CMakeLists.txt 2.0 KB
Newer Older
Y
Yan Chunwei 已提交
1 2 3 4 5 6
if(NOT LITE_WITH_CUDA)
    return()
endif()

message(STATUS "compile with lite CUDA kernels")

7
nv_library(mul_compute_cuda SRCS mul_compute.cc DEPS ${lite_kernel_deps} context)
8

9
lite_cc_library(io_copy_compute_cuda SRCS io_copy_compute.cc DEPS ${lite_kernel_deps})
10

11
nv_library(leaky_relu_compute_cuda SRCS leaky_relu_compute.cu DEPS ${lite_kernel_deps})
12
nv_library(yolo_box_compute_cuda SRCS yolo_box_compute.cu DEPS ${lite_kernel_deps})
13
nv_library(transpose_compute_cuda SRCS transpose_compute.cu DEPS ${lite_kernel_deps} ${math_cuda})
14
nv_library(nearest_interp_compute_cuda SRCS nearest_interp_compute.cu DEPS ${lite_kernel_deps})
15 16 17
nv_library(conv2d_cuda SRCS conv_compute.cc DEPS ${lite_kernel_deps} ${math_cuda})
nv_library(concat_compute_cuda SRCS concat_compute.cu DEPS ${lite_kernel_deps})
nv_library(elementwise_add_compute_cuda SRCS elementwise_add_compute.cu DEPS ${lite_kernel_deps})
18

19
nv_test(conv2d_cuda_test SRCS conv_compute_test.cc DEPS conv2d_cuda)
20 21 22
nv_test(nearest_interp_compute_cuda_test SRCS nearest_interp_compute_test.cc DEPS nearest_interp_compute_cuda)
nv_test(leaky_relu_compute_cuda_test SRCS leaky_relu_compute_test.cc DEPS leaky_relu_compute_cuda)
nv_test(yolo_box_compute_cuda_test SRCS yolo_box_compute_test.cc DEPS yolo_box_compute_cuda)
23
nv_test(transpose_compute_cuda_test SRCS transpose_compute_test.cc DEPS transpose_compute_cuda)
24 25
nv_test(concat_compute_cuda_test SRCS concat_compute_test.cc DEPS concat_compute_cuda)
nv_test(elementwise_add_compute_cuda_test SRCS elementwise_add_compute_test.cc DEPS elementwise_add_compute_cuda)
26

Z
Zhen Wang 已提交
27 28 29
nv_library(calib_compute_cuda SRCS calib_compute.cu DEPS ${lite_kernel_deps})
lite_cc_test(calib_compute_cuda_test SRCS calib_compute_cuda_test.cc DEPS calib_compute_cuda)

30
set(cuda_kernels
31
conv2d_cuda
32
mul_compute_cuda
33 34
io_copy_compute_cuda
leaky_relu_compute_cuda
35
nearest_interp_compute_cuda
36 37
concat_compute_cuda
elementwise_add_compute_cuda
38
yolo_box_compute_cuda
39
transpose_compute_cuda
40 41 42
)

set(cuda_kernels "${cuda_kernels}" CACHE GLOBAL "cuda kernels")