Run logcat monitor and device status check on bisect bots.

build/android/buildbot/bb_run_bot.py

Index: build/android/buildbot/bb_run_bot.py
diff --git a/build/android/buildbot/bb_run_bot.py b/build/android/buildbot/bb_run_bot.py
index c729ca2d5f87a0ab5243b24b3ea8b6533ae200a0..71b782adbe34643ecdca9c9bb63eafff227cb429 100755
--- a/build/android/buildbot/bb_run_bot.py
+++ b/build/android/buildbot/bb_run_bot.py
@@ -161,7 +161,8 @@ def GetBotStepMap():
         H(compile_step, extra_gyp='component=shared_library'),
         T(std_tests, ['--experimental', flakiness_server])),
       B('gpu-builder-tests-dbg', H(compile_step), T(['gpu'])),
-      B('perf-bisect-builder-tests-dbg', H(['bisect_perf_regression'])),
+      # Pass empty T([]) so that logcat monitor and device status check are run.
+      B('perf-bisect-builder-tests-dbg', H(['bisect_perf_regression']), T([])),

[navabi, 2013/12/06 19:13:58]

The empty T([]) will make bb_device_steps::MainTestWrapper run, where the device
status check is executed. This will also spawn logcat monitor. Since no other
options are passed, those are the only things that will be added by adding a T.

The bug was to add device status check. Ok if it comes with logcat monitor,
which will print the logcat of the device. Generally, this is good to have
anyway to investigate device issues.

       B('perf-tests-rel', H(std_test_steps),
         T([], ['--install=ChromiumTestShell'])),
       B('webkit-latest-webkit-tests', H(std_test_steps),