[OPEN II_1.3] Data connection couldn't be established when handover from GSM to UTRAN

UNCONFIRMED
Unassigned

Status

Firefox OS
RIL
P1
critical
UNCONFIRMED
4 years ago
2 years ago

People

(Reporter: Haicheng, Unassigned)

Tracking

Firefox Tracking Flags

(Not tracked)

Details

Attachments

(2 attachments)

426.19 KB, application/x-7z-compressed
Details
228.06 KB, application/x-7z-compressed
Details
(Reporter)

Description

4 years ago
Created attachment 8457803 [details]
OPEN_II_AP_logs.7z

User Agent: Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; Trident/4.0; SLCC2; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729; Media Center PC 6.0; InfoPath.2; .NET4.0C; .NET4.0E; TCO_20140717113824)

Steps to reproduce:

step 1. place a voice call in 2G only area
step 2. move to 3G area
step 3. open a simple website(google homepage) in Browser.


Actual results:

Failed to open this website. It said "network connection unavailable" in Broswer.


Expected results:

Open this website successful.
(Reporter)

Updated

4 years ago
Severity: normal → critical
Flags: needinfo?(vchen)
OS: All → Gonk (Firefox OS)
Priority: -- → P1
Hardware: All → ARM
Hi Ken -

Partner already check with QCT about this issue and QCT mentioned that everything works find in the Modem side, therefore could you help to find someone to see if we can check why the data connection isn't available in this scenario?

Thanks

Vance
Flags: needinfo?(vchen) → needinfo?(kchang)

Comment 2

4 years ago
Vance, can partner help get a log of network manager?
Jessica, please help this bug. Thanks.
Flags: needinfo?(vchen)
Flags: needinfo?(kchang)
Flags: needinfo?(jjong)
Sure, could you let me know how to get network manager log? is it just normal logcat log, or need special way to get it?

Thanks
Flags: needinfo?(vchen)
From the log, it is using qcril. It'd be better to have NetworkManager log enabled to identify where the problem is.
To turn on NetworkManager log:
1. adb pull /system/b2g/defaults/pref/user.js
2. Add "pref('network.debugging.enabled', true)" in user.js
3. adb remount
4. adb push user.js /system/b2g/defaults/pref/
5. adb shell sync && adb reboot

NetworkManager log will appear in normal logcat log. BTW, which version of ffos is this? Thanks.
Assignee: nobody → jjong
Flags: needinfo?(jjong)
(Reporter)

Comment 5

4 years ago
It is FFOS 1.3.

Thanks
(Reporter)

Comment 6

4 years ago
Created attachment 8458373 [details]
OPEN_II_APLogs_20140718.7z

Dear Mozilla Team,

Attached the logs which enable Network Manager.

I found abnormal logs below, please help to check.
07-17 11:50:32.130 V/DATA_CALL_TRACKER_QC_B2G(  295): UpdateNetworkManager
07-17 11:50:32.130 V/DATA_CALL_QC_B2G(  295): RegisterDataCall
07-17 11:50:32.220 D/DHCP    (  295): getaddrinfo failed: invalid address 
07-17 11:50:32.220 D/NetUtils(  295): failed to remove default route for rmnet0: No such process
07-17 11:50:32.450 D/DHCP    (  295): getaddrinfo failed: invalid address 

Thanks
ni Jessica since log is available now
Flags: needinfo?(jjong)
Actually, the NetworkManager log was not enabled in the logs, what you see is qcril's log. If NetworkManager log is enabled, you'll see lines like this:
I/Gecko   (  288): -*- NetworkManager: Network 1/rmnet0 changed state to 1
I/Gecko   (  288): -*- NetworkManager: Evaluating whether active network needs to be changed.

What I observe is that data call is unregistered when voice call is started and registered again when voice call is ended. If that is the case, we may need caf's help to check on this.
Please help get logs with NetworkManager log enabled to confirm this. Thanks!
Flags: needinfo?(jjong)
So sorry for this, I was reminded that 1.3 does not support enabling NetworkManager log dynamically. orz
Vance, could you ask help from partner to confirm the statement in comment 8? Thanks a lot.

(In reply to Jessica Jong [:jjong] [:jessica] from comment #8)
> Actually, the NetworkManager log was not enabled in the logs, what you see
> is qcril's log. If NetworkManager log is enabled, you'll see lines like this:
> I/Gecko   (  288): -*- NetworkManager: Network 1/rmnet0 changed state to 1
> I/Gecko   (  288): -*- NetworkManager: Evaluating whether active network
> needs to be changed.
> 
> What I observe is that data call is unregistered when voice call is started
> and registered again when voice call is ended. If that is the case, we may
> need caf's help to check on this.
> Please help get logs with NetworkManager log enabled to confirm this. Thanks!
Flags: needinfo?(vchen)
Hi Jessica, sorry, not sure what you want me to confirm with partner...confirm this ? 

