ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
2 217 коммитов 3 Ветки 0 Теги 130 MiB
41192eda33
Граф коммитов

2217 Коммитов

This Branch
This Branch
Все ветки
Автор SHA1 Сообщение Дата
Jeff Squyres
d767e7822f Forgot to destroy the mutex 
This commit was SVN r2331.
 2004-08-27 16:43:35 +00:00
Jeff Squyres
5e3a22b822 Fix minor memory leak 
This commit was SVN r2330.
 2004-08-27 16:35:11 +00:00
Rich Graham
21211f7dd3 debug the ompi_fifo functions. 
This commit was SVN r2329.
 2004-08-27 16:32:09 +00:00
Rich Graham
59f339f06b add test for ompi_fifo functions. 
This commit was SVN r2328.
 2004-08-27 16:31:27 +00:00
Ralph Castain
8291cfba9f Missing constant defined here. 
This commit was SVN r2327.
 2004-08-27 15:54:58 +00:00
Weikuan Yu
2ba24b4a0d fix elan4_probe_cmdq compatibility 
This commit was SVN r2326.
 2004-08-27 15:44:15 +00:00
Ralph Castain
1df5cca479 Clean up constructors/destructors for proper memory management. 
This commit was SVN r2325.
 2004-08-27 15:36:53 +00:00
Jeff Squyres
0e12409387 Oops -- this really was not necessary! 
This commit was SVN r2324.
 2004-08-27 15:00:20 +00:00
Jeff Squyres
15bb4154e7 Fix static compiles for "common" libraries -- requires a little extra 
mojo in common/*/Makefile.am files.  Also don't traverse into
src/mca/common and don't snarf the common convenience library if we're
not building staticly (remember: there is no "base" in common).

This commit was SVN r2323.
 2004-08-27 14:55:41 +00:00
Weikuan Yu
b5a8821cff -- Verify the VERSION problem and have it work in both quad and thor 
-- Add data integrity check into the test program
-- To go through tests with get:completion queue:threading

This commit was SVN r2322.
 2004-08-27 14:25:34 +00:00
Jeff Squyres
92c4d73082 Remove old build-the-tarball script; this script has been replaced 
with build_tarball.pl.

This commit was SVN r2321.
 2004-08-27 10:08:32 +00:00
Jeff Squyres
48c437208f Also have this component link to the sm common library 
This commit was SVN r2320.
 2004-08-27 10:07:29 +00:00
Jeff Squyres
fdc399ada5 Add "ompi_config.h" and "util/printf.h" 
This commit was SVN r2319.
 2004-08-27 09:32:08 +00:00
Jeff Squyres
07d9849e62 As was previously decided, "common" blocks are vanilla shared 
libraries -- not components.  So they shouldn't show up in ompi_info.

This commit was SVN r2318.
 2004-08-27 09:19:11 +00:00
Ralph Castain
f1ab634fab Massive update of the registry system to incorporate publish/subscribe notifications, including new "synchro" function that allows a barrier-like operation to be performed on a registry segment. Corresponding update to unit test for replica - system passes, but additional new functionality needs to be added to test. 
This commit was SVN r2317.
 2004-08-27 05:23:04 +00:00
Jeff Squyres
7ccf210a28 According to the AM manual (which I was re-reading today for another 
upcoming commit), this is the preferred way to do conditional source
files these days.  AM can then infer which files to pick up for
distribution without the need for EXTRA_DIST.  Also, for mpi/c/profile
and mpi/f77/profile, don't distribute any source files -- just let the
sym links be made by the user.  This makes the tarball just slightly
smaller.  ;-)

This commit was SVN r2316.
 2004-08-26 20:28:36 +00:00
George Bosilca
d892202a71 Run in troubles with the ompi_atomic_cmpset_32 function. Give a quick hack but have to talk with David about that. 
This commit was SVN r2315.
 2004-08-26 20:20:10 +00:00
Vishal Sahay
3b23ef1084 F77 bindings 
This commit was SVN r2314.
 2004-08-26 19:23:18 +00:00
Weikuan Yu
f4167d7d49 -- Create two threads per PTL, one for send/recv 
This commit was SVN r2313.
 2004-08-26 18:45:48 +00:00
Jeff Squyres
cc0fc842f9 Check to see if you don't have lex/flex and don't have the .c file 
that was generated by it (i.e., if you're in a developer checkout or
if you're in a distribution tarball).  If you don't have the .c file
and lex/flex, error.

