[Soot-list] Question about soot.jimple.toolkits.infoflow package

[Jiaqi Tan]

Hi Richard,

I am looking for a constant propagation analysis.

Yes, I am using Spark for points-to analysis.

I downloaded the latest branch of Soot from git and I got the
following error message:
Exception in thread "main" java.lang.NullPointerException
at soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.handleInvokeExpr(SmartMethodInfoFlowAnalysis.java:411)
at soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.addFlowToCdfg(SmartMethodInfoFlowAnalysis.java:826)
at soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.generateAbbreviatedInfoFlowGraph(SmartMethodInfoFlowAnalysis.java:154)
at soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.<init>(SmartMethodInfoFlowAnalysis.java:131)
at soot.jimple.toolkits.infoflow.ClassInfoFlowAnalysis.getMethodInfoFlowAnalysis(ClassInfoFlowAnalysis.java:57)
.... (my own code) ...

I'm not sure if the following change in
InfoFlowAnalysis.getInvokeInfoFlowSummary() would be acceptable:

Currently getInvokeInfoFlowSummary() can return null if the Stmt
passed to the method has no edges out of it. I modified
getInvokeInfoFlowSummary() to return a new HashMutableDirectedGraph if
ret is still null at the return statement. Is that consistent with
what getInvokeInfoFlowSummary() is trying to do?

Thanks,
Jiaqi

On Mon, Apr 8, 2013 at 2:28 PM, Richard L. Halpert <richhal22 at gmail.com> wrote:
> It looks like you might not be using the latest code.  This code was last
> updated in July 2010, and since then line 415 in
> SmartMethodInfoFlowAnalysis.java has been a blank line.
>
> Additionally, are you using Spark for points-to analysis?  You need to use
> Spark for the infoflow package to work, but I know that this precondition is
> not carefully checked.
>
> -Richard
>
>
> On Mon, Apr 8, 2013 at 10:56 AM, Richard L. Halpert <richhal22 at gmail.com>
> wrote:
>>
>> The code has been used a bit since then, but hasn't been carefully
>> maintained.  Jiangfan Shi (cc'd) is probably its most recent user.
>>
>> What are you trying to do?  The infoflow package can do things similar to
>> escape-analysis, but you're likely better off using Heros since it is a more
>> principled (and I suspect more efficient) approach to whole-program data
>> flow.
>>
>> -Richard
>>
>>
>>
>> On Sun, Apr 7, 2013 at 11:40 PM, Eric Bodden <eric.bodden at ec-spride.de>
>> wrote:
>>>
>>> Hi Jiaqi Tan.
>>>
>>> This code is from 2007 and AFAIK has not been used since. It could
>>> well be that it does not work any more.
>>>
>>> In case you are looking for a taint analysis: We will probably be
>>> releasing an implementation based on Heros and Soot in early May.
>>>
>>> Eric
>>>
>>> On 7 April 2013 22:10, Jiaqi Tan <tanjiaqi at gmail.com> wrote:
>>> > Hi,
>>> >
>>> > I have some questions about trying to get the various analyses in the
>>> > soot.jimple.toolkits.infoflow package to work.
>>> >
>>> > I noticed that in InfoFlowAnalysis, the doApplicationClassesAnalysis()
>>> > method is commented out, is there a bug or problem with the method?
>>> >
>>> > Also, I tried doing this in the internalTransform() of a
>>> > SceneTransformer:
>>> >
>>> > Iterator<SootClass> appClassIter =
>>> > Scene.v().getApplicationClasses().iterator();
>>> > while (appClassIter.hasNext()) {
>>> >      Iterator<SootMethod> mtdIter =
>>> > appClassIter.next().methodIterator();
>>> >      while (mtdIter.hasNext()) {
>>> >           SootMethod currMtd = mtdIter.next();
>>> >           DirectedGraph flowSummary =
>>> > ifa.getMethodInfoFlowSummary(currMtd);
>>> >           InfoFlowAnalysis.printInfoFlowSummary(flowSummary);
>>> >      }
>>> > }
>>> >
>>> > And I got an NPE in:
>>> > Exception in thread "main" java.lang.NullPointerException
>>> > at
>>> > soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.handleInvokeExpr(SmartMethodInfoFlowAnalysis.java:415)
>>> > at
>>> > soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.addFlowToCdfg(SmartMethodInfoFlowAnalysis.java:830)
>>> > at
>>> > soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.generateAbbreviatedInfoFlowGraph(SmartMethodInfoFlowAnalysis.java:154)
>>> > at
>>> > soot.jimple.toolkits.infoflow.SmartMethodInfoFlowAnalysis.<init>(SmartMethodInfoFlowAnalysis.java:131)
>>> > at
>>> > soot.jimple.toolkits.infoflow.ClassInfoFlowAnalysis.getMethodInfoFlowSummary(ClassInfoFlowAnalysis.java:89)
>>> > at
>>> > soot.jimple.toolkits.infoflow.InfoFlowAnalysis.getMethodInfoFlowSummary(InfoFlowAnalysis.java:112)
>>> > at
>>> > soot.jimple.toolkits.infoflow.InfoFlowAnalysis.getMethodInfoFlowSummary(InfoFlowAnalysis.java:108)
>>> >
>>> > How should I use the classes in soot.jimple.toolkits.infoflow?
>>> >
>>> > Thanks!
>>> >
>>> > Regards,
>>> > Jiaqi Tan
>>> > _______________________________________________
>>> > Soot-list mailing list
>>> > Soot-list at sable.mcgill.ca
>>> > http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>>>
>>>
>>>
>>> --
>>> Eric Bodden, Ph.D., http://sse.ec-spride.de/ http://bodden.de/
>>> Head of Secure Software Engineering Group at EC SPRIDE
>>> Tel: +49 6151 16-75422    Fax: +49 6151 16-72051
>>> Room 3.2.14, Mornewegstr. 30, 64293 Darmstadt
>>> _______________________________________________
>>> Soot-list mailing list
>>> Soot-list at sable.mcgill.ca
>>> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>>
>>
>