[Soot-list] Call Graph Performance and Alternatives

Eric Bodden bodden at st.informatik.tu-darmstadt.de
Thu Sep 30 03:11:24 EDT 2010


Hi all.

As Patrick correctly said, it is usually unsound to not analyze the
JDK. However, because the 1-2 minutes wait time is a know issue (that
cannot be avoided otherwise), I have recently included an option that
allows you to skip over the JDK during call-graph construction. Soot
is much faster then, but you may be missing some call edges (certainly
all for the JDK).

See here: http://www.bodden.de/2010/06/04/soot-no-jdk/

And you may want to use a nightly build for this:
http://plg.uwaterloo.ca/~olhotak/build/

Cheers,
Eric

--
Dr. Eric Bodden
Software Technology Group, Technische Universität Darmstadt, Germany
Tel: +49 6151 16-5478    Fax: +49 6151 16-5410
Mailing Address: S2|02 A209, Hochschulstraße 10, 64289 Darmstadt



On 30 September 2010 01:22, Patrick Lam <plam at cs.mcgill.ca> wrote:
> On 09/29/10 18:53, Kyle Patrick Holohan wrote:
>> - Modifying the Soot source to exclude java.* and javax.* classes from
>> call graph construction,
>> - Implementing a custom call graph generation library for the application, and
>> - Modifying the Soot source to remove unused call graph attributes to
>> try to cut down on overall memory usage.
>>
>> Any other suggestions would be extremely helpful. Thanks in advance.
>
> It's going to be hard to not consider the Java libraries in call graph
> construction, which does generally take a while. However, you might want
> to consider analyzing the software with the oldest JDK library that you
> can, since it will be smaller.
>
> pat
> _______________________________________________
> 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