Just got "Advanced Compiler Design Impementation"

Reinhard Wilhelm <wilhelm@cs.uni-sb.de>
9 Nov 1997 12:02:20 -0500

          From comp.compilers

Related articles
Just got "Advanced Compiler Design Impementation" opnek@@bgn.mindspring.com (1997-10-26)
Re: Just got "Advanced Compiler Design Impementation" greened@eecs.umich.edu (1997-10-29)
Re: Just got "Advanced Compiler Design Impementation" ssolyanik@icdc.com (Sergey Solyanik) (1997-11-07)
Just got "Advanced Compiler Design Impementation" wilhelm@cs.uni-sb.de (Reinhard Wilhelm) (1997-11-09)
| List of all articles for this month |

From: Reinhard Wilhelm <wilhelm@cs.uni-sb.de>
Newsgroups: comp.compilers
Date: 9 Nov 1997 12:02:20 -0500
Organization: Compilers Central
References: 97-10-123
Keywords: optimize, books

Jay L Cole wrote:
> I just received the book "Advanced Compiler Design Implementation" by
> Steven S. Muchnick. The book is very high quality by Morgan Kaufmann
> publishers. So far, based on the introduction and table of contents,
> it looks to be a very impressive tome on the other side of compilers.
> It is nice to see a book that emphasizes the code generation and
> optimization side rather than the same old lexing/parsing and
> intermediate code representation. The only other book I've read that
> is good on the code generation side is "Design of an optimizing
> compiler" -- Wolff. Out of print now.
>
> The bibliographical references are worth the book price alone :).


Well, if you think that compilers are an American invention and
unknown everywhere else, you may be content with the references.
However, if you assume that compiler research and implementation has
happened also outside of North America you should be less content.


The bibliography surprises with almost complete absence of non-
American references.


Not a single reference to Abstract Interpretation invented 20 years
ago (by some Frenchies), incorporated into quite a few compilers.
Only this theory provided the basis to understand and prove that some
of the data flow analyzes worked (many don't). No reference to the
still best paper on interprocedural data flow analysis (Sharir &
Pnueli, two Israelis). This is particularly annoying, since the paper
has been published 1981 in a book edited by Steve Muchnik and Neil
Jones.


Even the Dragon Book did a lot more justice to the rest of the world
in crediting their developments. cf. all those Russians who invented
fundamental algorithms for compiler backends, e.g. graph colouring or
the famous labeling algorithm.


Sorry, Steve, this had to be said.


Reinhard Wilhelm


FB Informatik
Universitaet des Saarlandes
66123 Saarbruecken
Germany
--


Post a followup to this message

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