Skip to content

Conversation

@roiedanino
Copy link
Contributor

1. SPML Base Infrastructure

  • Added mca_spml_base_team_t structure with pSync and pWrk work arrays
  • Added mca_spml_base_alloc_sync_array() and mca_spml_base_free_sync_array() helper functions
  • Provides consistent symmetric heap allocation/deallocation across all SPML components

2. UCX SPML Implementation

  • Modified mca_spml_ucx_team_t to inherit from mca_spml_base_team_t
  • Updated team_split_strided() to allocate work arrays with proper error handling
  • Updated team_destroy() to free work arrays

3. Predefined Teams Initialization

  • Changed SHMEM_TEAM_WORLD and SHMEM_TEAM_SHARED to statically-allocated instances
  • Added work array allocation during shmem_init()
  • Added work array cleanup during shmem_finalize()

Signed-off-by: Roie Danino <rdanino@nvidia.com>
…ync and pWrk

Signed-off-by: Roie Danino <rdanino@nvidia.com>
@roiedanino
Copy link
Contributor Author

Hi @manjugv, can you please review?

This should be "Phase 1"

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant