[OMPI users] vader_single_copy_mechanism

2020-02-24 Thread Bennet Fauber via users
We are getting errors on our system that indicate that we should export OMPI_MCA_btl_vader_single_copy_mechanism=none Our user originally reported > This occurs for both GCC and PGI. The errors we get if we do not set this > indicate something is going wrong in our communication which uses

[OMPI users] Help with One-Sided Communication: Works in Intel MPI, Fails in Open MPI

2020-02-24 Thread Matt Thompson via users
All, My guess is this is a "I built Open MPI incorrectly" sort of issue, but I'm not sure how to fix it. Namely, I'm currently trying to get an MPI project's CI working on CircleCI using Open MPI to run some unit tests (on a single node, so need some oversubscribe). I can build everything just fin

Re: [OMPI users] Help with One-Sided Communication: Works in Intel MPI, Fails in Open MPI

2020-02-24 Thread Gabriel, Edgar via users
I am not an expert for the one-sided code in Open MPI, I wanted to comment briefly on the potential MPI -IO related item. As far as I can see, the error message “Read -1, expected 48, errno = 1” does not stem from MPI I/O, at least not from the ompio library. What file system did you use for t

Re: [OMPI users] Help with One-Sided Communication: Works in Intel MPI, Fails in Open MPI

2020-02-24 Thread Nathan Hjelm via users
The error is from btl/vader. CMA is not functioning as expected. It might work if you set btl_vader_single_copy_mechanism=none Performance will suffer though. It would be worth understanding with process_readv is failing. Can you send a simple reproducer? -Nathan > On Feb 24, 2020, at 2:59 PM

Re: [OMPI users] Help with One-Sided Communication: Works in Intel MPI, Fails in Open MPI

2020-02-24 Thread Matt Thompson via users
On Mon, Feb 24, 2020 at 4:57 PM Gabriel, Edgar wrote: > I am not an expert for the one-sided code in Open MPI, I wanted to comment > briefly on the potential MPI -IO related item. As far as I can see, the > error message > > > > “Read -1, expected 48, errno = 1” > > does not stem from MPI I/O, at

Re: [OMPI users] Help with One-Sided Communication: Works in Intel MPI, Fails in Open MPI

2020-02-24 Thread Matt Thompson via users
Nathan, The reproducer would be that code that's on the Intel website. That is what I was running. You could pull my image if you like but...since you are the genius: [root@adac3ce0cf32 ~]# mpirun --mca btl_vader_single_copy_mechanism none -np 2 ./a.out Rank 0 running on adac3ce0cf32 Rank 1 runn

Re: [OMPI users] Help with One-Sided Communication: Works in Intel MPI, Fails in Open MPI

2020-02-24 Thread Adam Simpson via users
Calls to process_vm_readv() and process_vm_writev() are disabled in the default Docker seccomp profile. You can add the docker flag --cap-add=SYS_PTRACE or better yet modify the seccomp profile so that process_vm_readv and

[OMPI users] How to use OPENMPI with different Service Level in Infiniband Virtual Lane?

2020-02-24 Thread Kihang Youn via users
Hello, I am searching the option to apply different service level(SL) in Infiniband communication. For example, In Intel MPI, the environment variable named "DAPL_IB_SERVICE_LEVEL" can change the SL. I found a runtime options "btl_openib_ib_service_level" and "UCX_IB_SL" on the FAQ pages. Pl