1
1
Граф коммитов

1976 Коммитов

Автор SHA1 Сообщение Дата
George Bosilca
c9da782804 Keep only one function to get the size of a datatype.
This commit was SVN r12170.
2006-10-18 17:33:01 +00:00
George Bosilca
3db5c0487d typos.
This commit was SVN r12168.
2006-10-18 17:12:25 +00:00
George Bosilca
21ade43b96 Remove a non reacheable statement.
This commit was SVN r12166.
2006-10-18 16:43:55 +00:00
Rainer Keller
47b24a0603 - Now the branch is done, linearize access regarding
request handling.  Buys a little bit on IMB, no
   functional change, otherwise.

This commit was SVN r12165.
2006-10-18 16:11:50 +00:00
Ralph Castain
f4a458532b This doesn't totally resolve the comm_spawn problem, but it helps a little. I'll continue working on it and hope to resolve it completely shortly. The issue primarily centers on where to start mapping the child job's processes, and how to deal with oversubscription that might result. At the moment, I am trying to resolve the first issue first (hey, that even sounds right!).
This change does a couple of things:

1. Since the USE_PARENT_ALLOC attribute is a directive about regarding allocation of resources to a job, it more properly should be an attribute of the RAS. Change the name to reflect that and move the attribute define to the ras_types.h file.

2. Add the attributes list to the RMAPS map_job interface. This provides us with the desired flexibility to dynamically specify directives for mapping. The system will - in the absence of any attribute-based directive - default to the values provided in the MCA parameters (either from environment or command-line interface).

This commit was SVN r12164.
2006-10-18 14:01:44 +00:00
Gleb Natapov
252a9cea34 Fix bug in vma rcache.
This commit was SVN r12163.
2006-10-18 10:55:01 +00:00
George Bosilca
be27ee6fa0 Correct the bcast problem where we always did a bcast with segzise of 0.
Activate the reduce decision function.
Others small updates (mostly TAB to spaces).

This commit was SVN r12161.
2006-10-18 02:00:46 +00:00
George Bosilca
640178c4b3 Grepping through the source files I found these calls to the data-type engine
with the wrong type of arguments.

This commit was SVN r12148.
2006-10-17 21:05:04 +00:00
George Bosilca
6f5ec2390b pedantic...
This commit was SVN r12147.
2006-10-17 20:25:40 +00:00
George Bosilca
8852c00c36 Look like a big commit but in fact it address only one issue. The way we're working with
size and diplacement of data-type. After this patch all data can contain size_t bytes
and the displacements are defined as ptrdiff_t. All of the files I was able to compile
have been modified to match this requirement.

This commit was SVN r12146.
2006-10-17 20:20:58 +00:00
Andrew Friedley
16769e64fe Remove old UD BTL code.
The UD BTL isn't gone - the latest version is in my afriedle-ud branch.  This version on the trunk was very old, ompi_ignore'd, lacked performance, and probably contained bugs.  The maintained version on my branch is working solid, and will eventually come back, but not for v1.2.

This commit was SVN r12144.
2006-10-17 18:59:21 +00:00
Rainer Keller
668902c780 - trivial spelling
This commit was SVN r12139.
2006-10-17 16:34:52 +00:00
Ralph Castain
13227e36ab This commit looks a lot bigger than it is, so relax :-)
Fix the problem observed by multiple people that comm_spawned children were (once again) being mapped onto the same nodes as their parents. This was caused by going through the RAS a second time, thus overwriting the mapper's bookkeeping that told RMAPS where it had left off.

To solve this - and to continue moving forward on the ORTE development - we introduce the concept of attributes to control the behavior of the RM frameworks. I defined the attributes and a list of attributes as new ORTE data types to make it easier for people to pass them around (since they are now fundamental to the system, and therefore we will be packing and unpacking them frequently). Thus, all the functions to manipulate attributes can be implemented and debugged in one place.

