IMB-MPI1-GPU treats -h / --help as an invalid benchmark name and starts running benchmarks instead of printing help and exiting
Nobody has claimed this yet.
Assessment
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Newbie friendliness
- 65/100
Research direction
Start at the IMB-MPI1-GPU command-line entry point and reproduce the reported behavior with -h or --help. The fix is done when either option prints help and exits without reporting an invalid benchmark or running benchmarks.
Written by the indexing model from the issue text.
Description
qa_sharp@r-hpc-gpu01:~$ $HPCX_MPI_TESTS_DIR/imb/IMB-MPI1-GPU -h
#----------------------------------------------------------------
Intel(R) MPI Benchmarks 2021.10, MPI-1 part (GPU)
#----------------------------------------------------------------
Date : Mon Apr 20 13:47:06 2026
Machine : x86_64
System : Linux
Release : 6.12.0-55.9.1.el10_0.x86_64
Version : #1 SMP PREEMPT_DYNAMIC Tue Mar 25 09:14:09 EDT 2025
MPI Version : 3.1
MPI Thread Environment:
Calling sequence was:
/auto/qa_smg/IB/qa_sharp/HPCX_2.50_TP4/hpcx-v2.50-gcc-doca_ofed-redhat10-cuda13-x86_64/ompi5/tests/imb/IMB-MPI1-GPU -h
Minimum message length in bytes: 0
Maximum message length in bytes: 4194304
MPI_Datatype : MPI_BYTE
MPI_Datatype for reductions : MPI_FLOAT
MPI_Op : MPI_SUM
List of Benchmarks to run:
PingPong
PingPing
Sendrecv
Exchange
Allreduce
Reduce
Reduce_local
Reduce_scatter
Reduce_scatter_block
Allgather
Allgatherv
Gather
Gatherv
Scatter
Scatterv
Alltoall
Alltoallv
Bcast
Barrier
Invalid benchmark name -h
!! Benchmark PingPong invalid for 1 processes !!
!! Benchmark PingPing invalid for 1 processes !!
#-----------------------------------------------------------------------------
Benchmarking Sendrecv
#processes = 1
#-----------------------------------------------------------------------------
#bytes #repetitions t_min[usec] t_max[usec] t_avg[usec] Mbytes/sec
0 1000 0.14 0.14 0.14 0.00
1 1000 6960.38 6960.38 6960.38 0.00
2 1000 6960.27 6960.27 6960.27 0.00
- Dominant language
- C
- Stars
- 189
- Forks
- 70
- PR merge metrics
- No merged PRs in 30d
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from intel/mpi-benchmarks
-
Difficulty 4/5 3-5 days Newbie friendliness 35/100
intel/mpi-benchmarks#71 · 2 comments ·
-
Building out of the box IMB 2021.10 with out of the box oneapi 2025.2.0 with set compilers fails Open
Difficulty 3/5 1-2 days Newbie friendliness 35/100
intel/mpi-benchmarks#67 · 1 comment ·
-
IMB-NBC and IMB-IO Open
Difficulty 4/5 3-5 days Newbie friendliness 20/100
intel/mpi-benchmarks#51 · 3 comments ·
-
Hangs in some IMB-RMA tests when I_MPI_ROOT is set and when using a non-One API-provided libfabric Open
Difficulty 4/5 3-5 days Newbie friendliness 35/100
intel/mpi-benchmarks#48 · 4 comments ·
-
Difficulty 2/5 1-3 hours Newbie friendliness 35/100
intel/mpi-benchmarks#42 · 2 comments ·
All issues in intel/mpi-benchmarks
Similar issues
-
level/task module/gcp type/bug
Difficulty 2/5 1-3 hours Newbie friendliness 85/100
-
docs
Difficulty 1/5 Under an hour Newbie friendliness 85/100
-
P3 sonic-vpp
Difficulty 2/5 1-3 hours Newbie friendliness 88/100
sonic-net/sonic-buildimage#29662 ·
-
Difficulty 1/5 Under an hour Newbie friendliness 94/100
spack/spack-packages#6586 ·
-
category:port-update
Difficulty 2/5 1-3 hours Newbie friendliness 72/100