Our application does H.264 encoding of RGB4 (RGB32) frames. It does this by having a VPP stage before encoding that converts RGB4 to the NV12 that is required as the input to the encoding stage. The VPP stage also does deinterlacing if the input video is interlaced.

It has always worked fine, but now we have a report from one particular client that it fails on both of their computers which are:

1. micro PC/NUC manufactured by Intel "Intel Skull Canyon".

Graphics Devices:
        Name                                         Version             State
        Intel(R) Iris(R) Pro Graphics 580            23.20.16.4944       Active

System info:
        CPU:    Intel(R) Core(TM) i7-6770HQ CPU @ 2.60GHz
        OS:     Microsoft Windows 10 Pro
        Arch:   64-bit
 

2. Intel NUC

Graphics Devices:
        Name                                         Version             State
        Intel(R) Iris(R) Graphics 540                23.20.16.4982       08
        Microsoft Basic Display Adapter              10.0.16299.15       Active

System info:
        CPU:    Intel(R) Core(TM) i5-6260U CPU @ 1.80GHz
        CPU:    Intel(R) Core(TM) i5-6260U CPU @ 1.80GHz
        OS:     Microsoft Windows 10 Pro
        Arch:   64-bit

If it is fed with NV12 instead of RGB4 so that it does not have to do color conversion then encoding succeeds on both computers.

The deinterlacing works OK, so it seems to be the color conversion that is causing it to fail.

Are there some platforms that are not capable of doing RGB4 to NV12 conversion ?

I compile ffmpeg with qsv supporting (MEDIASDK 2018R1) and try run:

ffmpeg -i sintel_trailer-1080p.mp4 -vcodec h264_qsv -init_hw_device qsv:hw out_qsv.mp4 Twice and more time from root user and is works 

But when i run second process from any other user i have an error 

libva info: VA-API version 1.0.0
open lib: No such file or directory
libva info: va_getDriverName() returns 1
libva info: User requested driver 'iHD'
libva info: Trying to open /opt/intel/mediasdk/lib64/iHD_drv_video.so
libva info: Found init function __vaDriverInit_1_0
get chip id failed: -1 [13]
param: 4, val: 0
libva error: /opt/intel/mediasdk/lib64/iHD_drv_video.so init failed
libva info: va_openDriver() returns 18          

What does it mean ? How to fix ? 

CentOS Linux release 7.4.1708 (Core)

Linux localhost.localdomain 3.10.0-693.21.1.el7.x86_64 #1 SMP 

libva info: VA-API version 1.0.0
libva info: va_getDriverName() returns 0
libva info: User requested driver 'iHD'
libva info: Trying to open /opt/intel/mediasdk/lib64/iHD_drv_video.so
libva info: Found init function __vaDriverInit_1_0
libva info: va_openDriver() returns 0
vainfo: VA-API version: 1.0 (libva )
vainfo: Driver version: 16.8.69021-ubit
vainfo: Supported profile and entrypoints
      VAProfileNone                   : VAEntrypointVideoProc
      VAProfileNone                   : <unknown entrypoint>
      VAProfileMPEG2Simple            : VAEntrypointVLD
      VAProfileMPEG2Simple            : VAEntrypointEncSlice
      VAProfileMPEG2Main              : VAEntrypointVLD
      VAProfileMPEG2Main              : VAEntrypointEncSlice
      VAProfileH264Main               : VAEntrypointVLD
      VAProfileH264Main               : VAEntrypointEncSlice
      VAProfileH264Main               : <unknown entrypoint>
      VAProfileH264Main               : VAEntrypointEncSliceLP
      VAProfileH264High               : VAEntrypointVLD
      VAProfileH264High               : VAEntrypointEncSlice
      VAProfileH264High               : <unknown entrypoint>
      VAProfileH264High               : VAEntrypointEncSliceLP
      VAProfileVC1Simple              : VAEntrypointVLD
      VAProfileVC1Main                : VAEntrypointVLD
      VAProfileVC1Advanced            : VAEntrypointVLD
      VAProfileJPEGBaseline           : VAEntrypointVLD
      VAProfileJPEGBaseline           : VAEntrypointEncPicture
      VAProfileH264ConstrainedBaseline: VAEntrypointVLD
      VAProfileH264ConstrainedBaseline: VAEntrypointEncSlice
      VAProfileH264ConstrainedBaseline: <unknown entrypoint>
      VAProfileH264ConstrainedBaseline: VAEntrypointEncSliceLP
      VAProfileVP8Version0_3          : VAEntrypointVLD
      VAProfileHEVCMain               : VAEntrypointVLD
      VAProfileHEVCMain               : VAEntrypointEncSlice
      VAProfileHEVCMain               : <unknown entrypoint> 


