1
1
openmpi/ompi/mca
Ralph Castain 3e55fe6f6d Fold in the revised modex scheme. Move the ompi_proc_t modex portions to the RTE level since the daemons already have that info. Provide each process with the equivalent of a "nidmap" - both a map of what nodes are in the job, and a map of which node each process is on. This enables the use of static ports, though that hasn't been turned "on" in this commit.
Update the rsh tree spawn capability so we spawn the next wave of daemons before launching our own local procs.

Add an ability to encode nodenames for large clusters with contiguous node name numbering schemes - this allows communication of all node names in a few bytes instead of tens-of-bytes/node.

This commit was SVN r18338.
2008-04-30 19:49:53 +00:00
..
allocator Per long threads on the mailing list and much confusion discussion 2007-12-15 13:32:02 +00:00
bml Fix a checkpoint/restart bug that causes a restarted application to occasionally throw a SIGSEGV or SIGPIPE due to invalid socket descriptors. 2008-04-24 17:54:22 +00:00
btl The r17940 broke the XRC code. 2008-04-29 13:22:40 +00:00
coll accidentally made a change in the wrong place. 2008-04-23 17:32:05 +00:00
common include portals modex fun in the distro 2008-04-28 18:51:54 +00:00
crcp Shift the architecture calculation from the ompi/datatype engine to the opal/util area. This allows us to compute the architecture earlier in the launch and communicate it outside of the modex. 2008-04-17 20:43:56 +00:00
dpm Fix a bug that rpevented to use the same port (as returned by Open_port) for several Comm_accept) 2008-04-25 20:41:44 +00:00
io Restore a placeholder to make non-SVN SCM's happy. 2008-02-28 20:19:22 +00:00
mpool Remove the obsolete and largely unused orte_system_info structure. The only fields that were used in that struct were nodeid and nodename - these have been transferred to the orte_process_info structure. 2008-03-23 23:10:15 +00:00
mtl nasty memory bug... 2008-04-18 03:01:53 +00:00
osc Shift the architecture calculation from the ompi/datatype engine to the opal/util area. This allows us to compute the architecture earlier in the launch and communicate it outside of the modex. 2008-04-17 20:43:56 +00:00
pml Fold in the revised modex scheme. Move the ompi_proc_t modex portions to the RTE level since the daemons already have that info. Provide each process with the equivalent of a "nidmap" - both a map of what nodes are in the job, and a map of which node each process is on. This enables the use of static ports, though that hasn't been turned "on" in this commit. 2008-04-30 19:49:53 +00:00
pubsub Cleanup and fix bugs in the MPI dynamics section. Modify the dpm API so it properly takes ports instead of process names (as correctly identified by Aurelien). Fix race conditions in the use of ompi-server. Fix incompatibilities between the mpi bindings and the dpm implemenation that could cause segfaults due to uninitialized memory. 2008-04-16 14:27:42 +00:00
rcache Don't call free(), or library functions that may call free() inside (such as 2008-01-08 08:55:42 +00:00
topo Present state of MPI debugger work: 2008-03-05 12:22:34 +00:00