🚀 go-pugleaf

In article <1992Aug18.153253.3473@midway.uchicago.edu>  
eer1@midway.uchicago.edu(Paul Woods) writes:
> Hi -
> 
>  Last week I posted to .sysadmin and I received a couple of responses to  
the  
> following question (none of them were detailed enough).  The HostID on a  
NeXT  
> can be changed ( At least that is what I have been told ) and I need to  
be able  
> to change it on several of my NeXT's.  HOW DO I PATCH THE OS SO THAT IT  
WILL  
> ALLOW ME TO CHANGE THE HOSTID.  I am not a programmer (not really  
anyway) so  
> please keep it on a very simple level.
> 
> Thanks,
> 
> Paul-

Hi Paul.  You can't change the hostid of a Sun or a NeXT.  (Unless there  
is some sort of magical system patch to do it). I missed your previous  
posting so I don't really understand why you would want to.
 
Some people may have read the man page on the "hostid" command and been  
mislead.  The man page says you can change it.  It then refers to the  
gethostid(2) call for more info.  In that man page it states that the  
hostid is permanent and unique for every machine.  It currently returns a  
4-byte integer with the first byte set to 1, and the last three bytes set  
to the last three bytes of the ethernet address of your machine (which are  
guaranteed to be unique).

The system call sethostid(2) doesn't really seem to work (there's no man  
page for it either).  In generic BSD I assume it is possible to set your  
hostid.  In fact I know it is on some machines - I've done it.  However  
the fixed hostid scheme makes a lot more sense to me as it can be used to  
identify a particular CPU and be guaranteed it will never change.  Many  
licensing schemes make use of this.

--
David A. Ross, Programmer Analyst  | dave@nakiska.ucs.ualberta.ca
Computing and Network Services     |  (NeXTmail accepted)
352 Gen Serv Bldg, Univ of Alberta | Phone: (403) 492-9339
Edmonton, Alberta, Canada  T6G 2H1 | Fax:   (403) 492-1729

Why do you need to change the hostid?  Why would you want to change the hostid?

Inquiring minds want to know...

-r


In article <1992Aug18.153253.3473@midway.uchicago.edu>  
eer1@midway.uchicago.edu(Paul Woods) writes:
> Hi -
> 
>  Last week I posted to .sysadmin and I received a couple of responses to the  
> following question (none of them were detailed enough).  The HostID on a NeXT  
> can be changed ( At least that is what I have been told ) and I need to be  
able  
> to change it on several of my NeXT's.  HOW DO I PATCH THE OS SO THAT IT WILL  
> ALLOW ME TO CHANGE THE HOSTID.  I am not a programmer (not really anyway) so  
> please keep it on a very simple level.
> 
> Thanks,
> 
> Paul-

In article <1992Aug20.144252.6882@hot.com> Robert_La_Ferla@hot.com writes:
>Why do you need to change the hostid?  Why would you want to change the hostid?

Because you got your computer's board repaired/replaced and it came
back from the shop with a different hostid.  You then find that all of
the software that you've registered has stopped working.

louie

In article <1992Aug19.194102.3029@kakwa.ucs.ualberta.ca>, dave@nakiska.ucs.ualberta.ca (David Ross) writes:
|> In article <1992Aug18.153253.3473@midway.uchicago.edu>  
|> eer1@midway.uchicago.edu(Paul Woods) writes:

[munch]

|> > to change it on several of my NeXT's.  HOW DO I PATCH THE OS SO THAT IT  
|> WILL  
|> > ALLOW ME TO CHANGE THE HOSTID.  I am not a programmer (not really  
|> anyway) so  
|> > please keep it on a very simple level.
|> > 
|> > Thanks,
|> > 
|> > Paul-
|> 
|> Hi Paul.  You can't change the hostid of a Sun or a NeXT.  (Unless there  
|> is some sort of magical system patch to do it). I missed your previous  
|> posting so I don't really understand why you would want to.
|>  

[munch]

|> 
|> The system call sethostid(2) doesn't really seem to work (there's no man  
|> page for it either).  In generic BSD I assume it is possible to set your  
|> hostid.  In fact I know it is on some machines - I've done it.  However  

[munch]

Well, err. Not quite right. Mach actually has the ability to change
the host id with 'sethostid()', but NeXT chose to disable that feature.
Eric P. Scott found out which table entry has to be patched in NS 2.1.
The output of 'cmp -l /sdmach.old /sdmach' on my machine would give
(it does _not_, as there are some other patches as well...):

sdmach.binpatch:
---- CUT HERE ----
579602   0   1
579607 240 217
579608 214 206
---- CUT HERE ----

This will re-enable the 'sethostid()'-call. For information on how to
apply the patch and some other kernel hacks, see my recent postings
on comp.sys.next.software (topic: RAMDISK).


cheers, Uwe

Uwe Meyer-Gruhl                      "Close female relative wear combat boots"
Lehrstuhl Informatik IX
Technische Universitaet Muenchen     email:MeyerGru@Informatik.TU-Muenchen.DE
Orleansstr. 34, D-8000 Muenchen 80   tel: ++49 89 48095-209

Thinking about it, there seems to be three ways to change your hostid:

1) Fix the system call entry table so that the entry for the
sethostid() system call points to the actual code, rather than the
subroutine which returns an error.

2) Write a program to open /dev/kmem, seek to the right place, and
then write over the spot in kernel memory that the gethostid() system
call grabs the hostid from.

3) Patch the shared C library so that rather than invoking the
gethostid() system call, it just returns in D0 whatever number you
like.

I know the first two methods have been tried; I don't know about the
third.  Much of this would be unnecessary if two things were done:

- The ethernet address/hostid was on a socketed, removeable ROM which
could be easily moved from one CPU board to another when they are
repaired.

- It wasn't as much a pain in the $#@% to maintain and register
software that depends on using hostids as copy protection.  I think
that this mechanism needs to be replaced with something like a public
key cryptosystem scheme that would be useful for other things, too.

louie

In article <1992Aug21.135516.10733@ni.umd.edu>, louie@sayshell.umd.edu (Louis 
A. Mamakos) writes:
	[...material deleted...]
> - It wasn't as much a pain in the $#@% to maintain and register
> software that depends on using hostids as copy protection.  I think
> that this mechanism needs to be replaced with something like a public
> key cryptosystem scheme that would be useful for other things, too.
	Amen.

c.f.waltrip

Internet:	<waltrip@capsrv.jhuapl.edu>

Opinions expressed are my own.