drm/i915: Lift timeline ordering to await_dma_fence

Currently we filter out repeated use of the same timeline in the low
level i915_gem_request_await_request(), after having added the
dependency on the old request. However, we can lift this to
i915_gem_request_await_dma_fence() (before the dependency is added)
using the observation that requests along the same timeline are
explicitly ordered via i915_add_request (along with the dependencies).
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170503093924.5320-3-chris@chris-wilson.co.uk

drivers/gpu/drm/i915/i915_gem_request.c

@@ -687,6 +687,7 @@ i915_gem_request_await_request(struct drm_i915_gem_request *to,
 	int ret;
 
 	GEM_BUG_ON(to == from);
+	GEM_BUG_ON(to->timeline == from->timeline);
 
 	if (i915_gem_request_completed(from))
 		return 0;
@@ -699,9 +700,6 @@ i915_gem_request_await_request(struct drm_i915_gem_request *to,
 			return ret;
 	}
 
-	if (to->timeline == from->timeline)
-		return 0;
-
 	if (to->engine == from->engine) {
 		ret = i915_sw_fence_await_sw_fence_gfp(&to->submit,
 						       &from->submit,
@@ -767,6 +765,14 @@ i915_gem_request_await_dma_fence(struct drm_i915_gem_request *req,
 		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
 			continue;
 
+		/*
+		 * Requests on the same timeline are explicitly ordered, along
+		 * with their dependencies, by i915_add_request() which ensures
+		 * that requests are submitted in-order through each ring.
+		 */
+		if (fence->context == req->fence.context)
+			continue;
+
 		if (dma_fence_is_i915(fence))
 			ret = i915_gem_request_await_request(req,
 							     to_request(fence));