8086汇编实现快速排序（有bug，运行结果不正确）

  如果有热心大佬能够指出问题所在，感谢！

stack   segment stack
        db 512 dup(?)
stack   ends
    
data    segment
		arr		dw 10, 9, 23, 5, 88
        string1 db "Press any key to continue!", 0ah, 0dh, '$'
		string2 db "Before sort:", 0ah, 0dh, '$'
		string3 db "After sort:", 0ah, 0dh, '$'
		msg		db "Here!", 0ah, 0dh, '$'
data    ends
		
code 	segment
        assume ds: data, cs: code, ss: stack
main:
		mov ax, stack
        mov ss, ax
		mov ax, data
        mov ds, ax
		
		; test 1
		;lea ax, arr
		;push	ax
		;mov ax, 0
		;push	ax
		;mov ax, 8
		;push	ax
		;call	far ptr partition

		;push	ax
		;call	far ptr dispsiw
		;call	far ptr lineFeed
		
		lea dx, string1
		mov ah, 09h
		int 21h
		
		lea ax, arr
		push	ax
		mov ax, 5
		push	ax
		call	far ptr print_arr
		call	far ptr lineFeed
		pop		ax
		pop		ax
		
		lea ax, arr
		push	ax
		mov ax, 0
		push	ax
		mov ax, 8
		push	ax
		call	far ptr quickSort
		pop		ax
		pop		ax
		pop		ax
		
		lea dx, string1
		mov ah, 09h
		int 21h
		
		lea ax, arr
		push	ax
		mov ax, 5
		push	ax
		call	far ptr print_arr
		call	far ptr lineFeed
		pop		ax
		pop		ax
		
		jmp done

		;[bp + 10] 	-- right
		;[bp + 12]  -- left
		;[bp + 14]	-- Arr[]
partition	proc	far
		push	bp
		push	si
		push	di
		mov bp, sp
		
		; [di]{k} -- Arr[k/2]
		mov si, [bp + 14]
		xor di, di
		
		mov bx, [bp + 12]
		and bx, 1
		cmp bx, 1
		je left_is_odd
		jmp left_is_even
left_is_odd:
		mov bx, [bp + 12]
		add bx, 1
		mov [bp + 12], bx
		;jmp left_is_even
left_is_even:
		mov bx, [bp + 10]
		and bx, 1
		cmp bx, 1
		je right_is_odd
		jmp right_is_even
right_is_odd:
		mov bx, [bp + 10]
		add bx, 1
		mov [bp + 10], bx
		;jmp right_is_even
right_is_even:	
		mov bx, [bp + 12]			; bx -- left
		mov ax, [di + bx]			; ax -- pivot
		
		;
		; debugging
		;mov bx, [bp + 10]
		;mov ax, [di + bx]
		;call	far ptr dispsiw
		;
		
		mov dx, [bp + 10]			; dx -- right
		
		;
		; debugging
		;mov ax, bx
		;call	far ptr dispsiw
		;jmp partition_done
		;
		
		cmp bx, dx
		jl left_smaller_right		; left < right
		jmp partition_done
		
left_smaller_right:
		jmp t2_1
t2_1:		
		mov bx, [bp + 10]			; bx -- right
		mov dx, [di + bx]			; dx -- Arr[right]
		cmp ax, dx					; pivot <= Arr[right]
		jng t2_2					; jle
		jmp t3
		
t2_2:		
		mov bx, [bp + 12]			; bx -- left
		mov dx, [bp + 10]			; dx -- right
		cmp bx, dx					; left < right
		jl right_minus_1
		jmp t3

right_minus_1:
		mov dx, [bp + 10]			; dx -- right
		dec dx
		mov [bp + 10], dx			; right--;
		;jmp t3
t3:
		mov bx, [bp + 12]			; bx -- left
		mov dx, [bp + 10]			; dx -- right
		cmp bx, dx					; left != right
		jne swap_1
		jmp t4_1

