home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.compilers
- Path: sparky!uunet!news.gtech.com!noc.near.net!news.Brown.EDU!qt.cs.utexas.edu!yale.edu!jvnc.net!newsserver.jvnc.net!darwin.sura.net!bogus.sura.net!howland.reston.ans.net!sol.ctr.columbia.edu!eff!world!iecc!compilers-sender
- From: tchannon@black.demon.co.uk (Tim Channon)
- Subject: Re: Code quality
- Reply-To: tchannon@black.demon.co.uk (Tim Channon)
- Organization: Compilers Central
- Date: Thu, 7 Jan 1993 02:20:25 GMT
- Approved: compilers@iecc.cambridge.ma.us
- Message-ID: <93-01-026@comp.compilers>
- Keywords: optimize
- References: <93-01-017@comp.compilers>
- Sender: compilers-sender@iecc.cambridge.ma.us
- Lines: 35
-
- > How important is generated code quality these days? There are a lot of
- > good optimization techniques that seem to be adequate for ordinary
- > programming. But they still are at least 10% or 20% worse than the ideal.
- > Is there much of a market for another 10% in speed of generated code?
-
- It depends... I think that 10..20% is often very optimistic but will
- depend very much on the processor.
-
- A 2MHz 6502 will walk all over a 486/33 under certain cicumstances, same
- algorithm, no choice about the data format and high speed bit twiddling of
- the most awkward sort is not generally a HLL forte.
-
- I've seen this sort of thing time and time again.
-
- What does this mean: IMO understood problems can be solved by a compiler
- because the compiler author or language designer knew a good solution,
- move outside this area and we see the compiler for the dead thing it is.
-
- Then there is the choice of source language. In general these seem to
- cover a pretty narrow range of styles where the programmer has little
- chance to use his brain to come up with imaginative solutions because of
- the straightjacket of the HLL. How often do you see C programmers bending
- their code to try and persuade the compiler to generate certain code?
-
- I guess that some of you may be angry at me suggesting that compilers are
- no good because they cannot produce efficent code for ancient 8 bit
- processors. Look at it this way -- if you need large machine resources to
- produce what you call efficient code you aren't within 20% of optimum in
- the global sense.
-
- TC.
- E-mail: tchannon@black.demon.co.uk or tchannon@cix.compulink.co.uk
- --
- Send compilers articles to compilers@iecc.cambridge.ma.us or
- {ima | spdcc | world}!iecc!compilers. Meta-mail to compilers-request.
-