Hi,
I would like to use Horner's method for fast multiplication on MSP430f2234.
I found source files from:
http://focus.ti.com/general/docs/litabsmultiplefilelist.tsp?literatureNumber=slaa329
multiply_int.c:
/******************************************************************************
* MSP430FG439 Horner's method to perform efficient multiplication
*
* Description: This code performs a integer-integert multplication of numbers
* 41 and 441 to give the result 18081. The main C file calls
* four multiplication functions. The first is a IAR CLIB function
* followed by a conventional multiply algorithm, Horner's method
* and Horner's method with CSD format.
*
* Note: This code assumes 32.768kHz XTAL on LFXT1 on the MSP430FG439
*
* MSP430FG439
* -----------------
* | XIN|-
* | | 32.768kHz
* | XOUT|-
* | |
*
*
* K. Venkat
* Texas Instruments Inc.
* July 2006
* Built with IAR Embedded Workbench Version: 3.41A
********************************************************************************/
#include <stdio.h>
#include <msp430xG43x.h>
int mul_c_int(int,int);
int mul_hamacher(register int, register int);
int mul_horner_int(register int);
int mul_csd_int(register int);
int input=41,mult=441;
int result1, result2,result3,result4;
main()
{
WDTCTL = WDTPW + WDTHOLD; // Stop WDT.
FLL_CTL0 |= XCAP18PF; // Configure load caps.
result1=mul_c_int(input,mult); // C routine called for multiplication.
result2=mul_hamacher(input,mult); // MSP430 Assembly routine for
algorithm.
// explained for multiplication in
Hamacher etal.
result3=mul_horner_int(input); // MSP430 Assembly routine with
Horner's
// method for multiplication.
result4=mul_csd_int(input); // MSP430 Assembly routine with
Horner's
// method using CSD format for
multiplication.
printf("%d\n%d\n%d\n%d\n",result1,result2,result3,result4);
}
int mul_c_int(int x, int y) // C routine for multiplication
{
return(x*y);
}
horner_mul_int.s43
;******************************************************************************
; MSP430FG439 Integer-Integer multiplication using Horner's Algorithm
;
; Description: This code performs a 16-bit integer-integer multplication
; of numbers to return an 16-bit integer
;
;
; Note: This code assumes 32.768kHz XTAL on LFXT1 on the MSP430FG439
; MSP430FG439
; -----------------
; | XIN|-
; | | 32.768kHz
; | XOUT|-
; | |
;
;
; K. Venkat
; Texas Instruments Inc.
; July 2006
; Built with IAR Embedded Workbench Version: 3.41A
;*******************************************************************************/
public mul_horner_int
RSEG CODE
mul_horner_int
mov.w R12,R13 ; The operand "input" in register
R12
rla.w R13
add.w R12,R13 ; X1=X*2^1+X
rla.w R13
rla.w R13
add.w R12,R13 ; X2=X1*2^2+X
rla.w R13
add.w R12,R13 ; X3=X2*2^1+X
rla.w R13
add.w R12,R13 ; X4=X4*2^1+X
rla.w R13
rla.w R13
rla.w R13
add.w R12,R13 ; Final Result=X5=X4*2^3+X
mov.w R13,R12 ; The final answer returned to the
; calling function
ret
END
My question is what exactly asm mul_horner_int does? Why it has only one
argument - comment states that it is integer integer multiplication that
multiplies two numbers?
