Those attached are configure.log/make.log from a MPI build in Fedora 34
x86_64 where the error below occurred.

On 9/12/21 19:18, Antonio T. sagitter wrote:
> Okay. I will try to set correctly the DATAFILESPATH options.
>
> I see even this error:
>
> not ok
> vec_is_sf_tutorials-ex1_4+sf_window_sync-fence_sf_window_flavor-create #
> Error code: 68
>
> #    PetscSF Object: 4 MPI processes
>
> #      type: window
>
> #      [0] Number of roots=3, leaves=2, remote ranks=2
>
> #      [0] 0 <- (3,1)
>
> #      [0] 1 <- (1,0)
>
> #      [1] Number of roots=2, leaves=3, remote ranks=2
>
> #      [1] 0 <- (0,1)
>
> #      [1] 1 <- (2,0)
>
> #      [1] 2 <- (0,2)
>
> #      [2] Number of roots=2, leaves=3, remote ranks=3
>
> #      [2] 0 <- (1,1)
>
> #      [2] 1 <- (3,0)
>
> #      [2] 2 <- (0,2)
>
> #      [3] Number of roots=2, leaves=3, remote ranks=2
>
> #      [3] 0 <- (2,1)
>
> #      [3] 1 <- (0,0)
>
> #      [3] 2 <- (0,2)
>
> #      [0] Roots referenced by my leaves, by rank
>
> #      [0] 1: 1 edges
>
> #      [0]    1 <- 0
>
> #      [0] 3: 1 edges
>
> #      [0]    0 <- 1
>
> #      [1] Roots referenced by my leaves, by rank
>
> #      [1] 0: 2 edges
>
> #      [1]    0 <- 1
>
> #      [1]    2 <- 2
>
> #      [1] 2: 1 edges
>
> #      [1]    1 <- 0
>
> #      [2] Roots referenced by my leaves, by rank
>
> #      [2] 0: 1 edges
>
> #      [2]    2 <- 2
>
> #      [2] 1: 1 edges
>
> #      [2]    0 <- 1
>
> #      [2] 3: 1 edges
>
> #      [2]    1 <- 0
>
> #      [3] Roots referenced by my leaves, by rank
>
> #      [3] 0: 2 edges
>
> #      [3]    1 <- 0
>
> #      [3]    2 <- 2
>
> #      [3] 2: 1 edges
>
> #      [3]    0 <- 1
>
> #      current flavor=CREATE synchronization=FENCE MultiSF sort=rank-order
>
> #        current info=MPI_INFO_NULL
>
> #    [buildhw-x86-09:1135574] *** An error occurred in MPI_Accumulate
>
> #    [buildhw-x86-09:1135574] *** reported by process [3562602497,3]
>
> #    [buildhw-x86-09:1135574] *** on win rdma window 4
>
> #    [buildhw-x86-09:1135574] *** MPI_ERR_RMA_RANGE: invalid RMA address
> range
>
> #    [buildhw-x86-09:1135574] *** MPI_ERRORS_ARE_FATAL (processes in
> this win will now abort,
>
> #    [buildhw-x86-09:1135574] ***    and potentially your MPI job)
>
> #    [buildhw-x86-09.iad2.fedoraproject.org:1135567] 3 more processes
> have sent help message help-mpi-errors.txt / mpi_errors_are_fatal
>
> #    [buildhw-x86-09.iad2.fedoraproject.org:1135567] Set MCA parameter
> "orte_base_help_aggregate" to 0 to see all help / error messages
>
> Looks like an error related to OpenMPI-4*:
> https://github.com/open-mpi/ompi/issues/6374
> 

--
---
Antonio Trande
Fedora Project
mailto: sagitter@fedoraproject.org
GPG key: 0x29FBC85D7A51CC2F
GPG key server: https://keyserver1.pgp.com/