ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
18 209 коммитов 3 Ветки 0 Теги 130 MiB
e100b8d165
Граф коммитов

2968 Коммитов

Автор SHA1 Сообщение Дата
Ralph Castain
e100b8d165 don't need the return value, but should check for error 
This commit was SVN r28534.
 2013-05-16 15:15:02 +00:00
Jeff Squyres
128cc27417 Minor type fix (they're both enums/ints, so the compiler previously 
silently cast them).

This commit was SVN r28532.
 2013-05-16 00:47:37 +00:00
Ralph Castain
3a372a65b8 Mapping policies must be tested as equalities as they are values, not bitmasks 
This commit was SVN r28526.
 2013-05-15 13:45:00 +00:00
Ralph Castain
29e4b0cc50 Cannot test equality on mapping directives as it is a bitmask 
This commit was SVN r28525.
 2013-05-15 13:41:49 +00:00
Ralph Castain
04b11accd3 Silience a few warnings 
This commit was SVN r28515.
 2013-05-14 21:58:40 +00:00
Ralph Castain
5296099ecb Fix the cpus-per-rank when binding to hwthreads. Add cpus-per-rank to diag printout 
Thanks to Elena for reporting the problem

This commit was SVN r28508.
 2013-05-14 20:17:50 +00:00
Ralph Castain
427b6b0b47 Fix the verbosity of yet another framework...sigh. 
This commit was SVN r28481.
 2013-05-13 14:36:32 +00:00
Jeff Squyres
456df1c9f7 Remove redundant opal_output() messages from the module; the called 
functions will now show_help() their own error messages if something
goes wrong (per r28470).

This commit was SVN r28471.

The following SVN revision numbers were found above:
  r28470 --> open-mpi/ompi@2ff95a7739
 2013-05-10 15:12:07 +00:00
Jeff Squyres
2ff95a7739 Proper show_help error messages for LAMA. 
This commit was SVN r28470.
 2013-05-10 15:06:25 +00:00
Ralph Castain
f15fe5045e Ensure that debugger connect can occur by getting the rml contact info updated before calling init_after_spawn 
cmr:v1.7.3,reviewer=jsquyres

This commit was SVN r28455.
 2013-05-06 22:00:45 +00:00
Ralph Castain
c52b94af8b Revert r28453 and r28452 - wrong fix 
This commit was SVN r28454.

The following SVN revision numbers were found above:
  r28452 --> open-mpi/ompi@756ee4b5e0
  r28453 --> open-mpi/ompi@6da24143a2
 2013-05-06 21:52:17 +00:00
Ralph Castain
6da24143a2 Minor performance improvement 
This commit was SVN r28453.
 2013-05-06 20:27:16 +00:00
Ralph Castain
756ee4b5e0 Update the rml_uri for each proc so debuggers can attach 
This commit was SVN r28452.
 2013-05-06 20:18:14 +00:00
Ralph Castain
707d0e653a Must use equal and not & comparison for mapping directives 
This commit was SVN r28451.
 2013-05-06 15:07:12 +00:00
Ralph Castain
a0a6412545 Do a little cleanup on abnormal termination procedure - don't keep submitting forced exit events (one will do), no need to reset the abnormal termination pipe event in orterun, etc. 
This commit was SVN r28450.
 2013-05-05 17:39:45 +00:00
Jeff Squyres
42a9a4c62c After examining a '''lot''' of MTT output with Ralph, fix the cause of 
many, many MTT timeouts when running jobs under SLURM: send the right
command at the end to cause remote orteds to shut down.

This commit was SVN r28438.
 2013-05-02 00:23:53 +00:00
Ralph Castain
5d7a93c032 Add the ability to use an external version of libevent. Clearly not recommended at this time. I've verified that it works in limited scenarios, but more thorough testing and performance impacts need to be assessed. 
Interesting how many includes had to be fixed here and there to fill in missing dependencies :-)

