Closed Bug 980682 Opened 6 years ago Closed 6 years ago

[B2G][Calendar] Notifications for calendar event alarm reminders are not being received

Categories

(Core :: DOM: Device Interfaces, defect)

ARM
Gonk (Firefox OS)
defect
Not set

Tracking

()

RESOLVED FIXED
1.4 S4 (28mar)
blocking-b2g 1.4+
Tracking Status
firefox29 --- wontfix
firefox30 --- fixed
firefox31 --- fixed
b2g-v1.3 --- unaffected
b2g-v1.4 --- fixed
b2g-v2.0 --- fixed

People

(Reporter: demerick, Assigned: ochameau)

References

Details

(Keywords: regression, Whiteboard: [fxos-bug-bash-1.4], burirun1.4-1, burirun1.4-2)

Attachments

(2 files)

Attached file logcat.txt
Description:
The user does not receive a notification for any event alarm reminders (at time of event, 5 minutes before, 1 hour before, 1 day before, etc.)

Repro Steps:
1) Update a Buri to BuildID: 20140306134106
2) Open Calendar app from the home screen
3) Tap '+' in the upper right hand corner
4) Set the start time to 1 hour and 5 minutes from the current time
5) Make sure the end time is after the start time
6) Set a reminder alarm for '1 hour before'
7) Exit the calendar app and wait 5 minutes

Actual:
No alarm notification is received

Expected:
The user should receive an alarm reminder notification for the calendar event

1.4 Environmental Variables:
Device: Buri 1.4 MOZ
BuildID: 20140306134106
Gaia: 98cf46d6623b164845fe1fdc99a2a7bf64aa667d
Gecko: 8095b7dd8f58
Version: 30.0a1
Firmware Version: v1.2-device.cfg

Repro frequency: 100%
See attached: logcat

notes: This occurs for all calendar types (google, yahoo, caldav)

keywords: off, go, firing, pop, trigger,
This issue does not occur on the Buri v 1.3.0 Mozilla RIL

1.3 Environmental Variables:
Device: Buri 1.3 MOZ
BuildID: 20140306004002
Gaia: 8aed4fafbaeb86d6884d31ce7d3cbeb87bcbf837
Gecko: 3d2d84d52141
Version: 28.0
Firmware Version: v1.2-device.cfg

Notifications for calendar event alarm reminders were received by the user using a Buri v 1.3.0 Moz RIL device.
blocking-b2g: --- → 1.4?
Whiteboard: [fxos-bug-bash-1.4] → [fxos-bug-bash-1.4], burirun1.4-1
blocking-b2g: 1.4? → 1.4+
Regression Window using b2g inbound builds: 

Last Working Environmental Variables:
Device: Buri
BuildID: 20140207004704
Gaia: b545796e7dda648aaaf9f14415ec60dadc3ce574
Gecko: 83ca9463d106
Version: 30.0a1
Base Image: V1.2-device.cfg

First Broken Environmental Variables:
Device: Buri
BuildID: 20140207010805
Gaia: b545796e7dda648aaaf9f14415ec60dadc3ce574
Gecko: e4d4748c86f5
Version: 30.0a1
Base Image: V1.2-device.cfg

Last Working Gaia/First Broken Gecko: Issue DOES reproduce.
Gaia: b545796e7dda648aaaf9f14415ec60dadc3ce574
Gecko: e4d4748c86f5

First Broken Gaia/Last Working Gecko: Issue does NOT reproduce.
Gaia: b545796e7dda648aaaf9f14415ec60dadc3ce574
Gecko: 83ca9463d106

Gecko Pushlog: http://hg.mozilla.org/integration/b2g-inbound/pushloghtml?fromchange=83ca9463d106&tochange=e4d4748c86f5
QA Contact: jzimbrick
So this is a gecko bug, but I don't know which gecko bug is the cause. Someone will need to do an initial investigation here to point out what's failing here in gecko to figure out where this bug should be moved.
Push log looks incorrect here - there shouldn't be that many changesets present in a b2g-inbound push log. Reflagging the window.
I re-downloaded the builds from the pvt b2g-inbound folder, retested the issue and got the same results, pulled environmental variables, generated a new pushlog, and got all the same results as Comment 2.

