optimization fix
This commit is contained in:
@@ -5,7 +5,7 @@ CXX=g++ -ffast-math
|
||||
CXXFLAGS=-O3 -I$(CUDATK)/include -Iobjs_gpu/ -D_CUDA_
|
||||
#
|
||||
NVCC=nvcc
|
||||
NVCC_FLAGS=-O3 -arch=sm_35 -D_CUDA_ -I../ -Xptxas=-v -Iobjs_gpu/
|
||||
NVCC_FLAGS+=-O3 -arch=sm_35 -D_CUDA_ -I../ -Xptxas=-v -Iobjs_gpu/
|
||||
ifdef PTXCC_REGMAX
|
||||
NVCC_FLAGS += --maxrregcount=$(PTXCC_REGMAX)
|
||||
endif
|
||||
|
||||
@@ -1,6 +1,8 @@
|
||||
PROG=radixSort
|
||||
ISPC_SRC=radixSort.ispc
|
||||
|
||||
CU_SRC=radixSort.cu
|
||||
NVCC_FLAGS=-Xptxas=-O1
|
||||
CXX_SRC=radixSort.cpp radixSort.cpp
|
||||
PTXCC_REGMAX=64
|
||||
|
||||
|
||||
Reference in New Issue
Block a user