[Soot-list] pointer-analysis questions

[Eric Bodden]

Hi Dan.

> 2. Here's a simplification of what I would like to do: Suppose there's
> a special static method "foo" and I want a pointer-analysis that is
> context-sensitive in the sense that it computes separate results for
> when (one or more) "foo" calls are "on the stack" (foo has been called
> but not yet returned) versus when no such calls are.  This is scalable
> (it at most doubles the number of contexts), even though it supports
> arbitrary stack depth "between foo and a program point".
> 
> So: Is modifying Spark's notion of "context" to support this context-
> sensitivity (in addition to its other sensitivites) a reasonable
> pursuit?  I think so, but thought it worth checking with the
community.
>   Would I be better off pursuing this work with Paddle instead?

For this *particular* problem, you might actually want to have a look at
the static cflow analysis which is implemented in the AspectBench
Compiler (www.aspectbench.org). It uses Soot/Paddle to statically
analyze whether certain events must or may (not) happen within the
control flow of other events. The paper that describes this technique
can be found here:
http://abc.comlab.ox.ac.uk/papers#pldi2005

Cheers,
Eric

--
Eric Bodden
Sable Research Group, McGill University
Montreal, Canada
http://bodden.de