From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sheng Yang Subject: Re: [PATCH] KVM: VMX: Execute WBINVD to keep data consistency with assigned devices Date: Fri, 25 Jun 2010 18:20:00 +0800 Message-ID: <201006251820.00783.sheng@linux.intel.com> References: <1277452623-24046-1-git-send-email-sheng@linux.intel.com> <4C246EBB.1020909@siemens.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Cc: Avi Kivity , Marcelo Tosatti , kvm@vger.kernel.org, "Yaozu (Eddie) Dong" To: Jan Kiszka Return-path: Received: from mga01.intel.com ([192.55.52.88]:2839 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753410Ab0FYKVT (ORCPT ); Fri, 25 Jun 2010 06:21:19 -0400 In-Reply-To: <4C246EBB.1020909@siemens.com> Sender: kvm-owner@vger.kernel.org List-ID: On Friday 25 June 2010 16:54:19 Jan Kiszka wrote: > Sheng Yang wrote: > > Some guest device driver may leverage the "Non-Snoop" I/O, and explicitly > > WBINVD or CLFLUSH to a RAM space. Since migration may occur before WBINVD > > or CLFLUSH, we need to maintain data consistency either by: > > 1: flushing cache (wbinvd) when the guest is scheduled out if there is no > > wbinvd exit, or > > 2: execute wbinvd on all dirty physical CPUs when guest wbinvd exits. > > > > For wbinvd VMExit capable processors, we issue IPIs to all physical CPUs > > to do wbinvd, for we can't easily tell which physical CPUs are "dirty". > > wbinvd is a heavy weapon in the hands of a guest. Even if it is limited > to pass-through scenarios, do we really need to bother all physical host > CPUs with potential multi-millisecond stalls? Think of VMs only running > on a subset of CPUs (e.g. to isolate latency sources). I would suggest > to track the physical CPU usage of VCPUs between two wbinvd requests and > only send the wbinvd IPI to that set. OK, would try to make it more specific(and complex)... > > Also, I think the code is still too much vmx-focused. Only the trapping > should be vendor specific, the rest generic. OK, would consider it. -- regards Yang, Sheng > > Jan