FPMPI-2

Home

Docs: Routines

Download
Features
Documentation
- Routines
- Installation
- Changes
- Troubleshooting
- Bug Reporting
- FAQ
- Copyright
Developers

All routines listed have the number of times called as well as time spent during the call included in the profiled data, as well as the information listed:

Message data: Message length statistics with histograms of the number of calls with a particular message length and percentage of data with a particular message length
Synchronization data: Synchronization time as well as Message data
Communicator data: Information related to usage with MPI_COMM_WORLD and MPI_COMM_SELF

MPI_Allgather: Synchronization
MPI_Allgatherv: Synchronization
MPI_Allreduce: Synchronization
MPI_Alltoall: Synchronization
MPI_Alltoallv: Synchronization
MPI_Barrier
MPI_Bcast: Message
MPI_Comm_dup: Communicator
MPI_Comm_split: Communicator
MPI_Gather: Message
MPI_Gatherv: Message
MPI_Pack: Message
MPI_Unpack: Message
MPI_Reduce: Synchronization
MPI_Reduce_scatter: Synchronization
MPI_Scan: Message
MPI_Scatter: Message
MPI_Scatterv: Message
MPI_Recv: Message
MPI_Send: Message
MPI_Sendrecv: Message
MPI_Sendrecv_replace: Message
MPI_Bsend: Message
MPI_Ibsend: Message
MPI_Irecv: Message
MPI_Irsend: Message
MPI_Rsend: Message
MPI_Ssend: Message
MPI_Isend: Message
MPI_Issend: Message
MPI_Wait: Message
MPI_Waitall: Message
MPI_Waitany: Message
MPI_Waitsome: Message
MPI_Test: Message
MPI_Testall: Message
MPI_Testany: Message
MPI_Testsome: Message
MPI_Iprobe
MPI_Probe

And of course there is setup occurring in MPI_Init and output is generated in MPI_Finalize.