Age | Commit message (Collapse) | Author |
|
Refactors ImsManager and the ImsServiceProxy
to use the specific ImsFeature AIDLs instead
of a flat ImsService interface.
Test: Manual, Telephony IMS unit tests
Change-Id: I00155a0092f7e2527bbb96c196a59f0d0d4039eb
|
|
Replace the default value of Enhanced 4G LTE mode with carrier confg
to control it by carrier configuration.
Test: manual - Checked that the Enhanced 4G LTE mode can be controlled
by carrier config.
Bug: 67725875
Change-Id: Ib4b19692670e7343e52f1840e7d768dbd3abf8f8
|
|
Removes ImsService interfaces from Proxy
in preparation to move to feature based
AIDLs instead of a flat interface
Test: Manual
Change-Id: Id408f3d3f180aea7a2c706d703665ca4df8870a4
|
|
ImsServiceProxy is tightly coupled with the
ImsManager and should be in the same namespace.
Test: Manual
Change-Id: I1383fb114b3c23d2f3d80fa4abdb4b1091c7f879
|
|
Modifies ImsManager to use new ImsServiceProxy
class hierarchy.
Test: Manual
Change-Id: I0be48ff1b17c53933aabf0561957dee08f3205b0
|
|
-- Add isServiceReady to check if ImsService is ready to
send requests to lower layers or not.
Test: Manual
Bug: 63920021
Change-Id: I97e4af1e6eb428956d5f343a4628aba6857f37b8
|
|
Currently, if the user requests merge, and the call they request the merge
on is in the midst of some other pending operation (e.g. hold, resume), the
system will prevent the merge from taking place. The logic, however does
not account for the fact that the background call being merged into the
foreground call could also be engaged in some other update operation.
Added code to ensure that both calls are not actively engaged in some
pending operation.
Test: Manual
Change-Id: Icc2f3786865345435bbf138b54736fc6f09c0aa4
Merged-In: I5bb050b406bff514fa6132c395c2a4f6574b572b
Fixes: 63764631
|
|
am: 236b6f319e
Change-Id: Ia93cb15f9fdf3eac4edaebc9241bcb74c25ba1b1
|
|
|
|
am: 479f52cab3
Change-Id: Ic74b483dc520e0ff62cf2464a8fe5253069f8680
|
|
Adding "ForSlot" to ImsManager instance methods
was redundant and does not yield a clean API.
This change removes "ForSlot" from the methods.
Bug: 65063634
Test: Manual
Merged-In: Ida63b5c375df7af67bd3f7c5763d04f11fce92ea
Change-Id: Ida63b5c375df7af67bd3f7c5763d04f11fce92ea
|
|
am: 57b1189a54
Change-Id: I47152e032e21b24a2506712e776b69cd5d44a42d
|
|
am: 9a0280cf27 -s ours
Change-Id: Ib0812da1ac30ddc229a8dc58aa2dd6dc6e8a8319
|
|
MMTelFeature has two ways of registering registration callbacks,
through startSession and addRegistrationListener. startSession
requires a registration listener, which can cause double
registration if addRegistrationListener happens before
startSession in ImsManager.
This fix temporarily solves that problem by registering a Stub
implementation of the registration callback in startSession.
This will be fixed later by modifying the MMTelFeature
APIs to remove the registration listener from the arg list.
Bug: 62207372
Test: Telephony unit tests and manual IMS registration
Merged-In: Ieeff249e05777bdb9135e9ddd6e08a15416d5a1f
Change-Id: I00fdbae1baaf77cca5156a442ea109f998a8d5a5
|
|
Listeners were not being unregistered from
the ImsService. The listeners were restructured
to instead be held in ImsManager, instead of the
ImsService.
Bug: 38304804
Test: Run Rogers VoWiFi activation
Merged-In: I2a0d4b2ed7d232834be7c74c318a60fa11e64389
Change-Id: I2a0d4b2ed7d232834be7c74c318a60fa11e64389
|
|
When video is disabled (due to mobile data being off), we will automatically
reject any incoming requests to upgrade to video.
Test: Manual
Bug: 36900451
Change-Id: I9991cc9940aef417ec7f05cb61e777bf0d354c93
Merged-In: Ie32bc79c6d6443039e90f294b490a5535cc92d31
|
|
am: 7512e74110
Change-Id: Id0f5c20dd860c0178b3feb5f4109ab078e68e9c6
|
|
am: 3835d9d5f2
Change-Id: I3217fbfcf96d03be69dc0b5bf6b0f9a340f33fb8
|
|
If the request being sent to the video provider is to resume the video,
but the video stream is already resumed, do not send this request to the
modem.
