Workaround python time.strptime flake in Android tests

Workaround python time.strptime flake in Android tests

In general time.strptime is not threadsafe in python: [1]. This causes
flaky errors on various bots that test Android because catapult [2]
talks to devices from different threads (for multiple reasons).

One observation in the python bug ([1]) says that only importing parts
of strptime is racy, so importing it early enough before any threading
is started would remove this kind of flake.

This workaround is used in other parts of chromium and catapult itself,
so we can probably be more preventive by doing it in a central place.
For now let's just reduce the flake.

[1] time.strptime not thread safe
    https://bugs.python.org/issue7980

[2] Catapult is the home for several performance tools that span from
    gathering, displaying and analyzing performance data.
    https://github.com/catapult-project/catapult

BUG=724524
Review-Url: https://codereview.chromium.org/2891403002
Cr-Commit-Position: refs/heads/master@{#473545}
Committed: https://chromium.googlesource.com/chromium/src/+/bfb763ee9499a5f88993910c2401ac5550e9ac38

[pasko, 2017-05-19 16:58:51]

The CQ bit was checked by pasko@chromium.org to run a CQ dry run

[pasko, 2017-05-19 16:59:22]

Description was changed from

==========
Workaround python time.strptime flake in Android tests

In general time.strptime is not threadsafe in python: [1]. This causes
flaky errors on various bots that test Android because catapult [2]
talks to devices from different threads (for multiple reasons).

One observation in the python bug ([1]) says that only importing parts
of strptime is racy, so importing it early enough before any threading
is started would remove this kind of flake.

This workaround is used in other parts of chromium and catapult itself,
so we can probably be more preventive by doing it in a central place.
For now let's just reduce the flake.

Also presubmit complained about test_runner pydeps, so I regenerated
them with:

shell> build/print_python_deps.py --root build/android --output \
    build/android/test_runner.pydeps build/android/test_runner.py

[1] time.strptime not thread safe
    https://bugs.python.org/issue7980

[2] Catapult is the home for several performance tools that span from
    gathering, displaying and analyzing performance data.
    https://github.com/catapult-project/catapult

BUG=724524
==========

to

==========
Workaround python time.strptime flake in Android tests

In general time.strptime is not threadsafe in python: [1]. This causes
flaky errors on various bots that test Android because catapult [2]
talks to devices from different threads (for multiple reasons).

One observation in the python bug ([1]) says that only importing parts
of strptime is racy, so importing it early enough before any threading
is started would remove this kind of flake.

This workaround is used in other parts of chromium and catapult itself,
so we can probably be more preventive by doing it in a central place.
For now let's just reduce the flake.

Also presubmit complained about test_runner pydeps, so I regenerated
them with:

shell> build/print_python_deps.py --root build/android --output \
    build/android/test_runner.pydeps build/android/test_runner.py

[1] time.strptime not thread safe
    https://bugs.python.org/issue7980

[2] Catapult is the home for several performance tools that span from
    gathering, displaying and analyzing performance data.
    https://github.com/catapult-project/catapult

BUG=724524
==========

[pasko, 2017-05-19 16:59:22]

pasko@chromium.org changed reviewers:
+ jbudorick@chromium.org

[commit-bot: I haz the power, 2017-05-19 16:59:32]

Dry run: CQ is trying da patch.

Follow status at:
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[jbudorick, 2017-05-19 18:04:50]

lgtm w/ pydeps changes

https://codereview.chromium.org/2891403002/diff/1/build/android/test_runner.p...
File build/android/test_runner.pydeps (left):

https://codereview.chromium.org/2891403002/diff/1/build/android/test_runner.p...
build/android/test_runner.pydeps:117: ../../third_party/jinja2/__init__.py
This is wrong & is due to a bug in our importing of these two libraries.. Rebase
on https://codereview.chromium.org/2896583003/ & pull this out.

[commit-bot: I haz the power, 2017-05-19 18:25:28]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-05-19 18:25:29]

Dry run: Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[pasko, 2017-05-22 09:22:27]

Description was changed from

==========
Workaround python time.strptime flake in Android tests

In general time.strptime is not threadsafe in python: [1]. This causes
flaky errors on various bots that test Android because catapult [2]
talks to devices from different threads (for multiple reasons).

One observation in the python bug ([1]) says that only importing parts
of strptime is racy, so importing it early enough before any threading
is started would remove this kind of flake.

This workaround is used in other parts of chromium and catapult itself,
so we can probably be more preventive by doing it in a central place.
For now let's just reduce the flake.

