[Soot-list] cha, rta: ok but vta fails
Eric Bodden
eric.bodden at ec-spride.de
Tue Jan 8 05:02:00 EST 2013
Hi Pierre.
Try to remove *everything* from your main method except for:
PackManager.v().getPack("wjtp").add(
new Transform("wjtp.myTransform", new SceneTransformer() {
protected void internalTransform(String phaseName, Map options) {
//access call graph here with Scene.v().getCallGraph()
}
}));
soot.Main.main(args);
Then simply set all required options on the command line you use to
call your main class. If you really do with to hard-code those then
modify the "args" array you pass to Soot's main method.
Eric
On 7 January 2013 19:54, Pierre Caserta <pierre.caserta at loria.fr> wrote:
> Hi Eric,
> Thank you for your answer.
>
> To be able to explain correctly, I've written the code I have right bellow.
> Now my code look like this:
>
> soot.options.Options.v().set_keep_line_number(true);
> soot.options.Options.v().allow_phantom_refs();
> soot.options.Options.v().set_whole_program(true);
> soot.options.Options.v().setPhaseOption("cg","verbose:true");
> soot.options.Options.v().setPhaseOption("cg","safe-forname:false");
> soot.options.Options.v().setPhaseOption("cg","safe-newinstance:false");
>
> Scene.v().setSootClassPath(myClassPath);
>
> //pseudo code for this example (just to get the idea)
> for(all classes s) {
> Scene.v().addBasicClass(s);
> Scene.v().loadClassAndSupport(s);
> }
>
> //pseudo code for this example (just to get the idea)
> for(all jars j){
> Scene.v().setSootClassPath(Scene.v().getSootClassPath() + File.pathSeparator + j);
> }
>
> Scene.v().loadNecessaryClasses();
> Scene.v().loadBasicClasses();
>
> if( type.compareTo("cha") == 0 ){
>
> CHATransformer.v().transform();
>
> }else{
> //When type="rta" OR type="vta"
>
> HashMap opt = new HashMap();
> opt.put("on-fly-cg","false");
> opt.put("vta","false");
> opt.put("rta","false");
>
> opt.put( type ,"true");
>
> opt.put("enabled","true");
> opt.put("propagator","worklist"); //worklist iter none
> opt.put("set-impl","double");
> opt.put("double-set-old","hybrid");
> opt.put("double-set-new","hybrid");
> //SparkTransformer.v().transform("",opt);
>
> }
>
> //I'm not very confident about that
> PackManager.v().getPack("wjtp").add(
> new Transform("wjtp.myTransform", new SceneTransformer() {
> protected void internalTransform(String phaseName, Map options) {
> System.err.println(Scene.v().getApplicationClasses());
> }
> }));
> String[] mainClasses = {"abcd.Main"};
> soot.Main.main(mainClasses);
>
> CallGraph cg = Scene.v().getCallGraph();
>
> ---------
> If I remove the line: SparkTransformer.v().transform("",opt);
> I obtain the same result with type="cha", type="rta" and type="vta".
>
> So now that I've removed the line: SparkTransformer.v().transform("",opt);
> How do I specify that I want a call graph RTA or VTA ?
>
> ---------
> If I left the line: SparkTransformer.v().transform("",opt);
> I obtain:
>
> java.lang.NullPointerException
> at soot.SootMethod.getBodyFromMethodSource(SootMethod.java:82)
> at soot.SootMethod.retrieveActiveBody(SootMethod.java:315)
> at soot.jimple.toolkits.callgraph.OnFlyCallGraphBuilder.processNewMethod(OnFlyCallGraphBuilder.java:526)
> at soot.jimple.toolkits.callgraph.OnFlyCallGraphBuilder.processReachables(OnFlyCallGraphBuilder.java:426)
> at soot.jimple.toolkits.callgraph.CallGraphBuilder.build(CallGraphBuilder.java:84)
> at soot.jimple.spark.builder.ContextInsensitiveBuilder.build(ContextInsensitiveBuilder.java:80)
> at soot.jimple.spark.SparkTransformer.internalTransform(SparkTransformer.java:83)
> at soot.SceneTransformer.transform(SceneTransformer.java:39)
> at callgraph.BuildCG.setSparkPointsToAnalysis(BuildCG.java:256)
> at callgraph.BuildCG.startAnalysis(BuildCG.java:185)
> at callgraph.BuildCG.analyse(BuildCG.java:118)
> at callgraph.SootCallGraphGeneration.<init>(SootCallGraphGeneration.java:190)
> at callgraph.SootCallGraphGeneration.main(SootCallGraphGeneration.java:121)
>
> I think it is due to the soot.Main.main(mainClasses); instruction.
