It looks like it was a bug report for MVAPICH-1, not MVAPICH2, so it shouldn't apply to MPICH2. Rajeev On Feb 15, 2012, at 2:51 PM, Pavan Balaji wrote:

[Moving to mpich2-core from mpich2-maint, which is a deprecated list (and should probably be deleted)]

Not quite sure what you mean whether it applies to us -- mpirun_rsh can be used to launch MPICH2 jobs, so it does impact us in some way. But we don't distribute it, so I guess we don't care.

-- Pavan

On 02/15/2012 11:28 AM, Rajeev Thakur wrote:

...

Does this apply to us or only MVAPICH?

Rajeev

Begin forwarded message:

...

From: "PHILIPP, Axel, Dr."<[email protected]> Date: February 15, 2012 2:19:47 AM CST To:<[email protected]>,<[email protected]> Subject: [mpich2-maint] mpirun_rsh does not work with newer linux kernels

After upgrading to kernel 2.6.32.49-0.3-default (SLES11 SP1) mpirun_rsh and mpispawn fail with bind errors. Another problem arises with killing the remote processes because now rsh and ssh use the user's shell instead of /bin/sh to execute commands. Therefor the bash only syntax fails if the user's shell is ksh. The following patches (applied to package mvapich-1.2.0-3635) resolve both problems:

--- mpispawn.c.ori 2010-01-28 16:00:51.000000000 +0100 +++ mpispawn.c 2012-02-14 17:15:29.802596012 +0100 @@ -380,6 +380,7 @@ }

mc_sockaddr.sin_addr.s_addr = INADDR_ANY; + mc_sockaddr.sin_family = AF_INET; mc_sockaddr.sin_port = 0;

if(bind(*mc_socket, (struct sockaddr *)&mc_sockaddr, mc_sockaddr_len)< 0) { @@ -461,6 +462,7 @@ exit (EXIT_FAILURE); } c_sockaddr.sin_addr.s_addr = INADDR_ANY; + c_sockaddr.sin_family = AF_INET; c_sockaddr.sin_port = 0;

if (bind (c_socket, (struct sockaddr *)&c_sockaddr, sockaddr_len) < 0) {

--- mpirun_rsh.c.ori 2010-01-28 16:00:51.000000000 +0100 +++ mpirun_rsh.c 2012-02-14 17:08:59.074910138 +0100 @@ -488,6 +488,7 @@ exit(EXIT_FAILURE); } sockaddr.sin_addr.s_addr = INADDR_ANY; + sockaddr.sin_family = AF_INET; sockaddr.sin_port = 0; if (bind(s, (struct sockaddr *)&sockaddr, sockaddr_len)< 0) { perror("bind"); @@ -1484,7 +1485,7 @@ strcat(kill_cmd, tmp); }

- strcat(kill_cmd, ">&/dev/null"); + strcat(kill_cmd, ">/dev/null 2>&1");

if(use_rsh) { execl(RSH_CMD, RSH_CMD, pg->hostname, kill_cmd, NULL); @@ -1554,7 +1555,7 @@

if(!plist[i].remote_pid) exit(EXIT_SUCCESS);

- snprintf(kill_cmd, 80, "kill -s 9 %d>&/dev/null", + snprintf(kill_cmd, 80, "kill -s 9 %d>/dev/null 2>&1", plist[i].remote_pid);

if(use_rsh) {

Mit freundlichen Gruessen / Best Regards

Axel PHILIPP Geb. 044/557 Dr. rer. nat., Dipl. Phys.

MTU Aero Engines GmbH Informationswirtschaft/Entwicklungssysteme (FIE) Information Management/Engineering Systems (FIE) Dachauer Str. 665 80995 Muenchen Germany

Tel +49 (0)89 1489-4715 Fax +49 (0)89 1489-97533 mailto:[email protected]

-- MTU Aero Engines GmbH Geschaeftsfuehrung/Board of Management: Egon W. Behle, Vorsitzender/CEO; Dr. Rainer Martens, Dr. Stefan Weingartner, Reiner Winkler Vorsitzender des Aufsichtsrats/Chairman of the Supervisory Board: Klaus Eberhardt Sitz der Gesellschaft/Registered Office: Muenchen Handelsregister/Commercial Register: Muenchen HRB 154230

Diese E-Mail sowie ihre Anhaenge enthalten MTU-eigene vertrauliche oder rechtlich geschuetzte Informationen. Wenn Sie nicht der beabsichtigte Empfaenger sind, informieren Sie bitte den Absender und loeschen Sie diese E-Mail sowie die Anhaenge. Das unbefugte Speichern, Kopieren oder Weiterleiten ist nicht gestattet.

This e-mail and any attached documents are proprietary to MTU, confidential or protected by law. If you are not the intended recipient, please advise the sender and delete this message and its attachments. Any unauthorised storing, copying or distribution is prohibited.

_______________________________________________ mpich2-maint mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/mpich2-maint

_______________________________________________ mpich2-maint mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/mpich2-maint

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