Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rochester!kodak!elmgate!jdg
From: jdg@elmgate.UUCP (Jeff Gortatowsky)
Newsgroups: comp.sys.m68k
Subject: Re: 68020 bitblt (was 386 vs 020 and big benchmarks)
Message-ID: <622@elmgate.UUCP>
Date: Mon, 27-Apr-87 13:04:56 EDT
Article-I.D.: elmgate.622
Posted: Mon Apr 27 13:04:56 1987
Date-Received: Tue, 28-Apr-87 02:09:20 EDT
References: <930@intsc.UUCP> <513@omen.UUCP> <933@intsc.UUCP> <866@oakhill.UUCP> <650@desint.UUCP> <875@nrcvax.UUCP> <17320@sun.uucp>
Reply-To: jdg@aurora.UUCP (Jeff Gortatowsky)
Organization: Eastman Kodak Company, Rochester, NY
Lines: 26
Keywords: C assembler bitblit 68020
Summary: 26k assembly language programs are NOT that uncommon.

In article <17320@sun.uucp> david@sun.uucp (David DiGiacomo) writes:
>
>However, the biggest disadvantage of the bitfield instructions is that
>it's impossible to use them effectively from C.  The current Sun bitblt
>(libpixrect.a:mem_rop.o) is close to 26K of object code -- there's no
>way I could have written that in assembly language!
>David DiGiacomo, Sun Microsystems, Mt. View, CA sun!david david@sun.com

One thing that always makes me grin is the way some folks look down (pun
intended) at assembler.  26k of assembly code is not that uncommon for a
library or application written in assembly.   Further had you coded your
routines in assembly, it *may* have been significantly less than 26k.

Maintaining the code is another story all together.  How *WELL* written
assembler is maintainable.  One think that irk's me.  If my Sun is FAST
now, just think how fast it could have............(or buggy, umm.. well
then again I'm using 3.2, so much for that! :^) ).
I've always liked to get my code working in C, then recode the low level
stuff into assembler.  Things like pixrects..........


-- 
Jeff Gortatowsky       {seismo,allegra}!rochester!kodak!elmgate!jdg
Eastman Kodak Company  
These comments are mine alone and not Eastman Kodak's. How's that for a
simple and complete disclaimer?