Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: henry@utzoo.UUCP (Henry Spencer)
Newsgroups: net.unix-wizards
Subject: Re: Pids
Message-ID: <6669@utzoo.UUCP>
Date: Thu, 8-May-86 15:54:35 EDT
Article-I.D.: utzoo.6669
Posted: Thu May  8 15:54:35 1986
Date-Received: Thu, 8-May-86 15:54:35 EDT
References: <266@brl-smoke.ARPA>
Organization: U of Toronto Zoology
Lines: 12

Several solutions to the inefficient pid-duplication-search problem exist.
A few of them were described in the Usenix newsletter a while ago.  The
simplest approach is the one Chris Torek has mentioned:  it is trivial to
change the search loop so that instead of looking for an exact match, it
looks for the lowest pid >= than the one you're trying.  Then if you don't
get an exact hit (the = part of >=), you've got a range of pids which are
known to be unused, and the search needs to be repeated only when that
range is exhausted.  This idea has been independently re-invented several
times in different places.
-- 
Join STRAW: the Society To	Henry Spencer @ U of Toronto Zoology
Revile Ada Wholeheartedly	{allegra,ihnp4,decvax,pyramid}!utzoo!henry