Toggling in PDP-11 Bootstraps...

From: Aaron Christopher Finney <af-list_at_wfi-inc.com>
Date: Tue Mar 21 16:33:11 2000

Thanks Pete, I'll give this a try when I get home tonight. And thanks for
putting the comments in as well! It's nice to see what's going on, that's
kind of what I'm after by playing around with this stuff this way.

Cheers,

Aaron

On Tue, 21 Mar 2000, Pete Turnbull wrote:

> On Mar 21, 11:09, Aaron Christopher Finney wrote:
> >
> > Does anyone have any practical advice concerning this, something that
> > might save me a little time/frustration?
>
> At least to start with, I'd ignore the MSCP stuff and play with the 11/73
> and RXV21. It's easier to see what's going on, since you can single-step
> and look at the RXV21 registers; MSCP is a relatively complex protocol
> which uses command buffers in memory, and the registers won't tell you
> much.
>
> > I'd like to be able to set up the simplest system possible and try to
> > toggle in a bootstrap. I've played with the ones on metalab (RX01, MSCP)
> > without much real success so far. Trying to boot from the RX01's using
> the
> > bootstrap code on metalab results in the drive access light coming on,
> but
> > nothing else happening. Sending a break stops at 1044 every time, the
> > contents of that register is 1776. Trying to boot from the Dilog scsi
> > device #0 (should be DU0, right?) just hangs (again using the MSCP
> > bootstrap on metalab). The MSCP bootstrap is well-commented, so I can see
> > what's going on there...a little hint or push may be all I need to be
> able
> > to modify it to work for me. The RX01 bootstrap is not commented at all,
> > unfortunately.
>
> An RXV21 (RX02) bootstrap is different from an RXV11 (RX01). Below is one
> for an RXV21. I also have quite a few little routines to do things like
> echo to the console, test/size RAM, dump memory, find CSRs, do serial port
> loopback tests, etc; all small enough to enter using ODT. Mail me if you
> want any...
>
> 1 000000 ; RXV21_boot From Microcomputer Interfaces
> Handbook 1983-84, page 484
> 2 000000 ; Use ODT to enter, then set RS=340,
> R6=1000, R7=1000, then P
> 3 000000 ;
> 4 000000 ORG O1000
> 5 001000 ;
> 6 001000 012700 MOV #O100240,R0
> 6 001002 100240
> 7 001004 012701 MOV #O177170,R1 ; RXCSR
> 7 001006 177170
> 8 001010 005002 CLR R2
> 9 001012 012705 MOV #O200,R5
> 9 001014 000200
> 10 001016 012704 MOV #O401,R4 ; track 1,
> sector 1
> 10 001020 000401
> 11 001022 012703 MOV #O177172,R3 ; RXDBR
> 11 001024 177172
> 12 001026 030011 BIT R0,(R1)
> 13 001030 001776 BEQ $-4 ; wait for
> TransferReq or Done
> 14 001032 100437 BMI O1132 ; branch if
> ERR set
> 15 001034 012711 MOV #O407,(R1) ; set DDens,
> Read, Go
> 15 001036 000407
> 16 001040 030011 BIT R0,(R1) ; wait for TR
> 17 001042 001776 BEQ $-4
> 18 001044 100432 BMI O1132 ; branch if
> ERR set
> 19 001046 110413 MOVB R4,(R3) ; give sector
> number
> 20 001050 000304 SWAB R4 ; swap track
> and sector
> 21 001052 030011 BIT R0,(R1) ; wait for TR
> 22 001054 001776 BEQ $-4
> 23 001056 110413 MOVB R4,(R3) ; give track
> number
> 24 001060 000304 SWAB R4 ; swap sector
> and track
> 25 001062 030011 BIT R0,(R1) ; wait for TR
> 26 001064 001776 BEQ $-4
> 27 001066 100421 BMI O1132
> 28 001070 012711 MOV #O403,(R1) ; EmptyBuffer
> (DDens) command
> 28 001072 000403
> 29 001074 030011 BIT R0,(R1) ; wait for TR
> 30 001076 001776 BEQ $-4
> 31 001100 010414 MOV R4,(R4) ; save sector
> number
> 32 001102 010513 MOV R5,(R3) ; set word
> count=256
> 33 001104 030011 BIT R0,(R1)
> 34 001106 001776 BEQ $-4 ; wait for
> done
> 35 001110 100410 BMI O1132
> 36 001112 010213 MOV R2,(R3) ; set
> address=0
> 37 001114 060502 ADD R5,R2 ;
> increment...
> 38 001116 060502 ADD R5,R2 ; ...address
> 39 001120 122424 CMPB (R4)+,(R4)+ ; bump R4 by
> 2 and clear all flags
> 40 001122 120427 CMPB R4,#3 ; sectors 1
> and 3 get done
> 40 001124 000003
> 41 001126 003735 BLE O1022 ; loop if not
> finished
> 42 001130 012700 MOV #0,R0
> 42 001132 000000
> 43 001134 005007 CLR PC ; go to
> address zero
> 44 001136 120427 CMPB R4,#0 ; dummy
> operation, pipelined but not executed
> 44 001140 000000
>
> --
>
> Pete Peter Turnbull
> Dept. of Computer Science
> University of York
>
Received on Tue Mar 21 2000 - 16:33:11 GMT