swap_1:		
		mov bx, [bp + 10]			; bx -- right
		mov dx, [di + bx]			; dx -- Arr[right]
		mov bx, [bp + 12]			; bx -- left
		mov [di + bx], dx			; [di + bx] -- Arr[left], Arr[left] <= Arr[right];
		mov bx, [bp + 12]			; bx -- left
		inc bx
		mov [bp + 12], bx			; left++
		;jmp t4_1
t4_1:
		mov bx, [bp + 12]			; bx -- left
		mov dx, [di + bx]			; dx -- Arr[left]
		cmp dx, ax
		jng t4_2					; jle
		jmp t5

t4_2:
		mov bx, [bp + 12]			; bx -- left
		mov dx, [bp + 10]			; dx -- right
		cmp bx, dx					; left < right
		jl left_plus_1
		jmp t5
		
left_plus_1:
		mov bx, [bp + 12]			; bx -- left
		inc bx
		mov [bp + 12], bx			; left++
		;jmp t5

t5:
		mov bx, [bp + 12]			; bx -- left
		mov dx, [bp + 10]			; dx -- right
		cmp bx, dx					; left != right
		jne swap_2
		jmp get_pivot

swap_2:
		mov bx, [bp + 12]			; bx -- left
		mov dx, [di + bx]			; dx -- Arr[left]
		mov bx, [bp + 10]			; bx -- right
		mov [di + bx], dx			; [di + bx] -- Arr[right], Arr[right] <= Arr[left];
		mov bx, [bp + 10]			; bx -- right
		inc bx
		mov [bp + 10], bx			; right++
		;jmp get_pivot

get_pivot:
		mov bx, [bp + 12]			; bx -- left
		mov [di + bx], ax			; Arr[left] <= pivot
		jmp partition_done

partition_done:
		mov ax, [bp + 12]			; ax -- left
		pop di
		pop si
		pop	bp
		ret
partition	endp

		;[bp + 10] 	-- upper
		;[bp + 12]  -- lower
		;[bp + 14]	-- Arr[]
quickSort	proc	far
		push	bp
		push	si
		push	di
		mov bp, sp
		
		; [di]{k} -- Arr[k/2]
		mov si, [bp + 14]
		xor di, di
		
		mov bx, [bp + 10]		; bx -- upper
		and bx, 1
		cmp bx, 1
		je lower_is_odd
		jmp lower_is_even
lower_is_odd:
		mov bx, [bp + 10]		; bx -- upper
		add bx, 1
		mov [bp + 10], bx
		;jmp lower_is_even
lower_is_even:
		mov bx, [bp + 12]		; bx -- lower
		and bx, 1
		cmp bx, 1
		je upper_is_odd
		jmp upper_is_even
upper_is_odd:
		mov bx, [bp + 12]		; bx -- lower
		add bx, 1
		mov [bp + 12], bx
		;jmp upper_is_even
upper_is_even:
		mov ax, [bp + 14]		; ax -- Arr
		push	ax
		mov ax, [bp + 12]		; ax -- lower
		push	ax
		mov ax, [bp + 10]		; ax -- upper
		push	ax
		call	far ptr partition
		mov dx, ax
		pop		bx
		pop		bx
		pop		bx
		jmp v1
		
v1:	
		mov ax, dx			; ax -- pivot
		mov bx, [bp + 12]	; bx -- lower
		mov dx, ax			; dx -- pivot
		cmp bx, ax			
		jl lower_smaller_pivot	; lower(bx) < pivot(ax)
		jmp v2

lower_smaller_pivot:
		mov ax, [bp + 14]	; ax -- Arr
		push	ax
		mov ax, [bp + 12]	; ax -- lower
		push	ax
		mov ax, dx			; ax -- pivot
		dec ax				; ax -- pivot - 1
		push	ax
		call	far ptr quickSort
		pop		bx
		pop		bx
		pop		bx

