arm/arm64: KVM: Fix migration race in the arch timer

When a VCPU is no longer running, we currently check to see if it has a
timer scheduled in the future, and if it does, we schedule a host
hrtimer to notify is in case the timer expires while the VCPU is still
not running.  When the hrtimer fires, we mask the guest's timer and
inject the timer IRQ (still relying on the guest unmasking the time when
it receives the IRQ).

This is all good and fine, but when migration a VM (checkpoint/restore)
this introduces a race.  It is unlikely, but possible, for the following
sequence of events to happen:

 1. Userspace stops the VM
 2. Hrtimer for VCPU is scheduled
 3. Userspace checkpoints the VGIC state (no pending timer interrupts)
 4. The hrtimer fires, schedules work in a workqueue
 5. Workqueue function runs, masks the timer and injects timer interrupt
 6. Userspace checkpoints the timer state (timer masked)

At restore time, you end up with a masked timer without any timer
interrupts and your guest halts never receiving timer interrupts.

Fix this by only kicking the VCPU in the workqueue function, and sample
the expired state of the timer when entering the guest again and inject
the interrupt and mask the timer only then.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>

1 files changed, 36 insertions, 9 deletions

diff --git a/virt/kvm/arm/arch_timer.c b/virt/kvm/arm/arch_timer.c
index 6e54f3542126..98c95f2fcba4 100644
--- a/virt/kvm/arm/arch_timer.c
+++ b/virt/kvm/arm/arch_timer.c
@@ -85,13 +85,22 @@ static irqreturn_t kvm_arch_timer_handler(int irq, void *dev_id)
         return IRQ_HANDLED;
 }
 
+/*
+ * Work function for handling the backup timer that we schedule when a vcpu is
+ * no longer running, but had a timer programmed to fire in the future.
+ */
 static void kvm_timer_inject_irq_work(struct work_struct *work)
 {
         struct kvm_vcpu *vcpu;
 
         vcpu = container_of(work, struct kvm_vcpu, arch.timer_cpu.expired);
         vcpu->arch.timer_cpu.armed = false;
-        kvm_timer_inject_irq(vcpu);
+
+        /*
+         * If the vcpu is blocked we want to wake it up so that it will see
+         * the timer has expired when entering the guest.
+         */
+        kvm_vcpu_kick(vcpu);
 }
 
 static enum hrtimer_restart kvm_timer_expire(struct hrtimer *hrt)
@@ -102,6 +111,21 @@ static enum hrtimer_restart kvm_timer_expire(struct hrtimer *hrt)
         return HRTIMER_NORESTART;
 }
 
+bool kvm_timer_should_fire(struct kvm_vcpu *vcpu)
+{
+        struct arch_timer_cpu *timer = &vcpu->arch.timer_cpu;
+        cycle_t cval, now;
+
+        if ((timer->cntv_ctl & ARCH_TIMER_CTRL_IT_MASK) ||
+                !(timer->cntv_ctl & ARCH_TIMER_CTRL_ENABLE))
+                return false;
+
+        cval = timer->cntv_cval;
+        now = kvm_phys_timer_read() - vcpu->kvm->arch.timer.cntvoff;
+
+        return cval <= now;
+}
+
 /**
  * kvm_timer_flush_hwstate - prepare to move the virt timer to the cpu
  * @vcpu: The vcpu pointer
@@ -119,6 +143,13 @@ void kvm_timer_flush_hwstate(struct kvm_vcpu *vcpu)
          * populate the CPU timer again.
          */
         timer_disarm(timer);
+
+        /*
+         * If the timer expired while we were not scheduled, now is the time
+         * to inject it.
+         */
+        if (kvm_timer_should_fire(vcpu))
+                kvm_timer_inject_irq(vcpu);
 }
 
 /**
@@ -134,16 +165,9 @@ void kvm_timer_sync_hwstate(struct kvm_vcpu *vcpu)
         cycle_t cval, now;
         u64 ns;
 
-        if ((timer->cntv_ctl & ARCH_TIMER_CTRL_IT_MASK) ||
-                !(timer->cntv_ctl & ARCH_TIMER_CTRL_ENABLE))
-                return;
-
-        cval = timer->cntv_cval;
-        now = kvm_phys_timer_read() - vcpu->kvm->arch.timer.cntvoff;
-
         BUG_ON(timer_is_armed(timer));
 
-        if (cval <= now) {
+        if (kvm_timer_should_fire(vcpu)) {
                 /*
                  * Timer has already expired while we were not
                  * looking. Inject the interrupt and carry on.
@@ -152,6 +176,9 @@ void kvm_timer_sync_hwstate(struct kvm_vcpu *vcpu)
                 return;
         }
 
+        cval = timer->cntv_cval;
+        now = kvm_phys_timer_read() - vcpu->kvm->arch.timer.cntvoff;
+
         ns = cyclecounter_cyc2ns(timecounter->cc, cval - now, timecounter->mask,
                                  &timecounter->frac);
         timer_arm(timer, ns);