> ---------
>
> Any idea?
>
> Best,
> Pierre
>
> ----- Mail original -----
>> De: "Eric Bodden" <eric.bodden at ec-spride.de>
>> À: "Pierre Caserta" <pierre.caserta at loria.fr>
>> Cc: "Soot List" <soot-list at sable.mcgill.ca>
>> Envoyé: Jeudi 3 Janvier 2013 16:05:13
>> Objet: Re: [Soot-list] cha, rta: ok but vta fails
>>
>> Hi Pierre.
>>
>> You are right, the result is incorrect. The question is why this is
>> so. Looking at your code, it seems that you never call
>> soot.Main.main(...). This may be an issue. My recommendation is to
>> write code that:
>>
>> 1.) Sets arguments, as you do now.
>> 2.) Inserts a SceneTransformer into the wjtp pack.
>> 3.) Calls soot.Main.main(...) with your arguments.
>>
>> The SceneTransformer can then simply access the call graph via
>> Scene.v().getCallGraph() - no call to
>> SparkTransformer.v().transform("",opt) is necessary.
>>
>> Eric
>>
>> On 27 December 2012 11:47, Pierre Caserta <pierre.caserta at loria.fr>
>> wrote:
>> > I've clearly described my problem at this url:
>> > http://www.loria.fr/~casertap/vta_problem.html
>> >
>> > Can someone help me please?
>> > Best,
>> > Pierre
>> >
>> > ----- Mail original -----
>> >> De: "Pierre Caserta" <pierre.caserta at loria.fr>
>> >> À: "Quentin Sabah" <quentin.sabah at inria.fr>
>> >> Cc: "Soot List" <soot-list at sable.mcgill.ca>
>> >> Envoyé: Lundi 24 Décembre 2012 14:31:49
>> >> Objet: Re: [Soot-list] cha, rta: ok but vta fails
>> >>
>> >> Hi Quentin,
>> >> Thank you for your reply.
>> >>
>> >> I changed a little bit my example to introduce a virtual call
>> >> inside
>> >> the method m but the vta analysis still does not work.
>> >>
>> >> I've updated my web page to include the new example with the
>> >> resulting call graphs.
>> >>
>> >>
>> >> ----- Mail original -----
>> >> > De: "Quentin Sabah" <quentin.sabah at inria.fr>
>> >> > À: "Soot List" <soot-list at sable.mcgill.ca>
>> >> > Envoyé: Lundi 24 Décembre 2012 11:57:28
>> >> > Objet: Re: [Soot-list] cha, rta: ok but vta fails
>> >> >
>> >> > Hi Pierre,
>> >> >
>> >> > > Hi everybody,
>> >> > > I'm able to generate cha and rta call graphs but vta is not
>> >> > > working.
>> >> > >
>> >> > > To clearly expose my problem, I created a web page:
>> >> > >
>> >> > > http://www.loria.fr/~casertap/vta_problem.html
>> >> >
>> >> > First, I'm not familiar with the VTA implementation in Soot.
>> >> > What I can see is that virtual call edges to m(), and to
>> >> > initializers, are missing in the VTA analysis result. I can also
>> >> > see
>> >> > that these methods have an empty body.
>> >> >
>> >> > Would the VTA performs some kind of side-effect analysis and
>> >> > deduce
>> >> > these edges may be omitted? Or, because these methods provide no
>> >> > typing information at all, the VTA is not reporting them as they
>> >> > should be ?
>> >> >
>> >> > Have you the same kind of results when the methods are filled
>> >> > with
>> >> > useful statements ?
>> >> >
>> >> > --
>> >> > Quentin Sabah, CIFRE Ph.D. student
>> >> > Grenoble University
>> >> > INRIA-SARDES | STMicroelectronics/AST
>> >> > Montbonnot, France | Grenoble, France
>> >> > mailto:quentin.sabah at inria.fr | mailto:quentin.sabah at st.com
>> >> > phone: +33 476 61 52 42 | phone: +33 476 58 44 14
>> >> >
>> >> > _______________________________________________
>> >> > Soot-list mailing list
>> >> > Soot-list at sable.mcgill.ca
>> >> > http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>> >> >
>> >> _______________________________________________
>> >> Soot-list mailing list
>> >> Soot-list at sable.mcgill.ca
>> >> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
>> >>
>> > _______________________________________________
>> > 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
>>
--
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
More information about the Soot-list
mailing list