[Intel-gfx] [PATCH 09/10] drm/i915/gt: Decouple inflight virtual engines

[Chris Wilson]

Once a virtual engine has been bound to a sibling, it will remain bound
until we finally schedule out the last active request. We can not rebind
the context to a new sibling while it is inflight as the context save
will conflict, hence we wait. As we cannot then use any other sibliing
while the context is inflight, only kick the bound sibling while it
inflight and upon scheduling out the kick the rest (so that we can swap
engines on timeslicing if the previously bound engine becomes
oversubscribed).

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/gt/intel_lrc.c | 30 +++++++++++++----------------
 1 file changed, 13 insertions(+), 17 deletions(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index 34a8eadc2de3..4a1e23e51314 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -1409,9 +1409,8 @@ execlists_schedule_in(struct i915_request *rq, int idx)
 static void kick_siblings(struct i915_request *rq, struct intel_context *ce)
 {
 	struct virtual_engine *ve = container_of(ce, typeof(*ve), context);
-	struct i915_request *next = READ_ONCE(ve->request);
 
-	if (next == rq || (next && next->execution_mask & ~rq->execution_mask))
+	if (READ_ONCE(ve->request))
 		tasklet_hi_schedule(&ve->base.execlists.tasklet);
 }
 
@@ -1825,18 +1824,14 @@ first_virtual_engine(struct intel_engine_cs *engine)
 			rb_entry(rb, typeof(*ve), nodes[engine->id].rb);
 		struct i915_request *rq = READ_ONCE(ve->request);
 
-		if (!rq) { /* lazily cleanup after another engine handled rq */
+		/* lazily cleanup after another engine handled rq */
+		if (!rq || !virtual_matches(ve, rq, engine)) {
 			rb_erase_cached(rb, &el->virtual);
 			RB_CLEAR_NODE(rb);
 			rb = rb_first_cached(&el->virtual);
 			continue;
 		}
 
-		if (!virtual_matches(ve, rq, engine)) {
-			rb = rb_next(rb);
-			continue;
-		}
-
 		return ve;
 	}
 
@@ -5486,7 +5481,6 @@ static void virtual_submission_tasklet(unsigned long data)
 	if (unlikely(!mask))
 		return;
 
-	local_irq_disable();
 	for (n = 0; n < ve->num_siblings; n++) {
 		struct intel_engine_cs *sibling = READ_ONCE(ve->siblings[n]);
 		struct ve_node * const node = &ve->nodes[sibling->id];
@@ -5496,20 +5490,19 @@ static void virtual_submission_tasklet(unsigned long data)
 		if (!READ_ONCE(ve->request))
 			break; /* already handled by a sibling's tasklet */
 
+		spin_lock_irq(&sibling->active.lock);
+
 		if (unlikely(!(mask & sibling->mask))) {
 			if (!RB_EMPTY_NODE(&node->rb)) {
-				spin_lock(&sibling->active.lock);
 				rb_erase_cached(&node->rb,
 						&sibling->execlists.virtual);
 				RB_CLEAR_NODE(&node->rb);
-				spin_unlock(&sibling->active.lock);
 			}
-			continue;
-		}
 
-		spin_lock(&sibling->active.lock);
+			goto unlock_engine;
+		}
 
-		if (!RB_EMPTY_NODE(&node->rb)) {
+		if (unlikely(!RB_EMPTY_NODE(&node->rb))) {
 			/*
 			 * Cheat and avoid rebalancing the tree if we can
 			 * reuse this node in situ.
@@ -5549,9 +5542,12 @@ static void virtual_submission_tasklet(unsigned long data)
 		if (first && prio > sibling->execlists.queue_priority_hint)
 			tasklet_hi_schedule(&sibling->execlists.tasklet);
 
-		spin_unlock(&sibling->active.lock);
+unlock_engine:
+		spin_unlock_irq(&sibling->active.lock);
+
+		if (intel_context_inflight(&ve->context))
+			break;
 	}
-	local_irq_enable();
 }
 
 static void virtual_submit_request(struct i915_request *rq)
-- 
2.20.1