[PATCH 16/44] drm/i915: Prioritise non-busywait semaphore workloads

[Chris Wilson]

We don't want to busywait on the GPU if we have other work to do. If we
give non-busywaiting workloads higher (initial) priority than workloads
that require a busywait, we will prioritise work that is ready to run
immediately.

Testcase: igt/gem_exec_schedule/semaphore
Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_request.c   |  8 +++++++-
 drivers/gpu/drm/i915/i915_scheduler.c |  2 +-
 drivers/gpu/drm/i915/i915_scheduler.h | 11 +++++++----
 drivers/gpu/drm/i915/intel_lrc.c      |  5 +++--
 4 files changed, 18 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index 6d825cd28ae6..30cd8724d39f 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -844,7 +844,7 @@ emit_semaphore_wait(struct i915_request *to,
 		intel_ring_advance(to, cs);
 	}
 
-	to->sched.semaphore = true;
+	to->sched.semaphore |= I915_SCHED_HAS_SEMAPHORE;
 	return 0;
 }
 
@@ -867,6 +867,9 @@ i915_request_await_request(struct i915_request *to, struct i915_request *from)
 			return ret;
 	}
 
+	if (from->sched.semaphore && !i915_request_started(from))
+		to->sched.semaphore |= I915_SCHED_CHAIN_SEMAPHORE;
+
 	if (to->engine == from->engine) {
 		ret = i915_sw_fence_await_sw_fence_gfp(&to->submit,
 						       &from->submit,
@@ -1117,6 +1120,9 @@ void i915_request_add(struct i915_request *request)
 	if (engine->schedule) {
 		struct i915_sched_attr attr = request->gem_context->sched;
 
+		if (!request->sched.semaphore)
+			attr.priority |= I915_PRIORITY_NOSEMAPHORE;
+
 		/*
 		 * Boost priorities to new clients (new request flows).
 		 *
diff --git a/drivers/gpu/drm/i915/i915_scheduler.c b/drivers/gpu/drm/i915/i915_scheduler.c
index 7f9c4018530b..702204d4e729 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.c
+++ b/drivers/gpu/drm/i915/i915_scheduler.c
@@ -29,7 +29,7 @@ void i915_sched_node_init(struct i915_sched_node *node)
 	INIT_LIST_HEAD(&node->waiters_list);
 	INIT_LIST_HEAD(&node->link);
 	node->attr.priority = I915_PRIORITY_INVALID;
-	node->semaphore = false;
+	node->semaphore = 0;
 }
 
 static struct i915_dependency *
diff --git a/drivers/gpu/drm/i915/i915_scheduler.h b/drivers/gpu/drm/i915/i915_scheduler.h
index d764cf10536f..d84f09e8c248 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.h
+++ b/drivers/gpu/drm/i915/i915_scheduler.h
@@ -24,14 +24,15 @@ enum {
 	I915_PRIORITY_INVALID = INT_MIN
 };
 
-#define I915_USER_PRIORITY_SHIFT 2
+#define I915_USER_PRIORITY_SHIFT 3
 #define I915_USER_PRIORITY(x) ((x) << I915_USER_PRIORITY_SHIFT)
 
 #define I915_PRIORITY_COUNT BIT(I915_USER_PRIORITY_SHIFT)
 #define I915_PRIORITY_MASK (I915_PRIORITY_COUNT - 1)
 
-#define I915_PRIORITY_WAIT	((u8)BIT(0))
-#define I915_PRIORITY_NEWCLIENT	((u8)BIT(1))
+#define I915_PRIORITY_WAIT		((u8)BIT(0))
+#define I915_PRIORITY_NEWCLIENT		((u8)BIT(1))
+#define I915_PRIORITY_NOSEMAPHORE	((u8)BIT(2))
 
 struct i915_sched_attr {
 	/**
@@ -72,7 +73,9 @@ struct i915_sched_node {
 	struct list_head waiters_list; /* those after us, they depend upon us */
 	struct list_head link;
 	struct i915_sched_attr attr;
-	bool semaphore;
+	unsigned long semaphore;
+#define I915_SCHED_HAS_SEMAPHORE	BIT(0)
+#define I915_SCHED_CHAIN_SEMAPHORE	BIT(1)
 };
 
 struct i915_dependency {
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index f7f16b8d3422..4fcee493dddb 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -164,7 +164,7 @@
 #define WA_TAIL_DWORDS 2
 #define WA_TAIL_BYTES (sizeof(u32) * WA_TAIL_DWORDS)
 
-#define ACTIVE_PRIORITY (I915_PRIORITY_NEWCLIENT)
+#define ACTIVE_PRIORITY (I915_PRIORITY_NEWCLIENT | I915_PRIORITY_NOSEMAPHORE)
 
 static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 					    struct intel_engine_cs *engine,
@@ -608,7 +608,8 @@ static bool can_merge_rq(const struct i915_request *prev,
 	 * rules should mean that if this semaphore is preempted, its
 	 * dependency chain is preserved and suitably promoted via PI.
 	 */
-	if (prev->sched.semaphore && !i915_request_started(prev))
+	if (prev->sched.semaphore & I915_SCHED_HAS_SEMAPHORE &&
+	    !i915_request_started(prev))
 		return false;
 
 	if (!can_merge_ctx(prev->hw_context, next->hw_context))
-- 
2.20.1