Re: ReCrash

11 views
Skip to first unread message

Shay Artzi

unread,
Jun 5, 2008, 10:34:07 AM6/5/08
to ima...@cs.technion.ac.il, ar...@csail.mit.edu, hun...@csail.mit.edu, mer...@csail.mit.edu, rec...@googlegroups.com
Hi Itay-
> Anyway, the lecture's topic is ReCrash - very impressive, by the way
> - and I want to know how to pass optimization flags to the
> instrumentation engine (In my lecture I want to show the effect of
> different copying strategies on the generated test classes).
> Otherwise, if this things are hard-coded, could you send me a link to
> the source code?
The differences might not be visible in the generated tests, since any
amount of information is read from the stored object database and is not
visible in the generated tests. However, running the tests might expose
some difference as some of them might not produce the crash.

The parameters can be supplied as parameters to Transform:
transform <src> <desc> <options>
-verbose (prints information)
-skip_main (do not replace main)
-use_clone (shallow clone)
-use_shallow_clone (used field + shallow clone)
-use_serialization (deep copy)
-pname <program name> (Program name to display in test cases)
-mut_data <mutability info> (Mutability info in the format of Pidasa's
output)
-second_chance <second chance data file>

The default copying is reference copy. We are currently working on a new
distribution that will be available soon (with source) and the
parameters will better match the options in the paper.


Shay

Reply all
Reply to author
Forward
0 new messages