ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
7bf4c425ff
openmpi/ompi/mca/coll/ml/mca-coll-ml.config
Nathan Hjelm 1a021b8f2d coll/ml: add support for blocking and non-blocking allreduce, reduce, and 
allgather.

The new collectives provide a signifigant performance increase over tuned for
small and medium messages. We are initially setting the priority lower than
tuned until this has had some time to soak in the trunk. Please set
coll_ml_priority to 90 for MTT runs.

Credit for this work goes to Manjunath Gorentla Venkata (ORNL), Pavel Shamis (ORNL),
and Nathan Hjelm (LANL).

Commit details (for reference):

Import ORNL's collectives for MPI_Allreduce, MPI_Reduce, and MPI_Allgather.

We need to take the basesmuma header into account when calculating the
ptpcoll small message thresholds. Add a define to bcol.h indicating the
maximum header size so we can take the header into account while not
making ptpcoll dependent on information from basesmuma.

This resolves an issue with allreduce where ptpcoll overwrites the
header of the next buffer in the basesmuma bank.

Fix reduce and make a sequential collective launcher in coll_ml_inlines.h

The root calculation for reduce was wrong for any root != 0. There are
four possibilities for the root:

 - The root is not the current process but is in the current hierarchy. In
   this case the root is the index of the global root as specified in the
   root vector.

 - The root is not the current process and is not in the next level of the
   hierarchy. In this case 0 must be the local root since this process will
   never communicate with the real root.

 - The root is not the current process but will be in next level of the
   hierarchy. In this case the current process must be the root.

 - I am the root. The root is my index.

Tested with IMB which rotates the root on every call to MPI_Reduce. Consider
IMB the reproducer for the issue this commit solves.

Make the bcast algorithm decision an enumerated variable

Resolve various asset failures when destructing coll ml requests.

Two issues:

 - Always reset the request to be invalid before returning it to the
   free list. This will avoid an asset in ompi_request_t's destructor.
   OMPI_REQUEST_FINI does this (and also releases the fortran handle
   index).

 - Never explicitly construct or destruct the superclass of an opal
   object. This screws up the class function tables and will cause
   either an assert failure or a segmentation fault when destructing
   coll ml requests.

Cleanup allgather.

I removed the duplicate non-blocking and blocking functions and modeled
the cleanup after what I found in allreduce. Also cleaned up the code
somewhat.

Don't bother copying from the send to the recieve buffer in
bcol_basesmuma_allreduce_intra_fanin_fanout if the pointers are the
same.

The eliminates a warning about memcpy and aliasing and avoids an
unnecessary call to memcpy.

Alwasy call CHECK_AND_RELEASE on memsync collectives.

There was a call to OBJ_RELEASE on the collective communicator but
because CHECK_AND_RECYLCE was never called there was not matching call
to OBJ_RELEASE. This caused coll ml to leak communicators.

Make allreduce use the sequential collective launcher in coll_ml_inlines.h

Just launch the next collective in the component progress.

I am a little unsure about this patch. There appears to be some sort
of race between collectives that causes buffer exhaustion in some cases
(IMB Allreduce is a reproducer). Changing progress to only launch the
next bcol seems to resolve the issue but might not be the best fix.

Note that I see little-no performance penalty for this change.

Fix allreduce when there are extra sources.

There was an issue with the buffer offset calculation when there are
extra sources. In the case of extra sources == 1 the offset was set
to buffer_size (just past the header of the next buffer). I adjusted
the buffer size to take into accoun the maximum header size (see the
earlier commit that added this) and simplified the offset calculation.

Make reduce/allreduce non-blocking. This is required for MPI_Comm_idup
to work correctly.

This has been tested with various layouts using the ibm testsuite and
imb and appears to have the same performance as the old blocking version.

Fix allgather for non-contiguous layouts and simplify parsing the
topology.

Some things in this patch:

 - There were several comments to the effect that level 0 of the
   hierarchy MUST contain all of the ranks. At least one function
   made this assumption but it was not true. I changed the sbgp
   components and the coll ml initization code to enforce this
   requirement.

 - Ensure that hierarchy level 0 has the ranks in the correct
   scatter gather order. This removes the need for a separate
   sort list and fixes the offset calculation for allgather.

 - There were several passes over the hierarchy to determine
   properties of the hierarchy. I eliminated these extra passes
   and the memory allocation associated with them and calculate the
   tree properties on the fly. The same DFS recursion also handles
   the re-order of level 0.

All these changes have been verified with MPI_Allreduce, MPI_Reduce, and
MPI_Allgather. All functions now pass all IBM/Open MPI, and IMB tests.

coll/ml: correct pointer usage for MPI_BOTTOM

Since contiguous datatypes are copied via memcpy (bypassing the convertor) we
need to adjust for the lb of the datatype. This corrects problems found testing
code that uses MPI_BOTTOM (NULL) as the send pointer.

