847044 - [Buri][STK]can not rightly execute setup call command (disconnecting all other calls)

Status: RESOLVED WORKSFORME

(Firefox OS Graveyard :: Gaia::System, defect, P2)

Description

[sync-1]

Firefox v1.0.1,  build identifier： 20130228114642
 +++ This bug was initially created as a clone of Bug #409428 +++
 
 CONTACT INFO (Name,Phone number):
 
 DEFECT DESCRIPTION:
 [STN]can not rightly execute setup call command (disconnecting all other calls)
 
 REPRODUCING PROCEDURES:
 1. Setup a call on SET;
 2. Load and execute the test applet which can issue the setup call proactive
 command (CQ '04' = set up call, disconnecting all other calls (if any));
 3. Observe the screen display on SET. --K.O. The SET screen never shows the
 call connected information on screen, but from the network emulator, the call
 has already been setup, when try to setup another call, the previous call which
 is setup via setup call proactive command still not display.
 
 besides, this bug is in 3G network ;
 ME cannot connect with 2G network with simulated SIM card.
 but with white SIM card , ME can connect with 2G network, canot setup call
 
 EXPECTED BEHAVIOUR:
 Setup call should be represented correctly.
 
 
 ASSOCIATE SPECIFICATION:
 
 TEST PLAN REFERENCE:
 
 TOOLS AND PLATFORMS USED:
 
 USER IMPACT:
 
 REPRODUCING RATE:
 
 For FT PR, Please list reference mobile's behavior:
 
 ++++++++++ end of initial bug #409428 description ++++++++++

Comment 1

[Kai-Chih Hu [:khu]]

Looks like Gaia work. Could you help?

Comment 2

[Carol Yang [:cyang]]

Is there a GCF TC for this?

Also, please share the PDU used to test.

Comment 3

[jing.wu]

Hi Carol Yang ,
GCF case ID:27.22.4.13 Expected Sequence 1.5 (SET UP CALL, disconnecting all other calls, ME busy);
APDU:D0 3A 81 03 01 10 04 82 02 81 83 85 22 32 30 20 64 69 67 69 74 73 20 3E 3E 3E 20 30 31 32 33 34 35 36 37 38 39 30 31 32 33 34 35 36 37 38 39 86 0B 81 10 32 54 76 98 10 32 54 76 98 .

Comment 4

[jing.wu]

check on AU_LINUX_GECKO_ICS_STRAWBERRY_V1.01.00.01.19.044. setup call with CQ =04 ,disconnect all other calls, call can eatablish rightly.
but the alpha ID will prompt twice. 
if first alpha identifier is provide in command and not a null data object, the ME shall use it during the user confirmation phase. 
but now the first alpha ID will prompt twice, after user confirm, the call has established, then you enter to STK menu , the alpha ID display twice.

Comment 5

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

Can you provide some logs (logcat)?

STK GUI only orders the Modem (through a STK response code) to establish a phone call.

After that, some system messages are sent in order to show the correct UI (call in progress ;)

AFAIK, If other calls are established, is the communications app or the modem responsible to manage that. (put onhold, join them, ...)

Comment 6

[Carol Yang [:cyang]]

Hi Jing,

Can you please clarify your question? The original bug was opened with: "The SET screen never shows the call connected information on screen, but from the network emulator, the call has already been setup"

However, in Comment 4, you mentioned: "the alpha ID will prompt twice"

I'm assuming the original issue reported is no longer? Please confirm.

Comment 7

[Carol Yang [:cyang]]

Attachment: STKUI debug trace

Hi Fernando,

If I understand this bug correctly, I think Jing is reporting that she sees the popup show up twice for a given STK command. I just ran a quick test by setting up an STK menu and selecting the item would send a SET_UP_CALL proactive command. I only selected the menu item once but in the logs, I do see two of these messages with timestamps very close together.

1933:01-07 07:22:43.969   443   443 I Gecko   : -*- RILContentHelper: Received message 'RIL:StkCommand': {"commandNumber":1,"typeOfCommand":16,"commandQualifier":4,"options":{"confirmMessage":"Disconnect","address":"012340123456,1,2"}}
1934:01-07 07:22:43.989   130   130 I Gecko   : -*- RILContentHelper: Received message 'RIL:StkCommand': {"commandNumber":1,"typeOfCommand":16,"commandQualifier":4,"options":{"confirmMessage":"Disconnect","address":"012340123456,1,2"}}

The popup shows up, after I press OK, I do see a second popup. Pressing OK a second time, no more popups are displayed.

PDU used: D020810301100482028183850A446973636F6E6E6563748609911032042143651C2C

Logs attached.

