Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 9242b5b6 authored by Bandan Das's avatar Bandan Das Committed by Paolo Bonzini
Browse files

KVM: x86: Check for nested events if there is an injectable interrupt 



With commit b6b8a145 that introduced
vmx_check_nested_events, checks for injectable interrupts happen
at different points in time for L1 and L2 that could potentially
cause a race. The regression occurs because KVM_REQ_EVENT is always
set when nested_run_pending is set even if there's no pending interrupt.
Consequently, there could be a small window when check_nested_events
returns without exiting to L1, but an interrupt comes through soon
after and it incorrectly, gets injected to L2 by inject_pending_event
Fix this by adding a call to check for nested events too when a check
for injectable interrupt returns true

Signed-off-by: default avatarBandan Das <bsd@redhat.com>
Signed-off-by: default avatarPaolo Bonzini <pbonzini@redhat.com>
parent cd3de83f

arch/x86/kvm/x86.c

+12 −0
Original line number Original line Diff line number Diff line
@@ -5887,6 +5887,18 @@ static int inject_pending_event(struct kvm_vcpu *vcpu, bool req_int_win) 
			kvm_x86_ops->set_nmi(vcpu);

			kvm_x86_ops->set_nmi(vcpu);

		}

		}

	} else if (kvm_cpu_has_injectable_intr(vcpu)) {

	} else if (kvm_cpu_has_injectable_intr(vcpu)) {

		/*

		 * Because interrupts can be injected asynchronously, we are

		 * calling check_nested_events again here to avoid a race condition.

		 * See https://lkml.org/lkml/2014/7/2/60 for discussion about this

		 * proposal and current concerns.  Perhaps we should be setting

		 * KVM_REQ_EVENT only on certain events and not unconditionally?

		 */

		if (is_guest_mode(vcpu) && kvm_x86_ops->check_nested_events) {

			r = kvm_x86_ops->check_nested_events(vcpu, req_int_win);

			if (r != 0)

				return r;

		}

		if (kvm_x86_ops->interrupt_allowed(vcpu)) {

		if (kvm_x86_ops->interrupt_allowed(vcpu)) {

			kvm_queue_interrupt(vcpu, kvm_cpu_get_interrupt(vcpu),

			kvm_queue_interrupt(vcpu, kvm_cpu_get_interrupt(vcpu),

					    false);

					    false);