990669 - [B2G][Cost Control] Data usage bar in tray not functioning and no notification appears when user reaches mobile data limit set in usage app

Status: RESOLVED DUPLICATE of bug 988979

(Firefox OS Graveyard :: Gaia::Cost Control, defect)

Description

[Brogan Zumwalt [Inactive]]

Attachment: Screenshot

Description:
After user sets data limit in Usage app, reaching or exceeding this limit does not cause a notification to appear. Opening notification tray reveals that usage bar remains at zero. Looking inside Usage app shows proper information.

Repro Steps:
1) Updated Buri to BuildID: 20140401000202
2) Launch Usage app and enable Data Use Alert
3) Set usage limit to 0.1 MB
4) Disable Wi-Fi and launch Browser app
5) Navigate to media heavy site (e.g. Youtube.com)
6) After browsing for a time drag down notification bar

Actual:
No notification is received and usage bar in notification tray remains at zero.

Expected:
Notification is received when data limit is exceeded. Notification tray usage bar shows correct information.

Environmental Variables:
Device: Buri v1.4 Mozilla RIL
BuildID: 20140401000202
Gaia: 216cc2e5c8692ba71aa78a9f27e84e9da27952b8
Gecko: 9223243c65a2
Version: 30.0a2
Firmware Version: v1.2-device.cfg

Notes:
Repro frequency: 9/10, 90%
See attached: screenshot & logcat

Comment 1

[Brogan Zumwalt [Inactive]]

Attachment: Logcat

Comment 2

[Noemí Freire (:noemi)]

Hi,

kernel config issue, Hamachi vendor should enable some modules in the kernel to make alarms work (https://bugzilla.mozilla.org/show_bug.cgi?id=960974#c1).

Comment 3

[Jason Smith [:jsmith]]

(In reply to Noemí Freire (:noemi) from comment #2)
> Hi,
> 
> kernel config issue, Hamachi vendor should enable some modules in the kernel
> to make alarms work (https://bugzilla.mozilla.org/show_bug.cgi?id=960974#c1).
> 
> *** This bug has been marked as a duplicate of bug 960974 ***

Not a dupe. This only started reproducing recently, so this is a recent regression here.

Comment 4

[Matthew Vaughan]

Currently we are concentrating on getting a regression window for the issue where the usage tracking bar in the notification area does not update with the user's data usage, and therefore not matching the Usage app (as seen in the attached screenshot). 

After further investigation, we have found this issue to be intermittent, occurring about 1 out of 5 times which may affect the accuracy of the regression window. We believe that bug 988979 (where the SIM is not detected after rebooting/resetting) is playing a part in this. And since this issue can carry over to older builds when we flash, it is going to take us a bit of time to find a regression window.

Comment 5

[Jason Smith [:jsmith]]

(In reply to Matthew Vaughan from comment #4)
> Currently we are concentrating on getting a regression window for the issue
> where the usage tracking bar in the notification area does not update with
> the user's data usage, and therefore not matching the Usage app (as seen in
> the attached screenshot). 
> 
> After further investigation, we have found this issue to be intermittent,
> occurring about 1 out of 5 times which may affect the accuracy of the
> regression window. We believe that bug 988979 (where the SIM is not detected
> after rebooting/resetting) is playing a part in this. And since this issue
> can carry over to older builds when we flash, it is going to take us a bit
> of time to find a regression window.

To simplify this - can we check if the window from bug 988979 applies to this bug as well?

Comment 6

[Matthew Vaughan]

We managed to get a regression window that was two builds deeper than the window in bug 988979, so this issue (1st paragraph in comment 4) does look to coincide with that one. 

However, the result from performing the gaia/gecko swap was inconclusive due to the issue not reproducing on either of the builds. With that, I have provided both gaia and gecko push logs. The logs did not seem to indicate which inbound branch we could use to go deeper with the window, but they look to list the same 3 commits.

TINDERBOX:
- Last Working -
Device: Buri v1.4 MOZ RIL
BuildID: 20140326123050
Gaia: 29ed37f475ecfd292722f8d9c918c817c6e57e9b
Gecko: 2f010596adf2
Version: 30.0a2
Firmware Version: V1.2-device.cfg

- First Broken -
Device: Buri v1.4 MOZ RIL
BuildID: 20140326124649
Gaia: 7d716de0c186416b5b123baa1f3242e23d50529b
Gecko: 34078570fda0
Version: 30.0a2
Firmware Version: v1.2-device.cfg

**The gaia/gecko swap test was inconclusive**
last working gaia/first broken gecko = NO REPRO
Gaia: 29ed37f475ecfd292722f8d9c918c817c6e57e9b
Gecko: 34078570fda0

first broken gaia/last working gecko = NO REPRO
Gaia: 7d716de0c186416b5b123baa1f3242e23d50529b
Gecko: 2f010596adf2

Gaia push log: https://github.com/mozilla-b2g/gaia/compare/29ed37f475ecfd292722f8d9c918c817c6e57e9b...7d716de0c186416b5b123baa1f3242e23d50529b

Gecko push log: http://hg.mozilla.org/releases/mozilla-aurora/pushloghtml?fromchange=2f010596adf2&tochange=34078570fda0

Comment 7

[Jason Smith [:jsmith]]

In that case, I'm going to conclude that this is the same root cause, as the window seems to imply that.

