Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!cs.utexas.edu!asuvax!ukma!hsdndev!spdcc!iecc!compilers-sender
From: ssinghani@viewlogic.com (Sunder Singhani)
Newsgroups: comp.compilers
Subject: Fast Interpreted Code
Keywords: interpreter, threaded code
Message-ID: <1991Apr22.003548.14803@iecc.cambridge.ma.us>
Date: 17 Apr 91 20:40:16 GMT
References: <3035@redstar.cs.qmw.ac.uk> <1991Mar31.180635.5944@cs.rochester.edu>
Sender: compilers-sender@iecc.cambridge.ma.us
Reply-To: ssinghani@viewlogic.com (Sunder Singhani)
Organization: Viewlogic Systems, Inc., MA
Lines: 20
Approved: compilers@iecc.cambridge.ma.us

Recently, there have been a lot of postings on the comparison of
threaded code vs. full compilation performance. We have a system
which uses a threaded code implementation for interpreting this
particular higher level language. The performance is below our
requirements; hence I am looking at different types of IRs (SPC,
RTL, tuples etc.) and VMs to get good interpreted speed. So, I 
have the following question -

   Is threaded code the best form of interpreted code representation
   for performance ? If not, what are the other kinds and how do
   they compare with full compilation ? Any references to general
   comparisons between performance of interpreted code and compiled
   code systems are welcome.

---------------
Sunder Singhani
ssinghani@viewlogic.com
-- 
Send compilers articles to compilers@iecc.cambridge.ma.us or
{ima | spdcc | world}!iecc!compilers.  Meta-mail to compilers-request.