Installation and activation of the Linux cross development environment is described in Chapter 5 of the Linux Cortex-M User's Manual.

Generally speaking, our customers do not have many issues if any with installation and activation of the development environment. For some reason, the most frequently occurring error is that people forget to run the activation script. This obviously results in insufficient path settings and the compiler failing to build  sources for the Cortex-M target.

The activation script needs to be run any time you start a development session. The script resides at the top of the installation directory:

-bash-3.2$ cd linux-cortexm-1.14.2
-bash-3.2$ ls
A2F ACTIVATE.sh linux projects tools u-boot
-bash-3.2$

Here is how the script is executed. Note that you must type '.' followed by a space before the name of the script:

-bash-3.2$ . ./ACTIVATE.sh
-bash-3.2$

An alternative way to run the script is to perform the following command:

-bash-3.2$ source ACTIVATE.sh
-bash-3.2$

Whichever way you run the script, it sets a number of environment variables that are required for the cross development environment to function correctly. One of those variables is called MCU. The easy way to test if the cross development environment is active is to run the following command:

-bash-3.2$ echo $MCU
STM32F4
-bash-3.2$

If MCU is empty or set to something other than STM32F4, this means that you have forgotten to run the activation script or somehow run it incorrectly.

Some of our customers ask if the cross development environment can be run on a Windows host. We don't support Windows, however some of our customers have been successfully running the development environment on a virtual Linux machine under Windows. Again, we don't support that explicitly so if you want to try that, you are on your own.