Xref: utzoo comp.unix.sysv386:7840 comp.unix.wizards:25446 comp.unix.internals:2716
Newsgroups: comp.unix.sysv386,comp.unix.wizards,comp.unix.internals
Path: utzoo!utgpu!news-server.csri.toronto.edu!torsqnt!lethe!tvcent!comspec!scocan!john
From: john@sco.COM (John R. MacMillan)
Subject: Re: SIGPWR signal in system v
Organization: SCO Canada, Inc.
Date: Mon, 06 May 1991 21:05:23 GMT
Message-ID: <1991May06.210523.13080@sco.COM>
References: <1991May6.112253.5344@cs.tcd.ie>
Sender: news@sco.COM (News administration)

|Does anybody out there have information on the SYS V SIGPWR signal? The man
|says that this signal occurs if there's a power failure. I have a couple of
|questioned which I haven't been able to answer myself:

On the only system I've used that implemented SIGPWR, the signal is
raised when power is _restored_ not just before it goes away.  In the
event of a power loss the kernel alone had any warning and was able to
write memory contents and state info to non-volatile storage.  When
power was restored, the kernel restored the previous state, and sent
SIGPWR to all processes.  I can't remember for certain, but I think
the default handling may have been SIG_IGN, so many processeses
wouldn't even notice the power loss.  Those that caught SIGPWR could
refresh their screens, or whatever.