Context:
I built an a.out on my local system with a modern CMake project adding a dependency to a shared library:
find_package(GDAL QUIET)
if(GDAL_FOUND)
get_property(_loc TARGET GDAL::GDAL PROPERTY LOCATION)
message(STATUS "Found GDAL: ${_loc} (version ${GDAL_VERSION})")
add_library(gdal_external INTERFACE) # dummy
Everything works fine on my system. Now I want to distribute the a.out to systems that don't have the dependency (the .so).
Problem:
I copied the binary a.out on a cluster, and when I execute it, I get the following error: error while loading shared libraries: libgdal.so.26: cannot open shared object file: No such file or directory.
Possible solution and other considerations
Installing the dependencies on the cluster, but I don't have admin rights on this server
even if I had the dependency installed by the staff, the same problem will arise later when I will ship the binary to possible users
I can't ask future users to manage dependencies and build from sources.
It seems to me that I should embed all the dependencies in the same executable, that is using a static library rather than share
Executable size or memory issues are not the priority
How do I indicate CMake to use the static version of a library ?
Related
Our Android application consists of 40-some Android Library Modules (ALMs), each of which also builds a C++ shared library with externalNativeBuild and CMake. So far we had the dependencies between these libs set up like this:
The dependent ALM references the dependency ALM with api project(':lib')
The dependent CMake script references the dependency .so with add_library(SHARED IMPORTED lib) and set_target_properties(lib PROPERTIES IMPORTED_LOCATION ...) and a relative path.
Recently we had to upgrade to the latest Android API version. This started off a cascade because now we were getting deprecated warnings in code generated by the navigation-ktx library, but upgrading that requires upgrading Gradle and the Android Gradle plugin. After that I started getting errors like liblib.so, needed by 'project', missing and no known rule to make it.
It looks like the latest Gradle parallelizes build tasks more heavily, and this means the dependent CMake/Ninja builds are being started concurrently with their dependencies, resulting in this error because the dependency is not yet built. I figured out that what we were doing was not entirely supported, but there is a "supported" way to do that now, so I refactored our entire build to use Prefab.
Now I started getting other errors, alternating between:
1.
C++ build system [prefab] failed while executing ...
Usage: prefab [OPTIONS] PACKAGE_PATH...
Error: Invalid value for "PACKAGE_PATH": Directory ... is not readable.
ld: error: undefined symbol ...
I looked into build/intermediates and found that in the 2nd case, the cmake config script was generated incorrectly: instead of add_library(lib::lib SHARED IMPORTED ) it had add_library(lib::lib INTERFACE IMPORTED) like it it was a header only library, and there was no IMPORTED_LOCATION set in the file.
What am I doing wrong and what should I do to unbreak our build?
It looks like the toolchain support for prefab interdependencies within a project is not quite finished. Others are reporting the same kind of errors at https://issuetracker.google.com/issues/265544858:
This appears to be a race condition with generating prefab cmake files.
It says in https://issuetracker.google.com/issues/221231432 that the header-only cmake config is generated to satisfy Android Studio's IDE features (completion, etc.) before the library is actually built.
Treat as-yet-unconfigured modules as if they are Header-only libraries for Android Studio purposes. This works because Android Studio doesn't care about linker flags for the purposes of providing language services.
This question asks how to install a shared library with cmake which has been imported rather than being built by the current project:
Can I install shared imported library?
To repeat the issue:
add_library(libfoobar SHARED IMPORTED)
# this install command is illegal
install(TARGET libfoobar LIBRARY DESTINATION "${RPMBUILDROOT}${LIBDIR}")
This was raised as a [https://gitlab.kitware.com/cmake/cmake/issues/14311|issue] with cmake that has been closed, effectively with a resolution of will not fix. The grounds are, quite reasonably, that cmake does not know enough about an imported target to reliably install it.
One point the answer to that question misses is that install(TARGET) will automagically create links from libfoo.so to libfoo.so.major and libfoo.so.minor version on GNU/Linux and other unix-like platforms where this is required.
Is there a way to hijack cmake into treating a custom target as if it was built by the project or otherwise persuade it to create those links?
Something like:
add_library(libfoobar SHARED IMPORTED)
#? add_custom_target(X libfoobar)
install(TARGET X LIBRARY DESTINATION "${RPMBUILDROOT}${LIBDIR}")
What is a canonical way to do this?
