1
1
Граф коммитов

5372 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
323aff7b1e - First cut of a paffinity (processor affinity) framework in opal
- Simple components for getting and setting processor affinity of a
  process; does *not* include scheduling decisions
- No one in the OMPI code base invokes the framework yet
- Added linux component for using sched_setaffinity()
- Added shell solaris component that will use processor_bind()
  (currently .ompi_ignore'd)

This commit was SVN r6854.
2005-08-13 14:00:56 +00:00
Galen Shipman
067357617f add r2 to sources..
This commit was SVN r6853.

The following SVN revision numbers were found above:
  r2 --> open-mpi/ompi@58fdc18855
2005-08-13 05:32:01 +00:00
Brian Barrett
1b830beddb * move over changes from the /tmp/bwb-memory-hooks/copy-1 into the trunk.
This includes updates to the malloc_hook method and making everything
  components.

This commit was SVN r6852.
2005-08-13 01:08:34 +00:00
Josh Hursey
9bcb2989a5 Some general cleanup.
Added the FT-MPI interface functions. Many are hidden and not implemented,
but serve as place holders for future work.

This commit was SVN r6847.
2005-08-12 22:09:58 +00:00
Galen Shipman
c3fcf25508 fix to add_procs
This commit was SVN r6846.
2005-08-12 21:42:55 +00:00
Jeff Squyres
cf16a521c8 Ensure to get ompi/include/constants.h
This commit was SVN r6845.
2005-08-12 21:42:07 +00:00
Tim Woodall
0ab588e5d2 added support for sending eager messages directly from user buffer
if supported by btl - btl must set MCA_BTL_FLAGS_SEND_INPLACE

This commit was SVN r6844.
2005-08-12 21:33:01 +00:00
Jeff Squyres
cf2c8b45a8 - Use proper prefixes for all #include statements (opal/, orte/, and
ompi/).
- There's still a handful of places that have orte/ #include files;
  still need to clean those up
- A lot of places still use ompi/include/constants.h -- those need to
  be converted over to use OPAL_ return codes and then switch to the
  opal constants.h.  This commit is the first few steps towards
  that...

This commit was SVN r6843.
2005-08-12 20:46:25 +00:00
Josh Hursey
e370141ae3 add memory manager init
This commit was SVN r6842.
2005-08-12 20:45:46 +00:00
Jeff Squyres
090712ebf9 Remove two useless files (they weren't even listed in their respective
Makefile.am's)

This commit was SVN r6841.
2005-08-12 20:30:25 +00:00
Jeff Squyres
b2225ea8f3 Arrgh -- stupid shell curses problems ended up naming that last commit
poorly.  :-(

This commit was SVN r6840.
2005-08-12 19:43:58 +00:00
Jeff Squyres
6095735046 Arrgh. Broke the tree with the last commit -- need to rename this to
avoid #include confusion.

This commit was SVN r6839.
2005-08-12 19:40:17 +00:00
Tim Woodall
15c5fd6092 cleanup
This commit was SVN r6838.
2005-08-12 19:29:05 +00:00
Tim Woodall
faf146ec4c correct address count
This commit was SVN r6837.
2005-08-12 19:01:04 +00:00
Jeff Squyres
9947c71aa2 Starting to use this...
This commit was SVN r6836.
2005-08-12 18:54:09 +00:00
Galen Shipman
58f7c3abfb Makefile typo...
This commit was SVN r6834.
2005-08-12 17:01:30 +00:00
Galen Shipman
df78b8957c Misc fixes.. no need for procs hashtable in bml_r2..
This commit was SVN r6833.
2005-08-12 16:59:15 +00:00
Tim Woodall
5558c014b9 default TCP to only be used if self/sm/gm/mvapi.... are not available
This commit was SVN r6832.
2005-08-12 16:56:46 +00:00
Brian Barrett
a63f6e6ce6 re-add values for max tag and max cid to TEG and UNIQ.
This commit was SVN r6830.
2005-08-12 16:28:16 +00:00
Brian Barrett
ad383f5fcd * If the event library is going to be a noop (really, this only happens on
Red Storm), don't bother doing all the book keeping work to do calls
  into the event library

This commit was SVN r6829.
2005-08-12 16:21:17 +00:00
Brian Barrett
a211427eaa * add ignore file
This commit was SVN r6828.
2005-08-12 16:09:49 +00:00
Brian Barrett
57efade1e2 * bump the event_tick_rate count from 100 to 10k, per conversation with Tim.
Might want to bump this higher?
* remove assumption that progress functions registered with opal_progress
  are not reentrant.  There is still the assumption that the event loop
  is not reentrant (because it's not), so that part is still protected by
  a spinlock. 

This commit was SVN r6827.
2005-08-12 16:08:44 +00:00
Brian Barrett
070e376407 * don't build qsort workaround unless we actually have a broken qsort
This commit was SVN r6824.
2005-08-12 15:28:11 +00:00
Brian Barrett
a926a9b4fb * I'm sure any decent optimizing compiler would have figured this one out,
but there's no point in having the call_yield check performed if we
  don't have sched_yield to call in the first place.

This commit was SVN r6823.
2005-08-12 15:21:59 +00:00
Brian Barrett
aab684f159 * fix off by one error that was causing the event library to only be
triggered every other call to opal_progress when the TCP BTL/PTL
  were being used.

This commit was SVN r6822.
2005-08-12 15:20:37 +00:00
Galen Shipman
9cbcdae14b more BML related stuff.. static_components.h is autogen'd configure.params is
not. Compile warning fixed in pml_ob1_sendreq.c

This commit was SVN r6821.
2005-08-12 14:40:06 +00:00
Brian Barrett
2c44b2398d * remove the src/ directory for Portals BTL to bring it in line with the
other BTLs

This commit was SVN r6820.
2005-08-12 14:33:45 +00:00
Galen Shipman
b01ebf45c9 Fixed build error related to direct call (bml_direct_call.h). Misc bug fixes
and compiler warning issues. Fixed threaded build issue. 

This commit was SVN r6819.
2005-08-12 14:08:40 +00:00
Brian Barrett
c8865bd5c1 * remove bool from internal interface, since we might not want to include
ompi_config.h in some of the intercept mechanisms.
* Intercept munmap when called directly by the user when we are using
  ptmalloc2 (previously we only covered when the user called free()).
* Don't go through the locking and list traversal logic trying to fire
  callbacks until there is actually a callback to fire.  This is both
  a performance boost and a way to cope with the hook callback being
  triggered before opal_init.

This commit was SVN r6818.
2005-08-12 13:35:01 +00:00
Brian Barrett
c9f5e591b1 * make sure to try munmap when testing the hooks
* add check to see impact of our hooks with malloc/free timings

This commit was SVN r6817.
2005-08-12 13:29:26 +00:00
Ralph Castain
67bd42d167 Fix a few minor compiler warnings
This commit was SVN r6816.
2005-08-12 04:07:47 +00:00
Galen Shipman
c3c83aa3e1 BML (BTL Managment Layer). Allows BTL's to be used outside of the PML. See
bml.h and PML-OB1 for usage. 

This commit was SVN r6815.
2005-08-12 02:41:14 +00:00
Tim Prins
311efa5bcc fix a small 64bit problem....
This commit was SVN r6814.
2005-08-11 20:36:07 +00:00
Josh Hursey
22c7f2b3e0 Quite a range of small changes.
ns_replica.c
 - Removed the error logging since I use this function in orte_init_stage1 to 
   check if we have created a cellid yet or not.

ras_types.h & rase_base_node.h
 - This was an empty file. moved the orte_ras_node_t from base/ras_base_node.h
   to this file.
 - Changed the name of orte_ras_base_node_t to orte_ras_node_t to match the 
   naming mechanisms in place.

ras.h
 - Exposed 2 functions:
   - node_insert:
     This takes a list of orte_ras_base_node_t's and places them in the Node 
     Segment of the GPR. This is to be used in orte_init_stage1 for singleton 
     processes, and the hostfile parsing (see rds_hostfile.c). This just puts 
     in the appropriate API interface to keep from calling the 
     orte_ras_base_node_insert function directly.
   - node_query:
     This is used in hostfile parsing. This just puts in the appropriate API 
     interface to keep from calling the orte_ras_base_node_query function 
     directly.
 - Touched all of the implemented components to add reference to these new 
   function pointers

ras_base_select.c & ras_base_open.c
 - Add and set the global module reference

rds.h
 - Exposed 1 function:
   - store_resource:
     This stores a list of rds_cell_desc_t's to the Resource Segment. 
     This is used in conjunction with the orte_ras.node_insert function in 
     both the orte_init_stage1 for singleton processes and rds_hostfile.c

rds_base_select.c & rds_base_open.c
 - Add and set the global module reference

rds_hostfile.c
 - Added functionality to create a new cellid for each hostfile, placing 
   each entry in the hostfile into the same cellid. Currently this is 
   commented out with the cellid hard coded to 0, with the intention of 
   taking this out once ORTE is able to handle multiple cellid's
 - Instead of just adding hosts to the Node Segment via a direct call to 
   the ras_base_node_insert() function. First add the hosts to the Resource 
   Segment of the GPR using the orte_rds.store_resource() function then use 
   the API version of orte_ras.node_insert() to store the hosts on the Node 
   Segment.
 - Add 1 new function pointer to module as required by the API.