v2:
		mov ax, dx			; ax -- pivot
		mov bx, [bp + 10]	; bx -- upper
		cmp ax, bx			
		jl pivot_smaller_upper	; pivot(ax) < upper(bx)
		jmp quickSort_done

pivot_smaller_upper:
		mov ax, [bp + 14]	; ax -- Arr
		push	ax
		mov ax, dx			; ax -- pivot
		inc ax				; pivot + 1
		push	ax
		mov ax, [bp + 10]	; ax -- upper
		push	ax
		call	far ptr quickSort
		pop		bx
		pop		bx
		pop		bx

quickSort_done:
		
		pop di
		pop si
		pop	bp
		ret
quickSort	endp

read_arr	proc	far
		push	bp
		push	si
		push	di
		mov bp, sp
		
		mov si, [bp]
		xor di, di
		mov cx, [bp+10]

readArr:
		call	far ptr readsiw
		mov [di], ax
		add di, 2
		loop readArr
		
		pop di
		pop si
		pop	bp
		ret
read_arr	endp				

print_arr	proc	far
		push	bp
		push 	si
		push 	di
		mov	bp,sp
		
		mov si, [bp]
		xor di, di
		mov cx, [bp+10]			; why must to be [bp+10] here?
		
printArr:
		mov ax, [di]
		call	far ptr dispsiw
		;call	far ptr lineFeed
		call	far ptr space
		add di, 2
		loop printArr

		pop di
		pop si
		pop	bp
		ret
print_arr	endp

readsiw proc	far
		push bx
		push cx
		push dx
		xor bx, bx   
		xor cx, cx   
		mov ah, 1
		int 21h
		cmp al, '+'
		jz rsiw1
		cmp al, '-'
		jnz rsiw2
		mov cx, -1
rsiw1: 
		mov ah,  1
		int 21h
rsiw2:    
		cmp al, '0'
		jb rsiw3
		cmp al, '9'
		ja rsiw3
		sub al, 30h
		xor ah, ah
		shl bx, 1
		mov dx, bx
		shl bx, 1
		shl bx, 1
		add bx, dx
		add bx, ax
		jmp rsiw1
rsiw3:
		cmp cx, 0
		jz rsiw4
		neg bx
rsiw4:
		mov ax, bx
		pop dx
		pop cx
		pop bx
		ret
readsiw endp

dispsiw proc	far
		push ax
		push bx
		push dx
		test ax, ax 
		jnz dsiw1
		mov dl, '0' 
		mov ah, 2
		int 21h
		jmp dsiw5
dsiw1:
		jns dsiw2  
		mov bx, ax
		mov dl, '-'
		mov ah, 2
		int 21h
		mov ax, bx
		neg ax 
dsiw2:  
		mov bx, 10
		push bx 
dsiw3:
		cmp ax, 0
		jz dsiw4
		xor dx, dx 
		div bx  
		add dl, 30h  
		push dx     
		jmp dsiw3
dsiw4:
		pop dx
		cmp dl, 10
		je dsiw5
		mov ah, 2
		int 21h
		jmp dsiw4
dsiw5:
		pop dx
		pop bx
		pop ax
		ret
dispsiw endp

lineFeed proc	far
		push ax
		push dx
		mov dl, 0dh
		mov ah, 2
		int 21h
		mov dl, 0ah
		mov ah, 2
		int 21h
		pop dx
		pop ax
		ret
lineFeed endp

space	proc	far
		pushf
		push ax
		push dx
		
		mov ah,02h
		mov dl,' '
		int 21h
			
		pop dx
		pop ax
		popf
		ret
space	endp


done:		
		lea dx, string1
		mov ah, 09h
		int 21h
		
		mov ah, 07h
		int 21h
		
		
		mov ah, 4ch
		int 21h
code	ends
		end main

运行结果（错误）