1
1
openmpi/ompi/mca/mtl/portals
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
..
configure.m4 * Add new PML (CM) and network drivers (MTL) for high speed 2006-07-04 01:20:20 +00:00
configure.params * Add new PML (CM) and network drivers (MTL) for high speed 2006-07-04 01:20:20 +00:00
Makefile.am Per long threads on the mailing list and much confusion discussion 2007-12-15 13:32:02 +00:00
mtl_portals_component.c ... Delayed due to notifier commits earlier this day ... 2009-04-29 01:32:14 +00:00
mtl_portals_endpoint.h * Add new PML (CM) and network drivers (MTL) for high speed 2006-07-04 01:20:20 +00:00
mtl_portals_probe.c ... Delayed due to notifier commits earlier this day ... 2009-04-29 01:32:14 +00:00
mtl_portals_recv_short.c - On the way to get the BTLs split out and lessen dependency on orte: 2009-02-14 02:26:12 +00:00
mtl_portals_recv_short.h Added send-side optimizations (persistent zero-length md and copy blocks) 2007-11-21 21:31:37 +00:00
mtl_portals_recv.c - Split the datatype engine into two parts: an MPI specific part in 2009-07-13 04:56:31 +00:00
mtl_portals_recv.h Added send-side optimizations (persistent zero-length md and copy blocks) 2007-11-21 21:31:37 +00:00
mtl_portals_request.h - Split the datatype engine into two parts: an MPI specific part in 2009-07-13 04:56:31 +00:00
mtl_portals_send_short.c - On the way to get the BTLs split out and lessen dependency on orte: 2009-02-14 02:26:12 +00:00
mtl_portals_send_short.h Fix a bunch of compiler warnings. Refs trac:1458 2008-09-26 16:15:05 +00:00
mtl_portals_send.c - Split the datatype engine into two parts: an MPI specific part in 2009-07-13 04:56:31 +00:00
mtl_portals.c ... Delayed due to notifier commits earlier this day ... 2009-04-29 01:32:14 +00:00
mtl_portals.h - Split the datatype engine into two parts: an MPI specific part in 2009-07-13 04:56:31 +00:00