From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753878Ab1HCJj4 (ORCPT ); Wed, 3 Aug 2011 05:39:56 -0400 Received: from cn.fujitsu.com ([222.73.24.84]:56189 "EHLO song.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1753640Ab1HCJjw convert rfc822-to-8bit (ORCPT ); Wed, 3 Aug 2011 05:39:52 -0400 Message-ID: <4E3917DC.9040902@cn.fujitsu.com> Date: Wed, 03 Aug 2011 17:41:48 +0800 From: Xiao Guangrong User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.2.17) Gecko/20110428 Fedora/3.1.10-1.fc15 Thunderbird/3.1.10 MIME-Version: 1.0 To: Avi Kivity CC: Marcelo Tosatti , LKML , KVM Subject: Re: [PATCH v2 02/12] KVM: x86: tag the instructions which are used to write page table References: <4E37DA49.1040000@cn.fujitsu.com> <4E37DA73.7010908@cn.fujitsu.com> <4E37EB8E.6080207@redhat.com> <4E38E477.2000208@cn.fujitsu.com> <4E39021F.8070508@redhat.com> <4E3913CC.7010907@cn.fujitsu.com> <4E391417.6050503@redhat.com> In-Reply-To: <4E391417.6050503@redhat.com> X-MIMETrack: Itemize by SMTP Server on mailserver/fnst(Release 8.5.1FP4|July 25, 2010) at 2011-08-03 17:38:46, Serialize by Router on mailserver/fnst(Release 8.5.1FP4|July 25, 2010) at 2011-08-03 17:38:47 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 08/03/2011 05:25 PM, Avi Kivity wrote: > On 08/03/2011 12:24 PM, Xiao Guangrong wrote: >> > Maybe it's better to emulate if we can't find a fix for that. >> > >> > One way would be to emulate every 20 instructions; this breaks us out of the loop but reduces costly emulations to 5%. >> > >> >> After much thought about this, may be this optimization is not good since: >> - it is little complex >> - this optimization is only applied to the instruction emulation caused by #PF >> - it does not improve too much: >> if we emulate the instruction, we need to do: >> - decode instruction >> - emulate it >> - zap shadow pages >> And do this, it can return to the guest, the guest can run the next instruction >> >> if we retry the instruction, we need to do: >> - decode instruction >> - zap shadow pages >> then return to the guest and retry the instruction, however, we will get page fault >> again(since the mapping is still read-only), so we will get another VM-exit and need >> to do: >> # trigger page fault >> - handle the page fault and change the mapping to writable >> - retry the instruction >> until now, the guest can run the next instruction >> >> So, i do not think the new way is better, your opinion？ > > We can change to writeable and zap in the same exit, no? Basically call page_fault() again after zapping. > OK. :-) Um, how about cache the last eip when we do this optimization, if we meet the same eip, we can break out the potential infinite loop?