export PATH=/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/bin/
export LIBVA_DRIVERS_PATH=/opt/intel/mediasdk/lib64
export LIBVA_DRIVER_NAME=iHD
export MFX_HOME=/opt/intel/mediasdk

Hi, i'm trying to encode 10 bit HEVC with Intel GPU Accelerated Media SDK plugin, here my command line options

sample_multi_transcode -i::h264 "C:\WORKING (HEVC)\13\T1_Video - English.264" -o::h265 "C:\WORKING (HEVC)\LOSTS3E13.hevc" -hw -async 10 -pe e5400a06c74d41f5b12d430bbaa23d0b -b 5000 -u balanced -l 1 -vbr -gop_size 240 -dist 4 -num_ref 6 -bref -bpyr -MaxKbps 12000 -ec::p010 -dc::nv12 -w 1920 -h 1080 -WeightedBiPred::default -WeightedPred::default -amfs:on -extbrc:on -pp -sys -n 1000
Multi Transcoding Sample Version 8.3.26.352

External allocator will be used as some cmd line paremeters request it.
Session 0:
Preset-controlled parameters (Default):
RateControlMethod: 2 (original preset value: 2)
AdaptiveMaxFrameSize: 16 (original preset value: 0)
AsyncDepth: 10 (original preset value: 4)
BRefType: 2 (original preset value: 0)
EnableBPyramid: 1 (original preset value: 0)
GopRefDist: 4 (original preset value: 0)
IntRefCycleDist: 0
IntRefCycleSize: 0
IntRefQPDelta: 0
IntRefType: 0
LowDelayBRC: 0
TargetUsage: 4 (original preset value: 4)
WeightedBiPred: 1 (original preset value: 0)
WeightedPred: 1 (original preset value: 0)
GopPicSize: 240 (original preset value: 47)
MaxKbps: 12000 (original preset value: 0)
TargetKbps: 5000 (original preset value: 2748)
BufferSizeInKB: 5000
MaxFrameSize: 0
LookAheadDepth: 0

plugin_loader.h :173 [INFO] Plugin was loaded from GUID: { 0xe5, 0x40, 0x0a, 0x06, 0xc7, 0x4d, 0x41, 0xf5, 0xb1, 0x2d, 0x43, 0x0b, 0xba, 0xa2, 0x3d, 0x0b } (Intel (R) Media SDK GPU-Accelerated plugin for HEVC ENCODE)
Pipeline surfaces number (EncPool): 19
Pipeline surfaces number (DecPool): 10
Sample BRC is used
MFX HARDWARE Session 0 API ver 1.25 parameters:
Input  video: AVC
Output video: HEVC
MFX dll: C:\Program Files\Intel\Media SDK\libmfxhw64.dll

Session 0 was NOT joined with other sessions

Transcoding started
.........
Transcoding finished

Common transcoding time is 53.0178 sec
-------------------------------------------------------------------------------
*** session 0 PASSED (MFX_ERR_NONE) 53.0172 sec, 1000 frames
-i::h264 C:\WORKING (HEVC)\13\T1_Video - English.264 -o::h265 C:\WORKING (HEVC)\LOSTS3E13.hevc -hw -async 10 -pe e5400a06c74d41f5b12d430bbaa23d0b -b 5000 -u balanced -l 1 -vbr -gop_size 240 -dist 4 -num_ref 6 -bref -bpyr -MaxKbps 12000 -ec::p010 -dc::nv12 -w 1920 -h 1080 -WeightedBiPred::default -WeightedPred::default -amfs:on -extbrc:on -pp -sys -n 1000

-------------------------------------------------------------------------------

The test PASSED
plugin_loader.h :199 [INFO] MFXBaseUSER_UnLoad(session=0x0000011FF83C6920), sts=0

I have included a screenshot of the final hevc movie picture

System Specs:

Intel Kabylake 7700 with Intel HD 630 Graphics 

Windows 10 x64

Intel Media SDK 2018R1

Intel 23.20.16.4982 drivers

I have inherited a project, which is using IPP 8.1 static library.  I need to make my project support windows one core.  So I downloaded w_ipp_2018.2.185.exe, but it doesn't have static libraries after I install it.  How can I get the latest stable build of IPP static library that is compatible with windows one core?  Thanks!

