Mailing List Archive

Implementation of perlesec
Hi everyone, I'm a phd student and I'm working on dynamic program analysis and in particular information flow analysis.
I'm interested to understand how perlsec (aka the taint mode of perl) is implemented.
I've tried to find resources online but could only find user guides which describe what perlsec achieves but not how it does so.
I've spend some time in the source code but, being unfamiliar with perl, I found it hard to reverse engineer the design of perlsec.
So here is my questions to you guys:

* Is there online materials which describe the implementation of perlsec?
* If not, can you guys help me figure out the technique underlying perlsec?
- Does it relies on reflection?
- Does it relies on program instrumentation?
- Does it requires modification of the internals of the perl's interpreter?

Kind regards,
Laurent