I'm not sure that something like that could exist. Open MPI developers are
developing tools and are not engaged in the analysis of the performance or
complexity of algo. Maybe you can write some?)
One thing that can help analyze complexity of algo - is a brain and
profiling tools i think (like this
Under Windows you can use MS MPI, its not a openMPI, but all works good
(with MS VS). All info you can found in MSDN, here for example:
https://msdn.microsoft.com/en-us/library/bb524831%28v=vs.85%29.aspx
If you want use openMPI - you can use build for windows (see site openMPI).
All info about fucn
You install openmpi as root, and run executable with perm of root on user.
Don't you think something wrong? (hint: when you root mpirun work)
Try install openmpi for user.
Also, you use ubuntu - binary based distro, you can use apt-get to install
openmpi:
sudo apt-get install libopenmpi-dev openmp
Maybe here you found what need:
https://www.open-mpi.org/faq/?category=tuning#setting-mca-params
also you can try run gdb via mpirun to get debug info:
mpirun -np 2 xterm -e gdb ./your_program
2015-09-28 14:43 GMT+03:00 Sven Schumacher :
> Hello,
>
> I've set up our new cluster using Infiniband
MPI_Comm comm)
>
> so you have
>
> recvbuf = 0 (!)
> recvcount = tmpVar[i*matrixSize]
>
> i guess you meant to have recvcount = blockSize
> that being said, tmpVar[i*matrixSize] is and int and it should likely be
> an double *
>
> Cheers,
>
> Gilles
>
>
>
Hi everybody!
I try implement Fox algorithm via mpi, but got some errors (see below)
Can someone explain how fix it or give a way for search.
Source code attached to letter
errors:
[estri_mobile:6337] *** An error occurred in MPI_Gather
[estri_mobile:6337] *** reported by process [1826816001,0]
[e
