1
1
Граф коммитов

182 Коммитов

Автор SHA1 Сообщение Дата
George Bosilca
f0303a8b25 Indentation.
This commit was SVN r22254.
2009-12-02 22:03:52 +00:00
George Bosilca
3a2f071018 If the user asked for dynamic rules but "forget" to provide them, nicely
complain and switch back to the default behavior (fixed rules).

This commit was SVN r22109.
2009-10-19 17:58:47 +00:00
Rainer Keller
8e1b23779f - Replace combinations of
#if defined (c_plusplus)
          defined (__cplusplus)
   followed by
      extern "C" {
   and the closing counterpart by BEGIN_C_DECLS and END_C_DECLS.

   Notable exceptions are:
    - opal/include/opal_config_bottom.h:
      This is our generated code, that itself defines BEGIN_C_DECL and
      END_C_DECL
    - ompi/mpi/cxx/mpicxx.h:
      Here we do not include opal_config_bottom.h:                                 
    - Belongs to external code:                                                    
      opal/mca/backtrace/darwin/MoreBacktrace/MoreDebugging/MoreBacktrace.c        
      opal/mca/backtrace/darwin/MoreBacktrace/MoreDebugging/MoreBacktrace.h        
    - opal/include/opal/prefetch.h:
      Has C++ specific macros that are protected:                                  

    - Had #if ... } #endif  _and_ END_C_DECLS (aka end up with 2x
      END_C_DECLS)
      ompi/mca/btl/openib/btl_openib.h
    - opal/event/event.h has #ifdef __cplusplus as BEGIN_C_DECLS...
    - opal/win32/ompi_process.h: had extern "C"\n {...
      opal/win32/ompi_process.h: dito
    - ompi/mca/btl/pcie/btl_pcie_lex.l: needed to add *_C_DECLS
      ompi/mpi/f90/test/align_c.c: dito
    - ompi/debuggers/msgq_interface.h: used #ifdef __cplusplus
    - ompi/mpi/f90/xml/common-C.xsl: Amend

   Tested on linux using --with-openib and --with-mx

   The following do not contain either opal_config.h, orte_config.h or
   ompi_config.h
   (but possibly other header files, that include one of the above):
      ompi/mca/bml/r2/bml_r2_ft.h
      ompi/mca/btl/gm/btl_gm_endpoint.h
      ompi/mca/btl/gm/btl_gm_proc.h
      ompi/mca/btl/mx/btl_mx_endpoint.h
      ompi/mca/btl/ofud/btl_ofud_endpoint.h
      ompi/mca/btl/ofud/btl_ofud_frag.h
      ompi/mca/btl/ofud/btl_ofud_proc.h
      ompi/mca/btl/openib/btl_openib_mca.h
      ompi/mca/btl/portals/btl_portals_endpoint.h
      ompi/mca/btl/portals/btl_portals_frag.h
      ompi/mca/btl/sctp/btl_sctp_endpoint.h
      ompi/mca/btl/sctp/btl_sctp_proc.h
      ompi/mca/btl/tcp/btl_tcp_endpoint.h
      ompi/mca/btl/tcp/btl_tcp_ft.h
      ompi/mca/btl/tcp/btl_tcp_proc.h
      ompi/mca/btl/template/btl_template_endpoint.h
      ompi/mca/btl/template/btl_template_proc.h
      ompi/mca/btl/udapl/btl_udapl_eager_rdma.h
      ompi/mca/btl/udapl/btl_udapl_endpoint.h
      ompi/mca/btl/udapl/btl_udapl_mca.h
      ompi/mca/btl/udapl/btl_udapl_proc.h
      ompi/mca/mtl/mx/mtl_mx_endpoint.h
      ompi/mca/mtl/mx/mtl_mx.h
      ompi/mca/mtl/psm/mtl_psm_endpoint.h
      ompi/mca/mtl/psm/mtl_psm.h
      ompi/mca/pml/cm/pml_cm_component.h
      ompi/mca/pml/csum/pml_csum_comm.h
      ompi/mca/pml/dr/pml_dr_comm.h
      ompi/mca/pml/dr/pml_dr_component.h
      ompi/mca/pml/dr/pml_dr_endpoint.h
      ompi/mca/pml/dr/pml_dr_recvfrag.h
      ompi/mca/pml/example/pml_example.h
      ompi/mca/pml/ob1/pml_ob1_comm.h
      ompi/mca/pml/ob1/pml_ob1_component.h
      ompi/mca/pml/ob1/pml_ob1_endpoint.h
      ompi/mca/pml/ob1/pml_ob1_rdmafrag.h
      ompi/mca/pml/ob1/pml_ob1_recvfrag.h
      ompi/mca/pml/v/pml_v_output.h
      opal/include/opal/prefetch.h
      opal/mca/timer/aix/timer_aix.h
      opal/util/qsort.h
      test/support/components.h

