Cycles: Add an AVX2 CPU kernel.

This kernel is compiled with AVX2, FMA3, and BMI compiler flags. At the moment only Intel Haswell benefits from this, but future AMD CPUs will have these instructions as well.

Makes rendering on Haswell CPUs a few percent faster, only benchmarked with clang on OS X though.

Part of my GSoC 2014.

1 files changed, 53 insertions, 0 deletions

diff --git a/intern/cycles/device/device_cpu.cpp b/intern/cycles/device/device_cpu.cpp
index 71bf2d23d6e..7308d036fe3 100644
--- a/intern/cycles/device/device_cpu.cpp
+++ b/intern/cycles/device/device_cpu.cpp
@@ -62,6 +62,7 @@ public:
 		system_cpu_support_sse3();
 		system_cpu_support_sse41();
 		system_cpu_support_avx();
+		system_cpu_support_avx2();
 	}
 
 	~CPUDevice()
@@ -167,6 +168,28 @@ public:
 			int start_sample = tile.start_sample;
 			int end_sample = tile.start_sample + tile.num_samples;
 
+#ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX2
+			if(system_cpu_support_avx2()) {
+				for(int sample = start_sample; sample < end_sample; sample++) {
+					if (task.get_cancel() || task_pool.canceled()) {
+						if(task.need_finish_queue == false)
+							break;
+					}
+
+					for(int y = tile.y; y < tile.y + tile.h; y++) {
+						for(int x = tile.x; x < tile.x + tile.w; x++) {
+							kernel_cpu_avx2_path_trace(&kg, render_buffer, rng_state,
+													  sample, x, y, tile.offset, tile.stride);
+						}
+					}
+
+					tile.sample = sample + 1;
+
+					task.update_progress(tile);
+				}
+			}
+			else
+#endif
 #ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX
 			if(system_cpu_support_avx()) {
 				for(int sample = start_sample; sample < end_sample; sample++) {
@@ -293,6 +316,15 @@ public:
 		float sample_scale = 1.0f/(task.sample + 1);
 
 		if(task.rgba_half) {
+#ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX2
+			if(system_cpu_support_avx2()) {
+				for(int y = task.y; y < task.y + task.h; y++)
+					for(int x = task.x; x < task.x + task.w; x++)
+						kernel_cpu_avx2_convert_to_half_float(&kernel_globals, (uchar4*)task.rgba_half, (float*)task.buffer,
+															 sample_scale, x, y, task.offset, task.stride);
+			}
+			else
+#endif
 #ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX
 			if(system_cpu_support_avx()) {
 				for(int y = task.y; y < task.y + task.h; y++)
@@ -337,6 +369,15 @@ public:
 			}
 		}
 		else {
+#ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX2
+			if(system_cpu_support_avx2()) {
+				for(int y = task.y; y < task.y + task.h; y++)
+					for(int x = task.x; x < task.x + task.w; x++)
+						kernel_cpu_avx2_convert_to_byte(&kernel_globals, (uchar4*)task.rgba_byte, (float*)task.buffer,
+													   sample_scale, x, y, task.offset, task.stride);
+			}
+			else
+#endif
 #ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX
 			if(system_cpu_support_avx()) {
 				for(int y = task.y; y < task.y + task.h; y++)
@@ -390,6 +431,18 @@ public:
 		OSLShader::thread_init(&kg, &kernel_globals, &osl_globals);
 #endif
 
+#ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX2
+		if(system_cpu_support_avx2()) {
+			for(int x = task.shader_x; x < task.shader_x + task.shader_w; x++) {
+				for(int sample = 0; sample < task.num_samples; sample++)
+					kernel_cpu_avx2_shader(&kg, (uint4*)task.shader_input, (float4*)task.shader_output, task.shader_eval_type, x, sample);
+
+				if(task.get_cancel() || task_pool.canceled())
+					break;
+			}
+		}
+		else
+#endif
 #ifdef WITH_CYCLES_OPTIMIZED_KERNEL_AVX
 		if(system_cpu_support_avx()) {
 			for(int x = task.shader_x; x < task.shader_x + task.shader_w; x++) {