From mboxrd@z Thu Jan 1 00:00:00 1970 From: Waiman Long Subject: Re: [PATCH v10 18/19] pvqspinlock, x86: Enable PV qspinlock PV for KVM Date: Thu, 08 May 2014 13:54:07 -0400 Message-ID: <536BC4BF.9030603@hp.com> References: <1399474907-22206-1-git-send-email-Waiman.Long@hp.com> <1399474907-22206-19-git-send-email-Waiman.Long@hp.com> <20140507190706.GA16333@phenom.dumpdata.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20140507190706.GA16333@phenom.dumpdata.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: virtualization-bounces@lists.linux-foundation.org Errors-To: virtualization-bounces@lists.linux-foundation.org To: Konrad Rzeszutek Wilk Cc: linux-arch@vger.kernel.org, Rik van Riel , Raghavendra K T , Gleb Natapov , kvm@vger.kernel.org, Oleg Nesterov , Peter Zijlstra , Scott J Norton , x86@kernel.org, Paolo Bonzini , linux-kernel@vger.kernel.org, virtualization@lists.linux-foundation.org, Ingo Molnar , Chegu Vinod , David Vrabel , "H. Peter Anvin" , xen-devel@lists.xenproject.org, Thomas Gleixner , "Paul E. McKenney" , Linus Torvalds , Boris Ostrovsky List-Id: linux-arch.vger.kernel.org On 05/07/2014 03:07 PM, Konrad Rzeszutek Wilk wrote: >> Raghavendra KT had done some performance testing on this patch with >> the following results: >> >> Overall we are seeing good improvement for pv-unfair version. >> >> System: 32 cpu sandybridge with HT on (4 node with 32 GB each) >> Guest : 8GB with 16 vcpu/VM. >> Average was taken over 8-10 data points. >> >> Base = 3.15-rc2 with PRAVIRT_SPINLOCK = y >> >> A = 3.15-rc2 + qspinlock v9 patch with QUEUE_SPINLOCK = y >> PRAVIRT_SPINLOCK = y PARAVIRT_UNFAIR_LOCKS = y (unfair lock) >> >> B = 3.15-rc2 + qspinlock v9 patch with QUEUE_SPINLOCK = y >> PRAVIRT_SPINLOCK = n PARAVIRT_UNFAIR_LOCKS = n >> (queue spinlock without paravirt) >> >> C = 3.15-rc2 + qspinlock v9 patch with QUEUE_SPINLOCK = y >> PRAVIRT_SPINLOCK = y PARAVIRT_UNFAIR_LOCKS = n >> (queue spinlock with paravirt) > Could you do s/PRAVIRT/PARAVIRT/ please? > Sorry for the typo, I didn't check the text carefully enough when I cut-and-paste it from Raghavendra's email. >> Ebizzy %improvements >> ==================== >> overcommit A B C >> 0.5x 4.4265 2.0611 1.5824 >> 1.0x 0.9015 -7.7828 4.5443 >> 1.5x 46.1162 -2.9845 -3.5046 >> 2.0x 99.8150 -2.7116 4.7461 > Considering B sucks Yes, I don't expect the plain qspinlock code will perform well in a guest without either unfair or pvspinlock support. -Longman From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from g2t2352.austin.hp.com ([15.217.128.51]:51305 "EHLO g2t2352.austin.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754658AbaEHRyd (ORCPT ); Thu, 8 May 2014 13:54:33 -0400 Message-ID: <536BC4BF.9030603@hp.com> Date: Thu, 08 May 2014 13:54:07 -0400 From: Waiman Long MIME-Version: 1.0 Subject: Re: [PATCH v10 18/19] pvqspinlock, x86: Enable PV qspinlock PV for KVM References: <1399474907-22206-1-git-send-email-Waiman.Long@hp.com> <1399474907-22206-19-git-send-email-Waiman.Long@hp.com> <20140507190706.GA16333@phenom.dumpdata.com> In-Reply-To: <20140507190706.GA16333@phenom.dumpdata.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-arch-owner@vger.kernel.org List-ID: To: Konrad Rzeszutek Wilk Cc: Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , Peter Zijlstra , linux-arch@vger.kernel.org, x86@kernel.org, linux-kernel@vger.kernel.org, virtualization@lists.linux-foundation.org, xen-devel@lists.xenproject.org, kvm@vger.kernel.org, Paolo Bonzini , Boris Ostrovsky , "Paul E. McKenney" , Rik van Riel , Linus Torvalds , Raghavendra K T , David Vrabel , Oleg Nesterov , Gleb Natapov , Scott J Norton , Chegu Vinod Message-ID: <20140508175407.gI5PFITEcnGd-di27DESAS5GP05pou-zeYvUT4tvntU@z> On 05/07/2014 03:07 PM, Konrad Rzeszutek Wilk wrote: >> Raghavendra KT had done some performance testing on this patch with >> the following results: >> >> Overall we are seeing good improvement for pv-unfair version. >> >> System: 32 cpu sandybridge with HT on (4 node with 32 GB each) >> Guest : 8GB with 16 vcpu/VM. >> Average was taken over 8-10 data points. >> >> Base = 3.15-rc2 with PRAVIRT_SPINLOCK = y >> >> A = 3.15-rc2 + qspinlock v9 patch with QUEUE_SPINLOCK = y >> PRAVIRT_SPINLOCK = y PARAVIRT_UNFAIR_LOCKS = y (unfair lock) >> >> B = 3.15-rc2 + qspinlock v9 patch with QUEUE_SPINLOCK = y >> PRAVIRT_SPINLOCK = n PARAVIRT_UNFAIR_LOCKS = n >> (queue spinlock without paravirt) >> >> C = 3.15-rc2 + qspinlock v9 patch with QUEUE_SPINLOCK = y >> PRAVIRT_SPINLOCK = y PARAVIRT_UNFAIR_LOCKS = n >> (queue spinlock with paravirt) > Could you do s/PRAVIRT/PARAVIRT/ please? > Sorry for the typo, I didn't check the text carefully enough when I cut-and-paste it from Raghavendra's email. >> Ebizzy %improvements >> ==================== >> overcommit A B C >> 0.5x 4.4265 2.0611 1.5824 >> 1.0x 0.9015 -7.7828 4.5443 >> 1.5x 46.1162 -2.9845 -3.5046 >> 2.0x 99.8150 -2.7116 4.7461 > Considering B sucks Yes, I don't expect the plain qspinlock code will perform well in a guest without either unfair or pvspinlock support. -Longman