In the case of swapping between calls, the video appears to resume
automatically on the modem-side, so sending another resume request is
redundant.
Also, in VideoPauseTracker, correcting the case where we get a resume
request, but there are no remaining pause requests in the tracker.
Although that case shouldn't run up in reality, if it did we should still
let the resume request pass along (since it would otherwise be re-written
as a pause).
Test: Manual test.
Change-Id: Ib9b9acaf2d92b1485e4766a13701fd472d6c117d
Fixes: 63606238
|
|
ImsVideoCallProviderWrapper was modified recently to check for the case
where the video is paused, and the video state changes to an unpaused state.
It used this to trigger a clear of the video pause tracker.
It did this by checking if the video pause tracker thought the video was
paused. In reality, it makes more sense for a change in the videostate
from paused to unpaused to be detected, and THAT used as a basis to
clear the video pause tracker.
The previous functionality caused a problem when the user, in quick
succession:
1. Turned off the camera
2. Went to the background.
The two requests would hit the framework, causing the video pause
tracker to record the current video state as paused (as requested by
the user). Shortly thereafter, the modem reports the fact that the
camera is off with a state of Audio RX. This would be misinterpreted
and cause the pause tracker to be cleared.
Test: Manually tested the bug, as well as regression tested b/62784036.
Change-Id: Iaa736e93e05ef3ae5dec21cd8ebc18be464f18ae
Fixes: 63410964
|
|
am: 7649108831
Change-Id: I91458d5d26a5e919ed322e84c3241fecb41417fd
|
|
If user powered up with WFC editable SIM card and modify the
WFC preference, and rebooted the device with a non editable WFC
SIM card, then user will not have a choice to pickup the default
WFC mode.
To fix this issue, update default WFC mode during power up,
if WFC UI option is non editable for inserted SIM card.
Test: manual - Checked that WFC mode was changed to default with a non
editable WFC SIM card
Bug: 37621367
Change-Id: Ib47022e23fafc5ea1b4424605577cf11ed0a1b6b
|
|
am: be6cb67bd1
Change-Id: I6d84fc804617206a49f6db77103e0c467bd64282
|
|
ImsService initialization can take longer than
telephoyn initialization in some cases, which causes
ImsConfig queries to return with an error. This change
adds an exponential backoff for querying provisioning
results from ImsConfig if we get an error during
initialization.
Bug: 64540800
Test: Manual
Merged-In: Id92ae35af954fbc0427fabb46ec0e119e4d3a4eb
Change-Id: I2193e42736994de3cce6fb24ee1dcc9e936bebf0
|
|
am: b3d249ba99
Change-Id: I0a13d9215bcb6068fa83337833ffd9bcbe17f657
|
|
am: 4801a7a0a9
Change-Id: Id290b10c204301bd0eeabda53759f8148e144db0
|
|
am: e511a2031e
Change-Id: I46d1438bf8170418901b17933895d02ab6b88906
|
|
Make changes to only toggle IMS on/off if the actual TTY option
is on (i.e. UI preference is enabled and TTY accessory is plugged in)
Test: manual
Change-Id: Ieac0ca041a0bd026e8b2f07ebcadc9a6a1db75b9
Fixes: 63968802
|
|
am: fb38165516 -s ours
Change-Id: If843b1613311073cb87b6d30ea8180bdee9aaba2
|
|
am: c3d1880ed7 -s ours
Change-Id: I737c400c9dda11a94d8916d5fbeb02c5677546ea
|
|
am: a8f627deea -s ours
Change-Id: I21f67eafc9098eadb2e5a0119b1ad25c937941e0
|
|
1. Replace onImsConnected() with onImsConnected(int)
2. Replace onImsProgressing() with onImsProgressing(int)
3. Add registrationChangeFailed()
Bug: 33430556
Test: m
Merged-In: I14c171494bdcb0d82ed4be0caff8cb52d35421cc
Change-Id: I14c171494bdcb0d82ed4be0caff8cb52d35421cc
|
|
4c0a963a11 -s ours am: feb4c40b2f -s ours
am: bf46fafe6b -s ours
Change-Id: Icd614f00d976be163e918134392383d04991e1fd
|
|
4c0a963a11 -s ours
am: feb4c40b2f -s ours
Change-Id: I8bcbf272cf87bab70e3f253dc7392a5af3389104
|
|
am: 4c0a963a11 -s ours
Change-Id: Id0e47ff1a27f8676cbbb1d675a858a85854339fa
|
|
|
|
Scenario: A and B are in video call.
A goes to the background, pausing the video.
B attempts to turn off their camera.
The request fails to be interpreted by vendor code.
