It is really useful to understand how the compiler works and at least is good to know how to run from the CLI with some of the most useful options. For instance, did you know we can call the entire build process one step at a time? or the many options to compile, disassemble, or link.
A real world project is made of several files that needs to be first compiled and then linked, here is the most easy example to compile, just two files with no headers, again it does nothing useful but it works to illustrate what we want
Instead of writing the build instructions in our terminal we are going to create a new file with the name makefile and write all the build instructions there
Learn how the compiler organizes your program. It divides the code into sections: one for program code, one for global variables with initial values, another for uninitialized variables, and even custom sections. The linker later arranges these into your microcontroller’s memory
it is the time to learn how to declare memory regions, this is something can help us to make a better memory organization for our microcontrollers. By using it carefully, you can describe which memory regions may be used by the linker, and which memory regions it must avoid
In our previous post we look into how the compiler organize our code into sections, now is time to see how we can assign those section in different addresses. We are going to require an extra step because dealing with memory addresses is a job for the Linker, let's write the following code:
Like I mentioned in the previous post there is something more sophisticated to organize the memory for our program, the linker scripts allows us to do that. lets go back to the most
In the previous post we learn how to place the most common input sections from a object file inside output sections defined in a linker scripts, but what about the .rodata and the custom sections we can define, well it is basically the same thing
So far we been writing programs that are not possible to flash in our micro, it is because they lack several features in order to run property, but don’t worry everything what you learned on previous post prepares you to create you first program ever
After a reset, the microcontroller doesn’t jump to main. It reads the reset vector at address 0x08000004, which points to the code. The first flash address stores the stack’s initial address, usually the last SRAM memory address. We use a custom section for this setup
Before start talking about OpenOCD lets remember the different pieces of software and hardware you should have in order to successfully code, build and flash any microcontroller. The next diagram shows those different parts:
programs are more complex and requires global variables with and without initialization and to do that we need several things like assign their respective output sections in the linker file and initialize the SRAM memory.
Initialize the .data section is kind of similar but this time wee need to copy the .data section from FLASH to SRAM, The global initialized variables need to stored in flash first, otherwise they gonna lose their values on every power on reset, when we add the line > SRAM AT > FLASH
No decent program is complete without the C Standard library, and important par we have been ignore until now, we did on purpose because we gonna need to mess with the linker and add some extra missing symbols. Remove the flag -nostdlib from the makefile