Add fallback collectives for allreduce and reduce.

cmr=v1.7.5:reviewer=pasha

This commit was SVN r30363.
2014-01-22 15:39:19 +00:00

171 строка
4.7 KiB
Plaintext
Исходник Ответственный История

##################################
# ML collective configuration file
##################################
# NOTE (by Pasha):
# Since ML configuration infrastructure is limited on this stage we do not support some tunings, even so parser
# understands this values and keys, but we do not have place to load all this values.
# threshold - ML infrastructure does not handle multiple thresholds.
# fragmentation - ML infrastructure does not fragmentation tuning per collective.
##################################
# Defining collective section
[BARRIER]
# Defining message size section. We will support small/large. In future we may add more options. Barrier is very specific case, because it is only collective that does not transfer any data, so for this specific case we use small
<small>
# Since ML does not define any algorithm for BARRIER, we just use default. Later we have to introduce some algorithm name for Barrier
algorithm = ML_BARRIER_DEFAULT
# Hierarchy setup:
#
# full_hr - means all possible levels of hierarchy (list of possible is defined by user command line)
# full_hr_no_basesocket - means all possible levels of hierarchy (list of possible is defined by user command line)
# except the basesocket subgroup.
# ptp_only - only ptp hierarchy
# iboffload_only - only iboffload hierarhcy
hierarchy = full_hr
[IBARRIER]
<small>
algorithm = ML_BARRIER_DEFAULT
hierarchy = full_hr
[BCAST]
<small>
# bcast supports: ML_BCAST_SMALL_DATA_KNOWN, ML_BCAST_SMALL_DATA_UNKNOWN, ML_BCAST_SMALL_DATA_SEQUENTIAL
algorithm = ML_BCAST_SMALL_DATA_KNOWN
hierarchy = full_hr
<large>
# bcast supports: ML_BCAST_LARGE_DATA_KNOWN, ML_BCAST_LARGE_DATA_UNKNOWN, ML_BCAST_LARGE_DATA_SEQUENTIAL
algorithm = ML_BCAST_LARGE_DATA_KNOWN
hierarchy = full_hr
[IBCAST]
<small>
algorithm = ML_BCAST_SMALL_DATA_KNOWN
hierarchy = full_hr
<large>
algorithm = ML_BCAST_LARGE_DATA_KNOWN
hierarchy = full_hr
[GATHER]
<small>
# gather supports: ML_SMALL_DATA_GATHER
algorithm = ML_SMALL_DATA_GATHER
hierarchy = full_hr
<large>
# gather supports: ML_LARGE_DATA_GATHER
algorithm = ML_LARGE_DATA_GATHER
hierarchy = full_hr
[IGATHER]
<small>
# gather supports: ML_SMALL_DATA_GATHER
algorithm = ML_SMALL_DATA_GATHER
hierarchy = full_hr
<large>
# gather supports: ML_LARGE_DATA_GATHER
algorithm = ML_LARGE_DATA_GATHER
hierarchy = full_hr
[ALLGATHER]
<small>
# allgather supports: ML_SMALL_DATA_ALLGATHER
algorithm = ML_SMALL_DATA_ALLGATHER
hierarchy = full_hr
<large>
# allgather supports: ML_LARGE_DATA_ALLGATHER
algorithm = ML_LARGE_DATA_ALLGATHER
hierarchy = full_hr
[IALLGATHER]
<small>
# allgather supports: ML_SMALL_DATA_ALLGATHER
algorithm = ML_SMALL_DATA_ALLGATHER
hierarchy = full_hr
<large>
# allgather supports: ML_LARGE_DATA_ALLGATHER
algorithm = ML_LARGE_DATA_ALLGATHER
hierarchy = full_hr
[ALLTOALL]
<small>
# alltoall supports: ML_SMALL_DATA_ALLTOALL
algorithm = ML_SMALL_DATA_ALLTOALL
hierarchy = ptp_only
<large>
# alltoall supports: ML_LARGE_DATA_ALLTOALL
algorithm = ML_LARGE_DATA_ALLTOALL
hierarchy = ptp_only
[IALLTOALL]
<small>
# alltoall supports: ML_SMALL_DATA_ALLTOALL
algorithm = ML_SMALL_DATA_ALLTOALL
hierarchy = ptp_only
<large>
# alltoall supports: ML_LARGE_DATA_ALLTOALL
algorithm = ML_LARGE_DATA_ALLTOALL
hierarchy = ptp_only
[ALLREDUCE]
<small>
# allreduce supports: ML_SMALL_DATA_ALLREDUCE
algorithm = ML_SMALL_DATA_ALLREDUCE
hierarchy = full_hr
<large>
# allreduce supports: ML_LARGE_DATA_ALLREDUCE
algorithm = ML_LARGE_DATA_ALLREDUCE
hierarchy = full_hr
[IALLREDUCE]
<small>
# allreduce supports: ML_SMALL_DATA_ALLREDUCE
algorithm = ML_SMALL_DATA_ALLREDUCE
hierarchy = full_hr
<large>
# allreduce supports: ML_LARGE_DATA_ALLREDUCE
algorithm = ML_LARGE_DATA_ALLREDUCE
hierarchy = full_hr
[REDUCE]
<small>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_SMALL_DATA_REDUCE
hierarchy = full_hr
<large>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_LARGE_DATA_REDUCE
hierarchy = full_hr
[IREDUCE]
<small>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_SMALL_DATA_REDUCE
hierarchy = full_hr
<large>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_LARGE_DATA_REDUCE
hierarchy = full_hr
[SCATTER]
<small>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_SCATTER_SMALL_DATA_SEQUENTIAL
hierarchy = full_hr
<large>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_SCATTER_SMALL_DATA_SEQUENTIAL
hierarchy = full_hr
[ISCATTER]
<small>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_SCATTER_SMALL_DATA_SEQUENTIAL
hierarchy = full_hr
<large>
# scatter supports: ML_SCATTER_SMALL_DATA_SEQUENTIAL
algorithm = ML_SCATTER_SMALL_DATA_SEQUENTIAL
hierarchy = full_hr
Показать git blame Копировать постоянную ссылку