| /*---------------------------------------------------------------------------*\ | |
| ========= | | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox | |
| \\ / O peration | Website: https://openfoam.org | |
| \\ / A nd | Copyright (C) 2011-2022 OpenFOAM Foundation | |
| \\/ M anipulation | | |
| ------------------------------------------------------------------------------- | |
| License | |
| This file is part of OpenFOAM. | |
| OpenFOAM is free software: you can redistribute it and/or modify it | |
| under the terms of the GNU General Public License as published by | |
| the Free Software Foundation, either version 3 of the License, or | |
| (at your option) any later version. | |
| OpenFOAM is distributed in the hope that it will be useful, but WITHOUT | |
| ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
| FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
| for more details. | |
| You should have received a copy of the GNU General Public License | |
| along with OpenFOAM. If not, see <http://www.gnu.org/licenses/>. | |
| Class | |
| Foam::commSchedule | |
| Description | |
| Determines the order in which a set of processors should communicate | |
| with one another. | |
| The communication order should | |
| - have maximum overlap | |
| - allow blocking communication without deadlock | |
| Does a very simple scheduling which assumes same time for all operations. | |
| After construction: | |
| - schedule() gives the order in which the input communication should occur | |
| - procSchedule()[proci] gives per proci | |
| Does not care whether 'talking' is first send, second receive or maybe | |
| full swap. This is all responsibility of caller. See ProcessorTopology | |
| class for use in scheduling processor boundary swaps. | |
| SourceFiles | |
| commSchedule.C | |
| \*---------------------------------------------------------------------------*/ | |
| // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // | |
| namespace Foam | |
| { | |
| /*---------------------------------------------------------------------------*\ | |
| Class commSchedule Declaration | |
| \*---------------------------------------------------------------------------*/ | |
| class commSchedule | |
| { | |
| // Private Data | |
| //- Order in which input communication has been scheduled | |
| labelList schedule_; | |
| //- Per processor the order in which communication has been scheduled | |
| labelListList procSchedule_; | |
| // Private Member Functions | |
| //- Count the number of outstanding communications for a single | |
| // processor | |
| label outstandingComms(const labelList&, DynamicList<label>&) const; | |
| public: | |
| ClassName("commSchedule"); | |
| // Constructors | |
| //- Construct from wanted communication. Wanted communication is between | |
| // two processors. Can be a one-way communication or | |
| // two-way communication, that is up to the caller. This class just | |
| // determines an order for it such that any processor is only talking | |
| // to one other at a time. After construction: | |
| // - schedule is the order in which comms is done. | |
| // - procSchedule[proci] is for proci the order in which comms is done. | |
| commSchedule(const label nProcs, const List<labelPair>& comms); | |
| // Member Functions | |
| //- Order in which comms is scheduled | |
| const labelList& schedule() const | |
| { | |
| return schedule_; | |
| } | |
| //- Per processor the order in which communication has been scheduled | |
| const labelListList& procSchedule() const | |
| { | |
| return procSchedule_; | |
| } | |
| }; | |
| // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // | |
| } // End namespace Foam | |
| // * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * // | |
| // ************************************************************************* // | |