Cross compiling with CMake: linker errors - c++

I am attempting to cross compile a project for my raspberry pi (32 bit armv8) on my ubuntu machine. I set up a toolchain using crosstool-NG and compiling works, but linking fails. My CMake toolchain file is as follows:
SET(CMAKE_SYSTEM_NAME Linux)
SET(CMAKE_SYSTEM_VERSION 1)
set(CMAKE_SYSROOT /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/sysroot)
SET(CMAKE_C_COMPILER /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/bin/armv8-rpi3-linux-gnueabihf-gcc)
SET(CMAKE_CXX_COMPILER /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/bin/armv8-rpi3-linux-gnueabihf-g++)
SET(CMAKE_FIND_ROOT_PATH /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf)
SET(CMAKE_FIND_ROOT_PATH_MODE_PROGRAM NEVER)
# for libraries and headers in the target directories
SET(CMAKE_FIND_ROOT_PATH_MODE_LIBRARY ONLY)
SET(CMAKE_FIND_ROOT_PATH_MODE_INCLUDE ONLY)
set(CMAKE_FIND_ROOT_PATH_MODE_PACKAGE ONLY)
I have a few libraries that are not being found by the linker. The problem is with my sysroot. I copied over /usr and /lib from the pi to my sysroot folder that was created by crosstool-NG. The libraries that are being linked reside in sysroot/usr/lib/arm-linux-gnueabihf.
However, the linker is not checking that directory, it checks sysroot/usr/lib and a few others, but for some reason it does not check the arm-linux-gnueabihf directory.
On the pi, I can run ld -lasound --verbose and get the following output:
attempt to open //usr/local/lib/arm-linux-gnueabihf/libasound.so failed
attempt to open //usr/local/lib/arm-linux-gnueabihf/libasound.a failed
attempt to open //lib/arm-linux-gnueabihf/libasound.so failed
attempt to open //lib/arm-linux-gnueabihf/libasound.a failed
attempt to open //usr/lib/arm-linux-gnueabihf/libasound.so succeeded
so its finding it correctly on the pi.
When I do the same using the ld generated by crosstool-NG:
attempt to open /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/bin/../sysroot/home/stone/x-tools/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/lib/libasound.so failed
attempt to open /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/bin/../sysroot/home/stone/x-tools/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/lib/libasound.a failed
attempt to open /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/bin/../sysroot/usr/local/lib/libasound.so failed
attempt to open /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/bin/../sysroot/usr/local/lib/libasound.a failed
attempt to open /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/bin/../sysroot/lib/libasound.so failed
attempt to open /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/bin/../sysroot/lib/libasound.a failed
attempt to open /home/stone/development/raspberry_pi/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/bin/../sysroot/usr/lib/libasound.so failed
It fails to find it because it doesnt look in the arm-linux-gnueabihf directory. I copied the library .so file from sysroot/usr/lib/arm-linux-gnueabihf to sysroot/usr/lib and it compiles and links successfully, but I would like to make it so that I do not have to do that. How can I make the linker check that arm-linux-gnueabihf directory?
EDIT: I also copied /etc/ld.so.conf and /etc/ld.so.conf.d from the pi into my sysroot, but it does not seem to affect the linker.
EDIT: After further research it looks like this might be due to something with gcc-multiarch. I am not sure what that is but hopefully I can figure this out
EDIT: I verified that indeed the linker is not searching the arm-linux-gnueabihf path:
./armv8-rpi3-linux-gnueabihf-ld --verbose | grep -i "search"
outputs:
SEARCH_DIR("=/home/stone/x-tools/armv8-rpi3-linux-gnueabihf/armv8-rpi3-linux-gnueabihf/lib"); SEARCH_DIR("=/usr/local/lib"); SEARCH_DIR("=/lib"); SEARCH_DIR("=/usr/lib");

