svadev AT lists.siebelschool.illinois.edu

Subject: Svadev mailing list

List archive

[svadev] Call graph in C++

From: "Arushi Aggarwal" <arushi987 AT gmail.com>
To: <svadev AT cs.uiuc.edu>
Subject: [svadev] Call graph in C++
Date: Sun, 17 Oct 2010 03:18:56 -0500
List-archive: <http://lists.cs.uiuc.edu/pipermail/svadev>
List-id: <svadev.cs.uiuc.edu>

This refers to one of the reasons why 252.eon possibly fails with poolallocation.

In the reduced example attached, the call site in DerivedClass1::MyFunction() is never resolved. Similiarly the call site for MyFunction in main, is not resolved, either.

The call graph hence is wrong. It does manage to get the call targets for the call site in main, but for the call site in DerivedClass1::MyFunction(), the call graph has no information.

I have attached the DSGraph for main. The call site is not resolved, because the callee node for the function is incomplete, which is because the @ob node is incomplete(because it is a parameter at that call site).

Seems like a vicious cycle. What am I missing here?

Arushi

Attachment: inheritance.cpp
Description: Binary data

Attachment: bu.main.dot
Description: MS-Word document

[svadev] Call graph in C++, Arushi Aggarwal, 10/17/2010