I used those capabilities in two places:

1. Added an attribute list to the rmgr.spawn interface.

2. Added an attribute list to the ras.allocate interface. At the moment, the only attribute I modified the various RAS components to recognize is the USE_PARENT_ALLOCATION one (as defined in rmgr_types.h).

So the RAS components now know how to reuse an allocation. I have debugged this under rsh, but it now needs to be tested on a wider set of platforms.

This commit was SVN r12138.
2006-10-17 16:06:17 +00:00
George Bosilca
ed83927025 Don't reset the convertor when a persistent request complete. Instead reset it
next time then request is used. This will keep the execution path on the default
case (not persistent) shorter.

This commit was SVN r12134.
2006-10-17 05:01:47 +00:00
George Bosilca
ef66afe45c Another inner loop optimization. Only check for num_fails when prev_bytes is
equal to num_bytes.

This commit was SVN r12133.
2006-10-17 04:38:38 +00:00
George Bosilca
6a4cf1fef5 A request is inactive only if it is persistant. Therefore, we don't need
2 sequential ifs, we can reach the same result with 2 embedded ifs.

This commit was SVN r12132.
2006-10-17 04:27:00 +00:00
George Bosilca
e116a37482 My last commit was wrong. Here is the correct version.
This commit was SVN r12131.
2006-10-17 02:45:03 +00:00
George Bosilca
01f5b4007b Check the count. It has to be a positive number.
This commit was SVN r12130.
2006-10-17 02:40:17 +00:00
Jeff Squyres
0ebe687ed8 Refs trac:502, #503. This commit as a result of review of r12122.
* Update comments in some MPI_FILE_* functions to reflect that the
   MPI specs have different page numbers in the ps and pdf (woof!).
 * Update comments to say "Retain" where we meant retain (not "return)
 * Add a check in MPI_ERRHANDLER_FREE to raise an MPI exception if the
   user attempts to free an intrinsic errhandler *and* the refcount is
   1 (meaning that it would actually free the intrinsic).  This
   protects erroneous programs from segv'ing.
 * Remove lengthy comment from comm_get_errhandler.c which is no
   longer valid (because of the MPI-2 errata that says that users *do*
   have to call MPI_ERRHANDLER_FREE).

This commit was SVN r12128.

The following SVN revision numbers were found above:
  r12122 --> open-mpi/ompi@407b3cb788

The following Trac tickets were found above:
  Ticket 502 --> https://svn.open-mpi.org/trac/ompi/ticket/502
2006-10-17 00:18:35 +00:00
Pavel Shamis
d64cb58007 Tavor HCA vendor_id was changed to correct one: 23108
Reviewed by: Jeff

This commit was SVN r12127.
2006-10-16 14:46:39 +00:00
Brian Barrett
fdc8f69b84 * need to include iostream as well as stdio.h when doing the tricks with
MPI::SEEK_* because iostreams (well, ios_base, but I don't think that
    should be included directly) can use SEEK_* as values in an enum, which
    means that 'const int' is bad for them.
  * Remove now useless comments in the cxx example programs
  * include iostream after mpi.h so that our examples work with other MPI
    implementations that don't try to be as friendly with the constants.

Refs trac:387

This commit was SVN r12125.

The following Trac tickets were found above:
  Ticket 387 --> https://svn.open-mpi.org/trac/ompi/ticket/387
2006-10-16 14:20:31 +00:00
Jeff Squyres
407b3cb788 Fix some problems with errhandlers:
* Fix MPI-2 page number in comments for a specific reference in the
   spec
 * Allow getting/setting the errhandler on MPI_FILE_NULL
 * Allow freeing of intrinsic errhandlers, per MPI-2 errata (if you GET
   an errhandler on a communicator, you must be able to FREE it, even
   if it's an intrinsic).

Thanks to Lisandro Dalcin for reporting these problems.