"that data call is unregistered when voice call is started and registered again when voice call is ended"

If so, how to confirm except checking the logs? Observe the data transcation icon or check of PDP Activate?

Thanks
Flags: needinfo?(vchen) → needinfo?(jjong)
Oh, sorry for not being clearer. The following is what we observed from the logs: "data call is unregistered when voice call is started and registered again when voice call is ended".
I would like partner to confirm whether their code does this; if so, does their code do anything if radio tech changes while in call? if not, then it is expected that network won't work while in call. Thanks.

(In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #10)
> Hi Jessica, sorry, not sure what you want me to confirm with
> partner...confirm this ? 
> 
> "that data call is unregistered when voice call is started and registered
> again when voice call is ended"
> 
> If so, how to confirm except checking the logs? Observe the data transcation
> icon or check of PDP Activate?
> 
> Thanks
Flags: needinfo?(jjong) → needinfo?(vchen)
(In reply to Jessica Jong [:jjong] [:jessica] from comment #11)
> Oh, sorry for not being clearer. The following is what we observed from the
> logs: "data call is unregistered when voice call is started and registered
> again when voice call is ended".
> I would like partner to confirm whether their code does this; if so, does
> their code do anything if radio tech changes while in call? if not, then it
> is expected that network won't work while in call. Thanks.
> 

Hi Jessica, sorry, one last stupid question, so you want partner to check if they have modified the QCT RIL part such that the behavior is "data call is unregistered when voice call is started and registered
 again when voice call is ended"

Is that right?

Thanks
Flags: needinfo?(vchen) → needinfo?(jjong)
(In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #12)
> (In reply to Jessica Jong [:jjong] [:jessica] from comment #11)
> > Oh, sorry for not being clearer. The following is what we observed from the
> > logs: "data call is unregistered when voice call is started and registered
> > again when voice call is ended".
> > I would like partner to confirm whether their code does this; if so, does
> > their code do anything if radio tech changes while in call? if not, then it
> > is expected that network won't work while in call. Thanks.
> > 
> 
> Hi Jessica, sorry, one last stupid question, so you want partner to check if
> they have modified the QCT RIL part such that the behavior is "data call is
> unregistered when voice call is started and registered
>  again when voice call is ended"
> 
> Is that right?
> 
> Thanks

Yes, to know if that is the expected behavior from their code. Thanks.
Flags: needinfo?(jjong)
(In reply to Jessica Jong [:jjong] [:jessica] from comment #13)
> (In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #12)
> > (In reply to Jessica Jong [:jjong] [:jessica] from comment #11)
> > > Oh, sorry for not being clearer. The following is what we observed from the
> > > logs: "data call is unregistered when voice call is started and registered
> > > again when voice call is ended".
> > > I would like partner to confirm whether their code does this; if so, does
> > > their code do anything if radio tech changes while in call? if not, then it
> > > is expected that network won't work while in call. Thanks.
> > > 
> > 
> > Hi Jessica, sorry, one last stupid question, so you want partner to check if
> > they have modified the QCT RIL part such that the behavior is "data call is
> > unregistered when voice call is started and registered
> >  again when voice call is ended"
> > 
> > Is that right?
> > 
> > Thanks
> 
> Yes, to know if that is the expected behavior from their code. Thanks.

Thanks, already send an email to partner to check.

BTW, out of curiosity, could we see from the log who trigger the data call unregistered while establishing the voice call?

Thanks

Vance
Flags: needinfo?(jjong)
(In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #14)
> (In reply to Jessica Jong [:jjong] [:jessica] from comment #13)
> > (In reply to Vance Chen [:vchen][vchen@mozilla.com] from comment #12)
> > > (In reply to Jessica Jong [:jjong] [:jessica] from comment #11)
> > > > Oh, sorry for not being clearer. The following is what we observed from the
> > > > logs: "data call is unregistered when voice call is started and registered
> > > > again when voice call is ended".
> > > > I would like partner to confirm whether their code does this; if so, does
> > > > their code do anything if radio tech changes while in call? if not, then it
> > > > is expected that network won't work while in call. Thanks.
> > > > 
> > > 
> > > Hi Jessica, sorry, one last stupid question, so you want partner to check if
> > > they have modified the QCT RIL part such that the behavior is "data call is
> > > unregistered when voice call is started and registered
> > >  again when voice call is ended"
> > > 
> > > Is that right?
> > > 
> > > Thanks
> > 
> > Yes, to know if that is the expected behavior from their code. Thanks.
> 
> Thanks, already send an email to partner to check.
> 
> BTW, out of curiosity, could we see from the log who trigger the data call
> unregistered while establishing the voice call?
> 
> Thanks
> 
> Vance

Hi Vance,

I believe Jessica has done her best to investigate the issue. As this is not mozril, I think our partner should be the right person to answer your question. Thank you for the help!
Flags: needinfo?(jjong)
Assignee: jjong → nobody
You need to log in before you can comment on or make changes to this bug.