1004170 - [B2G][OpenC][Wifi] - Wi-Fi toggles on and off repeatedly after toggling airplane mode on and off

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: Vendcom, defect)

Description

[Joshua Mitchell (Inactive)]

Attachment: logcat.txt

Description:
When wi-fi is connected and user toggles on airplane mode wifi will turn off. When the user then turns off airplane mode, wifi turns back on, but will continuously toggle between on and off. If the user is watching the pull-down notification bar the wifi icon will switch between blue (on) and grey (off) or if the user is in the Wi-Fi area of the settings menu they will see the toggle switch back and forth by itself.
Notes: After 5-8 minutes of turning off and on the wifi is finally able to establish a lasting connection.

Repro Steps:
1) Update a Open C to BuildID: 20140430000201
2) Turn on Wi-Fi and connect to a network.
3) Pull down the notification menu and toggle ON airplane mode.
4) Wait 20 seconds then toggle OFF airplane mode.
5) Observe the behaviour of the Wi-Fi

Actual:
Wifi turns on and off
Expected:
(Describe the behavior you expected to have observed)

1.4 Environmental Variables:
Device: Open C 1.4 MOZ
BuildID: 20140430000201
Gaia: 81e97c3ca58be0487292011bc59efa4cebab30be
Gecko: 123485e733d5
Version: 30.0
Firmware Version: P821A10-Eng_20140410

Repro frequency: 50%
Link to failed test case: https://moztrap.mozilla.org/manage/case/1341/
See attached:logcat.txt

Comment 1

[Joshua Mitchell (Inactive)]

This DOES NOT repro on the 1.4 buri
1.4 Environmental Variables:
Device: Buri 1.4 MOZ
BuildID: 20140430000201
Gaia: 81e97c3ca58be0487292011bc59efa4cebab30be
Gecko: 123485e733d5
Version: 30.0
Firmware Version: V1.2-device.cfg

Comment 2

[Joshua Mitchell (Inactive)]

This DOES occur on the 2.0 Open C
2 Environmental Variables:
Device: Open C 2 MOZ
BuildID: 20140430040206
Gaia: badc73ee7f108fa631150bded0cc57e92aad810e
Gecko: e19812f56952
Version: 32.0a1
Firmware Version: P821A10-Eng_20140410

Comment 3

[Joshua Mitchell (Inactive)]

flagging this for  QAwanted to see if this occurs on the base build

Comment 4

[Chuck Lee [:chucklee, Triggered by NI, Review, and Feedback requests]]

I think it's caused by bug 997791, because Gaia part for 1.4 is not landed.
I'll make it landed today and please test again on tomorrows build, thanks.

Comment 5

[Derek Harris [:DerekH] - Dont need info. Last day 6/5/15]

I was unable to reproduce this issue on Open C 1.3

1.3 Environmental Variables:
Device: Open_C 1.3
BuildID: 20140505052400
Gaia: Unknown Git commit; build date shown here.
Gecko:
Version: 28.0
Firmware Version: FFOS_US_EBAY_P821A10V1.0.0B06_LOG_DL

Comment 6

[Jason Smith [:jsmith]]

(In reply to Chuck Lee [:chucklee] from comment #4)
> I think it's caused by bug 997791, because Gaia part for 1.4 is not landed.
> I'll make it landed today and please test again on tomorrows build, thanks.

Marking QA Wanted to retest per this comment on 1.4.

Comment 7

[Joshua Mitchell (Inactive)]

This issue DOES still repro on the latest Open_C 1.4
1.4 Environmental Variables:
Device: Open_C 1.4 MOZ
BuildID: 20140508000201
Gaia: 4ce973ef0732b0d52cb043210db598aa176b2ce9
Gecko: 16ab7f6b18f8
Version: 30.0
Firmware Version: FFOS_US_EBAY_P821A10V1.0.0B06_LOG_DL

Comment 8

[Chuck Lee [:chucklee, Triggered by NI, Review, and Feedback requests]]

Based on previous test report, this bug seems only happens on Open C after 1.4?

But I don't have the device for testing.

Comment 9

[Chuck Lee [:chucklee, Triggered by NI, Review, and Feedback requests]]

Please check again on Buri 1.4 and 2.0, thanks.

Comment 10

[Chuck Lee [:chucklee, Triggered by NI, Review, and Feedback requests]]

There's no wifi disable command "TERMINATE" in attached log, please attach complete log with "Settings->Developer->Wi-Fi output in adb" enabled, thanks.

