Revert of android: Move launcher_android.cc to helper

Revert of android: Move launcher_android.cc to helper (patchset #1 id:1 of https://codereview.chromium.org/2753383002/ )

Reason for revert:
Tentative revert to fix the following DCHECK failure:
[FATAL:test_launcher.cc(487)] Check failed: !g_launcher_delegate.
in many tests (e.g. content_browsertests) on chromium.android:Lollipop Tablet Tester and KitKat Tablet Tester

See https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=content_browsertests&builder=chromium.android%3ALollipop%20Tablet%20Tester

It seems the main function of the test processes is getting called more than once, and this and https://codereview.chromium.org/2753713002 are the only likely culprits in crrev.com/458066..458116

Original issue's description:
> android: Move launcher_android.cc to helper
> 
> The ultimate goal is to fix leaking mManagedConnections in
> BindingManagerImpl, which is leaking to support getting isOomProtected
> after a service has been disconnected. That requires saving that status
> in native when service is disconnected, which probably belongs on
> LauncherHelper. That implies Helper should talk directly java, and this
> is the first step.
> 
> This is meant to to be a straight-forward move with no changes, even if
> the existing code makes no sense once they live in the same file. Order
> of function is rearranged a bit though.
> 
> BUG=689758
> 
> Review-Url: https://codereview.chromium.org/2753383002
> Cr-Commit-Position: refs/heads/master@{#458114}
> Committed: https://chromium.googlesource.com/chromium/src/+/4ac59a56d54826dbdfc5b17b2e8875863065b0b8

TBR=jcivelli@chromium.org,boliu@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=689758

[johnme, 2017-03-21 16:17:24]

The CQ bit was checked by johnme@chromium.org

[johnme, 2017-03-21 16:17:24]

Created Revert of android: Move launcher_android.cc to helper

[johnme, 2017-03-21 16:17:29]

The CQ bit was unchecked by johnme@chromium.org

[johnme, 2017-03-23 11:53:47]

Description was changed from

==========
Revert of android: Move launcher_android.cc to helper (patchset #1 id:1 of
https://codereview.chromium.org/2753383002/ )

Reason for revert:
Tentative revert to fix the following DCHECK failure:
[FATAL:test_launcher.cc(487)] Check failed: !g_launcher_delegate.
in many tests (e.g. content_browsertests) on chromium.android:Lollipop Tablet
Tester and KitKat Tablet Tester

See
https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType...

It seems the main function of the test processes is getting called more than
once, and this and https://codereview.chromium.org/2753713002 are the only
likely culprits in crrev.com/458066..458116

Original issue's description:
> android: Move launcher_android.cc to helper
> 
> The ultimate goal is to fix leaking mManagedConnections in
> BindingManagerImpl, which is leaking to support getting isOomProtected
> after a service has been disconnected. That requires saving that status
> in native when service is disconnected, which probably belongs on
> LauncherHelper. That implies Helper should talk directly java, and this
> is the first step.
> 
> This is meant to to be a straight-forward move with no changes, even if
> the existing code makes no sense once they live in the same file. Order
> of function is rearranged a bit though.
> 
> BUG=689758
> 
> Review-Url: https://codereview.chromium.org/2753383002
> Cr-Commit-Position: refs/heads/master@{#458114}
> Committed:
https://chromium.googlesource.com/chromium/src/+/4ac59a56d54826dbdfc5b17b2e88...

TBR=jcivelli@chromium.org,boliu@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=689758
==========

to

==========
Revert of android: Move launcher_android.cc to helper (patchset #1 id:1 of
https://codereview.chromium.org/2753383002/ )

Reason for revert:
Tentative revert to fix the following DCHECK failure:
[FATAL:test_launcher.cc(487)] Check failed: !g_launcher_delegate.
in many tests (e.g. content_browsertests) on chromium.android:Lollipop Tablet
Tester and KitKat Tablet Tester

See
https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType...

It seems the main function of the test processes is getting called more than
once, and this and https://codereview.chromium.org/2753713002 are the only
likely culprits in crrev.com/458066..458116

Original issue's description:
> android: Move launcher_android.cc to helper
> 
> The ultimate goal is to fix leaking mManagedConnections in
> BindingManagerImpl, which is leaking to support getting isOomProtected
> after a service has been disconnected. That requires saving that status
> in native when service is disconnected, which probably belongs on
> LauncherHelper. That implies Helper should talk directly java, and this
> is the first step.
> 
> This is meant to to be a straight-forward move with no changes, even if
> the existing code makes no sense once they live in the same file. Order
> of function is rearranged a bit though.
> 
> BUG=689758
> 
> Review-Url: https://codereview.chromium.org/2753383002
> Cr-Commit-Position: refs/heads/master@{#458114}
> Committed:
https://chromium.googlesource.com/chromium/src/+/4ac59a56d54826dbdfc5b17b2e88...

TBR=jcivelli@chromium.org,boliu@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=689758
==========

[johnme, 2017-03-23 11:57:42]

Abandoning revert. Doesn't seem to be caused by this patch.
https://crbug.com/703708 is attempting to track down the true cause.