Closed Bug 997292 Opened 12 years ago Closed 12 years ago

[B2G][FTU/FTE] Re-entering Import from Gmail several times returns user to Homescreen

Categories

(Firefox OS Graveyard :: Gaia::First Time Experience, defect)

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(b2g-v1.3 unaffected, b2g-v1.3T affected, b2g-v1.4 unaffected)

RESOLVED WORKSFORME
Tracking Status
b2g-v1.3 --- unaffected
b2g-v1.3T --- affected
b2g-v1.4 --- unaffected

People

(Reporter: rkunkel, Unassigned)

Details

(Keywords: memory-footprint, perf, Whiteboard: 1.3tarakorun2, OOM, [MemShrink])

Attachments

(2 files)

Attached file logcat
Description: When the user enters / re-enters importing contacts from Gmail during the FTE, the user is sometimes sent to the device's Homescreen before finishing the remainder of the FTE. Repro Steps: 1) Update a Tarako device to v1.3T BuildID: 20140415004002 2) Launch the FTU through Developer Settings 3) Follow the FTE process, connecting to a WiFi network 4) On Import Contacts screen, tap on Gmail 5) Sign in with a valid username and password 6) Tap on Cancel ('x') 7) Repeat steps 4-6 until issue occurs Actual: User is returned to the Homescreen before finishing FTE Expected: User can enter / re-enter Import from Gmail during FTE v1.3T Environmental Variables: Device: Tarako v1.3T MOZ RIL BuildID: 20140415004002 Gaia: 44ff6248c28ff83b9ad1161847a176399f93d3bb Gecko: 28aea220e338 Version: 28.1 Firmware Version: sp6821 Notes: Currently the only way we have to test the FTU / FTE on the Tarako device is through Developer Settings Repro frequency: 100% See attached: logcat, firewatch
Attached file firewatch
Looks like OOM? The FTU seems to fall under communications.
Keywords: footprint, perf
Whiteboard: 1.3tarakorun2 → 1.3tarakorun2, OOM, [c=memory p= s= u=] [MemShrink]
Does this reproduce if you trigger this from initial flash of the device, rather than launching FTE from developer settings?
Keywords: qawanted
(In reply to Jason Smith [:jsmith] from comment #3) > Does this reproduce if you trigger this from initial flash of the device, > rather than launching FTE from developer settings? We are currently unable to test the FTE on the Tarako device unless we go through the developer settings. If there is a patch or workaround I can re-test under that environment and report with the results. However, When I tried to reproduce the issue on the Buri device for v1.3 and v1.4, I tested by going through an initial flash and was NOT able to reproduce the behaviour seen on the Tarako.
Keywords: qawanted
(In reply to Roland Kunkel [:RolandK] from comment #4) > (In reply to Jason Smith [:jsmith] from comment #3) > > Does this reproduce if you trigger this from initial flash of the device, > > rather than launching FTE from developer settings? > > We are currently unable to test the FTE on the Tarako device unless we go > through the developer settings. If there is a patch or workaround I can > re-test under that environment and report with the results. We really need that information. The user impact of this bug could be significantly different if we reproduce this during the FTE after an initial flash. Naoki - Can you try to figure out a way we can test this using the FTE post initial flash?
Flags: needinfo?(nhirata.bugzilla)
There's two ways to bring up the FTE post flash. 1) Reset the phone or 2) launch the FTE from the developer menu.
Flags: needinfo?(nhirata.bugzilla)
(In reply to Naoki Hirata :nhirata (please use needinfo instead of cc) from comment #6) > There's two ways to bring up the FTE post flash. > 1) Reset the phone or 2) launch the FTE from the developer menu. Can we try testing this FTE bug post resetting the phone?
Keywords: qawanted
MemShrink:P3 because re-entering import data is a rare case.
Whiteboard: 1.3tarakorun2, OOM, [c=memory p= s= u=] [MemShrink] → 1.3tarakorun2, OOM, [c=memory p= s= u=] [MemShrink:P3]
(In reply to Jason Smith [:jsmith] from comment #7) > (In reply to Naoki Hirata :nhirata (please use needinfo instead of cc) from > comment #6) > > There's two ways to bring up the FTE post flash. > > 1) Reset the phone or 2) launch the FTE from the developer menu. > > Can we try testing this FTE bug post resetting the phone? We are currently blocked from satisfying this request due to bug 991304 - [Tarako] [B2G][First Time Experience] No FTE/FTU appears when device is reset or reflashed. We will test this once that bug has been resolved. Leaving qawanted to keep track of this issue.
