Subject: hardware debug eprom
To: None <pc532@bungi.com, port-pc532@sun-lamp.cs.berkeley.edu>
From: Phil Nelson <phil@cs.wwu.edu>
List: port-pc532
Date: 04/25/1994 14:53:15
>Stuart has problem with `date` hanging on boot under NetBSD & Mach,
>(he has no dallas RTC chip) he used to be OK using date under Mach,
>but no longer, he asked me to post this for him tonight,
>he'll see any answer posted to this list
>(dunno what you can tell him tho' ;-)
>he'll hopefully post a fuller scenario tomorrow,

This is bizarre!  I wrote the NetBSD clock.c to check to see if there
was a RTC chip, and if it none was there, just "config it out".  It
is accessed ONLY at boot time, NOT when the date command is run.
In looking at the code, the only possible problem is that IF it doesn't
find a RTC, it doesn't set "time.tv_sec".  If that has a bad value in
it, it might cause problems, but I don't see how.  

date(1) just adjusts what the kernel thinks is the time and it doesn't
try to set the harware RTC at all.  (It should try at shutdown time,
but currently doesn't.)

I tried using date to set the clock and had no problems.

Q: What date does "date" report BEFORE trying to set the date?

--Phil

------------------------------------------------------------------------------