Android test_runner.py: Add --enable-device-cache to speed up file pushing

Android test_runner.py: Add --enable-device-cache to speed up file pushing

Speeds up repeated runs of the same test by at least 2 seconds. Much more for
tests with a large isolate payload.

BUG=540857
Committed: https://crrev.com/a538a143ab313bb1b89a0c283baac4fafa50c614
Cr-Commit-Position: refs/heads/master@{#353297}

[agrieve, 2015-10-08 03:08:11]

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

[agrieve, 2015-10-08 03:08:22]

On 2015/10/08 03:08:11, agrieve wrote:
> mailto:agrieve@chromium.org changed reviewers:
> + mailto:jbudorick@chromium.org

.

[jbudorick, 2015-10-08 14:34:49]

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
File build/android/pylib/local/device/local_device_environment.py (right):

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
build/android/pylib/local/device/local_device_environment.py:14: from pylib
import constants
nit: this should precede pylib.base.environment

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
build/android/pylib/local/device/local_device_environment.py:80: # Write the
cache even when not using it so that a bad cache will be fixed
I lost you on the reasoning here as written. I can understand writing the cache
s.t. the cache is kept valid on each run & can be used when going from disabled
-> enabled, but I'm not following the bad cache fix part.

The possible issue I see with that is the extra time in doing this when we
aren't using the device cache -- specifically, on the bots.

[agrieve, 2015-10-08 16:03:28]

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
File build/android/pylib/local/device/local_device_environment.py (right):

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
build/android/pylib/local/device/local_device_environment.py:14: from pylib
import constants
On 2015/10/08 14:34:49, jbudorick wrote:
> nit: this should precede pylib.base.environment

Done.

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
build/android/pylib/local/device/local_device_environment.py:80: # Write the
cache even when not using it so that a bad cache will be fixed
On 2015/10/08 14:34:49, jbudorick wrote:
> I lost you on the reasoning here as written. I can understand writing the
cache
> s.t. the cache is kept valid on each run & can be used when going from
disabled
> -> enabled, but I'm not following the bad cache fix part.
> 
> The possible issue I see with that is the extra time in doing this when we
> aren't using the device cache -- specifically, on the bots.

Reworded

[agrieve, 2015-10-09 00:28:32]

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
File build/android/pylib/local/device/local_device_environment.py (right):

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
build/android/pylib/local/device/local_device_environment.py:80: # Write the
cache even when not using it so that a bad cache will be fixed
On 2015/10/08 16:03:28, agrieve wrote:
> On 2015/10/08 14:34:49, jbudorick wrote:
> > I lost you on the reasoning here as written. I can understand writing the
> cache
> > s.t. the cache is kept valid on each run & can be used when going from
> disabled
> > -> enabled, but I'm not following the bad cache fix part.
> > 
> > The possible issue I see with that is the extra time in doing this when we
> > aren't using the device cache -- specifically, on the bots.
> 
> Reworded

Should also note that the extra time is negligible compared to everything else
we're doing (it's much faster than a single adb command)

[jbudorick, 2015-10-09 00:30:31]

lgtm

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
File build/android/pylib/local/device/local_device_environment.py (right):

https://codereview.chromium.org/1394173002/diff/1/build/android/pylib/local/d...
build/android/pylib/local/device/local_device_environment.py:80: # Write the
cache even when not using it so that a bad cache will be fixed
On 2015/10/09 00:28:31, agrieve wrote:
> On 2015/10/08 16:03:28, agrieve wrote:
> > On 2015/10/08 14:34:49, jbudorick wrote:
> > > I lost you on the reasoning here as written. I can understand writing the
> > cache
> > > s.t. the cache is kept valid on each run & can be used when going from
> > disabled
> > > -> enabled, but I'm not following the bad cache fix part.
> > > 
> > > The possible issue I see with that is the extra time in doing this when we
> > > aren't using the device cache -- specifically, on the bots.
> > 
> > Reworded
> 
> Should also note that the extra time is negligible compared to everything else
> we're doing (it's much faster than a single adb command)

Fair enough.

[agrieve, 2015-10-09 14:37:31]

The CQ bit was checked by agrieve@chromium.org

[commit-bot: I haz the power, 2015-10-09 14:37:56]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1394173002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1394173002/20001

[commit-bot: I haz the power, 2015-10-09 15:46:05]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-10-09 15:46:47]

Patchset 2 (id:??) landed as
https://crrev.com/a538a143ab313bb1b89a0c283baac4fafa50c614
Cr-Commit-Position: refs/heads/master@{#353297}