Re: Help: Appel's "Modern compiler implementation ..." project

Matthias Blume <see@my.sig>
2 Jun 1999 01:44:53 -0400

          From comp.compilers

Related articles
Help: Appel's "Modern compiler implementation ..." project stefan@bach.informatik.uni-ulm.de (Stefan Sarstedt) (1999-05-20)
Re: Help: Appel's "Modern compiler implementation ..." project will@ccs.neu.edu (William D Clinger) (1999-05-27)
Re: Help: Appel's "Modern compiler implementation ..." project see@my.sig (Matthias Blume) (1999-06-02)
Re: Help: Appel's "Modern compiler implementation ..." project march@bilix.cs.uoregon.edu (1999-06-02)
Re: Help: Appel's "Modern compiler implementation ..." project torbenm@diku.dk (Torben Mogensen) (1999-06-02)
| List of all articles for this month |

From: Matthias Blume <see@my.sig>
Newsgroups: comp.compilers
Date: 2 Jun 1999 01:44:53 -0400
Organization: Center for Information and Multimedia Studies, Kyoto University, JAPAN
References: 99-05-085 99-05-126
Keywords: books, courses

William D Clinger <will@ccs.neu.edu> writes:


> Stefan Sarstedt asked:
> > Has anybody done an implementation of Appel's Compiler project (from
> > his book "Modern compiler implementation in C/Java") for Sparc of
> > Intel maschines?
>
> In the compiler courses I taught during the winter quarter, my
> students implemented a slightly simplified version of the Fun-Tiger
> language, using the SPARC as the target architecture. We did not
> organize the projects as Appel suggests in his textbook, because that
> organization is too complex for a one-quarter compiler course.
> Another simplification came from using the assembler and run-time
> system (notably the generational garbage collector) provided by an
> existing implementation of Scheme.


Andrew Appel himself has taught the course that the book is based on
using Sparc machines. (The undergradutate labs at Princeton mainly
offer Sparcs.) As far as I remember, the languages that he actually
used at one time or the other were either ML or Java.


When I taught the course for one Semester, students were also
developing on Sparc machines. We used the ML text, and following
Andrew's example from past years we went through all of the "Basic
Techniques" chapters (ending up building a real compiler for Tiger
generating executable Sparc code) and through selected "advanced"
chapters. In fact, I believe I touched all the advanced chapters, but
I covered the material in only a cursory fashion.


Matthias


--
Matthias Blume <blume@k_u_r_i_m_s.k_y_o_t_o-u.a_c.j_p>
Kyoto University, Research Institute for Mathematical Sciences
(remove underscores in mail address; they are there to fight spam)





Post a followup to this message

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