[Soot-list] Confusing warnings

Marc-André Laverdière marc-andre.laverdiere-papineau at polymtl.ca
Wed Mar 26 22:22:54 EDT 2014


Hallo,

Phantom classes are classes that don't really exist, just like ghosts.

But since they are referenced in the application you are analyzing, Soot
needs to have some kind of handle on them. Soot does that by creating a
SootClass object for each of those unfound classes and sets its phantom
flag to true (and you get the warning).

Now, if you set your Soot class path such that you'd expect those class
definitions to be found, that means that something is wrong in your CP.

Good luck with your analysis.

Marc-André Laverdière-Papineau
Doctorant - PhD Candidate

On 03/26/2014 03:50 PM, Herr Thomas Vogel wrote:
> Hello everybody,
> 
> I have a quick question and since I'm not a Soot power-user, I send this
> question simply to the mailing list:
> 
> Every time when I run my analysis, Soot prints out the following lines
> and i don’t know what it means.
> 
> Soot started on …
> Warning: java.dyn.InvokeDynamic is a phantom class!
> Warning: java_cup.runtime.Symbol is a phantom class!
> Warning: java_cup.runtime.Scanner is a phantom class!
> Warning: java_cup.runtime.lr_parser is a phantom class!
> [Call Graph] For information on where the call graph may be incomplete,
> use the verbose option to the cg phase.
> 
> Funny fact about that: I do not use the Grappa package
> (http://www2.research.att.com/~john/Grappa/)...
> 
> Short explanation what I am doing:
> I am running a whole program analysis where I firstly extract specific
> Java annotations from the analyzed code by a customized SceneTransformer. 
> With the information of these annotations I am building so called
> environments for every field, method and class which are used inside the
> analyzed code. 
> Those environments will be used in the intra-procedural-part of
> my analysis, i.e. inside a customized BodyTransformer and a
> customized ForwardFlowAnalysis. 
> The overall goal is to find violations against a security-type-system. 
> 
> Regards and thanks for your reply!
> 
> Thomas
> –––––––––––––––––––––––––––––––––––––
> Thomas Vogel
> Mail: vogelt at informatik.uni-freiburg.de
> <mailto:vogelt at informatik.uni-freiburg.de>
> 
> 
> _______________________________________________
> Soot-list mailing list
> Soot-list at sable.mcgill.ca
> http://mailman.cs.mcgill.ca/mailman/listinfo/soot-list
> 


More information about the Soot-list mailing list