Comment 8

[jing.wu]

Attachment: QXDM and ADB logs part1

Comment 9

[jing.wu]

Attachment: QXDM and ADB logs part2

Comment 10

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

(In reply to Carol Yang from comment #7)
> Created attachment 729896 [details]
> STKUI debug trace
> 
> Hi Fernando,
> 
> If I understand this bug correctly, I think Jing is reporting that she sees
> the popup show up twice for a given STK command. I just ran a quick test by
> setting up an STK menu and selecting the item would send a SET_UP_CALL
> proactive command. I only selected the menu item once but in the logs, I do
> see two of these messages with timestamps very close together.
> 
> 1933:01-07 07:22:43.969   443   443 I Gecko   : -*- RILContentHelper:
> Received message 'RIL:StkCommand':
> {"commandNumber":1,"typeOfCommand":16,"commandQualifier":4,"options":
> {"confirmMessage":"Disconnect","address":"012340123456,1,2"}}
> 1934:01-07 07:22:43.989   130   130 I Gecko   : -*- RILContentHelper:
> Received message 'RIL:StkCommand':
> {"commandNumber":1,"typeOfCommand":16,"commandQualifier":4,"options":
> {"confirmMessage":"Disconnect","address":"012340123456,1,2"}}
> 
> The popup shows up, after I press OK, I do see a second popup. Pressing OK a
> second time, no more popups are displayed.
> 
> PDU used:
> D020810301100482028183850A446973636F6E6E6563748609911032042143651C2C
> 
> Logs attached.

Thanks Carol.

So the problem is in the platform. If the platform sends two messages, the GUI will process two messages.

Asking Yoshi to get platform feedback about this "two very close messages" ;)

Comment 11

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

(In reply to Carol Yang from comment #7) 
> 1933:01-07 07:22:43.969   443   443 I Gecko   : -*- RILContentHelper:
> Received message 'RIL:StkCommand':
> {"commandNumber":1,"typeOfCommand":16,"commandQualifier":4,"options":
> {"confirmMessage":"Disconnect","address":"012340123456,1,2"}}
> 1934:01-07 07:22:43.989   130   130 I Gecko   : -*- RILContentHelper:
> Received message 'RIL:StkCommand':
> {"commandNumber":1,"typeOfCommand":16,"commandQualifier":4,"options":
> {"confirmMessage":"Disconnect","address":"012340123456,1,2"}}

Hi Carol
The log you see is from RILContentHelper, also you can see the PID is different in each log, it's a normal situtation here.

Comment 12

[Carol Yang [:cyang]]

(In reply to Yoshi Huang[:allstars.chh][:yoshi] from comment #11)
> (In reply to Carol Yang from comment #7) 
> Hi Carol
> The log you see is from RILContentHelper, also you can see the PID is
> different in each log, it's a normal situtation here.

Hi Yoshi,

Can you please help clarify a few things:

1) Based on https://github.com/mozilla/mozilla-central/blob/b2g18_v1_0_1/dom/system/gonk/RadioInterfaceLayer.js#L1712, I was under the impression that broadcastMessage() is called to open Settings app and then sendTargetMessage() is what Gaia will use to process the STK command and display popups, etc. Is this correct?

2) When I comment sendTargetMessage() out, I still see the STK command popup on the UI. So broadcastMessage() is all that is needed? Or is this incorrect behavior?

3) When I comment broadcastMessage() out (and left sendTargetMessage uncommented), I was still seeing that RILContentHelper was receving two 'RIL:StkCommand' messages (with 2 different PIDs).  This doesn't seem right. Can you please comment on this?

Comment 13

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

(In reply to Carol Yang from comment #12)
> (In reply to Yoshi Huang[:allstars.chh][:yoshi] from comment #11)
> > (In reply to Carol Yang from comment #7) 
> > Hi Carol
> > The log you see is from RILContentHelper, also you can see the PID is
> > different in each log, it's a normal situtation here.
> 
> Hi Yoshi,
> 
> Can you please help clarify a few things:
> 
> 1) Based on
> https://github.com/mozilla/mozilla-central/blob/b2g18_v1_0_1/dom/system/gonk/
> RadioInterfaceLayer.js#L1712, I was under the impression that
> broadcastMessage() is called to open Settings app and then
> sendTargetMessage() is what Gaia will use to process the STK command and
> display popups, etc. Is this correct?
> 
sendTargerMessage is used to call the icc.onstkcommand callback on content side.

However, the content side might not start running yet,
so System Message is used to wake up the app and send the command to it.

> 2) When I comment sendTargetMessage() out, I still see the STK command popup
> on the UI. So broadcastMessage() is all that is needed? Or is this incorrect
> behavior?
> 
> 3) When I comment broadcastMessage() out (and left sendTargetMessage
> uncommented),

I've explained their purposes above.
Both will send a STK command to content side.
But I am not sure what happens if content side will listen both messages
(onstkcommand and icc-stkcommand system message)

> I was still seeing that RILContentHelper was receving two
> 'RIL:StkCommand' messages (with 2 different PIDs).  This doesn't seem right.
> Can you please comment on this?

RILContentHelper will have one instance in content side, and one in chrome side.
It's normal to have 2 instances of RILContentHelper.

I think you could check a normal STK log and you can also see two instances of RILContentHelper, no?

