How do you statically link matio, a library for reading mat-files, mainly used by matlab and octave into an executable?
If the file main.cpp holds matio functionality the compiler call
g++ -o main main.cpp -Imatio/include matio/lib/libmatio.a
fails with a bunch of error messages like: undefined reference to `inflateEnd'. This can be resolved by also adding zlib to the compiler call:
g++ -o main main.cpp -Imatio/include matio/lib/libmatio.a -lz
Now the error messages differ with something like undefined reference to `__intel_sse2_strlen'. So it appears that the zlib library is necessary for the comilation.
I now have the following questions:
What do you need to do to statically link the matio library in an executable?
Why do I need to add the zlib library even though I configured and compiled matio with ./configure --without-libz?
To build matio without zlib you apparently need to invoke configure with
./configure --with-zlib=no
(Checked this from configure.ac and config/matio_zlib.m4.)
In case you want to build matio with icc, Intel's developer pages tell that __intel_sse2_strlen is defined in libirc.a on Linux and libirc.lib on Windows.
To compile matio with the gcc do
./configure --with-zlib=no CC=gcc
afterwards, the matio library is statically linkable with the call posted in the question
Related
I made a little c++ program with this library: http://libtins.github.io
I compile with: g++ -o arp arp.cpp -ltins
It runs fine (Ubuntu 14.04 64 bit), however if i send the executable to a friend, he gets this error when running:
error while loading shared libraries: libtins.so.3.4: cannot open shared object file: No such file or directory
I've looked up on stackoverflow and found something about statically linking and dynamically linking, also tried some g++ arguments but no success.
Is it even possible, to create a binary that "contains" the shared object itself?
Thanks
You can use -static to link the libraries statically.
g++ -o arp arp.cc -ltins -static
To build libtins statically, you could refer to the following section in its README:
Static/shared build
Note that by default, only the shared object is compiled. If you would
like to generate a static library file, run:
cmake ../ -DLIBTINS_BUILD_SHARED=0
I'm trying to static link OpenSSL into my program.
It works fine when linking into the executable. I need to use OpenSSL in a shared library (so, or dll) that I dynamically load later on when the process executes.
Trying to statically link OpenSSL into the shared library causes errors due to OpenSSL not being compiled with -fPIC. Is it possible to do this without recompiling openssl?
Also, is there a better way to do this?
I'm trying to static link OpenSSL into my program.
In this case, its as simple as:
gcc prog.c /usr/local/lib/libssl.a /usr/local/lib/libcrypto.a -o prog.exe -ldl
It works fine when linking into the executable.
Devil's advocate... Does it work fine with Position Independent Code (PIE)? PIE on a program is the equivalent to PIC on a shared object (some hand waiving).
gcc -fPIE prog.c /usr/local/lib/libssl.a /usr/local/lib/libcrypto.a -o prog.exe -ldl
According to the GCC folks, you can compile with fPIC, and then build a shared object with -fPIC or a relocatable executable with -fPIE. That is, its OK to use -fPIC for both.
Trying to statically link OpenSSL into the shared library causes errors due to OpenSSL not being compiled with -fPIC.
That's easy enough to fix. You simply specify shared in configure:
./config shared no-ssl2 no-ssl3 no-comp --openssldir=/usr/local/ssl
make
sudo make install
I think you can also (notice the lack of shared):
export CFLAGS="-fPIC"
./config no-ssl2 no-ssl3 no-comp --openssldir=/usr/local/ssl
make
sudo make install
not being compiled with -fPIC. Is it possible to do this without recompiling openssl?
NO, you have to compile with PIC to ensure GCC generates relocatable code.
Also, is there a better way to do this?
Usually you just configure with shared. That triggers -fPIC, which gets you relocatable code.
There's other things you can do, but they are more intrusive. For example, you can modify Configure line (like linux-x86_64), and add -fPIC in the second field. The fields are separated by colons, and the second field is $cflags used by the OpenSSL build system.
You can see an example of modifying Configure at Build OpenSSL with RPATH?
I am trying to compile a program on my system with Debian Wheezy and g++4.7. I want it to be able to run on another system with Debian Squeeze (and no recent g++). I can't compile the program on the Squeeze, because I use certain C++11 features the old g++ does not support, as well as a new Boost version and libcrypto++9.
As far as I understand the usual way to get around this problem is to static link the libraries not supported at the other system, in my case libstdc, boost and crypto++.
My (linking) compiler call right now is
g++-4.7 .obj/btcmirco.o -Wl,-Bstatic -lboost_program_options -lboost_system -lcrypto++ -Wl,-Bdynamic -lcurl -static-libgcc -std=c++11 -o MyProgram
However I seem to have missed something, because it throws a lot of undefined reference errors. It works fine if I dynamic link to crypto++ (and only static link libstdc and boost).
