without this fix, an error handler invoked on pml_ucx request would
segfault while trying to dereference requests[i]->req_mpi_object.comm
Signed-off-by: Yossi Itigin <yosefe@mellanox.com>
Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
Correctly bubble up errors in NBC collective operations
Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
The error field of requests needs to be rearmed at start, not at create
Signed-off-by: Aurelien Bouteiller <bouteill@icl.utk.edu>
So far Vader is faster than OFI MTL for doing shared memory.
Therefore, let it run by default when all procs are local.
Reviewed-by: Spruit, Neil R <neil.r.spruit@intel.com>
Reviewed-by: Gopalakrishnan, Aravind <aravind.gopalakrishnan@intel.com>
Signed-off-by: Matias Cabral <matias.a.cabral@intel.com>
Now Open MPI requires a C99 compiler. Checking availability of
the following types is no more needed.
- `long long` (`signed` and `unsigned`)
- `long double`
- `float _Complex`
- `double _Complex`
- `long double _Complex`
Furthermore, the `#if HAVE_[TYPE]` style checking is not correct.
Availability of C types is checked by `AC_CHECK_TYPES` in `configure.ac`.
`AC_CHECK_TYPES` defines macro `HAVE_[TYPE]` as `1` in `opal_config.h`
if the `[TYPE]` is available. But it does not define `HAVE_[TYPE]`
(instead of defining as `0`) if it is not available. So even if we
need `HAVE_[TYPE]` checking, it should be `#if defined(HAVE_[TYPE])`.
I didn't remove `AC_CHECK_TYPES` for these types in `configure.ac`
since someone may use `HAVE_[TYPE]` macros somewhere.
Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
When the provider does not support FI_REMOTE_CQ_DATA, the OFI tag does not have
sizeof(int) bits for the rank. Therefore, unexpected behavior will occur when
this limit is crossed.
Check the max allowed number of ranks during add_procs() and return if there is
danger of exceeding this threshold.
Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
1. Remove debug output in iallgather (I have forgotten to remove it).
2. Remove an incorrect comment in description of ibcast
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
PR #5450 addresses MPI_IN_PLACE processing for basic collective algorithms.
But in conjunction with that, we need to check for MPI_IN_PLACE in tuned paths
as well before calling ompi_datatype_type_size() as otherwise we segfault.
MPI spec also stipulates to ignore sendcount and sendtype for Alltoall and
Allgatherv operations. So, extending the check to these algorithms as well.
Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
Some compilers complain when comparing signed and unsigned. romio321
was doing just this. The check is meant to check whether a size (which
is an ADIO_Offset-- a signed number) will work with memcpy which takes
a size_t. To silence the warning I added a new type (ADIO_Size) which
is an unsigned type and cast the ADIO_Offset to this new type.
Fixes#5951
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
The monitoring PML hides it's existence from the OMPI infrastructure by
removing itself from the list of PML loaded components, remaining hidden
until MPI_Finalize.
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
With this patch the best PML is selected earlier, before finalizing
the others PML. This provides a simpler mechanism to intercept and
highjack the PML (as done in the monitoring PML)
Signed-off-by: George Bosilca <bosilca@icl.utk.edu>
An implementation of R. Rabenseifner's algorithm for MPI_Iallreduce.
This algorithm is a combination of a reduce-scatter implemented with recursive vector halving
and recursive distance doubling, followed either by an allgather.
Limitations:
-- count >= 2^{\floor{\log_2 p}}
-- commutative operations only
-- intra-communicators only
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
check for providing a data representation that is actually supported
by ompio.
Add also one check for a non-NULL pointer in mpi/c/file_set_view
for the data representation.
Also fixes parts of issue #5643
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
Several fixes to string handling:
1. strncpy() -> opal_string_copy() (because opal_string_copy()
guarantees to NULL-terminate, and strncpy() does not)
2. Simplify a few places, such as:
* Since opal_string_copy() guarantees to NULL terminate, eliminate
some memsets(), etc.
* Use opal_asprintf() to eliminate multi-step string creation
There's more work that could be done; e.g., this commit doesn't
attempt to clean up any strcpy() usage.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
this ensures that all processes are done modifying a file
before syncing. Fixes an error in the testmpio testsuite.
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
return MPI_ERR_ARG if the size of the fileview is not a
multiple of the size of the etype provided.
