Introduction to isolation mechanisms
Some newer microcontrollers include support for isolation between trusted and non-trusted software running onboard. This mechanism is based on a CPU extension, available only on some specific architectures, which usually relies on a sort of physical separation inside the CPU itself between the two modes of execution. All the code running from a non-trusted zone in the system will have a restricted view of the RAM, devices, and peripherals, which must be dynamically configured by the trusted counterpart in advance.
Software running from the trusted area can also provide features that are not directly accessible from the non-trusted world, through special function calls that cross the secure/non-secure boundary.
Chapter 11, Trusted Execution Environment, explores the technology behind Trust Execution Environments (TEEs), as well as the software components involved in real embedded systems to provide a safe environment to run non-trusted modules and components.
