Linux CPU+GPU |
Mac OS CPU+GPU |
Windows CPU |
Android |
Apple IOS |
License Scan |
|---|---|---|---|---|---|
bazel.cmake is a seamless submodule which aims to mimic the behavior of bazel to simplify the usability of CMake for any mainstream operating systems (Linux, Mac OS, Windows, Android, IOS).
When launching a large open source project, We found that not everyone can adapt to the tremendous higher-level mechanisms CMake offered. We borrow the abstracts from Bazel to make it easy for us to do so! Back to 2017, Bazel has not yet matured to support manycore accelerators and even the most popular operating system - Windows. Maybe it's better now, but CMake is still the most powerful building tool for C++. Last but not least, developers can leverage the elegant abstracts from this module to reach the fine-grained compilation and test. So we will continue to support this project.
If you create a new project, first add git submodule bazel.cmake into your project.
project_dir$ git submodule add --force https://github.com/gangliao/bazel.cmake
project_dir$ git submodule update --init --recursiveJust like our test directory, you need to integrate bazel.cmake module into
current project's CMakeLists.txt as follows:
# CMakeLists.txt
set(CMAKE_MODULE_PATH ${CMAKE_MODULE_PATH} ${CMAKE_CURRENT_SOURCE_DIR}/bazel.cmake/cmake)
include(bazel)Then, you can use the built-in bazel abstracts to compile your code and run it under any mainstream operating system!
To compile the following code, you can invoke cc_test in CMakeLists.txt.
// bazel.cmake/test/c++/hello.cc
#include <stdio.h>
int main(int argc, char *argv[]) {
printf("Hello World!\n");
return 0;
}# bazel.cmake/test/c++/CMakeLists.txt
cc_test(hello SRCS hello.cc)
# If gtest is being used in your code, please explicitly specify gtest in
# cc_test(xxx SRCS xxx.cc DEPS gtest) so that all dependent libraries
# (libgtest.a and libgtest_main.a) could be linked.Then, issue the below commands to build an executable for Mac OS X or Linux.
project_dir$ mkdir build && cd build
project_dir/build$ cmake ..
project_dir/build$ make hello
# Build Output
Scanning dependencies of target hello
[ 50%] Building CXX object c++/CMakeFiles/hello.dir/hello.cc.o
[100%] Linking CXX executable hello
[100%] Built target helloYou can verify all test cases in your project by env GTEST_COLOR=1 ctest -j4 or make tests.
project_dir/build$ env GTEST_COLOR=1 ctest
Test project github/bazel.cmake/test/build
Start 1: cpu_id_test
1/4 Test #1: cpu_id_test ...................... Passed 0.00 sec
Start 2: hello
2/4 Test #2: hello ............................ Passed 0.00 sec
Start 3: vector_add_test
3/4 Test #3: vector_add_test .................. Passed 0.51 sec
Start 4: test_add_person
4/4 Test #4: test_add_person .................. Passed 0.26 sec
100% tests passed, 0 tests failed out of 4Your host machine must be Windows!
mkdir build && cd build
cmake .. -G "Visual Studio 14 2015" -A x64 -DCMAKE_GENERATOR_PLATFORM=x64 -Wno-dev
# Build code
cmake --build . -- -maxcpucount
# Run test
cmake -E env CTEST_OUTPUT_ON_FAILURE=1 cmake --build . --target RUN_TESTS# Download and decompress Android ndk
wget -c https://dl.google.com/android/repository/android-ndk-r17-darwin-x86_64.zip && unzip -q android-ndk-r17-darwin-x86_64.zip
ANDROID_STANDALONE_TOOLCHAIN=`pwd`/android-toolchain-gcc
android-ndk-r17/build/tools/make-standalone-toolchain.sh --force --arch=arm --platform=android-21 --install-dir=$ANDROID_STANDALONE_TOOLCHAIN
# Create the build directory for CMake.
mkdir build && cd build
PROJECT_DIR=......
