[Intel-gfx] [CI] drm/i915: Skip execlists_dequeue() early if the list is empty

[Chris Wilson]

Do an early read of the execlists' queue before we take the spinlock and
start checking. This is safe as the first writer to the execlists queue
will cause the tasklet to be run again after a memory barrier.

v2: Keep guc in sync with execlists queue changes

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin at linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
Cc: Michał Winiarski <michal.winiarski at intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin at linux.intel.com>
Reviewed-by: Michał Winiarski <michal.winiarski at intel.com>
---
 drivers/gpu/drm/i915/i915_guc_submission.c | 4 ++++
 drivers/gpu/drm/i915/intel_lrc.c           | 4 ++++
 2 files changed, 8 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
index a3636b31ebc3..78718a8e70c3 100644
--- a/drivers/gpu/drm/i915/i915_guc_submission.c
+++ b/drivers/gpu/drm/i915/i915_guc_submission.c
@@ -577,6 +577,10 @@ static bool i915_guc_dequeue(struct intel_engine_cs *engine)
 	struct rb_node *rb;
 	bool submit = false;
 
+	/* After execlist_first is updated, a new tasklet must be scheduled */
+	if (!READ_ONCE(engine->execlist_first))
+		return false;
+
 	spin_lock_irqsave(&engine->timeline->lock, flags);
 	rb = engine->execlist_first;
 	while (rb) {
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index becde55b02a3..7adc588d1a19 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -403,6 +403,10 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	struct rb_node *rb;
 	bool submit = false;
 
+	/* After execlist_first is updated, a new tasklet must be scheduled */
+	if (!READ_ONCE(engine->execlist_first))
+		return;
+
 	last = port->request;
 	if (last)
 		/* WaIdleLiteRestore:bdw,skl
-- 
2.11.0