From mboxrd@z Thu Jan  1 00:00:00 1970
From: Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH] KVM: nVMX: nested TPR shadow/threshold emulation
Date: Fri, 01 Aug 2014 08:35:03 +0200
Message-ID: <53DB3517.1010408@redhat.com>
References: <1406721880-5490-1-git-send-email-wanpeng.li@linux.intel.com> <53D90D5A.3060404@redhat.com> <20140731080353.GA3033@kernel> <53DA0B91.1010703@redhat.com> <A9667DDFB95DB7438FA9D7D576C3D87E0AB4D3F5@SHSMSX104.ccr.corp.intel.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Cc: Marcelo Tosatti <mtosatti@redhat.com>,
	Gleb Natapov <gleb@kernel.org>, Bandan Das <bsd@redhat.com>,
	"kvm@vger.kernel.org" <kvm@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
To: "Zhang, Yang Z" <yang.z.zhang@intel.com>,
	Wanpeng Li <wanpeng.li@linux.intel.com>,
	Jan Kiszka <jan.kiszka@siemens.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <A9667DDFB95DB7438FA9D7D576C3D87E0AB4D3F5@SHSMSX104.ccr.corp.intel.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org

Il 01/08/2014 02:57, Zhang, Yang Z ha scritto:
> > TPR_THRESHOLD will be likely written as zero, but the processor will
> > never use it anyway.  It's just a small optimization because
> > nested_cpu_has(vmcs12, CPU_BASED_TPR_SHADOW) will almost always be true.
> 
> Theoretically, you are right. But we should not expect all VMMs
> follow it. It is not worth to violate the SDM just for saving two or
> three instructions' cost.

Yes, you do need an "if (cpu_has_vmx_tpr_shadow())" around the
vmcs_write32.  But still, checking nested_cpu_has is not strictly
necessary.  Right now they both are a single AND, but I have plans to
change all of the cpu_has_*() checks to static keys.

Paolo