This commit was SVN r28411.
 2013-04-29 17:02:37 +00:00
Ralph Castain
3a354c4ea3 Cleanup the verbose output channel name 
This commit was SVN r28391.
 2013-04-24 23:44:02 +00:00
Ralph Castain
c5e1a7dc65 fix typo 
This commit was SVN r28390.
 2013-04-24 23:37:59 +00:00
Nathan Hjelm
55decca2b7 routed/debruijn: if the next hop for a message is unknown forward the message to the parent process 
This commit was SVN r28384.
 2013-04-24 19:05:25 +00:00
Ralph Castain
2e8946db0a Add some debug output 
This commit was SVN r28371.
 2013-04-23 23:11:22 +00:00
Ralph Castain
b6377a2138 Properly remove the object from the list prior to releasing it when an error is encountered 
This commit was SVN r28370.
 2013-04-23 22:44:52 +00:00
Ralph Castain
1a54e52da4 Per conversation with Josh H., remove the stale rsh component from the filem framework. Let the "raw" component do the job. 
This commit was SVN r28338.
 2013-04-16 20:39:48 +00:00
Jeff Squyres
349ee654c1 Fix some --without-hwloc compile errors. Also remove one 
assigned-but-not-used variable assignment.

This commit was SVN r28321.
 2013-04-10 15:08:31 +00:00
Ralph Castain
45af6cf59e The move of the orte_db framework to opal required that we create an opaque opal_identifier_t type as OPAL cannot know anything about the ORTE process name. However, passing a value down to opal and then having the db components reference it causes alignment issues on Solaris Sparc platforms. So pass the pointer instead and do the old "memcpy" trick to avoid the problem. 
This commit was SVN r28308.
 2013-04-08 23:34:16 +00:00
Ralph Castain
76426285f0 Cannot retain opal_buffer_t, so use a copy 
This commit was SVN r28302.
 2013-04-07 23:02:59 +00:00
Ralph Castain
698b4ad6e7 Fix the parameter handling so no-tree-spawn isn't getting reversed 
This commit was SVN r28300.
 2013-04-07 15:48:25 +00:00
Ralph Castain
1f011bef99 Cleanup the updated sys limits capability. Fix a few copy/paste bugs (my bad). Shift the limit set to the ODLS default module so that we sete the limits for all apps, even those that don't call opal_init. Leave it in opal_init as well to support direct-launch apps, but ensure we only set the limits once by removing the envar after launch by ODLS. 
Provide some nice error messages if we fail to set the limits. Since the user had to specifically request we set the limit, treat failure as an error-out situation.

This commit was SVN r28288.
 2013-04-04 16:00:17 +00:00
Ralph Castain
252147fba6 Cleanup error message if unknown host is given in -host and -hostfile options 
This commit was SVN r28262.
 2013-03-28 16:52:10 +00:00
Ralph Castain
e6ae088813 Cleanup error outputs when a daemon fails to start 
This commit was SVN r28261.
 2013-03-28 16:51:19 +00:00
Ralph Castain
21ee48de57 Add missing static declaration 
This commit was SVN r28247.
 2013-03-27 21:59:17 +00:00
Ralph Castain
1b5b9afa85 Silence warnings 
This commit was SVN r28244.
 2013-03-27 21:56:46 +00:00
Nathan Hjelm
17315bf360 Now that the entire codebase has been updated to use the MCA framework 
system remove the last calls to the MCA parameter system.

This commit was SVN r28242.
 2013-03-27 21:17:53 +00:00
Nathan Hjelm
c041156f60 Update ORTE frameworks to use the MCA framework system. 
This commit was SVN r28240.
 2013-03-27 21:14:43 +00:00
Nathan Hjelm
365cf48db5 Update OPAL frameworks to use the MCA framework system. 
This commit was SVN r28239.
 2013-03-27 21:11:47 +00:00
Nathan Hjelm
cf377db823 MCA/base: Add new MCA variable system 
Features:
 - Support for an override parameter file (openmpi-mca-param-override.conf).
   Variable values in this file can not be overridden by any file or environment
   value.
 - Support for boolean, unsigned, and unsigned long long variables.
 - Support for true/false values.
 - Support for enumerations on integer variables.
 - Support for MPIT scope, verbosity, and binding.
 - Support for command line source.
 - Support for setting variable source via the environment using
   OMPI_MCA_SOURCE_<var name>=source (either command or file:filename)
 - Cleaner API.
 - Support for variable groups (equivalent to MPIT categories).

