[Soot-list] speeding up p2 analysis

[Eric Bodden]

Hi Raul.

I think the answer to both your questions would be that yes, it's
certainly possible but unfortunately there's no support for this in
Soot yet. In other words, if you are willing to implement such
functionality then you would certainly be able to do so with Soot, but
currently there's no "out of the box" solution.

Eric

2009/1/22 Raul Santelices <raul at cc.gatech.edu>:
> Hi all,
>  For medium-size or large subjects, spark takes a long time doing
> points-to analysis at the beginning. The RTA and VTA alternatives are
> too imprecise for me. I have two questions in this respect:
>
> 1. Is it possible to make the p2 analysis work on the application
> classes only? I know this would produce unsafe results, but I would like
> to replace the effects of library calls with some hand-coded approximation.
> 2. Is it possible to serialize/deserialize the p2 results so they are
> not re-computed every time I run Soot on the same subject?
>
> Thanks,
> Raul
>
> _______________________________________________
> Soot-list mailing list
> Soot-list at sable.mcgill.ca
> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>



-- 
Eric Bodden
Sable Research Group, McGill University
Montréal, Québec, Canada