1
1
Граф коммитов

18450 Коммитов

Автор SHA1 Сообщение Дата
Ralph Castain
7a21661785 Silence a warning when --without-hwloc is used
This commit was SVN r28783.
2013-07-13 17:17:17 +00:00
Ralph Castain
5bdc4082ea Remove stale platform directory
This commit was SVN r28777.
2013-07-12 18:49:08 +00:00
Ralph Castain
ee3e3a3dd8 Fix typo for make tarball
This commit was SVN r28776.
2013-07-12 18:48:38 +00:00
Ralph Castain
0230a9d4f8 Hmmm...try and remove these again
This commit was SVN r28775.
2013-07-12 18:47:55 +00:00
Ralph Castain
af950eb8ac Update platform files
This commit was SVN r28774.
2013-07-12 18:45:03 +00:00
Ralph Castain
e093187ff6 Cleanup and rename of platform files
This commit was SVN r28773.
2013-07-12 18:42:16 +00:00
Nathan Hjelm
dfca3d4804 fix typos in the ugni and vader btls
This commit was SVN r28772.
2013-07-12 17:55:33 +00:00
Nathan Hjelm
8f9b7926ec mca/base: fix component selection negation. cmr:v1.7:reviewer=jsquyres
This commit was SVN r28770.
2013-07-12 17:55:20 +00:00
Dave Goodell
3741d62308 fix --without-hwloc build failure
All builds since r28682 configured with '--without-hwloc' fail at "make"
time without this fix.

Reviewed by rhc@

This commit was SVN r28769.

The following SVN revision numbers were found above:
  r28682 --> open-mpi/ompi@446e33a5d8
2013-07-12 17:21:14 +00:00
Ralph Castain
b001d31c27 Per RFC, remove libevent 2.0.19 and leave 2.0.21 as the default
This commit was SVN r28767.
2013-07-12 16:37:15 +00:00
Nathan Hjelm
1119cd3e8a Merge branch 'vader_fix'
This commit was SVN r28764.
2013-07-11 23:30:20 +00:00
Brian Barrett
2f19fc52de use the same multi-md workaround the rest of the Portals code is using.
This commit was SVN r28761.
2013-07-11 21:00:11 +00:00
Nathan Hjelm
b5281778b0 btl/vader: improve small message performance
This commit improved the small message latency and bandwidth when using
the vader btl. These improvements should make performance competative
with other MPI implementations.

This commit was SVN r28760.
2013-07-11 20:54:12 +00:00
Brian Barrett
bea54eeeb1 First take at a BTL for Portals 4
This commit was SVN r28759.
2013-07-11 20:47:08 +00:00
Jeff Squyres
9252afdcd9 Updates and tweaks to the documentation of the new MCA parameter
system (written in conjunction with Nathan).

This commit was SVN r28758.
2013-07-11 20:04:51 +00:00
Jeff Squyres
bdb45a2e4f Add an oh-so-slightly faster variant of the hotel "checkin" action
(since this is used in the fast path) for when you ''know'' that there
will be a room available:

 * Don't do the last_unoccupied_room check
 * Return void

This commit was SVN r28757.
2013-07-11 20:00:37 +00:00
Jeff Squyres
baa3182794 Per RFC
(http://www.open-mpi.org/community/lists/devel/2013/07/12534.php),
remove a bunch of dead code.

This commit was SVN r28756.
2013-07-11 17:34:28 +00:00
Rolf vandeVaart
858ef65142 Fix loop limit.
This commit was SVN r28755.
2013-07-11 17:15:43 +00:00
Rolf vandeVaart
5051cd53fd Use new API.
This commit was SVN r28754.
2013-07-11 17:06:14 +00:00
Joshua Ladd
16beaa3878 This fixes the nasty configure.m4 hack that was added long ago and not removed. My fault for not catching earlier. I've also removed the '.ompi_ignore' in coll/hcoll. Throwing this to Nathan for review. Upon successful review, this should be added to cmr:v1.7:reviewer=hjelmn
This commit was SVN r28753.
2013-07-11 09:55:46 +00:00
Nathan Hjelm
a694bcb6b6 Add support for the MCA variable information level to ompi_info.
Add an option to ompi_info (-l, --level) that takes a number in the
interval (1,9). Only MCA variables up to this level will be printed.
The default level is 1.

Print the level as part of both the parsable and readable output.

