drm/amdkfd: Improve concurrency of event handling

Use rcu_read_lock to read p->event_idr concurrently with other readers
and writers. Use p->event_mutex only for creating and destroying events
and in kfd_wait_on_events.

Protect the contents of the kfd_event structure with a per-event
spinlock that can be taken inside the rcu_read_lock critical section.

This eliminates contention of p->event_mutex in set_event, which tends
to be on the critical path for dispatch latency even when busy waiting
is used. It also eliminates lock contention in event interrupt handlers.
Since the p->event_mutex is now used much less, the impact of requiring
it in kfd_wait_on_events should also be much smaller.

This should improve event handling latency for processes using multiple
GPUs concurrently.

v2: Reschedule the worker periodically to avoid soft lockup warnings
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Sean Keely <Sean.Keely@amd.com> # v1
Tested-by: Sanjay Tripathi <sanjay.tripathi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drivers/gpu/drm/amd/amdkfd/kfd_events.h

@@ -59,6 +59,7 @@ struct kfd_event {
 
 	int type;
 
+	spinlock_t lock;
 	wait_queue_head_t wq; /* List of event waiters. */
 
 	/* Only for signal events. */

drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c

@@ -146,15 +146,24 @@ static void interrupt_wq(struct work_struct *work)
 	struct kfd_dev *dev = container_of(work, struct kfd_dev,
 						interrupt_work);
 	uint32_t ih_ring_entry[KFD_MAX_RING_ENTRY_SIZE];
+	long start_jiffies = jiffies;
 
 	if (dev->device_info.ih_ring_entry_size > sizeof(ih_ring_entry)) {
 		dev_err_once(dev->adev->dev, "Ring entry too small\n");
 		return;
 	}
 
-	while (dequeue_ih_ring_entry(dev, ih_ring_entry))
+	while (dequeue_ih_ring_entry(dev, ih_ring_entry)) {
 		dev->device_info.event_interrupt_class->interrupt_wq(dev,
 								ih_ring_entry);
+		if (jiffies - start_jiffies > HZ) {
+			/* If we spent more than a second processing signals,
+			 * reschedule the worker to avoid soft-lockup warnings
+			 */
+			queue_work(dev->ih_wq, &dev->interrupt_work);
+			break;
+		}
+	}
 }
 
 bool interrupt_is_wanted(struct kfd_dev *dev,