Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755147Ab0AUUcd (ORCPT ); Thu, 21 Jan 2010 15:32:33 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754966Ab0AUUcb (ORCPT ); Thu, 21 Jan 2010 15:32:31 -0500 Received: from mx1.redhat.com ([209.132.183.28]:63075 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754879Ab0AUUcb (ORCPT ); Thu, 21 Jan 2010 15:32:31 -0500 Date: Thu, 21 Jan 2010 21:32:07 +0100 From: Oleg Nesterov To: Martin Schwidefsky Cc: Roland McGrath , caiqian@redhat.com, Heiko Carstens , Jan Kratochvil , linux-kernel@vger.kernel.org, linux-s390@vger.kernel.org, utrace-devel@redhat.com Subject: Re: s390 && user_enable_single_step() (Was: odd utrace testing results on s390x) Message-ID: <20100121203207.GA20050@redhat.com> References: <1503844142.2061111261478093776.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> <1257887498.2061171261478252049.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> <20100104155225.GA16650@redhat.com> <20100104171626.22ea2d9c@mschwide.boeblingen.de.ibm.com> <20100104181412.GA21146@redhat.com> <20100104211147.4CC94D532@magilla.sf.frob.com> <20100105105030.66bb8a0a@mschwide.boeblingen.de.ibm.com> <20100106205633.700CC134D@magilla.sf.frob.com> <20100107100050.31724463@mschwide.boeblingen.de.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20100107100050.31724463@mschwide.boeblingen.de.ibm.com> User-Agent: Mutt/1.5.18 (2008-05-17) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1142 Lines: 28 On 01/07, Martin Schwidefsky wrote: > > On Wed, 6 Jan 2010 12:56:33 -0800 (PST) > Roland McGrath wrote: > > > In other circumstances with utrace, it is very possible to wind up with > > user_disable_single_step being called superfluously when there was no > > stop (and so not necessarily any context switch or other high overhead). > > On other machines, user_disable_single_step is pretty cheap even where > > user_enable_single_step is quite costly. Given how simple and cheap it > > is to short-circuit the excess work on s390, I think it is worthwhile. > > We could use the same compare of the control registers as the code in > __switch_to. See below. FYI, I tested your c3311c13adc1021e986fef12609ceb395ffc5014 commit which does this optimization (compared to the patch you sent previously), it works fine. But please see another email I am going to send... Oleg. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/