Cycles:
authorBrecht Van Lommel <brechtvanlommel@pandora.be>
Fri, 9 Sep 2011 12:04:39 +0000 (12:04 +0000)
committerBrecht Van Lommel <brechtvanlommel@pandora.be>
Fri, 9 Sep 2011 12:04:39 +0000 (12:04 +0000)
commitcfbd6cf154bbc653422f30b2bf8077545f5fb99c
treeb32a82345cf6d87f56ede78a8bdf1cd18ca2f0f3
parent9b31cba74e2bd84e9988ebdab723e6e43f9b8357
Cycles:
* OpenCL now only uses GPU/Accelerator devices, it's only confusing if CPU
  device is used, easy to enable in the code for debugging.
* OpenCL kernel binaries are now cached for faster startup after the first
  time compiling.
* CUDA kernels can now be compiled and cached at runtime if the CUDA toolkit
  is installed. This means that even if the build does not have CUDA enabled,
  it's still possible to use it as long as you install the toolkit.
intern/cycles/CMakeLists.txt
intern/cycles/blender/addon/engine.py
intern/cycles/blender/blender_python.cpp
intern/cycles/device/device_cuda.cpp
intern/cycles/device/device_opencl.cpp
intern/cycles/kernel/CMakeLists.txt
intern/cycles/kernel/kernel_compat_cuda.h
intern/cycles/render/session.cpp
intern/cycles/util/util_cuda.cpp
intern/cycles/util/util_path.cpp
intern/cycles/util/util_path.h