In order to ensure that my Linux builds are identical regardless of the distribution the build host uses, I have packaged up my compiler and the sysroot files into a relocatable tar file and checked that into source control.
So the first step in any build (or at least, a step that must be invoked before any compile step) must be to extract this tar file.
If I was using a makefile, this would be simple to do. However, the project is using cmake and I can't figure out any way to do it with cmake. It might even be that I need this extract step invoked before cmake starts to detect the compiler: I can hard-code the compiler name but if cmake fails if it can't find the compiler then I need the unpack to happen before that test.
Is this possible with cmake?
You can use execute_process to invoke cmake's cross-platform command mode (cmake -E tar). The command would be something like:
execute_process(COMMAND ${CMAKE_COMMAND} -E tar xvf MyCompiler.bz2)
The command which causes CMake to check for a valid compiler is project, so as long as you have your execute_process call before the project call, the unpacking will be done before the compiler check.
Related
Let's say I have a Python script which does something with just built executable. And I want CMake to rebuild that executable if the script was modified (actually it is enough to just re-run the script, but rebuild an executable is fine too).
add_executable(App src/main.cpp)
add_custom_command(
TARGET App
POST_BUILD
COMMAND "${Python3_EXECUTABLE}" ARGS "scripts/do_stuff.py" "$<TARGET_FILE:App>"
WORKING_DIRECTORY "${CMAKE_CURRENT_LIST_DIR}"
)
How can I achieve that? add_custom_command with TARGET argument doesn't support DEPENDS argument. add_dependency(App "scripts/do_stuff.py") produces an error, because "scripts/do_stuff.py" is not a target, but just a file.
Running the script is very important for correct working of the executable so I don't want define completely separate target via add_custom_command allowing bypass script execution by building just App target.
actually it is enough to just re-run the script
So the executable does not depend on the script. So re-run the script, not the executable.
add_executable(app src/main.cpp)
add_custom_command(
OUTPUT ${CMAKE_CURRENT_BINARY_DIR}/did_do_stuff
COMMAND "${Python3_EXECUTABLE}" "scripts/do_stuff.py" "$<TARGET_FILE:app>"
COMMAND ${CMAKE_COMMAND} -E touch ${CMAKE_CURRENT_BINARY_DIR}/did_do_stuff
DEPENDS "$<TARGET_FILE:app>"
"${CMAKE_CURRENT_LIST_DIR}/scripts/do_stuff.py"
WORKING_DIRECTORY "${CMAKE_CURRENT_LIST_DIR}"
)
add_custom_target(do_stuff
DEPENDS ${CMAKE_CURRENT_BINARY_DIR}/did_do_stuff
)
And build target do_stuff (or all) to run it.
My question is very similar to cmake: add_custom_command / add_custom_target ignoring dependency
But the answer specified does not solve my issue, and also it is for a newer version of cmake (3.20)
I want some files (shader files) to be copied to the executable directory every time the shader source changes
So I have the following code in cmake:
add_custom_command(
OUTPUT ${CMAKE_BINARY_DIR}/Shaders.txt
COMMAND ${CMAKE_COMMAND} -E echo "Actually Copying shaders"
COMMAND ${CMAKE_COMMAND} -E copy_directory ${CMAKE_CURRENT_SOURCE_DIR}/assets/ $<TARGET_FILE_DIR:Editor>/assets
COMMAND ${CMAKE_COMMAND} -E touch ${CMAKE_BINARY_DIR}/Shaders.txt
DEPENDS ${CMAKE_CURRENT_SOURCE_DIR}/assets
)
add_custom_target(CopyShaders ALL DEPENDS ${CMAKE_BINARY_DIR}/Shaders.txt)
Now, as I understand, CopyShaders will always be built (since it is always out of date), but Shaders.txt should only be built once (after the shader source is changed), after which it is up to date
I'm trying to build the target CopyShaders using cmake --build build --target CopyShaders
If I'm using MinGW, then "Actually copying shaders" never gets printed if Shaders.txt is present, even if the assets folder containing the shaders has been modified
If I'm using MSVC, I get the following warning on the terminal:
warning MSB8064: Custom build for item "D:\Acads\Programming\opengl\SummerOfCode\build\CMakeFiles\05fb3856b7a5e1f6ce1ea66ca9091779\Shaders.txt.rule" succeeded, but specified dependency "d:\acads\programming\opengl\summerofcode\editor\assets" does not exist. This may cause incremental build to work incorrectly. [D:\Acads\Programming\opengl\SummerOfCode\build\Editor\CopyShaders.vcxproj]
Again, I do have the folder existing
I figured it out
DEPENDS should only take as input files, not folders
Replacing it with all the files inside /assets works
I would like to run program to perform do extra installation tasks from CMake. My attempted solution, based on INSTALL(CODE ...) is (this is a real MWE):
macro(MY_EXTRA_STUFF ARG)
execute_process(...)
