20061115184315.GA5078@elte.hu> <455B611C.3010503@goop.org> <20061115184936.GA6389@elte.hu> <455B63D8.7000601@goop.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <455B63D8.7000601@goop.org> User-Agent: Mutt/1.4.2.2i X-ELTE-SpamScore: -4.4 X-ELTE-SpamLevel: X-ELTE-SpamCheck: no X-ELTE-SpamVersion: ELTE 2.0 X-ELTE-SpamCheck-Details: score=-4.4 required=5.9 tests=ALL_TRUSTED,AWL,BAYES_00 autolearn=no SpamAssassin version=3.0.3 -3.3 ALL_TRUSTED Did not pass through any untrusted hosts -2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1% [score: 0.0000] 1.5 AWL AWL: From: address is in the auto white-list X-ELTE-VirusStatus: clean Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org * Jeremy Fitzhardinge wrote: > Ingo Molnar wrote: > > i'd not even use glibc's %gs but set up two separate selectors. > > (that's a more controlled experiment - someone might run a non-TLS > > glibc, etc.) > > > > Well, in that case they probably don't care whether the kernel uses > %fs or %gs ;) > > But either way, this doesn't have much bearing on Eric's test; we'd be > only talking about a few ns per kernel exit, rather than 5% for > read/write. if the timings are different then it very much has bearing on the argument that i made against the current i386 PDA patchset, that mixed use segments are suboptimal. So i'm NAK-ing the i386 PDA patchset until this has been properly measured (and fixed if needed). Ingo