| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
| |
* Added a check to make the script does not work when there is not device
|
|
|
|
|
| |
* Correct UpdateStatusManager behavior after kOnValidUpdateReceived
* Added an initialization of flags in external proprietary mode
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
cachefile (#3820)
* sdl_read_value_navigation_and_phonecall_from_hmicapabilities_and_did_not_from_systemcapabilities
* Update src/appMain/hmi_capabilities.json
Co-authored-by: Jacob Keeler <jacob.keeler@livioradio.com>
Co-authored-by: Jacob Keeler <jacob.keeler@livioradio.com>
|
|
|
| |
fix validation of synonyms in QUERY_APPS system request
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
FinishSendingResponseToMobile (#3815)
* don't use null application ptr in RAI::CheckLanguage or FinishSendingResponseToMobile
* prevent application nullptr deref in SendRegisterAppInterfaceResponseToMobile
* check application valid in ApplicationDataShouldBeResumed
* restore ResumeCtrlImpl DCHECKs
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* do not modify applications array while iterating it
* remove unnecessary and dangerous iterator++
* fixup! do not modify applications array while iterating it
restore sending on language change to mobile
* deref app ptr while holding app list lock
* SendNotificationToMobile can also invalidate apps list iterator by calling unregister application
* apply fix to TTS and VR on language change
* Revert "Fix early IsRegistered() state of application (#3796)"
This reverts commit 0ed38e8d2f7cb3a61a303cca2806a06ce9a1d58d.
* revert #3799
* prevent crash with null app in rai response
* do not deactivate app before unregistering in case of vr language change
* make copy of ApplicationSet instead of pushing app_ids to vector
* address failure of 3797_3798 atf test in extprop policy mode
* fixup! address failure of 3797_3798 atf test in extprop policy mode
only send NO_APPS_REGISTERED when no apps are registered
* fix ext prop unit test and APPLICATION_NOT_REGISTERED log message
* fixup! fix ext prop unit test and APPLICATION_NOT_REGISTERED log message
fix style
* simplify code copying applicationset
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Fix ThreadedSocket disconnect logic
There was noticed an issue with ThreadedSocket connection
which often closes connection like an "unexpected disconnect"
calling `ConnectionAborted()` every time even if it is handled
disconnect. Also, there were no any events raised to Transport
Manager for rare cases when disconnect is expected.
To fix that, flow has been separated for handling expected and
unexpected disconnect. Expected disconnect will be handled in
case when external component calls `Disconnect()` and socket
thread is alive. In that case should be called `DisconnectDone()`
when thread is stopped.
Unexpected disconnect will be handled in case when internal thread
has been aborted and `Abort()` was called. In that case internal
thread `ConnectionAborted()` will be called from internal thread
to notify Transport Manager that physical disconnect has happened.
This issue is pretty hard to detect just because business logic
handles both disconnect types in the same way for now, but this
might be changed in the future.
* Update Boost Download URL
Co-authored-by: JackLivio <jack@livio.io>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Fix unstable policy handler unit tests
Looks like there was a missing async waiter in some
unit tests so Async thread is still working when
unit test is already finalized. That causes a core
crashes and random unit tests failures.
Was added `UnloadPolicyLibrary()` call to stop the
async thread before finalizing the unit test.
* Increase timeout for tests
|
|
|
|
|
|
|
| |
Looks like we have to properly shutdown message queue
in the unit test which emulates data sending over the
secured websocket session. Otherwise, unit test might
be terminated earlier and data sending will be failed
which causes a core crash.
|
|
|
| |
If this value is set to 1, no apps can register because SDL Core responds with UNSUPPORTED_VERSION to protocol V1 messages (besides the initial StartService)
|
|\ |
|
| |\
| | |
| | | |
Release 8.0.0
|
| | | |
|
|\ \ \
| | |/
| |/| |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* Add initial states for pending applications
This would prevent crashes for cases when app will be
attempted to be activated.
* Fix UpdateAppList content
* Fix checks in SDL.ActivateApp
* Fix unit tests
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* Removed the unit that interrupted the sequence of closing the connection to the device via usb
* Update logic for USB in transfer
* fixup! Update logic for USB in transfer
Co-authored-by: Andrii Kalinich <AKalinich@luxoft.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* Fix early IsRegistered state of application
Mark application as registered only when app registration
was finalized and all corresponding requests were sent to
HMI. This prevents data races in case when language has been
changed by HMI and application is in intermediate registration
state yet.