This commit was SVN r21855.

The following SVN revision numbers were found above:
  r2 --> open-mpi/ompi@58fdc18855
2009-08-20 11:42:18 +00:00
George Bosilca
23e8ce91ba Rework the selection logic for the tuned collectives. All supported collectives
now are able to use the dynamic rules. Moreover, these rules are loaded only once,
and stored at the component level. All communicators are able to use these rules
(not only MPI_COMM_WORLD as until now).
A lot of minor corrections, memory management issues and reduction in the amount
of memory used by the tuned collectives.

This commit was SVN r21825.
2009-08-14 21:06:23 +00:00
Rainer Keller
6c5532072a - Split the datatype engine into two parts: an MPI specific part in
OMPI
   and a language agnostic part in OPAL. The convertor is completely
   moved into OPAL.  This offers several benefits as described in RFC
   http://www.open-mpi.org/community/lists/devel/2009/07/6387.php
   namely:
    - Fewer basic types (int* and float* types, boolean and wchar
    - Fixing naming scheme to ompi-nomenclature.
    - Usability outside of the ompi-layer.
 - Due to the fixed nature of simple opal types, their information is
   completely
   known at compile time and therefore constified
 - With fewer datatypes (22), the actual sizes of bit-field types may be
   reduced
   from 64 to 32 bits, allowing reorganizing the opal_datatype
   structure, eliminating holes and keeping data required in convertor
   (upon send/recv) in one cacheline...
   This has implications to the convertor-datastructure and other parts
   of the code.
 - Several performance tests have been run, the netpipe latency does not
   change with
   this patch on Linux/x86-64 on the smoky cluster.
 - Extensive tests have been done to verify correctness (no new
   regressions) using:
   1. mpi_test_suite on linux/x86-64 using clean ompi-trunk and
    ompi-ddt:
    a. running both trunk and ompi-ddt resulted in no differences
       (except for MPI_SHORT_INT and MPI_TYPE_MIX_LB_UB do now run
       correctly).
    b. with --enable-memchecker and running under valgrind (one buglet
       when run with static found in test-suite, commited)
   2. ibm testsuite on linux/x86-64 using clean ompi-trunk and ompi-ddt:
      all passed (except for the dynamic/ tests failed!! as trunk/MTT)
   3. compilation and usage of HDF5 tests on Jaguar using PGI and
      PathScale compilers.
   4. compilation and usage on Scicortex.
 - Please note, that for the heterogeneous case, (-m32 compiled
   binaries/ompi), neither
   ompi-trunk, nor ompi-ddt branch would successfully launch.

This commit was SVN r21641.
2009-07-13 04:56:31 +00:00
Greg Koenig
60485ff95f This is a very large change to rename several #define values from
OMPI_* to OPAL_*.  This allows opal layer to be used more independent
from the whole of ompi.

NOTE: 9 "svn mv" operations immediately follow this commit.

This commit was SVN r21180.
2009-05-06 20:11:28 +00:00
Rainer Keller
9736af1191 - Fix Coverity CID 182:
Well, well, just do not "call" ompi_comm_rank twice but rather
   reuse variable...

 - Fix Coverity CID 1262:
   Using uninitialized value "(statuses[err_index]).MPI_ERROR"
   Sure, these statuses are only initialized after ompi_request_wait_all,
   so introduce a short-circuit label to jump to...

