From mboxrd@z Thu Jan 1 00:00:00 1970 From: Radim =?utf-8?B?S3LEjW3DocWZ?= Subject: Re: [PATCH v3 RESEND] kvm: vmx: Properly handle machine check during VM-entry Date: Tue, 11 Jul 2017 16:30:40 +0200 Message-ID: <20170711143039.GB3326@potion> References: <20170629200116.177531-1-jmattson@google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: kvm list , Xiao Guangrong To: Jim Mattson Return-path: Received: from mx1.redhat.com ([209.132.183.28]:44948 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755310AbdGKOan (ORCPT ); Tue, 11 Jul 2017 10:30:43 -0400 Content-Disposition: inline In-Reply-To: <20170629200116.177531-1-jmattson@google.com> Sender: kvm-owner@vger.kernel.org List-ID: 2017-06-29 13:01-0700, Jim Mattson: > vmx_complete_atomic_exit should call kvm_machine_check for any > VM-entry failure due to a machine-check event. Such an exit should be > recognized solely by its basic exit reason (i.e. the low 16 bits of > the VMCS exit reason field). None of the other VMCS exit information > fields contain valid information when the VM-exit is due to "VM-entry > failure due to machine-check event". > > Signed-off-by: Jim Mattson > --- > arch/x86/kvm/vmx.c | 15 +++++++++------ > 1 file changed, 9 insertions(+), 6 deletions(-) > > diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c > index ca5d2b93385c..42db3eb2d13b 100644 > --- a/arch/x86/kvm/vmx.c > +++ b/arch/x86/kvm/vmx.c > @@ -8591,17 +8591,20 @@ static void vmx_apicv_post_state_restore(struct kvm_vcpu *vcpu) > > static void vmx_complete_atomic_exit(struct vcpu_vmx *vmx) > { > - u32 exit_intr_info; > + u32 exit_intr_info = 0; > + u16 basic_exit_reason = (u16)vmx->exit_reason; > > - if (!(vmx->exit_reason == EXIT_REASON_MCE_DURING_VMENTRY > - || vmx->exit_reason == EXIT_REASON_EXCEPTION_NMI)) > + if (!(basic_exit_reason == EXIT_REASON_MCE_DURING_VMENTRY > + || basic_exit_reason == EXIT_REASON_EXCEPTION_NMI)) > return; > > - vmx->exit_intr_info = vmcs_read32(VM_EXIT_INTR_INFO); > - exit_intr_info = vmx->exit_intr_info; > + if (basic_exit_reason != EXIT_REASON_MCE_DURING_VMENTRY) I think this would be clearer as if (!(vmx->exit_reason & VMX_EXIT_REASONS_FAILED_VMENTRY)) because that is the actual reason why VM_EXIT_INTR_INFO is not useable. I'm ok with the original too, and will change that while applying if you agree, thanks. > + exit_intr_info = vmcs_read32(VM_EXIT_INTR_INFO); > + vmx->exit_intr_info = exit_intr_info; > > /* Handle machine checks before interrupts are enabled */ > - if (is_machine_check(exit_intr_info)) > + if (basic_exit_reason == EXIT_REASON_MCE_DURING_VMENTRY || > + is_machine_check(exit_intr_info)) > kvm_machine_check(); > > /* We need to handle NMIs before interrupts are enabled */ > -- > 2.13.2.725.g09c95d1e9-goog >