fix buffer overflow for dc_values and ac_values (v2)

The dc_values only have 12 bytes and ac_value only 162 bytes but the
memcpy did it for 16 bytes and 256 bytes copying thru hard code value.
To avoid the array index out of bound again, recommend move to use sizeof.

v2:
Fix commit message typo from 265 bytes to 256 bytes.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
(cherry picked from commit c36778ff264b3c45b538db4bbfe6aea38fcb165e)

Remove unused 'id' in va_wayland_drm_create

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit c36971c682d890681fe839bbaa8a348fe845aa42)

Remove unused local varibles in upload_source_YUV_once_for_all.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 4ea3e2f0b921cfbed30d0cfe795939df742e5cfd)

Remove unused variable 'id' in va_wayland_emgd_create.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit fb3d2ef558907d490e9d1183bba013ca8e75af9e)

Remove unused 'k' in test func and pre func.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 979894b6812d977d2784a3c7c3091ee517473da2)

Remove unused 'i' in va_TraceVAEncPackedHeaderParameterBufferType

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit d8397f79e0ded7c8a87098996af3c603bba2d61f)

Remove unused 'trace_index' in va_TraceInit

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 98ee039437488047529746956f9d4505f86fa74f)

Remove unused 'j' in va_TraceVAPictureParameterBufferVP9.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 6bfc10804d24e4ba49c2c07f15115f3b549323bd)

Remove unused 'check_sum' and 'j' in va_TraceSurface

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 42ab32a8042691988abb469bfa36467d7618b9ad)

Remove unused 'i' and 'buf_list' in va_TraceDestroyBuffer

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>
(cherry picked from commit 12beda5f22e0d223367ffd91d40c17e6ab2b631e)

libva 1.7.2.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit aa8df4c3d10d0609ffac235ffdc76c5e05763a23)

Conflicts:
configure.ac

va_trace: add traces for vp9encoder

add trace information for VAEncPictureParameterBufferVP9
and VAEncSequenceParameterBufferVP9

also add the case for VAEncMiscParameterBufferType on
va_TraceVP9Buf

Signed-off-by: Daniel Charles <daniel.charles@intel.com>

Bump VA-API version to 0.39.3

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

Add one flag to use the qp_delta instead of qp priority for ROI under non-CQP mode

Currently the roi_value in VAEncROI is used as the qp priority when
VAConfigAttribRateControl != VA_RC_CQP. But some customers hope to use
it as qp_delta like CQP mode. In such case one flag is added to indicate
whether it is qp_priority or not. At the same time driver can export it
in VAConfigAttribValEncROI to indicate whether the qp_delta is supported
when VAConfigAttribRateControl != VA_RC_CQP.

v2: update comments in va.h

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Signed-off-by: ceciliapeng <cecilia.peng@intel.com>
Signed-off-by: Pengfei Qu <Pengfei.Qu@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Properly terminate parsed environment values with '\0'.

The function strncpy() does not guarantee to nul terminate the
destination. In most cases, this cannot be triggered, but it is also
used to parse user environment variables. These are allowed to be longer
than 1023 characters, effectively resulting in an unterminated string.

I've adjusted other places as well, because it won't hurt.

https://bugs.freedesktop.org/show_bug.cgi?id=96677

Changed 'c' variable 'char' data type to 'int' data type (v3)

The getopt_long func is return integer data type is more than 1 bytes.

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

Clean up the duplicate checking for unsuported source YUV format

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

Fix the p->load_huffman_table[0] keep print first element in for loop

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

avoid U_row overwrite NULL value if src_fourcc and dst_fourcc is NV12

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

Fix infinite loop

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

Add assert check memory allocation potential NULL issue

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

Fix memory leak issue in open_display func

If the !d->display is true, will exit without free memory
block allocate for 'd'.

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

Fix test_06 failed and hitting assert check

Add in missing profile and entrypoint get from va.h

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

Avoid access invalid memory location huffman_table for index 2..3 (v2)

The size of array huffman_table only 2 in VAHuffmanTableBufferJPEGBaseline
default_huffman_table_param. The index in 2..3 in huffman_table[x] will
be access invalid memory location in for loop that looping 4 times.

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

Fix buffer overflow for array index out of bounds in test_07

The size of array surface_1 is 2, but it passed to test_unique_surfaces
func is 16 or 6 to do compare for surface_16. The index for surface_1
not in the array index range for surface_1 and access invalid memory
location.

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

add assert check for potential NULL issue in test/encode/

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

add assert to check for potential NULL issue in test_XX

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

check memory alloc to avoid NULL and initialize value in YUV_blend_with_pic (v2)

v2:
Add in second free(NULL) for pic_y and third free(NULL) for pic_u.
Put back 'allocated' missing consider about the pic_y, pic_u and pic_v
will be contains pic_y_old, pic_u_old and pic_v_old will be assign to
pic_y, pic_u and pic_v.

