Computer Science – Distributed – Parallel – and Cluster Computing
Scientific paper
2007-09-24
Computer Science
Distributed, Parallel, and Cluster Computing
accepted by HPC Asia 2007
Scientific paper
It is very common to use dynamic methods to detect deadlocks in MPI programs for the reason that static methods have some restrictions. To guarantee high reliability of some important MPI-based application software, a model of MPI synchronization communication is abstracted and a type of static method is devised to examine deadlocks in such modes. The model has three forms with different complexity: sequential model, single-loop model and nested-loop model. Sequential model is a base for all models. Single-loop model must be treated with a special type of equation group and nested-loop model extends the methods for the other two models. A standard Java-based software framework originated from these methods is constructed for determining whether MPI programs are free from synchronization communication deadlocks. Our practice shows the software framework is better than those tools using dynamic methods because it can dig out all synchronization communication deadlocks before an MPI-based program goes into running.
Ming-Xue Liao
Xiao-Xin He
Zhi-Hua Fan
No associations
LandOfFree
Static Deadlock Detection in MPI Synchronization Communication does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Static Deadlock Detection in MPI Synchronization Communication, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Static Deadlock Detection in MPI Synchronization Communication will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-276479