WeeklyTelcon_20160405

Open MPI Weekly Telcon

---

• Dialup Info: (Do not post to public mailing list or public wiki)

Attendees

• Brad Benton
• Geoffroy Vallee
• Howard
• Josh Hursey
• Joshua Ladd
• Nathan Hjelm
• Nysal
• Ralph
• Ryan Grant
• Todd Kordenbrock
• Geoff Paulsen

---

Agenda

Review 1.10

• Milestones: https://github.com/open-mpi/ompi-release/milestones/v1.10.3
• 1.10.3 in April - collecting some smaller fixes accumulating in the branch.
• waiting for discussion on one PR out there.

Review 2.0.x

• Wiki: https://github.com/open-mpi/ompi/wiki/Releasev20
• Blocker Issues: https://github.com/open-mpi/ompi/issues?utf8=%E2%9C%93&q=is%3Aopen+milestone%3Av2.0.0+label%3Ablocker
  • Issue 1505 - for v2.x merge - need PR on 2.0!
    • TCP BTL THREAD_MULTIPLE deadlock
      • Resolution: Bugfixing goodness.
    • New non-default feature: TCP async progress only in active if requested via requested via MCA param.
    • Ralph pointed out some compiler errors, that George just fixed yesterday, don't lose those.
    • Howard will merge to Master, and open a PR againt 2.0.0
  • Memory Hooks: PR 1495
    • move registration of memory manager to as late as possible.
    • MxM has it's own way of hooking calls, but doesn't play well with patcher methods put in.
      • Nathan should have a good understanding today.
      • Could turn off leave pinned for BTLs when using MxM (don't like this, but possible)
      • Should talk to Yoci. Setup a call, no reason to speculate.
    • Howard found another issue where need to set LEAVE_PINNED
    • If we can't fix this in a week, we may need to fall back to not pinning by default
    • Joshua will setup phone call with Yoci, Howard, Nathan, and Mark to discuss MxM.
  • -host merged to master - PR 1353
    • try to get into v2.0.0?
  • Failures in MTT - prob because USOCK component came in, and missing some commit, and causing failures on Master.
• Milestones: https://github.com/open-mpi/ompi-release/milestones/v2.0.0 *
• OMPI Release Open Pull Requests: https://github.com/open-mpi/ompi-release/pulls

Review Master?

• Master tests are failing.
• Josh question on PR1482.
  • legitimate concerns on mechanism.
  • Users confused why self

MTT status:

Status Updates:

1. Cisco
2. ORNL
3. UTK
4. NVIDIA

---

Status Update Rotation

1. Cisco, ORNL, UTK, NVIDIA
2. Mellanox, Sandia, Intel
3. LANL, Houston, IBM

---

Back to 2016 WeeklyTelcon-2016