Looks like testInputTextEvents_SetComposingText is somewhat flaky:https://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=org.chromium.content.browser.input.ImeTest%23testInputTextEvents_SetComposingText&testType=content_shell_test_apk

I saw this failure today on https://build.chromium.org/p/chromium.android/builders/Lollipop%20Tablet%20Tester/builds/4484

C  297.354s Main  [FAIL] org.chromium.content.browser.input.ImeTest#testInputTextEvents_SetComposingText:
C  297.354s Main  junit.framework.ComparisonFailure: expected:<...a),input,keyup(229),[compositionupdate(a),input,compositionend(a),selectionchange,selectionchange],selectionchange,sel...> but was:<...a),input,keyup(229),[selectionchange,selectionchange,compositionupdate(a),input,compositionend(a)],selectionchange,sel...>
C  297.354s Main  	at org.chromium.content.browser.input.ImeTest.waitForEventLogs(ImeTest.java:1264)
C  297.354s Main  	at org.chromium.content.browser.input.ImeTest.testInputTextEvents_SetComposingText(ImeTest.java:1179)
C  297.354s Main  	at android.test.InstrumentationTestCase.runMethod(InstrumentationTestCase.java:214)
C  297.354s Main  	at android.test.InstrumentationTestCase.runTest(InstrumentationTestCase.java:199)
C  297.354s Main  	at android.test.ActivityInstrumentationTestCase2.runTest(ActivityInstrumentationTestCase2.java:192)
C  297.354s Main  	at org.chromium.content_shell_apk.ContentShellTestBase.runTest(ContentShellTestBase.java:233)
C  297.354s Main  	at org.chromium.base.test.BaseTestResult.runParameterized(BaseTestResult.java:161)
C  297.355s Main  	at org.chromium.base.test.BaseTestResult.run(BaseTestResult.java:124)
C  297.355s Main  	at android.test.AndroidTestRunner.runTest(AndroidTestRunner.java:191)
C  297.355s Main  	at android.test.AndroidTestRunner.runTest(AndroidTestRunner.java:176)
C  297.355s Main  	at android.test.InstrumentationTestRunner.onStart(InstrumentationTestRunner.java:555)
C  297.355s Main  	at android.app.Instrumentation$InstrumentationThread.run(Instrumentation.java:1853)

Changwan, guessing you can help triage or route this appropriately. Do you have any idea what might be causing these flakes?  We saw an increase in flakyness in Contextual Search instrumentation tests about a week ago when the startup timing was shortened ( issue 635661 ).  That might be affecting this test too (wild guess).