hfp voicecall fails to register in ofono with iPhone voicemail

hfp voicecall fails to register in ofono with iPhone voicemail

[Mike]

[-- Attachment #1: Type: text/plain, Size: 3791 bytes --]

I'm experiencing an issue where the +CLCC response provided by an
iPhone is ignored in the case of a user utilizing visual voicemail.
The problem here is that the AT+CLCC is being sent in response to a
+CIEV indicating call setup, but by the time the +CLCC response comes
back, the call status is active.  This is a problem because the
callback, sync_dialing_cb, is specifically looking for calls in the
dialing state.  It looks like at least some bits from clcc_poll_cb
should be copied into sync_dialing_cb (or a helper function) because I
would think any time we get a +CLCC we would want to completely update
our call list (and therefore fix the issue I'm encountering).  I'm
still digging into this, but as I'm not familiar with all the quirks
of devices out there, I'm throwing this out there in case anyone has
some ideas on this.

Thanks,
Mike

> ACL data: handle 1 flags 0x02 dlen 141
    L2CAP(d): cid 0x0040 len 137 [psm 3]
      RFCOMM(d): UIH: cr 1 dlci 14 pf 0 ilen 132 fcs 0xa5
      0000: 0d 0a 2b 43 49 4e 44 3a  20 28 22 73 65 72 76 69  ..+CIND: ("servi
      0010: 63 65 22 2c 28 30 2d 31  29 29 2c 28 22 63 61 6c  ce",(0-1)),("cal
      0020: 6c 22 2c 28 30 2d 31 29  29 2c 28 22 63 61 6c 6c  l",(0-1)),("call
      0030: 73 65 74 75 70 22 2c 28  30 2d 33 29 29 2c 28 22  setup",(0-3)),("
      0040: 62 61 74 74 63 68 67 22  2c 28 30 2d 35 29 29 2c  battchg",(0-5)),
      0050: 28 22 73 69 67 6e 61 6c  22 2c 28 30 2d 35 29 29  ("signal",(0-5))
      0060: 2c 28 22 72 6f 61 6d 22  2c 28 30 2d 31 29 29 2c  ,("roam",(0-1)),
      0070: 28 22 63 61 6c 6c 68 65  6c 64 22 2c 28 30 2d 32  ("callheld",(0-2
      0080: 29 29 0d 0a                                       ))..

...

> ACL data: handle 1 flags 0x02 dlen 22
    L2CAP(d): cid 0x0040 len 18 [psm 3]
      RFCOMM(d): UIH: cr 1 dlci 14 pf 0 ilen 14 fcs 0xa5
      0000: 0d 0a 2b 43 49 45 56 3a  20 33 2c 32 0d 0a        ..+CIEV: 3,2..
< ACL data: handle 1 flags 0x02 dlen 16
    L2CAP(d): cid 0x7800 len 12 [psm 3]
      RFCOMM(d): UIH: cr 0 dlci 14 pf 0 ilen 8 fcs 0x7f
      0000: 41 54 2b 43 4c 43 43 0d                           AT+CLCC.
> HCI Event: Number of Completed Packets (0x13) plen 5
    handle 1 packets 1
> HCI Event: Connect Request (0x04) plen 10
    bdaddr 90:27:E4:58:B8:DC class 0x7a020c type eSCO
< HCI Command: Accept Synchronous Connection (0x01|0x0029) plen 21
    bdaddr 90:27:E4:58:B8:DC
    0000: 40 1f 00 00 40 1f 00 00  ff ff 60 00 ff 3f 00     @...@.....`..?.
> HCI Event: Command Status (0x0f) plen 4
    Accept Synchronous Connection (0x01|0x0029) status 0x00 ncmd 1
> HCI Event: Synchronous Connect Complete (0x2c) plen 17
    status 0x00 handle 257 bdaddr 90:27:E4:58:B8:DC type eSCO
    Air mode: CVSD
> ACL data: handle 1 flags 0x02 dlen 22
    L2CAP(d): cid 0x0040 len 18 [psm 3]
      RFCOMM(d): UIH: cr 1 dlci 14 pf 0 ilen 14 fcs 0xa5
      0000: 0d 0a 2b 43 49 45 56 3a  20 32 2c 31 0d 0a        ..+CIEV: 2,1..
> ACL data: handle 1 flags 0x02 dlen 22
    L2CAP(d): cid 0x0040 len 18 [psm 3]
      RFCOMM(d): UIH: cr 1 dlci 14 pf 0 ilen 14 fcs 0xa5
      0000: 0d 0a 2b 43 49 45 56 3a  20 33 2c 30 0d 0a        ..+CIEV: 3,0..
> ACL data: handle 1 flags 0x02 dlen 56
    L2CAP(d): cid 0x0040 len 52 [psm 3]
      RFCOMM(d): UIH: cr 1 dlci 14 pf 0 ilen 48 fcs 0xa5
      0000: 0d 0a 2b 43 4c 43 43 3a  20 31 2c 30 2c 30 2c 30  ..+CLCC: 1,0,0,0
      0010: 2c 30 2c 22 2a 2a 32 31  31 35 33 2a 2a 22 2c 31  ,0,"**21153**",1
      0020: 32 39 2c 22 56 6f 69 63  65 6d 61 69 6c 22 0d 0a  29,"Voicemail"..
> ACL data: handle 1 flags 0x02 dlen 14
    L2CAP(d): cid 0x0040 len 10 [psm 3]
      RFCOMM(d): UIH: cr 1 dlci 14 pf 0 ilen 6 fcs 0xa5
      0000: 0d 0a 4f 4b 0d 0a                                 ..OK..

