Xref: utzoo comp.unix.i386:5697 comp.sys.att:9750
Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uunet!codonics!bret
From: bret@codonics.COM (Bret Orsburn)
Newsgroups: comp.unix.i386,comp.sys.att
Subject: Re: Why is /dev/lp slow?
Message-ID: <1609@codonics.COM>
Date: 7 Jun 90 06:29:31 GMT
References: <110@westmark.UU.NET> <1990Jun1.184946.17448@cbnewsl.att.com>
Followup-To: comp.unix.i386
Distribution: usa
Organization: Codonics, Inc., Middleburg Heights, OH
Lines: 27

In article <1990Jun1.184946.17448@cbnewsl.att.com> rubin@cbnewsl.att.com (Mike Rubin) writes:
>In article <110@westmark.UU.NET> dave@westmark.UU.NET (Dave Levenson) writes:
>>Just upgraded from AT&T SysV/386r3.2 to SysV/386r3.2.2.  Nothing in
>>the hardware has changed....
>>Since upgrading to 3.2.2, the printer has slowed to about 30 cps or so.
>
>Any other board in the system (whether it has a Unix driver installed
>or not) using interrupt 7, will cause this behavior -- it will eat
>some of the interrupts that come from the lp port.
>
>Make sure you have no extraneous boards in your box.
>

A bad cable or connector can do this, too. (At least it could under
Xenix 286, which was the last place I saw this problem.)

If your /ACK signal (Centronics pin 10) is lost, your printer will still
work fine in a polled environment (i.e. DOS) but will slow to a crawl under
Unix (an interrupt-driven environment).



-- 
-------------------
bret@codonics.com
uunet!codonics!bret
Bret Orsburn