Path: utzoo!attcan!uunet!lll-winken!csd4.milw.wisc.edu!mailrus!cornell!uw-beaver!rice!sun-spots-request
From: anne@cvl.umd.edu (Anne Becker)
Newsgroups: comp.sys.sun
Subject: sockets and SIGINTs?
Message-ID: <8902172306.AA03951@cvl.umd.edu>
Date: 1 Mar 89 01:06:49 GMT
Sender: usenet@rice.edu
Organization: Sun-Spots
Lines: 29
Approved: Sun-Spots@rice.edu
Original-Date: Fri, 17 Feb 89 18:06:50 EST
X-Sun-Spots-Digest: Volume 7, Issue 173, message 5 of 16

First, I suppose I had better provide some background:  I am working on a
SunView tool under 3.4 on some 3/50s, a 3/160 and a 3/260. The proram
itself is a server, which receives its commands over a socket.  Right now
I am using UNIX domain sockets.  I REALLY do not want to go over to using
internet sockets unless there is absolutely no way of solving my problem.

The problem:  Due to what I understand to be a bug in the socket code, If
a ^C is sent to the server at an inappropriate time, i.e. when a socket is
connected, the entire Sun goes into rigor mortis. (At least, it appears
that this is when it happens.)  It also will halt if, while the socket is
open, I go up to the menu and select "quit."

I would like to know if simply ignoring ^C will actually stop this
"feature" from bringing our file server to its knees. If it will, I would
also like to request some pointers to the standard procedure for doing
this, since at this point, I don't have much experience in the area of
trapping signals.  I would very much appreciate any help from any of you
who are wise in the ways of Suns.

[[ You can inhibit ^C interrupts with the "signal" function.  Read the
manual page "signal(3)".  But I don't guarantee that that will prevent
system crashes.  Your best bet is probably to switch over to using
Internet sockets.  It's not that bad....really!  (Note:  "signal() is a
simplified interface to the more general sigvec(2) facility.")  --wnl ]]

	Thanks in advance,
	Anne Becker
	arpa: anne@cvl.umd.edu 
	uucp: uunet!mimsy!cvl!anne