This commit was SVN r21153.
2009-05-05 12:28:51 +00:00
Rainer Keller
221fb9dbca ... Delayed due to notifier commits earlier this day ...
- Delete unnecessary header files using
   contrib/check_unnecessary_headers.sh after applying
   patches, that include headers, being "lost" due to
   inclusion in one of the now deleted headers...

   In total 817 files are touched.
   In ompi/mpi/c/ header files are moved up into the actual c-file,
   where necessary (these are the only additional #include),
   otherwise it is only deletions of #include (apart from the above
   additions required due to notifier...)

 - To get different MCAs (OpenIB, TM, ALPS), an earlier version was
   successfully compiled (yesterday) on:
   Linux locally using intel-11, gcc-4.3.2 and gcc-SVN + warnings enabled
   Smoky cluster (x86-64 running Linux) using PGI-8.0.2 + warnings enabled
   Lens cluster (x86-64 running Linux) using Pathscale-3.2 + warnings enabled

This commit was SVN r21096.
2009-04-29 01:32:14 +00:00
George Bosilca
c5b1bdd57c Correctly deal with the error case. The problem is tricky: the MPI standard doesn't allow
MPI_ERR_IN_STATUS to be returned from any functions that return only one completed request
(few exception here: wait_some and wait_all and the test versions). As we use an wait_all
in these send_receive functions we should convert the MPI_ERR_IN_STATUS to the real
error, i.e. the one comming from the MPI_ERROR field in the status corresponding to the
failed request.

This commit was SVN r20907.
2009-03-31 23:44:59 +00:00
Rainer Keller
9dea63d63a - Last of intrusive commits (promised)... err for now.
Anyway, this is blocking the move: do not include pml.h
   if not really needed, aka none of the following used:
     mca_pml
     MCA_PML_CALL
     OMPI_ANY_TAG
     OMPI_ANY_SOURCE
     OMPI_PROC_NULL

 - Notable exceptions (deleting in one header->adding):
   - ompi/mca/mtl/psm/
   - ompi/mca/osc/rdma/
   - ompi/mca/btl/openib/btl_openib_endpoint.c depended on
     pml_base_sendreq.h

 - Tested on Linux/x86-64, this time including make check
   (thanks Jeff and Ralph)

This commit was SVN r20725.
2009-03-04 17:06:51 +00:00
Terry Dontje
0178b6c45f Added padding to predefined handle structures to maintain library version to
version compatibility.

This commit was SVN r20627.
2009-02-24 17:17:33 +00:00
Rainer Keller
d81443cc5a - On the way to get the BTLs split out and lessen dependency on orte:
Often, orte/util/show_help.h is included, although no functionality
   is required -- instead, most often opal_output.h, or               
   orte/mca/rml/rml_types.h                                           
   Please see orte_show_help_replacement.sh commited next.            

 - Local compilation (Linux/x86_64) w/ -Wimplicit-function-declaration
   actually showed two *missing* #include "orte/util/show_help.h"     
   in orte/mca/odls/base/odls_base_default_fns.c and                  
   in orte/tools/orte-top/orte-top.c                                  
   Manually added these.                                              

   Let's have MTT the last word.

This commit was SVN r20557.
2009-02-14 02:26:12 +00:00
George Bosilca
01adc999c5 Correctly forward the right module if we call another collective function. Kudos to
Edgar for figuring out this tricky bug.

This commit was SVN r20267.
2009-01-14 03:22:54 +00:00
Jeff Squyres
11b375f8b5 CIDs 1080-1090: assert() checks were not sufficient to check for
NEGATIVE_RETURNS from _reg_int() because those are not always
checked.  So replace them with real if() checks.

This commit was SVN r20195.
2009-01-03 15:56:25 +00:00
Nysal Jan
ee8ec6f6b5 Remove dead/redundant code. Minimize number of calloc invocations
This commit was SVN r20121.
2008-12-12 10:55:50 +00:00
Rolf vandeVaart
13e8975f83 In the case where we detect a value of 0 in the recvcount
array, fall back to the simpler algorithms.  This is not
the optimal solution, but it works.  

