home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!das-news.harvard.edu!spdcc!iecc!compilers-sender
- From: henry@zoo.toronto.edu (Henry Spencer)
- Newsgroups: comp.compilers
- Subject: Re: Is this a new idea?
- Keywords: parse, performance
- Message-ID: <92-11-032@comp.compilers>
- Date: 8 Nov 92 04:36:51 GMT
- Article-I.D.: comp.92-11-032
- References: <92-10-113@comp.compilers> <92-11-009@comp.compilers>
- Sender: compilers-sender@iecc.cambridge.ma.us
- Reply-To: henry@zoo.toronto.edu (Henry Spencer)
- Organization: U of Toronto Zoology
- Lines: 23
- Approved: compilers@iecc.cambridge.ma.us
-
- >[Scanning and parsing can be as much as half of the total time that a
- >compiler takes (so says Ken Thompson of his Plan 9 C compiler) and that is
- >quite amenable to incremental precalculation. -John]
-
- Although the Plan 9 compiler is a very interesting design and the paper on
- it is very much worth reading, some of the comments need to be taken with
- a grain of salt. He isn't doing ambitious optimization (note that he
- ignores `volatile' entirely), which will surely influence the distribution
- of time among phases.
-
- I also recall a case (not a compiler and not Ken Thompson) where an
- authoritative claim was made that a precompiled binary format was
- necessary because re-reading the text file each time was too slow... and
- on examination this appeared to be mostly the result of an input parser
- written so ineptly that it bordered on incompetence. For the code I had
- to write, I ignored the precompiled binary -- its format was complicated
- and inconvenient -- and re-read the text file. The performance impact was
- unnoticeable.
- --
- Henry Spencer @ U of Toronto Zoology, henry@zoo.toronto.edu utzoo!henry
- --
- Send compilers articles to compilers@iecc.cambridge.ma.us or
- {ima | spdcc | world}!iecc!compilers. Meta-mail to compilers-request.
-