Cycles: Add sample-based runtime profiler that measures time spent in various parts of the CPU kernel

This commit adds a sample-based profiler that runs during CPU rendering and collects statistics on time spent in different parts of the kernel (ray intersection, shader evaluation etc.) as well as time spent per material and object.

The results are currently not exposed in the user interface or per Python yet, to see the stats on the console pass the "--cycles-print-stats" argument to Cycles (e.g. "./blender -- --cycles-print-stats").

Unfortunately, there is no clear way to extend this functionality to CUDA or OpenCL, so it is CPU-only for now.

Reviewers: brecht, sergey, swerner

Reviewed By: brecht, swerner

Differential Revision: https://developer.blender.org/D3892

1 files changed, 9 insertions, 1 deletions

diff --git a/intern/cycles/render/session.h b/intern/cycles/render/session.h
index 56a69919a7a..db07c758427 100644
--- a/intern/cycles/render/session.h
+++ b/intern/cycles/render/session.h
@@ -20,6 +20,7 @@
 #include "render/buffers.h"
 #include "device/device.h"
 #include "render/shader.h"
+#include "render/stats.h"
 #include "render/tile.h"
 
 #include "util/util_progress.h"
@@ -55,6 +56,8 @@ public:
 	int pixel_size;
 	int threads;
 
+	bool use_profiling;
+
 	bool display_buffer_linear;
 
 	bool use_denoising;
@@ -89,6 +92,8 @@ public:
 		pixel_size = 1;
 		threads = 0;
 
+		use_profiling = false;
+
 		use_denoising = false;
 		denoising_passes = false;
 		denoising_radius = 8;
@@ -118,6 +123,7 @@ public:
 		&& start_resolution == params.start_resolution
 		&& pixel_size == params.pixel_size
 		&& threads == params.threads
+		&& use_profiling == params.use_profiling
 		&& display_buffer_linear == params.display_buffer_linear
 		&& cancel_timeout == params.cancel_timeout
 		&& reset_timeout == params.reset_timeout
@@ -159,7 +165,7 @@ public:
 	void set_samples(int samples);
 	void set_pause(bool pause);
 
-	void update_scene();
+	bool update_scene();
 	void load_kernels(bool lock_scene=true);
 
 	void device_free();
@@ -168,6 +174,8 @@ public:
 	 * (for example, when rendering with unlimited samples). */
 	float get_progress();
 
+	void collect_statistics(RenderStats *stats);
+
 protected:
 	struct DelayedReset {
 		thread_mutex mutex;