This commit was SVN r19702.
2008-10-07 19:44:51 +00:00
Rolf vandeVaart
0a0ddfc934 Handle MPI_IN_PLACE correctly in the ompi_coll_tuned_reduce_scatter_intra_ring function.
We were not adjusting the sendbuf in this case so we were reducing garbage.

This fixes ticket #1506.

This commit was SVN r19673.
2008-10-02 20:01:27 +00:00
George Bosilca
6a9514ee08 Make the code match the comment. I checked with Jelena, and based on the papers we
published this is the expected algorithm for the specified message and communicator
size.

This commit closes ticket #1330.

This commit was SVN r19563.
2008-09-15 23:28:40 +00:00
Rainer Keller
ee1fe9015a - Make sure, that the *param_index are > 0 (here, we don't pass
errors up...).
   Coverity CID 1080 - 1090
 - Really make sure, the user does not specify stupid negative values.

This commit was SVN r19233.
2008-08-11 11:21:04 +00:00
Rainer Keller
23c2292478 - Fix variable set but not used
Coverity CID1058

This commit was SVN r19184.
2008-08-06 13:57:38 +00:00
Jeff Squyres
0af7ac53f2 Fixes trac:1392, #1400
* add "register" function to mca_base_component_t
   * converted coll:basic and paffinity:linux and paffinity:solaris to
     use this function
   * we'll convert the rest over time (I'll file a ticket once all
     this is committed)
 * add 32 bytes of "reserved" space to the end of mca_base_component_t
   and mca_base_component_data_2_0_0_t to make future upgrades
   [slightly] easier
   * new mca_base_component_t size: 196 bytes
   * new mca_base_component_data_2_0_0_t size: 36 bytes
 * MCA base version bumped to v2.0
   * '''We now refuse to load components that are not MCA v2.0.x'''
 * all MCA frameworks versions bumped to v2.0
 * be a little more explicit about version numbers in the MCA base
   * add big comment in mca.h about versioning philosophy

This commit was SVN r19073.

The following Trac tickets were found above:
  Ticket 1392 --> https://svn.open-mpi.org/trac/ompi/ticket/1392
2008-07-28 22:40:57 +00:00
Ralph Castain
9613b3176c Effectively revert the orte_output system and return to direct use of opal_output at all levels. Retain the orte_show_help subsystem to allow aggregation of show_help messages at the HNP.
After much work by Jeff and myself, and quite a lot of discussion, it has become clear that we simply cannot resolve the infinite loops caused by RML-involved subsystems calling orte_output. The original rationale for the change to orte_output has also been reduced by shifting the output of XML-formatted vs human readable messages to an alternative approach.

I have globally replaced the orte_output/ORTE_OUTPUT calls in the code base, as well as the corresponding .h file name. I have test compiled and run this on the various environments within my reach, so hopefully this will prove minimally disruptive.

This commit was SVN r18619.
2008-06-09 14:53:58 +00:00
Ralph Castain
c992e99035 Remove the tags from orte_output_open and the filtering operation from orte_output - this will be handled differently to improve the XML output interface
This commit was SVN r18557.
2008-06-03 14:24:01 +00:00
Rolf vandeVaart
5baa733ad5 Fix another warning (using a variable before it was initialized.)
Thanks Jeff for pointing this out.

This commit was SVN r18489.
2008-05-23 13:57:55 +00:00
George Bosilca
c31cc5b270 Remove a warning about line being unused.
This commit was SVN r18472.
2008-05-21 20:46:22 +00:00
Rolf vandeVaart
74d0259480 Add new implentation of barrier. This shows better performance on some clusters.
However, no decision logic is changed by this commit so default behavior has not changed.  This
is only selectable by runtime parameters.

This commit was SVN r18464.
2008-05-20 17:37:41 +00:00
Jeff Squyres
7154776465 Removed unused variable / compiler warning.
This commit was SVN r18454.
2008-05-19 13:41:45 +00:00
Jeff Squyres
e7ecd56bd2 This commit represents a bunch of work on a Mercurial side branch. As
such, the commit message back to the master SVN repository is fairly
long.