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
return MPI_ERR_ACCESS if the user tries to read from a file
that was opened using MPI_MODE_WRONLY
return MPI_ERR_READ_ONLY if the user tries to write a file
that was opened using MPI_MODE_RDONLY
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
Implements recursive doubling algorithm for MPI_Iallgather.
The algorithm can be used only for power-of-two number of processes.
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
Make sure all pending communications are done on all ranks before
closing the window. This way it will be safe to close the endpoints when
closing the component.
Signed-off-by: Yossi Itigin <yosefe@mellanox.com>
An implementation of R. Rabenseifner's algorithm for MPI_Ireduce.
This algorithm is a combination of a reduce-scatter implemented with recursive vector halving
and recursive distance doubling, followed either by a gather.
Limitations:
-- count >= 2^{\floor{\log_2 p}}
-- commutative operations only
-- intra-communicators only
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
The Open MPI code base assumed that asprintf always behaved like
the FreeBSD variant, where ptr is set to NULL on error. However,
the C standard (and Linux) only guarantee that the return code will
be -1 on error and leave ptr undefined. Rather than fix all the
usage in the code, we use opal_asprintf() wrapper instead, which
guarantees the BSD-like behavior of ptr always being set to NULL.
In addition to being correct, this will fix many, many warnings
in the Open MPI code base.
Signed-off-by: Brian Barrett <bbarrett@amazon.com>
Change default provider selection logic for the OFI MTL. The
old logic was whitelist-only, so any new HPC NIC provider would
have to ask users to do extra work or wait for an OMPI release
to be whitelisted. The reason for the logic was to avoid
selecting a "generic" provider like sockets or shm that would
frequently have worse performance than the optimized BTL options
Open MPI supports.
With the change, we blacklist the (small, relatively static) list
of providers that duplicate internal capabilities. Users can use
one of thse blacklisted providers in two ways: first, they can
explicitly request the provider in the include list (which will
override the default exclude list) and second, the can set a new
empty exclude list.
Since most HPC networks require special libraries and therefore
an explicit build of libfabric, it is highly unlikely that this
change will cause users to use libfabric when they didn't want to
do so. It does, however, solve the whitelisting problem.
Signed-off-by: Brian Barrett <bbarrett@amazon.com>
Implements recursive doubling algorithm for MPI_Iexscan.
The algorithm preserves order of operations so it can be used both
by commutative and non-commutative operations.
The MCA parameter 'coll_libnbc_iexscan_algorithm' was added for dynamic
algorithm selection.
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
Implements recursive doubling algorithm for MPI_Iscan. The algorithm preserves order of operations so it can be used both by commutative and non-commutative operations.
The MCA parameter coll_libnbc_iscan_algorithm was added for dynamic algorithm selection.
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
The parameter passed to NBC_Return_handle() was incorrectly casted
and not dereferenced.
Thanks Yossi for the bug report.
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
Move includes to top of file. Set some #defines so that
monitoring_prof.c compiles without warning (as identified by gcc 8 on
MacOS). Also ensure to include the internal Open MPI "mpi.h" file
(not some random system <mpi.h> file).
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
Gcc 8 identified hb_tree_csearch() as an infinite recursion, and it
turns out that we never call this function, anyway. So just remove
it.
Fixes#5670.
Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
This commit updates the entire codebase to use specific opal types for
all atomic variables. This is a change from the prior atomic support
which required the use of the volatile keyword. This is the first step
towards implementing support for C11 atomics as that interface
requires the use of types declared with the _Atomic keyword.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
always initialize 'size'.
Only the a2a_sched_diss() alltoall algorithm is impacted,
and this algo is currently unused, so there is no need
to backport nor update the NEWS file for now.
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
file_delete triggers underneath the hood the full component selection
logic, since we do not have a file handle, just a file name.
As part of the selection logic, we have to however initiate the
framework-open of the fs component in case of ompio, since ompio
will call the delete function of the selected fs componentn, which
is based on the file system where the file is located.
This was not handled correctly so far. The problem however only
shows up if the first I/O operatin to be executed is a file_delete,
other wise the file_open will lead to the correct opening and initialization
of the fs framework. This commit ensures that we do the right thing
even if file_delete is the first file I/O operation in the application.
Fixes issue #5611
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
If an application opens a file for reading from multiple processes
using MPI_COMM_SELF (or another communicator that has distinct
process groups but the same comm-id, as can happen as the result
of comm_split), the naming chosen for the lockedfile or the mmapped
file used by the sharedfp/sm component would collide. This patch
ensures that the filename is different by integrating the process id
of rank 0 for each sub-communicator.
