825123 - [Settings][Cellular&Data] Not able to get data connection and not able to automatically fetch "Data settings" when roaming

Status: RESOLVED WONTFIX

(Firefox OS Graveyard :: General, defect)

Description

[Walter Chen[:ypwalter][:wachen]]

2012-12-27 date build from pvt server

STR:
 0. make sure you have a foreign data plan sim
 1. turn on the phone
 2. go to "Settings" app
 3. go to "Cellular & Data" 
 4. turn on "Data connection" & "Data Roaming"
 5. go to "Data settings"

expected result:
  a. "adb logcat -b Radio" it, you will see some action in step 4
  b. It should get data settings of current roaming service provider from database in step 5


actual result:
  a. "adb logcat -b Radio" it, you will see nothing in step 4
  b. There is no information in data settings in step 5

Comment 1

[Walter Chen[:ypwalter][:wachen]]

just another comment that I was using China Mobile sim card for this one.
not able to check using other sim card.

Comment 2

[Shian-Yow Wu [:swu]]

Could be related to bug 825075.

Comment 3

[Joe Cheng [:jcheng] (please needinfo)]

Tried an AT&T SIM card with Walter. Things seems to work well with AT&T SIM.
It seems to be a problem with not having a APN settings in our database for China Mobile

Comment 4

[Shian-Yow Wu [:swu]]

Per discussion with Kaze, the APN settings are retrieved base on MCC/MNC from SIM card.

The China Mobile SIM card been tested has a strange MCC/MNC.  So this issue could be related to SIM card.

I/Gecko   (  832): RIL Worker: Handling parcel as REQUEST_SIM_IO
I/Gecko   (  832): RIL Worker: AD: 0, 0, 0,
I/Gecko   (  832): RIL Worker: MCC: 204 MNC: 17924

Comment 5

[Fabien Cazenave [:kaze]]

Nice MNC. :-)

Note that you could add a specific setting for this operator in the /shared/resources/apn/apn_confs_local.xml file and rebuild apn.json accordingly. We already had to do it for a Canadian operator.

Comment 6

[Fabien Cazenave [:kaze]]

FTR, we have China Mobile entries in apn.json for MCC=460.

Comment 7

[Shian-Yow Wu [:swu]]

Hi Kaze,

Thanks for the info!

Hi Yoshi,

It seems MNC value was not correctly parsed due to missing MNC length information.

I/Gecko   (  832): -*- RadioInterfaceLayer: Received message from worker: {"isDisplayNetworkNameRequired":false,"isDisplaySpnRequired":true,"rilMessageType":"iccinfochange","imsi":"204043004929951","iccid":"89860312710103107205","ad":{"0":0,"1":0,"2":0},"mcc":204,"mnc":17924,"sst":{"0":255,"1":63,"2":255,"3":15,"4":63,"5":0,"6":60,"7":243,"8":0,"9":12},"spn":"中å<9b>½ç<94>µä¿¡"}

Comment 8

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

(In reply to Shian-Yow Wu from comment #7)
"imsi":"204043004929951",
"ad":{"0":0,"1":0,"2":0},

AD is mandatory EF and should not be 0 and should have at least 4 octets.

Comment 9

[Shian-Yow Wu [:swu]]

If it's SIM card problem, should we mark this bug as RESOLVED/WONTFIX?

Comment 10

[Joe Cheng [:jcheng] (please needinfo)]

(In reply to Shian-Yow Wu from comment #9)
> If it's SIM card problem, should we mark this bug as RESOLVED/WONTFIX?

better error handling? i believe it's a commercial SIM that's tested

Comment 11

[Andrew Overholt [:overholt]]

This doesn't sound like a blocker but getting better error handling is something that would be nice so please request uplift when a patch is ready.

Comment 12

[Walter Chen[:ypwalter][:wachen]]

(In reply to Shian-Yow Wu from comment #9)
> If it's SIM card problem, should we mark this bug as RESOLVED/WONTFIX?

I think we still need to fix it. Just it won't be a urgent case.
And, we might need to have better error handling since we don't know what other phone companies' sim card would do.

(It happened to CHT phone company in TW before that it's affecting bluetooth. I think bluetooth gecko/gaia team did fix it.)

Comment 13

[Walter Chen[:ypwalter][:wachen]]

Just btw, China Unicom mobile phone service provider's sim card is working fine.

Comment 14

[Walter Chen[:ypwalter][:wachen]]

Any news on this one?

Comment 15

[Shian-Yow Wu [:swu]]

Yoshi, do you think we can do better error handling on this?

Comment 16

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

I dont see any error from RIL.
Why do you keep saying "error handling"?

Comment 17

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

Marking as WONTFIX as it's defect on the SIM.
Also we haven't heard any China partner report this problem in these 3 months.

Comment 18

[Walter Chen[:ypwalter][:wachen]]

There is no China partner reporting this because this is a ROAMING issue.
And, for all the same company SIM card would act the same (means 1/4 of China mobile provider). Please reconsider to do a better handling?

Comment 19

[Yoshi Cheng-Hao Huang [:allstars.chh][:allstarschh][:yoshi]]

(In reply to Walter Chen from comment #18)
> There is no China partner reporting this because this is a ROAMING issue.
Operators have their machines worth million dollars to test that,
so they don't need to travel around the world.

> And, for all the same company SIM card would act the same (means 1/4 of
> China mobile provider). Please reconsider to do a better handling?

You need to discuss with the operator first.
From your previous log, their SIMs doesn't have a mandatory field called EF_AD. That's defined by 3GPP spec.
Without the spec we nor modem vendor have no idea how to read those values.
That's why I say it's a defect on that SIM.

An example is if you don't have a SIM card inserted in your phone, you cannot call your friends.
In this case, the SIM doesn't have EF_AD in it so RIL doesn't know how to retrieve those MCC/MNC.

Before we have one similar issue like this, the SIM from one of our partners doesn't have voice mail stored in it. (Although in that case , voice mail is optional). After we discuss with the partner, they responded immediately their SIMs don't support this.

Please provide a better reason or information for us to proceed, or even correct me if I got wrong in my analysis, if you still want to fix bug.