@logs,Video: http://go/chrome-androidlogs1/7/735626

There is a java exception here:
/ViewGroup(12617): addInArray been called, this = org.chromium.chrome.browser.snackbar.BottomContainer{41d10188 V.E..... ......ID 0,0-720,1230 #7f0f024d app:id/bottom_container}call stack =
D/ViewGroup(12617): java.lang.Throwable: addInArray
D/ViewGroup(12617): 	at android.view.ViewGroup.addInArray(ViewGroup.java:3786)
D/ViewGroup(12617): 	at android.view.ViewGroup.addViewInner(ViewGroup.java:3740)
D/ViewGroup(12617): 	at android.view.ViewGroup.addView(ViewGroup.java:3564)
D/ViewGroup(12617): 	at android.view.ViewGroup.addView(ViewGroup.java:3540)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.infobar.InfoBarContainer.addToParentView(InfoBarContainer.java:39)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.infobar.InfoBarContainer$2.onViewAttachedToWindow(InfoBarContainer.java:5)
D/ViewGroup(12617): 	at android.view.View.dispatchAttachedToWindow(View.java:12763)
D/ViewGroup(12617): 	at android.view.ViewGroup.dispatchAttachedToWindow(ViewGroup.java:2577)
D/ViewGroup(12617): 	at android.view.ViewGroup.dispatchAttachedToWindow(ViewGroup.java:2584)
D/ViewGroup(12617): 	at android.view.ViewGroup.addViewInner(ViewGroup.java:3757)
D/ViewGroup(12617): 	at android.view.ViewGroup.addView(ViewGroup.java:3564)
D/ViewGroup(12617): 	at android.view.ViewGroup.addView(ViewGroup.java:3540)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.compositor.layouts.phone.stack.Stack.startAnimation$5154ORRICSNM6Q3IDTMMITBD5THMGSJFDLIIUOJIDTRN6PBI5THMURBGDTPMIT3FE8NMOOBPDTQN8SPFE1K6URJ55TPN8OB3DCNL6T31CDLK2RJ9DLGN8QBFDOI4UTJ5E9R6IPBN85N6IRB1EHKMURIKF5O6AEQ995D2ILG_0(Stack.java:87)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.compositor.layouts.phone.StackLayout.onTabCreated(StackLayout.java:166)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.compositor.layouts.LayoutManagerChrome.tabCreated(LayoutManagerChrome.java:148)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.compositor.layouts.LayoutManagerChromePhone.tabCreated(LayoutManagerChromePhone.java:65)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.compositor.layouts.LayoutManagerChrome$LayoutManagerTabModelObserver.didAddTab(LayoutManagerChrome.java:35)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.tabmodel.TabModelImpl.addTab(TabModelImpl.java:117)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.tabmodel.ChromeTabCreator.createNewTab(ChromeTabCreator.java:73)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.tabmodel.ChromeTabCreator.createNewTab(ChromeTabCreator.java:11)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.tabmodel.ChromeTabCreator.launchUrl(ChromeTabCreator.java:92)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.tabmodel.ChromeTabCreator.launchUrl(ChromeTabCreator.java:88)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.tabmodel.TabCreatorManager$TabCreator.launchNTP(TabCreatorManager.java:3)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.ChromeTabbedActivity$3.onClick(ChromeTabbedActivity.java:3)
D/ViewGroup(12617): 	at org.chromium.chrome.browser.toolbar.ToolbarPhone.onClick(ToolbarPhone.java:106)
D/ViewGroup(12617): 	at android.view.View.performClick(View.java:4463)
D/ViewGroup(12617): 	at android.view.View$PerformClick.run(View.java:18789)
D/ViewGroup(12617): 	at android.os.Handler.handleCallback(Handler.java:808)
D/ViewGroup(12617): 	at android.os.Handler.dispatchMessage(Handler.java:103)
D/ViewGroup(12617): 	at android.os.Looper.loop(Looper.java:193)
D/ViewGroup(12617): 	at android.app.ActivityThread.main(ActivityThread.java:5299)
D/ViewGroup(12617): 	at java.lang.reflect.Method.invokeNative(Native Method)
D/ViewGroup(12617): 	at java.lang.reflect.Method.invoke(Method.java:515)
D/ViewGroup(12617): 	at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:825)
D/ViewGroup(12617): 	at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:641)
D/ViewGroup(12617): 	at dalvik.system.NativeStart.main(Native Method)

But to me it looks like the renderer is not starting.

@aelias || @boliu -- To me this looks like the renderer isn't starting based on the video and the stack to me looks like a red herring.  Looping you in before more investigation to make sure this isn't a known issue.

Is this a recent regression?  Does it happen in 58?

take a trace next time this happens

#C5 It doesn't happen on M58 and we will provide bisect info soon.

