This repository has been archived on 2023-10-09. You can view files and clone it. You cannot open issues or pull requests or push a commit.
Files
blender-archive/intern/cycles/device/opencl/opencl_mega.cpp
Lukas Stockner a2ebc5268f Cycles: Refactor Progress system to provide better estimates
The Progress system in Cycles had two limitations so far:
 - It just counted tiles, but ignored their size. For example, when rendering a 600x500 image with 512x512 tiles, the right 88x500 tile would count for 50% of the progress, although it only covers 15% of the image.
 - Scene update time was incorrectly counted as rendering time - therefore, the remaining time started very long and gradually decreased.

This patch fixes both problems:
First of all, the Progress now has a function to ignore time spans, and that is used to ignore scene update time.
The larger change is the tile size: Instead of counting samples per tile, so that the final value is num_samples*num_tiles, the code now counts every sample for every pixel, so that the final value is num_samples*num_pixels.

Along with that, some unused variables were removed from the Progress and Session classes.

Reviewers: brecht, sergey, #cycles

Subscribers: brecht, candreacchio, sergey

Differential Revision: https://developer.blender.org/D2214
2016-12-03 05:02:21 +01:00

155 lines
4.2 KiB
C++

/*
* Copyright 2011-2013 Blender Foundation
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#ifdef WITH_OPENCL
#include "opencl.h"
#include "buffers.h"
#include "kernel_types.h"
#include "util_md5.h"
#include "util_path.h"
#include "util_time.h"
CCL_NAMESPACE_BEGIN
class OpenCLDeviceMegaKernel : public OpenCLDeviceBase
{
public:
OpenCLProgram path_trace_program;
OpenCLDeviceMegaKernel(DeviceInfo& info, Stats &stats, bool background_)
: OpenCLDeviceBase(info, stats, background_),
path_trace_program(this, "megakernel", "kernel.cl", "-D__COMPILE_ONLY_MEGAKERNEL__ ")
{
}
virtual bool show_samples() const {
return true;
}
virtual void load_kernels(const DeviceRequestedFeatures& /*requested_features*/,
vector<OpenCLProgram*> &programs)
{
path_trace_program.add_kernel(ustring("path_trace"));
programs.push_back(&path_trace_program);
}
~OpenCLDeviceMegaKernel()
{
task_pool.stop();
path_trace_program.release();
}
void path_trace(RenderTile& rtile, int sample)
{
/* Cast arguments to cl types. */
cl_mem d_data = CL_MEM_PTR(const_mem_map["__data"]->device_pointer);
cl_mem d_buffer = CL_MEM_PTR(rtile.buffer);
cl_mem d_rng_state = CL_MEM_PTR(rtile.rng_state);
cl_int d_x = rtile.x;
cl_int d_y = rtile.y;
cl_int d_w = rtile.w;
cl_int d_h = rtile.h;
cl_int d_offset = rtile.offset;
cl_int d_stride = rtile.stride;
/* Sample arguments. */
cl_int d_sample = sample;
cl_kernel ckPathTraceKernel = path_trace_program(ustring("path_trace"));
cl_uint start_arg_index =
kernel_set_args(ckPathTraceKernel,
0,
d_data,
d_buffer,
d_rng_state);
#define KERNEL_TEX(type, ttype, name) \
set_kernel_arg_mem(ckPathTraceKernel, &start_arg_index, #name);
#include "kernel_textures.h"
#undef KERNEL_TEX
start_arg_index += kernel_set_args(ckPathTraceKernel,
start_arg_index,
d_sample,
d_x,
d_y,
d_w,
d_h,
d_offset,
d_stride);
enqueue_kernel(ckPathTraceKernel, d_w, d_h);
}
void thread_run(DeviceTask *task)
{
if(task->type == DeviceTask::FILM_CONVERT) {
film_convert(*task, task->buffer, task->rgba_byte, task->rgba_half);
}
else if(task->type == DeviceTask::SHADER) {
shader(*task);
}
else if(task->type == DeviceTask::PATH_TRACE) {
RenderTile tile;
/* Keep rendering tiles until done. */
while(task->acquire_tile(this, tile)) {
int start_sample = tile.start_sample;
int end_sample = tile.start_sample + tile.num_samples;
for(int sample = start_sample; sample < end_sample; sample++) {
if(task->get_cancel()) {
if(task->need_finish_queue == false)
break;
}
path_trace(tile, sample);
tile.sample = sample + 1;
task->update_progress(&tile, tile.w*tile.h);
}
/* Complete kernel execution before release tile */
/* This helps in multi-device render;
* The device that reaches the critical-section function
* release_tile waits (stalling other devices from entering
* release_tile) for all kernels to complete. If device1 (a
* slow-render device) reaches release_tile first then it would
* stall device2 (a fast-render device) from proceeding to render
* next tile.
*/
clFinish(cqCommandQueue);
task->release_tile(tile);
}
}
}
};
Device *opencl_create_mega_device(DeviceInfo& info, Stats& stats, bool background)
{
return new OpenCLDeviceMegaKernel(info, stats, background);
}
CCL_NAMESPACE_END
#endif