[Android] Nuke the test server port lockfile explicitly.

[Android] Nuke the test server port lockfile explicitly.

ports.ResetTestServerPortAllocation is called in two locations, once in a
multithreaded context in which it's holding the lock file and once in a
singlethreaded context in which it isn't. It currently nukes the lock file,
which is fine in a singlethreaded context but can cause duplicate port
allocation in a multithreaded context. This CL will allow that function
to stop nuking the lock file in a subsequent CL to fix that issue.

BUG=631018
Committed: https://crrev.com/fdec47a6f1b4e7db662796344a115a72b6c0805f
Cr-Commit-Position: refs/heads/master@{#407791}

[jbudorick, 2016-07-26 02:27:44]

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

[mikecase (wrong mikecase), 2016-07-26 03:07:05]

mikecase@google.com changed reviewers:
+ mikecase@google.com

[mikecase (wrong mikecase), 2016-07-26 03:07:06]

https://codereview.chromium.org/2176183004/diff/1/build/android/test_runner.py
File build/android/test_runner.py (right):

https://codereview.chromium.org/2176183004/diff/1/build/android/test_runner.p...
build/android/test_runner.py:823: os.unlink(ports._TEST_SERVER_PORT_LOCKFILE) #
pylint: disable=protected-access
do you have to add an...

if os.path.exists(ports._TEST_SERVER_PORT_LOCKFILE)?

It seems os.unlink raises an exception if the file is not found, and it looks
like
the file wont exist since it is currently deleted by
ports.ResetTestServerPortAllocation()

[jbudorick, 2016-07-26 03:24:31]

https://codereview.chromium.org/2176183004/diff/1/build/android/test_runner.py
File build/android/test_runner.py (right):

https://codereview.chromium.org/2176183004/diff/1/build/android/test_runner.p...
build/android/test_runner.py:823: os.unlink(ports._TEST_SERVER_PORT_LOCKFILE) #
pylint: disable=protected-access
On 2016/07/26 03:07:06, mikecase (wrong mikecase) wrote:
> do you have to add an...
> 
> if os.path.exists(ports._TEST_SERVER_PORT_LOCKFILE)?
> 
> It seems os.unlink raises an exception if the file is not found, and it looks
> like
> the file wont exist since it is currently deleted by
> ports.ResetTestServerPortAllocation()

Good catch. Fixed.

[mikecase (wrong mikecase), 2016-07-26 13:41:05]

lgtm

[jbudorick, 2016-07-26 13:41:50]

The CQ bit was checked by jbudorick@chromium.org

[commit-bot: I haz the power, 2016-07-26 13:42:06]

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, 2016-07-26 14:36:18]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-07-26 14:40:48]

Description was changed from

==========
[Android] Nuke the test server port lockfile explicitly.

ports.ResetTestServerPortAllocation is called in two locations, once in a
multithreaded context in which it's holding the lock file and once in a
singlethreaded context in which it isn't. It currently nukes the lock file,
which is fine in a singlethreaded context but can cause duplicate port
allocation in a multithreaded context. This CL will allow that function
to stop nuking the lock file in a subsequent CL to fix that issue.

BUG=631018
==========

to

==========
[Android] Nuke the test server port lockfile explicitly.

ports.ResetTestServerPortAllocation is called in two locations, once in a
multithreaded context in which it's holding the lock file and once in a
singlethreaded context in which it isn't. It currently nukes the lock file,
which is fine in a singlethreaded context but can cause duplicate port
allocation in a multithreaded context. This CL will allow that function
to stop nuking the lock file in a subsequent CL to fix that issue.

BUG=631018

Committed: https://crrev.com/fdec47a6f1b4e7db662796344a115a72b6c0805f
Cr-Commit-Position: refs/heads/master@{#407791}
==========

[commit-bot: I haz the power, 2016-07-26 14:40:49]

Patchset 2 (id:??) landed as
https://crrev.com/fdec47a6f1b4e7db662796344a115a72b6c0805f
Cr-Commit-Position: refs/heads/master@{#407791}