From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH 1/6] KVM: Check for pending events before attempting injection Date: Wed, 28 Jul 2010 19:31:03 +0300 Message-ID: <4C505B47.5090405@redhat.com> References: <1280236780-5847-1-git-send-email-avi@redhat.com> <1280236780-5847-2-git-send-email-avi@redhat.com> <20100728162119.GA25224@amt.cnet> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org To: Marcelo Tosatti Return-path: Received: from mx1.redhat.com ([209.132.183.28]:52965 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753352Ab0G1QbH (ORCPT ); Wed, 28 Jul 2010 12:31:07 -0400 Received: from int-mx01.intmail.prod.int.phx2.redhat.com (int-mx01.intmail.prod.int.phx2.redhat.com [10.5.11.11]) by mx1.redhat.com (8.13.8/8.13.8) with ESMTP id o6SGV7PM026402 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK) for ; Wed, 28 Jul 2010 12:31:07 -0400 In-Reply-To: <20100728162119.GA25224@amt.cnet> Sender: kvm-owner@vger.kernel.org List-ID: On 07/28/2010 07:21 PM, Marcelo Tosatti wrote: > On Tue, Jul 27, 2010 at 04:19:35PM +0300, Avi Kivity wrote: >> Instead of blindly attempting to inject an event before each guest entry, >> check for a possible event first in vcpu->requests. Sites that can trigger >> event injection are modified to set KVM_REQ_EVENT: >> >> - interrupt, nmi window opening >> - ppr updates >> - i8259 output changes >> - local apic irr changes >> - rflags updates >> - gif flag set >> - event set on exit >> >> This improves non-injecting entry performance, and sets the stage for >> non-atomic injection. >> >> Signed-off-by: Avi Kivity >> --- >> arch/x86/kvm/i8259.c | 1 + >> arch/x86/kvm/lapic.c | 12 ++++++++++-- >> arch/x86/kvm/svm.c | 8 +++++++- >> arch/x86/kvm/vmx.c | 6 ++++++ >> arch/x86/kvm/x86.c | 35 ++++++++++++++++++++++++++--------- >> include/linux/kvm_host.h | 1 + >> 6 files changed, 51 insertions(+), 12 deletions(-) >> >> @@ -4731,17 +4737,19 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) >> goto out; >> } >> >> - inject_pending_event(vcpu); >> + if (kvm_check_request(KVM_REQ_EVENT, vcpu)) { >> + inject_pending_event(vcpu); >> >> - /* enable NMI/IRQ window open exits if needed */ >> - if (vcpu->arch.nmi_pending) >> - kvm_x86_ops->enable_nmi_window(vcpu); >> - else if (kvm_cpu_has_interrupt(vcpu) || req_int_win) >> - kvm_x86_ops->enable_irq_window(vcpu); >> + /* enable NMI/IRQ window open exits if needed */ >> + if (vcpu->arch.nmi_pending) >> + kvm_x86_ops->enable_nmi_window(vcpu); >> + else if (kvm_cpu_has_interrupt(vcpu) || req_int_win) >> + kvm_x86_ops->enable_irq_window(vcpu); > Problem is it might not be possible to inject the event signalled by > KVM_REQ_EVENT, say an interrupt from an irqchip, if there is an event > that needs reinjection (or an exception). That can happen event now, no? A pending exception, interrupt comes along, injection picks up the exception but leaves the interrupt. Now the situation can be more complicated: - pending exception - injection - interrupt, sets KVM_REQ_EVENT - notices KVM_REQ_EVENT - drops KVM_REQ_EVENT, cancels exception (made pending again) - goes back - injection (injects exception again, interrupt is pending) as far as I can tell, this is all fine. > Perhaps moving atomic_set(&vcpu->guest_mode, 1) up to preemptible > section is safe, because kvm_vcpu_kick can only IPI stale vcpu->cpu > while preemption is enabled. In that case, it will hit > > if (!atomic_read(&vcpu->guest_mode) > > later. > I don't really follow. > Although the KVM_REQ_EVENT idea is nice. Can you think of a way > to fix the issue? -- I have a truly marvellous patch that fixes the bug which this signature is too narrow to contain.