Motorola DSP Developer's Kit

sort-c.asm

[ Y, I] = mot###_sort( X )

Description

This function sorts the complex input vector X

Input/Output

Input: Complex vector X (includes the real part Xr, and the imaginary part Xi)

Output: Complex vector Y, and vector I (index)

Algorithm

#define  IN        $0
#define  IN_REAL   x:$0
#define  IN_IMAG  y:$0

Algorithm

for(i = 0; i< size; i++)
{
    R[i] = (Xr[i]* Xr[i] + Xi[i]* Xi[i])>>1;
}

Sort vector R using heap sort algorithm, and then adjust vector X according to the index of vector R.

Generate vector R (and its index vector starting from 1), and then put the results in the address:

#IN+SIZE    (X and Y memory, respectively)

Sort vector R and adjust its index vector accordingly.

Adjust the complex data according to the index.

Please refer to sort-r1.asm for the heap sort algorithm.

Intermediate vector:

Vector R: R[i] = (Xr[i]*Xr[i] + Xi[i]*Xi[i])>>1

Start address of Vector R, and Vector I (index) will be calculated by the assembly code.

Vector I is located in Y memory.

Memory & Register

Memory allocation:

• X memory: Label IN stores the address of the real input of vector X
• X memory label IN_REAL & Y memory IN_IMAG are used by the MEX function
• P:$F000 stores the start address of vector R
• P:$F001 stores the start address of vector I

Register usage:

• R7 stores the address of the number of items in the array

Status Register

The assembly function sort-c.asm does not explicitly set any status registers/bits during the function execution.

Data Size Limit

The length of vector X can't be larger than the continuous available data memory size.

Data Range Limit

Input data vector X range [-1.0, +1.0]

Precision

In the case of DSP563, precision is 23 bits.

In the case of DSP566, precision is 15 bits.

sort-r2.asm

sqrt-sr.asm