This fixes one aspect of the problem reported in github issue 5593
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
The aggregation code in osc/rdma is currently broken and will likely
not be reused. This commit cleans it out.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
When an application is not using multiple threads to call into MPI, we can
safely ask for FI_THREAD_DOMAIN setting from the provider as it should
translate to the least amount of locking in provider.
Conversely, for applications using THREAD_MULTIPLE, explicitly ask for
FI_THREAD_SAFE to prevent race conditions.
Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
gcc complains about ret possibly being used uninitialized. That will
never happen but we should still quiet the warning. This commit sets
ret to a valid value.
Fixes#5513
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
the sharedfp component has to be selected and opened before
we set the default file view during file_open. Otherwise
there is a sperious error message from the sharefp_file_seek
operation that is called during the file_set_view.
Fixes Issue #5560
Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
Changes made:
- Create a new fs component for IME
- Create a new fbtl component for IME
- Modify the close function of OMPIO to finalize IME if necessary
Signed-off-by: Gaëtan Bossu <gbossu@ddn.com>
Signed-off-by: Sylvain Didelot <sdidelot@ddn.com>
Since progress entries array is globally allocated, it is susceptible
to race conditions when using multi-threaded applications. Allocating it
on the stack resolves any potential races as it is thread local by default.
Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
This commit updates the new custom matching code in pml/ob1 so it can
not be enabled with a configure option. This commit also renames the
fuzzy-matching headers to avoid potential name conflicts and removes
the use of C reserved identifiers.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
In the cleanup phase, it is possible for PtlMEUnlink() to return
PTL_IN_USE if the NIC is not done with the ME. This should not
be considered an error. This commit adds a retry loop around
PtlMEUnlink().
In some cases, the return value of PtlMEUnlink() and PtlCTFree()
was not checked at all. Check them with the same retry loop as
above.
Signed-off-by: Todd Kordenbrock <thkgcode@gmail.com>
When romio314 was first pulled in an extra patch was applied to it, see commit
92f6c7c1e2. Most of that patch is already present
in vanilla romio321, but the fix for MPIO_DATATYPE_ISCOMMITTED() isn't.
If that macro doesn't set err_ then some paths end up with a variable being used
uninitialized. In particular you can trace through romio321/romio/mpi-io/read.c
to see what happens with error_code. It's an uninitialized stack variable that goes
through three MPIO_CHECK_* macros none of which set it. The macros consistently set
error_code to a failure if they see something wrong, but they don't consistently
set it to success when things are fine.
And then in the last macro MPIO_CHECK_DATATYPE it tries to look at the value
of error_code that was never set.
Signed-off-by: Mark Allen <markalle@us.ibm.com>
The call of MPI_Allgatherv with sendbuf and sendtype parameters equal to MPI_IN_PLACE and NULL correspondingly, produces the segmentation fault.
The problem is that sendtype is used even when sendbuf value is MPI_IN_PLACE. But according to the standard, sendtype and sendcount parameters should be ignored in this case.
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
-Updated blocking send to directly call functionality and
set completion events expected to 0 initally. This allows for optimization for
providers that support fi_tinject up to larger sizes. This also reduces
latency on running the OFI mtl with smaller sizes without requiring
calls to progress given fi_tinject is required to complete the messaging
before returning and will not create any events in the Completion Queue.
-Updated non-blocking send to directly call fi_tsend and avoid calling
fi_tinject as the functionality should not wait on completions. This
resolves a bug where applications calling MPI_Isend can overrun the
TX buffer with small (inject) messages causing a deadlock. In addition
this improves performance in message rates by preventing
waiting on any size message to complete in non-blocking send messages.
-Created common ompi_mtl_ofi_ssend_recv function to post the ssend recv
which is common between isend and send code paths.
Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com>
- If a message for a recv that is being cancelled gets completed after
the call to fi_cancel, then the OFI mtl will enter a deadlock state
waiting for ofi_req->super.ompi_req->req_status._cancelled which will
never happen since the recv was successfully finished.
- To resolve this issue, the OFI mtl now checks ofi_req->req_started
to see if the request has been started within the loop waiting for the
event to be cancelled. If the request is being completed, then the loop
is broken and fi_cancel exits setting
ofi_req->super.ompi_req->req_status._cancelled = false;
Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com>
OFI providers may reserve some of the upper bits of the tag for
internal usage and expose it using mem_tag_format. Check for that
and adjust communicator bits as needed.
Signed-off-by: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@intel.com>
- in sine cases persistent request was deleted during completion
callback, this cause double free of linked UCX request (assert
in debug build or hang in release build)
- UCX request is freed prior completion calback
Signed-off-by: Sergey Oblomov <sergeyo@mellanox.com>
The call of MPI_Allgather with sendbuf and sendtype parameters equal to MPI_IN_PLACE and NULL correspondingly, produces the segmentation fault.
The problem is that sendtype is used even when sendbuf value is MPI_IN_PLACE. But according to the standard, sendtype and sendcount parameters should be ignored in this case.
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
OMPIO now uses the correct delete function depending on the fs
mca_common_ompio_file_delete now works this way instead
of calling POSIX unlink:
- create a minimal file handle with the given file name
- select the best fs component using this file handle
- call the component-specific file delete function
Signed-off-by: Gaëtan Bossu <gbossu@ddn.com>
It is needed because the fs components might be queried due to a MPI_File_delete call.
And in this case, we don't have a communicator value.
Signed-off-by: Gaëtan Bossu <gbossu@ddn.com>
instead of invoking ompi_request_test_all(), that will end up
calling opal_progress() recursively, manually check the status
of the requests.
the same method is used in ompi_comm_request_progress()
Refs open-mpi/ompi#3901
Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
- Added support in MTL_OFI_RETRY_UNTIL_DONE to handle -FI_EAGAIN
from the provider and correctly attempt to progress the OFI Completion
queue by calling ompi_mtl_ofi_progress.
- If events were pending that blocked OFI operations from being enqueued
they will be completed and the OFI operation will be retried once
ompi_mtl_ofi_progress has successfully completed.
- Updated MTL_OFI_RETRY_UNTIL_DONE to take a RETURN variable instead of
requiring the existance of a "ret" variable to pass back the return
value from completing the OFI operation.
Signed-off-by: Spruit, Neil R <neil.r.spruit@intel.com>
Move memory hooks init (for request based operation) in osc ucx to window
creation time, to avoid performance issue in MPI initialization.
Signed-off-by: Xin Zhao <xinz@mellanox.com>
This commit fixes a crash that occurs when using btl/vader as an RDMA
btl. This btl supports using CPU atomics and does not support using
the btl for self communication so we must use the local memory
optimizations in osc/rdma.
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
Data transferred by `MPI_BSEND` may corrupt if all of the following
conditions are met.
- The message size is less than the eager limit.
- The `btl_alloc` function in the BTL interface returns `NULL`
for some reason.
- The MPI program overwrites the send buffer after `MPI_BSEND`
returns.
The problem is in the way of pending a send request in ob1 PML.
The `mca_pml_ob1_send_request_start_copy` function retruns
`OMPI_ERR_OUT_OF_RESOURCE` if `mca_bml_base_alloc` function returns
`des = NULL`. In this case, the send request is added to the
`send_pending` list and `MPI_BSEND` returns immediately. Next time
the `mca_pml_ob1_send_request_start_copy` function tries sending,
the user buffer may have been overwritten by the MPI program.
Call hierarchy of `MPI_BSEND`:
```
MPI_Bsend
mca_pml_ob1_send
if (MCA_PML_BASE_SEND_BUFFERED == sendmode)
mca_pml_ob1_isend
MCA_PML_OB1_SEND_REQUEST_START_W_SEQ
mca_pml_ob1_send_request_start_seq
mca_pml_ob1_send_request_start_btl
if (size <= eager_limit)
if (req_send_mode == MCA_PML_BASE_SEND_BUFFERED)
mca_pml_ob1_send_request_start_copy
mca_bml_base_alloc
btl_alloc
if (OMPI_ERR_OUT_OF_RESOURCE == rc)
add_request_to_send_pending
ompi_request_free
```
To solve this problem, we should save the data to the buffer
attached by `MPI_BUFFER_ATTACH` before leaving `MPI_BSEND`.
This problem was introduced by ob1 optimization (commits 2b57f422
and a06e491c) in v1.8 series.
Signed-off-by: KAWASHIMA Takahiro <t-kawashima@jp.fujitsu.com>
This commit fixes an issue identified by MTT where we can have two
different sets of processes on the same node creating a shared memory
window with communicators sharing the same CID. To avoid this issue
the temporary filename now includes the creating processes vpid.
References #5363
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov>
