Ralph Castain 4bbc9a28d6 Try to resolve the simultaneous connection problem by being a little more careful about the choice of returned status when a connection is refused. As before, have the higher vpid of the two peers retry the connection, while the lower one waits. This can happen in a couple of places, so try to hit them all. Since this is hard to test, will ask Gilles to give it a try since he's the one who is seeing it. ...

cmr=v1.8.3:reviewer=rhc

This commit was SVN r32744.

2014-09-16 18:59:36 +00:00

..

dfs

Per the PMIx RFC:

2014-08-21 18:56:47 +00:00

errmgr

Resolve a race condition that could cause us to hang during abnormal terminations due to multi-counting num_terminated

2014-09-02 00:32:52 +00:00

ess

Re-enable the usock oob component. Ensure the TCP component promotes messages for other procs to the OOB base so that other components have a chance to send the relay. Seems to be passing MTT, so let's see how it works for others.

2014-08-30 19:33:46 +00:00

filem

Per the PMIx RFC:

2014-08-21 18:56:47 +00:00

grpcomm

Temporarily make the direct grpcomm component the default until we can debug the other modules

2014-09-11 14:47:54 +00:00

iof

Per the PMIx RFC:

2014-08-21 18:56:47 +00:00

odls

Resolve a race condition that could cause us to hang during abnormal terminations due to multi-counting num_terminated

2014-09-02 00:32:52 +00:00

oob

Try to resolve the simultaneous connection problem by being a little more careful about the choice of returned status when a connection is refused. As before, have the higher vpid of the two peers retry the connection, while the lower one waits. This can happen in a couple of places, so try to hit them all. Since this is hard to test, will ask Gilles to give it a try since he's the one who is seeing it.

2014-09-16 18:59:36 +00:00

plm

[Contribution from Artem - moved it to svn from git for him]

2014-09-15 18:00:46 +00:00

ras

Some minor uninitialized variable cleanups

2014-08-29 02:21:13 +00:00

rmaps

Per patch from Tetsuya, allow the user to bind-to none when specifying multiple pe's/rank as requested by Reuti. This allows the user to reserve multiple "slots" in the allocation for each process while mapping, but not to bind the process to specific processing elements on the node.

2014-09-10 15:52:18 +00:00

rml

[Contribution from Artem - moved it to svn from git for him]

2014-09-15 18:00:46 +00:00

routed

Per the PMIx RFC:

2014-08-21 18:56:47 +00:00

rtc

check-help-strings cleanup

2014-08-11 03:26:21 +00:00

snapc

Per the PMIx RFC:

2014-08-21 18:56:47 +00:00

sstore

use the orte attributes in the FT code to fix compile errors

2014-06-26 03:19:17 +00:00

state

Fix a race condition caused by a bad attribute flag that created an OR instead of an AND condition check

2014-08-22 22:48:16 +00:00