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

check memory allocation and initialize to zero value in save_recyuv (v2)

v2:
Remove not necessary check dst_Y against NULL for sencond free(NULL).
Remove not necessary check dst_Y and dst_U against NULL for second free(NULL)
and third free(NULL).

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

fix buffer overflow for dc_values and ac_values (v2)

The dc_values only have 12 bytes and ac_value only 162 bytes but the
memcpy did it for 16 bytes and 256 bytes copying thru hard code value.
To avoid the array index out of bound again, recommend move to use sizeof.

v2:
Fix commit message typo from 265 bytes to 256 bytes.

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

Remove unused 'id' in va_wayland_drm_create

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused local varibles in upload_source_YUV_once_for_all.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused variable 'id' in va_wayland_emgd_create.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused 'k' in test func and pre func.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused 'i' in va_TraceVAEncPackedHeaderParameterBufferType

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused 'trace_index' in va_TraceInit

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused 'j' in va_TraceVAPictureParameterBufferVP9.

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused 'check_sum' and 'j' in va_TraceSurface

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

Remove unused 'i' and 'buf_list' in va_TraceDestroyBuffer

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

libva 1.7.1

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

libva 1.7.2.pre1 for development

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

Update NEWS

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

fix check NULL issues

v2:
code clean up

v1:
initial

Signed-off-by: peng.chen <peng.c.chen@intel.com>

Bump VA-API version to 0.39.2

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

Add API for VP9 Encode

Signed-off-by: Wang, Ce <ce.wang@intel.com>
Signed-off-by: bzhao11 <bo.zhao@intel.com>
[Haihao: include va_enc_vp9.h in va core include file]
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 1c314d6a6e6aad9c73ec4837d78e4e3c99137f3b)

Fix missing parameter of vatrace DisplayAttribute print

[49270.199330] ==========va_TraceMaxNumDisplayAttributes
[49270.199334]  max_display_attributes = 6
[49270.199362]  num_attributes = 5
[49270.199365]  attr_list[-1429804448] =
[49270.199367]    typ = 0x00000000

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

trace: Cleanup some HEVC prints

Some fields in HEVC buffers had extraneous timestamps interspersed
with the traced data.

Signed-off-by: Scott D Phillips <scott.d.phillips@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

trace: Add `va_TracePrint`, `va_TraceVPrint` static functions

These functions are similar to va_TraceMsg() but without emitting
a timestamp.

Signed-off-by: Scott D Phillips <scott.d.phillips@intel.com>
Reviewed-by: Sean V Kelley <seanvk@posteo.de>

add the support of multiple context tracing

Libva has a limited tracing function before. If more than one context is created
in one instance, then this tracing can't work and output wrong log messages.
This patch fixes this problem. Tracing is possible while multiple contexts are
being used.

Each log file name is suffixed by thread id. This file has all log message from
this thread. Each surface or codedbuf dump file name is suffixed by context id.
This file has all surface or codedbuf data of this context.

Signed-off-by: peng.chen <peng.c.chen@intel.com>
Reviewed-by: Sean V Kelley <sean.v.kelley@intel.com>
Tested-by: Daniel Charles <daniel.charles@intel.com>

Bump VA-API version to 0.39.1

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

test/encode/avcenc: Add support for low power mode

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-By: Sean V Kelley <sean.v.kelley@intel.com>

test/encode/avcenc: use getopt_long_only() to parse optional command-line options

It's convenient to add new option later

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-By: Sean V Kelley <sean.v.kelley@intel.com>

Add Macroblock based rate control method.

Signed-off-by: pylee <penne.y.lee@intel.com>
Reviewed-By: Sean V Kelley <sean.v.kelley@intel.com>

Merge ROI patches from staging branch

Cherry-pick'ed from:
26cc3b0: Add Region-of-Interest (ROI) support for encoder based on user provided ROI rectangles.
84b90a3: ROI updates for encoder.

Signed-off-by: pylee <penne.y.lee@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Reviewed-By: Sean V Kelley <sean.v.kelley@intel.com>

Add the missing entrypoints in vainfo list

It is a part of 624838b in staging branch

Signed-off-by: Austin Yuan <shengquan.yuan@intel.com>
Reviewed-By: Sean V Kelley <sean.v.kelley@intel.com>

Added VAEntrypointEncSliceLP for implementations to expose a low power/high performance variant of slice level encoding entrypoint.

