Re: Where is SPAM compiler?

"Uncle Noah" <nkavv@skiathos.physics.auth.gr>
30 Nov 2006 02:07:03 -0500

          From comp.compilers

Related articles
Where is SPAM compiler? xiangdong.ji@gmail.com (sailer) (2006-11-29)
Re: Where is SPAM compiler? nkavv@skiathos.physics.auth.gr (Uncle Noah) (2006-11-29)
Re: Where is SPAM compiler? nkavv@skiathos.physics.auth.gr (Uncle Noah) (2006-11-30)
| List of all articles for this month |
From: "Uncle Noah" <nkavv@skiathos.physics.auth.gr>
Newsgroups: comp.compilers
Date: 30 Nov 2006 02:07:03 -0500
Organization: Compilers Central
References: 06-11-11606-11-126
Keywords: tools
Posted-Date: 30 Nov 2006 02:07:03 EST

Uncle Noah wrote:
> I have my local copies of both the SPAM 1.0 and 2.0 distros. I'll check
> (later tonight) the licensing issues.


Hi there,


it seems that SPAM 2.0 was back then (~2002) available via mail
confirmation. Still there are no licensing terms to be found in the
distro. But there were some licensing terms on the website that i can't
recall :(


You have to contact some of the SPAM authors:
ASHOK SUDARSANAM ashok@ee.princeton.edu
or
SHARAD MALIK sharad@ee.princeton.edu


There is a seminal SPAM compiler paper that you should know about:
"A Retargetable Compilation Methodology for Embedded Digital Signal
Processors Using a
Machine-Dependent Code Optimization Library",
Kluwer Academic Publishers, Design Automation for Embedded Systems, 4,
187-206 (1999).
Kluwer has been acquired by Springer (? correct me if i'm wrong) so the
new DAES should be somewhere at Springer's.


Browsing some of the internals of SPAM it looks like that this is
impressive work. (especially the OLIVE interface as well as the
bindings -- API-level -- for writing your own register allocator). If
you want to develop new ports, which do not resemble those in it (a TI
C25x and a 56k) then there is some significant work you have to do.


And it is over 9 years old.


Nikolaos Kavvadias
PS: Not sure, but it might need a working SUIF 1.x installation.


Post a followup to this message

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