/*******************************************************************************
Copyright(c) 2000 - 2002 Analog Devices. All Rights Reserved.
Developed by Joint Development Software Application Team, IPDC, Bangalore, India
for Blackfin DSPs  ( Micro Signal Architecture 1.0 specification).

By using this module you agree to the terms of the Analog Devices License
Agreement for DSP Software. 
********************************************************************************
Module Name     : close.asm
Label name      : __close
Version         :   1.2
Change History  :

                Version     Date          Author        Comments
                1.2         11/18/2002    Swarnalatha   Tested with VDSP++ 3.0
                                                        compiler 6.2.2 on 
                                                        ADSP-21535 Rev.0.2
                1.1         11/13/2002    Swarnalatha   Tested with VDSP++ 3.0
                                                        on ADSP-21535 Rev. 0.2
                1.0         06/29/2001    Raghavendra   Original 

Description     : This function performs close operation on  input binary 
                  image. This is done by calling  dilation function followed by 
                  erosion function Temporary results after calling dilation 
                  function is stored on stack. This temporary result is used as 
                  an input to erosion function and structured elements are 
                  stored as below

                                   | h00  h01  h02 |
                                   | h10  h11  h12 |
                                   | h20  h21  h22 |

Assumption      : 1. Number of columns of input image is even  AND  aligned to 
                     word or half word boundary 
                  2. Input and structured elements are in different banks.
                  3. Same structured elements are used for both dilation and 
                     erosion operation.

Prototype       : void _close (unsigned char* in, int row, int col, short * SE, 
                               unsigned char *out );
               
                   in   ->  Pointer to the input image.
                   row  ->  Number of rows in input image.
                   col  ->  Number of columns in input image.
                   SE   ->  Pointer to 3x3 structured element.
                   out  ->  Pointer to the output buffer.

Registers used  : R0-R7, P0-P5.

Performance     :  
                Code size   : 70 bytes
                Cycle count : 1276 cycles for 8x8 image
*******************************************************************************/
.section  L1_code;
.global   __close;
.align     8;
    
__close:
    [--SP] = (R7:4,P5:3);
                            // Push R7:4 and P5:3 
    R3 = R1.L *R2.L(IS)||P5 = [SP+40];
                            // Row *column and fetch address of SE 
    P3 = R3;                // offset to store temp. result on  stack
    R7 = R1;                // r7 = number of  Rows
    R6 = R2<<0|| P4 = [SP+44];
                            // Address of output array
    [--SP] = RETS;          // store RETS value on stack
    SP -= P3;               // Decrement stack to store temp. result
    R4 = SP;         
    SP += -20;              // decrement stack to store arguments
    [SP+12] = P5;           // Arguments for dilation function
    [SP+16] = R4; 
    CALL __dilation;      
    R0 = R4;                // Arguments for erosion function
    R1 = R7 << 0 || [SP+12] = P5;
    R2 = R6 << 0 || [SP+16] = P4;        
    CALL __erosion;
    SP += 20;               // Increment stack pointer to proper position
    SP = SP + P3;
    RETS = [SP++];          // Pop RETS value
    (R7:4,P5:3) = [SP++];   // Pop R7:4 and P5:3
    RTS;