| Commit message (Collapse) | Author | Age | Files | Lines |
|\
| |
| |
| |
| |
| |
| | |
switching resolutions while preview on" into jb-mr2-dev
* commit 'b7337d56a72c9b20abf57e853953126d0aa3dd1a':
camera3: Fix recording failure when switching resolutions while preview on
|
| |\
| | |
| | |
| | | |
preview on" into jb-mr2-dev
|
| | |
| | |
| | |
| | |
| | | |
Bug: 9069574
Change-Id: Ife2ae5467f799dbe37e66714b3a6c9a51e1fe328
|
|\ \ \
| |/ /
| | |
| | |
| | |
| | |
| | | |
than frame timestamp
* commit '1d60db8bfce9becee9d101eacc58e2a673221622':
Camera3: Fix Bug minTimestamp is never greater than frame timestamp
|
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Bug in logic since minTimestamp is initialized to -1, and this needs
to be updated for the first frame atleast for logic to work correctly
for subsequent frames.
Bug: 9007356
Change-Id: I70edc4d9f34ff67a141c027c67396dc2ca9e413a
Signed-off-by: Igor Murashkin <iam@google.com>
|
|\ \
| |/
| |
| |
| | |
* commit '722e4892e2afeb0c7abab84be1caa243d1692e28':
Camera2/3: Increase digital zoom steps
|
| |
| |
| |
| |
| |
| |
| |
| | |
Digitial zoom number of steps was 30 and a bit rough. Change it to 100 to make
smooth zoom smoother.
Bug: 9050908
Change-Id: I141bc0335d58cdcd6530c14cbbfe6c4ffb4c6dc6
|
|\ \
| |/
| |
| |
| |
| |
| | |
into jb-mr2-dev
* commit '074eb387091ca464257601a2d23020e3e10a6915':
update to Fence::waitForever() API change
|
| |
| |
| |
| |
| | |
Bug: 8988871
Change-Id: I24a2620d356f3bb6207ed834c79fd3e48b429695
|
| |\ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
A U by any other name may smell just as sweet, but the color's pretty
awful if you call it a V.
Henceforth they never will be UV.
Bug: 8949720
Change-Id: I4897e93d5a293e576bb1d7ba44d91792b869ad42
|
| |/
| |
| |
| |
| | |
Bug: 8969579
Change-Id: Ia51d4072725754fd3b6ca028232a605885376287
|
| |
| |
| |
| |
| | |
Bug: 8876221
Change-Id: Idf9490a48462fa44d49c4ebe99425c3149ae6378
|
| |\ |
|
| | |
| | |
| | |
| | |
| | | |
Bug: 8851039
Change-Id: Iaac2926bfa25dd6e9db8b307765d4fe709b88d21
|
| |\ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Covers cases where the HAL format and the API format have the same
chroma layout.
Bug: 8734880
Change-Id: Ia735082c260b5914fc14f12551f91917c4e53b01
|
| |\ \ \ |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Bug: 8484377
Change-Id: I5ffcc20b68dc92b502acc9898e57f12cadb92848
|
| | |/ /
| |/| |
| | | |
| | | |
| | | |
| | | |
| | | | |
One call to set orientation too many.
Bug: 8683719
Change-Id: I4b776fc8665eed940a1f18a6f617be4f1406e41a
|
| |\ \ \
| | | |/
| | |/|
| | | | |
error" into jb-mr2-dev
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug: 8622089
Change-Id: I4df0258d6803d460d4dd98f0a80829eb844ca30d
|
| |\ \ \ |
|
| | | |/
| | |/|
| | | |
| | | |
| | | | |
Bug: 8756080
Change-Id: I90b9166440e92f7675255a01ddc4a980c2e52201
|
| |\ \ \
| | |_|/
| |/| | |
|
| | |/
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When the HAL supports it, and the client asks for YV12 or NV21,
use the new flexible YUV format instead.
