[comp.unix.xenix.sco] Fun with the XENIX kernel

pgd@bbt.se (01/28/91)

In article <19002@rpp386.cactus.org> jfh@rpp386.cactus.org (John F Haugh II) writes:

>The value of _node is initialized when the kernel is built.  The
>copy from _node to _utsname is probably made when the system is first
>booted.

It is initialized from node[] as the first thing in the Xenix kernel
main() procedure, and also (one more time?) in the gdcsb() procedure.
You can also change the node name with the sysi86() system call,
function SETNAME.

>  Note that there are two different "utsname"s.  One is "utsname"
>and the other is "uts3name".  

utsname[] is returned from the uname() system call for sysV
executables, uts3name[] for sysIII executables.

nerd@percy.rain.com (Michael Galassi) (01/29/91)

In article <396@bria>  writes:

>One is trying to set the nodename for the machine by stuffing _node
>with the characters of my choice; in of itself, it isn't a problem.
>The problem is that uname(S) doesn't seem to load the utsname.nodename
>from _node.  Where is the nodename kept in the kernel?  And, yes,
>I know about /etc/systemid ... that ain't the point. :-)

My solution is to go to /usr/sys/conf and edit the master & xenixconf
files inserting the name you want in there.  Then type make and you
will have skirted SCO's longtime bug whereby certain names (3 & 4
characters in lengh I think) can not be configured into the kernel with
the 'configure' program.  If you do modify master and xenixconf make
sure you 'un-modify' them before you run configure next time, then
modify them again.

PS:
when will SCO (who recognizes this as a bug) fix this problem?  I
reported it along with the above workaround almost a year ago.

PPS:
I just installed fas 2.07 under sco Xenix 2.3.3 supporting 4 serial
ports (com1 - com4 in the ms-dog world), currently 3 devices are
16550s, the last one is a 16450, I use them to support two Intel
9600EXs and a Telebit TrailBlazer+ while waiting for my DigiBOARD
pc/8e to come in.  (Your board is in the mail say they!)  Overall
FAS is a very competent set of drivers.  I like!
If there is interest I will post the changes I had to make from the
documented procedure to make them run to this group.
-- 
Michael Galassi				| nerd@percy.rain.com
MS-DOS:  The ultimate PC virus.		| ...!tektronix!percy!nerd