From owner-freebsd-current  Mon Apr  7 15:45:00 1997
Return-Path: <owner-current>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id PAA19539
          for current-outgoing; Mon, 7 Apr 1997 15:45:00 -0700 (PDT)
Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50])
          by freefall.freebsd.org (8.8.5/8.8.5) with SMTP id PAA19532
          for <freebsd-current@freebsd.org>; Mon, 7 Apr 1997 15:44:56 -0700 (PDT)
Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id PAA02421; Mon, 7 Apr 1997 15:26:04 -0700
From: Terry Lambert <terry@lambert.org>
Message-Id: <199704072226.PAA02421@phaeton.artisoft.com>
Subject: Re: POLL & the Single FreeBSD'r
To: peter@spinner.DIALix.COM (Peter Wemm)
Date: Mon, 7 Apr 1997 15:26:04 -0700 (MST)
Cc: terry@lambert.org, freebsd-current@freebsd.org
In-Reply-To: <199704072215.GAA03730@spinner.DIALix.COM> from "Peter Wemm" at Apr 8, 97 06:15:09 am
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

> > scheduler can't keep up, then it can't keep up (the part of the man
> > page I was referrung to was the tv_usec reference). 
> 
> peter@spinner[6:07am]~src/sys/kern-132> man select | col -b | grep tv_
> peter@spinner[6:11am]~src/sys/kern-133>
> What tv_usec reference?  Are you on the wrong system? :-)
> 
> Our 4.4BSD derived man page only says the word "timeout" and mentions nothing 
> about tv_* at all

|    If timeout is a non-nil pointer, it specifies a maximum interval
|    to wait for the selection to complete.

And timeout specified a struct timeval.

It draws a distinction between a "minimally non-zero valued" and a
"zero valued" timeval as well:

|    To affect a poll, the timeout argument should be non-nil,
|    pointing to a zero-valued timeval structure.

(This last is implicit, but no less real for it).

Yeah, I'm picking nits, but you have to admit that those nits result
in useful behaviours under most circumstances.   For instance, a 200uS
timeout for a select on an X fd is sufficient for ~30FPS cell animation,
while still returning unused process time to the system instead of
buzz-looping (I know, I know: now I have self-identified as "one of
those graphics geeks"... 8-)).


Mostly, I've found the SunOS 4.x behaviour useful in the past, and
would like to see that code runable on FreeBSD without a round up
by a factor of 50 in the inter-event delay.  8-).


					Regards,
					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.