Hello,

could you please help to identify an issue when running FFMPEG with QSV under Ubuntu 17.10, Media Server Studio 2017?

I try to encode raw video using H264 codec using a command:

ffmpeg -y -i sample_1080p.y4m -c:v h264_qsv -preset fast -look_ahead 0 output_1080p.264

The h264_qsv codec prints a warning "Encoder will work with partial HW acceleration". The CPU load is pretty high and reaches about 70%. Meantime the performance is lower than expected and it's about 70fps. Adding some constraints like maximum bitrate and buffer can drop it to 20fps. I observe similar behaviour on both HD530 and Iris Pro 6200 iGPUs (i7-6700K and i7-5775C CPUs respectively).

If drop "-look_ahead 0" parameter, the codec prints error: "Selected ratecontrol mode is not supported by the QSV runtime. Choose a different mode.".

Checking the quality shows a PSNR(Y) difference may reach several dB comparing to the same parameters for libx264. Which I believe is not expected. Looks like I use QSV in some wrong way.

Could you guide me to the correct direction? What can a root cause be for such behaviour?

Thank you!

Hello, i'm trying to encode to HEVC GPU ACC with lookahead (la_ext ) plugin included wth Intel Meda SDK 2018 R1 and in release notes sayd

this about the HEVC GPU ACC

Version 1.15.3.62   MAIN 10 and MAIN 10 422 range extension profile support added.

 Look Ahead feature implementation added. 

     Added support for Extended Look Ahead plugin, featuring analyses for before encode or VPP session, and reuse analyses data for several encoding sessions in multi-session scenario, so significantly decreases memory utilization and improves performance. 

Extended look Ahead plugin added API support for B-pyramid analyses. 

TU4 support added featuring middle range encode quality/performance optimization.  

For general performance vs bitrates savings please check HEVC Whitepaper on Product web page.

How can i trascode with la_ext enabled? Can show me some command line options example for sample_multi_transcode?

I Always get this error

sample_multi_transcode -i::h264 "D:\LOST (WORKING)\1\T1_Video - English.h264" -o::h265 D:\LST01HLT1.hevc -hw_d3d11 -async 10 -pe e5400a06c74d41f5b12d430bbaa23d0b -b 2890 -u quality -l 1 -la_ext -lad 100 -vbr -gop_size 240 -dist 4 -num_ref 6 -bref -bpyr -MaxKbps 8000 -ec::nv12 -dc::nv12 -w 1280 -h 720 -WeightedBiPred::implicit -WeightedPred::implicit -amfs:on -extbrc:on -pp
Multi Transcoding Sample Version 8.3.26.352

External allocator will be used as some cmd line paremeters request it.
Session 0:
Preset-controlled parameters (Default):
RateControlMethod: 2 (original preset value: 2)
AdaptiveMaxFrameSize: 16 (original preset value: 0)
AsyncDepth: 10 (original preset value: 4)
BRefType: 2 (original preset value: 0)
EnableBPyramid: 1 (original preset value: 0)
GopRefDist: 4 (original preset value: 0)
IntRefCycleDist: 0
IntRefCycleSize: 0
IntRefQPDelta: 0
IntRefType: 0
LowDelayBRC: 0
TargetUsage: 4 (original preset value: 4)
WeightedBiPred: 3 (original preset value: 0)
WeightedPred: 3 (original preset value: 0)
GopPicSize: 240 (original preset value: 47)
MaxKbps: 8000 (original preset value: 0)
TargetKbps: 2890 (original preset value: 2420)
BufferSizeInKB: 2890
MaxFrameSize: 0
LookAheadDepth: 100 (original preset value: 0)

plugin_loader.h :173 [INFO] Plugin was loaded from GUID: { 0x58, 0x8f, 0x11, 0x85, 0xd4, 0x7b, 0x42, 0x96, 0x8d, 0xea, 0x37, 0x7b, 0xb5, 0xd0, 0xdc, 0xb4 } (Intel (R) Media SDK plugin for LA ENC)

