Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Topic/a2a inplace5.0 #9493

Merged
merged 5 commits into from
Oct 14, 2021
Merged

Topic/a2a inplace5.0 #9493

merged 5 commits into from
Oct 14, 2021

Commits on Oct 7, 2021

  1. coll/basic: fix MPI_Alltoallw(MPI_IN_PLACE) gap handling 

    The temporary buffer must be shifted by the true_extent on a
per type basis (since the various datatypes might have different
true_extent).

Thanks Heiko Bauke for reporting this.

Refs. open-mpi#9329

Signed-off-by: Gilles Gouaillardet <[email protected]>
(cherry picked from commit 0041ce8)
    @ggouaillardet @jsquyres
    ggouaillardet authored and jsquyres committed Oct 7, 2021
    Configuration menu
    Copy the full SHA
    eab455f View commit details
    Browse the repository at this point in the history

  2. Expose opal_datatype_compute_remote_size. 

    This function can be used to compute the packed size of a datatype on a
target architecture.

Signed-off-by: George Bosilca <[email protected]>
(cherry picked from commit 74049fc)
    @bosilca @jsquyres
    bosilca authored and jsquyres committed Oct 7, 2021
    Configuration menu
    Copy the full SHA
    570b5ae View commit details
    Browse the repository at this point in the history

  3. Reduce the amount of temporary memory needed for MPI_Alltoallw. 

    Dont copy the datatype into a buffer with the same extent, but instead
pack it and send it to the peer as packed.

Signed-off-by: George Bosilca <[email protected]>
(cherry picked from commit 447b289)
    @bosilca @jsquyres
    bosilca authored and jsquyres committed Oct 7, 2021
    Configuration menu
    Copy the full SHA
    71bb6a6 View commit details
    Browse the repository at this point in the history

  4. A better MPI_IN_PLACE alltoallw algorithm. 

    Provide optimized variant for the homogeneous case.

Signed-off-by: George Bosilca <[email protected]>
(cherry picked from commit dc4e2ce)
    @bosilca @jsquyres
    bosilca authored and jsquyres committed Oct 7, 2021
    Configuration menu
    Copy the full SHA
    97f9f61 View commit details
    Browse the repository at this point in the history

  5. Better INPLACE algorithm for all2all and all2allv 

    Signed-off-by: George Bosilca <[email protected]>
(cherry picked from commit b9012a3)
    @bosilca @jsquyres
    bosilca authored and jsquyres committed Oct 7, 2021
    Configuration menu
    Copy the full SHA
    531735c View commit details
    Browse the repository at this point in the history