[Android] Remove adb_commands from telemetry.

[Android] Remove adb_commands from telemetry.

...because it's an abomination.

BUG=476709
Committed: https://crrev.com/c951c9e158d55d2be18f8226200679aee55f336d
Cr-Commit-Position: refs/heads/master@{#332538}

[jbudorick, 2015-05-15 01:43:56]

jbudorick@chromium.org changed reviewers:
+ skyostil@chromium.org, sullivan@chromium.org

[jbudorick, 2015-05-15 01:43:57]

Note that this doesn't attempt to update usages of DeviceUtils (e.g. by adding
check_return=True to RunShellCommand calls) because this patch has enough moving
pieces as it is.

https://codereview.chromium.org/1141833003/diff/40001/tools/telemetry/telemet...
File tools/telemetry/telemetry/core/platform/android_platform_backend.py
(right):

https://codereview.chromium.org/1141833003/diff/40001/tools/telemetry/telemet...
tools/telemetry/telemetry/core/platform/android_platform_backend.py:63: def
_SetupPrebuiltTools(device):
note that this function and the one below were moved from adb_commands.py and
converted to device_utils

https://codereview.chromium.org/1141833003/diff/40001/tools/telemetry/telemet...
tools/telemetry/telemetry/core/platform/android_platform_backend.py:64: """Some
of the android pylib scripts we depend on are lame and expect
:(

[sullivan, 2015-05-15 02:09:01]

lgtm

Thanks for all the clean up! We will want to be a little careful submitting--too
soon and it'll be hard to distinguish errors from this vs the cl that removed
old_device. Too late and the bots could go red over the weekend. Sami, you're
perf sheriff, what do you think? Maybe you could cq it in the morning after
you've had a chance to look at whether the last cl broke any bots?

[jbudorick, 2015-05-15 02:10:49]

On 2015/05/15 02:09:01, sullivan wrote:
> lgtm
> 
> Thanks for all the clean up! We will want to be a little careful
submitting--too
> soon and it'll be hard to distinguish errors from this vs the cl that removed
> old_device. Too late and the bots could go red over the weekend. Sami, you're
> perf sheriff, what do you think? Maybe you could cq it in the morning after
> you've had a chance to look at whether the last cl broke any bots?

I have no problem with holding this one over the weekend so the last one has a
chance to rattle through the perf bots a bit.

[jbudorick, 2015-05-19 15:33:32]

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

[jbudorick, 2015-05-19 15:33:33]

+primiano for perf sheriff approval

This could cause some turbulence on the Android bots -- any objections to trying
to land this today?

[Primiano Tucci (use gerrit), 2015-05-19 16:39:30]

LGTM for the cleanup.
As regards sheriffing, the situation is the following:
everything is badly broken on Nexus 4,6 and 9 due to the flakiness.
On top of that, it's almost end of day here in London, so won't be able to take
too much look today.
Maybe can we wait landing this until the situation becomes a bit more quiet and
green (will it ever be the case?)

[jbudorick, 2015-05-19 16:41:34]

On 2015/05/19 16:39:30, Primiano Tucci wrote:
> LGTM for the cleanup.
> As regards sheriffing, the situation is the following:
> everything is badly broken on Nexus 4,6 and 9 due to the flakiness.
> On top of that, it's almost end of day here in London, so won't be able to
take
> too much look today.
> Maybe can we wait landing this until the situation becomes a bit more quiet
and
> green (will it ever be the case?)

fwiw, the N9s are experiencing a non-flaky device issue that I can explain in
more detail offline.

Whenever I land this, I intend to watch the perf waterfall as well.

[jbudorick, 2015-05-19 16:42:15]

On 2015/05/19 16:41:34, jbudorick wrote:
> On 2015/05/19 16:39:30, Primiano Tucci wrote:
> > LGTM for the cleanup.
> > As regards sheriffing, the situation is the following:
> > everything is badly broken on Nexus 4,6 and 9 due to the flakiness.
> > On top of that, it's almost end of day here in London, so won't be able to
> take
> > too much look today.
> > Maybe can we wait landing this until the situation becomes a bit more quiet
> and
> > green (will it ever be the case?)
> 
> fwiw, the N9s are experiencing a non-flaky device issue that I can explain in
> more detail offline.

er, N6s

> 
> Whenever I land this, I intend to watch the perf waterfall as well.

[Primiano Tucci (use gerrit), 2015-05-19 16:44:27]

SGTM if you keep an eye on the perf waterfall.
Many thanks!