= ORTE Job-Level Output Messages =

Add two new interfaces that should be used for all new code throughout
the ORTE and OMPI layers (we already make the search-and-replace on
the existing ORTE / OMPI layers):

 * orte_output(): (and corresponding friends ORTE_OUTPUT,
   orte_output_verbose, etc.)  This function sends the output directly
   to the HNP for processing as part of a job-specific output
   channel.  It supports all the same outputs as opal_output()
   (syslog, file, stdout, stderr), but for stdout/stderr, the output
   is sent to the HNP for processing and output.  More on this below.
 * orte_show_help(): This function is a drop-in-replacement for
   opal_show_help(), with two differences in functionality:
   1. the rendered text help message output is sent to the HNP for
      display (rather than outputting directly into the process' stderr
      stream)
   1. the HNP detects duplicate help messages and does not display them
      (so that you don't see the same error message N times, once from
      each of your N MPI processes); instead, it counts "new" instances
      of the help message and displays a message every ~5 seconds when
      there are new ones ("I got X new copies of the help message...")

opal_show_help and opal_output still exist, but they only output in
the current process.  The intent for the new orte_* functions is that
they can apply job-level intelligence to the output.  As such, we
recommend that all new ORTE and OMPI code use the new orte_*
functions, not thei opal_* functions.

=== New code ===

For ORTE and OMPI programmers, here's what you need to do differently
in new code:

 * Do not include opal/util/show_help.h or opal/util/output.h.
   Instead, include orte/util/output.h (this one header file has
   declarations for both the orte_output() series of functions and
   orte_show_help()).
 * Effectively s/opal_output/orte_output/gi throughout your code.
   Note that orte_output_open() takes a slightly different argument
   list (as a way to pass data to the filtering stream -- see below),
   so you if explicitly call opal_output_open(), you'll need to
   slightly adapt to the new signature of orte_output_open().
 * Literally s/opal_show_help/orte_show_help/.  The function signature
   is identical.

=== Notes ===

 * orte_output'ing to stream 0 will do similar to what
   opal_output'ing did, so leaving a hard-coded "0" as the first
   argument is safe.
 * For systems that do not use ORTE's RML or the HNP, the effect of
   orte_output_* and orte_show_help will be identical to their opal
   counterparts (the additional information passed to
   orte_output_open() will be lost!).  Indeed, the orte_* functions
   simply become trivial wrappers to their opal_* counterparts.  Note
   that we have not tested this; the code is simple but it is quite
   possible that we mucked something up.

= Filter Framework =

Messages sent view the new orte_* functions described above and
messages output via the IOF on the HNP will now optionally be passed
through a new "filter" framework before being output to
stdout/stderr.  The "filter" OPAL MCA framework is intended to allow
preprocessing to messages before they are sent to their final
destinations.  The first component that was written in the filter
framework was to create an XML stream, segregating all the messages
into different XML tags, etc.  This will allow 3rd party tools to read
the stdout/stderr from the HNP and be able to know exactly what each
text message is (e.g., a help message, another OMPI infrastructure
message, stdout from the user process, stderr from the user process,
etc.).

Filtering is not active by default.  Filter components must be
specifically requested, such as:

{{{
$ mpirun --mca filter xml ...
}}}

There can only be one filter component active.

= New MCA Parameters =

The new functionality described above introduces two new MCA
parameters:

 * '''orte_base_help_aggregate''': Defaults to 1 (true), meaning that
   help messages will be aggregated, as described above.  If set to 0,
   all help messages will be displayed, even if they are duplicates
   (i.e., the original behavior).
 * '''orte_base_show_output_recursions''': An MCA parameter to help
   debug one of the known issues, described below.  It is likely that
   this MCA parameter will disappear before v1.3 final.

