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=-15.3 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,NICE_REPLY_A,SPF_HELO_NONE,SPF_PASS,UNPARSEABLE_RELAY, USER_AGENT_SANE_1 autolearn=unavailable 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 AEB1BC432BE for ; Tue, 10 Aug 2021 10:35:00 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 8F58B60E97 for ; Tue, 10 Aug 2021 10:35:00 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239839AbhHJKfV (ORCPT ); Tue, 10 Aug 2021 06:35:21 -0400 Received: from out30-44.freemail.mail.aliyun.com ([115.124.30.44]:48209 "EHLO out30-44.freemail.mail.aliyun.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S238459AbhHJKfS (ORCPT ); Tue, 10 Aug 2021 06:35:18 -0400 X-Alimail-AntiSpam: AC=PASS;BC=-1|-1;BR=01201311R351e4;CH=green;DM=||false|;DS=||;FP=0|-1|-1|-1|0|-1|-1|-1;HT=e01e04420;MF=laijs@linux.alibaba.com;NM=1;PH=DS;RN=14;SR=0;TI=SMTPD_---0Uiayl.7_1628591694; Received: from C02XQCBJJG5H.local(mailfrom:laijs@linux.alibaba.com fp:SMTPD_---0Uiayl.7_1628591694) by smtp.aliyun-inc.com(127.0.0.1); Tue, 10 Aug 2021 18:34:55 +0800 Subject: Re: [PATCH V2 3/3] KVM: X86: Reset DR6 only when KVM_DEBUGREG_WONT_EXIT To: Paolo Bonzini , Lai Jiangshan , linux-kernel@vger.kernel.org Cc: Sean Christopherson , Vitaly Kuznetsov , Wanpeng Li , Jim Mattson , Joerg Roedel , Thomas Gleixner , Ingo Molnar , Borislav Petkov , x86@kernel.org, "H. Peter Anvin" , kvm@vger.kernel.org References: <20210809174307.145263-1-jiangshanlai@gmail.com> <20210809174307.145263-3-jiangshanlai@gmail.com> From: Lai Jiangshan Message-ID: <7a1ca89f-7b4e-7df2-e47a-ac5207137a05@linux.alibaba.com> Date: Tue, 10 Aug 2021 18:34:54 +0800 User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:78.0) Gecko/20100101 Thunderbird/78.11.0 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2021/8/10 18:14, Paolo Bonzini wrote: > On 09/08/21 19:43, Lai Jiangshan wrote: >> From: Lai Jiangshan >> >> The commit efdab992813fb ("KVM: x86: fix escape of guest dr6 to the host") >> fixed a bug by reseting DR6 unconditionally when the vcpu being scheduled out. >> >> But writing to debug registers is slow, and it can be shown in perf results >> sometimes even neither the host nor the guest activate breakpoints. >> >> It'd be better to reset it conditionally and this patch moves the code of >> reseting DR6 to the path of VM-exit and only reset it when >> KVM_DEBUGREG_WONT_EXIT which is the only case that DR6 is guest value. >> >> Signed-off-by: Lai Jiangshan >> --- >>   arch/x86/kvm/x86.c | 8 ++------ >>   1 file changed, 2 insertions(+), 6 deletions(-) >> >> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c >> index d2aa49722064..f40cdd7687d8 100644 >> --- a/arch/x86/kvm/x86.c >> +++ b/arch/x86/kvm/x86.c >> @@ -4309,12 +4309,6 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) >>       static_call(kvm_x86_vcpu_put)(vcpu); >>       vcpu->arch.last_host_tsc = rdtsc(); >> -    /* >> -     * If userspace has set any breakpoints or watchpoints, dr6 is restored >> -     * on every vmexit, but if not, we might have a stale dr6 from the >> -     * guest. do_debug expects dr6 to be cleared after it runs, do the same. >> -     */ >> -    set_debugreg(0, 6); >>   } >>   static int kvm_vcpu_ioctl_get_lapic(struct kvm_vcpu *vcpu, >> @@ -9630,6 +9624,8 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) >>           static_call(kvm_x86_sync_dirty_debug_regs)(vcpu); >>           kvm_update_dr0123(vcpu); >>           kvm_update_dr7(vcpu); >> +        /* Reset Dr6 which is guest value. */ >> +        set_debugreg(DR6_RESERVED, 6); >>       } >>       /* >> > > ... and this should also be done exclusively for VMX, in vmx_sync_dirty_debug_regs: > >     KVM: VMX: Reset DR6 only when KVM_DEBUGREG_WONT_EXIT >     The commit efdab992813fb ("KVM: x86: fix escape of guest dr6 to the host") >     fixed a bug by resetting DR6 unconditionally when the vcpu being scheduled out. >     But writing to debug registers is slow, and it can be visible in perf results >     sometimes, even if neither the host nor the guest activate breakpoints. >     Since KVM_DEBUGREG_WONT_EXIT on Intel processors is the only case >     where DR6 gets the guest value, and it never happens at all on SVM, >     the register can be cleared in vmx.c right after reading it. >     Reported-by: Lai Jiangshan >     Signed-off-by: Paolo Bonzini > > diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c > index 21a3ef3012cf..3a91302d05c0 100644 > --- a/arch/x86/kvm/vmx/vmx.c > +++ b/arch/x86/kvm/vmx/vmx.c > @@ -5110,6 +5110,12 @@ static void vmx_sync_dirty_debug_regs(struct kvm_vcpu *vcpu) > >      vcpu->arch.switch_db_regs &= ~KVM_DEBUGREG_WONT_EXIT; >      exec_controls_setbit(to_vmx(vcpu), CPU_BASED_MOV_DR_EXITING); > + > +    /* > +     * do_debug expects dr6 to be cleared after it runs, avoid that it sees > +     * a stale dr6 from the guest. > +     */ do_debug() is renamed. Maybe you can use "The host kernel #DB handler". > +    set_debugreg(DR6_RESERVED, 6); >  } > >  static void vmx_set_dr7(struct kvm_vcpu *vcpu, unsigned long val) > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c > index fbc536b21585..04c393551fb0 100644 > --- a/arch/x86/kvm/x86.c > +++ b/arch/x86/kvm/x86.c > @@ -4313,12 +4313,6 @@ void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) > >      static_call(kvm_x86_vcpu_put)(vcpu); >      vcpu->arch.last_host_tsc = rdtsc(); > -    /* > -     * If userspace has set any breakpoints or watchpoints, dr6 is restored > -     * on every vmexit, but if not, we might have a stale dr6 from the > -     * guest. do_debug expects dr6 to be cleared after it runs, do the same. > -     */ > -    set_debugreg(0, 6); >  } > >  static int kvm_vcpu_ioctl_get_lapic(struct kvm_vcpu *vcpu, > > > Paolo