When a library is built by CMake, it is CMake who assigns soversion numbers to it (according to project's settings).
When a library isn't built by CMake, CMake doesn't know soversion, so it cannot create symlinks for you.
If you bother about that CMake actually installs symlink instead of file, resolve symlinks before installing, like in that question.
Well, you may ask CMake to guess soversion of the library (e.g. by resolving symlinks and checking their names). But why you ever need symlinks?
The main purpose of soversion symlink is to resolve compatibility issues with the future library's updates. But updates are possible only when the library is installed by the project who creates it.
If your project installs library produced by other project, it is unlikely that you want to support updates for the local library's installation. So there is no needs for you to support soversions.
I am working on a larger C++ library that is using CMake and depends on Qt.
We moved from Qt4 to Qt5 and now I encounter a problem when using our lib
in an upstream project. As a minimal working example demonstrating the problem please have a look at this repo:
https://github.com/philthiel/cmake_qt5_upstream
It contains two separate CMake projects:
MyLIB: a tiny library that uses QString from Qt5::Core.
It generates and installs package configuration files
MyLIBConfig.cmake, MyLIBConfigVersion.cmake, and MyLIBTargets.cmake
in order to be searchable by CMake find_package()
MyAPP: a tiny executable depending on MyLIB
The project uses find_package(MyLIB) and creates an executable that uses MyLIB
The problem is that CMake gives me the following error message when configuring the MyAPP project:
CMake Error at CMakeLists.txt:11 (add_executable):
Target "MyAPP" links to target "Qt5::Core" but the target was not found.
Perhaps a find_package() call is missing for an IMPORTED target, or an
ALIAS target is missing?
The reason for this behaviour is that in the automatically generated MyLIBTargets.cmake file the INTERFACE_LINK_LIBRARIES entry for Qt5 Core is the Qt5::Core symbol. Using Qt4, the absolute path to the Qt core lib was specified here.
Now, I simply can resolve this by using
find_package(Qt5Core 5.X REQUIRED)
in the MyAPP project.
However, I would like to know if this is the intended/generic way to go, i.e. requesting upstream projects of our lib to search for the required transitive Qt5 dependencies themselves, or if I probably misuse CMake here and need to change my configuration procedure?
The CMake docu on package file generation
https://cmake.org/cmake/help/v3.0/manual/cmake-packages.7.html
mentions that macros can be provided by the package configuration files to upstream. Maybe this would be the correct place to search for imported targets like Qt5 and break upstream configuration runs when these dependencies are not found?
Best,
Philipp
[edit of the edit] Full Source Example
You need to deliver a CMake config file for your project, and probably the ConfigFile should be generated via CMake itself (because you cannot know for shure where the user will install your software).
Tip, use the ECM cmake modules to ease the creation of that:
find_package(ECM REQUIRED NO_MODULE)
include(CMakePackageConfigHelpers)
ecm_setup_version(${PROJECT_VERSION}
VARIABLE_PREFIX ATCORE
VERSION_HEADER "${CMAKE_CURRENT_BINARY_DIR}/atcore_version.h"
PACKAGE_VERSION_FILE "${CMAKE_CURRENT_BINARY_DIR}/KF5AtCoreConfigVersion.cmake"
SOVERSION 1
)
configure_package_config_file("${CMAKE_CURRENT_SOURCE_DIR}/KF5AtCoreConfig.cmake.in"
"${CMAKE_CURRENT_BINARY_DIR}/KF5AtCoreConfig.cmake"
INSTALL_DESTINATION ${CMAKECONFIG_INSTALL_DIR}
)
and the KF5AtCoreConfig.cmake.in:
#PACKAGE_INIT#
find_dependency(Qt5Widgets "#REQUIRED_QT_VERSION#")
find_dependency(Qt5SerialPort "#REQUIRED_QT_VERSION#")
find_dependency(KF5Solid "#KF5_DEP_VERSION#")
include("${CMAKE_CURRENT_LIST_DIR}/KF5AtCoreTargets.cmake")
This will generate the correct FindYourSortware.cmake with all your dependencies.
[edit] Better explanation on what's going on.