Alright I got it all figured out. The repo where I learned the tricks to getting this to work can be found here. I am grateful for this repo, without it I would not have gotten this to work.
The issue lies with the linker. The linker that comes with plain old binutils that gets pulled in by crosstool-ng does not search the arm-linux-gnueabihf subdirectories. These directories exist because of Debian multiarch, and in order to get a linker that looks in such a directory, binutils needs to be patched. The instructions that follow assume you have already generated a ct-ng config based on the rpi3 sample.
On the raspberry pi, I installed the binutils-source package:
$ sudo apt install binutils-source
This will give access to a patch file located in /usr/src/binutils/patches.
On the host machine, you need to add a patches directory in the directory where you are building your toolchain (where the ct-ng config file is). It must have a specific structure that reflects the structure of packages in the crosstool-ng repo (ie. inside patches/ you must have a directory with the package name, inside that you must have the version number, and inside that is where you place the patch.):
$ cd your_toolchain_directory
$ mkdir -p patches/binutils/2.31.1/
Now we can copy the patch file from the pi to the host. The specific patch we need is 129_multiarch_libpath.patch:
$ cd patches/binutils/2.31.1/
$ scp pi#raspberrypi:/usr/src/binutils/129_multiarch_libpath.patch
Now that we have the patch, we need to update the config file to tell ct-ng to include local patches as well as enable the multiarch flag for gcc. You can do this using ct-ng menuconfig so change back to the directory with the config file and run:
$ ct-ng menuconfig
In the paths and misc options you need to change the patches origin to bundled then local, and then add ${CT_TOP_DIR}/patches as the local patch directory. This should produce the following lines under the extracting section of the .config file:
CT_PATCH_BUNDLED_LOCAL=y
CT_PATCH_ORDER="bundled,local"
CT_PATCH_USE_LOCAL=y
CT_LOCAL_PATCH_DIR="${CT_TOP_DIR}/patches"
Next you need to add the --enable-multiarch flag in the gcc options. Again using menuconfig, go to the C compiler settings. Add --enable-multiarch to the gcc extra config setting. The config file should have the following in the gcc settings:
CT_CC_GCC_EXTRA_CONFIG_ARRAY="--enable-multiarch"
Save the config. The last thing we need to do is export a variable.
$ export DEB_TARGET_MULTIARCH=arm-linux-gnueabihf
Now we can build the toolchain:
$ ct-ng build
Once its done, you can change into the bin directory of the generated toolchain and run:
$ ./ld --verbose | grep -i "search"
and you should see that the arm-linux-gnueabihf directory is now in the search paths. What a process.

You need to use static library instead. then the executable will include the binary of the libraries.
But keep in mind the executable file will be bigger ( depends obviously which lib...).
When you using cross compile, you should check that the libraries you use exist in the other environment you execute the program.
The solution will be to add STATIC word according to CmakeList.txt file

Related

libSDL2-2.0.so.0: cannot open shared object file

I'm trying to build the SDL library from the source code. I've downloaded the compressed file (i.e. SDL2-2.0.3.tar.gz) and extracted it. I don't want to install the files in /usr/local. According to this link, I need to change the configure
The last command says "sudo" so we can write it to /usr/local (by
default). You can change this to a different location with the
--prefix option to the configure script. In fact, there are a LOT of good options you can use with configure! Be sure to check out its
--help option for details.
This is what I've done.
mkdir build
cd build
../configure
make
sudo make install
In install folder that I've created are the following files
share
lib
include
bin
Now I would like to run the test files. I've picked this testatomic.c and this is the command line
gcc testatomic.c -o test -I/home/xxxx/Desktop/SDL2-2.0.3/install/include/SDL2 -L/home/xxxx/Desktop/SDL2-2.0.3/install/lib -lSDL2 -lSDL2main
I get this error
error while loading shared libraries: libSDL2-2.0.so.0: cannot open shared object file: No such file or directory
In lib, these are the files
Where is the shared object file?
You're getting error when running resulting program because system's dynamic linker cannot find required library. Program requires libSDL2-2.0.so.0, linker looks for it in system-defined directories (/lib, /usr/lib, ..., - defined in /etc/ld.so.conf), but finds none - hence an error.
To inform linker where you want it to look for libraries, you can define LD_LIBRARY_PATH environment variable, e.g. in your case:
export LD_LIBRARY_PATH="$HOME/Desktop/SDL2-2.0.3/install/lib"
./test
Other ways is installing libraries in standard location, defining LD_LIBRARY_PATH in your .bashrc (or whatever shell you use), or using rpath, e.g. adding -Wl,-rpath=$HOME/Desktop/SDL2-2.0.3/install/lib at the end of your compilation line.
I was able to fix this problem with:
sudo apt install libsdl2-dev
I too had:
./01_hello_SDL: error while loading shared libraries: libSDL2-2.0.so.0: cannot open shared object file: No such file or directory
as a result of compiling the first C++ program (using the SDL headers) as part of the Lazy Foo tutorial. I found out that libSDL2-2.0.so.0 was just using the find command in the GUI. It turned out to be in /usr/local/lib
Then in terminal I typed:
export LD_LIBRARY_PATH="/usr/local/lib"
I checked the value of LD_LIBRARY_PATH using:
echo $LD_LIBRARY_PATH
I recompiled (don't know if that was necessary) and voila, it worked.

