[PPL-devel] [GIT] ppl/ppl(MPI): New commits

[Roberto Bagnara]

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=410bc1bd073874d00ac862c758bc2ced8902dc9d
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Thu Sep 16 17:49:42 2010 +0200

    Distributed_Sparse_Matrix: fix bug in exiting_index_reducer_functor, affecting performance.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=845523ef94033ab908ea709b87c6bc511235f1af
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Thu Sep 16 11:37:56 2010 +0200

    Distributed_Sparse_Matrix: print the number of pivot operations in the destructor.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=389d8c2fc3b5d63c6253696f19d0e1c35992b217
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 21:28:11 2010 +0200

    Revert "Distributed_Sparse_Matrix: use a dynamicly-allocated array instead of a vector for the reduce operation in float_entering_index."
    
    This reverts commit 4d5f95afb8b9a662e3b6f2160a15a49c971a2972.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=73999983bdc29fa3df0638a43b10f7929104e4cc
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 19:31:58 2010 +0200

    Distributed_Sparse_Matrix: optimize exact_entering_index(), exploiting the local copies of working_cost.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=4d5f95afb8b9a662e3b6f2160a15a49c971a2972
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 19:14:28 2010 +0200

    Distributed_Sparse_Matrix: use a dynamicly-allocated array instead of a vector for the reduce operation in float_entering_index.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=d03f80ce8f87b2f8416ac1502edc933fc70bce80
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 18:32:42 2010 +0200

    Distributed_Sparse_Matrix: optimize float_entering_index(), exploiting the local copies of working_cost.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=0fc1afaecf71d837aee24723a11d494c1b6a6633
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 17:57:01 2010 +0200

    Distributed_Sparse_Matrix: store the working_cost in worker nodes, add a set_working_cost() operation and update the nodes' working_cost in get_exiting_and_pivot().

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=46e1b901664295445047a8346beb12ead9cb253e
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 17:04:49 2010 +0200

    Distributed_Sparse_Matrix: optimize exact_entering_index() for sparse working_cost rows.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=d264fd25c3eb8103d4b0ad50448fcad8b46712a7
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 16:43:27 2010 +0200

    Distributed_Sparse_Matrix: optimize float_entering_index() for sparse working_cost rows.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=9a06a4a19080f101610b63cfdb8382b1dee7f6a6
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 16:37:56 2010 +0200

    Distributed_Sparse_Matrix: optimize compute_working_cost() for sparse working_cost rows.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=fe9a6af26af0bd688effb2f4e48cfe01cc32c3ae
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 16:36:52 2010 +0200

    Sparse_Row: optimize linear_combine() for the case (coeff1 == 1).

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=7eecade09567117227f2b54339fef4e28d2d08dc
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 16:36:23 2010 +0200

    Sparse_Row: avoid creating stored zeroes in combine_needs_second().

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=87f29dca86f2b717645976464d669c2fceb6f2d6
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 16:13:30 2010 +0200

    Distributed_Sparse_Matrix: optimize compute_working_cost__common() for sparse working_cost rows.

URL:    http://www.cs.unipr.it/git/gitweb.cgi?p=ppl/ppl.git;a=commit;h=307394ac54c4fffd8f0eeb7019825ba290bf1941
Author: Marco Poletti <poletti.marco at gmail.com>
Date:   Tue Sep 14 16:03:53 2010 +0200

    MIP_Problem: some little optimizations to exploit sparse workin_cost rows.