Sessions pr #59
Open
Sessions pr #59
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
hppritcha
force-pushed
the
sessions_pr
branch
5 times, most recently
from
262c9da
to
8513720
Compare
Signed-off-by: Mamzi Bayatpour <mbayatpour@nvidia.com> Co-authored-by: Tomislav Janjusic <tomislavj@nvidia.com> Co-authored-by: Manjunath Gorentla Venkata <manjunath@nvidia.com>
Ensure requested memory size can be accommodated by the current number
of memory allocator buckets
Code review comments, change fprintfs to opal_show_help()
Signed-off-by: William P. LePera <lepera@us.ibm.com>
Signed-off-by: William P. LePera <lepera@us.ibm.com>
Signed-off-by: Valentin Petrov <valentinp@nvidia.com>
Signed-off-by: Mikhail Kurnosov <mkurnosov@gmail.com>
Extend the OPAL_3RDPARTY_WITH code to support packages which can be disabled (notably, PRRTE) and remove a bunch of duplicated code in the PRRTE configure check. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
If the user specifies an external PMIX or PRRTE (or disables PRRTE), skip running the sub-configure script. The advantage of this change is that it gives the user a way out (an external build) if there is an error in our configure logic for the sub projects (which has recently happened). The down side is that "make dist" is broken whenever OMPI is configured against an external PMIX/PRRTE library or PRRTE is disabled. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
coll/base/alltoall: fixes send/recv offsets in rbuf for the case of MPI_IN_PLACE
OSHMEM v1.5 API header
With a previous commit, it is now possible that configure will not run the PMIX or PRRTE configure scripts. This means that "make dist" will not recurse into the package's directory and those packages will unexpectedly not be in the dist tarball. This patch avoids that situation by explicitly breaking "make dist" in those situations, printing the package that caused the issue. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
…scatterv coll/basic: correctly handle zero size datatypes in gatherv/scatterv
…-upgrade Add a devel document to guide FT/ULFM upgrade of components
Initialization of a lock does not require sequential memory ordering so relax it to memory_order_relaxed. Signed-off-by: Joseph Schuchart <schuchart@icl.utk.edu>
…heck
Check for allocator bucket overflow
…res-if-external Skip running PMIX/PRRTE configure when not needed
…so_osc_rdma_can_be_used_out_of_the_box btl/uct: set reasonable defaults
btl/uct: Bump supported version to UCX 1.11.x
According to open-mpi#9637 (comment) UCT has been stable from UCX 1.9.0 so there is no need to check the maximum version anymore. Signed-off-by: Joseph Schuchart <schuchart@icl.utk.edu>
…ring MPI4: add support for mpi_info_get_string
…etc) Signed-off-by: Mamzi Bayatpour <mbayatpour@nvidia.com>
coll/hcoll: fixes fallback on rooted ops
UCT: require at least UCX 1.9.0
OSHMEM Coverity Fixes
Signed-off-by: Valentin Petrov <valentinp@nvidia.com>
Fix resource leak in error handling that was identified by coverity (CID 1497839 and 1497838). Signed-off-by: Brian Barrett <bbarrett@amazon.com>
…ariants MPI4: implement mpi_isendrecv and variant
…ity-fixes sharedfp/sm: Fix resource leak
…ld-cuda-plugin hwloc: enable --enable-plugins when appropriate
Remove the type-specific implementation of min in the btl active message code and use opal_min instead. While this loses type safety in the general case, most compilers used in development support C generics, so we'll still get type safety checks. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
Rename internal functions in the btl am-rdma code to be consistently prefixed by am_rdma_. Previously, there was a mixture of prefixes, all of which were unnecessarily long for private functions. There is no external-visible change in this commit. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
Disable the underlying flush() call when using the am-rdma interfce. The current code does not properly handle flushing between the am-rdma and underlying code properly, so it is quite easy to end up with flush not behaving the way users expect. Since there are currently no callers that rely on a btl having both flush and am-rdma interfaces, explicitly disable the case for now. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
Add per-btl storage for the am-rdma interface, allowing the interface to make one-time decisions about whether or not to use the underlying put/get interfaces. This is a necessary step in allowing users of the am-rdma interface to require different semantics (completion, registration, etc.) than what is natively provided by the BTL. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
Add a mode for the am-rdma interface where the existing BTL is not updated, but the wrapper functions are exported directly to the caller. This interface allows for specifying required flags (although REMOTE_COMPLETION is the only one supported today) and disabling memory registration requirements for the returned interface. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
The front end functions for am-rdma atomic operations only allow a 4 or 8 byte atomic. Add an assert on the progress function that the received size is less than a uint8_t can hold before assignment from a size_t. Signed-off-by: Brian Barrett <bbarrett@amazon.com>
Make base directories compile only the bare minimum when disabling ompio. These can make mca_ompio calls, which will lead to a linking error. Signed-off-by: Austen Lauria <awlauria@us.ibm.com>
Fix --disable-ompio builds.
Clean up atomics interface
…factor Allow explicit use of the Active Message RDMA interface
fix a missing ISENDRECV Signed-off-by: Howard Pritchard <howardp@lanl.gov>
see review comments for PR open-mpi#9667 Signed-off-by: Howard Pritchard <howardp@lanl.gov>
per review feedback for PR open-mpi#9667 Signed-off-by: Howard Pritchard <howardp@lanl.gov>
Signed-off-by: Austen Lauria <awlauria@us.ibm.com>
Coverity CID 1497862: Fix missing unlock in early return.
…eate_opt coll/ucc team create optimizations
…endrecv_replace manpages:mpi_isendrecv_replace
…ce_fix isendrecv_rplace: fix problem with req status
spc: fix up an update for isendrecv, etc.
…oxes btl/sm: avoid using xpmem for fast boxes
- suppressed compilation warnings Signed-off-by: Sergey Oblomov <sergeyo@nvidia.com>
…essed-warnings MCA/COMMON/UCX: suppressed compilation warnings
Signed-off-by: Nathan Hjelm <hjelmn@lanl.gov> Signed-off-by: Howard Pritchard <howardp@lanl.gov> Signed-off-by: Howard Pritchard <howardp@lanl.gov>
Signed-off-by: Howard Pritchard <howardp@lanl.gov>
hppritcha
force-pushed
the
sessions_pr
branch
from
536bd20
to
3aa5213
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The PR contains changes to Open MPI to support the MPI 4 Sessions API (see chapter 11 sections 3 and 4, and chapter 7 sections 3.2 and 4.2, and parts of chapter 9 concerning errorhandlers). The MPI 4 standard is available at https://www.mpi-forum.org/docs/mpi-4.0/mpi40-report.pdf.
A set of sessions tests has been added to the OMPI public tests repo:https://github.com/open-mpi/ompi-tests-public.
A set of issues concerning this PR are at https://github.com/hpc/ompi/issues
Major changes associated with this PR are:
MPI_Initand friends and after callingMPI_Finalizeand friends.A paper describing the prototype (at least at an early stage) is available at https://ieeexplore.ieee.org/abstract/document/8891002
There will be extensive squashing of commits prior to merging so it would probably be best to focus on the changes introduced by this PR as a whole rather than reviewing individual commits.