endmacro()
install(CODE "MY_EXTRA_STUFF(${SOME_ARG})")
but CMake complains when I run ninja install (or make install, depending on generator in use):
[0/1] Install the project...
-- Install configuration: ""
CMake Error at cmake_install.cmake:41 (MY_EXTRA_STUFF):
Unknown CMake command "MY_EXTRA_STUFF".
FAILED: CMakeFiles/install.util
cd /tmp && /usr/bin/cmake -P cmake_install.cmake
ninja: build stopped: subcommand failed.
Is there a way to smuggle my own code into the install stage? The code is too long to fit inside install(CODE "...") nicely. A bonus to do it without an external file. Thanks!
The code passed to install(CODE) is executed as standalone CMake code, thus it shouldn't use definitions (functions,macros, variables) from the rest of CMakeLists.txt.
That is, install(CODE) behaves similar as install(SCRIPT) with a standalone script containing given code.
The thing is that configuration stage (when you call cmake to configure your project) and installation stage, which, as you can see, calls /usr/bin/cmake -P cmake_install.cmake, are separate cmake invocations. These invocations parse different files, so they unaware about context of each other.
I'm trying to write generic way to run executable resulted after build using CMake's way.
git clone git#github.com:gargamel/ihatesmurfs.git
cmake -E make_directory build
cmake -Sihatesmurfs -Bbuild
cmake --build build
cmake -E chdir build
Now I want to start executable but on *nix, it's like:
./output
and on Windows:
output.exe
Is there a way to escape this with any possible CMake command?
Expanding on my comment a bit, you can modify the CMakeLists.txt file of the project to include add_custom_command. If your CMake creates an executable named HateSmurfs, you can add the custom command to run the executable after compilation completes:
add_executable(HateSmurfs smurfs.cpp)
# Add this piece of code to run the executable after it is built.
add_custom_command(
TARGET HateSmurfs
POST_BUILD
COMMAND HateSmurfs
)
According to add_custom_command documentation:
COMMAND
If COMMAND specifies an executable target name (created by the add_executable() command) it will automatically be replaced by the location of the executable created at build time.
I'd like to run a command on a target after installing it. I see "cmake run script for install target?", which appears to be about running a single script after installing everything. My question is a per-target script.
What I want to do is to run patchelf on an installed binary to change the interpreter. This is much like how cmake will change the RPATH on the installed binary. Looking into how this is done, I see stuff this in cmake_install.cmake (edited for brevity):
file(INSTALL DESTINATION "/bin" TYPE EXECUTABLE FILES "program")
file(RPATH_CHANGE
FILE "$ENV{DESTDIR}/bin/program"
OLD_RPATH "build-dir"
NEW_RPATH "")
if(CMAKE_INSTALL_DO_STRIP)
execute_process(COMMAND "/usr/bin/strip" "$ENV{DESTDIR}/bin/program")
endif()
This is changing the rpath and also stripping the binary after it is installed. It seems like all I need to do is get one more execute_process put in there to run my patchelf command. Yet I can't find any way to get cmake to do that.
It appears add_custom_command(POST_BUILD ...) is very close, but I do not want to modify the binary in the build directory, only after install, like how cmake modifies the rpath.