Fix race between nmi injection and enabling nmi window in a simpler way.
Signed-off-by: Marcelo Tosatti <[email protected]>
(cherry picked from commit c2dd554dc61173ecb6b3741b680d2ae4c245d2ba)
---
arch/x86/kvm/x86.c | 11 ++++++++++-
1 files changed, 10 insertions(+), 1 deletions(-)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index bcc0efc..6ee962c 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -5148,6 +5148,7 @@ static void kvm_put_guest_xcr0(struct kvm_vcpu *vcpu)
static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
{
int r;
+ bool nmi_pending;
bool req_int_win = !irqchip_in_kernel(vcpu->kvm) &&
vcpu->run->request_interrupt_window;
@@ -5191,11 +5192,19 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
if (unlikely(r))
goto out;
+ /*
+ * An NMI can be injected between local nmi_pending read and
+ * vcpu->arch.nmi_pending read inside inject_pending_event().
+ * But in that case, KVM_REQ_EVENT will be set, which makes
+ * the race described above benign.
+ */
+ nmi_pending = ACCESS_ONCE(vcpu->arch.nmi_pending);
+
if (kvm_check_request(KVM_REQ_EVENT, vcpu) || req_int_win) {
inject_pending_event(vcpu);
/* enable NMI/IRQ window open exits if needed */
- if (vcpu->arch.nmi_pending)
+ if (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);
--
1.7.4.4
_______________________________________________
stable mailing list
[email protected]
http://linux.kernel.org/mailman/listinfo/stable
