Re: [OMPI users] mpirun w/ enable-mpi-threads spinning up cputime when app path is invalid

[Ralph H Castain]

Sorry - went to one of your links to get that info.

We know OMPI 1.2.x isn't thread safe. This is unfortunately another example
of it. Hopefully, 1.3 will be better.

Ralph



On 7/2/08 11:01 AM, "Ralph H Castain"  wrote:

> Out of curiosity - what version of OMPI are you using?
> 
> 
> On 7/2/08 10:46 AM, "Steve Johnson"  wrote:
> 
>> If mpirun is given an application that isn't in the PATH, then instead of
>> exiting it prints the error that it failed to find the executable and then
>> proceeds spins up cpu time.  strace shows an endless stream of sched_yield().
>> 
>> For example, if "blah" doesn't exist:
>> mpirun -np 16 blah
>> Ditto if ./blah doesn't exist and mpirun is called as
>> mpirun -np 16 ./blah
>> 
>> OS: CentOS 5.1
>> Kernel: 2.6.18-92.1.1.el5.centos.plus
>> Arch: x86_64
>> glibc/pthread: glibc-2.5-18.el5_1.1
>> GCC: 4.1.2-14.el5
>> 
>> CC=gcc
>> CXX=g++
>> F77=gfortran
>> FC=gfortran
>> ./configure --with-tm --prefix=$HOME/openmpi --libdir=$HOME/openmpi/lib64
>> --enable-mpi-threads
>> 
>> A qsig -s 15 will terminate the mpirun processes.
>> 
>> ompi_info is at http://isc.tamu.edu/~steve/ompi_info.txt
>> config.log.bz is at http://isc.tamu.edu/~steve/ompi_config.log.bz2
>> 
>> Also confirmed this on openSUSE 10.2, 2.6.18.8-0.9-default, x86_64,
>> glibc-2.5-34.7, gcc-4.1.3-29.
>> 
>> // Steve
>> ___
>> users mailing list
>> us...@open-mpi.org
>> http://www.open-mpi.org/mailman/listinfo.cgi/users
> 
> 
> ___
> users mailing list
> us...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/users

Re: [OMPI users] mpirun w/ enable-mpi-threads spinning up cputime when app path is invalid

[Ralph H Castain]

Out of curiosity - what version of OMPI are you using?


On 7/2/08 10:46 AM, "Steve Johnson"  wrote:

> If mpirun is given an application that isn't in the PATH, then instead of
> exiting it prints the error that it failed to find the executable and then
> proceeds spins up cpu time.  strace shows an endless stream of sched_yield().
> 
> For example, if "blah" doesn't exist:
> mpirun -np 16 blah
> Ditto if ./blah doesn't exist and mpirun is called as
> mpirun -np 16 ./blah
> 
> OS: CentOS 5.1
> Kernel: 2.6.18-92.1.1.el5.centos.plus
> Arch: x86_64
> glibc/pthread: glibc-2.5-18.el5_1.1
> GCC: 4.1.2-14.el5
> 
> CC=gcc
> CXX=g++
> F77=gfortran
> FC=gfortran
> ./configure --with-tm --prefix=$HOME/openmpi --libdir=$HOME/openmpi/lib64
> --enable-mpi-threads
> 
> A qsig -s 15 will terminate the mpirun processes.
> 
> ompi_info is at http://isc.tamu.edu/~steve/ompi_info.txt
> config.log.bz is at http://isc.tamu.edu/~steve/ompi_config.log.bz2
> 
> Also confirmed this on openSUSE 10.2, 2.6.18.8-0.9-default, x86_64,
> glibc-2.5-34.7, gcc-4.1.3-29.
> 
> // Steve
> ___
> users mailing list
> us...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/users

[OMPI users] mpirun w/ enable-mpi-threads spinning up cputime when app path is invalid

[Steve Johnson]

If mpirun is given an application that isn't in the PATH, then instead of 
exiting it prints the error that it failed to find the executable and then 
proceeds spins up cpu time.  strace shows an endless stream of sched_yield().


For example, if "blah" doesn't exist:
mpirun -np 16 blah
Ditto if ./blah doesn't exist and mpirun is called as
mpirun -np 16 ./blah

OS: CentOS 5.1
Kernel: 2.6.18-92.1.1.el5.centos.plus
Arch: x86_64
glibc/pthread: glibc-2.5-18.el5_1.1
GCC: 4.1.2-14.el5

CC=gcc
CXX=g++
F77=gfortran
FC=gfortran
./configure --with-tm --prefix=$HOME/openmpi --libdir=$HOME/openmpi/lib64 
--enable-mpi-threads


A qsig -s 15 will terminate the mpirun processes.

ompi_info is at http://isc.tamu.edu/~steve/ompi_info.txt
config.log.bz is at http://isc.tamu.edu/~steve/ompi_config.log.bz2

Also confirmed this on openSUSE 10.2, 2.6.18.8-0.9-default, x86_64, 
glibc-2.5-34.7, gcc-4.1.3-29.


// Steve