Re: Is it just me or... (Zvi Lamm)
25 Jan 1997 22:13:03 -0500

          From comp.compilers

Related articles
Is it just me or... (1997-01-22)
Re: Is it just me or... (1997-01-25)
Re: Is it just me or... (1997-01-25)
Re: Is it just me or... (1997-01-25)
Re: Is it just me or... (1997-01-26)
Re: Is it just me or... (1997-01-29)
Re: Is it just me or... (Tim Roberts) (1997-01-30)
Re: Is it just me or... (1997-02-07)
Re: Is it just me or... (1997-02-08)
Re: Is it just me or... (Carl Cerecke) (1997-02-11)
[3 later articles]
| List of all articles for this month |

From: (Zvi Lamm)
Newsgroups: comp.compilers
Date: 25 Jan 1997 22:13:03 -0500
Organization: Hebrew University, Jeruslem, Israel
References: 97-01-180
Keywords: courses

Jay Cole ( wrote:

: 3) Are there any other books out there besides "The Design of an
: Optimizing Compiler" by Wulf? Even though this book is out of print
: and I had to get my copy through UMI, I have not been able to find
: another decent code generation book "theory-wise" than this one. His
: ideas on register allocation and jump optimization were a great
: stepping off point for me. Any other recommendations out there on
: this type of code optimization/generation or perhaps other techniques,
: methods?

I always had the feeling that the only big success secret in complier
construction is code generation, and as such - the methods are usually

Parsing as taught in books, is almost a closed subject. Sometimes you
can play some hacks during lexing, or even during the parsing it self,
but most of time you can rely on the standard methods, and use
pre-built tools.

Code generation as taught in books is, however, not enough. It is not
fast enough, easy enough to use, and doesn't produce great code unless
you tweak it in many ways.

If one wants to learn how to produce good code, one has to look to
commercial compilers, in order to learn some real methods. This, by
nature, are usually proprietry.

Since code generation is more of an art than a science, most methods
are not general, and are specific to a certain enviornment. This makes
life tougher.

For these reasons, I advise people interested in compilers to produce
a working parser fast - and concentrate on code generation and
run-time enviornments. If you can the best way to learn these topics
is by dissecting working compilers....
Ehud Lamm

Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.