openmpi

dfs

The final step of the RFC: convert the <foo>libdir and friends to fit their respective code areas, and equate them all at the top. Note that we can't entirely separate things as the opal_install_dirs framework can't handle separated locations for the various trees.

2014-05-08 02:01:35 +00:00

errmgr

Resolve the MTT timeout problem. This appears to have largely been caused by missing sigchld notifications, thus causing the daemons to believe that not all procs had exited. Let comm failure also serve as notification of process termination, and add appropriate flags/attributes to avoid multiple reporting of proc termination.

2014-06-21 17:09:02 +00:00

ess

Fix potential issue if opal_hwloc_topology is NULL

2014-06-19 18:52:41 +00:00

filem

This isn't as big a change as it appears - a change in one place caused a whole bunch of files to require updated #include's due to some arcane linkage. Rework the orte_wait code to reflect the introduction of the state machine. If we are in cleanup mode and just want to kill all our local children, then there is no reason to be polite about it as that introduces *very* long delays at scale. Just kill the procs and move on.

2014-06-17 17:57:51 +00:00

grpcomm

Cleanup a memory leak in the daemons - thanks to Artem for spotting it

2014-06-09 17:14:02 +00:00

iof

This isn't as big a change as it appears - a change in one place caused a whole bunch of files to require updated #include's due to some arcane linkage. Rework the orte_wait code to reflect the introduction of the state machine. If we are in cleanup mode and just want to kill all our local children, then there is no reason to be polite about it as that introduces *very* long delays at scale. Just kill the procs and move on.

2014-06-17 17:57:51 +00:00

odls

Resolve the MTT timeout problem. This appears to have largely been caused by missing sigchld notifications, thus causing the daemons to believe that not all procs had exited. Let comm failure also serve as notification of process termination, and add appropriate flags/attributes to avoid multiple reporting of proc termination.

2014-06-21 17:09:02 +00:00

oob

Per the OMPI developer conference, remove the last vestiges of OMPI_USE_PROGRESS_THREADS

2014-06-24 17:05:11 +00:00

plm

Mark the proc as alive so waitpid callback system doesn't immediately activate the callback

2014-06-18 14:04:55 +00:00

ras

Cleanup compile issues - missing updates to some plm components and the slurm ras component

2014-06-01 17:59:06 +00:00

rmaps

Don't release the node_name field as it gets used in the slots parsing - will be released at newline detection

2014-06-20 13:18:46 +00:00

rml

This isn't as big a change as it appears - a change in one place caused a whole bunch of files to require updated #include's due to some arcane linkage. Rework the orte_wait code to reflect the introduction of the state machine. If we are in cleanup mode and just want to kill all our local children, then there is no reason to be polite about it as that introduces *very* long delays at scale. Just kill the procs and move on.

2014-06-17 17:57:51 +00:00

routed

This isn't as big a change as it appears - a change in one place caused a whole bunch of files to require updated #include's due to some arcane linkage. Rework the orte_wait code to reflect the introduction of the state machine. If we are in cleanup mode and just want to kill all our local children, then there is no reason to be polite about it as that introduces *very* long delays at scale. Just kill the procs and move on.

2014-06-17 17:57:51 +00:00

rtc

Fix static builds by renaming conflicting type

2014-06-14 17:39:28 +00:00

snapc

This isn't as big a change as it appears - a change in one place caused a whole bunch of files to require updated #include's due to some arcane linkage. Rework the orte_wait code to reflect the introduction of the state machine. If we are in cleanup mode and just want to kill all our local children, then there is no reason to be polite about it as that introduces *very* long delays at scale. Just kill the procs and move on.

2014-06-17 17:57:51 +00:00

sstore

Mark the proc as alive so waitpid callback system doesn't immediately activate the callback

2014-06-18 14:04:55 +00:00

state

Resolve the MTT timeout problem. This appears to have largely been caused by missing sigchld notifications, thus causing the daemons to believe that not all procs had exited. Let comm failure also serve as notification of process termination, and add appropriate flags/attributes to avoid multiple reporting of proc termination.

2014-06-21 17:09:02 +00:00