va_trace.c : Fixes incorrect hardcode printfs message.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
(cherry picked from commit 70b80c0dd2effb4956b208775641f7c68a67a9df)

va_trace.c: add vp9 decoder trace info
(cherry picked from commit c9ebdf432f154691464359caa7f37f8a2034f0ff)

Conflicts:

va/va_trace.c

Signed-off-by: Xu Guangxin <Guangxin.Xu@intel.com>
[Support VAProfileVP9Profile0 only on master]
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 319b0fcf6c1d7faf7d2cad9baf1a45d2a8f20880)

va_trace.c: fixes incorrect use of signed decimal conversion in printfs.

Signed-off-by: Jonathan Bian <jonathan.bian@intle.com>
(cherry picked from commit 27812e7144d59603bd1c2d6dfc32dbcd0a37bc7a)
(cherry picked from commit 1c47a86c36733d6b7bb4ea9df58309d6d1fd0f18)

Va_trace changes to dump the HEVC encode parameters
(cherry picked from commit c24b5da1a8296ca13aceb233b32eaf8bc1d47802)

Conflicts:

va/va_trace.c
(cherry picked from commit 762aea574b596fcb5b8b1d335e9010fc343d9acc)

vatrace: print frame_type for VP8 encoding

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit 1a011ce5bb0b80506797a25a988854f3f81ce909)
(cherry picked from commit 739d3ed0c15f93fe7404d3925d0ce2598effeb20)

vatrace: dump the right parameters for VP8 encode

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit b7494802d94bf8af3577fbb5145f144892283585)

Conflicts:

va/va_trace.c
(cherry picked from commit 92a9f37aaf241cb3ac585284dfa97834a7e13987)

Enhance LIBVA_TRACE to dump SPS/PPS/ for VP8 encode.

Enhance LIBVA_TRACE tool to dump VP8 encode related parameters based on the new libva interface.

Change-Id: I2c731761bb180061e18dac6b63d77572ca1abef6
Signed-off-by: Liu Bolun <bolun.liu@intel.com>
(cherry picked from commit bc49c42bcbfb24286c06b3e7cdb171b2a01c6f80)

Conflicts:

va/va_trace.c
(cherry picked from commit f4a05455a763e7918c23e64218ba42caab3db065)

libva 1.6.1.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 6a5785c7581ad89c219c477236e347bf0dbd31ed)

vatrace: add IVF header for VP8 coded buffer dump

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>

va_trace.c : Fix VAIQMatrixBufferHEVC tracing

Add VP9 profile to vainfo

Add VP9 profile to vainfo so that it returns the
profile correctly for supported platforms.

Signed-off-by: Sameer Kibey <sameer.kibey@intel.com>

va_trace.c : Fixes incorrect hardcode printfs message.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>

va_trace.c: add vp9 decoder trace info
(cherry picked from commit c9ebdf432f154691464359caa7f37f8a2034f0ff)

Conflicts:

va/va_trace.c

Signed-off-by: Xu Guangxin <Guangxin.Xu@intel.com>
[Support VAProfileVP9Profile0 only on master]
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

va_trace.c: fixes incorrect use of signed decimal conversion in printfs.

Signed-off-by: Jonathan Bian <jonathan.bian@intle.com>
(cherry picked from commit 27812e7144d59603bd1c2d6dfc32dbcd0a37bc7a)

Va_trace changes to dump the HEVC encode parameters
(cherry picked from commit c24b5da1a8296ca13aceb233b32eaf8bc1d47802)

Conflicts:

va/va_trace.c

vatrace: print frame_type for VP8 encoding

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit 1a011ce5bb0b80506797a25a988854f3f81ce909)

vatrace: dump the right parameters for VP8 encode

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit b7494802d94bf8af3577fbb5145f144892283585)

Conflicts:

va/va_trace.c

Enhance LIBVA_TRACE to dump SPS/PPS/ for VP8 encode.

Enhance LIBVA_TRACE tool to dump VP8 encode related parameters based on the new libva interface.

