[m-rev.] for (postcommit) review: Tracing allocated regions

Quan Phan quan.phan at cs.kuleuven.be
Wed Apr 22 19:24:17 AEST 2009


On Wed, Apr 22, 2009 at 10:44:07AM +1000, Ralph Becket wrote:
> Quan Phan, Tuesday, 21 April 2009:
> > Hi,
> > 
> > I've committed this diff.
> 
> Hi Quan,
> 
> do you have any performance results for RBMM?  I'd be interested to see
> them.

Hi Ralph,

We reported them in a paper at ISMM 08 titled 'Runtime support for region-based
memory management in Mercury'.

The broader and more detailed results and the benchmarks we used are in
cvs papers/rbmm/benchmarks.

If you have questions while looking at them, please ask.

Overall the results are quite promising, we got speedup for almost all
of the benchmarks. In some cases the gain is rather big.

Regards,
Quan.
> 
> Cheers,
> -- Ralph
> --------------------------------------------------------------------------
> mercury-reviews mailing list
> Post messages to:       mercury-reviews at csse.unimelb.edu.au
> Administrative Queries: owner-mercury-reviews at csse.unimelb.edu.au
> Subscriptions:          mercury-reviews-request at csse.unimelb.edu.au
> --------------------------------------------------------------------------
--------------------------------------------------------------------------
mercury-reviews mailing list
Post messages to:       mercury-reviews at csse.unimelb.edu.au
Administrative Queries: owner-mercury-reviews at csse.unimelb.edu.au
Subscriptions:          mercury-reviews-request at csse.unimelb.edu.au
--------------------------------------------------------------------------



More information about the reviews mailing list