KVM: fix OOPS on flush_work

The conversion done by commit 3706fea ("KVM: Remove deprecated
create_singlethread_workqueue") is broken.  It flushes a single work
item &irqfd->shutdown instead of all of them, and even worse if there
is no irqfd on the list then you get a NULL pointer dereference.
Revert the virt/kvm/eventfd.c part of that patch; to avoid the
deprecated function, just allocate our own workqueue---it does
not even have to be unbound---with alloc_workqueue.

Fixes: 3706fea
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

virt/kvm/eventfd.c

@@ -42,6 +42,7 @@
 
 #ifdef CONFIG_HAVE_KVM_IRQFD
 
+static struct workqueue_struct *irqfd_cleanup_wq;
 
 static void
 irqfd_inject(struct work_struct *work)
@@ -167,7 +168,7 @@ irqfd_deactivate(struct kvm_kernel_irqfd *irqfd)
 
 	list_del_init(&irqfd->list);
 
-	schedule_work(&irqfd->shutdown);
+	queue_work(irqfd_cleanup_wq, &irqfd->shutdown);
 }
 
 int __attribute__((weak)) kvm_arch_set_irq_inatomic(
@@ -554,7 +555,7 @@ kvm_irqfd_deassign(struct kvm *kvm, struct kvm_irqfd *args)
 	 * so that we guarantee there will not be any more interrupts on this
 	 * gsi once this deassign function returns.
 	 */
-	flush_work(&irqfd->shutdown);
+	flush_workqueue(irqfd_cleanup_wq);
 
 	return 0;
 }
@@ -591,7 +592,7 @@ kvm_irqfd_release(struct kvm *kvm)
 	 * Block until we know all outstanding shutdown jobs have completed
 	 * since we do not take a kvm* reference.
 	 */
-	flush_work(&irqfd->shutdown);
+	flush_workqueue(irqfd_cleanup_wq);
 
 }
 
@@ -621,8 +622,23 @@ void kvm_irq_routing_update(struct kvm *kvm)
 	spin_unlock_irq(&kvm->irqfds.lock);
 }
 
+/*
+ * create a host-wide workqueue for issuing deferred shutdown requests
+ * aggregated from all vm* instances. We need our own isolated
+ * queue to ease flushing work items when a VM exits.
+ */
+int kvm_irqfd_init(void)
+{
+	irqfd_cleanup_wq = alloc_workqueue("kvm-irqfd-cleanup", 0, 0);
+	if (!irqfd_cleanup_wq)
+		return -ENOMEM;
+
+	return 0;
+}
+
 void kvm_irqfd_exit(void)
 {
+	destroy_workqueue(irqfd_cleanup_wq);
 }
 #endif
 

virt/kvm/kvm_main.c

@@ -3846,7 +3846,12 @@ int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
 	 * kvm_arch_init makes sure there's at most one caller
 	 * for architectures that support multiple implementations,
 	 * like intel and amd on x86.
+	 * kvm_arch_init must be called before kvm_irqfd_init to avoid creating
+	 * conflicts in case kvm is already setup for another implementation.
 	 */
+	r = kvm_irqfd_init();
+	if (r)
+		goto out_irqfd;
 
 	if (!zalloc_cpumask_var(&cpus_hardware_enabled, GFP_KERNEL)) {
 		r = -ENOMEM;
@@ -3928,6 +3933,7 @@ int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
 	free_cpumask_var(cpus_hardware_enabled);
 out_free_0:
 	kvm_irqfd_exit();
+out_irqfd:
 	kvm_arch_exit();
 out_fail:
 	return r;