Signed-off-by: Jonathan Bian <jonathan.bian@intel.com>
Reviewed-By: Sean V Kelley <sean.v.kelley@intel.com>
(cherry picked from commit 71030701cef0c1099e87d112f3207cb78a6ce711)

gcc suggests not putting ; at end of if statements

dri2_util.c: In function 'dri2Close':
dri2_util.c:171:33: warning: suggest braces around empty body in an 'if'
   statement [-Wempty-body]
     if (dri_state->base.fd >= 0);
                                 ^

Introduced-by: commit fffeffb8a78fffbb6a May 30 2011
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
Reviewed-by: Zhao Yakui <yakui.zhao@intel.com>

libva 1.7.1.pre1 for development

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

libva 1.7.0

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

Allow loading the backend driver built against VAAPI 0.38

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

Update NEWS

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

Bump VA-API to 0.39

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

Merge remote-tracking branch 'fdo/v1.6-branch' into fdo--master

glx: check OpenGL 3.1 extensions

I successfully implemented hardware decoding with VAAPI via FFmpeg
by copying the data back to the CPU's memory; however, when I tried
to use the data directly from the GPU (instead of copying them back)
by using VA/GLX to convert the decoded VASurfaces to OpenGL textures,
I ran into some issues.

With the latest version of libva, vaCreateSurfaceGLX fails all the time
when I set OpenGL to 3.1 or above. If I set it to 3.0, then it doesn't
fail, and everything works correctly. I downloaded the sources of libva
and it fails when checking for GL extensions because of the use of
glGetString(GL_EXTENSIONS) which is deprecated. I have implemented a fix
on my end (which I've attached for convenience) which works well for me
but I haven't done a lot of testing, and I would rather use the official
version of libva.

Signed-off-by: Cyril Drouet <cyril.drouet@immerex.com>
Signed-off-by: Víctor Jáquez <vjaquez@igalia.com>

driver_name_map: add nouveau and radeon for Mesa Gallium

This patch allows va_getDriverName to succeed.
It is useful in vaInitialize to select the vaapi
driver provided by Mesa Gallium:
lib/dri/gallium_drv_video.so

Signed-off-by: Julien Isorce <j.isorce@samsung.com>

add the support of enum VAProfileVP9Profile1~3 for vainfo

Signed-off-by: peng.chen <peng.c.chen@intel.com>
(cherry picked from commit 09b2f1f637390318236c11312898e3f4906c9581)

Add support for 10b vp9 decode

v2:
recover VAProfileVP9Profile1~3 from the original patch

v1：
initial

Signed-off-by: Sun, Maggie <maggie.sun@intel.com>
Signed-off-by: peng.chen <peng.c.chen@intel.com>
(cherry picked from commit 000622dad919387ac40f3ad48157d5db925c1ce9)

Avoid reaching div by zero

Signed-off-by: Lim Siew Hoon <siew.hoon.lim@intel.com>
(cherry picked from commit cb418f6ef36b6e6f0d632bf7c0aaf035a271400b)

libva 1.6.3.pre1 for development

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 6ef997f03eba8fcabfe3a2ac1546bbaf924e519f)

Conflicts:
configure.ac

add the support of enum VAProfileVP9Profile1~3 for vainfo

Signed-off-by: peng.chen <peng.c.chen@intel.com>

Add support for 10b vp9 decode

v2:
recover VAProfileVP9Profile1~3 from the original patch

v1：
initial

Signed-off-by: Sun, Maggie <maggie.sun@intel.com>
Signed-off-by: peng.chen <peng.c.chen@intel.com>

Avoid reaching div by zero

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

libva 1.6.2

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

libva 1.6.3.pre1 for development

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

libva 1.6.2.pre3

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

Fix error caused by 'make distclean' with wayland disabled

This is based on Mesa's 426a23af147720ae3b89995ffee792a29e8ae2db.

v2: cleanup va/wayland/Makefile.am and nuke va/wayland/protocol/Makefile.am (Emil)

Cc: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit d6d1e231ebd76e5ee353f460b01ceda43f13e155)

Generate wayland related files even if wayland disabled

This fixes the regression caused by commit 883ce46

Making all in debian.upstream
make[2]: Entering directory '/home/uaeoff/Work/workspace/media/build/libva/debian.upstream'
make[2]: *** No rule to make target 'libva-wayland1.install', needed by 'all-am'.  Stop.
make[2]: Leaving directory '/home/uaeoff/Work/workspace/media/build/libva/debian.upstream'
Makefile:453: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/home/uaeoff/Work/workspace/media/build/libva'
Makefile:384: recipe for target 'all' failed
make: *** [all] Error 2

Reported-by: Eoff, Ullysses A <ullysses.a.eoff@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit af199eb763afcce467a66689049070bb416aeaa4)

