And speaking of fast compilers...

pardo@cs.washington.edu (David Keppel)
Thu, 12 Nov 1992 06:36:59 GMT

          From comp.compilers

Related articles
And speaking of fast compilers... pardo@cs.washington.edu (1992-11-12)
Re: And speaking of fast compilers... sasdrf@unx.sas.com (1992-11-16)
Re: And speaking of fast compilers... preston@dawn.cs.rice.edu (1992-11-17)
Re: And speaking of fast compilers... cheryl@gallant.apple.com (1992-11-17)
Re: And speaking of fast compilers... pardo@cs.washington.edu (1992-11-17)
Range Checking (was: And speaking of ...) grover@brahmand.Eng.Sun.COM (1992-11-19)
Re: And speaking of fast compilers... pardo@cs.washington.edu (1992-11-23)
[2 later articles]
| List of all articles for this month |

Newsgroups: comp.compilers
From: pardo@cs.washington.edu (David Keppel)
Organization: Computer Science & Engineering, U. of Washington, Seattle
Date: Thu, 12 Nov 1992 06:36:59 GMT
Keywords: performance, Ada, design

[In contrast to lexing and parsing taking 50% of the compile time...]


I was chatting recently with somebody who does a lot of Ada work. He
tried compiling a 7,000 line program using a couple of compilers on a
couple of platforms. The program is almost entirely made up of small
(one-page) functions.


The brand X compiler on a 50Mhz '486 box took 1 minute 48 seconds to
compile. A somewhat older version of the brand X compiler, running on a
SPARCstation, took about 6 minutes to compile. He noted that the SPARC
compiler used 1.1Mb of swap and the assembler used more than 3Mb of swap.


The brand Y Ada compiler, running on the same SPARCstation took 40 minutes
to compile without optimization and used 20Mb of swap. The front end of
the compiler was about 20 minutes of that time. With optimization turned
on, he killed it when it still hadn't finished compiling after an hour and
a half. His guess is that the compiler uses an algorithm that is
superlinear in the number of functions in a module.


He also said that at SIGPLAN '89, some English software engingeers (with a
10,000 line/minute COBOL compiler) said they had 200,000 lines of Ada code
that took a *week* to compile with the brand Z Ada compiler of that day.


BTW, brand X's compiler is a preprocessor that generates C code and calls
the C compiler, the reported times include the C compilation.


;-D on ( Exponential breakage ) Pardo
--


Post a followup to this message

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