Rainer Keller
ea77f52d89
Set the softlink to the correct source file, when configuring
...
in a different then the toplevel OpenMPI directory.
This commit was SVN r4131.
2005-01-25 18:41:23 +00:00
Jeff Squyres
4ee895c243
When on Windows systems, use ";" as a path delimiter
...
This commit was SVN r4125.
2005-01-24 22:53:28 +00:00
Prabhanjan Kambadur
e0e47f37a4
Fixing the hardcoding of TMPDIR to the windows defualt
...
This commit was SVN r4124.
2005-01-24 22:51:15 +00:00
Rainer Keller
7d37584e67
Allow MPI_ANY_TAG to probe for when MPI_PARAM_CHECKing.
...
This commit was SVN r4123.
2005-01-24 22:23:15 +00:00
Rainer Keller
1d8b5fa865
Added the top_build_dir/include directory for compilation for ompi_config.h
...
when compiling out of the main src-dir.
This commit was SVN r4119.
2005-01-24 19:02:47 +00:00
Craig E Rasmussen
ea6163cd6b
Added choice annonotation to MPI_Pack arguments.
...
This commit was SVN r4117.
2005-01-22 21:06:05 +00:00
George Bosilca
37721f96a3
Correctly handle the homogeneous pack. We support multiple iovec, and we respect the ioc's size as well as the total amount that should be packed as requested by the user. Several minor bugs have been removed.
...
This commit was SVN r4116.
2005-01-22 20:15:28 +00:00
George Bosilca
5ca7b580c4
Remove useless code.
...
This commit was SVN r4115.
2005-01-22 20:13:51 +00:00
George Bosilca
af5c7a34bd
Add a pack/unpack test. There is no validation of the operation yet but I will add it in few days.
...
This commit was SVN r4114.
2005-01-22 20:13:19 +00:00
George Bosilca
11b21152ba
Remove a useless (and additionally incorrect) assert.
...
This commit was SVN r4113.
2005-01-22 20:12:29 +00:00
George Bosilca
6fcc85a70f
Indentation + set the maximum amount of data packed/unpacked to zero if we call the convertor after
...
we finish the conversion.
This commit was SVN r4112.
2005-01-22 20:11:59 +00:00
George Bosilca
e8a418830b
Change the name of the DUMP_STACK to DDT_DUMP_STACK
...
Remove all useless labels.
Remove all "unsigned compared to signed" warnings
Chack that the datatypes are correctly committed before any usage of pack/unpack/copy functions (The reason is that
on the commit step we add a fake record at the end of the datatype representation. This record contain a END_LOOP
corresponding to the count used for the convertor. This allow us to remove few tests in the critical path, and to
slighty improuve the performances of these functions).
Remove useless code.
Correct the ompi_ddt_copy_to_same_ddt function (now the exit condition is correctly detected in all cases).
Correct the ompi_ddt_dump_stack function (The stack is already pointing to the current position not to the begining
of the stack).
This commit was SVN r4103.
2005-01-21 21:07:14 +00:00
George Bosilca
8961ca568d
Enable all the test cases, and correct the test regarding to the RELEASE bug (cannot release a datatype that have been
...
used to create another datatype).
This commit was SVN r4102.
2005-01-21 21:02:03 +00:00
George Bosilca
9b7a66ec87
Correct the while exit condition.
...
This commit was SVN r4101.
2005-01-21 21:00:51 +00:00
Ralph Castain
cd00fed56b
Bring the error manager over to the main branch so people can start to use it in their code. Please see mca/errmgr/errmgr.h for the documentation on this function - very little implemented at the moment. Only function so far is to output the error message, file, and line via the ompi_output function.
...
Note: in bringing this across, I have obeyed the mca-prefix-rule in defining functions and variables. However, I left the name of the component call as "orte_errmgr.xxx" so that any use of the code will directly "merge" with the new RTE branch without changes.
This commit was SVN r4099.
2005-01-21 17:49:14 +00:00
Jeff Squyres
76f2125c98
Fix some compiler warnings (specifically, from icc v8.0)
...
This commit was SVN r4093.
2005-01-21 13:54:38 +00:00
Jeff Squyres
46d7f9a7ed
Ensure that compilers that pretend to be gcc don't actually get the
...
gcc-specific compiler flags (because they don't work -- @#$%@#$%!!!)
This commit was SVN r4092.
2005-01-21 13:53:48 +00:00
Jeff Squyres
4f0adc3ab2
- Add a missing #define
...
- Change the params of 2 functions to match what they're always called
with
This commit was SVN r4091.
2005-01-21 13:52:57 +00:00
Prabhanjan Kambadur
824eed35d0
Forgot to add socket_errno.h to Makefile.am
...
This commit was SVN r4079.
2005-01-20 15:39:24 +00:00
Prabhanjan Kambadur
15acf6b07a
Moving the comments to the right places
...
This commit was SVN r4073.
2005-01-20 04:44:07 +00:00
Prabhanjan Kambadur
f03bb71273
For some reason forgot to include the comments properly
...
This commit was SVN r4071.
2005-01-20 00:40:22 +00:00
Prabhanjan Kambadur
9ac9f15537
These are the changes after the review with Jeff. Mostly are fixes for OOB and TCP
...
This commit was SVN r4070.
2005-01-20 00:03:23 +00:00
Tim Woodall
b922a56938
cleanup warnings
...
This commit was SVN r4044.
2005-01-18 21:31:33 +00:00
Tim Woodall
77c87b34d6
shut up warnings
...
This commit was SVN r4043.
2005-01-18 21:30:23 +00:00
Craig E Rasmussen
12f7bf3737
Completed annotations for functions with choice arguments.
...
This commit was SVN r4042.
2005-01-18 21:28:12 +00:00
Jeff Squyres
962bfa796f
Remove pointless comparison (size_t is unsigned, so it's always >=
...
0).
This commit was SVN r4037.
2005-01-18 20:24:13 +00:00
Tim Woodall
0d2784568e
fixed I/O forwarding cleanup
...
This commit was SVN r4033.
2005-01-18 17:32:54 +00:00
Tim Woodall
4bd9538bca
temporarily disable I/O forwarding
...
This commit was SVN r4031.
2005-01-18 16:43:47 +00:00
Craig E Rasmussen
6323a0ac4f
Initial version.
...
This commit was SVN r4030.
2005-01-18 16:15:55 +00:00
Edgar Gabriel
e62c6a4fd6
the parameter list for ompi_parse_port hast changed again back to the original version
...
This commit was SVN r4028.
2005-01-18 13:56:21 +00:00
Edgar Gabriel
ba31e74ec3
removing the current version of the mpi_appnum paramter, since it will be handled now using the registry
...
This commit was SVN r4027.
2005-01-18 13:55:48 +00:00
George Bosilca
72a28e7796
Remove the compilation warnings. Change the my_smp_rank to uint32_t and it's defaul
...
t value from -1 to 0xffffffff. Remove the volatile from the locks definition, as th
ey are useless ie. only the internal integer should be volatile (and that's already the case).
Left something inside for the SM maintainer:
- one warning in ptl_sm.c. For me it look like a bug (line 856). The PTL does not h
ave to signal that the request is completed. That's the PML job. Anyway, the ompi_r
equest_complete is called with a mca_pml_base_send_request_t, and that's not exactl
y how it should work.
- in ptl_sm.c line 122 there is a definition (volatile char** tmp_ptr). As far as I
understand the code it's not the char** who should be volatile, but the pointer to
char ...
This commit was SVN r4024.
2005-01-17 21:37:56 +00:00
George Bosilca
20ddf644e2
The standard C (ISO C) does not allow macros returning values. Switch to an inline function.
...
This commit was SVN r4023.
2005-01-17 21:26:17 +00:00
Brian Barrett
653b259d75
* add uio.h so that it compiles on OS X
...
This commit was SVN r4019.
2005-01-17 15:51:13 +00:00
Craig E Rasmussen
a48ff994e4
Fixed MPI_STATUS_SIZE usage to correctly dimension integer :: status variables. Also removed MPI_SIZEOF.
...
This commit was SVN r4017.
2005-01-15 16:36:48 +00:00
Craig E Rasmussen
aef0f3c503
Fixed MPI_STATUS_SIZE usage to correctly dimension integer :: status variables.
...
This commit was SVN r4016.
2005-01-15 16:36:13 +00:00
Jeff Squyres
b00522566b
- Make MPI_FINALIZE cleanup mpool properly
...
- Change mca_mpool_base_close() to know that mpool modules with a NULL
finalize should be skipped (i.e., it's legal)
- Fix a few function names in accorance to the prefix rule
This commit was SVN r4015.
2005-01-15 13:20:26 +00:00
Jeff Squyres
91b186234d
Add more missing .h files to Makefile.am's so that distribution
...
tarballs can be built.
This commit was SVN r4014.
2005-01-15 03:33:02 +00:00
Craig E Rasmussen
eaf6f99d61
Continuing improvements, mostly due to testing and adding problem procedures.
...
This commit was SVN r4013.
2005-01-14 23:47:39 +00:00
Craig E Rasmussen
de03176303
Added logical types.
...
This commit was SVN r4012.
2005-01-14 23:45:21 +00:00
Craig E Rasmussen
95a4b10d94
Added logical types and removed unused variable.
...
This commit was SVN r4011.
2005-01-14 23:44:33 +00:00
Craig E Rasmussen
dac75b0909
Added logical and rank 0 (elemental) types.
...
This commit was SVN r4010.
2005-01-14 23:11:50 +00:00
Craig E Rasmussen
f1634ae698
Added logical and rank 0 (elemental) types. Also removed output of methods with @kind='No_F90' (2 choice args, for example).
...
This commit was SVN r4009.
2005-01-14 23:03:36 +00:00
Craig E Rasmussen
ae9dcc2ab7
Added logical and rank 0 (elemental) types.
...
This commit was SVN r4007.
2005-01-14 22:07:15 +00:00
Jeff Squyres
a057a89943
Ignore "compile" script (from Automake)
...
This commit was SVN r4006.
2005-01-14 21:59:04 +00:00
Jeff Squyres
15f0cb6112
Use the right type when closing down the mpool components.
...
This commit was SVN r4005.
2005-01-14 21:58:02 +00:00
Jeff Squyres
6e12d4c378
Add missing .h files into Makefile.am so that the tarball builds
...
correctly
This commit was SVN r4003.
2005-01-14 19:46:13 +00:00
Brian Barrett
1fb96505a4
* There are a couple of platforms that do dead code elimination on linking,
...
which was resulting in the pthreads functions never getting linked in.
Which would cause false positives on the threading tests. Make sure
that there is a call chain from main() -> pthread_create(), which should
be enough to make the test happy.
This commit was SVN r3998.
2005-01-14 16:08:30 +00:00
Tim Woodall
035b55bf81
cleanup threading issues
...
This commit was SVN r3996.
2005-01-14 15:43:15 +00:00
George Bosilca
95784408c4
Update the ompi_output to match the type of the arguments.
...
This commit was SVN r3995.
2005-01-14 05:29:27 +00:00