| 742f098783e5ecd6370dcf9ef7648e82f64181bb |
|
14-Jan-2014 |
Shuzhen Wang <shuzhenw@codeaurora.org> |
Camera3: Clean up ALOGE calls in non-error conditions [DO NOT MERGE]
Use ALOGI/ALOGD instead.
Bug: 12489802
Change-Id: Id7fc9147395af3a9a0fe135d1017a155aad174fb
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 50906e4f6b6c8058ae8d2c59eccf142918621904 |
|
09-Oct-2013 |
Shuzhen Wang <shuzhenw@codeaurora.org> |
Camera3: Add poll fd in QBUF as asynchronous call
Without this change, there is deadlock when both Data_notify and
Qbuf are coming at same time i.e, data_notify is waiting for buf_lock
mutex which acquired by qbuf and qbuf-> Add poll fd is waiting for
data_notify to finish.
Adding Poll fd in asynchronous way i.e, POLL FD need not to wait for
data_notify cb's to finish.
Bug: 10998587
Change-Id: I574b25378639f5bc10b753af06b6d729261ee11d
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| b16c077cdee23becd2ec0c353b8edde1368a8001 |
|
17-Sep-2013 |
Mansoor Aftab <maftab@codeaurora.org> |
Camera3: Stop polling when no buffers are queued to the kernel
Bug: 10577807
Change-Id: I69fbb4139107ee21a834343fe039657ab298d769
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 6a225c824c49ff9ecdc4b67c802f1af272569e03 |
|
06-Sep-2013 |
Mansoor Aftab <maftab@codeaurora.org> |
Camera3: Name HAL threads
Name the HAL threads to make debugging easier
Bug: 10609699
Change-Id: Ic953947eca7e4fde7c87bd7c890d1811cc3ff2fb
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 025ab8ca7a6d9228b3c8cc04a4f1741defe526ed |
|
29-Jul-2013 |
Nagesh Subba Reddy <nageshsreddy@codeaurora.org> |
Camera3: Use different stream types for different streams
Add a different stream type for app callback streams, as well as
distinguish between zsl snapshot stream and non-zsl snapshot stream
by using different stream types.
Change-Id: I0d68ad61e1a3409bd9555af90895da95a8c9628d
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 7ee407aa1024789089aad411c372c79423699976 |
|
30-Jul-2013 |
Ajay Dudani <adudani@codeaurora.org> |
Camera: Initial commit for reprocess
1) Create a new channel for reprocessing
2) Make changes in postprocessor to enable reprocessing
3) Pass metadata buffer to reprocess
Enable reprocessing path for WNR
Change-Id: I0afd20be5ed087f04eabc4324a413659d9b0e7f1
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 013ea05a4d0b11d7e76e0a2a791d38441f2a3336 |
|
04-Jun-2013 |
Mansoor Aftab <maftab@codeaurora.org> |
Camera3: Disable debug logs
Bug: 9001388
Change-Id: I66ded6bd95cf5da1a22f8e2c93f0813b1aef6dd5
Signed-off-by: Iliyan Malchev <malchev@google.com>
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| b2b7907666ba95f2aa4aa19ad8533b265fdc7d0e |
|
17-May-2013 |
Shuzhen Wang <shuzhenw@codeaurora.org> |
Camera3: Fix JPEG encoding for different orientation
For A family chips, hardware encoder needs input buffer's height to
be multiple of 16. And to better support rotation, there needs to be
16x padding both on top and at the bottom.
A family hardware JPEG encoder supports rotation, so still capture
rotation is done by encoder instead of by other hardware component
(such as C2D).
