Re: [Javapathfinder-devel] Parallel version of JPF.

[Peter C. M. <Pet...@na...>]

the problem would be synchronization bottlenecks. If you have  
distributed searches and a single state repository for matching/ 
restoring, this would require an amount of synchronization that would  
probably slow things down considerably. What could work is to allow  
some redundancy, i.e. let nodes proceed independently of each other,  
but IMHO this requires some knowledge about the shape of the state  
space, to decide which branch points should spawn remote searches.

I think partitioning into execution, state management and report  
generation is more promising in the near to mid future. For example,  
one of the major memory sinks is simply storing traces (the Transition  
Steps). I think that could easily be factorized out into a different  
process/node, not requiring much synchronization

-- Peter 


On Oct 9, 2008, at 8:21 AM, Mateusz Ujma wrote:

> Hi,
>
>    Would it be possible to write parallel version of JPF that will use
> MPI ?
>
> Yours sincerely
>
>       Mateusz Ujma
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's  
> challenge
> Build the coolest Linux based applications with Moblin SDK & win  
> great prizes
> Grand prize is a trip for two to an Open Source event anywhere in  
> the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Javapathfinder-devel mailing list
> Jav...@li...
> https://lists.sourceforge.net/lists/listinfo/javapathfinder-devel