Re: Why do we still assemble?

Lars Wirzenius <wirzeniu@cc.helsinki.fi>
Sat, 16 Apr 1994 12:43:48 GMT

          From comp.compilers

Related articles
[30 earlier articles]
Re: Why do we still assemble? mps@dent.uchicago.edu (1994-04-14)
Re: Why do we still assemble? bill@amber.ssd.csd.harris.com (1994-04-14)
Re: Why do we still assemble? hbaker@netcom.com (1994-04-14)
Re: Why do we still assemble? djohnson@arnold.ucsd.edu (1994-04-15)
Re: Why do we still assemble? philw@tempel.research.att.com (1994-04-15)
Re: Why do we still assemble? pardo@cs.washington.edu (1994-04-15)
Re: Why do we still assemble? wirzeniu@cc.helsinki.fi (Lars Wirzenius) (1994-04-16)
Re: Why do we still assemble? hbaker@netcom.com (1994-04-16)
Re: Why do we still assemble? hrubin@b.stat.purdue.edu (1994-04-17)
Re: Why do we still assemble? bevan@cs.man.ac.uk (Stephen J Bevan) (1994-04-18)
| List of all articles for this month |

Newsgroups: comp.compilers
From: Lars Wirzenius <wirzeniu@cc.helsinki.fi>
Keywords: performance, linker, comment
Organization: University of Helsinki
References: 94-04-032 94-04-073
Date: Sat, 16 Apr 1994 12:43:48 GMT

conway@munta.cs.mu.OZ.AU (Thomas Charles CONWAY) writes:
> The unix linker is old technology. Are there newer linkers (not just for
> unix) that are better? What is the state of the art in linkers?


I read this for a course I'm taking; I don't know whether the authors
are the inventors or whether this is state of the art, but what they
describe sounds very nice.


Quong, Russel W., Linton, Mark A.,
"Linking Programs Incrementally",
ACM Transactions on Programming Languages and Systems,
Vol 13, No 1, Jan 1991, p. 1--20.


They've written an incremental linker (Inclink, they call it), which works
by modifying an executable when it needs to be linked again. To do this
efficiently, they overallocate to make room for growth; the overallocation
is 24 %.


According to their measurements, their Inclink is up to 70 times faster
than the standard linker on their system (some version of UNIX). Their
linker works in time proportional to the the size of the change, while a
normal linker works in time proportional to the resulting executable.


(I've never seen Inclink, I've only read the article.)
--
Lars.Wirzenius@helsinki.fi (finger wirzeniu@klaava.helsinki.fi)
[We discussed this here in compilers. If you're willing to take the space
hit, it apparently works well. -John]
--


Post a followup to this message

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