error while loading shared libraries: libgsl.so.0: cannot open shared object file: No such file or directory

I use gsl.
After I compiled my .cpp file and run it, I faced with below error:
error while loading shared libraries: libgsl.so.0: cannot open shared object file: No such file or directory
I found same as this problem in:
https://groups.google.com/forum/#!topic/cortex_var/6vluX7pP0Sk
&
Linux error while loading shared libraries: cannot open shared object file: No such file or directory
&
http://www.gnu.org/software/gsl/manual/html_node/Shared-Libraries.html
And I have done as in the above links wrote but the error is still remained.
Can anyone help me?
To make it work do the following steps
Start Borne Shell
$LD_LIBRARY_PATH= path to your gsl lib folder inside the gsl installation folder
$export LD_LIBRARY_PATH
now run your executable
It should work fine.
First, you need to locate the file (libgsl.so.0). You can do this, for example, by using the find command:
sudo find / -name "libgsl.so.0"
Let us assume, the file is located in /usr/local/lib.
(If the file has not been found, install the corresponding package or download the source, build it and install it.)
Now, you have two options:
(1) Quick & Dirty:
LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib
export LD_LIBRARY_PATH
This adds the path of the library to an environment variable. The disadvantage of this option is, that it is only valid for the current session. It will not work for other users. It will not work once you log off and on again.
(2) Permanent:
Review your /etc/ld.so.conf. If /usr/local/lib is not listed there, add it. Now, run ldconfig to detect the shared object file and add it to some system-wide index.
I got the same error with Krita on Arch Linux. I made a symlink with
ln /usr/lib/libgsl.so /usr/lib/libgsl.so.0
and that fixed it.
In my experience, fastStructure depends on gsl 1.6 but not the latest version.
wget http://gnu.mirror.vexxhost.com/gsl/gsl-1.6.tar.gz
tar -zxvf gsl-1.6.tar.gz
cd gsl-1.16
./configure
make
sudo make install
Add these lines to your .bashrc file on your home directory.
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib
export CFLAGS="-I/usr/local/include"
export LDFLAGS="-L/usr/local/lib"
then, run source ~/.bashrc to set these environment variables.
It works fine when I change the version from the latest to the 1.6.
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/shg047/software/gsl/lib/
such as: to-mr: error while loading shared libraries: libgsl.so.19: cannot open shared object file: No such file or directory
Have you tried updating your library? The program I was trying to run simply needed a newer version of gsl (I had 1.9.5 while it needed 2.0.0 or newer).
If you are on arch you can run:
yaourt gsl
and select the appropriate one.
You can use gsl-config --libs in you makefile or in the command line when you link the gsl library. Just type gsl-config you can find the options it offers to you. Choose the options you need, you will find compile and link process much easier than before. As a result, when I type gsl-config --libs in my terminal, I get -L/usr/local/lib -lgsl -lgslcblas -lm. Although it is very simple, first you should know where you gsl is installed. You can add the directory to the PATH environment variable or use the absolute path to execute gsl-config .
I needed libgsl.so.19:
/snap/inkscape/current/bin/inkscape: error while loading shared libraries: libgsl.so.19: cannot open shared object file: No such file or directory
I solved it with:
Installing Anaconda
searched for libgsl.so.19 and found it in ~/anaconda3/lib
run LD_LIBRARY_PATH=$LD_LIBRARY_PATH:~/anaconda3/lib (best add it to ~/.basrc)

Cyassl-2.4.2 missing libcyassl.a file

I'm have built and installed Cyassl-2.4.2 on my unix machine to use with a project of mine; however, when I try to compile my c project, the compiler cannot find the libcyassl.a file. I have searched many times in the Cyassl.2.4.2 folder and am unable to find the file myself. Where is this file locate? Have I missed a step in bulding/installing Cyassl.2.4.2? Thanks!
I had successfully built and installed CyaSSL 2.4.2 here and this is what I got:
~/cyassl$ find . -iname "*.a"
./src/.libs/libcyassl.a
~/cyassl$ ls /usr/local/lib/
libcyassl.a libcyassl.la libcyassl.so libcyassl.so.3 libcyassl.so.3.0.3
The building process used was:
./autogen.sh
./configure
make
sudo make install
The installation prints out:
Libraries have been installed in:
/usr/local/lib
If you ever happen to want to link against installed libraries
in a given directory, LIBDIR, you must either use libtool, and
specify the full pathname of the library, or use the -LLIBDIR'
flag during linking and do at least one of the following:
- add LIBDIR to theLD_LIBRARY_PATH' environment variable
during execution
- add LIBDIR to the LD_RUN_PATH' environment variable
during linking
- use the-Wl,-rpath -Wl,LIBDIR' linker flag
- have your system administrator add LIBDIR to `/etc/ld.so.conf'
See any operating system documentation about shared libraries for
more information, such as the ld(1) and ld.so(8) manual pages.

