[Soot-list] ForwardBranchedFlowAnalysis visits units in the "wrong" order?

[Eric Bodden]

Hi Attila.

> By the way, I think, in the presence of loops, the order does matter to some
> extent, unless you accept that your analysis finds itself in an endless
> loop...

No, if you have a finite lattice and define your meet operation
correctly then you are guaranteed to reach a fixed point after a
finite number of iterations (and usually you will reach it already
after two or three iterations).

> I am working on loops for quite a while now. I find it quite challenging.
> Do you know any analysis specifically dealing with loops, particularly
> dataflow analysis?

I am not aware of any such analyses that come directly shipped with
Soot, except for
soot.jimple.toolkits.annotation.logic.LoopInvariantFinder.

But there may be some projects that use Soot to do things with loops.
I don't really know. I once wrote a loop-based analysis myself with
Soot but it never really worked effectively and hence it never ended
up becoming a part of Soot.

I think the main problem is to first define precisely what exactly the
analyis should compute and how (on paper). Then, if you can proof
termination etc., from my experience the implementation is not such a
big deal any more.

Hope that helps,
Eric