= Known Issues =

 * The XML filter component is not complete.  The current output from
   this component is preliminary and not real XML.  A bit more work
   needs to be done to configure.m4 search for an appropriate XML
   library/link it in/use it at run time.
 * There are possible recursion loops in the orte_output() and
   orte_show_help() functions -- e.g., if RML send calls orte_output()
   or orte_show_help().  We have some ideas how to fix these, but
   figured that it was ok to commit before feature freeze with known
   issues.  The code currently contains sub-optimal workarounds so
   that this will not be a problem, but it would be good to actually
   solve the problem rather than have hackish workarounds before v1.3 final.

This commit was SVN r18434.
2008-05-13 20:00:55 +00:00
Rolf vandeVaart
0e32dd1022 Add MPI_Alltoallv to tuned collectives and add a pairwise implementation of MPI_Alltoallv. However, do not change the default behavior for now. The only way to use new pairwise implementation is via mca parameters.
This commit was SVN r18394.
2008-05-07 02:31:24 +00:00
Jeff Squyres
213b5d5c6e Per long threads on the mailing list and much confusion discussion
about linkers, have all OPAL, ORTE, and OMPI components '''not'' link
against the OPAL, ORTE, or OMPI libraries.

See ttp://www.open-mpi.org/community/lists/users/2007/10/4220.php for
details (or https://svn.open-mpi.org/trac/ompi/wiki/Linkers for a
better-formatted version of the same info).

This commit was SVN r16968.
2007-12-15 13:32:02 +00:00
George Bosilca
1e7a791349 Remove some of the problems identified by Coverty.
This commit was SVN r16112.
2007-09-12 20:13:26 +00:00
Shiqing Fan
a0660f4deb - Just some type casts.
This commit was SVN r16100.
2007-09-12 15:29:58 +00:00
Brian Barrett
af4e86c25f Update collectives selection logic to allow for multiple components to be
used at nce (up to one unique collective module per collective function).
Matches r15795:15921 of the tmp/bwb-coll-select branch

This commit was SVN r15924.

The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
  r15795
  r15921
2007-08-19 03:37:49 +00:00
Jelena Pjesivac-Grbovic
9bd9c92dbd Making sure that the decision function for scatter and gather correctly
computes everything for MPI_IN_PLACE case.

This commit was SVN r15841.
2007-08-13 17:35:50 +00:00
Jelena Pjesivac-Grbovic
b558e820cb removing compiler wraning
This commit was SVN r15803.
2007-08-08 15:22:01 +00:00
Jelena Pjesivac-Grbovic
daa10b277e modifying scatter decision function to use binomial algorithm for
small message sizes.

This commit was SVN r15798.
2007-08-07 22:16:13 +00:00
Jelena Pjesivac-Grbovic
1b66a52c50 Modifying type of binomial tree used for binomial reduce:
switching:
       0                         0
     / \ \                     / \ \
	1    \ \         -->       4   \ \
  /      \ \                 /     \ \
 3       2  \               3       2 \
             4                         1
(duh).  The first form is the bmtree suitable for bcast, but the latter is better for reduce.
Updating default decision function accordingly.

This commit was SVN r15422.
2007-07-13 21:07:51 +00:00
Jelena Pjesivac-Grbovic
d677db9b5f cleaning up alltoall implementation:
- removing MPI_* calls from bruck implementation
- simplifying 2 process case
- identation, etc.

This commit was SVN r15301.
2007-07-07 01:06:19 +00:00
Jelena Pjesivac-Grbovic
483222085e Fixing compiler warnings.
In gather, the ptmp += incr is irrelevant, since ptmp is set within the loop.

This commit was SVN r15293.
2007-07-05 20:40:50 +00:00
Jelena Pjesivac-Grbovic
3b0a52a104 adding tuned allgatherv implementation using bruck, ring, and neighbor-exchange algorithms.
The implementations passed intel and imb tests up to 40 processes.

This commit was SVN r15280.
2007-07-03 23:33:12 +00:00
Jelena Pjesivac-Grbovic
d55b415bb0 fixing typo
This commit was SVN r15240.
2007-06-28 20:56:55 +00:00
Jelena Pjesivac-Grbovic
8fc8b44d11 Modifying reduce decision function for large, single element reduces (again).
Binary algorithm without segmentation tends to outperform binomial algorithm 
in this case.

