1
1
Граф коммитов

6944 Коммитов

Автор SHA1 Сообщение Дата
Jeff Squyres
2cceba7230 Fix to George's and my bugs. :-) A comedy of errors made the test
case still compile properly, even though these bindings were wrong.  :-(

- Both interface functions are necessary.
- I accidentally had a subroutine interface named the same thing as the
  back-end F77 function -- a definite no-no.  Ensure that all F90
  interface functions have a suffix to make them different than the
  back-end F77 function names.
- Also parameterize the output() subroutine on the type of the argvs
  parameter to switch between the character arrays and the integer.

This commit was SVN r9202.
2006-03-04 19:36:13 +00:00
George Bosilca
a649ebc7c6 MPI_Comm_spawn_multiple does not need to have 2 definitions. Moreover,
this function is not allowed to call itself recursively.

This commit was SVN r9201.
2006-03-04 19:22:55 +00:00
Jeff Squyres
a192af34e7 Shame on me for not looking at the expanded collective datatype tables
for the C++ bindings in MPI-2 p276-278 to see that MPI_BOOL should
work with MPI_LAND, MPI_LOR, and MPI_LXOR.  Thanks to Andy Selle for
pointing this out.

This commit was SVN r9200.
2006-03-04 18:35:33 +00:00
George Bosilca
bd109c90f5 One of the , was missing.
This commit was SVN r9199.
2006-03-04 17:41:57 +00:00
Jeff Squyres
9d119109b6 Change the types of some of the "special" constants in mpif.h so that
they will match the prototypes in the [styictly-typed] MPI F90
bindings.  Specifically, fix up MPI_COMM_SPAWN and
MPI_COMM_SPAWN_MULTIPLE so that the constants MPI_ARGV_NULL,
MPI_ERRCODES_IGNORE, and MPI_ARGVS_NULL can be used in the F90
bindings.  Thanks to Michael Kluskens for pointing this out to us.

Some work still remains in the F90 bindings -- we are missing all
places where choice buffers can be of type CHARACTER.

This commit was SVN r9198.
2006-03-04 14:24:44 +00:00
Rainer Keller
42f4a6fe03 - use AC_C_BIGENDIAN instead of deprecate AC_WORDS_BIGENDIAN
- set the highest debugging levels for macro debug support on gcc/gdb

This commit was SVN r9197.
2006-03-04 13:59:14 +00:00
Rainer Keller
9e1c5716b6 - opal_cube_dim does not return an error
This commit was SVN r9196.
2006-03-04 13:47:24 +00:00
Jeff Squyres
c31e515db1 Since we've had problems recognizing the "special" Fortran constants,
add an explicit test case that checks for them.  It is necessary to
put this test function here because the OMPI_IS* macros are only
defined in this directory.

This commit was SVN r9195.
2006-03-04 13:23:19 +00:00
Jeff Squyres
adc6753fa9 Check for ARGVS_NULL, not ARGV_NULL
This commit was SVN r9194.
2006-03-04 13:22:12 +00:00
Rainer Keller
5627fb5627 - Get rid of compiler warning: init type to DT_MAX_PREDEFINED
and assert it is below when accessing ompi_ddt_basicDatatypes
 - datatype in ompi_ddt_pack_description_length is and will not be
   changed.

This commit was SVN r9193.
2006-03-04 13:16:04 +00:00
Tim Woodall
274ee03df6 work in progress
This commit was SVN r9192.
2006-03-04 00:36:16 +00:00
George Bosilca
ca116e3cf3 Where we create a duplicate data-type for the get_content MPI function we have to attach the
args from the old data-type. As this args are only read-only, I don't want to have several
copies around, so I add a reference count to them and simply attach them to the new
datatype. The args array now get freed only when the last reference dissapear.

This commit was SVN r9191.
2006-03-03 20:51:36 +00:00
Galen Shipman
4e430b0428 fix warnings, other misc
This commit was SVN r9190.
2006-03-03 04:01:10 +00:00
Rainer Keller
b8d1189091 - Use the actual _count parameter passed through the macro
This commit was SVN r9189.
2006-03-02 20:23:07 +00:00
Tim Woodall
8bf6ed7a36 - corrected locking in gm btl - gm api is not thread safe
- initial support for gm progress thread
- corrected threading issue in pml
- added polling progress for a configurable number of cycles to wait for threaded case

This commit was SVN r9188.
2006-03-02 00:39:07 +00:00
Galen Shipman
84d3055db5 Make sure everything is imediatly acked, even if not matched
Buffer first descriptor on the sendreq until postive ACK 
Set bytes delivered only after postive ACK, removed num_acks, etc, in general
trying to remove as much state as possible so that rolling things back isn't
such a nightmare 

This commit was SVN r9187.
2006-03-01 22:37:10 +00:00
Brian Barrett
c2fded586a * Andrew F. pointed out I can't spell...
This commit was SVN r9186.
2006-03-01 03:13:22 +00:00
Brian Barrett
7fa1c0eeb2 * if the default compile mode is 64bit, config.guess will call a MIPS
machine mips64

This commit was SVN r9185.
2006-03-01 02:25:09 +00:00
Brian Barrett
1479a90b39 * assert() that endianness doesn't need to change if we are sending RDMA headers
around, since OB1 currently doesn't do the right thing there, but that should
  not happen in the near future because the R2 BML should not make any RDMA
  networks available between machines with different architectures
* Clean up the #ifs a little bit so that we don't do unneeded work when
  on big endian machines and heterogeneous support is disabled...

This commit was SVN r9184.
2006-02-28 19:54:46 +00:00
Jeff Squyres
80bc1850bf Ensure that --prefix takes precedence over /path/to/orterun
This commit was SVN r9183.
2006-02-28 14:44:40 +00:00
Rainer Keller
a2a26525b3 - Release the output-stream in mca_base_component_close
- Destroy the verbose-stream upon finalize, not when closing.
   Also free the output_prefix and output_dir variables.

This commit was SVN r9182.
2006-02-28 12:33:42 +00:00
Jeff Squyres
88b3e6f8bd - Fix bug in orterun where --prefix didn't show up in the help output
(reported by Cisco)
- While in orterun, add a feature that multiple users have asked for:
  if you specify an absolute pathname to orterun, such as
  "/path/to/bin/orterun ...", it's equivalent to "orterun --path
  /path/to ..."

This commit was SVN r9181.
2006-02-28 11:52:12 +00:00
Rainer Keller
8568bb9892 - Allow MPI_ERRCODES_IGNORE when checking is on.
This commit was SVN r9180.
2006-02-28 10:16:42 +00:00
Brian Barrett
579e74290f * make gm wire-up endian safe
This commit was SVN r9179.
2006-02-28 02:03:46 +00:00
Galen Shipman
d9fd35d399 add acked items to datastructure,
fix compile issue. 

This commit was SVN r9178.
2006-02-28 01:07:35 +00:00
Galen Shipman
c6b4cc4417 Add data structure to track ACK's
This commit was SVN r9177.
2006-02-27 22:56:43 +00:00
Brian Barrett
e865a751bd * First whack at making the onesided component endian safe. Needs a endian-safe
datatype engine to really give it a whirl ;).

This commit was SVN r9176.
2006-02-27 18:47:00 +00:00
Galen Shipman
db6b1db548 use pml level datatype, someone else already cleaned this up in ob1.
This commit was SVN r9174.
2006-02-27 18:20:49 +00:00
Galen Shipman
2aa7b129a6 don't use ptl datatypes!
This commit was SVN r9173.
2006-02-27 18:07:38 +00:00
Brian Barrett
bfd49d248b * (hopefully) fix MPI_BOTTOM for portals, same way as oll the other RDMA btls from
eons ago...

