[TUHS] v6on286

[TUHS] v6on286

[Steve Nickolas]

>From: Maciek Bieszczad <macbiesz at optonline.net>
>Sent: Tue, 08 Jul 2003 11:07:48 -0400
>To: tuhs at minnie.tuhs.org
>Subject: RE: [TUHS] v6on286

>This might help:

>http://nibbly.york.ac.uk/mirrors/TUHS/Other/V6on286/README

>(I'm not sure why it was removed from TUHS)

I did read it.  (Hence, my knowledge to use BC3) :)

I was aware of the / bug and didn't even make it that far.  I was
hoping (still am) that someone did build it and make it that far.

-uso.

kirei-na pinku-na E-MAIL-saito
___________________________________________________________
Get your own Web-based E-mail Service at http://www.zzn.com

[TUHS] v6on286

[JStevens]

FYI the reason why it locks is the 286 multi-tasking is busted.. if you
comment out the iret in ken/slp.c it goes further, but it's busted none the
less...
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20030811/8d4b6ce6/attachment.html>

[TUHS] v6on286

[JStevens]

I just downloaded the source, and after much scrounging around I found my
old Borland 3.1 compiler.. I built the un.exe, and it does the same thing..
blank  screen, and off to nowhere... so I started to add some printf's...
 
Such as this bit in ken\main.c
 
            craftproc();
            printf("main.c craftproc()\n");
            initdevices();
            printf("main.c initdevices()\n");
            cinit();
            printf("main.c cinit()\n");
            binit();
            printf("main.c binit()\n");
 
 
and I modified dmr\cov.c
 
initvid()
        {
//      if (video==1)
//              (gdt_beg+6)->base_l =0;
        outbyte (VIDCA,VIDCSTART);
        outbyte (VIDCD,0);
        outbyte (VIDCA,VIDCEND);
        outbyte (VIDCD,14);
        outbyte (VIDCA,VIDADDRL);
        outbyte (VIDCD,0);
        outbyte (VIDCA,VIDADDRH);
        outbyte (VIDCD,0);
//      clear_screen(vd_attr);
        cosilence();
        }
 
so I could see some more.. it seems to initialize ok under virtual PC.. I
think the next thing is to dig for some more info on the 1st task.. My Lyons
book is somewhere as I want to create a dummy task that prints "a", and a
second one that prints "b", so I can test the scheduler and get
"abababababba" or something like that..
 
Another thing I noticed after trying to build the userland is that crt0.asm
is missing.. I'm too much of a newbie to construct that though.. Anyways
I'll try to get further this weekend either with my Lyons book, or with v6
on simh.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20030808/de0ca265/attachment.html>

[TUHS] v6on286

[Steve Nickolas]

>From: Kenneth Stailey <kstailey at yahoo.com>
>Sent: Wed, 9 Jul 2003 05:37:09 -0700 (PDT)
>To: Steve Nickolas <usagi.tsukino at pinku.zzn.com>
>Subject: Re: RE: [TUHS] v6on286

>The README says:

><< The kernel makes heavy use of the special 286 protected mode 
>features >>

>Try bochs set to be a 286.

I figure that a Celeron is a superset of the 386 - ergo, of the 286
also - so there shouldn't be a problem.  Maybe I'll do that though,
it's safer in a sandbox.

>I am wondering if Cygwin could be used to build the code.  I see that
>ancient C stuff like "=+" was eliminated already.

!!

I think if you converted the ASM to some other format, you could use
Turbo C++ to build it, though...haven't tried that, I don't grok ASM.

>Plus check this site out:
>
>http://www.thefreecountry.com/compilers/cpp.shtml
LOL, I have 6 working C compilers on the Windows/DOS side of my box
already :) (Turbo C++ 1, Borland C++ 3, Watcom C 11, djgpp, MinGW32,
Cygwin)

-uso.

kirei-na pinku-na E-MAIL-saito
___________________________________________________________
Get your own Web-based E-mail Service at http://www.zzn.com

[TUHS] v6on286

[Maciek Bieszczad]

Ah, silly of me. If I used FTP, that could have saved a couple wasted
hours compiling v6on286 :)

--
Maciek (macbiesz at optonline.net)

-----Original Message-----
From: Warren Toomey [mailto:wkt@tuhs.org] 
Sent: Tuesday, July 08, 2003 8:25 PM
To: Maciek Bieszczad
Subject: Re: [TUHS] v6on286

?! It's still here at ftp://minnie.tuhs.org/UnixArchive/Other/V6on286/
but I'm not sure why Apache hides the README when the same directory
is viewed with http://www.tuhs.org/Archive/Other/V6on286/, I'd better
check that out.

