|Code Generation by Coagulation? firstname.lastname@example.org (1992-08-12)|
|Re: Code Generation by Coagulation? grover@brahmand.Eng.Sun.COM (1992-08-13)|
|Re: Code Generation by Coagulation? email@example.com (1992-08-18)|
|From:||firstname.lastname@example.org (Jeremy Fitzhardinge)|
|Organization:||Softway Pty Ltd|
|Date:||Wed, 12 Aug 1992 09:27:45 GMT|
|Summary:||Has anyone used Code Generation by Coagulation?|
I've been looking through the paper "Code Generation by Coagualtion" by
Michael Karr, published in 1984. In it, a code generation algorithm is
describes that does instruction selection simultaneously with register
allocation. The code generator has access to runtime flow statistics, and
it generates code for most critical blocks first to least critical last.
It also generates linkage code between them so that most common flow paths
have optimial interfaces.
It all looks quite interesting, but the conclusion says that none of the
ideas has been implemented, so whether it is a good idea is entirely
hypothetical. I'm wondering if this work has been followed up or further
investigated. I'm looking for papers, actual code etc..
email@example.com ph:+61 2 698 2322-x122 fax:+61 2 699 9174
[Someone asked a similar question in 1987 but there weren't any answers.
Return to the
Search the comp.compilers archives again.