When using CMake on unix I dont have any issues. I can use CLion to do a cmake setup, cmake build and cmake install, open a different project and it will find the previously built library when using find_package. On windows this does not seem to be possible. By default it tries to install the build code into strange directories (like C:\Program Files). I have added a CMAKE_INSTALL_PREFIX to both my library CMakeLists.txt and the appliation CMakeLists.txt, however when using find_package(SDL2)CMake still complains there is no config file for CMake and SDL2. When checking the following file exists:
U:\various\cmake-cache\Program Files (x86)\SDL2\cmake\SDL2Config.cmake
The directory U:\various\cmake-cache was used as CMAKE_INSTALL_PREFIX for both SDL2 and my application. Yet it still refuses to compile.
What can I do to make CMake at least somewhat useful on windows? On Unix things work great, but it feels like a huge PITA on Windows so far... It seems like all the concepts dont work there. I would really like to have one central location that is used by every CMake build on my system and everything is installed there and when another project uses a library it is searched there. Is this possible?
JetBrains has spoiled me. I'm familiar with the standard UNIX make file and the make, make install routine normally associated with installing software with traditional make files, but I'm not as familiar with cmake since CLion does it all for me.
I want to distribute my code to others and provide simple instructions for building it via cmake so that they have a binary they can execute. The official cmake tutorial shows writing install rules in the CMakeLists.txt file but it isn't clear if this is supported by CLion (or even if it needs to be).
For a simple, single-file (main.cpp) application, what would be an example of how to build it using cmake (assuming those it is distributed to don't have CLion nor use another IDE, they just want to build and use it)?
To build code that comes with a CMakeLists.txt file, you run cmake to generate a Makefile (or other build configuration file):
cmake <path_to_CMakeLists.txt>
Then you run
make;make install
as usual. (or as described in the comment, you can type cmake --build . instead of make - useful if you're on a platform with a different build system)
You don't want to check in the Makefile into your source control, though, as it needs to be generated on the computer that will actually be doing the building.
I can't build MinGW binaries in OpenCV. I am using Code::Blocks and need to include OpenCV libraries in my C++ project.
I searched the web for hours for a solution, but i can't find any. I tried CMake, but get the following error:
CMake Error: CMake was unable to find a build program corresponding
to "MinGW Makefiles". CMAKE_MAKE_PROGRAM is not set. You probably
need to select a different build tool.
I searched the web for this error, but all the proposed solutions are too advanced for me.
Is there a way to build MinGW without CMake? My reason for not want to use CMake is that it seems like i will get too much trouble with it. If it's not possible (or more difficult) to not use CMake, is anyone knows how to get around with aforementioned error?
for around 5 consecutive days i have been trying to set up my computer with the c++ environment for programming with libraries such as sdl,glm,opengl. its important for us to be able to run it on unix machines on presentations so im running with cmake.
i finally got it to work with the cmake-gui, i wont even bother trying anymore with any IDE.
i specified my folder project and where to build the binaries, i got a folder "CMakeFiles" along with a txt "CMakeCache", a CMAKE file "cmake_install.cmake" and a file "Makefile". also in my folder "CMakeFiles" there are lots of other folders such as "CMakeTmp", "CompilerIdC", "CompilerIdCXX etc" and in both folders "Compiler*" has each an .exe which doesnt work! so where is my wanted executable?
i opened cmd and navigated to my folder and tried to write "make" as we are supposed to do according to the intruction. alas, it didnt work very well. hoping you could share your wisdom and help a newbie like me!
so what exactly is needed for compiling projects containing additional libraries? so far i have a compiler, Mingw32, the latest CMake and using the cmake-gui for extracting the binaries but gets makefiles.