I also thought it was odd when pulling that pushlog that the number of changes was so high using inbound builds, but these are in fact the two builds in the inbound folder that are closest together, and as far as I know this is the correct template for generating an inbound-b2g pushlog.

Inbound Regression Window: 

Last Working Environmental Variables:
Device: Buri
BuildID: 20140207004704
Gaia: b545796e7dda648aaaf9f14415ec60dadc3ce574
Gecko: 83ca9463d106
Version: 30.0a1
Base Image: V1.2-device.cfg

First Broken Environmental Variables:
Device: Buri
BuildID: 20140207010805
Gaia: b545796e7dda648aaaf9f14415ec60dadc3ce574
Gecko: e4d4748c86f5
Version: 30.0a1
Base Image: V1.2-device.cfg

Gecko Pushlog: https://hg.mozilla.org/integration/b2g-inbound/pushloghtml?fromchange=83ca9463d106&tochange=e4d4748c86f5
This is still an incorrect window. Please provide the m-c push log & I'll point out the problem here.
Tinderbox Regression Window:

Last Working Environmental Variables:
Device: Buri
BuildID: 20140206132902
Gaia: ea4a1f0d94a995486ed219f47132949071ecc172
Gecko: 262e73a6b7cd
Version: 30.0a1
Base Image: V1.2-device.cfg

First Broken Environmental Variables:
Device: Buri
BuildID: 20140207005201
Gaia: ea4a1f0d94a995486ed219f47132949071ecc172
Gecko: 1ca0ce406aad
Version: 30.0a1
Base Image: V1.2-device.cfg

Gecko Pushlog: http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=262e73a6b7cd&tochange=1ca0ce406aad
Okay - comment 7 has the right window here.
Assignee: nobody → evanxd
Target Milestone: --- → 1.4 S4 (28mar)
Component: Gaia::Calendar → DOM: Device Interfaces
Product: Firefox OS → Core
Hi all,
It is a gecko issue, change the component.
Hi Alexandre, it's highly possible bug 963258 causes this regression, which is the only change on Alarm API in the recent half year. Could you please take a look? I'd suggest let's back out bug 963258 to go back to a clear stage.
Flags: needinfo?(poirot.alex)
After investigating, some weird things happened.
The mozAlarms did NOT work between 17:00 to 18:00 yesterday(3/18),
but it could work between 10:30 to 12:00 today(3/19).

And we could see the http://hg.mozilla.org/mozilla-central/rev/72da2b2f029c patch about mozAlarms in gecko pushlog at Comment 7.
It might be the root cause.

Hope these clues could help.
Thanks.
Thanks Evan for the investigation!
Sorry about that regression, I'm quite scared to have broke such simple usecase
without having hurt any test.

This time I tested clock AND calendar. Is there any other major usage of mozAlarms?
Attachment #8393460 - Flags: review?(gene.lian)
The e-mail app uses mozAlarms to wake itself up to perform periodic sync.  But its only entrypoint is index.html.
Comment on attachment 8393460 [details] [diff] [review]
Ensure mozAlarms doesn't throw on chrome documents, and fix alarms set on non-index.html pages.

Review of attachment 8393460 [details] [diff] [review]:
-----------------------------------------------------------------

Thanks Alexandre!
Attachment #8393460 - Flags: review?(gene.lian) → review+
Assignee: evanxd → poirot.alex
Keywords: checkin-needed
Permafail for burirun1.4-2.
Whiteboard: [fxos-bug-bash-1.4], burirun1.4-1 → permafail
Whiteboard: permafail → [fxos-bug-bash-1.4], burirun1.4-1, burirun1.4-2
https://hg.mozilla.org/mozilla-central/rev/399ae6ca9286
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.