usr/bin/ld: cannot find -l<nameOfTheLibrary>

I'm trying to compile my program and it returns this error :
usr/bin/ld: cannot find -l<nameOfTheLibrary>
in my makefile I use the command g++ and link to my library which is a symbolic link to my library located on an other directory.
Is there an option to add to make it work please?
To figure out what the linker is looking for, run it in verbose mode.
For example, I encountered this issue while trying to compile MySQL with ZLIB support. I was receiving an error like this during compilation:
/usr/bin/ld: cannot find -lzlib
I did some Googl'ing and kept coming across different issues of the same kind where people would say to make sure the .so file actually exists and if it doesn't, then create a symlink to the versioned file, for example, zlib.so.1.2.8. But, when I checked, zlib.so DID exist. So, I thought, surely that couldn't be the problem.
I came across another post on the Internets that suggested to run make with LD_DEBUG=all:
LD_DEBUG=all make
Although I got a TON of debugging output, it wasn't actually helpful. It added more confusion than anything else. So, I was about to give up.
Then, I had an epiphany. I thought to actually check the help text for the ld command:
ld --help
From that, I figured out how to run ld in verbose mode (imagine that):
ld -lzlib --verbose
This is the output I got:
==================================================
attempt to open /usr/x86_64-linux-gnu/lib64/libzlib.so failed
attempt to open /usr/x86_64-linux-gnu/lib64/libzlib.a failed
attempt to open /usr/local/lib64/libzlib.so failed
attempt to open /usr/local/lib64/libzlib.a failed
attempt to open /lib64/libzlib.so failed
attempt to open /lib64/libzlib.a failed
attempt to open /usr/lib64/libzlib.so failed
attempt to open /usr/lib64/libzlib.a failed
attempt to open /usr/x86_64-linux-gnu/lib/libzlib.so failed
attempt to open /usr/x86_64-linux-gnu/lib/libzlib.a failed
attempt to open /usr/local/lib/libzlib.so failed
attempt to open /usr/local/lib/libzlib.a failed
attempt to open /lib/libzlib.so failed
attempt to open /lib/libzlib.a failed
attempt to open /usr/lib/libzlib.so failed
attempt to open /usr/lib/libzlib.a failed
/usr/bin/ld.bfd.real: cannot find -lzlib
Ding, ding, ding...
So, to finally fix it so I could compile MySQL with my own version of ZLIB (rather than the bundled version):
sudo ln -s /usr/lib/libz.so.1.2.8 /usr/lib/libzlib.so
Voila!
If your library name is say libxyz.so and it is located on path say:
/home/user/myDir
then to link it to your program:
g++ -L/home/user/myDir -lxyz myprog.cpp -o myprog
There does not seem to be any answer which addresses the very common beginner problem of failing to install the required library in the first place.
On Debianish platforms, if libfoo is missing, you can frequently install it with something like
apt-get install libfoo-dev
The -dev version of the package is required for development work, even trivial development work such as compiling source code to link to the library.
The package name will sometimes require some decorations (libfoo0-dev? foo-dev without the lib prefix? etc), or you can simply use your distro's package search to find out precisely which packages provide a particular file.
(If there is more than one, you will need to find out what their differences are. Picking the coolest or the most popular is a common shortcut, but not an acceptable procedure for any serious development work.)
For other architectures (most notably RPM) similar procedures apply, though the details will be different.
Compile Time
When g++ says cannot find -l<nameOfTheLibrary>, it means that g++ looked for the file lib{nameOfTheLibrary}.so, but it couldn't find it in the shared library search path, which by default points to /usr/lib and /usr/local/lib and somewhere else maybe.
To resolve this problem, you should either provide the library file (lib{nameOfTheLibrary}.so) in those search paths or use -L command option. -L{path} tells the g++ (actually ld) to find library files in path {path} in addition to default paths.
Example: Assuming you have a library at /home/taylor/libswift.so, and you want to link your app to this library. In this case you should supply the g++ with the following options:
g++ main.cpp -o main -L/home/taylor -lswift
Note 1: -l option gets the library name without lib and .so at its beginning and end.
Note 2: In some cases, the library file name is followed by its version, for instance libswift.so.1.2. In these cases, g++ also cannot find the library file. A simple workaround to fix this is creating a symbolic link to libswift.so.1.2 called libswift.so.
Runtime
When you link your app to a shared library, it's required that library stays available whenever you run the app. In runtime your app (actually dynamic linker) looks for its libraries in LD_LIBRARY_PATH. It's an environment variable which stores a list of paths.
Example: In case of our libswift.so example, dynamic linker cannot find libswift.so in LD_LIBRARY_PATH (which points to default search paths). To fix the problem you should append that variable with the path libswift.so is in.
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/home/taylor
During compilation with g++ via make define LIBRARY_PATH if it may not be appropriate to change the Makefile with the -Loption. I had put my extra library in /opt/lib so I did:
$ export LIBRARY_PATH=/opt/lib/
and then ran make for successful compilation and linking.
To run the program with a shared library define:
$ export LD_LIBRARY_PATH=/opt/lib/
before executing the program.
First, you need to know the naming rule of lxxx:
/usr/bin/ld: cannot find -lc
/usr/bin/ld: cannot find -lltdl
/usr/bin/ld: cannot find -lXtst
lc means libc.so, lltdl means libltdl.so, lXtst means libXts.so.
So, it is lib + lib-name + .so
Once we know the name, we can use locate to find the path of this lxxx.so file.
$ locate libiconv.so
/home/user/anaconda3/lib/libiconv.so # <-- right here
/home/user/anaconda3/lib/libiconv.so.2
/home/user/anaconda3/lib/libiconv.so.2.5.1
/home/user/anaconda3/lib/preloadable_libiconv.so
/home/user/anaconda3/pkgs/libiconv-1.14-0/lib/libiconv.so
/home/user/anaconda3/pkgs/libiconv-1.14-0/lib/libiconv.so.2
/home/user/anaconda3/pkgs/libiconv-1.14-0/lib/libiconv.so.2.5.1
/home/user/anaconda3/pkgs/libiconv-1.14-0/lib/preloadable_libiconv.so
If you cannot find it, you need to install it by yum (I use CentOS). Usually you have this file, but it does not link to right place.
Link it to the right place, usually it is /lib64 or /usr/lib64
$ sudo ln -s /home/user/anaconda3/lib/libiconv.so /usr/lib64/
Done!
ref: https://i-pogo.blogspot.jp/2010/01/usrbinld-cannot-find-lxxx.html
When you compile your program you must supply the path to the library; in g++ use the -L option:
g++ myprogram.cc -o myprogram -lmylib -L/path/foo/bar
I had this problem with compiling LXC on a fresh VM with Centos 7.8. I tried all the above and failed. Some suggested removing the -static flag from the compiler configuration but I didn't want to change anything.
The only thing that helped was to install glibc-static and retry. Hope that helps someone.
Check the location of your library, for example lxxx.so:
locate lxxx.so
If it is not in the /usr/lib folder, type this:
sudo cp yourpath/lxxx.so /usr/lib
Done.
Apart from the answers already given, it may also be the case that the *.so file exists but is not named properly. Or it may be the case that *.so file exists but it is owned by another user / root.
Issue 1: Improper name
If you are linking the file as -l<nameOfLibrary>
then library file name MUST be of the form lib<nameOfLibrary>
If you only have <nameOfLibrary>.so file, rename it!
Issue 2: Wrong owner
To verify that this is not the problem - do
ls -l /path/to/.so/file
If the file is owned by root or another user, you need to do
sudo chown yourUserName:yourUserName /path/to/.so/file
Here is Ubuntu information of my laptop.
lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 18.04.2 LTS
Release: 18.04
Codename: bionic
I use locate to find the .so files for boost_filesystem and boost_system
locate libboost_filesystem
locate libboost_system
Then link .so files to /usr/lib and rename to .so
sudo ln -s /usr/lib/x86_64-linux-gnu/libboost_filesystem.so.1.65.1 /usr/lib/libboost_filesystem.so
sudo ln -s /usr/lib/x86_64-linux-gnu/libboost_system.so.1.65.1 /usr/lib/libboost_system.so
Done! R package velocyto.R was successfully installed!
This error may also be brought about if the symbolic link is to a dynamic library, .so, but for legacy reasons -static appears among the link flags. If so, try removing it.
The library I was trying to link to turned out to have a non-standard name (i.e. wasn't prefixed with 'lib'), so they recommended using a command like this to compile it -
gcc test.c -Iinclude lib/cspice.a -lm
I encountered the same error message.
I built the cmocka as a so and tried to link it to my executable.
But ld always complains below:
/usr/bin/ld: cannot find -lcmocka
It turns out that there are 3 files generated after cmocka is built:
libcmocka.so
libcmocka.so.0
libcmocka.so.0.7.0
1 and 2 are symbol links and only 3 is the real file.
I only copied the 1 to my library folder, where ld failed to find the 3.
After I copied all 3, ld works.