Reason: This SHOULD really be a request to go from/to:
Audio/TX/RX/Paused --> Audio/RX/Paused
However, it MUST be sent as:
Audio/TX/RX/Paused --> Audio/RX
The introduction of the VideoPauseTracker in N caused the request to be
sent in the former correct format rather than the latter incorrect format.
The VideoPauseTracker attempts to ensure that a request to pause sent by
the framework vs the incall ui are kept in sync (we use pause to disable
video on a call when the data limit is reached). In the process of ensuring
pause and resume requests were handled properly, this code was
fixing the malformed request.
Added a workaround to the code to ensure the requests remain in the same
broken format vendor code depends on.
Test: Manual, unit
Bug: 35304446
Merged-In: I9b974542234d3f567aba3f2996a815e39bc8963e
Change-Id: I9b974542234d3f567aba3f2996a815e39bc8963e
|
|
am: fa3df0bc9e
am: c41dd5079d -s ours
Change-Id: I7c4c942a78db9e3c428dc301db5612b3824614ca
|
|
am: fa3df0bc9e
Change-Id: I11551ae94cd69532a81b8c0d43f5cfbc901f28b5
|
|
am: 138b4a620d
Change-Id: Ib617c2dfe895286a817fb06675c5510d071012c5
|
|
If the ImsService goes down, the stale binder interfaces
to the old service still get returned by ImsManager. This
change triggers the ImsManager to fetch the binder interfaces
again if the cached one is dead.
Bug: 62723694
Test: Manual, test procedure outlined in bug
Merged-In: I90e3df344785a5bc3818bc9bb24a2735f068058a
Change-Id: I64c1c668602d88663220a50c2e45ae4f970ac302
|
|
am: 28552fbbf0
Change-Id: I4992a41ae4a1099a08ff6a2cef32b1200bd8a3c2
|
|
am: 567a70c96d
Change-Id: I35a50f702a8055ece40ee444a645f38515553681
|
|
If the request being sent to the video provider is to resume the video,
but the video stream is already resumed, do not send this request to the
modem.
In the case of swapping between calls, the video appears to resume
automatically on the modem-side, so sending another resume request is
redundant.
Also, in VideoPauseTracker, correcting the case where we get a resume
request, but there are no remaining pause requests in the tracker.
Although that case shouldn't run up in reality, if it did we should still
let the resume request pass along (since it would otherwise be re-written
as a pause).
Test: Manual test.
Change-Id: Ib9b9acaf2d92b1485e4766a13701fd472d6c117d
Fixes: 63606238
|
|
ImsVideoCallProviderWrapper was modified recently to check for the case
where the video is paused, and the video state changes to an unpaused state.
It used this to trigger a clear of the video pause tracker.
It did this by checking if the video pause tracker thought the video was
paused. In reality, it makes more sense for a change in the videostate
from paused to unpaused to be detected, and THAT used as a basis to
clear the video pause tracker.
The previous functionality caused a problem when the user, in quick
succession:
1. Turned off the camera
2. Went to the background.
The two requests would hit the framework, causing the video pause
tracker to record the current video state as paused (as requested by
the user). Shortly thereafter, the modem reports the fact that the
camera is off with a state of Audio RX. This would be misinterpreted
and cause the pause tracker to be cleared.
Test: Manually tested the bug, as well as regression tested b/62784036.
Change-Id: Iaa736e93e05ef3ae5dec21cd8ebc18be464f18ae
Fixes: 63410964
|
|
am: efa7be99a6
Change-Id: I99b53926105c689ac19c8805035061e106a8f0a1
|
|
In N-MR2 we made a performance optimization to not bother sending
no-op video requests to the vendor IMS layer. In concept this performance
optimization makes sense; if a higher layer tells us to change from
the current video state to the current video state, then don't send it.
It turns out that the vendor IMS layer does not interpret the delta
style video requests as they were intended.
To fix this, we'll remove the performance optization and opt to make the
contract between the framework and the vendor IMS layer more explicit in
the future to avoid this kind of problem.
This fix corrects an issue where:
A and B are in a video call.
A pauses the video
B pauses the video
A un-pauses the video
When A un-pauses the video B would also unpause, even though it should
not. As it turns out the performance optimization was resulting in B
not issuing the pause request to the modem.
Test: Manual test
Change-Id: Ia29471b3f13d5b104affcf634e1315fb7c71451b
Fixes: 62954065
|
|
am: 06f97e8f98
Change-Id: Icc519c271c14330862c65659f2b8fb3cc2354115
|
|
oc-dr1-dev
am: 06869d6fd8
Change-Id: I90e3df344785a5bc3818bc9bb24a2735f068058a
|
|
|