This commit was SVN r12122.
2006-10-16 12:58:40 +00:00
Brian Barrett
e7a7a64e4c Implement MPI::SEEK_{SET, END, POS} for the C++ bindings, working around
some issues with the C #defines SEEK_{SET, END, POS}.  The workaround
involves some hackery that should work in almost every common use case
for the C stdio constants (and all the legal issues of the MPI constants).
The one issue is that the C stdio constants are now const ints instead
of #defines, which means that #ifdef checks will fail for the constants.

Behavior can be disabled at either configure time or build time.

Refs trac:387

This commit was SVN r12121.

The following Trac tickets were found above:
  Ticket 387 --> https://svn.open-mpi.org/trac/ompi/ticket/387
2006-10-15 23:50:24 +00:00
Brian Barrett
9adde4f7b8 Allow multilib capability based on compiler flags. See:
https://svn.open-mpi.org/trac/ompi/wiki/compilerwrapper3264
for more information.

Refs trac:374

This commit was SVN r12120.

The following Trac tickets were found above:
  Ticket 374 --> https://svn.open-mpi.org/trac/ompi/ticket/374
2006-10-15 21:21:08 +00:00
Brian Barrett
14f338b7df Fix for lock/unlock epoch issues. Previously, we did not handle the case
where a window was in both the passive and active side of a lock sequence.

Refs trac:488

This commit was SVN r12112.

The following Trac tickets were found above:
  Ticket 488 --> https://svn.open-mpi.org/trac/ompi/ticket/488
2006-10-12 22:52:13 +00:00
George Bosilca
c75cfd3efc Remove all warnings from the TotalView interface.
This commit was SVN r12091.
2006-10-11 17:29:29 +00:00
Gleb Natapov
afa26fc4f9 Remove empty lines at the end of the file.
This commit was SVN r12083.
2006-10-11 11:38:38 +00:00
Ralph Castain
1f7a5da3ce Bring singleton comm_spawn online.
This commit was SVN r12081.
2006-10-10 23:59:48 +00:00
Galen Shipman
7102d39415 Use OMPI_FREE_LIST_GET not the blocking OMPI_FREE_LIST_WAIT, this allows GM
to take advantage of PML OB1 resource management.. Tested with intel test
suite p2p_c on 4 nodes. 

This commit was SVN r12075.
2006-10-10 14:56:55 +00:00
George Bosilca
33f300f636 I don't know what it was supposed to do but I'm quite sure it didn't do it correctly.
Just follow inc_num and you will understand. Now _resize will grow the list to match
the required number of elements as described in the comment in the .h file.

This commit was SVN r12074.
2006-10-10 14:47:51 +00:00
Pavel Shamis
e400da01bc Fix of typo error introduced in revision 12053.
Reviewed by: Jeff

This commit was SVN r12071.
2006-10-10 11:58:29 +00:00
George Bosilca
179067dfb5 Correct a type that break the PSM build.
This commit was SVN r12069.
2006-10-09 23:14:22 +00:00
Brian Barrett
51b2a0fd3f A couple of changes to improve shared memory behavior when resources get
constrained:

  * Make sure we always have a number of eager fragments available
    that scales with the number of processes communicating with
    a given proc over shared memory
  * Use FREE_LIST_GET instead of FREE_LIST_WAIT to return an
    error to the PML when resource exhaustion occurs
  * Don't dereference the frag during alloc unless we're sure
    it's not NULL

Reviewed by: Galen

Refs trac:413

This commit was SVN r12053.

The following Trac tickets were found above:
  Ticket 413 --> https://svn.open-mpi.org/trac/ompi/ticket/413
2006-10-06 21:13:49 +00:00
George Bosilca
2411ad74e4 Fix the bug #315. If there are multiple threads waiting for a free_list item
then use broadcast in order to wake them up. If there is only one then use signal
(which is supposed to be faster) and of course if there are no threads
waiting then just continue.

