Re: SSA, Java and Points-to Analysis

Diego Novillo <dnovillo@redhat.com>
28 Jan 2006 15:11:00 -0500

          From comp.compilers

Related articles
SSA, Java and Points-to Analysis vidyut.vidyut@gmail.com (2006-01-26)
Re: SSA, Java and Points-to Analysis shreyas76@gmail.com (shrey) (2006-01-28)
Re: SSA, Java and Points-to Analysis dnovillo@redhat.com (Diego Novillo) (2006-01-28)
Re: SSA, Java and Points-to Analysis liekweg@ipd.info.uni-karlsruhe.de (Florian Liekweg) (2006-01-28)
Re: SSA, Java and Points-to Analysis jsinger@cs.man.ac.uk (Jeremy Singer) (2006-01-28)
| List of all articles for this month |

From: Diego Novillo <dnovillo@redhat.com>
Newsgroups: comp.compilers
Date: 28 Jan 2006 15:11:00 -0500
Organization: Red Hat Canada
References: 06-01-080
Keywords: Java, analysis

vidyut.vidyut@gmail.com wrote:


> Is there any paper/(any other resource) that describes the extension of
> SSA form for object oriented languages ?
>
You may want to take a look at GCC. We represent aliasing information
directly in the IR (a mixed blessing which is in the process of being
transformed), including structure fields. The representation is similar
to what is used in the SGI compiler.


You may find some high-level descriptions in the GCC summit proceedings
from 2004 and 2005:


Structure aliasing in GCC, Daniel Berlin, GCC Summit 2005
(http://www.gccsummit.org/2005/2005-GCC-Summit-Proceedings.pdf)


Design and Implementation of Tree SSA, Diego Novillo, GCC Summit 2004
(http://www.gccsummit.org/2004/2004-GCC-Summit-Proceedings.pdf)




> (2) Does the use of SSA for Java help in points to analysis ?
>
It does. You almost automatically get flow sensitivity if you do your
analysis on SSA names. We are also working on context sensitive algorithms.



Post a followup to this message

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