Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!news.mira.net.au!news.netspace.net.au!news.mel.connect.com.au!munnari.OZ.AU!comp.vuw.ac.nz!ak.netlink.co.nz!manawatu.planet.co.nz!manawatu.gen.nz!news.express.co.nz!news.triton.net!news2.agis.net!agis!news4.agis.net!news5.agis.net!agis!newspeer.santaclara.agis.net!agis!su-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!rill.news.pipex.net!pipex!tank.news.pipex.net!pipex!news.utell.co.uk!usenet
From: brian@shift.utell.net (Brian Somers)
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: Re: sio?: silo overflow
Date: 10 Apr 1997 14:20:44 GMT
Organization: Awfulhak Ltd.
Lines: 47
Message-ID: <5iisvs$d4p@ui-gate.utell.co.uk>
References: <5ig6ju$rmo@winx03.informatik.uni-wuerzburg.de>
Reply-To: brian@awfulhak.org, brian@utell.co.uk
NNTP-Posting-Host: shift.utell.net
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Newsreader: knews 0.9.8
Xref: euryale.cc.adfa.oz.au comp.unix.bsd.freebsd.misc:38923
In article <5ig6ju$rmo@winx03.informatik.uni-wuerzburg.de>,
token@cip.informatik.uni-wuerzburg.de (Matthias Buelow) writes:
> Hello folks,
>
> I've been configuring a 486dx2-80/8MB to route tcp/ip between a winNT
> box connected via ethernet, an amiga/060 connected by slip on cuaa1
> (115200 bps) and a modem (28k8, on cuaa0, speed 115200) which connects
> the stuff with the outer world using pppd. The operating system is
> FreeBSD 2.2.1R. It all works nicely except some occasional messages
> syslog spits all over the root console explaining that the sio-driver
> detected some silo overflow on the serial ports. Both serial ports
> are 16550A and have a fifo. The one that attaches slip to the amiga
> gets some more overflows (7 so far in a few hours yesterday) while the
> modem port is almost neglectable (2 overflows). We had some more trouble
> when using the onboard ports (we disabled & replaced them with a board)
> especially that the device kept sending junk data when the modem was
> already off or even disconnected (we discovered that with minicom and
> ppp-term). That doesn't happen anymore with the seperate ports on the
> i/o-board.
> I've looked up the driver source and saw that the message is printed when
> the serial control-line indicates a data register overrun. I'm not very
> familiar with serial port programming on PCs so I don't know where exactly
> the problem is. When does this overflow happen? On incoming or outgoing
> data? The manpage just says "problem with the interrupt handler" or
> similar. Does it indicate some weak hardware or can I fix it? It's not
> really a problem but somewhat annoying.
This would indicate an incoming character was missed. An outgoing
character wouldn't be sent 'till the last one was gone, so there's
no concept of "missing" an outgoing character.
This shouldn't really happen unless your machine is *really* busy,
but even if it does, it's no great loss. The ip checksum will fail
and the packet will be dropped (and resent later). If you get a lot
of overflows, you're basically losing bandwidth due to resends.
Having said that, it happens occasionally. I got a dropped character
the other day on a PP200 w/ 16550 UART. All it was doing was
receiving data. I guess it's possible that the modem gets a
load of data that was compressed at 4x and swamps the uart... not
likely though :| especially by only one character.
I guess it was a cosmic glitch.
--
Brian <brian@awfulhak.org> <brian@freebsd.org>
<http://www.awfulhak.demon.co.uk>
Don't _EVER_ lose your sense of humour !