Build Boost on MAC with Xcode

How do you install Boost on MacOS?

Download MacPorts, and run the following command:

sudo port install boost 

Include boost in xcode

I'm using boost with Xcode 3.2.1 on mac os 10.6.8

I had problems with using boost and Xcode today, finally I managed to make it work so I hope this helps:

  1. download boost, I got the 1.52.0 today, untar it

  2. install boost:

    at first I couldn't do the ./bootstrap.sh, it stopped after the first few lines. After some googling, I got the answer, it was because when I installed Xcode a long time ago, for some reason I didn't check the “Unix development" in the install options. At this point, reinstall a newer version of Xcode and check the "unix development line"

    after that, the ./bootstrap and ./b2 install worked fine

  3. the lambda example:

    at this point, the first example should work fine, if it doesn't try adding /usr/local/include in the project settings/header search path

  4. using thread or something else that requires to be built

    threads require a specific dylib to work: libboost_system.dylib. on my system it was located in /usr/local/lib after the ./b2 install thingy

    in the left part of the GUI, you can right click anywhere and select Add Existing File

    if the file is hidden you can find a shortcut to get to usr/local/lib in the directory /Developer/SDKs/MacOSX10.6.sdk/usr/local

once the libboost_system.dylib does appear in the Groups and Files part of the GUI, it should work

How do I compile boost for OS X 64b platforms with stdlibc++?

Downloaded Boost 1.55, bootstrapped using:

./bootstrap.sh --prefix=/usr/local/boost155 cxxflags="-arch i386 -arch x86_64" \
address-model=32_64 threading=multi macos-version=10.8 stage

Built using:

./b2 threading=multi link=static runtime-link=static \
cxxflags="-stdlib=libstdc++" linkflags="-stdlib=libstdc++"

Yields in libboost_chrono.a:

     U std::string::_Rep::_M_destroy(std::allocator<char> const&)
U std::string::_Rep::_S_empty_rep_storage
U std::string::append(char const*, unsigned long)
U std::string::append(std::string const&)
U std::string::assign(char const*, unsigned long)
U std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(char const*, std::allocator<char> const&)
U std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(std::string const&)

Which implies that the library was built with the option -stdlib=libstdc++ - i.e. it's linked against the gnu version of the C++ runtime.

We purge the build using:

find . -name \*.o -print0 | xargs -0 rm
find . -name \*.a -print0 | xargs -0 rm

If we don't do that then it doesn't rebuild, and you end up with the same code as before. Next we build using:

./b2 threading=multi link=static runtime-link=static \
cxxflags="-stdlib=libc++" linkflags="-stdlib=libc++"

Yields in libboost_chrono.a:

     U std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >::append(char const*)
U std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >::append(char const*, unsigned long)
U std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >::assign(char const*)
U std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >::basic_string(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&)
U std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >::~basic_string()

Which implies that it's built against libc++.

This can be verified by using a simple test c++ program (to indicate the linking):

#include <string>

int
main(int argc, char **argv)
{
std::string s("Hello World");
return 0;
}

$ make test
c++ test.cpp -o test
$ nm ./test | c++filt
U std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >::__init(char const*, unsigned long)
U std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >::~basic_string()

$ rm test
$ make test CXXFLAGS=-stdlib=libstdc++
c++ -stdlib=libstdc++ test.cpp -o test
$ nm ./test | c++filt
U std::basic_string<char, std::char_traits<char>, std::allocator<char> >::basic_string(char const*, std::allocator<char> const&)
U std::basic_string<char, std::char_traits<char>, std::allocator<char> >::~basic_string()

so yes, it's compiling with the relevant flag. What it does indicate that you have to pass the -stdlib=libstdc++ to everything you're compiling if you're using XCode 5 as it now defaults to using -stdlib=libc++. This means that any C++ based libraries that depend on c++ stdlib that you depend on also have to be compiled with the same flag.

Be careful with an incremental build of boost - if you don't purge the .o and .a files, they don't get recompiled based on the changed flags, which keeps the files as compiled, so if they were miscompiled then you encounter the problem.

How do I build simple boost program on Mac OS (Lion)

You need to link with Boost.System, which should be in /opt/local/lib/libboost_system (with some suffix, that depends on how you built boost)

Add that to your Xcode project.

linking Boost library to Xcode project

Adding the path is correct but you also need to specify the libraries you need. On the commandline you would use -l for that in Xcode you can add them to Other Linker Flags.

The libraries you need are boost_filesystem and boost_system.



Related Topics



Leave a reply



Submit