forked from bartvdbraak/blender
27d660ad20
Currently only summed number of traversal steps and intersections used by the camera ray intersection pass is implemented, but in the future we will support more debug passes which would help checking what things makes the scene slow. Example of such extra passes could be number of bounces, time spent on the shader tree evaluation and so. Implementation from the Cycles side is pretty much straightforward, could only mention here that it's a build-time option disabled by default. From the blender side it's implemented as a PASS_DEBUG with several subtypes possible. This way we don't need to create an extra DNA pass type for each of the debug passes, saving us a bits. Reviewers: campbellbarton Reviewed By: campbellbarton Differential Revision: https://developer.blender.org/D813
149 lines
6.0 KiB
Python
149 lines
6.0 KiB
Python
#!/usr/bin/env python
|
|
#
|
|
# ***** BEGIN GPL LICENSE BLOCK *****
|
|
#
|
|
# This program is free software; you can redistribute it and/or
|
|
# modify it under the terms of the GNU General Public License
|
|
# as published by the Free Software Foundation; either version 2
|
|
# of the License, or (at your option) any later version.
|
|
#
|
|
# This program is distributed in the hope that it will be useful,
|
|
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
# GNU General Public License for more details.
|
|
#
|
|
# You should have received a copy of the GNU General Public License
|
|
# along with this program; if not, write to the Free Software Foundation,
|
|
# Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
|
|
#
|
|
# The Original Code is Copyright (C) 2011, Blender Foundation
|
|
# All rights reserved.
|
|
#
|
|
# The Original Code is: all of this file.
|
|
#
|
|
# Contributor(s): Nathan Letwory.
|
|
#
|
|
# ***** END GPL LICENSE BLOCK *****
|
|
|
|
from os import path
|
|
Import('env')
|
|
|
|
cycles = env.Clone()
|
|
|
|
cycles.Depends('../../source/blender/makesrna/intern/RNA_blender_cpp.h', 'makesrna')
|
|
|
|
sources = cycles.Glob('bvh/*.cpp') + cycles.Glob('device/*.cpp') + cycles.Glob('kernel/*.cpp') + cycles.Glob('render/*.cpp') + cycles.Glob('subd/*.cpp') + cycles.Glob('util/*.cpp') + cycles.Glob('blender/*.cpp')
|
|
|
|
sources.remove(path.join('util', 'util_view.cpp'))
|
|
sources.remove(path.join('kernel', 'kernel_sse2.cpp'))
|
|
sources.remove(path.join('kernel', 'kernel_sse3.cpp'))
|
|
sources.remove(path.join('kernel', 'kernel_sse41.cpp'))
|
|
sources.remove(path.join('kernel', 'kernel_avx.cpp'))
|
|
sources.remove(path.join('kernel', 'kernel_avx2.cpp'))
|
|
|
|
incs = []
|
|
defs = []
|
|
cxxflags = Split(env['CXXFLAGS'])
|
|
|
|
defs.append('GLEW_STATIC')
|
|
|
|
defs.append('CCL_NAMESPACE_BEGIN=namespace ccl {')
|
|
defs.append('CCL_NAMESPACE_END=}')
|
|
|
|
defs.append('WITH_OPENCL')
|
|
defs.append('WITH_MULTI')
|
|
defs.append('WITH_CUDA')
|
|
|
|
if env['WITH_BF_CYCLES_OSL']:
|
|
defs.append('WITH_OSL')
|
|
defs.append('OSL_STATIC_LIBRARY')
|
|
incs.append(cycles['BF_OSL_INC'])
|
|
|
|
if env['WITH_BF_CYCLES_DEBUG']:
|
|
defs.append('WITH_CYCLES_DEBUG')
|
|
|
|
incs.extend('. bvh render device kernel kernel/osl kernel/svm util subd'.split())
|
|
incs.extend('#intern/guardedalloc #source/blender/makesrna #source/blender/makesdna #source/blender/blenlib'.split())
|
|
incs.extend('#source/blender/blenloader ../../source/blender/makesrna/intern'.split())
|
|
incs.extend('#extern/glew/include #extern/clew/include #extern/cuew/include #intern/mikktspace'.split())
|
|
incs.append(cycles['BF_OIIO_INC'])
|
|
incs.append(cycles['BF_BOOST_INC'])
|
|
incs.append(cycles['BF_OPENEXR_INC'].split())
|
|
incs.extend(cycles['BF_PYTHON_INC'].split())
|
|
|
|
if env['OURPLATFORM'] in ('win32-vc', 'win64-vc'):
|
|
cxxflags.append('-D_CRT_SECURE_NO_WARNINGS /fp:fast /EHsc'.split())
|
|
else:
|
|
cxxflags.append('-ffast-math'.split())