Change-Id: I2c731761bb180061e18dac6b63d77572ca1abef6
Signed-off-by: Liu Bolun <bolun.liu@intel.com>
(cherry picked from commit bc49c42bcbfb24286c06b3e7cdb171b2a01c6f80)

Conflicts:

va/va_trace.c

libva 1.6.1.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Merge branch 'v1.6-branch' into master

libva 1.6.0

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

libva 1.6.0 pre2

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

test/decode/tinyjpeg: make sure the pointer is valid before dereferencing it

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 8455834161bab3374fe9756fd4a28d919027daf7)

test/decode/tinyjpeg: use the fourcc format Y800 for grayscale images

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Beauchesne, Gwenole <gwenole.beauchesne@intel.com>
(cherry picked from commit 62085e11d30db590038a32fffeb04aa8bc0afd67)

test/decode/tinyjpeg: create VA surface with right render target format

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Beauchesne, Gwenole <gwenole.beauchesne@intel.com>
(cherry picked from commit 54f686f4a371bf34007ce1ff70e7b74cd9986ad7)

Enhanced JPEG VA test app

Added support for YUV422, YUV444 and YUV420 and multi-scan images.

Changes
1. Added new functions for finding the start of image and end of image.
   This will be called while parsing the JIFF. This helps to detect
   the next image. The check for the bogus image format is removed as it
   is no longer valid with these new functions.
2. Since different images in the file has different width and height.
   The height and width will be stored in an array. These array variables
   will be passed while creating surfaces and contexts.
3. Added code to determine the format of the image from the h_sampling factor
   and v_sampling factor. This helps to support
   different image formats.

(cherry picked from commit ef53340b19746589079d7ed5f9c67970fcc40401)

test/decode/tinyjpeg: make sure the pointer is valid before dereferencing it

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 8455834161bab3374fe9756fd4a28d919027daf7)

test/decode/tinyjpeg: use the fourcc format Y800 for grayscale images

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Beauchesne, Gwenole <gwenole.beauchesne@intel.com>
(cherry picked from commit 62085e11d30db590038a32fffeb04aa8bc0afd67)

test/decode/tinyjpeg: create VA surface with right render target format

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Beauchesne, Gwenole <gwenole.beauchesne@intel.com>
(cherry picked from commit 54f686f4a371bf34007ce1ff70e7b74cd9986ad7)

Enhanced JPEG VA test app

Added support for YUV422, YUV444 and YUV420 and multi-scan images.

Changes
1. Added new functions for finding the start of image and end of image.
   This will be called while parsing the JIFF. This helps to detect
   the next image. The check for the bogus image format is removed as it
   is no longer valid with these new functions.
2. Since different images in the file has different width and height.
   The height and width will be stored in an array. These array variables
   will be passed while creating surfaces and contexts.
3. Added code to determine the format of the image from the h_sampling factor
   and v_sampling factor. This helps to support
   different image formats.

(cherry picked from commit ef53340b19746589079d7ed5f9c67970fcc40401)

Update NEWS

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Update NEWS

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 05a1e4bc98ce02d2d87f0cabea7d575183639dd7)

va_enc_hevc.h: Fix the description for slice_segment_address

The slice_segment_address field indicates the starting CTU address
for the corresponding slice.

Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>

va_dec_hevc.h: Fix the misleading documentation

The slice data buffer should always include the start code prefix.

The comment section of VASliceParameterBufferHEVC is saying that the
accompanying slice_data_buffer should include start code prefix.
But the comment above slice_data_byte_offset is contradicting.

Signed-off-by: Sreerenj Balachandran <sreerenj.balachandran@intel.com>

Bump VA API to 0.38

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Add one API to allow the user to specify the loaded backend driver

Sometimes there exist multiple backend vaapi drivers on one system. In such case
user hopes to load the specific driver based on the request. Currently it can
be specified by using LIBVA_DRIVER_NAME environment. But it is not convenient.
So one API is added to allow the user to specify the loaded backend vaapi driver.
And it is called by using the following sequence:
   >vaGetDisplay(
   >vaSetDriverName(va_dpy, driver_name)
   >vaInitialize

V1->V2: Use the reserved space to store override_driver_name based on Haihao's
suggestion.

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>

HEVC: type delta_chroma_log2_weight_denom as int8_t to match the spec

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 612fc3824158278336e6eb36d9af8028b839a569)

Add VAConfigAttribEncSkipFrame and corresponding structures to support sending skip frame parameters to the encoder's rate control, when the user has externally skipped frames.

(cherry picked from commit 959f462709c9c4cba3243805031cf8d836a940ed)

Conflicts:
va/va.h

Bump VA API to 0.37.1

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Merge VP9 decode API from staging branch

Cherry-pick'ed from:
960352a: VP9 decode API
4b34eb6: Minor changes to the VP9 decode API to remove fields that are not needed
29fa9ec: Made VASliceParameterBufferVP9 consistent with other codecs
0302393: Rename symbols/fields in VP9 API (part of 0302393)

Signed-off-by: Wang, Ce <ce.wang@intel.com>
Signed-off-by: Jonathan Bian <jonathan.bian@intel.com>
Signed-off-by: Sun, Maggie <maggie.sun@intel.com>

libva:Add the bits mask of color-sapce conversion standards

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit e3f83e524264e073a9815c245a9f194a1651cff6)

Conflicts:
va/va.h

Add support for querying slice decoding mode supported by the underlying driver

This is a part of commit d65d7157da6dcfab44e639bee4905f793d869a1c

Added normalization and scaling of the quality factor and Quantization tables

When the header is packed before sending to the driver through LibVA,
It is neccesary to normalize the quality factor and scale the Quantization
tables based on this quality factor.

HEVC encode API

Signed-off-by: Wang, Ce <ce.wang@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 9d128cf241830f7ead49fa5b104248b44c680873)

Conflicts:
va/Android.mk
va/Makefile.am
va/va.h

Include sub core APIs in va.h as well

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 012838e07d99531c67e3798235b00437f5b67c02)

Move VAHuffmanTableBufferJPEGBaseline to va.h

VAHuffmanTableBufferJPEGBaseline is also used for JPEG encoding,
so move it to va.h as we discussed before.

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 504478e0bbc27cc529c96b168cae1fb9dd12c4a5)

Conflicts:
va/va.h

libva 1.5.2.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

x11: dri2: fix double Unlocks/SyncHandle

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>

Change comment for num_token_partionions

Signed-off-by: Zhong Li <zhong.li@intel.com>
(cherry picked from commit d9626b77f865ce7bdbc8c78988a86d3e4f55d67b)

libva 1.5.1.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 9ff92931c684f5b069433af18269d0f7cedbbe3c)

libva 1.5.0

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 2dcbf9d11b9d4bedf3a6a58b4562aab76a4a120c)

Conflicts:
configure.ac

Update NEWS

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit b1a1edcf1fcb4403aaac56e673c4040d65ca24b0)

vainfo: Add support for HEVC profiles

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit aa562960b5b3554ba2af065d77bd84a3a54c1f08)

Conflicts:
test/vainfo/vainfo.c

(cherry picked from commit b97c3d7dff67fe0a025e79c03c398d394bf55635)

test/encode/jpegenc: Add jpegenc_utils.h to EXTRA_DIST

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 8249dc912bf316f9a556b2851e0f1b5a90a48c97)

libva 1.5.0.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 2fb37255e7c9ccb76e65ea74466f5ffafe9c68c0)

Conflicts:
configure.ac

JPEG Encode: Added support for UYVY/YUY2/Y8 formats

Signed-off-by: Sirisha Muppavarapu <sirisha.muppavarapu@intel.com>
(cherry picked from commit 33cf544c2c381adeb4b119af1015b8e653c8b724)

Add VA trace support for HEVC

BZ:189819

Change-Id: I505a9e638191494f7f69c9ef9995fabc9d82dffb
Signed-off-by: Tang, XiaoxiX <xiaoxix.tang@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Added HEVC decode API

Signed-off-by: Wang, Ce <ce.wang@intel.com>
Signed-off-by: Jonathan Bian <jonathan.bian@intel.com>
Signed-off-by: Penne <penne.y.lee@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

x11: check x pixmap is still alive before DRI2DestroyDrawable

Calling vaTerminate() after the pixmaps, given to vaPutSurface(), have
been destroyed results in errors coming from the X server.

These errors happens because LibVA tries to destroy the DRI2Drawables
related to the XPixmaps passed to vaPutSurface(). VA_DRI2Error() has
some code to catch these errors and ignore them. The problem is that
LibX11 doesn't trigger the extension error handlers unless you're
waiting for a response using _XReply(), and there is no defined answer
to a DRI2DestroyDrawable message in the DRI2 protocol.

The attached patch tries to work around that issue by checking whether
the XPixmap is still alive before proceeding with DRI2DestroyDrawable.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>

Fix while loop

Otherwise i may be 32 after enter the loop and access
dri_state->drawable_hash[32].

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Encoding/avcenc: Optimize the allocated VA surfaces to reduce the memory pressure

When it needs to encode one new frame, it will try to reuse the previously
used surface for the reconstructed frame. In such case it is helpfl to
reduce the memory pressure under multi-channel 4K encoding.

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>

JPEG Encode: Added Unit Test for testing jpeg encode feature.

(cherry picked from commit 9d01b4468f89391892e31efc5bbec774ea781242)

vatrace: trace JPEG encode parameters

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit e468728f9e49603f91ccbce3a232aa3bd98d1fd4)

Conflicts:
va/va_trace.c
va/va_trace.h

Extended existing JPEG encoding data structures and added configuration attribute for JPEG encoding.

It is the combination of 1d93d44 and 970064cf

Signed-off-by: Jonathan Bian <jonathan.bian@intel.com>
Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
[Resovle conflicts & remove irrelevant attributes]
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

libva 1.4.2.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Conflicts:
configure.ac

va: hide most private va_Trace* symbols

Hide all but va_TracePutSurface symbols from the user.
The latter one is used internally by va_x11 and va_android.

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

va_x11: hide private symbols

The following are not part of the API thus should be hidden.
 - dri2Close
 - VA_FGLRXGetClientDriverName
 - VA_NVCTRLQueryDirectRenderingCapable
 - VA_NVCTRLGetClientDriverName

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

glx: Hide private symbol va_glx_init_context

There is a note below the symbol ATTRIBUTE_HIDDEN,
additionally the header providing it is not installed.

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

debian: distclean the *install files

MAINTCLEAN wraps around DISTCLEAN, so we can drop the files from the
former.

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

configure: use top_builddir for wayland_scanner_rules

Autotools already copies the files appropriately, and using
top_srcdir causes the files to be generated in the source
tree rather than the build one.

Spotted by 'make dist'

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

Fix out of tree builds

Prefix the libva.sym file or it will not be found when doing out of
tree builds.

Spotted by 'make distcheck'.

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

.gitignore: add mpeg2vaenc

Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>

Bump libva to libva 1.4.1.pre1

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

libva 1.4.0

Don't support DRM Render-Nodes in this version of libva

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 8c48c3b44a56d8a5af5c5fa7cb8bee131f933569)

NEWS: updates.

Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>

doc: fix scan order specification for quantization tables.

Fix the VA-API specification to define all quantization tables, resp.
matrices, to be supplied in zig-zag scan order, except for the H.264
codec where scaling lists shall be provided in raster scan order.

This clarifies the existing usage inherited from the primary VA driver
implementation (Intel Poulsbo), and subsequently adopted by the major
multimedia frameworks.

Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>

doc: add Encoding, VPP and some of Core APIs.

Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>

doc: improve Doxyfile configuration.

doc: don't add timestamps in generated documentation.

tests: try a render-node device first for VA/DRM displays.

Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
Reviewed-by: Sean V Kelley <sean.v.kelley@intel.com>

drm: add support for render nodes.

Allow vaGetDisplayDRM() to accept DRM Render-Nodes file descriptors,
thus skipping authentication burdens in that case. This also allows
remote users, connected through ssh for instance, to enjoy hardware
acceleration even if there is a local display opened (user logged in).

Signed-off-by: Andrey Larionov <andrey.larionov@intel.com>
[checked if fd is a render-node, fixed VA_DISPLAY_DRM_RENDERNODES]
Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
Reviewed-by: Sean V Kelley <sean.v.kelley@intel.com>
Tested-by: Rafal A Sapala <rafal.a.sapala@intel.com>

Bump VA API to 0.36

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Test/encode/avcenc: Follow the H264 spec to fix wrong SEI parameter setting for CBR HRD

V1->V2: Follow Haihao's comments to add some comments and remove some
magic number.

