CS计算机代考程序代写 ################################################################################

################################################################################
# Homework 3: Matrix-matrix multiplication
# CSCI-2500 Fall 2021
################################################################################
## Data segment
################################################################################
.data
newline: .asciiz “\n”
tab: .asciiz “\t”

################################################################################
## Text segment
################################################################################
.text
.align 2 # instructions must be on word boundaries
.globl main # main is a global label
.globl mm_read # read in NxM matrix
.globl mm_alloc # allocate space for NxM matrix
.globl mm_print # print out formatted NxM matrix
.globl mm_mult # multiple two non-square matrices

################################################################################
# Main segment
################################################################################
main:
# save return address on the stack
sub $sp, $sp, 4
sw $ra, 0($sp)

# Read in a matrix 1 of some NxM size
# Return in $v0 – N
# Return in $v1 – M
# Return in 0($sp) – base address of matrix
sub $sp, $sp, 4 # make space on stack for return value
jal mm_read

# save matrix 1 values
move $s0, $v0
move $s1, $v1
lw $s2, 0($sp)
add $sp, $sp, 4 # restore stack

# Setup arguments and print out matrix 1
move $a0, $s0
move $a1, $s1
move $a2, $s2
jal mm_print

# Read in matrix 2 of some NxM size
sub $sp, $sp, 4 # make space on stack for return value
jal mm_read

# save matrix 2 values
move $s3, $v0
move $s4, $v1
lw $s5, 0($sp)
add $sp, $sp, 4 # restore stack

# Setup arguments and print out matrix 2
move $a0, $s3
move $a1, $s4
move $a2, $s5
jal mm_print

# Setup arguments and call mm_mult
# Note: 5 arguments, so we need to use stack
# Note: num cols matrix 1 == num rows matrix 2
move $a0, $s0 # num rows matrix1
move $a1, $s1 # num cols matrix1
move $a2, $s2 # address of matrix 1
move $a3, $s4 # num cols matrix2
sub $sp, $sp, 4
sw $s5, 0($sp) # address of matrix 2
jal mm_mult
add $sp, $sp, 4

# print the result
move $a0, $s0
move $a1, $s4
move $a2, $v0
jal mm_print

# restore $ra, free stack space, and return
lw $ra, 0($sp)
add $sp, $sp, 4
jr $ra

################################################################################
# mm_read: Read in a NxM matrix from standard input
# Note: You will be calling mm_alloc from this function, so you’ll need to save
# $ra on the stack. Also, you’ll be returning 3 values back to main, one
# value (the matrix base address), will need to be saved on the stack.
# Matrix format is:
# First line: N = num rows
# Second line: M = num cols
# Subsequent lines contain one value each, filling each row in turn
# (same format as hw01, except N,M on separate lines)
# Input:
# None
# Return:
# v0 = N
# v1 = M
# 0($sp) = base address of matrix
################################################################################
mm_read:
# save return address and any saved registers on the stack, if necessary

# get N

# get M

# Setup up arguments and call mm_alloc – v0 is returned as base address

# do nested loops to read in values

# setup up return values
# Note: third return value goes on the stack *after* restoration below

# restore stack, ra, and any saved registers, if necessary

# return to main
jr $ra

################################################################################
# mm_alloc: Allocate an NxM matrix
# Note: You can just allocate a single array of length N*M.
# Input:
# a0 = N
# a1 = M
# Return:
# v0 = base address of allocated space
################################################################################
mm_alloc:
# save return address and any saved registers on the stack, if necessary

# Allocate sufficient space

# set return value

# restore stack, ra, and any saved registers, if necessary

# return to main
jr $ra

################################################################################
# mm_print: Output the formatted contents of NxM matrix
# Note: Within a row, separate each value by tab
# End each row with a newline
# Input:
# a0 = N
# a1 = M
# a2 = base pointer
# Return:
# None
################################################################################
mm_print:
# save return address and any saved registers on the stack, if necessary

# do nested loops to print out values

# restore stack, ra, and any saved registers, if necessary

# return to main
jr $ra

################################################################################
# mm_mult: Perform the actual NxM matrix multiplication
# Input:
# a0 = N for matrix 1
# a1 = M for matrix 1 <-> N for matrix 2
# a2 = base address for matrix 1
# a3 = M for matrix 2
# 0($sp) = base address for matrix 2
# Return:
# v0 = base address for result matrix
################################################################################
mm_mult:
# save return address and any saved registers on the stack, if necessary

# set up and call mm_alloc for result matrix

# three nested loops for matrix-matrix multiplication

# set return value

# restore stack, ra, and any saved registers, if necessary

# return to main
jr $ra