Bisect info:

Good Build:59.0.3063.0
Bad Build:59.0.3064.0
Regression range: https://chromium.googlesource.com/chromium/src/+log/59.0.3063.0..59.0.3064.0?pretty=fuller&n=10000

Good commit:462157
Bad commit:462158
Suspect CL: https://chromium.googlesource.com/chromium/src/+/52a60f52642d164a8f83cdf2a8c37beb8e26c898

Culprit is boliu@'s "android: Post warmup to launcher thread"

This is quite worrying: a severe stability issue new in M59 which probably does not appear in our crash stats because it is not a crash, so we have no way of estimating the commonness.  I worry this BLU phone repro may be just the tip of the iceberg, so marking this as a last-minute M59 blocker.

logs don't contain anything useful. oddly I don't see any log that shows any chrome child process starting.

give me the device and show me the repro please

also does this still happen in 60/61? there's been a lot more refactoring in this area

#C13 We are not able to repro this issue on M60 and M61 

launcher thread blocked waiting for the warm up signal. it should be signaled on by onServiceConnected/Disconnected which at that time still runs on the UI thread. hmm...

"Chrome_ProcessLauncherThread" prio=5 tid=13 WAIT
  | group="main" sCount=1 dsCount=0 obj=0x41d85268 self=0x60325aa0
  | sysTid=10230 nice=0 sched=0/0 cgrp=apps/bg_non_interactive handle=1615283120
  | state=S schedstat=( 33338618 23911615 93 ) utm=2 stm=1 core=0
  at java.lang.Object.wait(Native Method)
  - waiting on <0x41d84188> (a java.lang.Object)
  at java.lang.Object.wait(Object.java:364)
  at org.chromium.content.browser.ChildProcessLauncher.startInternal(ChildProcessLauncher.java:127)
  at org.chromium.content.browser.ChildProcessLauncher.start(ChildProcessLauncher.java:114)
  at org.chromium.content.browser.ChildProcessLauncherHelper.<init>(ChildProcessLauncherHelper.java:12)
  at org.chromium.content.browser.ChildProcessLauncherHelper.create(ChildProcessLauncherHelper.java:9)
  at org.chromium.base.SystemMessageHandler.nativeDoRunLoopOnce(Native Method)
  at org.chromium.base.SystemMessageHandler.handleMessage(SystemMessageHandler.java:7)
  at android.os.Handler.dispatchMessage(Handler.java:110)
  at android.os.Looper.loop(Looper.java:193)
  at android.os.HandlerThread.run(HandlerThread.java:61)

afaict, this is a legit android bug, bindService returning true, but does not result in onServiceConnected being called

yay, good job android

logs show sandboxed service process got killed by ActivityManager before it started running any chromium code. and presumably this only happens in first run because this is like a low memory device, and the first run activity will cause the service to be killed for memory. so android has no api to notify us that this happened, and onServiceConnected will never happen.

if I add IMPORATNT to the initial binding flags, then this doesn't happen anymore, but that's too big a hammer. I was thinking maybe we should just turn off warmup in for low memory devices, but this device has 1G of ram, so isn't even considered low ram.

I'm guessing my CL triggered this because it moved the warm up a little bit earlier, so maybe I can revert only that bit.

I'm not really convinced this is not happening in M60 though. It's just the result won't be as catastrophic. In m59, this will hang the entire launcher thread. But I think in m60, only the first renderer will never launch, and user can just swipe the tab away and everything will keep working.

oh, and my suspicion is that this only happens on first run because the first run activity somehow causes the activity manager to kill the service..

Partial revert apparently fixes whatever timing issue my CL seem to be tickling: https://codereview.chromium.org/2948293002

Partial revert discussed offline with boliu@ approved for M59 branch 3071.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/aefe4fdabf53cb59ca93dadefc6a63f8af2c1706

commit aefe4fdabf53cb59ca93dadefc6a63f8af2c1706
Author: Bo Liu <boliu@chromium.org>
Date: Thu Jun 22 20:06:34 2017

[Fork M59] android: Warmup after library load

This is partial revert of r462158 in an attempt to workaround an android
bug that consistently kill the warmup service without notifying us.

It looks like if warmup is delayed until after native library is loaded,
then this corner case doesn't happen anymore, so revert that part.

BUG= 735626 
R=tedchoc@chromium.org

Review-Url: https://codereview.chromium.org/2948293002 .
Cr-Commit-Position: refs/branch-heads/3071@{#819}
Cr-Branched-From: a106f0abbf69dad349d4aaf4bcc4f5d376dd2377-refs/heads/master@{#464641}

[modify] https://crrev.com/aefe4fdabf53cb59ca93dadefc6a63f8af2c1706/chrome/android/java/src/org/chromium/chrome/browser/init/AsyncInitTaskRunner.java
[modify] https://crrev.com/aefe4fdabf53cb59ca93dadefc6a63f8af2c1706/content/public/android/java/src/org/chromium/content/browser/ChildProcessLauncher.java

Filed follow up  crbug.com/736066 

also -RVG, nothing private here

Need same fix for M60, right?

This bug requires manual review: M60 has already been promoted to the beta branch, so this requires manual review
Please contact the milestone owner if you have questions.
Owners: amineer@(Android), cmasso@(iOS), josafat@(ChromeOS), bustamante@(Desktop)

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Hmm, I see the comment https://bugs.chromium.org/p/chromium/issues/detail?id=735626#c13 that there is no repro in M60.  But personally I'd be more comfortable just restoring the old timing to M60 as well until the root cause is better understood.  Unless the code has been refactored so much the fix doesn't apply anymore?

I'll do all of that in follow up bug.

Verified on chrome:59.0.3071.117 Device:BLU LIFE PLAY/KOT49H