View on GitHub JACK Logo - Home

How do I build JACK from the source code?

Important Caveats

There can be a variety of reasons to want to build, install and use a source code version of JACK. However, it is vitally important that you NEVER EVER have two versions of JACK installed on a machine at the same time. If you have an existing installation of JACK (for example, from a Linux distribution) then you must either

(For Linux only) tmpfs file system

JACK’s performance on Linux is much improved if it can use a “tmpfs” (shared memory) file system for certain purposes. Specifically, it will try to use a tmpfs filesystem mounted on /dev/shm by default. You can test if you have this already by running the following command in a terminal:

mount | grep shm

If it generates any output, you are all set. Most modern Linux systems will be this way. If yours is not then we are going to assume that you know what you are doing and understand how to set up a new tmpfs filesystem and how to tell JACK to use it.

For JACK1

Note that if you are installing JACK “on top” of an existing version, you need to add --prefix /usr or similar to the ./configure command, to ensure that the install step will overwrite the existing version. By default these commands will install JACK to /usr/local/bin whereas any system-installed version will likely exist in /usr/bin.

./configure
make
make install

If you are using JACK from svn, you will need to run sh ./autogen.sh before the configure step.

For JACK2

Note that if you are installing JACK “on top” of an existing version, you need to add --prefix /usr or similar to the ./waf configure command, to ensure that the install step will overwrite the existing version. By default these commands will install JACK to /usr/local/bin whereas any system-installed version will likely exist in /usr/bin.

./waf configure
./waf
./waf install