There are four kinds of toolchains you could encounter.

First off, you must understand the following: when it comes to compilers there are up to four machines involved:

  1. the machine configuring the toolchain components: the config machine
  2. the machine building the toolchain components: the build machine
  3. the machine running the toolchain: the host machine
  4. the machine the toolchain is generating code for: the target machine

We can most of the time assume that the config machine and the build machine are the same. Most of the time, this will be true. The only time it isn’t is if you’re using distributed compilation (such as distcc). Let’s forget this for the sake of simplicity.

So we’re left with three machines:

  • build
  • host
  • target

Any toolchain will involve those three machines. You can be as pretty sure of this as “2 and 2 are 4”. Here is how they come into play:

  1. build == host == target (“native”)

    This is a plain native toolchain, targeting the exact same machine as the one it is built on, and running again on this exact same machine. You have to build such a toolchain when you want to use an updated component, such as a newer gcc for example.

  2. build == host != target (“cross”)

    This is a classic cross-toolchain, which is expected to be run on the same machine it is compiled on, and generate code to run on a second machine, the target.

  3. build != host == target (“cross-native”)

    Such a toolchain is also a native toolchain, as it targets the same machine as it runs on. But it is build on another machine. You want such a toolchain when porting to a new architecture, or if the build machine is much faster than the host machine.

  4. build != host != target (“canadian”)

    This one is called a “Canadian Cross”¹ toolchain, and is tricky. The three machines in play are different. You might want such a toolchain if you have a fast build machine, but the users will use it on another machine, and will produce code to run on a third machine.

    The term Canadian Cross was coined because at the time that these issues were all being hashed out, Canada had three national political parties (per Wikipedia).

crosstool-NG can build all these kinds of toolchains, or is aiming at it, anyway. There are a few caveats, though.

While building a “native” toolchain, crosstool-ng will currently still compile new version of libc for the target. There is currently no way to use the system libc and/or system kernel headers as a part of the toolchain. This may work if you choose a compatible version (i.e., the applications compiled with the toolchain will load the system libc).

A “cross-native” toolchain can be built as a trivial case of the “canadian” toolchain. It is suboptimal, as it makes crosstool-NG build the tools targeting the host machine twice (first, as a separate toolchain which is a prerequisite for all canadian builds; and second, as a part of temporary toolchain created as a part of the canadian build itself). This will likely be improved in the future.

To build a “canadian” toolchain, you must build a toolchain that runs on build and targets the host as a prerequisite (i.e., a simple cross). Then, add the /bin directory of this prerequisite to the $PATH environment variable and configure the canadian, specifying the target of the prerequisite toolchain as the host of the new toolchain.

There are a few samples of canadian toolchains shipped with crosstool-NG. The names of the canadian samples consist of two comma-separated parts, i.e. HOST,TARGET. They require HOST sample as a prerequisiste. For example:

ct-ng x86_64-w64-mingw32
ct-ng build
PATH=~/x-tools/x86_64-w64-mingw32/bin:$PATH
ct-ng x86_64-w64-mingw32,x86_64-pc-linux-gnu
ct-ng build

Note that you will not be able to run the binaries from the canadian toolchain on your build machine! You need to transfer them to a machine running the OS configured as the host.