Category : Files from Magazines
Archive   : OCT91.ZIP
Filename : 2N10050A

 
Output of file : 2N10050A contained in archive : OCT91.ZIP
.386p
;---------------------------------------------------------------;
; CPU_Type determines the CPU type in the system. ;
;---------------------------------------------------------------;
; Written by: ;
; Robert Collins ;
; 215 Avenida Espana ;
; San Jose, CA 95139 ;
;---------------------------------------------------------------;
; Input: None ;
; Output: AX = CPU type ;
; 0 = 8086/8088 ;
; 1 = 80186/80188 ;
; 2 = 80286 ;
; 3 = 80386 ;
; 4 = 80486 ;
; FFFF = Unknown CPU type ;
; Register(s) modified: AX, BX, CX, EDX ;
;---------------------------------------------------------------;
; Macro definitions ;
;---------------------------------------------------------------;
; 80486 instruction macro -- because MASM 5.1 doesn't support ;
; the 80486! ;
;---------------------------------------------------------------;
XADD macro ;
db 0fh,0C0h,0D2h ; 80486 instruction macro ;

ENDM ;
;
;
;---------------------------------------------------------------;
CPU_Type proc near ;
;---------------------------------------------------------------;
; Determine the CPU type by testing for differences in the CPU ;
; in the system. ;
;---------------------------------------------------------------;
; To determine if we are a 8086/8088, or 80186/80188, test the
; value of SP after it is placed on the stack. The algorithm
; for "PUSH SP" differs from 8086/80186 to 80286+. The
; algorithm difference is as follows:
;
; 8086/80186 80286+
; { {
; SP = SP - 2 TEMP = SP
; SS:SP = SP SP = SP - 2
; } SS:SP = TEMP
; }
;
; Thus for the 8086/80186, the value of SP that gets pushed on
; the stack is the value after SP is decremented. Hence, the
; value on the stack does not reflect the value of SP before the
; "PUSH" instruction. Therefore, all we have to do to
; categorize the CPU as 8086/8088 or 80186/80188 is to "PUSH SP"
; and compare the value on the stack image to the value in SP.
;---------------------------------------------------------------
xor ax,ax ; clear CPU type return register
push sp ; save SP on stack to look at
pop bx ; get SP saved on stack
cmp bx,sp ; if 8086/8088 these values will
; differ
jz short @Not_8086 ; nope, must be other CPU type

;---------------------------------------------------------------
; If this test passes, then we need some other means to differ-
; entiate between 8088/8088 and 80186/80188. This method I will
; use comes from "80186/188, 80C186/C188 Hardware Reference
; Manual" from Intel, PN# 270788, page A-2: "When a word write
; is performed at offset FFFFh in a segment, the 8086 will write
; one byte at offset FFFFh, and the other at offset 0, while an
; 80186 family processor will write one byte at offset FFFFh,
; and the other at offset 10000h (one byte beyond the end of the
; segment).
;---------------------------------------------------------------
; Before we can blast a value out to FFFFh, we must save
; anything there, so we don't crash anybody else's data.
;---------------------------------------------------------------
push es
mov bx,ds ; get original DS
inc bx
mov es,bx
mov bl,ds:[0] ; get byte @ 0
mov bh,es:[0fff0h] ; get byte @ 10000h
mov ds:[0ffffh],0aaaah ; write signature at
; test location
cmp byte ptr ds:[0],0aah ; 8086?
mov ds:[0],bl ; restore original value
mov es:[0fff0h],bh
pop es
je short CPU_8086_Exit
inc ax
jmp short CPU_8086_Exit

;---------------------------------------------------------------
; When we get here, we know that we aren't a 8086/80186. And
; since all subsequent processors will trap invalid opcodes via
; INT6, we will determine which CPU we are by trapping an
; invalid opcode.
; We are an 80486 if: XADD DX,DX executes correctly
; 80386 if: MOV EDX,CR0 executes correctly
; 80286 if: SMSW DX executes correctly
;---------------------------------------------------------------
; Setup INT6 handler
;---------------------------------------------------------------
@Not_8086:
enter 4,0 ; create stack frame
mov word ptr INT6,offset INT6_handler
mov INT6+2,cs
call set_INT6_vector ; set pointer for INT6 handler
mov ax,4 ; initialize CPU flag=4 (80486)
xor cx,cx ; initialize semaphore

;---------------------------------------------------------------
; Now, try and determine which CPU we are by executing invalid
; opcodes. The instructions I chose to invoke invalid opcodes,
; are themselves rather benign. In each case, the chosen
; instruction modifies the DX register, and nothing else. No
; system parameters are changed, e.g. protected mode, or other
; CPU dependant features.
;---------------------------------------------------------------
; The 80486 instruction 'XADD' xchanges the registers, then adds
; them. The exact syntax for a '486 compiler would be:
; XADD DX,DX.
;---------------------------------------------------------------
XADD ;DX,DX ; 80486
jcxz CPU_exit
dec ax ; set 80386 semaphore
inc cx ; CX=0

;---------------------------------------------------------------
; For a description on the effects of the following instructions,
; look in the Intel Programmers Reference Manual's for the 80186,
; 80286, or 80386.
;---------------------------------------------------------------
mov edx,cr0 ; 80386
jcxz CPU_exit
dec ax ; set 80286 semaphore
inc cx ; CX=0

smsw dx ; 80286
jcxz CPU_exit
sub ax,3 ; set UNKNOWN_CPU semaphore

CPU_exit:
call set_INT6_vector
leave

CPU_8086_exit:
ret


;---------------------------------------------------------------
; Set the INT6 vector by exchanging it with the one currently on
; the stack.
;---------------------------------------------------------------
set_INT6_vector:
push ds
push ABS0 ; save interrupt vector segment
pop ds ; make DS=INT vector segment

ASSUME DS:ABS0
mov dx,word ptr ds:INT_6 ; get offset of INT6
xchg INT6,dx ; set new INT6 offset
mov word ptr ds:INT_6,dx
mov dx,word ptr ds:INT_6[2] ; get segment of INT6
xchg INT6+2,dx ; set new INT6 segment
mov word ptr ds:INT_6[2],dx
pop ds ; restore register
ret ; split
ASSUME DS:_TEXT


;---------------------------------------------------------------
; INT6 handler sets a semaphore (CX=FFFF) and adjusts the return
; address to point past the invalid opcode.
; [BP]
;---------------------------------------------------------------
INT6_handler:
enter 0,0 ; create new stack frame
dec cx ; make CX=FFFF
add word ptr ss:[bp][2],3 ; point past invalid
; opcode
leave
iret
CPU_Type endp


  3 Responses to “Category : Files from Magazines
Archive   : OCT91.ZIP
Filename : 2N10050A

  1. Very nice! Thank you for this wonderful archive. I wonder why I found it only now. Long live the BBS file archives!

  2. This is so awesome! 😀 I’d be cool if you could download an entire archive of this at once, though.

  3. But one thing that puzzles me is the “mtswslnkmcjklsdlsbdmMICROSOFT” string. There is an article about it here. It is definitely worth a read: http://www.os2museum.com/wp/mtswslnk/