(In reply to Jason Smith [:jsmith] from comment #7) > (In reply to Naoki Hirata :nhirata (please use needinfo instead of cc) from > comment #6) > > There's two ways to bring up the FTE post flash. > > 1) Reset the phone or 2) launch the FTE from the developer menu. > > Can we try testing this FTE bug post resetting the phone? This issue *does* reproduce after resetting the phone using the 04/25/14 1.3T build. It will also reproduce after flashing the phone, and still reproduces after launching the FTE from within the Developer menu. Device: Tarako v1.3T MOZ BuildID: 20140425014003 Gaia: 293b056683b506ec1b4787d0fbf6033cc30fa953 Gecko: d91121fbca90 Version: 28.1 Firmware Version: sp6821a-4-18
Keywords: qawanted
What's the least amount of times you have to do steps 4 - 6 to reproduce the bug?
Keywords: qawanted
(In reply to Jason Smith [:jsmith] from comment #11) > What's the least amount of times you have to do steps 4 - 6 to reproduce the > bug? It varies. One time I was able to reproduce this on the 1st try after resetting the phone. Other times it could take up to 11 times of repeating steps 4 to 6 to reproduce the issue. 1 attempt = Resetting the phone via Settings > Device Info > More Info > Reset Phone. Attempt 1 - 1st try (without repeating steps 4 to 6) Attempt 2 - 8th try Attempt 3 - 2nd try Attempt 4 - 2nd try Attempt 5 - 3rd try Attempt 6 - 11th try Tested on: Device: Tarako 1.3T MOZ BuildID: 20140429014002 Gaia: b5adc5a943d3abbd6ab070a47c847f2c24891cc5 Gecko: e9890f5d4709 Version: 28.1 Firmware Version: sp6821a_gonk4.0_user.pac
Keywords: qawanted
The main concern I have here for the blocking nomination is that this can actually cause the user to leave the FTE app earlier than expected outside of their control, which seems like something that should never happen. We also have confirmation that this can happen on the 1st try, which means a user could have a bad first time usage of the phone due to hitting the OOM seen in this bug. Historically, we've blocked heavily on bugs involving ensuring that the FTE app was solid, as we felt that it was critical to ensure that the user has a good first time usage of the phone.
blocking-b2g: --- → 1.3T?
I'm also sending this back into MemShrink & perf triage as recent testing has shown that this can happen on the 1st try of just trying to import contacts, which is a realistic case for a user to hit.
Priority: P3 → --
Whiteboard: 1.3tarakorun2, OOM, [c=memory p= s= u=] [MemShrink:P3] → 1.3tarakorun2, OOM, [MemShrink]
I think we need more info here and sorry if I missed it in one of the previous comments. How many contacts do you import each time? Are pictures included? Seems like the child process (communication) is leaking something or our peak memory consumption is too high and we have to chunk the import. Are you sure that the previous import finished before you start the next import? Towards the end I see NAME PID PPID CPU(s) NICE USS PSS RSS VSIZE OOM_ADJ USER b2g 83 1 114.6 0 7.9 11.1 16.4 144.5 0 root (Nuwa) 328 83 2.2 0 0.0 0.9 3.2 49.4 0 root Communications 1083 328 75.9 1 38.3 41.4 46.5 106.3 2 app_1083 (Preallocated a 1109 328 1.4 0 0.0 1.3 4.6 56.4 1 root Can we get an about:memory profile when you see the communication app somewhere above 42MB RSS?
Pi, can you and Matthew answer the question in comment 14 please?
Flags: needinfo?(pcheng)
Keywords: qawanted
I can no longer reproduce this issue with today's build. Repro rate is 0 out of 80 attempts, with every 20 attempts I reset the phone via Settings. Tested on: Device: Tarako 1.3T MOZ BuildID: 20140506014003 Gaia: aef737f4eae863949d4b42cd6c17339aec3a5fa0 Gecko: cd0c51278ae5 Version: 28.1 Firmware Version: sp6821a_gonk4.0_4-29 (In reply to Gregor Wagner [:gwagner] from comment #15) > How many contacts do you import each time? Are pictures included? I have only 2 contacts on my test account. Pictures are not included. This comment and comment 12 were both tested with the same test account. Matthew is no longer working with us so I can't speak for him. > Are you sure that the previous import finished before you start the next > import? The import is canceled at each repro. Read step 6 of original reporter's steps. > Can we get an about:memory profile when you see the communication app > somewhere above 42MB RSS? I am guessing this is no longer needed since it's no longer repro'ing? Let me know.
Flags: needinfo?(pcheng)
Keywords: qawanted
Status: NEW → RESOLVED
blocking-b2g: 1.3T? → ---
Closed: 12 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: