1
1

420 Коммитов

Автор SHA1 Сообщение Дата
George Bosilca
58c9c82dab Add a macro to mark the MPI request as completed (MCA_PML_BASE_REQUEST_MPI_COMPLETE) and
broadcast the request condition if required. This macro should be called with the request's
mutex locked.

This commit was SVN r8811.
2006-01-25 23:15:36 +00:00
Galen Shipman
1e0ea9dd6d Major fixes for the RDMA registration cache (leave_pinned).
This commit fixes issues with HPL runs on node counts > 4. 

This commit was SVN r8793.
2006-01-23 22:51:50 +00:00
George Bosilca
83cd8fac9d Now that the free list are correctly destroyed these tests are useless.
This commit was SVN r8792.
2006-01-23 20:33:40 +00:00
Jeff Squyres
bd2524378a Remove unused variable
This commit was SVN r8784.
2006-01-22 12:20:01 +00:00
Rainer Keller
a437fd3aaf - As with George's r8771, move lock/unlock of ompi_request_lock into the
if-clause, getting rid of local schedule variable.

This commit was SVN r8778.

The following SVN revision numbers were found above:
  r8771 --> open-mpi/ompi@2fadddebc8
2006-01-21 21:02:35 +00:00
Rainer Keller
7442a641ea - Fix locking, not related to the current deadlocks.
This commit was SVN r8777.
2006-01-21 13:43:36 +00:00
George Bosilca
2fadddebc8 Decrease the lock granularity and remove it from the path where we don't need it.
Reorder the code to get rid of some local variables.

This commit was SVN r8771.
2006-01-20 17:19:22 +00:00
George Bosilca
d247436bea Make the opal_atomic happy by using a signed int instead of an unsigned one.
This commit was SVN r8759.
2006-01-19 19:54:51 +00:00
Brian Barrett
5fdf753e2d remove unneeded second copy of the PML_CALL macros
This commit was SVN r8668.
2006-01-10 18:44:21 +00:00
George Bosilca
68b0ba0f58 We need the copyright ...
This commit was SVN r8625.
2005-12-31 15:51:57 +00:00
Jeff Squyres
06b01b3b76 Fix uninitialized variable
This commit was SVN r8605.
2005-12-23 13:02:36 +00:00
Tim Woodall
e1a49793f7 improve diagnostics when registration fails
This commit was SVN r8599.
2005-12-22 16:32:12 +00:00
Tim Woodall
0c57e2d091 correct typo
This commit was SVN r8593.
2005-12-22 14:28:13 +00:00
Jeff Squyres
efe84971ce Correct copyrights and some typos
This commit was SVN r8588.
2005-12-22 05:37:28 +00:00
Tim Woodall
1f9a559245 mising include
This commit was SVN r8579.
2005-12-21 14:26:56 +00:00
Tim Woodall
8c1027d974 first cut at ack/retrans protocol
This commit was SVN r8570.
2005-12-20 21:42:58 +00:00
Rainer Keller
a7ef9fc379 - The PML uniq is different only in a few places from teg.
Get the diff down as close as possible, so only whitespace and a few
   rearrangements of code (in mca_pml_teg_add_procs, uniq is nicer to
read), so awked diff of both is minimal.

This commit was SVN r8489.
2005-12-13 09:45:01 +00:00
Tim Woodall
ebbac05f3c remove questionable sched_yield - probably left over from debugging
This commit was SVN r8457.
2005-12-12 16:00:27 +00:00
George Bosilca
5c8c939713 Move the comment at the right place.
This commit was SVN r8445.
2005-12-10 23:25:29 +00:00
George Bosilca
c6eb429a9a Wondows work:
- remove windows socket initialization (it's already in the TCP component)
 - protect all used header files
 - remove the unused ones.

This commit was SVN r8434.
2005-12-10 21:38:48 +00:00
Jeff Squyres
6fbd321442 Fix a bunch of install locations for header files
This commit was SVN r8406.
2005-12-08 00:54:44 +00:00
Tim Woodall
bce1d2e220 volatile not required for non-threaded case
This commit was SVN r8392.
2005-12-06 18:27:56 +00:00
Tim Woodall
e135f850af backed out to much :-)
This commit was SVN r8356.
2005-12-01 17:32:27 +00:00
Tim Woodall
394bf196bd back out changes to match only one probe - consensus was we should
allow this if multiple threads post multiple probes

This commit was SVN r8353.
2005-12-01 17:17:06 +00:00
Tim Woodall
53a33f3bed dont allow fragment to match more than one probe
This commit was SVN r8352.
2005-12-01 17:06:40 +00:00
Tim Woodall
d7c1c23e3f corrections for handling probe with out of order delivery
- when processing out of order list - reset match to null on each iteration
- check matched request type and if probe - complete probe and queue fragment
  on unexpected list

This commit was SVN r8339.
2005-11-30 17:57:59 +00:00
George Bosilca
011e46017f Match the latest changes on the TEG PML (increase ompi_request_completed when a request is
finished).