|
|
|
|
# Warnings
|
|
# XXX Not supported by gcc < 4.9, since we do not have any 'supported flags' test as in cmake,
|
|
# simpler to comment for now.
|
|
#if env['C_COMPILER_ID'] == 'gcc':
|
|
# cxxflags.append(['-Werror=float-conversion'])
|
|
|
|
if env['OURPLATFORM'] in ('win32-vc', 'win32-mingw', 'linuxcross', 'win64-vc', 'win64-mingw'):
|
|
incs.append(env['BF_PTHREADS_INC'])
|
|
|
|
# optimized kernel. we compile the kernel multiple times with different
|
|
# optimization flags, at runtime it will choose the optimal kernel
|
|
kernel_flags = {}
|
|
|
|
if env['OURPLATFORM'] == 'win32-vc':
|
|
# there is no /arch:SSE3, but intrinsics are available anyway
|
|
kernel_flags['sse2'] = '/arch:SSE /arch:SSE2 -D_CRT_SECURE_NO_WARNINGS /fp:fast /Ox /GS-'
|
|
kernel_flags['sse3'] = kernel_flags['sse2']
|
|
|
|
elif env['OURPLATFORM'] == 'win64-vc':
|
|
# /arch:AVX only available from visual studio 2012
|
|
kernel_flags['sse2'] = '-D_CRT_SECURE_NO_WARNINGS /fp:fast /Ox /GS-'
|
|
kernel_flags['sse3'] = kernel_flags['sse2']
|
|
|
|
if env['MSVC_VERSION'] >= '12.0':
|
|
kernel_flags['sse41'] = kernel_flags['sse3']
|
|
kernel_flags['avx'] = kernel_flags['sse41'] + ' /arch:AVX'
|
|
kernel_flags['avx2'] = kernel_flags['sse41'] + ' /arch:AVX /arch:AVX2'
|
|
else:
|
|
# -mavx only available with relatively new gcc/clang
|
|
kernel_flags['sse2'] = '-ffast-math -msse -msse2 -mfpmath=sse'
|
|
kernel_flags['sse3'] = kernel_flags['sse2'] + ' -msse3 -mssse3'
|
|
kernel_flags['sse41'] = kernel_flags['sse3'] + ' -msse4.1'
|
|
|
|
if (env['C_COMPILER_ID'] == 'gcc' and env['CCVERSION'] >= '4.6') or (env['C_COMPILER_ID'] == 'clang' and env['CCVERSION'] >= '3.1'):
|
|
kernel_flags['avx'] = kernel_flags['sse41'] + ' -mavx'
|
|
kernel_flags['avx2'] = kernel_flags['avx'] + ' -mavx2 -mfma -mlzcnt -mbmi -mbmi2'
|
|
|
|
for kernel_type in kernel_flags.keys():
|
|
defs.append('WITH_KERNEL_' + kernel_type.upper())
|
|
|
|
for kernel_type in kernel_flags.keys():
|
|
kernel_source = path.join('kernel', 'kernel_' + kernel_type + '.cpp')
|
|
kernel_cxxflags = Split(env['CXXFLAGS'])
|
|
kernel_cxxflags.append(kernel_flags[kernel_type].split())
|
|
kernel_defs = defs[:]
|
|
kernel_env = cycles.Clone()
|
|
|
|
if env['OURPLATFORM'] == 'darwin' and env['C_COMPILER_ID'] == 'gcc' and env['CCVERSION'] >= '4.6':
|
|
# use Apple assembler for avx , gnu-compilers do not support it ( gnu gcc-4.6 or higher case )
|
|
kernel_env.BlenderLib('bf_intern_cycles_' + kernel_type, [kernel_source], incs, kernel_defs,
|
|
libtype=['intern'], priority=[10], cxx_compileflags=kernel_cxxflags,
|
|
cc_compilerchange='/usr/bin/clang', cxx_compilerchange='/usr/bin/clang++')
|
|
else:
|
|
kernel_env.BlenderLib('bf_intern_cycles_' + kernel_type, [kernel_source], incs, kernel_defs,
|
|
libtype=['intern'], priority=[10], cxx_compileflags=kernel_cxxflags)
|
|
|
|
cycles.BlenderLib('bf_intern_cycles', sources, incs, defs, libtype=['intern'], priority=[0], cxx_compileflags=cxxflags)
|
|
|
|
# OSL shaders
|
|
if env['WITH_BF_CYCLES_OSL']:
|
|
oso_files = SConscript(['kernel/shaders/SConscript'])
|
|
cycles.Depends("kernel/osl/osl_shader.o", oso_files)
|
|
|
|
SConscript(['kernel/osl/SConscript'])
|
|
|
|
# cuda kernel binaries
|
|
if env['WITH_BF_CYCLES_CUDA_BINARIES']:
|
|
kernel_binaries = SConscript(['kernel/SConscript'])
|
|
cycles.Depends("device/device_cuda.o", kernel_binaries)
|
|
|