[Soot-list] Does ForwardBranchedFlowAnalysis need UnitGraphs?

[Daniel Popescu]

On Wed, Feb 9, 2011 at 1:09 AM, Eric Bodden
<bodden at st.informatik.tu-darmstadt.de> wrote:
>> If the new interface could have a getUnits() method,
>> ForwardBranchedFlowAnalysis would not have to indirectly access the
>> unit chain via the UnitGraph's body.
>
> Yes, but you may be missing the point that there's an implicit
> assumption about the order in which getUnits() returns the units:
> ForwardBranchedFlowAnalysis assumes that when you fall through a unit
> the program will next execute the next unit in the chain that
> getUnits() returns, and if a unit branches then the program will next
> execute some other unit. I am not sure what that would look like in
> your case with multiple bodies.
Thanks Eric!

I have adapted my super graph class. The adapted super graph class
extends UnitGraph and contains a mock body object that points to the
patching unit chain in the super graph. This solution provides the
types that the ForwardBranchedFlowAnalysis requires.

I will also try to make sure that the units in the super graph adhere
to the order constraints. In general, the super graph incorporates
most of the original order of the bodies into its unit chains. I have
only added a couple of new node types around method invocations and
return nodes.

Thanks again.

Daniel

>
> Eric
>
> --
> Dr. Eric Bodden, http://bodden.de/
> Principal Investigator in Secure Services at CASED
> Coordinator of the CASED Advisory Board of Study Affairs
> PostDoc at Software Technology Group, Technische Universität Darmstadt
> Tel: +49 6151 16-5478    Fax: +49 6151 16-5410
> Mailing Address: S2|02 A209, Hochschulstraße 10, 64289 Darmstadt
>