This commit was SVN r28750.
2013-07-10 18:52:36 +00:00
Jeff Squyres
28dac8010b The hcoll component configure.m4 commits multiple sins, and breaks
many builds.  I am temporarily .ompi_ignore'ing this component until
it can be fixed by its owner.

 * It calls AC_MSG_ERROR, which configure.m4 scripts are ''never''
   supposed to do.  If you don't want to build, then call $2.
 * All static and --disable-dlopen builds are broken; they fall afoul
   of whatever test configure.m4 is doing and therefore error out of
   configure entirely (vs. simply disabling the hcoll component).
 * There appear to be multiple shell scripting errors in the
   configure.m4.  Here's the output of "./configure --disable-dlopen":
{{{
--- MCA component coll:hcoll (m4 configuration macro)
checking for MCA component coll:hcoll compile mode... static
checking --with-hcoll value... simple ok (unspecified)
./configure: line 421: test: basic: integer expression expected
configure: error: Can not use coll/hcoll and coll/ml (static build)
   simultaneously. You have two options:
                1. Use static build & disable ml with:
   --enable-mpi-no-build=coll-ml
                2. Use dso build for ML & disable ml at runtime: -mca
   coll self
./configure: line 310: return: basic: numeric argument required
./configure: line 320: exit: basic: numeric argument required
}}}

Finally, all of these configure.m4 errors aside, I don't understand
why there is a ''compile-time'' exclusion between the hcoll and ml
components.  Why isn't this a ''run-time'' decision?  Having what
seems to be an unnecessary compile-time exclusion goes against the
general Open MPI philosophy.

Note: Open MPI 1.7 is also broken in all the same ways.  I suggest
that the RM's .ompi_ignore hcoll over there, too.

Mellanox: please fix.

This commit was SVN r28748.
2013-07-10 16:03:15 +00:00
Jeff Squyres
80145742a3 Fix typo in comment
This commit was SVN r28747.
2013-07-10 15:13:08 +00:00
Jeff Squyres
ea94936531 First cut at assigning some fine-grained "levels" to MCA parameters
for the SM and TCP BTLs, as well as the mca_btl_base_param_register()
function (which registers MCA params for all BTLs).

The guidelines in
https://svn.open-mpi.org/trac/ompi/wiki/MCAParamLevels were used to
pick these levels.

This commit was SVN r28746.
2013-07-10 00:47:52 +00:00
Jeff Squyres
04da831f04 The gm BTL was removed in June of 2012; there's no need for
ompi_check_gm.m4 any more.

This commit was SVN r28744.
2013-07-09 20:57:12 +00:00
Dave Goodell
5626371196 update-my-copyright.pl now works with Git
This script now takes command line options:
```
./update-my-copyright.pl [options]

--help | -h          This help message
--quiet | -q         Only output critical messages to stdout
--check-only         exit(111) if there are files with copyrights to edit
--search-name=NAME   Set search name to NAME
--formal-same=NAME   Set formal name to NAME
```

The `--check-only` and `--quiet` options are suitable for use in a git
pre-commit script to check for out of date copyright headers.

Reviewed by jsquyres

This commit was SVN r28742.
2013-07-09 14:39:41 +00:00
Ralph Castain
028f5ee7a6 Cleanup some bitrot from moving the db framework to opal and from the new mca param system
This commit was SVN r28741.
2013-07-09 14:37:08 +00:00
Ralph Castain
9366fda374 Fix names in test - still generating warnings
This commit was SVN r28740.
2013-07-09 02:58:58 +00:00
Aurelien Bouteiller
e1066143a4 rename ompi_free_list operations to _mt, as per discussions at last face to face meeting
This commit was SVN r28734.
2013-07-08 22:07:52 +00:00
Brian Barrett
ecbbf888d3 * Update Portals 4 MTL's multi-md code to be a bit cleaner (no if statements
in the path) and not create MDs due to boundary crossing
* Add the same logic to the Coll component

This commit was SVN r28733.
2013-07-08 21:27:37 +00:00
Ralph Castain
315da8125d Remove stale headers
cmr:v1.7.3:reviewer=jsquyres

