CCS/MSP430F5529: Proper function calls in assembly

I forgot to mention it is all to be done in assembly language for this processor.

When you use a C compiler, it handles the details of parameter passing. Some are passed in registers and if that doesn't provide enough space, the stack is used. In assembly, it is up to you and you can do it any way you like.

The compiler groups the registers into two groups. One group must be save and restored before exit so that the calling programs data isn't mangled. The other group can be clobbered with no consequences. You can do something similar but it is up to you to manage it.

All the details belong to you.

One detail you have messed up is sections. You start out with text (.text directive) but switch to the data section and never return to text. Which means your code is going into a data section.

Data sections are typically placed in SRAM. Which is going to be a problem for your code. In addition, initialized data will be a problem. The C runtime copies initial values into variables at startup but you are on your own.

Before doing anything so complex, you should first get a program to blink an LED running. Then add to it.

I have already done the LED blinking, and it is very different from what I am doing here.  I have made some changes to my code to try and do it with the stack pointer instead.  I am only interested in making the HW multiplier work as a nested subroutine.  I will work on the SW multiplier after this is working.

        .cdecls C,LIST,"msp430.h"       ;Include device header file

        .def    RESET                   ;Export program entry-point to
                                        ;make it known to linker.

		.text                           ;Assemble into program memory.
		.retain                         ;Override ELF conditional linking
        		                        ;and retain current section.
		.retainrefs                     ;And retain any sections that have
        		                        ;references to current section.
calc_power:
 							; save the registers on the stack
		push 	R7 			; save R7, temporal product
		push 	R6 			; save R6, array length
		push 	R4 			; save R4, pointer to array
		clr.w 	R7 			; clear R7
		mov.w 	10(SP), R6	; retrieve array length
		mov.w 	12(SP), R4 	; retrieve starting address
lnext:
		mov.w 	@R4+, R7	; move to next element
		cmp 	#0, R7 		; is it null
		jeq 	lend 		; if yes, go to end
		call	HW_Mult
		mov.w	#val, R4
		jnz		lnext
lend:
		pop 	R4			; restore R4
		pop 	R6 			; restore R6
		pop 	R7 			; restore R7
		ret 				; return

HW_Mult:
		mov.w	b, &MPY		;move b to slot 1
		mov.w	R7, &OP2	;move R7 to slot 2
		nop					;3 clock cycles
		nop
		nop
		mov		RESLO, &val	;put product in val
		ret

		.data
b:		.int 	2						;Create variable and initialize it to 2
val:	.int	0						;Create variable for product placement init 0

RESET:  mov.w   #__STACK_END,SP         ;Initialize stack pointer
        mov.w   #WDTPW|WDTHOLD,&WDTCTL  ;Stop watchdog timer

;-------------------------------------------------------------------------------
; Main loop
;-------------------------------------------------------------------------------
main:
		push 	#hwarr 		; push the address of hwarr
		push 	#5 			; push the number of elements
		sub.w 	#2, SP 		; allocate space for the product
		call 	#calc_power
		mov.w 	@SP, &P1OUT ; store the sum in P2OUT&P1OUT
		add.w 	#6,SP 		; collapse the stack
next:
		push 	#swarr ; push the address of hwarr
		push 	#7 ; push the number of elements
		sub 	#2, SP ; allocate space for the sum
		call 	#calc_power
		mov.w 	@SP, &P3OUT ; store the sume in P4OUT&P3OUT
		add.w 	#6,SP ; collapse the stack

		jmp		$
		nop

hwarr: .int		2, 2, 2, 2, 2 ; the first array
swarr: .int	 	3, 3, 3, 3, 3 ; the second array

;-------------------------------------------------------------------------------
; Stack Pointer definition
;-------------------------------------------------------------------------------
        .global __STACK_END
        .sect   .stack

;-------------------------------------------------------------------------------
; Interrupt Vectors
;-------------------------------------------------------------------------------
         .sect   ".reset"               ; MSP430 RESET Vector
         .short  RESET
         .end

Oh, trust me, I would never use shift-and-add if it wasn't a requirement for calculating the efficiency of the hw multiplier over the sw multiplication.  Doing binary math in this program is completely inefficient.