Next: Systematic mapping of algorithms to hardware
Up: Research Subfields
Previous: Microkernel Technology Support to Parallel Applications
Index: Contents Page


Operating Systems for Parallel Machines

Within this topic, we have worked in several areas. First, we have designed and implemented PAROS a micro-kernel operating syststem for distributed memory multiprocessors. Second, we are also working on processor scheduling policies aimed for message passing applications focussing on a temporal sharing of coarse and fine grain communications. The application of these can be done both in massively parallel and workstation networks. Finally, we are working on a parallel/distributed file system with a cooperative cache aimed for parallel machines or networks of workstations.

We have also developped a set of analisys and visualization tools which ease the study of the above topics. All these tools are part of the DiP project.

PAROS

  • Cristina Pujol, Jesús Labarta, Luis Gregoris, and Sergi Girona. Paros Kernel Free Distribution v1.2 User Manual . Technical Report UPC-CEPBA-94-07, European Center for Parallelism of Barcelona (UPC), 1994. Also published as UPC-DAC-94-12.

  • Jesús Labarta, Judit Giménez, Cristina Pujol, and Teodor Jové. Multiprogramming parallel applications on the Paros operating system kernel. In Euromicro Workshop on Parallel and Distributed Processing , pp. 474-480, Gran Canaria (Spain), January 1993.

  • Jesús Labarta, Cristina Pujol, Teodor Jové, and José I. Navarro. Paros: Operating System Kernel for Distributed Memory Multiprocessors. In Parallel Computing and Transputer Applications (PACTA'92) , pp. 673-682, Barcelona (Spain), September 1992.

Processor Scheduling

  • Jesús Labarta, Sergi Girona, and Toni Cortés. Analyzing scheduling policies using Dimemas. In Environments and Tools For Parallel Scientific Computing III (ETPSC III) , Faverges de la Tour (France), August 1996.

  • Sergi Girona, Toni Cortés, Jesús Labarta, Vincent Pillet, Andrés Pérez, and Elena García. Resource Management Policies, 1994. Deliverable OPS4A of the project "Basic research APPARC".

Parallel/Distributed File Systems

  • Toni Cortés, Sergi Girona, and Jesús Labarta. PACA: A Cooperative File System Cache for Parallel Machines. In 2nd International Euro-Par Conference (Euro-Par'96) , pp. 477-486, Lyon (France), August 1996. Lecture Notes in Computer Science #1123.

  • Toni Cortés, Sergi Girona, and Jesús Labarta. PACA: a Cooperative File System Cache. In VII Jornadas de Paralelismo , pp. 245-258, Santiago de Compostela (Spain), September 1996.

Analisys and Visualization Tools

  • Jesús Labarta, Sergi Girona, Vincent Pillet, Toni Cortés, and Luis Gregoris. DiP: A Parallel Program Development Environment. In 2nd International Euro-Par Conference (Euro-Par'96) , pp. 665-674, Lyon (France), August 1996. Lecture Notes in Computer Science #1124.

  • Vincent Pillet, Jesús Labarta, Toni Cortés, and Sergi Girona. PARAVER: PARAllel Visualization and Events Representation. In The 18th Technical Meeting of WoTUG , pp. 17-31, Amsterdam (The Netherlands), April 1995.

We are also working in the interaction between the operating system and the compiler. The compiler can provide the operating system with information (obtained from an static analysis of the code) that is not offered by the user and can not be obtained at run-time by the operating system. This information may be useful to effectively exploit both parallelism in an application and multiprogrammed execution in tightly-coupled systems.

  • Xavier Martorell, Jesús Labarta, José I. Navarro, and Eduard Ayguadé. A Library Implementation of the Nano-Threads Programming Model. In 2nd International Euro-Par Conference (Euro-Par'96) , pp. 644-649, Lyon (France), August 1996. Lecture Notes in Computer Science #1124.

Next: Systematic mapping of algorithms to hardware
Up: Research Subfields
Previous: Microkernel Technology Support to Parallel Applications
Index: Contents Page


 

Inici | Presentació | Docència | Recerca | Centres de Recerca | Novetats Inici

Darrera actualització: 12 de novembre del 2004
Copyright © 2000-2005 Departament d'Arquitectura de Computadors