It looks to me as though the code is accessing the SRAM at address 0x017B02. I've seen that kind of thing for memory-mapped devices, but I'm not sure why anyone would use the above approach instead of letting the compiler and linker locate the variables wherever it saw fit for simple memory management.

Unless, perhaps, that 'C' code was based on earlier legacy assembler code. Perhaps an original assembler version located certain memory elements at certain places in SRAM. When they migrated to 'C' they had to make sure the C program accessed the same memory locations for compatability.

But if the code was originally written in 'C' I sure can't figure out why they did it that way. :)

Craig Steiner