1046956 - FX OS crash in BackgroundProcessLRUPool::UpdateAvailableIndexInLRUPool

Status: RESOLVED WORKSFORME

(Core :: DOM: Core & HTML, defect)

Description

[Marcia Knous [:marcia]]

This bug was filed from the Socorro interface and is 
report bp-4ce0108e-f1ef-4468-b1d1-71dcf2140731.
=============================================================

During testing today I hit this crash several times when the device was idle.

Flame while running:

Gaia   6548f8a89654727eb3a65200ef14298e106eb099
SourceStamp aa176fcc56b8
BuildID 20140731040203
Version 34.0a1
Base: v122

Comment 1

[Marcia Knous [:marcia]]

This seems to be happening after I launch the Twitter app.

Comment 2

[Marcia Knous [:marcia]]

Adding Alan since it appears he touched this code recently.

Comment 3

[Florin Strugariu [:Bebe]]

I can reproduce this fail using gaia UI automation.

This issue is breaking our build almost every time.

STR:

1. run https://github.com/mozilla-b2g/gaia/blob/master/tests/python/gaia-ui-tests/gaiatest/tests/functional/messages/test_add_to_an_existing_contact_from_messages.py
2. after the test ends wait for a few min.

Comment 4

[Florin Strugariu [:Bebe]]

Attachment: logcat of the issue

Comment 5

[Florin Strugariu [:Bebe]]

This is reproducing on today master build:

Gaia      04ea7e1a4034a50d4a7a4f5b95a04a2ed8313908
Gecko     https://hg.mozilla.org/mozilla-central/rev/104254bd1fc8
BuildID   20140801040326
Version   34.0a1
ro.build.version.incremental=109
ro.build.date=Mon Jun 16 16:51:29 CST 2014
B1TC00011220


I can't manually reproduce this issue


The manual steps of the test:

1. add a contact
2. send a SMS to a number (different than the contact number)
3. Tap on the phone number
4. tap add to the existing contacts and add to contact
5. close the messages app
6. Open the contacts app
7. verify the modified contact

Comment 6

[Jason Smith [:jsmith]]

[Blocking Requested - why for this release]:

This is causing our entire automated test suite for non-smoketests to be down right now due to this crash.

Comment 7

[Jason Smith [:jsmith]]

Might be caused by bug 874353.

Kyle - What do you think?

Comment 8

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

It's plausible, I guess.  Back it out and see what happens?

Do you have Level 3 access?  If you don't we should get that set up for you.

Comment 9

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

Oh, yes, it's almost certainly that.

Comment 10

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Agreed, khuey, we should get jsmith level 3 access.

Comment 11

[KTucker [:KTucker][Inactive 3/4/2016]]

Adding qaurgent because we encountered this issue twice during 2.1 smoketest. We need solid reproducible steps on this.

Comment 12

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

khuey backed out of this change: http://hg.mozilla.org/mozilla-central/rev/ff1e09666c42

Comment 13

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

We should verify that the backout actually makes this go away before marking it fixed.

Comment 14

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Fair enough.  I'm making my own build with the backout currently.

Comment 15

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Actually tinderbox builds may be faster.

Comment 16

[Cody Roesch [:croesch]]

We can't seem to get reliable steps here. It's very random and unpredictable.

We've seen the crash:
-During a Reboot
-Opening and closing rocketbar
-Opening marketplace and letting it sit in the background
-Opening settings menu in contacts then tapping done

We are seeing this using 319mb Flame devices. We had a tester on 512mb Flame using today's build but he mentioned he had not seen the crash all day during testing.

Sounds like there may be a fix for the issue so we'll see how that turns out.

Tested with
Flame 2.1 using today's 8/1 build. Mem set to 319mb.

Environmental Variables:
Device: Flame Master
Build ID: 20140801085315
Gaia: 9689218473b6fc4dd927ad6aa7b06c05f0843824
Gecko: 44e5072476b7
Version: 34.0a1 (Master)
Firmware Version: v122

Comment 17

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Seems to work on inbound build w/o crashing : 20140801133010
Will wait for a moz-central build to set to fixed.

Comment 18

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Turns out it's still crashing I can't tell which crash it is since inbound builds doesn't push symbols to Socorro.

Comment 19

[Zac C (:zac)]

We are still seeing some heavy crashes in the automation too but we haven't replciated it locally yet.

Comment 20

[Martin Shuman [:Marty]]

We did not encounter this crash on today's 2.1 Flame 319MB Smoketest.

Environmental Variables:
Device: Flame Master (319MB)
Build ID: 20140804040204
Gaia: 5fd14b8bc428f87f9b5cf9cc49f9a4f362a970fb
Gecko: e6614d8d85f9
Version: 34.0a1 (Master)
Firmware Version: v123
User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0

Comment 21

[KTucker [:KTucker][Inactive 3/4/2016]]

Closing this issue since we are not encountering it anymore on our manual smoketests.