Pull Request Overview

After reading through the cortex-m assembly 100 times, I think we should capture what is actually going on so future readers can get up to speed quicker. This also tries to make the format consistent since rustfmt doesn't help here.

I also changed assembly like:

movw LR, #0xFFF9
movt LR, #0xFFFF

to:

ldr lr, =0xFFFFFFF9

which asks the assembler to figure out the best way to load that 32 bit immediate into the register. Since we are mostly doing this with numbers which are mostly 0xf-s, the assembler uses the mvn instruction:

mvn lr, #6

which loads the bitwise not of the immediate into the register.

Note! The psuedoinstruction has to use the = symbol, as the # does something very different.

ldr lr, =0xFFFFFFF9  // works as expected
ldr lr, #0xFFFFFFF9  // does something very different and wrong

Testing Strategy

Running on hail.

TODO or Help Wanted

n/a

Documentation Updated

• Updated the relevant files in /docs, or no updates are required.

Formatting

• Ran make prepush.