ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
3 251 коммит 3 Ветки 0 Теги 130 MiB
48b0b10464
Граф коммитов

3251 Коммитов

This Branch
This Branch
Все ветки
Автор SHA1 Сообщение Дата
George Bosilca
d3df61dc9e Do not always destroy the first status. More and more tests get passed. 
This commit was SVN r3029.
 2004-10-10 01:13:51 +00:00
George Bosilca
a62dafec84 Behave as expected by the MPI layer. 
Use the defines from the mpi.h
Release the correct request.
Set the correct status
And return MPI_UNDEFINED if all requests are MPI_REQUEST_NULL.

This commit was SVN r3028.
 2004-10-10 01:12:42 +00:00
George Bosilca
950e46e086 Correct a bug on the CHECK_DDT macros. Now the datatype.h match what's expected by the mpi layer. 
This commit was SVN r3022.
 2004-10-10 00:22:23 +00:00
Jeff Squyres
722e51aa2c Oops -- assignment was the wrong way around. 
This commit was SVN r3021.
 2004-10-10 00:03:42 +00:00
Jeff Squyres
0021b2e7ce Fix typos in the default help message. 
No need to re-run autogen.sh -- this will be picked up the next time
you naturally run autogen.sh.

This commit was SVN r3020.
 2004-10-09 23:29:11 +00:00
George Bosilca
152b58de04 Handle the 0 count copy_ddt_to_ddt. In same time I add some small optimizations. 
This commit was SVN r3018.
 2004-10-09 21:40:15 +00:00
George Bosilca
356cc37c13 Handle the 0 count communications through the self module. 
This commit was SVN r3017.
 2004-10-09 21:38:42 +00:00
George Bosilca
c19d1d330e Correct the get_count function. Now we pass a lot more tests. 
This commit was SVN r3016.
 2004-10-09 21:29:34 +00:00
Jeff Squyres
143e83d3f2 I screwed up on the commit earlier :-( -- am rolling back until I have 
time to fix.  Sorry folks...

This commit was SVN r3015.
 2004-10-09 21:29:23 +00:00
George Bosilca
1777c361e9 If the rc variable is not initialized then there is a execution path who return whatever is in rc. Of course everything fail afterward as rc NEVER contain what we really expect by default. 
This commit was SVN r3014.
 2004-10-09 21:20:04 +00:00
Jeff Squyres
7bd8f14978 Fix minor logic error for when you run autogen.sh in a directory other 
than the top-level OMPI directory.

This change does *NOT* mean that anyone needs to re-run autogen.sh
right now.

This commit was SVN r3013.
 2004-10-09 14:46:51 +00:00
Jeff Squyres
9bb95ce949 After thinking about this a little more on the plane, my hackaround 
was actually the correct solution.  So change the comment to reflect
why that #if is there, and why it is correct.

This commit was SVN r3012.
 2004-10-09 14:42:27 +00:00
Rich Graham
eaae910b5b add matching logic that assumes in-order delivery with no 
dropped packets (shared memory, for example).  Reviewed by Tim W.

This commit was SVN r2999.
 2004-10-08 22:24:57 +00:00
Jeff Squyres
6b48c878a1 Arf! Didn't check this *without* enabling the fortran bindings. 
This is a hackaround so that people can compile; I need to move the
macro definitions elsewhere to do this right.  Will do so tonight.

This commit was SVN r2998.
 2004-10-08 19:24:12 +00:00
Tim Woodall
dded02fdcb add appropriate cast 
This commit was SVN r2997.
 2004-10-08 18:18:07 +00:00
Jeff Squyres
80b38390ab Enable proper f2c <--> c2f MPI_Request translation. Pick up f2c <--> 
c2f MPI_Status translation along the way.  This should enable Fortran
MPI apps that use non-blocking communication to start working.

This commit was SVN r2996.
 2004-10-08 17:12:36 +00:00
Tim Woodall
530dbf5b1e slight reorder of shutdown 
This commit was SVN r2995.
 2004-10-08 17:03:40 +00:00
Tim Woodall
76f98a54d8 use consistent "hex" format for jobid in registrations 
This commit was SVN r2994.
 2004-10-08 17:00:57 +00:00
Brian Barrett
96b3c503f9 * Add useful error messages to bootproxy for when it fails 
* don't push pcmclient env variables to clients, since they don't have useful
  information in them
* add check to comm_spawn code to make sure we can spawn

This commit was SVN r2993.
 2004-10-08 16:22:35 +00:00
George Bosilca
fc0797ab9d Correctly handle synonyms (like MPI_REAL4 <=> MPI_FLOAT). 
Correctly define the bool type and the CXX_*complex.
Remove all memory leaks ...

This commit was SVN r2983.
 2004-10-07 21:39:13 +00:00
Prabhanjan Kambadur
4c9f883a9b Changing the logic in ompi_comm_split to pass on the topo information only when required. Also had to change other files to go along wit this. Tested the intel test suites (relevant tests ... MPI_Cart_sub_c MPI_Comm_split1_c MPI_Comm_split2_c ... everything seems to work fine 
This commit was SVN r2982.
 2004-10-07 19:14:27 +00:00
George Bosilca
ecd7ef0bb0 Add the destruction of attributed when we release the last ref on a datatype. But I still have a problem ... it does not really look right to me to need attribute/attribute.h as an include in the datatype ... I still have to investigate this one :) 
This commit was SVN r2970.
 2004-10-07 01:50:28 +00:00
George Bosilca
0b605049e9 Indentation of 4 spaces + a small change on the way we optimize the data. 
This commit was SVN r2969.
 2004-10-07 01:38:31 +00:00
Jeff Squyres
101a83ab5f Change attribute hashes to only be non-NULL on MPI objects that 
actually have attributes cached on them.  If there are no attributes
cached on the object, the keyhash will be NULL.

