1
1
Граф коммитов

2715 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
57d09c10f7 Avoid some compiler warnings that come up ''every day'' in MTT (and
have been for eons): make a symbol be used in a dumb but harmless way.

This commit was SVN r15158.
2007-06-21 15:42:06 +00:00
Gleb Natapov
b88b7dedfe Rename btl_rdma_offset to btl_pipeline_send_length.
This commit was SVN r15153.
2007-06-21 07:12:40 +00:00
Jeff Squyres
84487f5c4b Update and correct the help messages for the generic BTL MCA
parameters.  Hopefully, they now make more sense to the mostly naieve
user...

This commit was SVN r15147.
2007-06-20 16:37:50 +00:00
George Bosilca
1772229535 Make the error messages a bit more clear (i.e. don't merge the send and receive
queues).

This commit was SVN r15141.
2007-06-19 22:34:20 +00:00
Rainer Keller
8d24934a80 - Add the missing parts: add MPI_REAL2 to the end of the list
of Fortran datatypes (mpif-common.h) and the list of registered
   datatypes: MOOG(REAL2).
   Configure and Compilation with ia32/gcc just finished, naturally
   without real2.

This commit was SVN r15137.
2007-06-19 20:41:28 +00:00
Jeff Squyres
930a9b7682 Make the help messages for if_include/if_exclude a little better.
This commit was SVN r15134.
2007-06-19 13:38:58 +00:00
Rainer Keller
d3372729bb - Support for opt. MPI_REAL2 (who has that?)
to make checks for MPI-implementations fail in the right way ,-]
 - check in configure.ac
 - BINARY INCOMPATIBLE change to mpif-common.h
   (if implemented the *right* way)
   Actually OMPI_F90_CHECK takes two arguments, not three.
 - Only have corresponding C-Type, if the opt. Fortran
   type is really supported,
   Otherwise pass ompi_mpi_unavailable to DECLARE_MPI_SYNONYM_DDT;
 - Reviewed by George and Jeff

This commit was SVN r15133.
2007-06-19 05:03:11 +00:00
Josh Hursey
7fd1805e97 Fix a couple of compile warnings that Tim P brought to by attention.
This commit was SVN r15132.
2007-06-19 00:46:16 +00:00
Rainer Keller
1feb5fb21a - Initializaton fixes of structure (o_f_to_c_index)...
- Mainly indentation, except for ompi_op_create, here just
   dont nest into ifs...

This commit was SVN r15131.
2007-06-18 23:03:56 +00:00
Rainer Keller
4a462eed3d - Should make declare only, when we do have long double.
This commit was SVN r15130.
2007-06-18 22:59:21 +00:00
Rainer Keller
9b91ac2fa3 Thanks to Daniel Spangberg for noticing this problem in the f77
interface:
 - Fix the handling of MPI_BOTTOM in various places
   Update of r15030
 - While being at it, handle MPI_IN_PLACE in the same way.
   Convert OMPI_ADDR -> OMPI_F2C_BOTTOM
   Convert OMPI_IN_PLACE -> OMPI_F2C_IN_PLACE
   and have them converted before the actual call.
 - Approved by George and tested with icc and simple f77 mpi-program and
   with program by Daniel.

This commit was SVN r15129.

The following SVN revision numbers were found above:
  r15030 --> open-mpi/ompi@15f9e58c68
2007-06-18 22:33:54 +00:00
Rainer Keller
858735c53c - Get rid of temporary variable -- we dont check for error and
therefore dont need it.

This commit was SVN r15128.
2007-06-18 21:34:12 +00:00
Rainer Keller
0cd2f99711 - Get rid of pendatic C++ namespace warning.
This commit was SVN r15127.
2007-06-18 21:29:09 +00:00
Gleb Natapov
643037907f Convert all #ifdef OMPI_ENABLE_DEBUG to #if.
This commit was SVN r15117.
2007-06-17 07:14:47 +00:00
George Bosilca
10a017d1bf For a obscure reason this have to be defined on Windows. The obscure
reason it's that we don't have the nice configure stuff, so detecting
when to enable the CR PML it's kind of hard. Keep it defined and at
least it compile smoothly.

This commit was SVN r15116.
2007-06-17 05:01:09 +00:00
George Bosilca
ceb8abe9c1 OMPI_ENABLE_DEBUG require an #if not an #ifdef
This commit was SVN r15107.
2007-06-15 19:22:19 +00:00
Josh Hursey
6cdfefad87 Fix portals BTL and cnos RML.
Both were failing due to interface changes that were never 
applied to them properly.

This commit was SVN r15082.
2007-06-14 18:49:41 +00:00
Jeff Squyres
2399b9a535 Ensure to initialize the variable so that we don't segv.
This commit was SVN r15078.
2007-06-14 13:59:28 +00:00
Gleb Natapov
7b9ae49fe1 This time correctly calculate local BTL rank among all BTLs in a subnet.
This commit was SVN r15073.
2007-06-14 10:27:11 +00:00
Jeff Squyres
1e18265c16 Bring over the functionality from the /tmp/jnysal-openib-wireup
branch:

 * Support btl_openib_if_include and btl_openib_if_exclude MCA
   parameters, similar to those supported by other BTLs.  Each take a
   comma-delimited lists of identifiers.  Identifiers can be HCA
   interface names (e.g., ipath0, mthca1, etc.)  or an HCA interface
   name and port numbers (e.g., ipath0:1, mthca1:2, etc.).  It is an
   error to specify both _include and _exclude.  If you specify a
   non-existant (or non-ACTIVE) HCA and/or port, you'll get a warning
   unless you disable the warning by setting the MCA parameter
   btl_openib_warn_nonexistent_if to 0.
 * Start updating to use BEGIN_C_DECLS and END_C_DECLS
 * A few other minor fixes that were picked up along the way.

This commit was SVN r15063.
2007-06-14 01:59:25 +00:00
Rainer Keller
ca09aae2cc - Get PERUSE compile again with latest RDMA changes in r14768/r14842.
This commit was SVN r15042.

The following SVN revision numbers were found above:
  r14768 --> open-mpi/ompi@3401bd2b07
  r14842 --> open-mpi/ompi@10266fb467
2007-06-13 12:47:47 +00:00
Gleb Natapov
8164723014 Allow to configure bandwidth and latency with finer granularity.
Set bandwidth for all ports of mthca0:
--mca btl_openib_bandwidth_mthca0 1000

Set bandwidth for port 1 of mthca1:
--mca btl_openib_bandwidth_mthca1:1 1000

Set latency for port 2 lid 123 on mthca0:
--mca btl_openib_latency_mthca0:2:123 20

This commit was SVN r15041.
2007-06-13 12:47:38 +00:00
Gleb Natapov
5c3f511451 Properly determine btl's rank among all btls withing the same subnet.
This commit was SVN r15038.
2007-06-13 11:15:58 +00:00
Brian Barrett
b71b2b4b0d Make the aio detection work with cross compiling. The tests no longer
even look at the status code and basically guarantee that the aio
function was never called, so there's really no point in AC_TRY_RUN
over AC_COMPILE_IFELSE...

This commit was SVN r15033.
2007-06-13 03:16:32 +00:00
George Bosilca
15f9e58c68 Convert all buffers from Fortran to C. This allow us to catch the
MPI_BOTTOM. Let the C functions do the checkings for where MPI_BOTTOM
is allowed and where not.

This commit was SVN r15030.
2007-06-12 23:13:56 +00:00
Brian Barrett
84d1512fba Add the potential for doing some basic error checking on mutexes during
single threaded builds.  In its default configuration, all this does
is ensure that there's at least a good chance of threads building
based on non-threaded development (since the variable names will be
checked).  There is also code to make sure that a "mutex" is never
"double locked" when using the conditional macro mutex operations.
This is off by default because there are a number of places in both
ORTE and OMPI where this alarm spews mega bytes of errors on a
simple test.  So we have some work to do on our path towards
thread support.

Also removed the macro versions of the non-conditional thread locks,
as the only places they were used, the author of the code intended
to use the conditional thread locks.  So now you have upper-case
macros for conditional thread locks and lowercase functions for
non-conditional locks.  Simple, right? :).

This commit was SVN r15011.
2007-06-12 16:25:26 +00:00
Galen Shipman
8e7cce813e don't update MPI_ERROR
This commit was SVN r15004.
2007-06-11 21:40:29 +00:00
Galen Shipman
406b05bdc3 update copyright..
This commit was SVN r15003.
2007-06-11 21:17:49 +00:00
Galen Shipman
798cc2c5b8 handle MPI_STATUS_IGNORE in iprobe for the MTLs
This commit was SVN r15002.
2007-06-11 20:19:31 +00:00
Brian Barrett
27ad954265 Fix a couple of problems with the way we were using orte_process_name_t
structures in the system.  Instead of using memcmp, use the ns function.
This won't cause a problem as long as all three elements of the name are
ints, but if they have different sizes, alignment and padding rules
can cause memcmp() to compare padding space, which rarely holds a sane
value.

This commit was SVN r14998.
2007-06-11 19:12:11 +00:00
George Bosilca
e2dd0a50fc A better version alowing for multi-rails or clusters of clusters. A lot of cleanups.
This commit was SVN r14963.
2007-06-08 20:37:20 +00:00
George Bosilca
c66cf32ee2 Cleaning up. Removing all unused variables and fields in the MX BTL and
component structures.

This commit was SVN r14957.
2007-06-07 21:02:18 +00:00
George Bosilca
5d6c958066 Enable the MTLs to be compiled in a visibility featured environment.
This commit was SVN r14955.
2007-06-07 20:14:53 +00:00
Gleb Natapov
423f404c34 Shut up compiler warning. Ugly, but I can see better way except changing
converter to use uint64_t(ssize_t?) for offset.

This commit was SVN r14950.
2007-06-07 11:33:28 +00:00
Gleb Natapov
9f9b64db4e Revert r14947 as this doesn't solve the problem.
This commit was SVN r14949.

The following SVN revision numbers were found above:
  r14947 --> open-mpi/ompi@5b9fe28e3f
2007-06-07 11:24:24 +00:00
Gleb Natapov
5b9fe28e3f Fix warning on 32bit systems.
This commit was SVN r14947.
2007-06-07 08:57:34 +00:00
Tim Prins
06bf4c3f3b fix some printf warnings
This commit was SVN r14934.
2007-06-06 22:37:26 +00:00
George Bosilca
6a5e039466 Allow smart connection to be setup. Each peer now has attached to it thea unique
id based on the last half of the mapper MAC. This allow us to figure out how
to connect peers. This allow the MX BTL to be used in a cluster of cluster 
configuration where each cluster have MX internally as well as on a multi
rail MX system.

This commit was SVN r14932.
2007-06-06 21:42:11 +00:00
Galen Shipman
5340f5e320 Try to cleanup the flow control logic a bit
Renamed a few variables 
Inialize the reserve receive buffers to 1, prior to this they were initialized
to zero. 

This commit was SVN r14919.
2007-06-06 18:51:09 +00:00
Brian Barrett
7c9bae5eb7 need to protect these includes as well...
This commit was SVN r14912.
2007-06-06 16:25:18 +00:00
Gleb Natapov
de58336c45 Let rdma_pipeline_offset to be set to zero.
This commit was SVN r14900.
2007-06-06 11:54:25 +00:00
George Bosilca
768ac4a0d8 Allow creation of a datatype from a packed description on 64 bits environment.
Correct the alignment macros.

This commit was SVN r14898.
2007-06-06 07:30:34 +00:00
Rich Graham
e276f7bcc7 undo my error.
This commit was SVN r14890.
2007-06-05 23:32:47 +00:00
George Bosilca
6ccffb0d3e Handle all convertor flags in one common place. Cleanup a little bit the
flags handling.

This commit was SVN r14889.
2007-06-05 23:26:07 +00:00
Bill D'Amico
9b5f73976d Bring Portable Linux Processor Affinity into trunk.
Changes paffinity interface to use a cpu mask for available/preferred cpus
rather than the current coarse grained paffinity that lets the OS choose
which processor.

Macros for setting and clearing masks are provided.

Solaris and windows changes have not been made. Solaris subdirectory has some
suggested changes - however the relevant man pages for the Solaris 10 APIs
have some ambiguity regarding order in which one create and sets a processor
set. As we did not have access to a solaris 10 machine we could not test to
see the correct way to do the work under solaris.

This commit was SVN r14887.
2007-06-05 22:07:30 +00:00
Rich Graham
ce0e9ac329 initialize lock properly.
This commit was SVN r14881.
2007-06-05 20:34:11 +00:00
Donald Kerr
8ecbc71ed2 add support for connection private data, off by default
This commit was SVN r14878.
2007-06-05 19:29:50 +00:00
George Bosilca
ee5552249c Dont set the same flag twice.
This commit was SVN r14875.
2007-06-05 18:01:34 +00:00
George Bosilca
df5394cdc2 Rich found a double call to the _PREPARE macro. Removing it give us some
performance improvements.

This commit was SVN r14873.
2007-06-05 15:41:29 +00:00
Shiqing Fan
efa74f7bfe The label name "ERROR" is defined as a flag in Visual Studio platformSDK. Using "ERROR" as a label causes conflicts. Changing it into lowercase will solve the problem.
This commit was SVN r14869.
2007-06-05 14:32:27 +00:00