Signed-off-by: Zhao, Yakui <yakui.zhao@intel.com>
(cherry picked from commit 51901e340cc467a2d562491121e0de38f5a066fe)

test/encode/avcenc: Follow the idea in h264encode to generalize avcenc test case

The following is generalized:
a. the switch between encoding order and displaying order.
b. Reference list0/1
c. It can insert more IDR frames intead of one IDR frame.

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit 615c2a0f3210ec567fde5b0bac60c3a17cff3aed)

Conflicts:
test/encode/avcenc.c

test/encode/avcenc: update the SPS parameter for encoding

It will simplify the calculation of POC and frame_num.

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit d794d040b8f7311d7b19d62826fbd3725d3bc7b9)

test/avcenc: Fix the incorrect parameter setting for SPS under CBR mode

Currently the parameter setting is incorrect for CBR, which causes that
it will complain the underflow/overflow during encoding.

V1->V2: Follow Haihao's comment to update some incorrect usage about the
frame_bit_rate.

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit 5e10a246287fa2dd3fae2bf756a5b33087cc097c)

test/avcenc: Remove the dynamic shrink of coded buffer

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit 296e5a548b260fcca5713bbb3ed6477f63ac977f)

vatrace: fix segfault of VASurfaceAttribExternalBufferDescriptor dumpping

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit e286c3264ebf7582792ff56537dd14c9f49903cf)

vafool: fix the vaMapBuffer for the buffer from vaDeriveImage

It is not desired to run into FOOL path for the vaMapBuffer
from vaDeriveImage

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit 2896166fefa36772ea6946ebc5df85024928dd6c)

vatrace: fix a overwrite

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit 7f1b4be6623f5fe65063581252d2b68bd77fd076)

vafool: simplify VAFOOL

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit ce00603a8d58f2e25c7a6bb312f60fda0833e042)

vatrace: make LIBVA_TRACE_SURFACE work even without LIBVA_TRACE

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit e39ebdd95062995e4bdf082d50bd7d9005a3be6c)

vatrace: make standalone LIBVA_TRACE_CODEDBUF work w/o LIBVA_TRACE

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
(cherry picked from commit 22925d9b7d1730468fc7435bee96d9d9c39f5987)

Enable 64bit build for Android, compatible with 32bit.

Refined the Android.mk to compatible with both 32 and 64 bit android
build.

Signed-off-by: Tinghong Li <tinghong.li@intel.com>
Signed-off-by: Qiming Shi <qiming.shi@intel.com>
(cherry picked from commit 03de01dce6f2b88b9c5ee54705aeebe17c6c5919)

API: add support for buffer exports.

Add interfaces for low-level buffer exports to suport interop with
external APIs like EGL or OpenCL (OCL). Theory of operations:

- vaAcquireBufferHandle(): locks buffer for external API usage. This
  is a synchronization point. Upon successful return, useful buffer
  details are returned in a somewhat implementation-specific way to
  support interop with external APIs.

- vaReleaseBufferHandle(): unlocks buffer after usage from external
  API, and deallocates any resources that were needed.

Signed-off-by: Gwenole Beauchesne <gwenole.beauchesne@intel.com>
(cherry picked from commit 544d31d2fc096a838c21f4684cec05746cfdd7b6)

Conflicts:
va/va.h

Update comment for quality_level

Allow user set VAEncMiscParameterBufferQualityLevel::quality_level
to 0 to use default quality level

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit b05e38f64c5aa2d6edfc2bd1d6e1f4339a54095a)

Conflicts:
va/va.h

Added VAConfigEncQualityRange attribute and VAEncMiscParameterBufferQualityLevel to support encoder quality settings.

(cherry picked from commit 24a14f9d13e6fe0c25d4094c4884aa4076ed05bc)

Conflicts:
va/va.h

Added OTC hybrid UMD driver to the list of supported drivers

This will be used for hybrid codecs like VP8 h-encode. No change
to existing KMD interface.

Signed-off-by: Sameer Kibey <sameer.kibey@intel.com>
(cherry picked from commit 98669760e62f50ca3e38e99835692aee5f89a4bf)

avcenc: only support chroma_format_idc with 1

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

h264encode: check driver supports the RateControl control mode or not

https://bugs.freedesktop.org/show_bug.cgi?id=76720

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>