Skip to content
This repository has been archived by the owner on Jan 3, 2023. It is now read-only.

Fixes NGRAPH-1030: segfault in train_cifar10.py #181

Merged
merged 1 commit into from
Mar 26, 2018

Conversation

cconvey
Copy link
Contributor

@cconvey cconvey commented Mar 26, 2018

  • Works around problem in which if mxnet::engine::OpenMP::OpenmP()
    calls omp_set_num_threads(...) then a segfault occurs in later.
    The workaround only addresses cases where libmxnet.so is accessed
    via our Python module.

Description

(Brief description on what this PR is about)

Checklist

Essentials

  • Passed code style checking (make lint)
  • Changes are complete (i.e. I finished coding on this PR)
  • All changes have test coverage:
  • Unit tests are added for small changes to verify correctness (e.g. adding a new operator)
  • Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore)
  • Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL)
  • Code is well-documented:
  • For user-facing API changes, API doc string has been updated.
  • For new C++ functions in header files, their functionalities and arguments are documented.
  • For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable
  • To the my best knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Changes

  • Feature1, tests, (and when applicable, API doc)
  • Feature2, tests, (and when applicable, API doc)

Comments

  • If this change is a backward incompatible change, why must this change be made.
  • Interesting edge cases to note here

@cconvey cconvey requested a review from mbrookhart March 26, 2018 16:48
Copy link
Contributor

@mbrookhart mbrookhart left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍 on base.py and the readme, but why is setup.py changed?

* Works around problem in which if `mxnet::engine::OpenMP::OpenmP()`
  calls `omp_set_num_threads(...)` then a segfault occurs in later.
  The workaround only addresses cases where `libmxnet.so` is accessed
  via our Python module.
@cconvey cconvey force-pushed the cconvey/ngmx-392-cifar10-segfault branch from 6b499c8 to 33272a4 Compare March 26, 2018 17:14
@cconvey cconvey changed the title Fixes NGMX-392: segfault in train_cifar10.py Fixes NGRAPH-1030: segfault in train_cifar10.py Mar 26, 2018
@cconvey
Copy link
Contributor Author

cconvey commented Mar 26, 2018

@mbrookhart :

👍 on base.py and the readme, but why is setup.py changed?

The python code now requires the psutil module, which isn't installed by default. That change adds the dependency. The source-formatting change was due to the line getting really long.

@mbrookhart
Copy link
Contributor

TESTNOW

@cconvey cconvey merged commit cdf879c into ngraph-integration-dev Mar 26, 2018
@mbrookhart mbrookhart deleted the cconvey/ngmx-392-cifar10-segfault branch March 26, 2018 18:20
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants