Skip to content

Conversation

@edgargabriel
Copy link
Member

data sieving has to occur for any offset provided that is larger
or equal zero for this implementation to work correctly.

Signed-off-by: Edgar Gabriel egabriel@central.uh.edu
(cherry picked from commit da640f9)

data sieving has to occur for any offset provided that is larger
or equal zero for this implementation to work correctly.

Signed-off-by: Edgar Gabriel <egabriel@central.uh.edu>
(cherry picked from commit da640f9)
@jsquyres
Copy link
Member

Per 2018-03-13 webex:

We talked about the series-of-3 PRs from Edgar/George today (for 2.x., 3.0.x, 3.1.x):

  • George: datatype fix
  • Edgar: ROMIO fix
  • Edgar: OMPI fix (sieving thing)

All 3 are necessary to make the latest HDF5 test suite pass.

Edgar's 2 fixes are small, self-contained, low risk. You merged the ROMIO fix on v3.0.x and v3.1.x yesterday.

The datatype fix seems higher risk.

We decided:

  • Let's merge Edgar's 2 fixes across all of 2.x, 3.0.x, 3.1.x. Simple, low risk. Fixes some things.
  • Let's wait on the datatype fix (but not hold up releases). Let's get more testing. If people need these fixes, they can get them in the nightlies after the release.

Only after all 3 fixes are in will the HDF5 test suite work. But we don't think that's a blocker for all of our immanent releases.

@bwbarrett bwbarrett merged commit 169afcd into open-mpi:v3.1.x Mar 13, 2018
@edgargabriel edgargabriel deleted the pr/v3.1.x-two_phase_data_sieving_fix branch March 21, 2018 19:38
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.

3 participants