Can anyone tell me whats wrong, or if there is a fundamental error in my approach?
The linker errors I get are (shorted):
`.text._ZN8CryptoPP22BufferedTransformationD2Ev' referenced in section `.text._ZN8CryptoPP22BufferedTransformationD1Ev[_ZN8CryptoPP22BufferedTransformationD1Ev]' of /usr/lib/gcc/x86_64-linux-gnu/4.7/../../../../lib/libcrypto++.a(cryptlib.o): defined in discarded section `.text._ZN8CryptoPP22BufferedTransformationD2Ev[_ZN8CryptoPP22BufferedTransformationD5Ev]' of /usr/lib/gcc/x86_64-linux-gnu/4.7/../../../../lib/libcrypto++.a(cryptlib.o)
`.text._ZN8CryptoPP25MessageAuthenticationCodeD2Ev' referenced in section `.text._ZN8CryptoPP25MessageAuthenticationCodeD1Ev[_ZN8CryptoPP25MessageAuthenticationCodeD1Ev]' of /usr/lib/gcc/x86_64-linux-gnu/4.7/../../../../lib/libcrypto++.a(cryptlib.o): defined in discarded section `.text._ZN8CryptoPP25MessageAuthenticationCodeD2Ev[_ZN8CryptoPP25MessageAuthenticationCodeD5Ev]' of /usr/lib/gcc/x86_64-linux-gnu/4.7/../../../../lib/libcrypto++.a(cryptlib.o)
I experienced the same problem and this has to do with the fact that you are trying to mix code generated by g++-4.7 (your program) with code generated by a previous version of g++ (cryptopp library).
The reason behind this is that when you execute compile the library executing make command, it uses the default version of g++ set up for your system, usually the one that comes with the OS.
In order to solve the issue what you should do is compile cryptopp library with g++-4.7.
For that, compile the library by executing make CXX=g++-4.7. The resulting static library shouldn't give you the error when being linked with your code.
I've been working on this for a while now, and can't seem to make sense of the situation - partly bceause I don't fully understand what's going on (which is why I came here).
I'm doing a sort of boost hello world as follows:
#include <boost/thread/thread.hpp>
#include <cstdio>
void helloworld() {
std::printf("HELLO FROM A BOOST THREAD!");
}
int main(int argc, char **argv) {
boost::thread t(&helloworld);
t.join();
}
This is on Windows. I stored the Boost directory in C:\Boost. I ran bootstrap and bjam, and now have a stage/lib folder that contains all the .lib files. The lib files relating to the boost/thread library are:
libboost_thread-vc100-mt.lib
libboost_thread-vc100-mt-1_46_1.lib
libboost_thread-vc100-mt-gd.lib
libboost_thread-vc100-mt-gd-1_46_1.lib
Now I compile:
g++ -c main.cpp -I/Boost
that line works fine, I get main.o. Then:
g++ -o test.exe main.o -L/Boost/stage/lib -llibboost_thread-vc100-mt
And that's where the trouble happens. First of all, If I didn't type the -l argument the way I did, MinGW couldn't even find the file. Meaning, if I tried:
-lboost_thread-vc100-mt
instead of the way I typed it above (and how I thought it should be done), ld would exit with no such file. Anyway, this is now the output I'm getting from that line:
main.o:main.cpp:(.text+0x47): undefined reference to `_imp___ZN5boost6thread4joinEv'
main.o:main.cpp:(.text+0x55): undefined reference to `_imp___ZN5boost6threadD1Ev'
main.o:main.cpp:(.text+0x70): undefined reference to `_imp___ZN5boost6threadD1Ev'
main.o:main.cpp:(.text$_ZN5boost6threadC1IPFvvEEET_NS_10disable_ifINS_14is_convertibleIRS4_NS_6detail13thread_move_tIS4_EEEEPNS0_5dummyEE4typeE[boost::thread::thread<void (*)()>(void (*)(), boost::disable_if<boost::is_convertible<void (*&)(), boost::detail::thread_move_t<void (*)()> >, boost::thread::dummy*>::type)]+0x23): undefined reference to `_imp___ZN5boost6thread12start_threadEv'
collect2: ld returned 1 exit status
Now somewhere in there, I can tell that these are apparently the functions I'm supposed to be getting from boost/thread, and apparently it does find the lib file, so why isn't it linking correctly?
Thank you very much for any help!
EDIT:
I've rebuilt boost using the bjam "stage" option
bjam toolset=gcc stage
Now, after the build completes, I'm left with a stage/lib folder with .a files, as is to be expected. These are the boost/thread related libraries:
libboost_thread-mgw45-mt-1_46_1.a
libboost_thread-mgw45-mt-d-1_46_1.a
However, linking as follows:
g++ -o test.exe main.o -L/Boost/stage/lib -lboost_thread-mgw45-mt-1_46_1
outputs the exact same errors. Also tried:
g++ -o test.exe main.o -L/Boost/stage/lib -lboost_thread-mgw45-mt-1_46_1 -static
I'm at a loss, still.
Solved the problem. Boost's headers are configured to be dynamically linked, but the dynamic libraries (dll's) are not built unless you specify:
--build-type=complete
when invoking bjam. After that, copy the appropriate dll to your application directory, but still use the
-L/BOOST_DIR/stage/lib -lname
when linking.
This set of library files:
libboost_thread-vc100-mt.lib
libboost_thread-vc100-mt-1_46_1.lib
libboost_thread-vc100-mt-gd.lib
libboost_thread-vc100-mt-gd-1_46_1.lib
are for the Visual Studio 2010 compiler. They won't work with GCC. If you want to use gcc/MinGW, you'll need to download/build a set of boost libraries for that compiler. Alternatively you can install VS 2010 and use that compiler (the free VC++ 2010 Express version should work fine if cost is an issue).
You can get a MinGW distribution with Boost already in the package from http://nuwen.net/mingw.html (32-bit target only, I believe).
To answer about getting the errors with using the MinGW libs:
The _imp_ prefixes on the symbols is an indication that g++ is looking to link to a a dll/shared library. The .lib file you have are for static libraries (which is what also what I get when doing a straightforward bjam build of the libraries). If you look in boost/thread/detail/config.hpp you'll see that for Win32 builds it defaults to building against a DLL library unless the MSVC or Intel compiler is being used.
I'm not even sure exactly how to build the DLL libraries - I'll have to look it up. In the meantime, you can use the following command to build your example such that it'll link against the static library. The BOOST_THREAD_USE_LIB macro build the .cpp file such that it'll expect to link against the static library:
g++ -I/Boost -DBOOST_THREAD_USE_LIB -c main.cpp
I would like to utilize the UnitTest++ library in a testing file. However, I am having some difficulty getting the library to be included at compile time. So here is my current directory structure:
tests/
UnitTests++/
libUnitTest++.a
src/
UnitTests++.h
unit/
test.cpp
I have just used the UnitTest++ getting started guide to just get the library setup. Here is test.cpp:
// test.cpp
#include <UnitTest++.h>
TEST(FailSpectacularly)
{
CHECK(false);
}
int main()
{
return UnitTest::RunAllTests();
}
And I am currently trying to compile with:
gcc -lUnitTest++ -L../UnitTest++/ -I../UnitTest++/src/ test.cpp
I am currently getting a bunch output with ld: symbol(s) not found at the end. So how would I be able to get the UnitTest++ library properly included when this program is compiled? I am on a Mac and I'd also like for there to be an easy way for people on a Linux machine to run these same tests.
Whew, I hope this provides enough information, if not please let me know.
I was able to build it in the following manner
gcc -L../UnitTest++/ -I../UnitTest++/src/ test.cpp -lUnitTest++ -lstdc++
or
g++ -L../UnitTest++/ -I../UnitTest++/src/ test.cpp -lUnitTest++
that links to libstdc++ automatically.
GCC documentation says:
-llibrary
-l library
Search the library named library when linking. (The second alternative with the library as a separate argument is only for POSIX compliance and is not recommended.)
It makes a difference where in the command you write this option; the linker searches and processes libraries and object files in the order they are specified.
Thus, foo.o -lz bar.o' searches libraryz' after file foo.o but before bar.o. If bar.o refers to functions in `z', those functions may not be loaded.
I guess that's why the library symbols are not found when you first specify -lUnitTest++ and then test.cpp
Compile test.cpp to get test.o
and use
g++ test.o libUnitTest++.a -o ./exectest
to get the ./exectest executable
libUnitTest++.a is just an archive of all the object files of UnitTest++. You just need to link all the object files (your test object file + libUnitTest++.a)
Try editing the makefile that came with unittest++ and make it suitable for your case
The message ld: symbol(s) not found means you haven't compiled the library. So you need to go in the UnitTest++ folder, compile and install it.
I've never worked on a MAC, but in linux, libraries are usually compiled and installed with:
./configure
make
make install
In the UnitTest++ link you posted, you should simply:
make install
Then you will have the UnitTest++.so library in the libraries folder of your OS. Now the library can be linked with your program with the -lUnitTest++ command.
Usually you have to put -Lsomething before the -lsomething that requires it.