* fixup! Fix early IsRegistered state of application
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* Added RC plugin extension after PTU
* Add new event `kAppHmiTypesChanged`
Co-authored-by: OlhaVorobiova <olha.vorobiova@dxc.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
There was found a misprint for a VR interface in
`ChangeRegistration` request. Because of that,
there might be a situations when HMI responds to one
part of request faster than another part was actually
sent.
All required ordering and checks were updated.
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* Fix wrong behavior of RCOnRemoteControlSettingsNotification
Looks like `kAllowed` parameter was swithing on/off the RC functionality
even if it was not present and previous value should be taken.
Processing of `kAllowed` and `kAccessMode` was split on two separate
and independent functions make it really isolated.
Processing of both parameters was aligned, so now core does not perform
extra actions if one of parameter is missing.
|
| | | |
|
| | |
| | |
| | |
| | |
| | | |
* Check specifically for invalid HMI level before suspending notifications
Checking for main window ID caused a very rare race condition where the notification was not suspended when necessary
|
| | |
| | |
| | | |
Co-authored-by: OlhaVorobiova <olha.vorobiova@dxc.com>
|
| | | |
|
| | |
| | |
| | | |
Co-authored-by: OlhaVorobiova <olha.vorobiova@dxc.com>
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
* Removes 'memory_kb'
* Removes 'device_certificates'
Co-authored-by: OlhaVorobiova <olha.vorobiova@dxc.com>
|
| | |
| | |
| | | |
Co-authored-by: OlhaVorobiova <olha.vorobiova@dxc.com>
|
| | |
| | |
| | | |
Co-authored-by: OlhaVorobiova <olha.vorobiova@dxc.com>
|
| | |
| | |
| | | |
Co-authored-by: Andrii Kalinich <AKalinich@luxoft.com>
|
|/ / |
|
| |
| |
| |
| | |
(#3782)
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* Add missing internal error codes to security manager
* Add internal error notification to SSL handshake failure
* Remove unused handshake result
* Add unit tests for new error codes
Co-authored-by: JackLivio <jack@livio.io>
|
| |
| |
| |
| |
| | |
* WebSocketServerTransportAdapter::Store write to last state
* hold lock until we set dict or decide not to
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* clear any parent reference to the state being erased even when it is the begin state
* Apply suggestions from code review
Co-authored-by: Jacob Keeler <jacob.keeler@livioradio.com>
Co-authored-by: Jacob Keeler <jacob.keeler@livioradio.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* fix defect with stoping video without audio streaming
* Isolate audio and video streaming
* Remove app unregistration if EndServiceAck not received
Also, application should not be switched to NONE in case
when at least one streaming is already approved and active
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* Change command request inheritance structure.
Fix data races during requests event processing.
* Change commands according to new inheritance structure.
* Change unit tests according new inheritance structure.
* Fix review comments
* Fix build after merge
* Fix format style
* Fix recursion in SetGlobalProperties
* Add delay before request releasing
That gives to event dispatcher some time to figure
out that event should not be rasied for a finalized request
and this request will be released soon. Also, this allows
to avoid calling some virtual functions on pointer to
invalidated object.
* Fix data races in EventDispatcher
Now reference counter is increased indirectly
by event dispatcher before calling `HandleOnEvent()`
of specific observer.
This will guarantee that during `HandleOnEvent` the
command object is still be alive.
* Revert changes for SetGlobalProperties
Co-authored-by: v-seme12 <vsemenyuk@kiev.luxoft.com>
Co-authored-by: Andrii Kalinich <AKalinich@luxoft.com>
|
| |
| |
| |
| |
| | |
* Send REQUEST type and generated sequence number in SendHandshakeData
* Update query type in tests
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
* Add method for delete empty message
* Add WARNINGS code when some parameters were cut off
Co-authored-by: Jacob Keeler <jacob.keeler@livioradio.com>
|
| |
| |
| |
| |
| |
| |
| | |
* remove invalid language code es-en from preloaded pt
* remove es-en from other pt files
* remove en-ie language
|