Bug: 8734880
Change-Id: Ib0129d9c26a6b30f3be7aa624c2439c6edba1bbd
|
| |\ \
| | |/
| |/|
| | | |
disconnect" into jb-mr2-dev
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Fixes an issue where a client could unconditionally transition to PRESENT
after a client disconnects, even though the underlying HAL status was actually
NOT_PRESENT or ENUMERATING.
Bug: 8780114
Change-Id: I68adb5fc819eec3b046ddcb2507b84bedc999a0f
|
| |\ \
| | |/
| |/|
| | | |
once" into jb-mr2-dev
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Finishing all stream configuration immediately is good, but when a stream is
both input and output it attempted to finish configuring bidi streams twice.
Since all ZSL streams are bidi, when we had a ZSL stream active preview would
immediately stop working.
Bug: 8563838
Change-Id: Iec998f11f6405fc15f3f31bd7cd29f03a7968d14
|
| |\ \
| | |/
| | |
| | | |
into jb-mr2-dev
|
| | |
| | |
| | |
| | |
| | | |
Bug: 8790212
Change-Id: I4bf7ac5ab856b0ea1738844182efb86137e14083
|
| |\ \
| | |/
| |/| |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Still FOV can vary as a function of the output sizes, with
the function depending on the HAL version.
Calculate and update the FOV whenever the output sizes might change.
Bug: 8484377
Change-Id: I56f2cc768a1e128a159b326588be55b2752db06e
|
| |\ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug: 8563838
Change-Id: Ie053655f5583d20003abdabccee68980d7cdcbfd
|
| |\ \ \
| | |/ /
| | | |
| | | |
| | | |
| | | |
| | | | |
* changes:
Camera3: Add ZSL stream support
camera: Add new RingBufferConsumer to keep a ring buffer of acquired frames
Camera3: Add input stream support
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug: 8563838
Change-Id: I2feda142ff5172aba17ade5c8d502f9bb5d5b766
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug: 8563838
Change-Id: I5a95e0be94e5388b30639905efae42d3c3279f72
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
- Untested with actual CAMERA3_STREAM_INPUT streams.
Bug: 8629088
Change-Id: Ia0c21ef0a2c951e401ea8babd15d3cceb4bb25a1
|
| | |/
| |/|
| | |
| | |
| | | |
Bug: 8765726
Change-Id: I78f5f98315068aaad3a39d56698b9c8a18cc8f4a
|
| |\ \
| | | |
| | | |
| | | | |
into jb-mr2-dev
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug: 8554573
Change-Id: Ie909908a4cab3700bd622282e8342e8fa5b72376
|
| |\ \ \
| | |/ /
| |/| /
| | |/ |
jb-mr2-dev
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
When the binder connection dies and is the only holder of a strong
pointer to the Camera2Client, disconnect is called from the
destructor. At this point, all weak pointers to Camera2Client are no
longer promotable, and lots of cleanup code paths are broken as a
result.
Rework all such code paths to not need the client pointer, and to
discard image buffers that arrive during shutdown.
Bug: 8696047
Change-Id: Ic0672ecde7c1baaf65079f925a45bd5be45f1fb3
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
While registering them later would be nice for startup time, current
hardware will have to reinitialize everything, resulting in glitches.
Bug: 8657740
Change-Id: I1ed1f0a65d648f219f5228c8df15ffcf2a4b272e
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
To simplify HAL implementation, allow the results for a single
capture request to be sent from the HAL across multiple calls
to process_capture_request.
This requires that the HAL must call notify() with the SHUTTER message
to indicate start of exposure time, before the first call to
process_capture_result for a given frame.
Bug: 8565103
Change-Id: I6a61449725f98af88769fafa07736848a226dad2
|
| |\ \ |
|
| | |/
| | |
| | |
| | |
| | |
| | |
| | | |
In preparation for supporting fragmented result calls.
Bug: 8565103
Change-Id: Id834637d7cbecf7b550333beedab3281b5400748
|
| |/
| |
| |
| |
| | |
Bug: 8580410
Change-Id: If493d87d60d71be664ad75b140c62acadb75b0d0
|