Comment 8

[Noemí Freire (:noemi)]

Hi Jason,

Just to confirm that everything is fine with the kernel configuration, could you please check that the following modules are enabled in config file?:

For that:
*adb pull /proc/config.gz
*unzip the file
*check that the following modules are enabled:
  CONFIG_IP_NF_RAW
  CONFIG_IP6_NF_TARGET_LOG
  CONFIG_IP6_NF_FILTER
  CONFIG_IP6_NF_MANGLE
  CONFIG_IP6_NF_TARGET_REJECT
  CONFIG_IP6_NF_TARGET_REJECT_SKERR
  CONFIG_IP6_NF_RAW
  CONFIG_IP6_NF_IPTABLES

Thanks!

Comment 9

[Brogan Zumwalt [Inactive]]

Attachment: modules config

Attached unzipped config file

Many lines mentioned in Comment 8 appear to not be enabled and others are not present in file.

My findings appear after "//" below

# CONFIG_IP_NF_RAW is not set  // Line 865 Appears to be commented out
# CONFIG_IP6_NF_TARGET_LOG is not set // Line 885 Appears to be commented out 
# CONFIG_IP6_NF_FILTER is not set // Line 886 Appears to be commented out
# CONFIG_IP6_NF_MANGLE is not set // Line 887 Appears to be commented out
// Could not find string "CONFIG_IP6_NF_TARGET_REJECT"
// Could not find string "CONFIG_IP6_NF_TARGET_REJECT_SKERR"
# CONFIG_IP6_NF_RAW is not set // Line 888 Appears to be commented out
CONFIG_IP6_NF_IPTABLES=y // Line 876 is Enabled

Comment 10

[Brogan Zumwalt [Inactive]]

Environmental Variables for info provided in comment 9:
Device: Buri v1.4 Mozilla RIL
BuildID: 20140403000210
Gaia: e0186439b9c2c03acd78e9ae011a0071865e7ffb
Gecko: 5bccd264a0e3
Version: 30.0a2
Firmware Version: v1.2-device.cfg

Comment 11

[Noemí Freire (:noemi)]

(In reply to Brogan Zumwalt from comment #9)
> Created attachment 8401458 [details]
> modules config
> 
> Attached unzipped config file
> 
> Many lines mentioned in Comment 8 appear to not be enabled and others are
> not present in file.
> 
> My findings appear after "//" below
> 
> # CONFIG_IP_NF_RAW is not set  // Line 865 Appears to be commented out
> # CONFIG_IP6_NF_TARGET_LOG is not set // Line 885 Appears to be commented
> out 
> # CONFIG_IP6_NF_FILTER is not set // Line 886 Appears to be commented out
> # CONFIG_IP6_NF_MANGLE is not set // Line 887 Appears to be commented out
> // Could not find string "CONFIG_IP6_NF_TARGET_REJECT"
> // Could not find string "CONFIG_IP6_NF_TARGET_REJECT_SKERR"
> # CONFIG_IP6_NF_RAW is not set // Line 888 Appears to be commented out
> CONFIG_IP6_NF_IPTABLES=y // Line 876 is Enabled

(In reply to Brogan Zumwalt from comment #10)
> Environmental Variables for info provided in comment 9:
> Device: Buri v1.4 Mozilla RIL
> BuildID: 20140403000210
> Gaia: e0186439b9c2c03acd78e9ae011a0071865e7ffb
> Gecko: 5bccd264a0e3
> Version: 30.0a2
> Firmware Version: v1.2-device.cfg

so we would need Hamachi vendor to enable those modules in the kernel to make alarms work. Notice that this issue started happening in v1.4 once the alarms code was landed, v1.3 branch is not affected at all due to the alarms code was not landed there.
We'd say this bug is a dupe of bug 960974.

Comment 12

[Jason Smith [:jsmith]]

(In reply to Noemí Freire (:noemi) from comment #11)
> (In reply to Brogan Zumwalt from comment #9)
> > Created attachment 8401458 [details]
> > modules config
> > 
> > Attached unzipped config file
> > 
> > Many lines mentioned in Comment 8 appear to not be enabled and others are
> > not present in file.
> > 
> > My findings appear after "//" below
> > 
> > # CONFIG_IP_NF_RAW is not set  // Line 865 Appears to be commented out
> > # CONFIG_IP6_NF_TARGET_LOG is not set // Line 885 Appears to be commented
> > out 
> > # CONFIG_IP6_NF_FILTER is not set // Line 886 Appears to be commented out
> > # CONFIG_IP6_NF_MANGLE is not set // Line 887 Appears to be commented out
> > // Could not find string "CONFIG_IP6_NF_TARGET_REJECT"
> > // Could not find string "CONFIG_IP6_NF_TARGET_REJECT_SKERR"
> > # CONFIG_IP6_NF_RAW is not set // Line 888 Appears to be commented out
> > CONFIG_IP6_NF_IPTABLES=y // Line 876 is Enabled
> 
> (In reply to Brogan Zumwalt from comment #10)
> > Environmental Variables for info provided in comment 9:
> > Device: Buri v1.4 Mozilla RIL
> > BuildID: 20140403000210
> > Gaia: e0186439b9c2c03acd78e9ae011a0071865e7ffb
> > Gecko: 5bccd264a0e3
> > Version: 30.0a2
> > Firmware Version: v1.2-device.cfg
> 
> so we would need Hamachi vendor to enable those modules in the kernel to
> make alarms work. Notice that this issue started happening in v1.4 once the
> alarms code was landed, v1.3 branch is not affected at all due to the alarms
> code was not landed there.
> We'd say this bug is a dupe of bug 960974.

As I said before, this isn't true. The window here clearly points to a loss of SIM on reboot being the cause, as we only recent started seeing this problem.