Using MPI: portable parallel programming with the message-passing interface
Using MPI: portable parallel programming with the message-passing interface
Static Scheduling Using Task Replication for LogP and BSP Models
Euro-Par '98 Proceedings of the 4th International Euro-Par Conference on Parallel Processing
Communicating Functional Agents and Their Application to Graphical User Interfaces
Proceedings of the Second International Andrei Ershov Memorial Conference on Perspectives of System Informatics
An Active Layer Extension to MPI
Proceedings of the 5th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Fuctional Message Passing with OPAL-MPI
Proceedings of the 5th European PVM/MPI Users' Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface
Hi-index | 0.00 |
This paper describes the implementation of a thread communication library on top of MPI. It allows light-weight threads to communicate with each other both locally between threads within the same process as well as globally between threads on different processors. The interface is similar to MPI with the use of thread identifiers instead of processor ranks. Problems occur in the implementation of global communication operations. Due to limited tag space we are not able to specify source and target thread identifiers in a call to MPI_Recv. As a result we may receive messages from the wrong thread which has to be resolved explicitly.