Comment 14

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

(In reply to Yoshi Huang[:allstars.chh][:yoshi] from comment #13)
> 
> I've explained their purposes above.
> Both will send a STK command to content side.
> But I am not sure what happens if content side will listen both messages
> (onstkcommand and icc-stkcommand system message)
> 

Hi Yoshi,

Currently, icc-stkcommand is used in system:
https://github.com/mozilla-b2g/gaia/blob/master/apps/system/js/icc_cache.js#L37

and stkcommand in settings:
https://github.com/mozilla-b2g/gaia/blob/master/apps/settings/js/icc.js#L84

Do you assure me that all STK Commands will be received through icc-stkcommand?

In that case, I can drop the settings listener, but I need to be sure that all messages will be driven through icc-stkcommand

Comment 15

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

(In reply to Fernando R. Sela [:frsela] from comment #14)
> Currently, icc-stkcommand is used in system:
> https://github.com/mozilla-b2g/gaia/blob/master/apps/system/js/icc_cache.
> js#L37
> 
> and stkcommand in settings:
> https://github.com/mozilla-b2g/gaia/blob/master/apps/settings/js/icc.js#L84
> 
> Do you assure me that all STK Commands will be received through
> icc-stkcommand?
> 
> In that case, I can drop the settings listener, but I need to be sure that
> all messages will be driven through icc-stkcommand

Yes, All STK commands will be notified through system message "icc-stkcommand".

Comment 16

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

Attachment: WIP: All STK messages attended in system and managed in settings only with the mozsettings observer

Thanks Yoshi for your fast response.
Carol, can you test this WIP patch?

Thanks in advance !

Comment 17

[Carol Yang [:cyang]]

Fernando/Yoshi- Thanks for the explanations!

I ran two different tests with the same PDU:

Test1) Do not launch the Settings app and wait for STK command while on homescreen
Test2) Launch the Settings app and then wait for STK command on homescreen

Prior to applying attachment 731764 [details], this is what I observed:
- Test1: One popup is displayed and I saw one 'RIL:StkCommand' messages received by RILContentHelper
- Test2: Two popups are displayed and I saw two 'RIL:StkCommand' messages received by RILContentHelper

After I applied attachment 731764 [details], this is what I observed:
- Test1: The empty SIM Toolkit menu is launched and no popup comes up at all, with one 'RIL:StkCommand' received by RILContentHelper
Test2: One popup is displayed and I saw two 'RIL:StkCommand' messages received by RILContentHelper

So, to better summarize, the original issue of the same popup coming up twice for one STK command was only happening for Test2. Looks like with your fix, it broke Test1 which was working already.

Comment 18

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

(In reply to Carol Yang from comment #17)
> Fernando/Yoshi- Thanks for the explanations!
> 
> I ran two different tests with the same PDU:
> 
> Test1) Do not launch the Settings app and wait for STK command while on
> homescreen
> Test2) Launch the Settings app and then wait for STK command on homescreen
> 
> Prior to applying attachment 731764 [details], this is what I observed:
> - Test1: One popup is displayed and I saw one 'RIL:StkCommand' messages
> received by RILContentHelper
> - Test2: Two popups are displayed and I saw two 'RIL:StkCommand' messages
> received by RILContentHelper
> 
> After I applied attachment 731764 [details], this is what I observed:
> - Test1: The empty SIM Toolkit menu is launched and no popup comes up at
> all, with one 'RIL:StkCommand' received by RILContentHelper
> Test2: One popup is displayed and I saw two 'RIL:StkCommand' messages
> received by RILContentHelper
> 
> So, to better summarize, the original issue of the same popup coming up
> twice for one STK command was only happening for Test2. Looks like with your
> fix, it broke Test1 which was working already.

I'll check. Thanks for the feedback !

Comment 19

[buri.blff]

Has the new patch been landed on  v1.0.1???

Comment 20

[buri.blff]

Dear frsela:
   Is there any new patch for this bug??

Comment 21

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

(In reply to buri.blff from comment #20)
> Dear frsela:
>    Is there any new patch for this bug??

No yet, I'm working on tef+ and other urgent issues. No tef+ has less priority. Sorry.

I'll do my best to advance this as fast as possible. If you consider this critical, please propose it as tef?

Comment 22

[Peter Dolanjski [:pdol]]

Kev, do you know if this would be a blocker for your partner?

Comment 23

[Peter Dolanjski [:pdol]]

Moving NI to Wilfred as Kev is on vacation.

Comment 24

[Kev Needham [:kev]]

Has any related info been filed related to this issue with the IOT process? Further, would this occur in a typical user setting? From what I've read/can see, this is unlikely to impact real world usage unless there was a STK app bundled with the device which is used during a call, which I do not believe is planned for 1.0.1. 

Joe, can you confirm with TCL, and Yoshi, could you also confirm that this is unlikely to occur?

I don't think I've seen this on the IOT blockers thus far.

Comment 25

[Alex Keybl [:akeybl]]

Let's ask for a renom if necessary in that case.

Comment 26

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

Popup messages twice should be fixed with master branch + this patch:

https://bugzilla.mozilla.org/attachment.cgi?id=768857

from bug 886808

Can you check if with this patches this bug is solved? Thanks in advance

Comment 27

[Fernando R. Sela (no CC, needinfo please) [:frsela]]

Work for me.
If continues failing, reopen it. Thanks