Re: hfp voicecall fails to register in ofono with iPhone voicemail

[Denis Kenzior]

[-- Attachment #1: Type: text/plain, Size: 1099 bytes --]

Hi Mike,

On 04/05/2012 11:28 PM, Mike wrote:
> I'm experiencing an issue where the +CLCC response provided by an
> iPhone is ignored in the case of a user utilizing visual voicemail.
> The problem here is that the AT+CLCC is being sent in response to a
> +CIEV indicating call setup, but by the time the +CLCC response comes
> back, the call status is active.  This is a problem because the
> callback, sync_dialing_cb, is specifically looking for calls in the
> dialing state.  It looks like at least some bits from clcc_poll_cb
> should be copied into sync_dialing_cb (or a helper function) because I
> would think any time we get a +CLCC we would want to completely update
> our call list (and therefore fix the issue I'm encountering).  I'm
> still digging into this, but as I'm not familiar with all the quirks
> of devices out there, I'm throwing this out there in case anyone has
> some ideas on this.

Are you dialing from the HF or the AG side?  And yes, the logic in
find_dialing() might indeed need to be changed to also find single
active calls.

Regards,
-Denis

Re: hfp voicecall fails to register in ofono with iPhone voicemail

[Mike]

[-- Attachment #1: Type: text/plain, Size: 1242 bytes --]

Hi Denis,

On Fri, Apr 6, 2012 at 9:36 AM, Denis Kenzior <denkenz@gmail.com> wrote:
> Hi Mike,
>
> On 04/05/2012 11:28 PM, Mike wrote:
>> I'm experiencing an issue where the +CLCC response provided by an
>> iPhone is ignored in the case of a user utilizing visual voicemail.
>> The problem here is that the AT+CLCC is being sent in response to a
>> +CIEV indicating call setup, but by the time the +CLCC response comes
>> back, the call status is active.  This is a problem because the
>> callback, sync_dialing_cb, is specifically looking for calls in the
>> dialing state.  It looks like at least some bits from clcc_poll_cb
>> should be copied into sync_dialing_cb (or a helper function) because I
>> would think any time we get a +CLCC we would want to completely update
>> our call list (and therefore fix the issue I'm encountering).  I'm
>> still digging into this, but as I'm not familiar with all the quirks
>> of devices out there, I'm throwing this out there in case anyone has
>> some ideas on this.
>
> Are you dialing from the HF or the AG side?  And yes, the logic in
> find_dialing() might indeed need to be changed to also find single
> active calls.

Dialing is on the AG side.