Also presubmit complained about test_runner pydeps, so I regenerated
them with:

shell> build/print_python_deps.py --root build/android --output \
    build/android/test_runner.pydeps build/android/test_runner.py

[1] time.strptime not thread safe
    https://bugs.python.org/issue7980

[2] Catapult is the home for several performance tools that span from
    gathering, displaying and analyzing performance data.
    https://github.com/catapult-project/catapult

BUG=724524
==========

to

==========
Workaround python time.strptime flake in Android tests

In general time.strptime is not threadsafe in python: [1]. This causes
flaky errors on various bots that test Android because catapult [2]
talks to devices from different threads (for multiple reasons).

One observation in the python bug ([1]) says that only importing parts
of strptime is racy, so importing it early enough before any threading
is started would remove this kind of flake.

This workaround is used in other parts of chromium and catapult itself,
so we can probably be more preventive by doing it in a central place.
For now let's just reduce the flake.

[1] time.strptime not thread safe
    https://bugs.python.org/issue7980

[2] Catapult is the home for several performance tools that span from
    gathering, displaying and analyzing performance data.
    https://github.com/catapult-project/catapult

BUG=724524
==========

[pasko, 2017-05-22 09:23:41]

thank you for review, John

landing..

https://codereview.chromium.org/2891403002/diff/1/build/android/test_runner.p...
File build/android/test_runner.pydeps (left):

https://codereview.chromium.org/2891403002/diff/1/build/android/test_runner.p...
build/android/test_runner.pydeps:117: ../../third_party/jinja2/__init__.py
On 2017/05/19 18:04:49, jbudorick wrote:
> This is wrong & is due to a bug in our importing of these two libraries..
Rebase
> on https://codereview.chromium.org/2896583003/ & pull this out.

Done.

[pasko, 2017-05-22 09:23:47]

The CQ bit was checked by pasko@chromium.org

[pasko, 2017-05-22 09:23:48]

The patchset sent to the CQ was uploaded after l-g-t-m from
jbudorick@chromium.org
Link to the patchset: https://codereview.chromium.org/2891403002/#ps20001
(title: "revert test_runner.pydeps to where the master is")

[commit-bot: I haz the power, 2017-05-22 09:23:58]

CQ is trying da patch.

Follow status at:
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-05-22 10:46:01]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1495445027942770,
"parent_rev": "e630002e04d5b0e9436a325f224df82fa694bc26", "commit_rev":
"bfb763ee9499a5f88993910c2401ac5550e9ac38"}

[commit-bot: I haz the power, 2017-05-22 10:46:10]

Description was changed from

==========
Workaround python time.strptime flake in Android tests

In general time.strptime is not threadsafe in python: [1]. This causes
flaky errors on various bots that test Android because catapult [2]
talks to devices from different threads (for multiple reasons).

One observation in the python bug ([1]) says that only importing parts
of strptime is racy, so importing it early enough before any threading
is started would remove this kind of flake.

This workaround is used in other parts of chromium and catapult itself,
so we can probably be more preventive by doing it in a central place.
For now let's just reduce the flake.

[1] time.strptime not thread safe
    https://bugs.python.org/issue7980

[2] Catapult is the home for several performance tools that span from
    gathering, displaying and analyzing performance data.
    https://github.com/catapult-project/catapult

BUG=724524
==========

to

==========
Workaround python time.strptime flake in Android tests

In general time.strptime is not threadsafe in python: [1]. This causes
flaky errors on various bots that test Android because catapult [2]
talks to devices from different threads (for multiple reasons).

One observation in the python bug ([1]) says that only importing parts
of strptime is racy, so importing it early enough before any threading
is started would remove this kind of flake.

This workaround is used in other parts of chromium and catapult itself,
so we can probably be more preventive by doing it in a central place.
For now let's just reduce the flake.

[1] time.strptime not thread safe
    https://bugs.python.org/issue7980

[2] Catapult is the home for several performance tools that span from
    gathering, displaying and analyzing performance data.
    https://github.com/catapult-project/catapult

BUG=724524

Review-Url: https://codereview.chromium.org/2891403002
Cr-Commit-Position: refs/heads/master@{#473545}
Committed:
https://chromium.googlesource.com/chromium/src/+/bfb763ee9499a5f88993910c2401...
==========

[commit-bot: I haz the power, 2017-05-22 10:46:11]

Committed patchset #2 (id:20001) as
https://chromium.googlesource.com/chromium/src/+/bfb763ee9499a5f88993910c2401...