This commit was SVN r9172.
2006-02-27 17:07:24 +00:00
Brian Barrett
9b19e3fef0 * remove some debugging output that shouldn't have been committed. Doh!
This commit was SVN r9171.
2006-02-27 16:23:52 +00:00
Rainer Keller
24f4b282f8 - Free the segment string.
This commit was SVN r9170.
2006-02-27 16:07:35 +00:00
Rainer Keller
4b1194056f - Upon Finalize, also release the memory of predefined attributes.
This commit was SVN r9169.
2006-02-27 15:15:48 +00:00
Rainer Keller
af0002c794 - replace lam with ompi
This commit was SVN r9168.
2006-02-27 15:00:33 +00:00
Jeff Squyres
c12b38a13a Sync README with v1.0 README.
This commit was SVN r9166.
2006-02-27 11:42:21 +00:00
Rainer Keller
5102571c02 - Get rid of the temporary reachability bitmap.
This commit was SVN r9163.
2006-02-27 11:06:01 +00:00
Brian Barrett
ee73cbb25c * remove unused F90 range and precision tests
This commit was SVN r9162.
2006-02-27 05:19:13 +00:00
George Bosilca
426c587dcc Dont use exit(0) when we try to run a simple C++ program. Use return instead
to avoid having problems on Windows with the C++ mangling scheme.

This commit was SVN r9161.
2006-02-27 05:14:35 +00:00
George Bosilca
a76213f352 No group and lossy signals on Windows.
This commit was SVN r9160.
2006-02-27 05:11:44 +00:00
Jeff Squyres
4a0d9bd46f Turns out that $(MCA_io_romio_STATIC_LTLIBS) is a throwback to a
previous incarnation of the configure/build system and isn't defined
anymore.  So it can be removed.

This commit was SVN r9158.
2006-02-27 03:28:40 +00:00
Jeff Squyres
d068c1516c A few minor Makefile.am fixes
This commit was SVN r9157.
2006-02-27 03:18:57 +00:00
George Bosilca
0320551262 Update include path.
This commit was SVN r9156.
2006-02-27 02:15:53 +00:00
George Bosilca
7abccaa819 Update the includes.
This commit was SVN r9155.
2006-02-27 02:14:31 +00:00
Brian Barrett
d45b2b77d1 * don't reset AR or RANLIB - libtool will do the "right thing" for us
This commit was SVN r9151.
2006-02-26 20:32:47 +00:00
Brian Barrett
1832a30c92 * htonl requires arpa/inet.h on most unix platforms...
This commit was SVN r9150.
2006-02-26 19:10:26 +00:00
Brian Barrett
68f5ad074a * need synch.h as well as threads.h for Solaris threads.
This commit was SVN r9149.
2006-02-26 17:08:47 +00:00
Jeff Squyres
bd04d8f09e Add missing <thread.h> for Solaris
This commit was SVN r9148.
2006-02-26 16:56:38 +00:00
Jeff Squyres
28a1610453 - Add mssing <netdb.h>
- Change #if to #ifdef

This commit was SVN r9146.
2006-02-26 16:06:58 +00:00
Brian Barrett
285581dff2 More endian-related cleanups:
- moved hton64 and ntoh64 from the bunch of places it had been copied
    into one header file
  - properly set and use the btl_tcp's nbo option to put things in
    network byte order on the wire if both sides don't have the same
    endianness
  - Put the OB1 PML's headers (with a couple exceptions I need to discuss
    with Tim) in network byte order on the wire if both sides don't have
    the same endianness
  - since it was needed for the TCP BTL, move the orte_process_name_t
    HTON and NTOH macros from the TCP OOB to ns_types.h

This commit was SVN r9145.
2006-02-26 00:45:54 +00:00
Brian Barrett
88448cf40a * fix setting of remote proc architecture integers. We were setting all
procs architecture ints to the architecture of the last proc we got 
  in the list from the gpr, which is probably not what we intended to
  do...

This commit was SVN r9144.
2006-02-26 00:05:25 +00:00