[Intel-gfx] [PATCH v5 02/35] drm/i915: Prelude to splitting i915_gem_do_execbuffer in two

[John.C.Harrison at Intel.com]

From: John Harrison <John.C.Harrison at Intel.com>

The scheduler decouples the submission of batch buffers to the driver
with their submission to the hardware. This basically means splitting
the execbuffer() function in half. This change rearranges some code
ready for the split to occur.

v5: Dropped runtime PM calls as they conflict with the mutex lock.
Instead of being done at the lowest submission level, they are now
left right at the top driver entry level. [feedback from Chris Wilson]

For: VIZ-1587
Signed-off-by: John Harrison <John.C.Harrison at Intel.com>
Reviewed-by: Jesse Barnes <jbarnes at virtuousgeek.org>
---
 drivers/gpu/drm/i915/i915_gem_execbuffer.c | 41 ++++++++++++++++++------------
 drivers/gpu/drm/i915/intel_lrc.c           | 18 +++++++++----
 2 files changed, 38 insertions(+), 21 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index e3e3666..d09130f 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -966,10 +966,7 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
 	if (flush_domains & I915_GEM_DOMAIN_GTT)
 		wmb();
 
-	/* Unconditionally invalidate gpu caches and ensure that we do flush
-	 * any residual writes from the previous batch.
-	 */
-	return intel_ring_invalidate_all_caches(req);
+	return 0;
 }
 
 static bool
@@ -1237,17 +1234,6 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 	u32 instp_mask;
 	int ret;
 
-	ret = i915_gem_execbuffer_move_to_gpu(params->request, vmas);
-	if (ret)
-		return ret;
-
-	ret = i915_switch_context(params->request);
-	if (ret)
-		return ret;
-
-	WARN(params->ctx->ppgtt && params->ctx->ppgtt->pd_dirty_rings & (1<<ring->id),
-	     "%s didn't clear reload\n", ring->name);
-
 	instp_mode = args->flags & I915_EXEC_CONSTANTS_MASK;
 	instp_mask = I915_EXEC_CONSTANTS_MASK;
 	switch (instp_mode) {
@@ -1281,6 +1267,30 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 		return -EINVAL;
 	}
 
+	ret = i915_gem_execbuffer_move_to_gpu(params->request, vmas);
+	if (ret)
+		return ret;
+
+	i915_gem_execbuffer_move_to_active(vmas, params->request);
+
+	/* To be split into two functions here... */
+
+	/*
+	 * Unconditionally invalidate gpu caches and ensure that we do flush
+	 * any residual writes from the previous batch.
+	 */
+	ret = intel_ring_invalidate_all_caches(params->request);
+	if (ret)
+		return ret;
+
+	/* Switch to the correct context for the batch */
+	ret = i915_switch_context(params->request);
+	if (ret)
+		return ret;
+
+	WARN(params->ctx->ppgtt && params->ctx->ppgtt->pd_dirty_rings & (1<<ring->id),
+	     "%s didn't clear reload\n", ring->name);
+
 	if (ring == &dev_priv->ring[RCS] &&
 	    instp_mode != dev_priv->relative_constants_mode) {
 		ret = intel_ring_begin(params->request, 4);
@@ -1317,7 +1327,6 @@ i915_gem_ringbuffer_submission(struct i915_execbuffer_params *params,
 
 	trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags);
 
-	i915_gem_execbuffer_move_to_active(vmas, params->request);
 	i915_gem_execbuffer_retire_commands(params);
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 5657924..53aeb55 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -678,10 +678,7 @@ static int execlists_move_to_gpu(struct drm_i915_gem_request *req,
 	if (flush_domains & I915_GEM_DOMAIN_GTT)
 		wmb();
 
-	/* Unconditionally invalidate gpu caches and ensure that we do flush
-	 * any residual writes from the previous batch.
-	 */
-	return logical_ring_invalidate_all_caches(req);
+	return 0;
 }
 
 int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request)
@@ -957,6 +954,18 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 	if (ret)
 		return ret;
 
+	i915_gem_execbuffer_move_to_active(vmas, params->request);
+
+	/* To be split into two functions here... */
+
+	/*
+	 * Unconditionally invalidate gpu caches and ensure that we do flush
+	 * any residual writes from the previous batch.
+	 */
+	ret = logical_ring_invalidate_all_caches(params->request);
+	if (ret)
+		return ret;
+
 	if (ring == &dev_priv->ring[RCS] &&
 	    instp_mode != dev_priv->relative_constants_mode) {
 		ret = intel_logical_ring_begin(params->request, 4);
@@ -981,7 +990,6 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
 
 	trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags);
 
-	i915_gem_execbuffer_move_to_active(vmas, params->request);
 	i915_gem_execbuffer_retire_commands(params);
 
 	return 0;
-- 
1.9.1