Why can't libcudart.so.4 be found when compiling the CUDA samples under Ubuntu?

I'm trying to get my Cuda SDK samples running, but I get the following error:
./bandwidthTest: error while loading shared libraries:
libcudart.so.4: cannot open shared object file:
No such file or directory
Why can I compile the example successfully, but not run it? Is there a way to specify the path to the CUDA runtime library manually?
try:
32-bit: sudo ldconfig /usr/local/cuda/lib
64-bit: sudo ldconfig /usr/local/cuda/lib64
cheers
First these that you need is to concatenate the paths to the CUDA binaries and libraries. This is simply done by adding the following lines to your .bashrc file.
export PATH=$PATH:/usr/local/cuda/bin
export LD_LIBRARY_PATH=:/usr/local/cuda/lib64
If you are using a 32-bit operating system change lib64 to lib
Second, there should have been some shared object files installed in /usr/lib or /usr/lib64, depending on your operating system. These object files should be contained in a directory called "nvidia". The two files we are concerned with are names libcuda.so.drivernumber and libOpenCL.so.somenumber. To differentiate between the actual shared object files just use ls -l. The symbolic links will show what they are actually linking to.
As root, execute the following commands:
ln -s /usr/lib64/nvidia/libcuda.so.somenumber /usr/lib64/libcuda.so
ln -s /usr/lib64/nvidia/libOpenCL.so.somenumber /usr/lib64/libOpenCL.so
That should allow you to compile all the sources in the SDK.
As of Cuda 5.5 and Ubuntu 12.04/12.10, the command above becomes (notice the Ubuntu and Cuda directory changes) for 64bit
ln -s /usr/local/cuda/lib64/libcuda.so.5.5 /usr/lib/libcuda.so.5.5
That is, the lib folders on Ubuntu as of 12.04 are lib32 and lib; the 64 is implicit, and cuda 5.5 and greater now installs to a different directory.
1 error while loading shared libraries: libcudart.so.6.0: cannot open shared object file: No such file or directory
32-bit: sudo ldconfig /usr/local/cuda/lib
64-bit: sudo ldconfig /usr/local/cuda/lib64
(refer: http://blog.csdn.net/shenchong721/article/details/21529295)
Works for me!
LD_LIBRARY_PATH is strongly deprecated. It may mess up other programs, and others may reset it. It should only be used to temporarily override the permanent paths for testing purposes (don't take my word, google it).
Instead, add a line with your cuda lib directory on it to /etc/ld.so.conf, after any existing lines.
For example, if you installed on /usr/local/cuda, you will need to add
32-bit : /usr/local/cuda/lib
64-bit : /usr/local/cuda/lib64
Save, and run ldconfig. This should permanently fix the problem.
The symbolic links are probably already set up by the installation. If not, then add them as Alex advised.
Note - I received errors referencing /lib, but I needed to add lib64 to fix them.
create a nvidia_settings.conf file in /etc/ld.so.conf.d/ and add the path to the libs in the file nvidia_settings.conf
/usr/local/cuda/lib64
/usr/local/cuda/lib
Now to update the changes run the following command:
sudo ldconfig
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib
or if you are running cuda-5.0 on a 64-bit machine
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-5.0/lib64
the system find library with ld tool. as the top answer says, 64-bit: sudo ldconfig /usr/local/cuda-xx/lib64 ;;xx is the cuda libraryedition
In my case I was running an application using MPI. The error was:
libcudart.so.7: cannot open shared object file
CUDA was properly installed in all nodes. Also, as in the previous answers, the variables $PATH and $LD_LIBRARY_PATH were pointing to the binary and libraries respectively.
Passing the $PATH and $LD_LIBRARY_PATH in the MPI command solved the issue.
mpirun -x PATH=$PATH -x LD_LIBRARY_PATH=$LD_LIBRARY_PATH ...