This commit was SVN r28732.
2013-07-08 18:26:58 +00:00
Ralph Castain
2ccc0438af On some systems, pthread_kill is actually in the "signals.h" header, so include it
This commit was SVN r28731.
2013-07-08 17:40:38 +00:00
Ralph Castain
eac174e624 For purposes of testing the RFC, make libevent2021 the default for now so it gets tested by MTT
This commit was SVN r28730.
2013-07-05 23:14:22 +00:00
Brian Barrett
84aeb6a6a5 Update request alloc to use free list get instead of free list wait.
This commit was SVN r28729.
2013-07-05 20:24:43 +00:00
Brian Barrett
ea9cee73c1 Per RFC, remove darwin backtrace, since OS X since 10.5 has supported the
execinfo() interface (which has been the default for OMPI to use on Darwin)

This commit was SVN r28727.
2013-07-05 19:06:27 +00:00
George Bosilca
dc9352faf6 Remove some unused variables.
This commit was SVN r28726.
2013-07-05 13:31:54 +00:00
George Bosilca
8b01c3da33 Slightly reorder the code.
This commit was SVN r28725.
2013-07-05 13:29:29 +00:00
Jeff Squyres
b417095639 Do not destroy the sub-communicator until we have freed its attributes,
per the reason cited in the comment in the code.

This commit was SVN r28724.
2013-07-05 12:15:03 +00:00
George Bosilca
483ed8da8c Remove an unused variable resulting from the removal of the last parameter of
the OMPI_FREE_LIST_GET macro.

This commit was SVN r28723.
2013-07-04 09:19:00 +00:00
George Bosilca
c9e5ab9ed1 Our macros for the OMPI-level free list had one extra argument, a possible return
value to signal that the operation of retrieving the element from the free list
failed. However in this case the returned pointer was set to NULL as well, so the
error code was redundant. Moreover, this was a continuous source of warnings when
the picky mode is on.

The attached parch remove the rc argument from the OMPI_FREE_LIST_GET and
OMPI_FREE_LIST_WAIT macros, and change to check if the item is NULL instead of
using the return code.

This commit was SVN r28722.
2013-07-04 08:34:37 +00:00
Ralph Castain
21c8041a40 Update libevent 2021 component so it also only warns once when detecting reentrant behavior
This commit was SVN r28721.
2013-07-04 04:41:04 +00:00
Ralph Castain
62378209f0 Even if we don't find the default hostfile, and nothing else was provided, then use all the known nodes.
cmr:v1.7.3:#3653:reviewer=jsquyres
cmr:v1.6.6:#3654:reviewer=jsquyres

This commit was SVN r28718.
2013-07-03 22:31:32 +00:00
Ralph Castain
443a6802b9 If the default hostfile is empty, we need to pickup all the known nodes, not just the head node.
cmr:v1.7.3:reviewer=jsquyres
cmr:v1.6.6:reviewer=jsquyres

This commit was SVN r28717.
2013-07-03 22:25:51 +00:00
Ralph Castain
bd65937bf3 If we enable ipv6, we resolve a hosts addresses and check them all against our local interfaces to determine if the given host is us. However, if we don't enable ipv6, we only checked the first address returned. This can cause us to incorrectly identify a hostname as "not us".
Make -disable-ipv6 behave the same as --enable-ipv6 by checking all the returned addresses.

This commit was SVN r28716.
2013-07-03 21:41:36 +00:00
Brian Barrett
d3b49535b5 Only allow communication from the same user, since we don't have job-level
protection.

This commit was SVN r28715.
2013-07-03 17:29:02 +00:00
Ralph Castain
45fad1ddcc We really should be closing the event framework when told to do so.
cmr:v1.7.3,reviewer=jsquyres

This commit was SVN r28714.
2013-07-03 16:57:14 +00:00
Jeff Squyres
d1ce64f049 Fix some "malloc of 0 bytes" warnings
This commit was SVN r28713.
2013-07-03 12:05:33 +00:00
Ralph Castain
9166a8cc95 Per telecon today, add a flag so we only warn once about reentrant libevent loops - this will allow developers to better diagnose the problem as we won't swamp filesystems with warning messages.
This commit was SVN r28712.
2013-07-03 04:51:36 +00:00
Ralph Castain
243f9ef586 Set ignores
This commit was SVN r28711.
2013-07-03 04:47:33 +00:00
Jeff Squyres
ad16bcd6d1 Followup from Justin Bronder: Looks like I spoke too soon. The
sandbox team has informed me that they are getting rid of SANDBOX_PID
in the future and that using SANDBOX_ON would be preferred.

This commit was SVN r28708.
2013-07-03 01:38:26 +00:00