[jbudorick, 2015-05-20 04:20:10]

On 2015/05/19 16:44:27, Primiano Tucci wrote:
> SGTM if you keep an eye on the perf waterfall.
> Many thanks!

obviously didn't get a chance to land this today. will fix the windows trybot
failure + attempt to land tomorrow w/ same perf bot watching promise

[nednguyen, 2015-05-22 21:43:30]

nednguyen@google.com changed reviewers:
+ dtu@chromium.org, nednguyen@google.com

[jbudorick, 2015-06-02 21:10:07]

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

[jbudorick, 2015-06-02 21:10:08]

+picksi as current perf sheriff

Telemetry folks, is it ok if I try to land this?

[sullivan, 2015-06-02 21:18:05]

sullivan@chromium.org changed reviewers:
+ zhenw@chromium.org

[sullivan, 2015-06-02 21:18:06]

lgtm

+Zhen, tomorrow's perf sheriff. This is okay to land, but Zhen, if you see
massive breakages please revert.

[jbudorick, 2015-06-02 21:23:03]

On 2015/06/02 at 21:18:06, sullivan wrote:
> lgtm
> 
> +Zhen, tomorrow's perf sheriff. This is okay to land, but Zhen, if you see
massive breakages please revert.

I'll be watching the perf waterfalls, too.

[jbudorick, 2015-06-02 21:31:04]

The CQ bit was checked by jbudorick@chromium.org

[jbudorick, 2015-06-02 21:31:05]

The patchset sent to the CQ was uploaded after l-g-t-m from
primiano@chromium.org, sullivan@chromium.org
Link to the patchset: https://codereview.chromium.org/1141833003/#ps240001
(title: "(last?) rebase")

[commit-bot: I haz the power, 2015-06-02 21:32:10]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1141833003/240001

[commit-bot: I haz the power, 2015-06-03 00:46:50]

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

[commit-bot: I haz the power, 2015-06-03 00:46:52]

Try jobs failed on following builders:
  mac_chromium_rel_ng on tryserver.chromium.mac (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_rel_...)

[jbudorick, 2015-06-03 00:47:08]

The CQ bit was checked by jbudorick@chromium.org

[commit-bot: I haz the power, 2015-06-03 00:50:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1141833003/240001

[commit-bot: I haz the power, 2015-06-03 03:22:42]

Committed patchset #13 (id:240001)

[commit-bot: I haz the power, 2015-06-03 03:23:53]

Patchset 13 (id:??) landed as
https://crrev.com/c951c9e158d55d2be18f8226200679aee55f336d
Cr-Commit-Position: refs/heads/master@{#332538}

[alancutter (OOO until 2018), 2015-06-03 05:38:44]

A revert of this CL (patchset #13 id:240001) has been created in
https://codereview.chromium.org/1145153005/ by alancutter@chromium.org.

The reason for reverting is: On 2015/06/03 at 03:23:53, commit-bot wrote:
> Patchset 13 (id:??) landed as
https://crrev.com/c951c9e158d55d2be18f8226200679aee55f336d
> Cr-Commit-Position: refs/heads/master@{#332538}

This change breaks --browser=list when non-rooted Android devices are attached.


$ tools/perf/run_benchmark --browser=list                                       
                                
ERROR:root:Fail to create platform instance for Android device 04f120591b9223e1.

Traceback (most recent call last):
  <module> at tools/perf/run_benchmark:20
    sys.exit(benchmark_runner.main(config))
  main at tools/telemetry/telemetry/benchmark_runner.py:433
    options, args = parser.parse_args()
  ParseArgs at tools/telemetry/telemetry/core/browser_options.py:179
    device)
  Cacher at tools/telemetry/telemetry/decorators.py:30
    cacher.__cache[key] = obj(*args, **kwargs)
  GetAllAvailableBrowsers at
tools/telemetry/telemetry/core/browser_finder.py:147
    browser_finder.FindAllAvailableBrowsers(options, device))
  FindAllAvailableBrowsers at
tools/telemetry/telemetry/core/backends/chrome/android_browser_finder.py:217
    android_platform = platform.GetPlatformForDevice(device, finder_options)
  GetPlatformForDevice at tools/telemetry/telemetry/core/platform/__init__.py:58
    finder_options))
  CreatePlatformForDevice at
tools/telemetry/telemetry/core/platform/android_platform_backend.py:186
    platform_backend = AndroidPlatformBackend(device, finder_options)
  __init__ at
tools/telemetry/telemetry/core/platform/android_platform_backend.py:148
    self._device.EnableRoot()
  TimeoutRetryWrapper at build/android/pylib/device/decorators.py:57
    return timeout_retry.Run(impl, timeout, retries)
  Run at build/android/pylib/utils/timeout_retry.py:161
    thread_group.JoinAll(child_thread.GetWatcher())
  JoinAll at build/android/pylib/utils/reraiser_thread.py:143
    self._JoinAll(watcher)
  _JoinAll at build/android/pylib/utils/reraiser_thread.py:130
    thread.ReraiseIfException()
  run at build/android/pylib/utils/reraiser_thread.py:76
    self._ret = self._func(*self._args, **self._kwargs)
  RunOnTimeoutThread at build/android/pylib/utils/timeout_retry.py:150
    ret[0] = func(*args, **kwargs)
  impl at build/android/pylib/device/decorators.py:51
    return f(*args, **kwargs)
  EnableRoot at build/android/pylib/device/device_utils.py:289
    'Cannot enable root in user builds.', str(self))
CommandFailedError: (device: 04f120591b9223e1) Cannot enable root in user
builds.