	Warren

[TUHS] v6on286

[Steve Nickolas]

I was finally able to download a good copy of v6on286 from minnie...

I have Borland C++ 3.1, the existing version was built with 3.0. 
There are no binaries in the v6on286 package for the Unix itself,
AFAICT, but I did get a successful MAKE.

This is the weird thing, and I'm not sure if it's pilot error, the
fact I'm running Windoze, or a glitch in the code.

C:\UNIX>un

Screen goes blank except for a block flashing cursor.  The keyboard
does not respond - not even the lights - indicating that the machine
is either in a PM loop or completely hung.

Has anyone had better luck than me?

Or have I had better luck than everyone else (author excluded)?

Thx.

BTW...I wonder, could some old C compiler be bootstrapped on v6on286
and then V7 ported on it?

-uso.

kirei-na pinku-na E-MAIL-saito
___________________________________________________________
Get your own Web-based E-mail Service at http://www.zzn.com

[TUHS] v6on286

[Maciek Bieszczad]

This might help:

http://nibbly.york.ac.uk/mirrors/TUHS/Other/V6on286/README

(I'm not sure why it was removed from TUHS)

--
Maciek (macbiesz at optonline.net)

-----Original Message-----
From: tuhs-bounces@minnie.tuhs.org [mailto:tuhs-bounces@minnie.tuhs.org]
On Behalf Of Steve Nickolas
Sent: Tuesday, July 08, 2003 1:22 AM
To: tuhs at tuhs.org
Subject: [TUHS] v6on286

I was finally able to download a good copy of v6on286 from minnie...

I have Borland C++ 3.1, the existing version was built with 3.0. 
There are no binaries in the v6on286 package for the Unix itself,
AFAICT, but I did get a successful MAKE.

This is the weird thing, and I'm not sure if it's pilot error, the
fact I'm running Windoze, or a glitch in the code.

C:\UNIX>un

Screen goes blank except for a block flashing cursor.  The keyboard
does not respond - not even the lights - indicating that the machine
is either in a PM loop or completely hung.

Has anyone had better luck than me?

Or have I had better luck than everyone else (author excluded)?

Thx.

BTW...I wonder, could some old C compiler be bootstrapped on v6on286
and then V7 ported on it?

-uso.

kirei-na pinku-na E-MAIL-saito
___________________________________________________________
Get your own Web-based E-mail Service at http://www.zzn.com
_______________________________________________
TUHS mailing list
TUHS at minnie.tuhs.org
http://minnie.tuhs.org/mailman/listinfo/tuhs

[TUHS] v6on286

[SZIGETI Szabolcs]

> Screen goes blank except for a block flashing cursor.  The keyboard
> does not respond - not even the lights - indicating that the machine
> is either in a PM loop or completely hung.
>
> Has anyone had better luck than me?
>
> Or have I had better luck than everyone else (author excluded)?


Hi,

Sorry, I don't have the soruces with me right now, so I can't be exact, but
I'll check next week.

You need to have a root disk (floppy or hdd), with /etc/init and stuff.

To
do this, first compile unix with the "built in shell". Check te Makefile for
the #define! Also set the root device in the kernel (floppy or hdd).

Make a floppy or hdd (be careful with the sector numbers, mkfs works with
absolute sectors, so it's easy to overwrite another partition) with the dos
based mkfs. Start the kernel with built in shell. It should create /dev/tty
automatically, and then you can populate the rest of the /dev directory.

There is a tool (sorry, I don't remenber the names, but check the source) to
transfer files, via an ather floppy, which is a horrible kludge, first you
write one file on it using one of the utilities, then start the kernel, and
using one of the built in commands you write it into the filesystem. This
has to be done one-by-one.

You can exit the kernel via shift-escape (?), but do a sync before.
Once you have the root floppy, recompile the kernel without the shell.

Now about the bug: I think I broke something in the inode code (I once
wanted to rewrite the code in ansi c, and then lost the original), so
currently, it won't mount the root fs. With plenty of printf's in the kernel
it's possible to debug :-).

Regards,

    Szabolcs