Per feedback from dmail@, rephrased to clarify that the intent is
fired at the time when a USB device/accessory is attached, not a
general "a USB device happens to be attached now" notification.
Staged to:
http://go/dac-stage/reference/android/hardware/usb/UsbManager.html
Bug: 36672976
Test: make ds-docs
Change-Id: I9f1c202d0a6a6a5b25cf775e2b11119a26625bf1
Part of the redesign for P, designed to support multiple clients and
easier usage.
Bug: 67734082
Test: make from root
Change-Id: I1e12a0484350fabfc13af32269961ad0fc69f9b1
Per bug, these were mislabeled as broadcast actions, when they're
actually activity intents. This has consequences post-O, since
broadcast actions are now subject to limitations.
Doc is staged to:
http://go/dac-stage/reference/android/hardware/usb/UsbManager.html
Bug: 36672976
Test: make ds-docs
Change-Id: I2b515b1ac152de389f952af21166ed2ba7843044
display need to be destroyed once the display is removed.
Then apply this flag to VR virtual display.
Change-Id: Icb5aa1280b6ebac5941afda1cc1b94b258ce44f4
Bug: 67420451
Test: Manually. "adb shell am stack list" does not show contents
from virtual display anymore after it's released.
Fix following types of issue
* over 100 char/line
* missing space
* trailing spaces
* array bracket position
* variable naming convention (some cases)
* indentation
* long expression wrapping operator position
* missing curly brace for one line scope
* function modifier order
Things not changed in this CL:
* Redundant public modifiers.
* A few local variable names.
* Suggested @deprecate but function is actually @removed.
Test: m and m docs
Change-Id: I5c22648888487edaa5954026a166cfd810a8a912
Java doc was left out when addressing API reviewer comments.
This CL fix the discrepancy between the actual logic and java doc.
Bug: 36550285
Test: compiles
Change-Id: I6406892ecdcc5d02f11966fa3fb0b81ed8d3b285
Merged-In: I6406892ecdcc5d02f11966fa3fb0b81ed8d3b285
- Since mHandle (UID) was not recorded in writeToParcel, it always
returned 0 as the UID. This resulted in error messages claiming
mismatches between the reported and actual UID of nanoapps when called
from an external process.
- Also fixes the Parcel constructor to actually read the mContexthubId
field.
Bug: 37785458
Test: Compile and load, verify error message doesn't appear
Change-Id: I2ff75f510aae549c741786f5746862256ea5a877
Explictly use Locale.US since EXIF/TIFF date/time fields require it.
Test: Switch to Arabic locale, verify no crash and resulting DNGs have valid metadata,
plus CTS DngCreatorTest passes.
Bug: 65748590
Change-Id: I729ee6d849f7a39e5d396cfd8dd5d44325381822
The deprecated camera API throws a lot of RuntimeExceptions, and many of the
methods don't mention this.
Test: Docs build and manual review
Bug: 64725721
Change-Id: I1c6dc2f984d0be8b0d4b5394bb0d7473dfec7949
This adds FINGERPRINT_ERROR_USER_CANCELED to the list of possible errors
returned by FingerprintManager.authenticate().
Fixes b/64540509
Test: documentation builds
Change-Id: I66b46d385ab10eb0bc8e982c06e7a60570a79165
SurfaceView created for on-screen preview sometimes loses
transformation information in the Camera2 API with HALv1
environment, and therefore produced wrongly transformed frames.
It's because in RequestThreadManager.configureOutputs(), the
transformation set by LegacyCameraDevice.setSurfaceOrientation()
is cleared at mGLThreadManager.setConfigurationAndWait().
The bug is fixed by setting transformation information again
after mGLThreadManager.setConfigurationAndWait().
BUG=63313861
TEST=Check on minnie with GCA and CameraIQTesting app. Preview always
displays correct transformation in Camera2<=>HALv1 environment.
Change-Id: Ic395047a120a21b3968dedb4314bdbe519c932e6
onRequestQueueEmpty callback needs to be supported for legacy HAL1
devices.
Test: Camera CTS
Bug: 64483624
Change-Id: I268a1fe7577788c4e4e13dfecd7bb7abe0d82e6c