EDIT:
hrrm. is it only me who gets these kind of problems? i can add that i have look thorough about 10 tutorials and 90% of the steps are similar (if compiling with VS which i tried at first):
Download latest SDL
Make a folder on e.g C:\SDL with two folders, include and lib
Copy the libs and includes from the downloaded SDL
Make new VS project, open VC++ directories and add lib/incl folder on e.g C:\SDL
Add to linker SDL.lib and SDLmain.lib (i made sure they got linked, no problem here)
Change system to WINDOWS (optional if you dont want two windows)
Added include to "additional libraries"
Put the SDL.dll file (which i got from the latest SDL) in my C:\windows\system32(64SysWoW)
and also in my project file.
so what i am actually looking for is gettning the CMake to work, since it generates and builds sources successfully (with the gui) and i feel im closing in. do i need to add any additional libraries from sdl to my compiler mingw32 and/or cmake?
if you run cmake by command:
cmake -G "Visual Studio 14 Win64" path\to\source\dir
you need to run this command to continue(in Visual Studio Command Prompt):
msbuild Project.sln
either if you run cmake:
cmake -G "NMake Makefiles" path\to\source\dir
you need to run this cmd to continue(in Visual Studio Command Prompt):
nmake
You were almost there with Visual Studio. Select Visual Studio as target. Open the generated project in Visual Studio, build it. (just like you alread did). Then, instead of trying to run BUILD_ALL, run a real project that creates an executable, it should also be in that list. Just right click it and 'play' it.
If you still get errors, post them in detail including what you did before the error. Note: a carefully configured cross platform CMake project (aka the CMakeLists.txt) should not require any fiddling with VC++ directories. It should work automagically, especially with well known libs such as SDL.
If I understood it correctly you want to use CMake in your project. I'm using CMake in all my projects. I won't give you exact step-by-step howto, since I use Arch Linux but I used it in Windows 7 too.
To make CMake find the libraries, it is often needed to set up the CMAKE_PREFIX_PATH environment variable so it points to the directories where dependencies of your project are installed.
Set you PATH environment varible so you can invoke you compiler and make just by calling by calling eg. make. I think you need to do than manually for Mingw32, for Visual Studio you can use the "Visual Studio Command Propt" which has these variables already set.
Run CMake with desired generator. To select the generator from command line use the -G switch. You will probably use one of the following (the ... means other options you want to pass to cmake)
For GNU make used in MinGW use cmake -G "MinGW Makefiles" ...
For NMake from visual studio use cmake -G "NMake Makefiles" ...
It is also possible to create a Visual Studio project but I do not recommend it, since it quite difficult to set up automatic builds then. I also had some problems with dependencies when I tried to use VS project.
change directory to your build directory (ie. the one where you called cmake, it contains the CMakeCache file) and run make or nmake
Quoting from "CMake support in Visual Studio":
Visual Studio 2017 introduces built-in support for handling CMake projects. This makes it a lot simpler to develop C++ projects built with CMake without the need to generate VS projects and solutions from the command line. This post gives you an overview of the CMake support, how to easily get started and stay productive in Visual Studio.
Statistics...
My System: Windows Vista 64 bit
Library: Bullet Physics v2.78
Makefile Generator: CMake
Build system: MinGW command line, MinGW + CodeBlocks
Makefiles were implemented by command line and through CMake-generated Code::Blocks project
In both cases, the build fails near 30%. NOTE: I did have to switch the build executable in the Code::Blocks project from make.exe to mingw32-make.exe
First Failure: Demos\OpenGL\GLDebugFont.cpp -> GL/glut.h: No such file or directory
Second Failure (after commenting out the #include from the first) -> Demos/OpenGL/GlutStuff.h: same error
These failures happen when I build command-line OR through the generated C::B project.
Why can't it find GL/glut.h? "bullet-2.78\Glut\GL\glut.h" exists. Maybe there's a way I can tell it to find glut there?
NOTE: During the CMake makefile generation, CMake did tell me that "You are using the obsolete GLU package, please use OpenGL instead." It continued as normal with the Makefile generation. Maybe I need to define some environment variables? Maybe I need to configure something in Windows that CMake is looking for?
I would add a compiler search path in bullet-2.78/Glut, but Code::Blocks doesn't allow that if you're using a custom makefile (like a CMake-generated one).
UPDATE: I have been able to build the library itself, by building specific targets in the Code::Blocks project. However, I have been unable to build any of the demos or the benchmark test, since they all use OpenGL (and apparently glut.h). I would still like to compile those.