Thanks,
Mike

Re: hfp voicecall fails to register in ofono with iPhone voicemail

[Mike]

[-- Attachment #1: Type: text/plain, Size: 1553 bytes --]

On Fri, Apr 6, 2012 at 10:53 AM, Mike <puffy.taco@gmail.com> wrote:
> Hi Denis,
>
> On Fri, Apr 6, 2012 at 9:36 AM, Denis Kenzior <denkenz@gmail.com> wrote:
>> Hi Mike,
>>
>> On 04/05/2012 11:28 PM, Mike wrote:
>>> I'm experiencing an issue where the +CLCC response provided by an
>>> iPhone is ignored in the case of a user utilizing visual voicemail.
>>> The problem here is that the AT+CLCC is being sent in response to a
>>> +CIEV indicating call setup, but by the time the +CLCC response comes
>>> back, the call status is active.  This is a problem because the
>>> callback, sync_dialing_cb, is specifically looking for calls in the
>>> dialing state.  It looks like at least some bits from clcc_poll_cb
>>> should be copied into sync_dialing_cb (or a helper function) because I
>>> would think any time we get a +CLCC we would want to completely update
>>> our call list (and therefore fix the issue I'm encountering).  I'm
>>> still digging into this, but as I'm not familiar with all the quirks
>>> of devices out there, I'm throwing this out there in case anyone has
>>> some ideas on this.
>>
>> Are you dialing from the HF or the AG side?  And yes, the logic in
>> find_dialing() might indeed need to be changed to also find single
>> active calls.
>
> Dialing is on the AG side.

It looks like simply adding "	clcc_poll_cb(ok, result, user_data);" to
the end of sync_dialing_cb fixes this problem.  We end up parsing the
result twice, but it was an easy test.  Does that seem like the right
path?

Mike

Re: hfp voicecall fails to register in ofono with iPhone voicemail

[Denis Kenzior]

[-- Attachment #1: Type: text/plain, Size: 1994 bytes --]

Hi Mike,

On 04/10/2012 05:53 PM, Mike wrote:
> On Fri, Apr 6, 2012 at 10:53 AM, Mike <puffy.taco@gmail.com> wrote:
>> Hi Denis,
>>
>> On Fri, Apr 6, 2012 at 9:36 AM, Denis Kenzior <denkenz@gmail.com> wrote:
>>> Hi Mike,
>>>
>>> On 04/05/2012 11:28 PM, Mike wrote:
>>>> I'm experiencing an issue where the +CLCC response provided by an
>>>> iPhone is ignored in the case of a user utilizing visual voicemail.
>>>> The problem here is that the AT+CLCC is being sent in response to a
>>>> +CIEV indicating call setup, but by the time the +CLCC response comes
>>>> back, the call status is active.  This is a problem because the
>>>> callback, sync_dialing_cb, is specifically looking for calls in the
>>>> dialing state.  It looks like at least some bits from clcc_poll_cb
>>>> should be copied into sync_dialing_cb (or a helper function) because I
>>>> would think any time we get a +CLCC we would want to completely update
>>>> our call list (and therefore fix the issue I'm encountering).  I'm
>>>> still digging into this, but as I'm not familiar with all the quirks
>>>> of devices out there, I'm throwing this out there in case anyone has
>>>> some ideas on this.
>>>
>>> Are you dialing from the HF or the AG side?  And yes, the logic in
>>> find_dialing() might indeed need to be changed to also find single
>>> active calls.
>>
>> Dialing is on the AG side.
> 
> It looks like simply adding "	clcc_poll_cb(ok, result, user_data);" to
> the end of sync_dialing_cb fixes this problem.  We end up parsing the
> result twice, but it was an easy test.  Does that seem like the right
> path?
> 

Yes, sync_dialing_cb was likely a holdover from when we didn't think
we'd need the expensive clcc polling logic.  Hence a somewhat optimized
version that doesn't seem to work in case a call skips the alerting stage.

Removing sync_dialing_cb and using clcc_poll_cb seems reasonable to me.
 Care to test this theory and submit a patch?

Regards,
-Denis