[mpich2-dev] Changes in host list handling

[Pavan Balaji]

It's hard to tell what the issue is, but most of the changes are in the 
stdout/stderr forwarding part when "-l" is given. Did you try without that?

  -- Pavan

On 10/27/2010 10:13 AM, Justin M Wozniak wrote:
> Hello
>   	I have a patch I'm using that makes some changes to Hydra
> bootstrapping but I'm having trouble applying it successfully in the
> current trunk.  My patch is based on r7255.  I'm trying to figure out what
> changed.
>
> Here's an example that doesn't rely on my patch to show different
> behavior.  This test program simply copies an input file to an output
> file:
>
> In r7255:
>
> mpiexec -l -n 2 -bootstrap fork -hosts A,B ./mpi-cp input.txt output.txt
> [1] recving
> [0] read: input.txt
> [0] size: 6
> [1] recvd
> [1] size: 6
> [1] write: output.txt
>
> However, using trunk, I get:
>
> mpiexec -l -n 2 -bootstrap fork -hosts A,B ./mpi-cp input.txt output.txt
> [1] recving
> [0] read: input.txt
> [0] size: 6
> [0] Fatal error in MPI_Send: Other MPI error, error stack:
> [0] MPI_Send(173)..........................: MPI_Send(buf=0x7ffff85bb534,
> count=1, MPI_INT, dest=1, tag=1, MPI_COMM_WORLD) failed
> [0] MPIDI_EagerContigShortSend(262)........: failure occurred while
> attempting to send an eager message
> [0] MPID_nem_tcp_iStartContigMsg(304)......:
> [0] MPID_nem_tcp_connect(864)..............:
> [0] MPID_nem_tcp_get_addr_port_from_bc(484): Missing ifname or invalid
> host/port description in business card
> APPLICATION TERMINATED WITH THE EXIT STRING: Hangup (signal 1)
>
> Does anyone have an idea on what the relevant change is?
>
>   	Thanks
>

-- 
Pavan Balaji
http://www.mcs.anl.gov/~balaji