svadev AT lists.siebelschool.illinois.edu

Subject: Svadev mailing list

List archive

Re: [svadev] Merging Information from Globals into Graphs for Individual functions

From: John Criswell <criswell AT illinois.edu>
To: svadev AT cs.uiuc.edu
Subject: Re: [svadev] Merging Information from Globals into Graphs for Individual functions
Date: Tue, 19 Oct 2010 10:24:36 -0500
List-archive: <http://lists.cs.uiuc.edu/pipermail/svadev>
List-id: <svadev.cs.uiuc.edu>
Organization: University of Illinois

On 10/19/10 12:23 AM, Arushi Aggarwal wrote:

In the given C++ example, the indirect function call site in main is resolved, after BU as the information about the object is merged back into main's graph.
The same call sequence in func, does not get resolved, as the information from the globals graph, is never merged back in.

It seems like extra overhead, to merge the information from the globals graph into each of the graphs after BU. But it seems to be needed for correct call graph construction.

Views?

Is the DSCallGraph correct? It may be correct behavior for the indirect call site to not be resolved as long as call graph construction is correct.

-- John T.

arushi

[svadev] Merging Information from Globals into Graphs for Individual functions, Arushi Aggarwal, 10/19/2010
- Re: [svadev] Merging Information from Globals into Graphs for Individual functions, John Criswell, 10/19/2010
  - Re: [svadev] Merging Information from Globals into Graphs for Individual functions, Arushi Aggarwal, 10/19/2010