1
1
Граф коммитов

6317 Коммитов

Автор SHA1 Сообщение Дата
George Bosilca
b0def3f6bf MX has 2 limitations regarding the iovecs. First they do not support iovec witha total size
larger than 32K for inter-nodes transfert ... and then they do not support iovecs larger than
16K for inter-node transfert. Therefore we have to set the size of our first fragment to
16K to match both cases.

This commit was SVN r7926.
2005-10-28 20:37:43 +00:00
Jeff Squyres
8503fce61b Remove debugging message
This commit was SVN r7924.
2005-10-28 18:53:20 +00:00
Jeff Squyres
7bdfe6557b - Update the checks in REDUCE, ALLREDUCE, SCAN, EXSCAN, and
REDUCE_SCATTER to more thoroughly check the datatype/op combination
  to see if it's valid or not.  If it's not, print a meaningful error
  message rather than "Invalid MPI_Op" indicating what specifically
  was wrong (therefore hopefully helping users track down where in the
  code the problem is, and/or telling us that there's a reduction
  operation combo that we don't support that we should)
- The check for whether a datatype is intrinsic needed to be updated
  -- it's not sufficient to check that dtype->id < DT_MAX_PREDEFINED;
  you really need to check the PREDEFINED flag on the datatype.
  Thanks to George for this fix (only intrinsics have a meaningful
  value in dtype->id).

This commit was SVN r7923.
2005-10-28 16:47:32 +00:00
Brian Barrett
4b35a11da8 * must use local labels for jumps instead of absolute labels so that bad things
don't happen when inlining the functions.

  This should go to the 1.0 branch

This commit was SVN r7921.
2005-10-28 15:52:09 +00:00
Brian Barrett
5ad49fac1f * when using gcc inline assembly, labels must be in the local number form or there
will be complaints about duplicate labels.  Fixes the rest of the compile issues
  in Opal on IRIX

This commit was SVN r7919.
2005-10-28 15:29:12 +00:00
Brian Barrett
d90a47c529 * remove typo in label name
This commit was SVN r7918.
2005-10-28 14:07:46 +00:00
Jeff Squyres
ce78b76598 Quick fix from Ralph -- this escape committing last night.
This commit was SVN r7917.
2005-10-28 14:03:26 +00:00
Ralph Castain
afeeacd76d Complete hookup of the registry proxy for the get_conditional command.
This commit was SVN r7915.
2005-10-28 05:35:07 +00:00
George Bosilca
ab97bde177 Rainer pointer out that the convertor already have the CONTIGUOUS flag is the
data is contiguous (set in ompi_convertor_prepare).
For unpack reinforce the limits of the pack for contiguous types.

This commit was SVN r7914.
2005-10-28 05:27:40 +00:00
George Bosilca
5355765d81 Cleanup has to reset the stack position.
This commit was SVN r7913.
2005-10-28 05:25:08 +00:00
Ralph Castain
ad9de4ca3b Restore the pointer arrays to the registry dictionaries. Revise the system so that the itag is equivalent to the index into the pointer array. It already was, but it wasn't obvious before (several functions relied upon it, but others "hid" the relationship) - now, make it explicitly clear. Set things up so lookups occur at max speed by just indexing into the dictionary array.
This commit was SVN r7912.
2005-10-28 04:56:06 +00:00
George Bosilca
d916e0c5b4 The (I hope) final solution for the convertor problem. As all the PML inherit
the base send and receive request from the pml_base, we can solve our problem
if we construct the convertor attached to any request in the pml_base_construct
function. At the end of the life time for each request (here life time is 
related to one utilisation, without taking in account the cache) we release
all information attached to the convertors in the _FINI macro by calling the
ompi_convertor_cleanup.

This commit was SVN r7910.
2005-10-28 03:26:36 +00:00
Jeff Squyres
f8ae2767cf Update README to match new compiler wrapper flags.
This commit was SVN r7909.
2005-10-28 03:15:39 +00:00
Ralph Castain
eebda71a0b Add a new API to the registry for conditional data retrievals. The new API allows you to retrieve data from registry containers that have key-value pairs where the value matches the specified one. The requested keys are then retrived from that container.
This commit was SVN r7907.
2005-10-28 00:30:58 +00:00
Brian Barrett
1c47b56163 * add ability to add compiler / linker flags to the MPI wrapper compilers
from configure (in addition to all the ones that the configure script
  finds)

This commit was SVN r7906.
2005-10-27 23:23:08 +00:00
Brian Barrett
bf67c9387b * initialize send request convertor with the correct type (convertor instead
of request).  This fixes at least the bug with NetPIPE in 64bit land that
  Troy was seeing. 