This commit was SVN r12049.
2006-10-06 16:17:50 +00:00
George Bosilca
8e246ce9db Add th headers to the make dist.
This commit was SVN r12045.
2006-10-06 14:52:44 +00:00
Jeff Squyres
77a5eb3b69 Fixes trac:236
Doing pointer math properly (e.g., incrementing by the right amount)
helps you not overflow buffers, cause random chaos, and contribute to
the heat death of the universe.  Sigh.

This commit was SVN r12015.

The following Trac tickets were found above:
  Ticket 236 --> https://svn.open-mpi.org/trac/ompi/ticket/236
2006-10-05 16:43:10 +00:00
George Bosilca
b2b2ade6c9 Remove the last underscore. Look like a typo ...
This commit was SVN r12006.
2006-10-05 05:41:57 +00:00
George Bosilca
3fc0e969b7 Cast in order to return the expected bool.
This commit was SVN r11992.
2006-10-05 05:11:05 +00:00
George Bosilca
b27f1814c6 If the function is expected to return a bool then let's return only
true or false.

This commit was SVN r11991.
2006-10-05 05:10:34 +00:00
George Bosilca
4689c56210 Always cast the return of malloc.
This commit was SVN r11990.
2006-10-05 05:07:43 +00:00
George Bosilca
5fb0c63959 Keep the ompi_config.h in the begining of the include list. This fixes
the #365. I found another way around on Windows.

This commit was SVN r11989.
2006-10-05 05:02:27 +00:00
George Bosilca
bf1d4ee259 Cast to the uint64_t before playing with the << operator in order to make sure we keep
all usefull informations.

This commit was SVN r11986.
2006-10-04 23:58:59 +00:00
George Bosilca
9e4b64ce81 Cleaning up the file, removing useless asserts. Commenting all calls to
opal_output.

This commit was SVN r11985.
2006-10-04 23:57:05 +00:00
George Bosilca
54476382e6 Remove all useless asserts and correct one of the comments.
Correctly compute the length to be unpacked.
Protect all outputs (only on debug mode and only if mpi_ddt_unpack is set).

This commit was SVN r11984.
2006-10-04 23:56:10 +00:00
George Bosilca
66e039496d This is the minimal patch for #365 (BLACS problem). Correctly compute the displacement
of the elements when we unroll a loop.

This commit was SVN r11983.
2006-10-04 23:54:22 +00:00
Jeff Squyres
e08c6e81f5 * Fixes trac:338: Only look at root-significant values at the root (e.g.,
recvbuf in MPI_GATHER).
 * Minor style updates (constants on the left of == and !=)
 * Fix a minor buglet that crept in r11904: had a recvbuf where it
   should have been recvcount.  Thankfully, this would have only
   affected erroneous programs.  ;-)

This commit was SVN r11980.

The following SVN revision numbers were found above:
  r11904 --> open-mpi/ompi@17539dc154

The following Trac tickets were found above:
  Ticket 338 --> https://svn.open-mpi.org/trac/ompi/ticket/338
2006-10-04 22:36:01 +00:00
George Bosilca
e4df4285b1 Reorder the enum in order to allow some compilers to optimize the big switch in
the header analisys.

This commit was SVN r11975.
2006-10-04 20:03:28 +00:00
George Bosilca
4d18d208d4 Getting closer to a full Total View support.
This commit was SVN r11974.
2006-10-04 20:01:33 +00:00
George Bosilca
e9572c7aee Mea culpa ... that was an weird error. When playing with 64 bits types we have to
the constant before using the << operator ... otherwise everything is 32 bits ...
which does not give the right answer.

This commit was SVN r11972.
2006-10-04 16:23:51 +00:00
George Bosilca
d05f492901 With the lastest changes for heterogeneous environments we have to
compute the hetero flag a little bit different. The hetero flag is now
attached to a convertor if and only if we can use the optimized conversion
functions. It's a little bit broader than before (the 2 architectures
has to be identical).

This commit was SVN r11962.
2006-10-03 20:25:30 +00:00