rds_hostfile_component.c
 - Converted this to use the new MCA parameter registration

orte_init_stage1.c
 - It is possible that a cellid was not created yet for the current environment. 
   So I put in some logic to test if the cellid 0 existed. If it does then 
   continue, otherwise create the cellid so we can properly interact with the 
   GPR via the RDS.
 - For the singleton case we insert some 'dummy' data into the GPR. The RAS 
   matches this logic, so I took out the duplicate GPR put logic, and 
   replaced it with a call to the orte_ras.node_insert() function.
 - Further before calling orte_ras.node_insert() in the singleton case, 
   we also call orte_rds.store_resource() to add the singleton node to the 
   Resource Segment.

Console:
 - Added a bunch of new functions. Still experimenting with many aspects of the
   implementation. This is a checkpoint, and has very limited functionality.
 - Should not be considered stable at the moment.

This commit was SVN r6813.
2005-08-11 19:51:50 +00:00
Tim Woodall
e083c706dc removed debug output
This commit was SVN r6812.
2005-08-11 19:25:45 +00:00
Greg Watson
88c4505e6e Enable component by default.
This commit was SVN r6811.
2005-08-11 19:14:24 +00:00
Greg Watson
ad958e3f19 Fix build problem caused by these functions just "going away".
This commit was SVN r6810.
2005-08-11 19:13:09 +00:00
Josh Hursey
bf2789fc66 Added some error logging for unimplemented daemon commands.
This commit was SVN r6809.
2005-08-11 16:57:06 +00:00
Tim Prins
36435cbe4a a small change to make this configure test right.
This component is ompi_ignored so no need to autogen.....

This commit was SVN r6807.
2005-08-11 14:29:40 +00:00
Josh Hursey
c92c3effb6 Made a global base_module of function pointers for both frameworks.
This commit was SVN r6806.
2005-08-10 22:49:41 +00:00
Josh Hursey
abe16e3fa4 Convert to use new MCA param reg functions
This commit was SVN r6805.
2005-08-10 21:05:13 +00:00
Jeff Squyres
f0003fc122 Update the code counter to account for the new directory structure.
This commit was SVN r6804.
2005-08-10 20:39:33 +00:00
Josh Hursey
52cf1e69e0 update makefile for addition of new header file
This commit was SVN r6803.
2005-08-10 20:03:20 +00:00
Josh Hursey
afe7e687cb A bit of cleanup and a couple of bug fixes for remote orted launching
using orteprobe.

Created a header file for orte_setup_hnp. [HNP = Head Node Process]

General cleanup and added a bit of documentation in orte_setup_hnp.c
Also fixed a cellid tokens issue (circa line 285)
Changed the launched scope from private to public

In orteprobe:
- added reference to orted.h to avoid duplicate header contents in orteprobe.h
- removed the version tag, and put in a verbose argument
- Fixed a buffer packing problem that was causing the parent from receiving the
  proper contact information for the new daemon.

This commit was SVN r6802.
2005-08-10 20:01:25 +00:00
Jeff Squyres
b405316075 These should no longer be here.
This commit was SVN r6801.
2005-08-10 19:49:10 +00:00
Brian Barrett
7b42267631 * need sys/types.h for size_t
This commit was SVN r6800.
2005-08-10 19:20:47 +00:00
Jeff Squyres
51de13bb97 Arrgh -- re-indenting this morning accidentally used real tabs instead
of spaces (curses!  indent(1) had been updated with a new option that
I did not use).  This commit simply converts tabs to real spaces.

This commit was SVN r6799.
2005-08-10 17:53:43 +00:00
George Bosilca
988f70edd8 Correct gatherv and allgatherv collective communications. Always do any internal send even if the count
is zero as the receive is always posted.

This commit was SVN r6798.
2005-08-10 17:30:46 +00:00
Brian Barrett
d9e5e3343d * random code cleanup
This commit was SVN r6797.
2005-08-10 15:26:15 +00:00
Brian Barrett
d4aa9c702c * remove unused files
This commit was SVN r6796.
2005-08-10 15:11:05 +00:00