You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Why is that there is an uninitialized mutex error with more than one MPI process per node?
I am running on an x86_64 system with the Intel 16 toolchain.
$ mpirun -n 1 -env MAD_NUM_THREADS=60 ./test_world.mpi
MADNESS runtime initialized with 59 threads in the pool and affinity -1 -1 -1
$ mpirun -n 4 -env MAD_NUM_THREADS=60 ./test_world.mpi
MADNESS runtime initialized with 59 threads in the pool and affinity -1 -1 -1
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 1
!! MADNESS ERROR: Mutex::lock() failed acquiring mutex
!! MADNESS ERROR: Mutex error EINVAL
!! MADNESS ERROR: The value specified by mutex does not refer to an initialized mutex object.
!! MADNESS ERROR: Mutex::lock() failed acquiring mutex
!! MADNESS ERROR: Mutex error EINVAL
!! MADNESS ERROR: The value specified by mutex does not refer to an initialized mutex object.
The text was updated successfully, but these errors were encountered:
Curious. Probably not directly due to multiple processes per node but a
timing issue. I will try to reproduce.
On Nov 6, 2015 7:14 PM, "Jeff Hammond" notifications@github.com wrote:
Why is that there is an uninitialized mutex error with more than one MPI
process per node?
I am running on an x86_64 system with the Intel 16 toolchain.
$ mpirun -n 1 -env MAD_NUM_THREADS=60 ./test_world.mpi
MADNESS runtime initialized with 59 threads in the pool and affinity -1 -1 -1
$ mpirun -n 4 -env MAD_NUM_THREADS=60 ./test_world.mpi
MADNESS runtime initialized with 59 threads in the pool and affinity -1 -1 -1
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 1
!! MADNESS ERROR: Mutex::lock() failed acquiring mutex
!! MADNESS ERROR: Mutex error EINVAL
!! MADNESS ERROR: The value specified by mutex does not refer to an initialized mutex object.
!! MADNESS ERROR: Mutex::lock() failed acquiring mutex
!! MADNESS ERROR: Mutex error EINVAL
!! MADNESS ERROR: The value specified by mutex does not refer to an initialized mutex object.
—
Reply to this email directly or view it on GitHub #163.
Why is that there is an uninitialized mutex error with more than one MPI process per node?
I am running on an
x86_64
system with the Intel 16 toolchain.The text was updated successfully, but these errors were encountered: