[Soot-list] Inter-procedure constant propagation

[P Lee]

Thanks Eric and Guru for the insights and the headsup of the IFDS
implmentation!

ReachingDefinition analysis is definitely the critical part I've been
looking for for my purpose. I've been considering Inter-procedure slicing
on system graph of inter procedures. Does Soot have any implementation of
System Graph?

Thanks,
Paul


On Mon, Oct 21, 2013 at 3:27 AM, Bodden, Eric <eric.bodden at sit.fraunhofer.de
> wrote:

> > There is an example implementation in
> soot.jimple.toolkits.ide.exampleproblems.IFDSReachingDefinitions.
> >
> > Currently, I use it to find all reaching definitions. Looks like you
> could use that information or supplement that information for your purpose.
> There may be a better way to do it though.
>
> Right. But keep in mind that IFDS implements "may" problems. Paul
> apparently wants to know when a value "must" be a constant. That makes the
> problem harder, as one would typically need to track all non-constant
> values to have a may-not-be-constant analysis (which then has a
> must-be-constant-analysis as its inverse).
>
> Cheers,
> Eric
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.mcgill.ca/pipermail/soot-list/attachments/20131021/07158d73/attachment.html