Notes:
 - Variables must be created with a backing store (char **, int *, or bool *)
   that must live at least as long as the variable.
 - Creating a variable with the MCA_BASE_VAR_FLAG_SETTABLE enables the use of
   mca_base_var_set_value() to change the value.
 - String values are duplicated when the variable is registered. It is up to
   the caller to free the original value if necessary. The new value will be
   freed by the mca_base_var system and must not be freed by the user.
 - Variables with constant scope may not be settable.
 - Variable groups (and all associated variables) are deregistered when the
   component is closed or the component repository item is freed. This
   prevents a segmentation fault from accessing a variable after its component
   is unloaded.
 - After some discussion we decided we should remove the automatic registration
   of component priority variables. Few component actually made use of this
   feature.
 - The enumerator interface was updated to be general enough to handle
   future uses of the interface.
 - The code to generate ompi_info output has been moved into the MCA variable
   system. See mca_base_var_dump().

opal: update core and components to mca_base_var system
orte: update core and components to mca_base_var system
ompi: update core and components to mca_base_var system

This commit also modifies the rmaps framework. The following variables were
moved from ppr and lama: rmaps_base_pernode, rmaps_base_n_pernode,
rmaps_base_n_persocket. Both lama and ppr create synonyms for these variables.

This commit was SVN r28236.
 2013-03-27 21:09:41 +00:00
Ralph Castain
e7ac6c9bde Don't build rank_file if you can't use it anyway 
This commit was SVN r28233.
 2013-03-27 15:12:40 +00:00
Ralph Castain
256414121e Protect the cpus-per-rank MCA param registration so that --without-hwloc will build 
This commit was SVN r28232.
 2013-03-27 14:53:30 +00:00
Ralph Castain
317915225c Finish the binding cleanup by removing the no-longer-used binding level scheme. This proved to be fallible as there is no guarantee that the hierarchy it used matched physical reality of the machine (e.g., is L3 "above" the socket or not). Still have to complete the ppr update, but get the rest of it correct. 
This commit was SVN r28223.
 2013-03-26 20:09:49 +00:00
Ralph Castain
24b91839aa Ensure the process knows it local cpuset early enough to perform the locality computation 
This commit was SVN r28221.
 2013-03-26 19:14:23 +00:00
Ralph Castain
6ee32767d4 Restore the cpus-per-proc option for byslot and bynode mapping. Remove the bind_idx (which recorded the index of the hwloc object where the proc was bound) as this would no longer be unique, and just use the bitmap as the standard reference for location. Update the relative locality computation to take bitmaps as its argument. 
This commit was SVN r28219.
 2013-03-26 18:27:50 +00:00
Ralph Castain
2f43989d22 Add debug and handle the use-case where someone (a) uses a hostfile while in a managed allocation to sub-allocate runs, and (b) includes the HNP's node in one of those hostfiles. 
cmr:v1.7

This commit was SVN r28203.
 2013-03-22 00:53:33 +00:00
Jeff Squyres
63d17ce901 Fix CID 968581: ensure that the string read from the socket is always 
\0-terminated so that strlen() and strstr() can be used without fear.
Also fix some insignificant mem leaks (which is somewhat moot, because
as soon as we leave those error conditions, the process will be
terminating, but what the heck, might as well fix these while I was in
the file for the \0-termination issue...).

This commit was SVN r28199.
 2013-03-21 16:05:50 +00:00
Jeff Squyres
562db0dd11 Fix CID 741328: remove some dead code 
This commit was SVN r28192.
 2013-03-21 11:15:06 +00:00
Ralph Castain
fa13d27238 Avoid double-release in error path 
This commit was SVN r28190.
 2013-03-20 21:00:59 +00:00
Ralph Castain
147c6ff9e7 Clean out the cruft leftover from the use_common_ports experiment 
cmr:v1.7

This commit was SVN r28184.
 2013-03-20 15:07:43 +00:00
Ralph Castain
a4b6fb241f Remove all remaining vestiges of the Windows integration 
This commit was SVN r28137.
 2013-02-28 17:31:47 +00:00
Ralph Castain
cf9796accd Remove the old configure option for disabling full rte support - we now use the OMPI rte framework for such purposes 
This commit was SVN r28134.
 2013-02-28 01:35:55 +00:00
Ralph Castain
347df93cd4 Handle the case of someone specifying a directory for the application. Ensure we get a non-zero exit status and clarify the error message. 
cmr:v1.7

This commit was SVN r28119.
 2013-02-27 01:36:21 +00:00
Ralph Castain
f36312ee6f Continue cleanup - this time, start working on the "without full support" flags in ORTE. Remove no-longer-needed configure.m4 files from the ess and errmgr. In the former case, since all priorities are now the same (given the removal of the cnos component), configure priorities are no longer required. 
This commit was SVN r28118.
 2013-02-26 21:27:48 +00:00