《Constructing a bootloader STM32F207ZG-SK》.pdf

By Shawn Prestridge, IAR Systems Software Inc. Co-author Colin Flournoy, IAR Systems Software Inc. Constructing a bootloader with IAR Embedded Workbench and the STM32F207ZG-SK • Download application project (zip) As software becomes more and more complex, the potential for defects to exist in shipped code increases exponentially. It is therefore becoming quite common for design specifications to call for the ability to dynamically update a device’s firmware in the field. This is done commonly via a bootloader. In this article, we will discuss some of the issues involved with developing a bootloader and show you how to set one up with the Embedded Workbench on the STM32F207ZG-SK board. The pragmatics of designing a bootloader are somewhat difficult to discuss because there can be so many different requirements placed on it, e.g. the mechanism for getting the new application into the MCU: for some, it will be through the USB peripheral of the MCU, for others, it might be through the Ethernet peripheral and TCP/IP and still others might choose to do it through the venerable RS232 port. Because a bootloader is a very common thing to do, it is highly advisable that you check with your silicon provider’s Field Applications Engineer to see if they have already created a sample bootloader project for your MCU which can greatly simplify your construction of a bootloader to fit your needs. There also are subtle differences when creating a bootloader depending on the architecture you choose. For example the reset vector table may differ slightly whether you are using an ARM device versus a Cortex-M3. Also some chip manufacturers allow you to relocate the vector table while others may not. It is also not within the scope of this article to point out these subtleties, but to, for the sake of brevity, choose one as a demonstration and explain the setup. Our test setup incl