From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-8.5 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_PASS,URIBL_BLOCKED, USER_AGENT_MUTT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 0D3C9C10F0E for ; Mon, 15 Apr 2019 16:23:51 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id DB2A62075B for ; Mon, 15 Apr 2019 16:23:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727456AbfDOQXt (ORCPT ); Mon, 15 Apr 2019 12:23:49 -0400 Received: from mga07.intel.com ([134.134.136.100]:17178 "EHLO mga07.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726034AbfDOQXt (ORCPT ); Mon, 15 Apr 2019 12:23:49 -0400 X-Amp-Result: UNSCANNABLE X-Amp-File-Uploaded: False Received: from orsmga001.jf.intel.com ([10.7.209.18]) by orsmga105.jf.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 15 Apr 2019 09:23:49 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.60,354,1549958400"; d="scan'208";a="223720949" Received: from sjchrist-coffee.jf.intel.com (HELO linux.intel.com) ([10.54.74.181]) by orsmga001.jf.intel.com with ESMTP; 15 Apr 2019 09:23:49 -0700 Date: Mon, 15 Apr 2019 09:23:49 -0700 From: Sean Christopherson To: Liran Alon Cc: Paolo Bonzini , Radim =?utf-8?B?S3LEjW3DocWZ?= , kvm@vger.kernel.org, Wanpeng Li Subject: Re: [PATCH 4/7] KVM: lapic: Allow user to override auto-tuning of timer advancement Message-ID: <20190415162348.GD24010@linux.intel.com> References: <20190412201834.10831-1-sean.j.christopherson@intel.com> <20190412201834.10831-5-sean.j.christopherson@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.24 (2015-08-30) Sender: kvm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm@vger.kernel.org On Sun, Apr 14, 2019 at 02:35:39PM +0300, Liran Alon wrote: > > > > On 12 Apr 2019, at 23:18, Sean Christopherson wrote: > > > > Disable auto-tuning the timer advancement if the user specifies an > > explicit value via the module param. Aside from the obvious override > > capability, this also allows the KVM admin to set the advancement > > beyond the internally-capped max of 5000ns. > > > > Cc: Wanpeng Li > > Fixes: 3b8a5df6c4dc6 ("KVM: LAPIC: Tune lapic_timer_advance_ns automatically") > > Signed-off-by: Sean Christopherson > > I agree we should allow admin the ability to disable the auto-tuning. > However, I think we should keep the semantic of lapic_timer_advance_ns value. > Whether this value is used as initial-value for auto-tuning or is auto-tuning > disabled is a different knob for admin in my opinion. Therefore, I think we > should just add another module parameter which basically set the initial > value for apic->lapic_timer.timer_advance_adjust_done. I waffled on whether to add a param or do the implicit override. I opted for the implicit behavior primarily because I think most people would expect that setting lapic_timer_advance_ns to some specific value would fix the advancement at said value, as opposed to acting as a hint to the autotuning logic. In other words, I again don't have a strong opinion :-) > One could also wonder if it makes sense that whether auto-tuning will be > enabled or not varies between VMs and should not be a global variable of KVM > module? I have no opinion on this one as I don't have any direct visibility into how the auto-tuning will be used in real world deployments. > > -Liran > > > --- > > arch/x86/kvm/lapic.c | 2 ++ > > arch/x86/kvm/lapic.h | 2 ++ > > arch/x86/kvm/x86.c | 2 +- > > 3 files changed, 5 insertions(+), 1 deletion(-) > > > > diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c > > index 70a0acd98e9e..c43cd26f040b 100644 > > --- a/arch/x86/kvm/lapic.c > > +++ b/arch/x86/kvm/lapic.c > > @@ -2286,6 +2286,8 @@ int kvm_create_lapic(struct kvm_vcpu *vcpu, u32 timer_advance_ns) > > HRTIMER_MODE_ABS_PINNED); > > apic->lapic_timer.timer.function = apic_timer_fn; > > apic->lapic_timer.timer_advance_ns = timer_advance_ns; > > + if (timer_advance_ns != LAPIC_TIMER_ADVANCE_DEFAULT_NS) > > + apic->lapic_timer.timer_advance_adjust_done = true; > > > > /* > > * APIC is created enabled. This will prevent kvm_lapic_set_base from > > diff --git a/arch/x86/kvm/lapic.h b/arch/x86/kvm/lapic.h > > index 3e97f8a68967..c7233629c05b 100644 > > --- a/arch/x86/kvm/lapic.h > > +++ b/arch/x86/kvm/lapic.h > > @@ -16,6 +16,8 @@ > > #define APIC_BUS_CYCLE_NS 1 > > #define APIC_BUS_FREQUENCY (1000000000ULL / APIC_BUS_CYCLE_NS) > > > > +#define LAPIC_TIMER_ADVANCE_DEFAULT_NS 1000 > > + > > enum lapic_mode { > > LAPIC_MODE_DISABLED = 0, > > LAPIC_MODE_INVALID = X2APIC_ENABLE, > > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c > > index b303a21a2bc2..709a8bf5ae0e 100644 > > --- a/arch/x86/kvm/x86.c > > +++ b/arch/x86/kvm/x86.c > > @@ -137,7 +137,7 @@ static u32 __read_mostly tsc_tolerance_ppm = 250; > > module_param(tsc_tolerance_ppm, uint, S_IRUGO | S_IWUSR); > > > > /* lapic timer advance (tscdeadline mode only) in nanoseconds */ > > -static u32 __read_mostly lapic_timer_advance_ns = 1000; > > +static u32 __read_mostly lapic_timer_advance_ns = LAPIC_TIMER_ADVANCE_DEFAULT_NS; > > module_param(lapic_timer_advance_ns, uint, S_IRUGO | S_IWUSR); > > > > static bool __read_mostly vector_hashing = true; > > -- > > 2.21.0 > > >