This commit was SVN r15226.
2007-06-27 22:01:56 +00:00
Jelena Pjesivac-Grbovic
0ecef1750d Modifying the default reduce decision function to use binomial algorithm
for single-element reduce (segmented algorithms make no sense in this case
and can cause performance degradation). 

This commit was SVN r15209.
2007-06-26 20:14:03 +00:00
Jelena Pjesivac-Grbovic
567b40b9a9 Modifying the default broadcast decision function to use binomial algorithm
for single-element broadcasts (segmented algorithms make no sense in this case
and can cause performance degradation).

This commit was SVN r15208.
2007-06-26 20:08:31 +00:00
Jelena Pjesivac-Grbovic
3740640711 Modifying MPI_Gather in tuned module:
- adding linear algorithm with synchronization for gather.
  This algorithm prevents congestion at root process, but introduces 
  synchronization (serializes non-root processes, but allows messages 
  to arrive from two processes at the same time).  
  It performed better than binomial and linear algorithms for large message, 
  and intermediate and large communicator sizes.
- Updating MPI_Gather decision function to reflect performance results
  from MX.  I will perform more measurements though - so this one can 
  change.

This commit was SVN r15165.
2007-06-21 20:00:36 +00:00
Jelena Pjesivac-Grbovic
625c6739ab Removing warning about unsed variable
This commit was SVN r14579.
2007-05-03 20:26:41 +00:00
Jelena Pjesivac-Grbovic
9eff74ad4d Modifying generalized reduce "synchronized" behavior:
- Removing "small" message size limit because it really does not relate to the eager size
accross the board.
Now, the leaf nodes in generalized reduce will use blocking send (DEFAULT/ORIGINAL BEHAVIOR) 
either when the maximum number of outstanding requests is 0 or 
when the total number of segments is less than the maximum number of outstanding requests.
Otherwise, it will send messages using non-blocking synchronized send operation.

This commit was SVN r14572.
2007-05-02 21:42:45 +00:00
George Bosilca
69642a9cd4 Remove 2 warnings about ptrdiff_t to unsigned long implicit conversion.
This commit was SVN r14565.
2007-05-01 19:47:33 +00:00
Jelena Pjesivac-Grbovic
3eac49aa59 Adding flow control for leaf nodes in generalized reduce structure.
This "feature" is disabled by default and it should not affect the current performance.

In case when the message size is large and segment size is smaller than eager size for particular interface,
the leaf nodes in generalized reduce function can overflood parent nodes by sending all segments without 
any synchronization.  This can cause the parent to have HIGH number of unexpected messages (think 16MB 
message with 1KB segments for example).  In case of binomial algorithm root node always has at least one
child which is leaf, so this can potentially affect the root's performance significantly [Especially in 
large communicators where root may have quite a few children (binomial tree for example)].
When the segment size is bigger than the eager size, rendezvous protocol ensures that this does 
not happen so it is not necessary.
Originally, the problem was exposed in "infinite" bucket allocator clean up time for "small" segment sizes
(which may explain some "deadlocks" on Thunderbird tests).

To prevent this, we allow user to specify mca parameter "--mca coll_tuned_reduce_algorithm_max_requests NUM"
this limits number of outstanding messages from a leaf node in generalized reduce to the parent to NUM.
Messages are sent as non-blocking synchrnous messages, so syncronization happens at "wait" time.
The synchronization actually improved performance of pipeline and binomial algorithm for large message sizes
with 1KB segments over MX, but I need to test it some more to make sure it is consistent.

Since there is no easy way to find out what is "the eager" size for particular btl, I set the limit to 4000B.
If message/individual segment size is greater than 4000B - we will not use this feature.  This variable may
or may not be exposed as mca parameter later...

I did not have any problems running it and both "default" and "synchronous" tests passed Intel Reduce* tests 
up to 80 processes (over MX).

This commit was SVN r14518.
2007-04-25 20:39:53 +00:00
Jelena Pjesivac-Grbovic
53cbec7a09 Make coll/tuned dynamic rules more verbose (when promted with --mca coll_base_verbose 1)
This commit was SVN r14469.
2007-04-23 16:34:52 +00:00