Comment 11

[Chuck Lee [:chucklee, Triggered by NI, Review, and Feedback requests]]

There's unknown error, not sure if it's related to this bug:

W/SocketClient(274): write error (Broken pipe)

Comment 12

[Joshua Mitchell (Inactive)]

Attachment: logcat.txt

Adding a new logcat with all the appropriate settings enabled.  I was able to reproduce this on todays 1.4 Open_C build however the bug was slightly different. After the 4th time of turning airplane mode ON the wifi became stuck in a loop of enabling and disabling whereas before the wifi loop would begin when turning airplane mode to off.
1.4 Environmental Variables:
Device: Open_C 1.4 MOZ
BuildID: 20140509000203
Gaia: f19735d288d9bf1c6ee0c0ecc7941421365037c7
Gecko: 33aff135dc42
Version: 30.0
Firmware Version: FFOS_US_EBAY_P821A10V1.0.0B06_LOG_DL

Comment 13

[Joshua Mitchell (Inactive)]

(In reply to Chuck Lee [:chucklee] from comment #9)
> Please check again on Buri 1.4 and 2.0, thanks.

This does NOT reproduce on the Buri 1.4

1.4 Environmental Variables:
Device: Buri 1.4 MOZ
BuildID: 20140508000201
Gaia: 4ce973ef0732b0d52cb043210db598aa176b2ce9
Gecko: 16ab7f6b18f8
Version: 30.0
Firmware Version: v1.2-device.cfg

This also does NOT reproduce on the Buri 2.0

2.0 Environmental Variables:
Device: Buri 2.0 MOZ
BuildID: 20140508040203
Gaia: 1e0574b8f6b8a2a8d9d468878ce2b4c283fc9a84
Gecko: 2a03b34c8953
Version: 32.0a1
Firmware Version: v1.2-device.cfg

Comment 14

[Jason Smith [:jsmith]]

Can we see if we can reproduce this problem on Flame?

Comment 15

[Joshua Mitchell (Inactive)]

This issue DOES NOT reproduce on 1.4 Flame

1.4 Environmental Variables:
Device: Flame 1.4 MOZ
BuildID: 20140509000203
Gaia: f19735d288d9bf1c6ee0c0ecc7941421365037c7
Gecko: 33aff135dc42
Version: 30.0
Firmware Version: v10E

Comment 16

[Jason Smith [:jsmith]]

Chuck - Why is this only present on Open C on 1.4 or later? Seems strange to me that this works fine on Open C 1.3, Flame, and Buri, but fails on Open C 1.4 or later.

Comment 17

[Chuck Lee [:chucklee, Triggered by NI, Review, and Feedback requests]]

Basically I have no clue.

The log didn't show much incoming state from wpa_supplicant, this might cause internal state error in gecko.
It might related to these lines of "W/SocketClient(274): write error (Broken pipe)".

Also, it seems wpa_supplicant is closed by some unknown way: command TERMINATE didn't show up in the log, and it seems supplicant is closed by next enable command.
And becomes strange behavior that wpa_supplicant is disabling and enabling at same time.
05-09 14:05:51.713: W/SocketClient(274): write error (Broken pipe)
05-09 14:05:52.553: D/QualcommSoftapCmd(274): Got softap fwreload command we are passing on
05-09 14:05:52.553: D/SoftapController(274): Softap fwReload - Ok
05-09 14:05:54.563: I/Gecko(281): -*- WifiWorker component: Event coming in: CTRL-EVENT-TERMINATING  - connection closed
05-09 14:05:54.563: E/WifiHW(281): ZQL wifi_start_supplicant with p2p_supported=0
05-09 14:05:54.563: E/WifiHW(281): ZQL supplicant_name is wpa_supplicant and supplicant_prop_name is init.svc.wpa_supplicant
05-09 14:05:54.583: I/Gecko(281): -*- WifiWorker component: State change: INITIALIZING -> DISCONNECTED
05-09 14:05:54.603: I/Gecko(281): -*- WifiWorker component: Supplicant died!

maybe Open C 1.4 uses different gonk as Open 1.3, and the behavior is different in wifi, and mentioned logs above gonk-related.
I think partner should check on this error first.

Comment 18

[BMO Automation]

Firefox OS is not being worked on