[ERROR], sts=MFX_ERR_UNSUPPORTED(-3), TranscodingSample::CTranscodingPipeline::EncodePreInit, m_pmfxENC->Query failed at src\pipeline_transcode.cpp:495

[ERROR], sts=MFX_ERR_UNSUPPORTED(-3), TranscodingSample::CTranscodingPipeline::Init, EncodePreInit failed at src\pipeline_transcode.cpp:3615

[ERROR], sts=MFX_ERR_UNSUPPORTED(-3), TranscodingSample::Launcher::Init, pThreadPipeline->pPipeline->Init failed at src\sample_multi_transcode.cpp:353
plugin_loader.h :199 [INFO] MFXBaseUSER_UnLoad(session=0x0000028ECE2127D0), sts=0

[ERROR], sts=MFX_ERR_UNSUPPORTED(-3), wmain, transcode.Init failed at src\sample_multi_transcode.cpp:91

How to use la_ext plugin in par file multi session?

I'm using Kabylake processor i7 7700

Regards

When following the steps as laid out in the HEVC Interlaced whitepaper(included in the 2018R1 release) to use the CBR SW bit rate controller(MFX_RATECONTROL_CBR), initially everything seems to work fine.

However, after somewhere between 8-16 hours of continuous real-time encoding it seems that the bitrate controller goes crazy and the encoded video drops. In our case we started encoding a 1080i stream in about 6 Mbps but after the 8 hours it looks rather like the encoding is done at a very, very low bitrate i.e. it's completely unwatchable. The resulting elementary stream does however keep the correct overall bit rate but the encoding looks looks bad. 

We have tested using the same rate controller on pure progressive material and are getting the exact same result.

Has anyone seen this behvaiour?
 

I'm trying to decode the following interlaced video using sample_decode.exe. The video looks terrible  (blocks of video missing etc), but it looks fine when using a different decoder (NVIDIA for example). If I look at the mfxFrameData information, many of the frames are marked as corrupt. Any ideas? Is this a problem with the decoder, or is the information I'm giving it formatted in such a way that is unsupported?

Thanks,

Erin

Hi!

I am trying to use IPP functionality to resize Yuv image using ippiResizeYUV420Super_8u_P2R API.

I did use ippiResizeYUV420GetSize, allocated memory for pSpec and pBuffer as per output of this API using malloc,  called ippiResizeYUV420SuperInit followed by ippiResizeYUV420GetBufferSize.

Everything is fine so far and all of these APIs return status 0, but when I actually use ippiResizeYUV420Super_8u_P2R, my application crashes at I9_ownSS1_8u.

Any advice? Also, I am not sure what  should be my destination offset in ippiResizeYUV420Super_8u_P2R and how to calculate that.

Please note that I am using this API for downsizing my image.

Thanks,

Pratibha

The documentation states that video memory is faster than system memory for encoding, but it isn't, at least in Win10 + Kaby Lake + DX9. If the frame is in system memory in YV12 format, it's faster to copy-convert it to NV12 into a system surface and send that to the encoder than to convert it to NV12 to a video surface.

An encoder doesn't actually re-read old frames like the decoder does. If a frame is a reference to another, all comparisons are against the lossy coded version of the reference frame, not the original, so it makes sense not to sweat where the original frame is. What am I missing?

Bruno

Since my feature update to Windows 10, version 1803, I can't run Visual Studio 2017 anymore. After showing splash screen it aborts silently.
My event logger reports the following error:

Application: devenv.exe
Framework Version: v4.0.30319
Description: The process was terminated due to an unhandled exception.
Exception Info: System.AccessViolationException
   at Intel.CodeBuilderSessionManager.UnsafeNativeMethods.KBAPI_Init(System.String)
   at Intel.CodeBuilderSessionManager.KBAPIDataService..ctor()
   at Intel.CodeBuilderSessionManager.KBAPIDataService.get_Instance()
   at Intel.CodeBuilderSessionManager.CodeBuilderSessionManagerPackage.Initialize()
   at Microsoft.VisualStudio.Shell.Package.Microsoft.VisualStudio.Shell.Interop.IVsPackage.SetSite(Microsoft.VisualStudio.OLE.Interop.IServiceProvider)

So I tried to uninstall Intel SDK OpenCL and after uninstalling the SDK VS 2017 worked well again. I tried to reinstall the newest version of the OpenCL SDK (Version 7.0.0.2567), but end up in the same situation: VS 2017 doesn't start anymore.

I am trying to use the VPP component to just do a color space conversion from UYVY to NV12, no de-interlacing or scaling, or other filters.  When I use the exact same input parameters with MFX_FOURCC_YUY2 everything works.  When I change mfxVideoParam.vpp.In.FourCC to MFX_FOURCC_UYVY, MFXVideoVPP::QueryIOSurf returns MFX_ERR_INVALID_VIDEO_PARAM.  I am using API level 1.25.

Are there any additional restrictions on converting from UYVY to NV12 that I am missing and may be causing this?

Hi All,

I have code which is decodes camera live frames. On giving first frame it return MFX_ERR_MORE_DATA . 

Then i give next frame and it return continuously MFX_ERR_NOT_INITIALIZED error.

On Fist frame I have initilize my decoder and video params.

Please help me to fix this.

Hi,  need to know how to set in sample_multi_transcode and sample_encode general_progressive_source_flag to 1, can someone help me out?

Regards

Dear all,

sorry for the questions, probably for most of you will be easy. I hope to not bother you to much.

I have just installed intel for linux: ifort, icc, mkl ans so on.

I know that I have to change my bash file in order to be able to use them.

How can I do it? I have just read some guide but I am not able to find it.

Really, Really Thanks

Hi ,

I have decoded my camera  live stream using intel media sdk and its give format NV12. but my renderer application needs RGB565 Format. also I m using hardware acceleration for decoding.

After little research i found we can not convert NV12 to RGB565 directly using VPP. 

Questions : 

1. Can Intel Media SDK decode directly into RGB565 ?

2. If not then how can I convert NV12 to RGB565 ? 

https://downloadmirror.intel.com/27803/eng/ReleaseNotes_24.20.100.6094.pdf​ mentions
"Improved video encoding/decoding in Handbrake*, VLC*, and other similar applications".
What changed? All codecs?

Hi, i'm need to encode interlaced yuv source to interlaced hevc CQP, but it gives erros, doesnt matter what type software or hardware here the command line options

D:\Intel Media SDK Plugins>sample_encode h265 -i C:\ffmpeg\Teste.yuv -o D:\WTRS1E1.hevc -tff -w 704 -h 528 -bref -idr_interval 1 -f 50 -cqp -qpi 20 -qpp 22 -qpb 24 -u balanced -r 4 -x 6 -g 50 -hw -p e5400a06c74d41f5b12d430bbaa23d0b -gpb:off -num_slice 1 -pp
Preset-controlled parameters (Default):
AdaptiveMaxFrameSize: 0
AsyncDepth: 4
BRefType: 2 (original preset value: 0)
GopPicSize: 50 (original preset value: 100)
GopRefDist: 4 (original preset value: 0)
IntRefCycleDist: 0
IntRefCycleSize: 0
IntRefQPDelta: 0
IntRefType: 0
LookAheadDepth: 0
LowDelayBRC: 0
MaxFrameSize: 0
RateControlMethod: 3 (original preset value: 2)
TargetUsage: 4 (original preset value: 4)
WeightedBiPred: 0
WeightedPred: 0

plugin_loader.h :173 [INFO] Plugin was loaded from GUID: { 0xe5, 0x40, 0x0a, 0x06, 0xc7, 0x4d, 0x41, 0xf5, 0xb1, 0x2d, 0x43, 0x0b, 0xba, 0xa2, 0x3d, 0x0b } (Intel (R) Media SDK GPU-Accelerated plugin for HEVC ENCODE)
Encoding Sample Version 8.3.26.0