This commit was SVN r7904.
2005-10-27 23:08:27 +00:00
Galen Shipman
4a15761732 add support for srq limit reached async event, even though it doesn't appear
to  be supported by mellanox vapi.. perhaps this will be supported in the near 
future, for now it doesn't hurt to have it in the trunk


Also cleanup the receive descriptor posting macro's.. 

This commit was SVN r7903.
2005-10-27 22:47:19 +00:00
Tim Woodall
3fd351117a removed debug
This commit was SVN r7902.
2005-10-27 21:07:49 +00:00
Tim Woodall
3bd5b81dfa Submitted: Gleb Natapov
This commit was SVN r7899.
2005-10-27 17:48:40 +00:00
Tim Woodall
c484185f19 loop over poll event entries if the total number is larger than supported
This commit was SVN r7898.
2005-10-27 17:12:42 +00:00
Tim Woodall
793836da57 removed debug
This commit was SVN r7897.
2005-10-27 17:10:49 +00:00
Tim Woodall
7300112564 removed debug
This commit was SVN r7896.
2005-10-27 17:08:30 +00:00
Tim Woodall
4fc5b2105a this is currently an int - we shouldn't restrict it unless required
This commit was SVN r7895.
2005-10-27 17:06:58 +00:00
Tim Woodall
60754acae8 - modified rmaps data structures to point directly to ras node
- modified rsh to NOT query for each nodes mapping, as all data is
  already available in the rmaps structures

This commit was SVN r7894.
2005-10-27 17:04:10 +00:00
Tim Woodall
c0124fecdd changed segment dictionary to hash table to improve
search time for reverse lookup

This commit was SVN r7893.
2005-10-27 17:00:47 +00:00
Tim Woodall
13409ec53b correction for hang, check for additional fragments before callback,
which may queue a new fragment

This commit was SVN r7889.
2005-10-27 01:39:39 +00:00
Graham Fagg
5bb0d4a053 enable allreduce to be selected
This commit was SVN r7888.
2005-10-26 23:55:37 +00:00
Graham Fagg
2587d7ade9 added some more linear functions.
minor corrections on naming and debug info

This commit was SVN r7887.
2005-10-26 23:51:56 +00:00
Graham Fagg
c3e1dc410d Started to add basic linear functions
Also started to add the allreduce algorithms as I test them
(i.e. if it goes in its after testing from now on)

This commit was SVN r7886.
2005-10-26 23:11:32 +00:00
Jeff Squyres
d47ce065e9 Minor Makefile.am fix for static builds.
This commit was SVN r7882.
2005-10-26 15:57:58 +00:00
Jeff Squyres
2b66d53fe4 Put a note in that you have to compile all your MPI apps with the
Intel / PGI compiler flags for the logical issue.

This commit was SVN r7881.
2005-10-26 14:24:28 +00:00
Edgar Gabriel
ba3bf6592f fixing some warnings. No idea yet why the static builds fail...
This commit was SVN r7879.
2005-10-26 12:56:56 +00:00
Jeff Squyres
94d38b1812 Add header file for non-g++ compilers
This commit was SVN r7875.
2005-10-25 21:59:01 +00:00
Jeff Squyres
3f01c33f8a Updates about compiler notes
This commit was SVN r7871.
2005-10-25 19:29:46 +00:00
Jeff Squyres
23ab9e0277 A better solution to the previous commit -- RETAIN/RELEASE the MPI_Op
at the top-level MPI API function.  This allows two kinds of
scenarios:

1. MPI_Ireduce(..., op, ...);
   MPI_Op_free(op);
   MPI_Wait(...);

For the non-blocking collectives that we're someday planning -- to
make them analogous to non-blocking point-to-point stuff.

2. Thread 1:
   MPI_Reduce(..., op, ...);
   Thread 2:
   MPI_Op_free(op);

Granted, for #2 to occur would tread a fine line between a correct and
erroneous MPI program, but it is possible (as long as the Op_free was
*after* MPI_reduce() had started to execute).  It's more realistic
with case #1, where the Op_free() could be executed in the same thread
or a different thread.

This commit was SVN r7870.
2005-10-25 19:20:42 +00:00
Edgar Gabriel
d009d8de57 opening the hierarchical collective component to the public. I am at this
stage fairly confident that 
- it works in most scenarious (with symmetric hierarchies, with asymmetric
  hierarchies, wihout hierarchies - it just removes itself)
- it does not create too many problems (I am not aware of any at least)
- it does not slow down startup anymore dramatically (thanks to the fixes of
  Brian, Jeff, Tim and a significant reduction in the number of collective
  operations in the comm_query)
Any feedback is highly welcome.

This commit was SVN r7868.
2005-10-25 18:38:43 +00:00
Edgar Gabriel
00c04ab56a moving the hierarch collective component to the new parameter
registration interface.

This commit was SVN r7867.
2005-10-25 18:34:47 +00:00
Edgar Gabriel
3633605010 moving op_init further up in ompi_mpi_init, since it is required when
quering some of the collective components. Up to now, it just worked
somehow, but now with correct reference counting for ops in place, it
refused :-)

This commit was SVN r7866.
2005-10-25 18:33:48 +00:00
Jeff Squyres
ef09e768e0 Ensure to OBJ_DESTRUCT to free memory during finalize (caught by
Brian). 

This commit was SVN r7864.
2005-10-25 17:27:58 +00:00
Jeff Squyres
f8fd10715c - Minor style fix
- Be sure to properly OBJ_CONSTRUCT the intrinsic MPI_Op's
- RETAIN/RELEASE the op's when used in the invoke function

This commit was SVN r7863.
2005-10-25 16:24:00 +00:00
Tim Woodall
b60bea9ada dont allow callbacks to processed recursively - appear to be blowing away the stack
This commit was SVN r7862.
2005-10-25 13:48:08 +00:00
Tim Woodall
4eca6e22bd use persistent non-blocking receives
This commit was SVN r7861.
2005-10-25 13:38:13 +00:00
Jeff Squyres
a9f04c7573 Only do the extra va_* stuff if we're compiling with the compiler that
cares about it (PGI).

This commit was SVN r7860.
2005-10-25 13:08:52 +00:00
Jeff Squyres
c41c77848e Update the --enable-f* options to be --enable-mpi-f* and add
documentation about --enable-mpi-cxx.

This commit was SVN r7859.
2005-10-25 12:18:49 +00:00
Graham Fagg
382f05c7ad Infastructure changes.
started to add static (fixed if) statement based decision rules based on gigE numbers
added mca params so that a user can force a certain algorithm/segment/topo on a per collective basis
(this is not in the fixed call path but only in the dynamic (at com create) call path).
(these params can be used by test suites such as OCC to choice which algorithm they are using).

This commit was SVN r7854.
2005-10-25 03:55:58 +00:00
Brian Barrett
65bcc283c0 * Change the --enable-{cxx,f77,f90} options to
--enable-mpi-{cxx,f77,f90} so that people aren't confused about what
  they are actually disabling.

  This should go to the 1.0 branch

This commit was SVN r7851.
2005-10-25 02:53:54 +00:00
Graham Fagg
d8e32464cb ops. setting/reading mca option from the right varible would help.
This commit was SVN r7850.
2005-10-24 21:33:48 +00:00
Brian Barrett
1e2f7d6a3d * make sure to expose ompi_op_t as an object
This commit was SVN r7848.
2005-10-24 20:31:14 +00:00
Rainer Keller
d6120d32d6 - Only minor white-space changes, to clean up
This commit was SVN r7843.
2005-10-24 10:36:16 +00:00
Andrew Friedley
f92185c43b Complete George's fix - this is a problem I caused :(
This takes care of Troy's first segfault problem, and compile errors that will likely happen as soon as Ken applies George's patch and runs make again.

This commit was SVN r7833.
2005-10-21 21:06:20 +00:00