Fix error caused by 'make distclean' with wayland disabled

This is based on Mesa's 426a23af147720ae3b89995ffee792a29e8ae2db.

v2: cleanup va/wayland/Makefile.am and nuke va/wayland/protocol/Makefile.am (Emil)

Cc: Emil Velikov <emil.l.velikov@gmail.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

Generate wayland related files even if wayland disabled

This fixes the regression caused by commit 883ce46

Making all in debian.upstream
make[2]: Entering directory '/home/uaeoff/Work/workspace/media/build/libva/debian.upstream'
make[2]: *** No rule to make target 'libva-wayland1.install', needed by 'all-am'.  Stop.
make[2]: Leaving directory '/home/uaeoff/Work/workspace/media/build/libva/debian.upstream'
Makefile:453: recipe for target 'all-recursive' failed
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory '/home/uaeoff/Work/workspace/media/build/libva'
Makefile:384: recipe for target 'all' failed
make: *** [all] Error 2

Reported-by: Eoff, Ullysses A <ullysses.a.eoff@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>

libva 1.6.2.pre2

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

Bump VA API to 0.38.1

We added some new flags

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 46ffebbf3c1ff123c47ef915ce49a59a5f6faf6e)

Added 10-bit YUV RT format value and fourcc codes

(cherry picked from commit 52ed6a80c332fd744a39442c4053df3befade570)

Add the new YV16 fourcc code

Some input sources require YV16, which is described as the following:
Planar YUV4:2:2; 8 bit Y plane, followed by 8 bit 2x1 subsampled
V and U planes.

V1->V2: Follow Gwenole and Haihao to update the comment

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit 238b914bba8cc41076c14eb2d37938f2f705e78f)

Skip the check related with wayland if wayland is not used or found

https://bugs.freedesktop.org/show_bug.cgi?id=79478

Signed-off-by: Thomas Orgis <obukus@sourcemage.org>
Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit 883ce465933e772c08ad77e4b7f053dc8cb85fba)

Bump VA API to 0.38.1

We added some new flags

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

Added 10-bit YUV RT format value and fourcc codes

(cherry picked from commit 52ed6a80c332fd744a39442c4053df3befade570)

Add the new YV16 fourcc code

Some input sources require YV16, which is described as the following:
Planar YUV4:2:2; 8 bit Y plane, followed by 8 bit 2x1 subsampled
V and U planes.

V1->V2: Follow Gwenole and Haihao to update the comment

Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
(cherry picked from commit 238b914bba8cc41076c14eb2d37938f2f705e78f)

Revert "Added 10-bit YUV RT format value and fourcc codes"

VA_FOURCC_YV16 was added in another commit

This reverts commit 280c9511f38156ed238fa9cf7526f3a5e4d04fe9.

Skip the check related with wayland if wayland is not used or found

https://bugs.freedesktop.org/show_bug.cgi?id=79478

Signed-off-by: Thomas Orgis <obukus@sourcemage.org>
Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>

Added 10-bit YUV RT format value and fourcc codes

(cherry picked from commit 52ed6a80c332fd744a39442c4053df3befade570)

Conflicts:
va/va.h

Update NEWS

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

LIBVA_DRIVER_NAME needs to be used if set at all

va_GetDriverName fails to find the vdpau driver returning
something != SUCCESS and the driver name is not used at all.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72822
Change-Id: I29ef398bf066badedc25de10873975ad0479dccf
Signed-off-by: Philippe Coval <rzr@gna.org>

x11_screen driver context member initialization for X11 output

https://bugs.freedesktop.org/show_bug.cgi?id=61603

Signed-off-by: Alexandr <pankoAA@ya.ru>

vainfo: Add option --display

User can use 'vainfo --display help' to get available display
list and 'vainfo --display <display name>' to show available
profile/entroypoint pairs with specified name

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

va/x11: Don't crash the application if it does not have access to the DRI2 device.

Applications run in a different account than the logged in user will not have permission
to access /dev/dri/card0 but may be able to work without it, particularly if VA-API was
brought in through a general framework such as GStreamer.

Signed-off-by: Francois Gouget <fgouget@free.fr>

test/encode: remove unused variables and functions

This gets rid of compiler warnings caused by various unused
variables and functions.

Signed-off-by: U. Artie Eoff <ullysses.a.eoff@intel.com>
Signed-off-by: Sean V Kelley <seanvk@posteo.de>

avcenc: support large source file

Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
(cherry picked from commit 29ad40444f3ed68e5f54ddd70dd256ea41e0456f)

Conflicts:
test/encode/avcenc.c

libva 1.6.2.pre1 for development

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

Merge branch 'v1.6-branch' into fdo--master

Conflicts:
configure.ac

libva 1.6.1

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