If you are providing a library that will use Qt, and that would also need to find the Qt5 library before compilling the user's source, you need to provide yourself a FindYourLibrary.cmake code, that would call
find_package(Qt5 REQUIRED COMPONENTS Core Gui Widgets Whatever)
Now, if it's your executable that needs to be linked, use the Components instead of the way you are doing it now.
find_package(Qt5 REQUIRED COMPONENTS Core)
then you link your library with
target_link_libraries(YourTarget Qt5::Core)
I am trying to build an OS X library using CMake. This libary also includes some amount of resource files (.pdfs used as icon images). In my initial attempts, I have been building the library and the resource files separately as libGeneric.a and generic-Resources.bundle - with, the bundle hosting all the relevant images that libGeneric.a uses.
set (SRC srcfile1.m srcfile2.m)
set (HEADERS srcfile1.h srcfile2.h)
set (ICONS icon1.pdf icon2.pdf)
set (SOURCE_FILE_PROPERTIES ${ICONS} PROPERTIES MACOSX_PACKAGE_LOCATION "Resources")
add_library(generic ${SRC} ${HEADERS})
add_library(generic-Resources MODULE ${ICONS})
set_target_properties(generic-Resources PROPERTIES LINKER_LANGUAGE C) # to suppress CMake error of not able to determine linker language for libary
set_target_properties(generic-Resources PROPERTIES BUNDLE TRUE)
This worked fine, except, I was not able to figure out how to directly include the .bundle in the build process of applications that was using libGeneric.a. The only way I could get CMake to load the bundle was to add it as a source file in the target application. But, since, the bundle had not yet been compiled while running CMake, it would complain that the source file did not exist. As a workaround, I resorted to manually adding the .bundle into xcode after CMake generated the App.xcodeproj file (and I actually compiled the bundle). As this was getting to be cumbersome, I figured I'd try to build a Mac OS X framework instead (to house both the library and the code)
set (SRC srcfile1.m srcfile2.m)
set (HEADERS srcfile1.h srcfile2.h)
set (ICONS icon1.pdf icon2.pdf)
set (SOURCE_FILE_PROPERTIES ${ICONS} PROPERTIES MACOSX_PACKAGE_LOCATION "Resources")
add_library(generic SHARED ${SRC} ${HEADERS} ${ICONS})
set_target_properties(generic PROPERTIES FRAMEWORK TRUE)
set_target_properties(generic PROPERTIES MACOSX_RPATH TRUE) #to suppress cmake warnings on rpath
However, this is creating a framework with just the Icons inside a Resources directory. The code library is missing. I would appreciate some assistance in getting this framework to build correctly. How, do I
get CMake to actually build a library with the indicated source file
and place it inside the framework
get CMake to copy the headers appropriately in the framework.
I am just setting rpath to true now, as I haven't really figured out what to actually do with it. What do I set this to? The objective is to build a private framework, that I would then bundle automatically with my application.
Alternatively, is there an easy way to get CMake to build the bundle file created in my earlier process, and then load that built file into my applications build process.
I am trying to run cmake to generate makefiles. In the minimum working example, I have three files and 1 build directory.
File 1 is CMakeLists.txt, containing exactly:
add_library (MathFunctions SHARED mysqrt.cxx)
install (TARGETS MathFunctions LIBRARY DESTINATION lib)
File 2 is MathFunctions.h containing the function prototype, function relates to mysqrt.cxx.
File 3 is mysqrt.cxx containing include statement and a function definition.
When I create a build sub-directory and run "cmake ..", I am getting
CMake Error at CMakeLists.txt:2 (install):
install Library TARGETS given no DESTINATION!
Isn't my add_library, then install statement grammar correct? If I remove both SHARED and LIBRARY, cmake builds without errors.
Thanks for your help.
The problem is likely down to you running this on what CMake calls a "DLL platform" and how CMake classifies a shared library on such a platform.
From the docs for install:
For DLL platforms the DLL part of a shared library is treated as a RUNTIME target and the corresponding import library is treated as an ARCHIVE target. All Windows-based systems including Cygwin are DLL platforms.
So, try changing your command to something like:
install (TARGETS MathFunctions
ARCHIVE DESTINATION lib
LIBRARY DESTINATION lib
RUNTIME DESTINATION bin)