Change-Id: Ia6c85f2379c9c4e17d2c7819533f96ed243373ee
Signed-off-by: Iliyan Malchev <malchev@google.com>
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 0a686661965fcbbfabf2d5e3c2768c9582828226 |
|
07-May-2013 |
Mansoor Aftab <maftab@codeaurora.org> |
Camera3: Fix Corruption for Snapshot
-Add cache cleaninvalidate call for Jpeg output buffer
-Remove additional Scanline as result of height padding
Change-Id: If075718f38fcf0615b8c57842a9e75cb63f08439
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 744f5403bebb0e27140d16a5df7a341713c24a13 |
|
05-May-2013 |
Nagesh Subba Reddy <nageshsreddy@codeaurora.org> |
Camera3: Update HAL code from AU 221 to AU281
Change-Id: Ie24a113648b00c2069db4e79c805d96df977bfd7
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| d5d5a90863708ac1cdbef25c41853b6f6f05d4e2 |
|
05-May-2013 |
Shuzhen Wang <shuzhenw@codeaurora.org> |
Camera3: Code clean up and bug fix
- Stop all streams before removing any of them. This is due to
hardware limitation.
- Remove compilation warnings.
- Eliminate log flooding.
- Remove conditional compiled-out code.
Change-Id: I606bf495f76c0cc67ebceed699b371ab939d73a4
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 6c6d887243f26c81543d33c2f8aa169862324b40 |
|
24-Apr-2013 |
Shuzhen Wang <shuzhenw@codeaurora.org> |
Camera2: Prepare for Camera HAL3 support
- Move utility classes out of HAL folder into util folder.
- Remove redundant code from mm_camera_stream.
- Add new definitions for cam_intf.h and cam_types.h for HAL3 support.
Change-Id: Ia336aa35f9d45167d2b45d3b58606e3d05a1746c
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 4b5791669581534e633ee7ac4b37a5215437f4f4 |
|
28-Mar-2013 |
Muhua Li <muhual@codeaurora.org> |
QCamera2: add two more generic 8-bit raw format.
Add two more generic 8-bit raw formats.
Change-Id: I35e9c4bdef4c6138609a3abe3b3f675cdaac9b06
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 634d42519c54c3effadde23cfe03ca1e03888a96 |
|
27-Mar-2013 |
Emilian Peev <epeevs@codeaurora.org> |
QCamera2: Invalidates buffer caches before queue op.
- The cache invalidate call needs to go right before
we en-queue the buffers. This way buffers that get
registered for the first time will get invalidated
as well.
- The change contains small code style fixes as well.
CRs-Fixed: 466986
Change-Id: I7bc61f178b11514bb6fafe8ff32a08c2c1e97acc
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 7fde32d818742141a39fda194e66b84109182aff |
|
25-Mar-2013 |
Emilian Peev <epeevs@codeaurora.org> |
QCamera2: Adds support for NV16 snapshots
- This change adds support for image capture using
NV16 pixelformat. This pixelformat will be added
to the supported list of picture formats if the
capabilities include it and can be configured by
the camera client.
Change-Id: Ica7859c00601ea32f2a96691b9502744fa2e70ca
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| a1724bc599bec5b2fbe3f4a34d0eca2406ba4c5f |
|
20-Mar-2013 |
Ivan Evlogiev <ivanevlogiev@codeaurora.org> |
QCamera2: Trigger buffer cache clean invalidate just after dequeuing
- This change moves the buffer cache clean invalidation
right after the buffer is dequeued from the camera.
This is done in order to avoid any race conditions
that could occur when buffers are processed in
different contexts at the same time.
Change-Id: I590e89bb4175afb57cb873631b210de6b7da48c6
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|
| 6f83d735d8e3b918da42e6b559fcd0efb78133e5 |
|
03-Apr-2013 |
Iliyan Malchev <malchev@google.com> |
qcom/camera: new camera HAL implementation
This HAL is taken from the following project:
remote: git://codeaurora.org/platform/hardware/qcom/camera.git
SHA: 72226d4d1cda595b55ded993571de114c8ec9893
In addition, the makefiles have been patched so that it builds within AOSP.
Change-Id: I87125fd134838f41c75a0027da98440f63cfedb5
Signed-off-by: Iliyan Malchev <malchev@google.com>
/hardware/qcom/camera/QCamera2/stack/mm-camera-interface/src/mm_camera_stream.c
|