This commit was SVN r8259.
2005-11-26 21:06:37 +00:00
Tim Woodall
b06335abe2 start of a pml for data reliability
This commit was SVN r8236.
2005-11-22 17:24:47 +00:00
George Bosilca
bec02702b7 Remove a static build conflict between UNIQ and TEG. Change the mca_pml_base_ptl_t to mca_pml_uniq_ptl_t
This commit was SVN r8227.
2005-11-22 03:44:18 +00:00
Brian Barrett
20cea60b82 * fix "make distclean" error in PML
* turns out (duh!) that there was a reason that the <projectdir>dir
  variable was set in the AM conditional.  If not, stupid directories
  are created and not needed...  duh.

This commit was SVN r8205.
2005-11-20 07:41:09 +00:00
Brian Barrett
8faa1884f0 * The last of the build system optimizations. Combine the component and
component/base Makefile.am files, reducing the time configure spends
  stamping out Makefiles at the end
* Install base_impl.h file when devel-headers are being installed

This commit was SVN r8200.
2005-11-20 01:03:01 +00:00
George Bosilca
a8d2b70d63 Small fix fot the collision between the TEG and UNIQ PML.
This commit was SVN r8145.
2005-11-13 23:03:36 +00:00
George Bosilca
c4e843ae08 A little add to the comment.
This commit was SVN r8143.
2005-11-13 23:01:41 +00:00
Tim Woodall
607f62accd - pass a flag to the peer indicating wether data is contiguous at the soure
- only attempt to schedule rdma if contiguous at both src/dst
- need to review this for next release 

This commit was SVN r8119.
2005-11-11 15:33:25 +00:00
Tim Woodall
4c7c277b0a improve the scalability of MPI_Waitall ...
note that any code that sets a request to a completed state must
now increment a counter for every completed request

This commit was SVN r8073.
2005-11-10 00:45:27 +00:00
Tim Woodall
2f6d50e0c6 init rdma count
This commit was SVN r8071.
2005-11-10 00:04:25 +00:00
Tim Woodall
78c98386d7 should reset the count (for persistent requests)
This commit was SVN r8064.
2005-11-09 22:02:48 +00:00
Tim Woodall
58b46d2da0 return mpool resources when request completes rather than in free
This commit was SVN r8063.
2005-11-09 21:59:01 +00:00
Jeff Squyres
42ec26e640 Update the copyright notices for IU and UTK.
This commit was SVN r7999.
2005-11-05 19:57:48 +00:00
Tim Woodall
e45f4744ee do not return these descriptors to cache
This commit was SVN r7986.
2005-11-03 23:20:38 +00:00
Tim Woodall
26003bc952 fix from release branch - don't use get protocol if more
than one btl is available

This commit was SVN r7984.
2005-11-03 20:52:56 +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
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
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
George Bosilca
75bc3dd43c Dont mess around with the OBJ_DESTRUCT on the communicator. It's quicker (and safer) to call
directly the communicator cleanup function (ompi_convertor_cleanup).

This commit was SVN r7814.
2005-10-19 21:28:52 +00:00
George Bosilca
1d75b7972f Solve thee problem with the reference count on the datatype (RT bug 1492). The problem is that the
convertor (when prepared) increase the reference count on the used datatype. This reference count
will be released only when the OBJ_DESTRUCT is called on a convertor. However, having to call
OBJ_CONSTRUCT and OBJ_DESTRUCT on each request every time we want to use it (even when it come
from the cache) is an expensive operation. This can be avoided is the OBJ_DESTRUCT will leave the
convertor in exactly the same state as OBJ_CONSTRUCT. With this approach we just have to call
OBJ_CONSTRUCT for each convertor once when we initially create the request.

This commit was SVN r7813.
2005-10-19 20:57:39 +00:00
George Bosilca
d2f831cd18 Construct the convertor attached to the receive request. This should happens only on the first allocation of a request object.
This commit was SVN r7807.
2005-10-18 21:53:05 +00:00
Jeff Squyres
f9974f72e0 construct/destruct convertor when requests are
constructed and allocated to free lists

This commit was SVN r7791.
2005-10-18 12:19:43 +00:00
Brian Barrett
1302cb4072 The next in a long line of crazed build system changes from Brian. This was
originally suggested by Ralf Wildenhues, to try to speed autogen, configure,
and make (and possibly even make install).  Use automake's include directive
to drastically reduce the number of Makefile files (although the number of
Makefile.am files is the same - most are just included in a top-level
Makefile.am).  Also use an Automake SUBDIRs feature to eliminate the
dynamic-mca tree, which was no longer really needed.  This makes adding
a framework easier (since you don't have to remember the dynamic-mca
tree) and makes building faster (as make doesn't have to recurse through
the dynamic-mca tree)

This commit was SVN r7777.
2005-10-17 00:21:10 +00:00
Edgar Gabriel
5d7fbd9d2e minor change in bml_r2_add_procs: the memory for the bml_endpoints structure
has to be allocated outside of the routine. Thus, the update version of pml/ob1/oml_ob1.c

This commit was SVN r7739.
2005-10-12 20:59:25 +00:00