KVM: PPC: Book3S HV: Make secondary threads more robust against stray IPIs

Currently on POWER7, if we are running the guest on a core and we don't
need all the hardware threads, we do nothing to ensure that the unused
threads aren't executing in the kernel (other than checking that they
are offline).  We just assume they're napping and we don't do anything
to stop them trying to enter the kernel while the guest is running.
This means that a stray IPI can wake up the hardware thread and it will
then try to enter the kernel, but since the core is in guest context,
it will execute code from the guest in hypervisor mode once it turns the
MMU on, which tends to lead to crashes or hangs in the host.

This fixes the problem by adding two new one-byte flags in the
kvmppc_host_state structure in the PACA which are used to interlock
between the primary thread and the unused secondary threads when entering
the guest.  With these flags, the primary thread can ensure that the
unused secondaries are not already in kernel mode (i.e. handling a stray
IPI) and then indicate that they should not try to enter the kernel
if they do get woken for any reason.  Instead they will go into KVM code,
find that there is no vcpu to run, acknowledge and clear the IPI and go
back to nap mode.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Avi Kivity <avi@redhat.com>

1 files changed, 46 insertions, 3 deletions

diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c
index 01294a5099d..e87f6196d22 100644
--- a/arch/powerpc/kvm/book3s_hv.c
+++ b/arch/powerpc/kvm/book3s_hv.c
@@ -569,6 +569,45 @@ static void kvmppc_remove_runnable(struct kvmppc_vcore *vc,
 	list_del(&vcpu->arch.run_list);
 }
 
+static int kvmppc_grab_hwthread(int cpu)
+{
+	struct paca_struct *tpaca;
+	long timeout = 1000;
+
+	tpaca = &paca[cpu];
+
+	/* Ensure the thread won't go into the kernel if it wakes */
+	tpaca->kvm_hstate.hwthread_req = 1;
+
+	/*
+	 * If the thread is already executing in the kernel (e.g. handling
+	 * a stray interrupt), wait for it to get back to nap mode.
+	 * The smp_mb() is to ensure that our setting of hwthread_req
+	 * is visible before we look at hwthread_state, so if this
+	 * races with the code at system_reset_pSeries and the thread
+	 * misses our setting of hwthread_req, we are sure to see its
+	 * setting of hwthread_state, and vice versa.
+	 */
+	smp_mb();
+	while (tpaca->kvm_hstate.hwthread_state == KVM_HWTHREAD_IN_KERNEL) {
+		if (--timeout <= 0) {
+			pr_err("KVM: couldn't grab cpu %d\n", cpu);
+			return -EBUSY;
+		}
+		udelay(1);
+	}
+	return 0;
+}
+
+static void kvmppc_release_hwthread(int cpu)
+{
+	struct paca_struct *tpaca;
+
+	tpaca = &paca[cpu];
+	tpaca->kvm_hstate.hwthread_req = 0;
+	tpaca->kvm_hstate.kvm_vcpu = NULL;
+}
+
 static void kvmppc_start_thread(struct kvm_vcpu *vcpu)
 {
 	int cpu;
@@ -588,8 +627,7 @@ static void kvmppc_start_thread(struct kvm_vcpu *vcpu)
 	smp_wmb();
 #if defined(CONFIG_PPC_ICP_NATIVE) && defined(CONFIG_SMP)
 	if (vcpu->arch.ptid) {
-		tpaca->cpu_start = 0x80;
-		wmb();
+		kvmppc_grab_hwthread(cpu);
 		xics_wake_cpu(cpu);
 		++vc->n_woken;
 	}
@@ -639,7 +677,7 @@ static int kvmppc_run_core(struct kvmppc_vcore *vc)
 	struct kvm_vcpu *vcpu, *vcpu0, *vnext;
 	long ret;
 	u64 now;
-	int ptid;
+	int ptid, i;
 
 	/* don't start if any threads have a signal pending */
 	list_for_each_entry(vcpu, &vc->runnable_threads, arch.run_list)
@@ -686,12 +724,17 @@ static int kvmppc_run_core(struct kvmppc_vcore *vc)
 	vc->napping_threads = 0;
 	list_for_each_entry(vcpu, &vc->runnable_threads, arch.run_list)
 		kvmppc_start_thread(vcpu);
+	/* Grab any remaining hw threads so they can't go into the kernel */
+	for (i = ptid; i < threads_per_core; ++i)
+		kvmppc_grab_hwthread(vc->pcpu + i);
 
 	preempt_disable();
 	spin_unlock(&vc->lock);
 
 	kvm_guest_enter();
 	__kvmppc_vcore_entry(NULL, vcpu0);
+	for (i = 0; i < threads_per_core; ++i)
+		kvmppc_release_hwthread(vc->pcpu + i);
 
 	spin_lock(&vc->lock);
 	/* disable sending of IPIs on virtual external irqs */