Cycles / CUDA Kernel compile:

* Added option "WITH_BF_CYCLES_CUDA_THREADED_COMPILE" for the people who have much RAM (8 or more) and can compile several kernels at the same time. If enabled, it uses the general BF_NUMJOBS flag.
* The option is off per default.

diff --git a/build_files/scons/tools/btools.py b/build_files/scons/tools/btools.py
index 98c1b74d5ccc0a344a10e153526ac833c2aa2e1e..05e28ab1813f281066908a5dc9710b955e423707 100644 (file)
--- a/build_files/scons/tools/btools.py
+++ b/build_files/scons/tools/btools.py
@@ -157,7 +157,7 @@ def validate_arguments(args, bc):
             'WITH_BF_JEMALLOC', 'WITH_BF_STATICJEMALLOC', 'BF_JEMALLOC', 'BF_JEMALLOC_INC', 'BF_JEMALLOC_LIBPATH', 'BF_JEMALLOC_LIB', 'BF_JEMALLOC_LIB_STATIC',
             'BUILDBOT_BRANCH',
             'WITH_BF_3DMOUSE', 'WITH_BF_STATIC3DMOUSE', 'BF_3DMOUSE', 'BF_3DMOUSE_INC', 'BF_3DMOUSE_LIB', 'BF_3DMOUSE_LIBPATH', 'BF_3DMOUSE_LIB_STATIC',
-            'WITH_BF_CYCLES', 'WITH_BF_CYCLES_CUDA_BINARIES' 'BF_CYCLES_CUDA_NVCC', 'BF_CYCLES_CUDA_NVCC',
+            'WITH_BF_CYCLES', 'WITH_BF_CYCLES_CUDA_BINARIES' 'BF_CYCLES_CUDA_NVCC', 'BF_CYCLES_CUDA_NVCC', 'WITH_BF_CYCLES_CUDA_THREADED_COMPILE',
             'WITH_BF_OIIO', 'WITH_BF_STATICOIIO', 'BF_OIIO', 'BF_OIIO_INC', 'BF_OIIO_LIB', 'BF_OIIO_LIB_STATIC', 'BF_OIIO_LIBPATH',
             'WITH_BF_BOOST', 'WITH_BF_STATICBOOST', 'BF_BOOST', 'BF_BOOST_INC', 'BF_BOOST_LIB', 'BF_BOOST_LIB_STATIC', 'BF_BOOST_LIBPATH'
             ]
@@ -545,6 +545,7 @@ def read_opts(env, cfg, args):
     localopts.AddVariables(
         (BoolVariable('WITH_BF_CYCLES', 'Build with the Cycles engine', True)),
         (BoolVariable('WITH_BF_CYCLES_CUDA_BINARIES', 'Build with precompiled CUDA binaries', False)),
+        (BoolVariable('WITH_BF_CYCLES_CUDA_THREADED_COMPILE', 'Build several render kernels at once (using BF_NUMJOBS)', False)),
         ('BF_CYCLES_CUDA_NVCC', 'CUDA nvcc compiler path', ''),
         ('BF_CYCLES_CUDA_BINARIES_ARCH', 'CUDA architectures to compile binaries for', []),
 

diff --git a/intern/cycles/kernel/SConscript b/intern/cycles/kernel/SConscript
index 7749f94818e794a27e7e2b5a3fe229a643fd4467..eb99680efe4eb8a2516da085520f2857e76eb889 100644 (file)
--- a/intern/cycles/kernel/SConscript
+++ b/intern/cycles/kernel/SConscript
@@ -46,11 +46,12 @@ if env['WITH_BF_CYCLES_CUDA_BINARIES']:
         kernel.Depends(cubin_file, dependencies)
 
         kernel_binaries.append(cubin_file)
-
-        # trick to compile one kernel at a time to reduce memory usage
-        if last_cubin_file:
-            kernel.Depends(cubin_file, last_cubin_file)
-        last_cubin_file = cubin_file
+        
+        if not env['WITH_BF_CYCLES_CUDA_THREADED_COMPILE']:
+            # trick to compile one kernel at a time to reduce memory usage
+            if last_cubin_file:
+                kernel.Depends(cubin_file, last_cubin_file)
+            last_cubin_file = cubin_file
 
 Return('kernel_binaries')