From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH v2 0/3] Fix task switches into/out of VM86 Date: Mon, 30 Jan 2012 15:27:41 +0200 Message-ID: <4F269ACD.80003@redhat.com> References: <1327692216-20450-1-git-send-email-kwolf@redhat.com> <20120127195239.GB2938@redhat.com> <4F265961.3070501@redhat.com> <20120130085541.GI30469@redhat.com> <4F267254.7040900@redhat.com> <4F2674BB.2000608@redhat.com> <20120130105022.GC23536@redhat.com> <4F268626.5090209@redhat.com> <20120130121604.GG23536@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Kevin Wolf , kvm@vger.kernel.org, joerg.roedel@amd.com, yoshikawa.takuya@oss.ntt.co.jp, mtosatti@redhat.com To: Gleb Natapov Return-path: Received: from mx1.redhat.com ([209.132.183.28]:5370 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751722Ab2A3N1r (ORCPT ); Mon, 30 Jan 2012 08:27:47 -0500 In-Reply-To: <20120130121604.GG23536@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 01/30/2012 02:16 PM, Gleb Natapov wrote: > On Mon, Jan 30, 2012 at 01:59:34PM +0200, Avi Kivity wrote: > > On 01/30/2012 12:50 PM, Gleb Natapov wrote: > > > On Mon, Jan 30, 2012 at 12:45:15PM +0200, Avi Kivity wrote: > > > > On 01/30/2012 12:35 PM, Kevin Wolf wrote: > > > > > Am 30.01.2012 09:55, schrieb Gleb Natapov: > > > > > > On Mon, Jan 30, 2012 at 09:48:33AM +0100, Kevin Wolf wrote: > > > > > >> Am 27.01.2012 20:52, schrieb Gleb Natapov: > > > > > >>> On Fri, Jan 27, 2012 at 08:23:33PM +0100, Kevin Wolf wrot= e: > > > > > >>>> I believe this should work with both VMX and SVM now. Gl= eb, J=C3=B6rg, can one of > > > > > >>>> you test this with SVM? I did some testing on my buggy p= rocessor and it looks > > > > > >>>> as good as it gets, but it would be better if you could = confirm. > > > > > >>>> > > > > > >>> You forgot to set cpl to 3 in vmcb in svm_set_rflags() wh= en vm86 is enabled, no? > > > > > >> > > > > > >> SVM updates the CPL when the segment selector for CS is lo= aded. From a > > > > > >> svm.c POV, segment selectors are updated immediately after= set_rflags, > > > > > >> so it wouldn't really make a difference to do it twice. > > > > > >> > > > > > > It is too subtle to rely on that. The fact is that checking= cpl after > > > > > > set_rflags provides incorrect value. This better be fixed. > > > > > > > > > > Depends on what value you consider to be correct between relo= ading > > > > > eflags and reloading cs. I think it's logical and more consis= tent to say > > > > > that CPL only changes when cs is reloaded, but you could argu= e that it's > > > > > effective with the reload of rflags. It doesn't make a differ= ence to > > > > > guests, so we can decide to choose whatever we like. > > > >=20 > > > > It's best to make it independent (like svm, and force vmx to em= ulate > > > > this behaviour). Real mode forces cpl to 0, vm86 forces cpl to= 3, > > > > protected mode (and I think long mode) uses cs.rpl. > > > This is what vmx does, not svm.=20 > >=20 > > That's the architectural definition, except for mode switch sequenc= es.=20 > > vmx implements it directly which means that mode switch sequences > > sometimes fail, either in guest software (setting cr0.pe while cs &= 3 !=3D > > 0) or in "microcode" (emulate.c). > >=20 > > > svm checks vmcb->cpl that can be > > > outdated during emulation. > >=20 > > This decoupling is actually helpful, since you can defer the cpl ch= ange > > until the end of the switch, and avoid inconsistencies like those > > checked by cs_ss_rpl_check(). > >=20 > I am not saying it is not helpful. The fact that it exists tells us > that dpl and cpl are not always the same. But cpl change should not = be > delayed until the end of the switch! Mode switch happens in the middl= e of > a task switch. Task switch happens in 3 stages according to the spec.= If > error happens during the first one (steps 1-11) it is handled by an o= ld > task, if error happens during second stage (12 this is where mode cha= nge > happens) then anything can happen (we may kill vcpu till reset if we = wish) > after that new task is running and all errors are handled by a new ta= sk. Agree. > To model this accurately we need to do task switch in this three stag= es > too and do a full register writeback after stage 2 before stage 3. Or > alternatively emulator should never access vcpu state during emulatio= n. > Entire vcpu state should be in emulation ctx. But this is more > complicated and slow.=20 Speed is immaterial here, but I agree about the complexity. I guess we should do full writeback after stage 2. --=20 error compiling committee.c: too many arguments to function