This commit fixes typos on the C side of the request-based RMA binding. We
were not returning the request on success but on failure. Thanks to
@alazzaro for reporting and @ggouaillardet, and @vondele for tracking
this down.
Fixes part of open-mpi/ompi#1869
Signed-off-by: Nathan Hjelm <hjelmn@me.com>
This commit introduces a new algorithm for MPI_Comm_split_type. The
old algorithm performed an allgather on the communicator to decide
which processes were part of the new communicators. This does not
scale well in either time or memory.
The new algorithm performs a couple of all reductions to determine the
global parameters of the MPI_Comm_split_type call. If any rank gives
an inconsistent split_type (as defined by the standard) an error is
returned without proceeding further. The algorithm then creates a
communicator with all the ranks that match the split_type (no
communication required) in the same order as the original
communicator. It then does an allgather on the new communicator (which
should be much smaller) to determine 1) if the new communicator is in
the correct order, and 2) if any ranks in the new communicator
supplied MPI_UNDEFINED as the split_type. If either of these
conditions are detected the new communicator is split using
ompi_comm_split and the intermediate communicator is freed.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit simplifies the communicator context ID generation by
removing the blocking code. The high level calls: ompi_comm_nextcid
and ompi_comm_activate remain but now call the non-blocking variants
and wait on the resulting request. This was done to remove the
parallel paths for context ID generation in preperation for further
improvements of the CID generation code.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
the coll_array functions are truly only used by the fcoll modules, so move
them to fcoll/base. There is currently one exception to that rule (number of aggreagtors
logic), but that function will be moved in a long term also to fcoll/base.
This commit fixes a bug in the RDMA compare-and-swap implementation
that caused the origin value to always be written even if the compare
should have failed.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
as reported by Coverity with CIDs 1363349-1363362
Offset temporary buffer when a non zero lower bound datatype is used.
Thanks Hristo Iliev for the report
(cherry picked from commit 0e393195d9)
- correctly handle non commutative operators
- correctly handle non zero lower bound ddt
- correctly handle ddt with size > extent
- revamp NBC_Sched_op so it takes two buffers and matches ompi_op_reduce semantic
- various fix for inter communicators
Thanks Yuki Matsumoto for the report
It was introduced in PR https://github.com/open-mpi/ompi/pull/1228
in particular in commit 041a6a9f53.
Original solution was using "flexible array member" called "mxm_base"
to "fall-through" to the "mxm" send/recv member that located in the
outer structure.
After changing number of elements in "mxm_base" from 0 to 1 we actually
allocating 2 mxm_req_base_t elements which leads to increased overal
size and harms cache performance.
It also brakes "mca_pml_yalla_check_request_state" function.
* Matches the blocking versions of these interfaces
- `iallreduce.c` to match `allreduce.c`
- `ireduce.c` to match `reduce.c`
- `ireduce_scatter.c` to match `reduce_scatter.c`
* Workaround for IMB-NBC benchmark, similar to the workaround
in place for the IMB-MPI1 benchmark for the blocking collectives.
* If hcoll is given a negative priority, but not enabled=0 then
the module is constructed, but then destructed before calling
it's query(). So the previous pointers are not initialized.
If we try to OBJ_RELEASE them in a debug build an assert will fire.
This commit adds some protection against that and initializes
the _module pointers to NULL.
* Print a verbose message if the component was disqualified because of
a negative priority.
* If a disqualified component provided a module, release it.
* Display list of selected components in priority order
- During the process of volunteering collective functions for a
communicator, print the component name and priority. This will
cause the verbose messages to be displayed in reverse priority
order (lowest priority first, up to highest). This is helpful
when determining which collective components are active in which
order for a given communicator.
To see the messages you need the following MCA parameter set to 9
or higher: `-mca coll_base_verbose 9`
* Adjust verbose for commonly needed verbose output from 10 to 9 to
make it easier to access this information.
This commit fixes a hang reported by @nysal which happens when a
request is completed after a sync object is created but before the
sync object can be assigned to the request. In this case we need to
set the sync signaling field to false to ensure WAIT_SYNC_RELEASE does
not hang.
Fixesopen-mpi/ompi#1828
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit makes bml/r2 more restrictive on which endpoints end up in the rdma
endpoint list. Before this commit an endpoint was added if it supported either
put or get. This was done to ensure that endpoints are available for RMA.
Thought it is possible to support put or get endpoints we only currently
support endpoints that have put, get, and amos. bml/r2 now reflects this
support.
Signed-off-by: Nathan Hjelm <hjelmn@me.com>
The OPAL_ENABLE_MULTI_THREADS macro is always defined as 1. This was
causing us to always use the multi-thread path for synchronization
objects. The code has been updated to use the opal_using_threads()
function. When MPI_THREAD_MULTIPLE support is disabled at build time
(2.x only) this function is a macro evaluating to false so the
compiler will optimize out the MT-path in this case. The
OPAL_ATOMIC_ADD_32 macro has been removed and replaced by the existing
OPAL_THREAD_ADD32 macro.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
This commit fixes a performance regression introduced by the request
rework. We were always using the multi-thread path because
OPAL_ENABLE_MULTI_THREADS is either not defined or always defined to 1
depending on the Open MPI version. To fix this I removed the
conditional and added a conditional on opal_using_threads(). This path
will be optimized out in 2.0.0 in a non-thread-multiple build as
opal_using_threads is #defined to false in that case.
Fixesopen-mpi/ompi#1806
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Need to increment the total size after checking the local offset not
before. This typo causes large allocations with MPI_Win_allocate() to
fail.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Fixed an error where if there were no MPI exceptions, a
JNI error could still exist and not get handled.
Signed-off-by: Nathaniel Graham <nrgraham23@gmail.com>
It is valid for any rank to deviate on the split_type argument if they
specify MPI_UNDEFINED. The code was incorrectly not allowing this
condition. Changed the split_type uniformity check and allow
local_size to be 0 if the local split_type is MPI_UNDEFINED.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
