Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!rutgers!usc!wuarchive!uunet!world!iecc!compilers-sender
From: preston@libya.rice.edu (Preston Briggs)
Newsgroups: comp.compilers
Subject: Re: Span-Dependent Instructions
Keywords: assembler, optimize
Message-ID: <1991Jan3.231845.21646@rice.edu>
Date: 3 Jan 91 23:18:45 GMT
References: <9101022304.AA22609@iecc.cambridge.ma.us> <1991Jan3.060915.3641@rice.edu>
Sender: compilers-sender@iecc.cambridge.ma.us
Reply-To: preston@libya.rice.edu (Preston Briggs)
Organization: Rice University, Houston
Lines: 17
Approved: compilers@iecc.cambridge.ma.us

I wrote:
>Has anyone ever measured the efficacy of any of these algorithms?

and the moderator noted
>[Szymanski's papers have some effectiveness results.  His first paper reports
>that in some pile of PDP-11 code, his algorithm shortened considerably more
>branches than the one the assembler used. -John]

Actually, I meant more along the lines of runtime improvements.  My personal
speculation is that this whole area is a source of only insignificant
improvements.  I could easily be wrong though -- that's why I was wondering
about empirical measurements.

Preston Briggs
-- 
Send compilers articles to compilers@iecc.cambridge.ma.us or
{ima | spdcc | world}!iecc!compilers.  Meta-mail to compilers-request.