From owner-freebsd-current  Fri Jun  6 21:03:27 1997
Return-Path: <owner-current>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id VAA27461
          for current-outgoing; Fri, 6 Jun 1997 21:03:27 -0700 (PDT)
Received: from bmccane.uit.net (bmccane.uit.net [208.129.189.48])
          by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA27456
          for <current@FreeBSD.ORG>; Fri, 6 Jun 1997 21:03:13 -0700 (PDT)
Received: from bmccane.uit.net (localhost.mccane.com [127.0.0.1])
	by bmccane.uit.net (8.8.5/8.8.5) with ESMTP id XAA26092;
	Fri, 6 Jun 1997 23:02:11 -0500 (CDT)
Message-Id: <199706070402.XAA26092@bmccane.uit.net>
X-Mailer: exmh version 2.0gamma 1/27/96
To: Michael Smith <msmith@atrad.adelaide.edu.au>
cc: straka@inficad.com (Richard Straka), bde@zeta.org.au, current@FreeBSD.ORG
Subject: Re: sio driver performance 
In-reply-to: Your message of "Fri, 06 Jun 1997 16:26:30 +0930."
             <199706060656.QAA02486@genesis.atrad.adelaide.edu.au> 
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Fri, 06 Jun 1997 23:02:10 -0500
From: Wm Brian McCane <root@bmccane.uit.net>
Sender: owner-current@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

> Richard Straka stands accused of saying:
> > 
> > I have noticed that the communications work great until I interrupt the
> > process on the box which is tranmitting the data.  At that point, the box
> > which was receiving suddenly elevates to 100% CPU usage (mostly system) as
> > if in a polling loop.
> 
> Read the manpage for the read() system call, and pay particular
> attention to the return value for EOF.  I suspect that when the sender
> is killed it is lowering whatever control signal is driving DCD at the
> receiving end.  Without seeing the rest of your code, it's hard to be
> sure if this is the case.
> 
> > Also when the receive process is started, before the transmit process is
> > started on the other box, the read seems to periodically return -1.
> > Shouldn't the read indefinitely block if c_cc[VMIN]=1 and c_cc[VTIME]=0?
> 
> Check errno on these occasions; I would guess that you are probably
> seeing EINTR.
> 
> > The "tty-level buffer overflows" only occurs after the transmit process is
> > started, then halted, then restarted. This could be a real problem for my
> > remote test equipment as the serial lines may be inadvertently
> > disconnected and reconnected from time to time.
> 
> I suspect that this is an artifact of the receiver having lost its brain.
> 
> >  Richard Straka 
> 
I think Mike is right on this one.  I would suggest adding a select call to 
your program with a read and exception fd_set.  If the bit gets set in the 
exception fd_set, close the port and reopen it.  This way you will wait until
DCD is high before continuing your reads.  It also makes a nice spring-board 
for when someone asks you to monitor more ports at once.

	brian