Locals:
  retries : 3
  timeout : 30
.

[jbudorick, 2015-06-03 06:14:02]

On 2015/06/03 05:38:44, alancutter wrote:
> A revert of this CL (patchset #13 id:240001) has been created in
> https://codereview.chromium.org/1145153005/ by mailto:alancutter@chromium.org.
> 
> The reason for reverting is: On 2015/06/03 at 03:23:53, commit-bot wrote:
> > Patchset 13 (id:??) landed as
> https://crrev.com/c951c9e158d55d2be18f8226200679aee55f336d
> > Cr-Commit-Position: refs/heads/master@{#332538}
> 
> This change breaks --browser=list when non-rooted Android devices are
attached.
> 
> 
> $ tools/perf/run_benchmark --browser=list                                     
 
>                                 
> ERROR:root:Fail to create platform instance for Android device
04f120591b9223e1.
> 
> Traceback (most recent call last):
>   <module> at tools/perf/run_benchmark:20
>     sys.exit(benchmark_runner.main(config))
>   main at tools/telemetry/telemetry/benchmark_runner.py:433
>     options, args = parser.parse_args()
>   ParseArgs at tools/telemetry/telemetry/core/browser_options.py:179
>     device)
>   Cacher at tools/telemetry/telemetry/decorators.py:30
>     cacher.__cache[key] = obj(*args, **kwargs)
>   GetAllAvailableBrowsers at
> tools/telemetry/telemetry/core/browser_finder.py:147
>     browser_finder.FindAllAvailableBrowsers(options, device))
>   FindAllAvailableBrowsers at
> tools/telemetry/telemetry/core/backends/chrome/android_browser_finder.py:217
>     android_platform = platform.GetPlatformForDevice(device, finder_options)
>   GetPlatformForDevice at
tools/telemetry/telemetry/core/platform/__init__.py:58
>     finder_options))
>   CreatePlatformForDevice at
> tools/telemetry/telemetry/core/platform/android_platform_backend.py:186
>     platform_backend = AndroidPlatformBackend(device, finder_options)
>   __init__ at
> tools/telemetry/telemetry/core/platform/android_platform_backend.py:148
>     self._device.EnableRoot()
>   TimeoutRetryWrapper at build/android/pylib/device/decorators.py:57
>     return timeout_retry.Run(impl, timeout, retries)
>   Run at build/android/pylib/utils/timeout_retry.py:161
>     thread_group.JoinAll(child_thread.GetWatcher())
>   JoinAll at build/android/pylib/utils/reraiser_thread.py:143
>     self._JoinAll(watcher)
>   _JoinAll at build/android/pylib/utils/reraiser_thread.py:130
>     thread.ReraiseIfException()
>   run at build/android/pylib/utils/reraiser_thread.py:76
>     self._ret = self._func(*self._args, **self._kwargs)
>   RunOnTimeoutThread at build/android/pylib/utils/timeout_retry.py:150
>     ret[0] = func(*args, **kwargs)
>   impl at build/android/pylib/device/decorators.py:51
>     return f(*args, **kwargs)
>   EnableRoot at build/android/pylib/device/device_utils.py:289
>     'Cannot enable root in user builds.', str(self))
> CommandFailedError: (device: 04f120591b9223e1) Cannot enable root in user
> builds.
> 
> Locals:
>   retries : 3
>   timeout : 30
> .

Did you observe this on a bot, or just locally?

[alancutter (OOO until 2018), 2015-06-03 07:18:46]

On 2015/06/03 at 06:14:02, jbudorick wrote:
> 
> Did you observe this on a bot, or just locally?

Observed locally on Linux Ubuntu with either Nexus 4 or Nexus 10 connected.