diff options
author | Brecht Van Lommel <brechtvanlommel@pandora.be> | 2012-09-04 17:29:07 +0400 |
---|---|---|
committer | Brecht Van Lommel <brechtvanlommel@pandora.be> | 2012-09-04 17:29:07 +0400 |
commit | adea12cb01e4c4f18f345dfbbf49e9e622192e4e (patch) | |
tree | b43018344c696e4d59437fabc7f17f5b9d6a8e80 /intern/cycles/device/device_cpu.cpp | |
parent | 68563134d4800be4eb46aa6b598fd719cdaf2980 (diff) |
Cycles: merge of changes from tomato branch.
Regular rendering now works tiled, and supports save buffers to save memory
during render and cache render results.
Brick texture node by Thomas.
http://wiki.blender.org/index.php/Doc:2.6/Manual/Render/Cycles/Nodes/Textures#Brick_Texture
Image texture Blended Box Mapping.
http://wiki.blender.org/index.php/Doc:2.6/Manual/Render/Cycles/Nodes/Textures#Image_Texture
http://mango.blender.org/production/blended_box/
Various bug fixes by Sergey and Campbell.
* Fix for reading freed memory in some node setups.
* Fix incorrect memory read when synchronizing mesh motion.
* Fix crash appearing when direct light usage is different on different layers.
* Fix for vector pass gives wrong result in some circumstances.
* Fix for wrong resolution used for rendering Render Layer node.
* Option to cancel rendering when doing initial synchronization.
* No more texture limit when using CPU render.
* Many fixes for new tiled rendering.
Diffstat (limited to 'intern/cycles/device/device_cpu.cpp')
-rw-r--r-- | intern/cycles/device/device_cpu.cpp | 69 |
1 files changed, 49 insertions, 20 deletions
diff --git a/intern/cycles/device/device_cpu.cpp b/intern/cycles/device/device_cpu.cpp index 070b20aec49..4c54671b0d0 100644 --- a/intern/cycles/device/device_cpu.cpp +++ b/intern/cycles/device/device_cpu.cpp @@ -27,6 +27,8 @@ #include "osl_shader.h" +#include "buffers.h" + #include "util_debug.h" #include "util_foreach.h" #include "util_function.h" @@ -141,28 +143,56 @@ public: OSLShader::thread_init(kg); #endif -#ifdef WITH_OPTIMIZED_KERNEL - if(system_cpu_support_optimized()) { - for(int y = task.y; y < task.y + task.h; y++) { - for(int x = task.x; x < task.x + task.w; x++) - kernel_cpu_optimized_path_trace(kg, (float*)task.buffer, (unsigned int*)task.rng_state, - task.sample, x, y, task.offset, task.stride); + RenderTile tile; + + while(task.acquire_tile(this, tile)) { + float *render_buffer = (float*)tile.buffer; + uint *rng_state = (uint*)tile.rng_state; + int start_sample = tile.start_sample; + int end_sample = tile.start_sample + tile.num_samples; - if(task_pool.cancelled()) - break; +#ifdef WITH_OPTIMIZED_KERNEL + if(system_cpu_support_optimized()) { + for(int sample = start_sample; sample < end_sample; sample++) { + if (task.get_cancel() || task_pool.cancelled()) + 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_optimized_path_trace(kg, render_buffer, rng_state, + sample, x, y, tile.offset, tile.stride); + } + } + + tile.sample = sample + 1; + + task.update_progress(tile); + } } - } - else + else #endif - { - for(int y = task.y; y < task.y + task.h; y++) { - for(int x = task.x; x < task.x + task.w; x++) - kernel_cpu_path_trace(kg, (float*)task.buffer, (unsigned int*)task.rng_state, - task.sample, x, y, task.offset, task.stride); - - if(task_pool.cancelled()) - break; + { + for(int sample = start_sample; sample < end_sample; sample++) { + if (task.get_cancel() || task_pool.cancelled()) + 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_path_trace(kg, render_buffer, rng_state, + sample, x, y, tile.offset, tile.stride); + } + } + + tile.sample = sample + 1; + + task.update_progress(tile); + } } + + task.release_tile(tile); + + if(task_pool.cancelled()) + break; } #ifdef WITH_OSL @@ -228,8 +258,7 @@ public: /* split task into smaller ones, more than number of threads for uneven * workloads where some parts of the image render slower than others */ list<DeviceTask> tasks; - - task.split(tasks, TaskScheduler::num_threads()*10); + task.split(tasks, TaskScheduler::num_threads()+1); foreach(DeviceTask& task, tasks) task_pool.push(new CPUDeviceTask(this, task)); |