This commit was SVN r2312.
 2004-08-26 16:17:20 +00:00
Tim Woodall
af28551365 removed bogus entry in struct initializer 
This commit was SVN r2311.
 2004-08-26 14:49:15 +00:00
Weikuan Yu
78b5f1abcd update some routine names 
This commit was SVN r2310.
 2004-08-26 08:48:35 +00:00
Weikuan Yu
262331690c Add stub threading functions 
This commit was SVN r2309.
 2004-08-26 08:30:29 +00:00
Jeff Squyres
b053a53823 - In an attempt for better compile times (i.e., less dependencies), 
strip down cmd_line.h; remove necessity of including
  include/constants.h and util/argv.h.
- This means that a bunch of other files in the source tree that
  depended on those files having already been included needed to be
  updated to include the relevant file(s)
- In fixing these (^^), added both /* $HEADER$ */ and #include
  "ompi_config.h" in a few files that were missing them.
- change from ompi_cmd_line_create() / ompi_cmd_line_free() to
  OBJ_NEW() / OBJ_RELEASE()
- update mpirun2.cc to allow "-np"
- fix minor problem in mpirun2.cc where two command line params were
  accepting "-h"
- slightly revamped cmd_line.c to use the OBJ_NEW/OBJ_RELEASE
  interface.  Added new function ompi_cmd_line_make_opt3() that allows
  "single dash" parameters.
- Greatly expanded the ompi_cmd_line_t doxygen docs.
- The diff for ompi_cmd_line_t is artificially large because I also
  took the opprotunity to reformat to 4 space tabs.  Sorry.  :-\

This commit was SVN r2308.
 2004-08-26 07:56:45 +00:00
Jeff Squyres
ca8768ef53 A few more changes for older perl versions 
This commit was SVN r2307.
 2004-08-25 18:37:14 +00:00
Jeff Squyres
f578d8cf52 Updates so that it will work with older versions of perl 
This commit was SVN r2306.
 2004-08-25 18:31:43 +00:00
Tim Woodall
3fd97e3c82 initialize local variable before use 
This commit was SVN r2305.
 2004-08-25 18:10:42 +00:00
George Bosilca
f9114b22a6 Removing the complaint about the attribute functions when we compile a C++ file. 
Second level commit log: "The Great and Powerful Jeff Squyres figured out and fixed this bug remotely. All praise Jeff." ;) yup yup

This commit was SVN r2304.
 2004-08-25 17:44:06 +00:00
Tim Woodall
3d754c4941 - modifications to oob to support bringing up oob before anything else (ns) 
- changed ns/gpr to register callback in component init (and ignore errors if not supported)

This commit was SVN r2303.
 2004-08-25 17:39:08 +00:00
Weikuan Yu
cbb53f305d -- Change the calling semantics for elan4_alloc_cmdq again 
-- Still need to verify the exact version numbers.

This commit was SVN r2302.
 2004-08-25 17:11:28 +00:00
Tim Woodall
db1eb110cc fix my misuse of #ifdef 
This commit was SVN r2301.
 2004-08-25 14:55:35 +00:00
Weikuan Yu
2e000b4a1b --Adapt to the changes of atomic interface 
This commit was SVN r2300.
 2004-08-25 14:26:16 +00:00
Tim Woodall
514f2b7d16 fix issue w/ use of #ifdef rather than #if 
This commit was SVN r2299.
 2004-08-25 14:24:16 +00:00
Weikuan Yu
1aa1ebd721 -- Introduce configure options for customized path to kernel header files 
-- Adjust routine synmatics according to the different qsnet version.
-- Remove an unnecessary reference to tcookie field in ELAN4_PRIVATESTATE

This commit was SVN r2298.
 2004-08-25 13:23:13 +00:00
Jeff Squyres
05b3b8d8f0 Remove some unused symbols, and therefore remove some compiler warnings 
This commit was SVN r2297.
 2004-08-25 11:52:40 +00:00
Jeff Squyres
15c01284e6 Remove unused variables to eliminate compiler warnings 
This commit was SVN r2296.
 2004-08-25 11:49:19 +00:00
Jeff Squyres
b5b873bb59 Eliminate some compiler warnings 
This commit was SVN r2295.
 2004-08-25 11:46:56 +00:00
Jeff Squyres
6c47f6823e - No, I really do want that debug output. :-) 
- Make the "uname" stderr go to /dev/null -- it's not the same on all
  platforms (grumble)

This commit was SVN r2294.
 2004-08-25 11:39:03 +00:00
Jeff Squyres
e4c0c75a20 - Fix an infinite loop (oops, forgot ++$i) 
- Remove debugging output
- Minor re-ordering of e-mail report

This commit was SVN r2293.
 2004-08-25 11:02:39 +00:00
Jeff Squyres
844f043ffa /usr/bin/env is portable; /bin/env is not :-) 
This commit was SVN r2292.
 2004-08-25 08:31:48 +00:00
Ralph Castain
83db216e40 Reserves a spot for the console program - no code in it yet. Attempts to fix earlier problem when compiling/linking the universe_setup_file_io routines. Will test further on new checkout - appears to work fine under OSX right now. 
This commit was SVN r2291.
 2004-08-25 02:01:35 +00:00
Ralph Castain
ee1f0b13f4 Add a new function - synchro - that allows the caller to be notified when a specified number of objects have been placed on a segment. The tokens describing the objects must also be provided, which means you can intermix objects on the segment and still be notified when you reach the specified number of objects meeting your description. 
Code compiles, but has not been functionally validated yet.

This commit was SVN r2290.
 2004-08-25 01:59:36 +00:00
Jeff Squyres
7d4905cf5b - Remove a few debugging lines 
- Fix problem with builds with no configure options

This commit was SVN r2289.
 2004-08-24 22:27:54 +00:00
Rich Graham
2ed427505c get the code in ompi_fifo to compile. 
This commit was SVN r2288.
 2004-08-24 22:25:59 +00:00
Jeff Squyres
4a1a26c35b Revamped test_build.pl script to be a bit more robust. Also committed 
sample configuration file (the comments are much longer than the
configurations! :-).

This commit was SVN r2286.
 2004-08-24 21:30:42 +00:00
Rich Graham
21402b68bb change circular buffer fifo calls to reflect changes. 
This commit was SVN r2285.
 2004-08-24 20:18:29 +00:00
Weikuan Yu
979a3c1165 Commit more code for shared send completion queue support 
This commit was SVN r2284.
 2004-08-24 20:17:57 +00:00
Rich Graham
8e348220eb add code to use link lists of circular buffer fifo's. 
change some of the ompi_circular buffer fifo's to provide
functionality needed for this.  Reviewed by Ralph and Denis.
ompi_fifo.h not yet tested.

This commit was SVN r2283.
 2004-08-24 20:17:44 +00:00
Brian Barrett
00e29e92ed * clarify some comments for the pcm parts of the rte interface. No code 
changes.

This commit was SVN r2281.
 2004-08-24 19:12:59 +00:00
Ralph Castain
c79ecdbd8e Add a new function to the name server - assign_cellid_to_process - that looks up the cellid for a given process and inserts it into the provided process name. 
Well....at least, that's what it will do eventually! For now, it just sticks a zero in the field.

Hey - at least it does *something*! :-)

This commit was SVN r2280.
 2004-08-24 19:05:04 +00:00