On Sat, 14 May 2011, SF Markus Elfring wrote:
> > Access to the disk is a limited resource.
>
> I imagine that the processing of included (header) files can benefit from
> parallelisation despite of speed limitations from physical device accesses.
Sure, I proposed this already. But the reading itself would be limited by
the access to the disk.
> > Coccinelle has nothing to do with either petri nets or finiste state
> > machines.
>
> I guess that a reuse of such techniques would enable further optimisation
> opportunities.
Coccinelle is based on temporal logic. Model checking is not easy to
parallelize because analyzed paths can both split apart and come together.
Parallel processes might not be aware of paths coming together, and thus
duplicate work. The other techniques you mention might suffer from
similar problems.
@inproceedings{DBLP:conf/kbse/HolzmannJG08,
author = {Gerard J. Holzmann and
Rajeev Joshi and
Alex Groce},
title = {Swarm Verification},
booktitle = {ASE},
year = {2008},
pages = {1-6},
ee = {http://dx.doi.org/10.1109/ASE.2008.9},
crossref = {DBLP:conf/kbse/2008},
bibsource = {DBLP, http://dblp.uni-trier.de}
}
@proceedings{DBLP:conf/kbse/2008,
title = {23rd IEEE/ACM International Conference on Automated
Software
Engineering (ASE 2008), 15-19 September 2008, L'Aquila,
Italy},
booktitle = {ASE},
publisher = {IEEE},
year = {2008},
bibsource = {DBLP, http://dblp.uni-trier.de}
}
julia
_______________________________________________
Cocci mailing list
[email protected]
http://lists.diku.dk/mailman/listinfo/cocci
(Web access from inside DIKUs LAN only)
