From owner-freebsd-hackers  Wed Jan 22 16:39:05 1997
Return-Path: <owner-hackers>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id QAA12659
          for hackers-outgoing; Wed, 22 Jan 1997 16:39:05 -0800 (PST)
Received: from watermarkgroup.com (lor.watermarkgroup.com [38.246.139.30])
          by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id QAA12646
          for <freebsd-hackers@freebsd.org>; Wed, 22 Jan 1997 16:39:00 -0800 (PST)
Received: from 38.246.139.33 (Kim.watermarkgroup.com) by watermarkgroup.com (4.1/SMI-4.1)
	id AA18850; Wed, 22 Jan 97 19:37:51 EST
Message-Id: <32E6B2C2.124B@watermarkgroup.com>
Date: Wed, 22 Jan 1997 19:37:29 -0500
From: Luoqi Chen <luoqi@watermarkgroup.com>
Organization: The Watermark Group
X-Mailer: Mozilla 3.01Gold (Macintosh; I; PPC)
Mime-Version: 1.0
To: Bruce Evans <bde@zeta.org.au>
Cc: freebsd-hackers@freebsd.org
Subject: Re: My sportster internal modem doesn't work after switched to 2.2-BETA
References: <199701211839.FAA14689@godzilla.zeta.org.au>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

Bruce Evans wrote:
> 
> >After some poking around, I found that the if I sent multiple bytes
> >to the modem, some would get lost. So I went to sio.c and changed
> >com->tx_fifo_size to 1 from 16, and everything works fine now.
> >But I am still wondering if this is a problem with the modem, or with
> >the software. It seems that NT is doing the same, sending no more than
> >one
> >byte at a time. This is not a satisfactory solution, the output FIFO
> >is underutilized and CPU is servicing 16 times more interrupts.
> 
> Which version did it work in?  The sio driver streams output better
sio.c:
     $Id: sio.c,v 1.147.2.5 1996/12/23 19:59:08 bde Exp $

> than it used to.  There used to be a small amount of dead time every
> 256(?)  characters.  Perhaps the internal modem can't quite keep up
> and its flow control doesn't work or you're not using flow control.
> Reducing com->tx_fifo_size gives flow control a better chance of working.
> 
> Try other values < 16 for com->tx_fifo_size.
I tried other values, only 1 worked. How does the flow control inside
an internal modem work? Do you need to check MSR_CTS each time you push
a byte into the fifo? Or you could check once and fill up the fifo if
CTS is set?

> 
> Bruce