This commit was SVN r2968.
 2004-10-07 01:26:40 +00:00
George Bosilca
d710ed9a49 Correct a mistake that I made 2 days ago. 
This commit was SVN r2967.
 2004-10-07 01:22:58 +00:00
Jeff Squyres
03dc8e21fb Silence meaningless compiler warning 
This commit was SVN r2966.
 2004-10-06 23:40:30 +00:00
Rich Graham
145ce7d68c set pointer to fragment matching logic routine, if this 
not already set by the ptl - reviewed by tim.

This commit was SVN r2964.
 2004-10-06 22:10:43 +00:00
Rich Graham
725afde500 just initialize some of the function pointers to null, so that when 
mca_pml_teg_add_ptls is called, the value of NULL can be used to
determine if this needs to be set, or a previously set value needs
to be used.

This commit was SVN r2963.
 2004-10-06 22:06:52 +00:00
Rich Graham
7fef6135d9 just initialize some of the function pointers to null, so that when 
mca_pml_teg_add_ptls is called, the value of NULL can be used to
determine if this needs to be set, or a previously set value needs
to be used.

This commit was SVN r2962.
 2004-10-06 22:06:02 +00:00
Rich Graham
281411f698 just initialize some of the function pointers to null, so that when 
mca_pml_teg_add_ptls is called, the value of NULL can be used to
determine if this needs to be set, or a previously set value needs
to be used.

This commit was SVN r2961.
 2004-10-06 22:05:15 +00:00
Tim Woodall
93a54d248f init pml callback function pointers to null 
This commit was SVN r2960.
 2004-10-06 21:08:54 +00:00
Prabhanjan Kambadur
13d6aaaffe this is the complete list of files which are included using <> braces but are not present in windows 
This commit was SVN r2951.
 2004-10-06 15:31:59 +00:00
Jeff Squyres
f09bc8f488 Change <> to "" -- just for clarification and help in transmorgifying 
for the Windoze port

This commit was SVN r2950.
 2004-10-06 15:24:48 +00:00
Prabhanjan Kambadur
fd839da034 These are the files which show which are the local headers and which are the files which include them using <> brackets. These files can be deleted whenever wanted by anyone as it can be generated by find_offenders.pl 
This commit was SVN r2949.
 2004-10-06 06:35:39 +00:00
Prabhanjan Kambadur
2e1dc70fb9 Checks for the presence of header files which are from the lcal tree but are included as <> i.e., using angular brackets 
This commit was SVN r2948.
 2004-10-06 06:32:35 +00:00
Prabhanjan Kambadur
35e236c0f0 Changing the format of the file a bit to suit the perl file 
This commit was SVN r2947.
 2004-10-06 06:18:51 +00:00
George Bosilca
4b6f26f561 Did I forget to remove my debug printf statements ? 
This commit was SVN r2946.
 2004-10-06 03:55:12 +00:00
Jeff Squyres
95e5fa016d Oops -- typo in Makefile.am. 
This commit was SVN r2945.
 2004-10-06 03:48:06 +00:00
Prabhanjan Kambadur
07c10f1a67 This commit was SVN r2944. 2004-10-05 23:55:57 +00:00
Prabhanjan Kambadur
496be79013 this is the file containing a list of all headers which are currently used in ompi but not present in Windows. This is generated by test_headers_in_ompi.pl, but thought I would check it in so that one can run fix_headers.pl from Linux boxes as well 
This commit was SVN r2943.
 2004-10-05 23:55:18 +00:00
Prabhanjan Kambadur
8c5626cd66 Some perl scripts to fix headers in windows. Not yet perfect 
This commit was SVN r2942.
 2004-10-05 23:52:35 +00:00
Prabhanjan Kambadur
a8764a7155 Changing graph create 
This commit was SVN r2941.
 2004-10-05 23:25:59 +00:00
Prabhanjan Kambadur
ffbe23e01b This file really should not be here 
This commit was SVN r2940.
 2004-10-05 22:27:54 +00:00
Prabhanjan Kambadur
baf370d0b8 The silliest of all cut and paste errors :-( .... had forgotten to change from dest to source 
This commit was SVN r2939.
 2004-10-05 22:26:35 +00:00
Prabhanjan Kambadur
02958a5a4e Making changes to clean up creation of a cart which is smaller than its parent proc 
This commit was SVN r2938.
 2004-10-05 21:17:14 +00:00
Tim Woodall
52a9776ab1 updating docs 
This commit was SVN r2937.
 2004-10-05 20:58:01 +00:00
Jeff Squyres
8961189cbf It's amazing how you can read the diff ahead of time and think "yes, 
that's good.  Go ahead -- commit that."  But then as soon as you read
the diff on the mailing list, you realize how borked it was.  [sigh]
- add MPI_INTEGER16 at George's request
- don't forget to commit acinclude.m4
- correctly name the REAL checks (accidentally had 2 sets of INTEGER
  checks)

This commit was SVN r2936.
 2004-10-05 19:38:56 +00:00
Jeff Squyres
1a076a5a17 Oops -- remove debugging output. #$%@#$% 
This commit was SVN r2935.
 2004-10-05 19:34:34 +00:00
Jeff Squyres
df98ca0010 Add some checks for optional fortran datatypes, required by the ddt 
engine

This commit was SVN r2934.
 2004-10-05 19:31:18 +00:00
Prabhanjan Kambadur
4655fe3943 Correcting code which allows topo to function properly ... also adding code to make comm_split and comm_dup work properly 
This commit was SVN r2933.
 2004-10-05 16:32:52 +00:00