📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1293f1f0240000

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/149d933903e5b2d04393f14e680304fee86acf33

commit 149d933903e5b2d04393f14e680304fee86acf33
Author: Ashley Enstad <ashleymarie@chromium.org>
Date: Mon May 14 19:16:32 2018

Disabling system_health.memory_mobile/background:news:nytimes

CC=perezju@chromium.org
BUG= 842731 
NOTRY=true
TBR=eyaich@chromium.org

Change-Id: Idbc0a87a843c3647e1ad15d13a5429dda6e9d0cf
Reviewed-on: https://chromium-review.googlesource.com/1058068
Reviewed-by: Ashley Enstad <ashleymarie@chromium.org>
Commit-Queue: Ashley Enstad <ashleymarie@chromium.org>
Cr-Commit-Position: refs/heads/master@{#558407}
[modify] https://crrev.com/149d933903e5b2d04393f14e680304fee86acf33/tools/perf/expectations.config

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/1293f1f0240000

cc::Surfaces for Video path handles null context_provider by lethalantidote@chromium.org
https://chromium.googlesource.com/chromium/src/+/df92bfed734a92c8f839af121218917ae599157d

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Range coincides with same story also failing on Go device: issue 843547.

When the issue is fixed please also try to re-enable on both Android One and Go.

Hi, I don't think my change is responsible here. I ran a pinpoint with the test reenabled but did not find the regressions. 

https://pinpoint-dot-chromeperf.appspot.com/job/17908638240000

Hmm... let's try again.
https://pinpoint-dot-chromeperf.appspot.com/job/12b40c1c240000

Actually. No. The revert in your pinpoint job is just to re-enable the story. But this has no effect, because pinpoint forces stories to run even if they are disabled.

The results in #6 just show that the test remains broken on ToT.

On the other hand, the results in #3 look pretty definitive. So this is pretty likely due to some change in your CL.

Anyway, if you really wanted to double-check, what you need is to create a patch with a revert of r557996, and run a job to test whether such revert fixes the test.

I would suggest anyway to start investigating, e.g. looking through the logs or reproducing locally. Note from the logs of failed runs, this *is* causing Chrome to crash:

	UI dump
	- (no package):
	  - (no id)
	- android:
	  - (no id)
	  - android:id/button1['OK']
	  - android:id/buttonPanel
	  - android:id/content
	  - android:id/contentPanel
	  - android:id/message['Unfortunately, Chromium has stopped.']
	  - android:id/parentPanel
	  - android:id/scrollView

	W/google-breakpad( 1435): ### ### ### ### ### ### ### ### ### ### ### ### ###
	W/google-breakpad( 1435): Chrome build fingerprint:
	W/google-breakpad( 1435): Developer Build
	W/google-breakpad( 1435): 1
	W/google-breakpad( 1435): ### ### ### ### ### ### ### ### ### ### ### ### ###
	F/google-breakpad( 1650): -----BEGIN BREAKPAD MICRODUMP-----
	F/google-breakpad( 1650): V Chrome_Android:68.0.3428.0
	F/google-breakpad( 1650): O A arm 04 armv7l google/4560MMX_b/4560MMX_b_sprout:5.1.1/LMY47W/1837160:userdebug/dev-keys

https://chromium-swarm.appspot.com/task?id=3d79fa78fde7fb10&refresh=10&show_raw=1

Unless a fix is found soon, the better course of action would be to revert the offending CL and re-enable the test while you keep investigating.

System health no longer running on this bot.