Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!sri-unix!sri-spam!ames!lll-tis!ohlone!lll-winken!csustan!csun!psivax!nrcvax!ihm
From: ihm@nrcvax.UUCP (Ian H. Merritt)
Newsgroups: comp.protocols.tcp-ip
Subject: Re: TCP performance limitations
Message-ID: <1218@nrcvax.UUCP>
Date: Thu, 8-Oct-87 12:05:47 EDT
Article-I.D.: nrcvax.1218
Posted: Thu Oct  8 12:05:47 1987
Date-Received: Tue, 13-Oct-87 03:04:25 EDT
References: <MAB%CORNELLC.BITNET@wiscvm.wisc.edu> <871006093545.3.DCP@KOYAANISQATSI.S4CC.Symbolics.COM>
Reply-To: ihm@minnie.UUCP (Ian Merritt)
Organization: Network Research Corporation (The Frobbozz Magic Network Company)
Lines: 16

>There is a fourth way that we (Symbolics) have done which you did not
>mentioned:
>
>(a) Pick a compile-time unrolling factor, usually a power of 2, say 16 = 2^4.
>(b) Divide the data length by the unrolling factor, obtaining a quotient
>    and remainder.  When the unrolling factor is a power of two, the
>    quotient is a shift and the remainder is a logical AND.
>(c) Write a unrolled loop whose length is the unrolling factor.  Execute
>    this loop <quotient> times.
>(d) Write an un-unrolled loop (whose length is therefore 1).  Execute
>    this loop <remainder> times.

Or if you have memory to burn (which is fast becoming a common
condition), just unroll the loop for the maximum condition and branch
into it at the appropriate point to process the length of the actual
packet.
					--i