cmake -DCMAKE_TOOLCHAIN_FILE=$PROJECT_DIR/bazel.cmake/third-party/android-cmake/android.toolchain.cmake \
-DANDROID_STANDALONE_TOOLCHAIN=$ANDROID_STANDALONE_TOOLCHAIN \
-DANDROID_ABI="armeabi-v7a with NEON FP16" \
-DANDROID_NATIVE_API_LEVEL=21 \
-DWITH_PYTHON=OFF \
-DHOST_CXX_COMPILER=/usr/bin/c++ \
-DHOST_C_COMPILER=/usr/bin/cc \
..
make -j4mkdir build && cd build
PROJECT_DIR=......
cmake -DCMAKE_TOOLCHAIN_FILE=$PROJECT_DIR/bazel.cmake/third-party/ios-cmake/toolchain/iOS.cmake \
-DIOS_PLATFORM=OS \
-DWITH_PYTHON=OFF \
-DHOST_CXX_COMPILER=/usr/bin/c++ \
-DHOST_C_COMPILER=/usr/bin/cc \
..
make "-j$(sysctl -n hw.ncpu)"make helps can dump the manuals on the screen!
| APIs | Linux | Windows | Android | Mac OS X | Apple IOS |
|-----------------------------------------------------|-------|---------|---------|----------|-----------|
| cc_library(lib_name SRCS src1.cc... [DEPS lib1...]) | yes | yes | yes | yes | yes |
| cc_library(lib_name DEPS lib1...) | yes | TBD | yes | yes | yes |
| cc_test(bin_name SRCS src1.cc... [DEPS lib1...]) | yes | yes | yes | yes | yes |
| cc_binary(bin_name SRCS src1.cc... [DEPS lib1...]) | yes | yes | yes | yes | yes |
| nv_library(lib_name SRCS src1.cc... [DEPS lib1...]) | yes | yes | no | yes | no |
| nv_test(bin_name SRCS src1.cc... [DEPS lib1...]) | yes | yes | no | yes | no |
| nv_binary(bin_name SRCS src1.cc... [DEPS lib1...]) | yes | yes | no | yes | no |
| proto_library(lib_name SRCS src.proto [DEPS lib1]) | yes | yes | yes | yes | yes |
| py_test(name SRCS src.py ARGS arg1 [DEPS lib1]) | yes | yes | no | yes | no |
Note: [DEPS lib1...] is optional syntax rules.
# To build a static library example.a from example.cc using the system
# compiler (like GCC):
#
cc_library(example SRCS example.cc)
#
# To build a static library example.a from multiple source files
# example{1,2,3}.cc:
#
cc_library(example SRCS example1.cc example2.cc example3.cc)
#
# To build a shared library example.so from example.cc:
#
cc_library(example SHARED SRCS example.cc)
#
# To build a library using Nvidia's NVCC from .cu file(s), use the nv_
# prefixed version:
#
nv_library(example SRCS example.cu)
#
# To specify that a library new_example.a depends on other libraies:
#
cc_library(new_example SRCS new_example.cc DEPS example)
#
# Static libraries can be composed of other static libraries:
#
cc_library(composed DEPS dependent1 dependent2 dependent3)
#
# To build an executable binary file from some source files and
# dependent libraries:
#
cc_binary(example SRCS main.cc something.cc DEPS example1 example2)
#
# To build an executable binary file using NVCC, use the nv_ prefixed
# version:
#
nv_binary(example SRCS main.cc something.cu DEPS example1 example2)
#
# To build a unit test binary, which is an executable binary with
# GoogleTest linked:
#
cc_test(example_test SRCS example_test.cc DEPS example)
#
# To build a unit test binary using NVCC, use the nv_ prefixed version:
#
nv_test(example_test SRCS example_test.cu DEPS example)
#
# It is pretty often that executable and test binaries depend on
# pre-defined external libaries like glog and gflags defined in
# /cmake/external/*.cmake:
#
cc_test(example_test SRCS example_test.cc DEPS example glog gflags)
#
# To generate protobuf cpp code using protoc and build a protobuf library.
# It will also generate protobuf python code.
proto_library(example SRCS example.proto DEPS dependent1)
#
# To run a unit test using [env var=value python example.py args1 args2] after dependent1 completed.
py_test(example_test ENVS var=value SRCS example.py ARGS args1 args2 DEPS dependent1)
#