From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marcelo Tosatti Subject: Re: [PATCH v2 3/3] KVM: Non-atomic interrupt injection Date: Tue, 20 Jul 2010 21:55:33 -0300 Message-ID: <20100721005533.GA9989@amt.cnet> References: <1279631827-4016-1-git-send-email-avi@redhat.com> <1279631827-4016-4-git-send-email-avi@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: kvm@vger.kernel.org To: Avi Kivity Return-path: Received: from mx1.redhat.com ([209.132.183.28]:47535 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1761701Ab0GUA40 (ORCPT ); Tue, 20 Jul 2010 20:56:26 -0400 Received: from int-mx03.intmail.prod.int.phx2.redhat.com (int-mx03.intmail.prod.int.phx2.redhat.com [10.5.11.16]) by mx1.redhat.com (8.13.8/8.13.8) with ESMTP id o6L0uQdh008110 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK) for ; Tue, 20 Jul 2010 20:56:26 -0400 Content-Disposition: inline In-Reply-To: <1279631827-4016-4-git-send-email-avi@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On Tue, Jul 20, 2010 at 04:17:07PM +0300, Avi Kivity wrote: > Change the interrupt injection code to work from preemptible, interrupts > enabled context. This works by adding a ->cancel_injection() operation > that undoes an injection in case we were not able to actually enter the guest > (this condition could never happen with atomic injection). > > Signed-off-by: Avi Kivity > --- > arch/x86/include/asm/kvm_host.h | 1 + > arch/x86/kvm/svm.c | 12 ++++++++++++ > arch/x86/kvm/vmx.c | 11 +++++++++++ > arch/x86/kvm/x86.c | 27 ++++++++++++++------------- > 4 files changed, 38 insertions(+), 13 deletions(-) > > --- a/arch/x86/kvm/x86.c > +++ b/arch/x86/kvm/x86.c > @@ -4709,6 +4709,19 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) > if (unlikely(r)) > goto out; > > + 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); > + > + if (kvm_lapic_enabled(vcpu)) { > + update_cr8_intercept(vcpu); > + kvm_lapic_sync_to_vapic(vcpu); > + } > + > preempt_disable(); > > kvm_x86_ops->prepare_guest_switch(vcpu); > @@ -4727,23 +4740,11 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu) > smp_wmb(); > local_irq_enable(); > preempt_enable(); > + kvm_x86_ops->cancel_injection(vcpu); > r = 1; > goto out; > } > > - 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); > - > - if (kvm_lapic_enabled(vcpu)) { > - update_cr8_intercept(vcpu); > - kvm_lapic_sync_to_vapic(vcpu); > - } > - > srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx); > > kvm_guest_enter(); This breaks int kvm_lapic_find_highest_irr(struct kvm_vcpu *vcpu) { struct kvm_lapic *apic = vcpu->arch.apic; int highest_irr; /* This may race with setting of irr in __apic_accept_irq() and * value returned may be wrong, but kvm_vcpu_kick() in * __apic_accept_irq * will cause vmexit immediately and the value will be * recalculated * on the next vmentry. */ (also valid for nmi_pending and PIC). Can't simply move atomic_set(guest_mode, 1) in preemptible section as that would make it possible for kvm_vcpu_kick to IPI stale vcpu->cpu. Also should undo vmx.rmode.* ?