[Soot-list] Fwd: Re: May Happen In Parallel -- Please help

[Richard L. Halpert]

Forwarding my response to Bharath to the list so others can see...

-Richard
---------- Forwarded message ----------
From: "Richard L. Halpert" <richardlhalpert at gmail.com>
Date: Mar 11, 2012 8:30 AM
Subject: Re: May Happen In Parallel -- Please help
To: "Bharath Reddy A." <bharath270392 at gmail.com>

The analysis is based on a run-once, run-many analysis, and it makes use of
the call graph generated by Spark.  It completely ignores
locks/synchronization, and computes a conservative solution.

The analysis is described in detail in our PACT2007 paper, and in
excruciating detail in my Masters thesis:

http://www.sable.mcgill.ca/publications/papers/#pact2007

http://www.sable.mcgill.ca/publications/thesis/#richardsMastersThesis

-Richard
On Mar 11, 2012 12:10 AM, "Bharath Reddy A." <bharath270392 at gmail.com>
wrote:

> Hi Richard,
>              I am currently working on Soot and trying to understand how
> MHP information is exactly computed in SynchObliviousMhpAnalysis.java . I
> saw in in the mailing list(
> http://www.sable.mcgill.ca/pipermail/soot-list/2008-October/002031.html)
> that call to  MhpAnalysis which implements Nauvomich's Algorithm has been
> removed. Can you please explain what MHP information is now computed in
> SynchObliviousMhpAnalysis.java .
>
> Thanks in advance
>
> --
> Bharath Reddy A.
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.mcgill.ca/pipermail/soot-list/attachments/20120312/6b8daef4/attachment.html