Input file format       YUV420
Output video            HEVC
Source picture:
        Resolution      704x544
        Crop X,Y,W,H    0,0,704,528
Destination picture:
        Resolution      704x272
        Crop X,Y,W,H    0,0,704,264
Frame rate      50.00
QPI     20
QPP     22
QPB     24
Gop size        50
Ref dist        4
Ref number      6
Idr Interval    1
Target usage    balanced
Memory type     system
Media SDK impl          hw
Media SDK version       1.26

Processing started

[ERROR], sts=MFX_ERR_MORE_SURFACE(-11), CEncodingPipeline::Run, m_pmfxENC->EncodeFrameAsync failed at c:\users\thx4ever\documents\intel« media sdk 2018 r1 - media samples 8.3.26.352\sample_encode\src\pipeline_encode.cpp:1892

[ERROR], sts=MFX_ERR_MORE_SURFACE(-11), wmain, pPipeline->Run failed at c:\users\thx4ever\documents\intel« media sdk 2018 r1 - media samples 8.3.26.352\sample_encode\src\sample_encode.cpp:1308
Frame number: 0
Encoding fps: -0
plugin_loader.h :199 [INFO] MFXBaseUSER_UnLoad(session=0x0000023BB23B1380), sts=0

I use  intel media sdk 2017 to realize hardware encoding function,it works well on win7 64 bit computer,but cant work on win7 32 bit. the two. According the requirement,it only has the intel graphic card.

     the 32bit cpu is intel i3 4th processor,the 64 bit cpu is intel i3 6th processor.

so i don't know  sdk surpport 32bit hardware encoding?if surpport,what shoud i do to solve this problem?thanks.

Hello?

In a Linux environment, video transcoding is performed using multi-threading in an account with a video group.
The first thread is processing video transcoding 'va_getDriverName() failed with operation failed, driver_name = i965' when the second thread calls the CreateVAAPIDevice() function
However, when calling the CreateAVAPIDevice() function when executing the process with the root account, the iHD_drv_video.so library is called and processed normally

First thread..

libva info: VA-API version 0.35.0
libva info: va_getDriverName() returns 0
libva info: User requested driver 'iHD'
libva info: Trying to open /opt/intel/mediasdk/lib64/iHD_drv_video.so
libva info: Found init function __vaDriverInit_0_32
libva info: va_openDriver() returns 0

Second thread..

libva info: VA-API version 0.35.0
libva info: va_getDriverName() returns 1
libva error: va_getDriverName() failed with operation failed,driver_name=i965
terminate called after throwing an instance of 'std::bad_alloc'
  what():  std::bad_alloc

I want to run the program from a user other than the root account
Should a user in the video group call the CreateAVAPIDevice () function only once when the application is running?

Thanks in advance.

My Linux environment..

$ ll /opt/intel/mediasdk/lib64

drwxr-xr-x  2 root root       76  4월 28  2015 ./
drwxr-xr-x 10 root root     4096  5월 11 01:47 ../
-rwxr-xr-x  1 root root 25162960  2월 17  2015 iHD_drv_video.so*
-rwxr-xr-x  1 root root 10605896  2월 17  2015 libmfxhw64-p.so.1.14*
lrwxrwxrwx  1 root root       20  4월 28  2015 libmfxhw64.so -> libmfxhw64-p.so.1.14*

$ id
uid=201(k3) gid=200(informix) groups=200(informix),39(video)

$ ll /dev/dri

drwxr-xr-x  2 root root       100  5월 28 02:40 ./
drwxr-xr-x 19 root root      3220  5월 28 02:40 ../
crw-rw-rw-  1 root video 226,   0  5월 28 02:40 card0
crw-rw-rw-  1 root video 226,  64  5월 28 02:40 controlD64
crw-rw-rw-  1 root video 226, 128  5월 28 02:40 renderD128

$ vainfo
error: can't connect to X server!
libva info: VA-API version 0.35.0
libva info: va_getDriverName() returns 0
libva info: User requested driver 'iHD'
libva info: Trying to open /opt/intel/mediasdk/lib64/iHD_drv_video.so
libva info: Found init function __vaDriverInit_0_32
libva info: va_openDriver() returns 0
vainfo: VA-API version: 0.35 (libva 1.3.1)
vainfo: Driver version: 16.4.32088-ubit
vainfo: Supported profile and entrypoints
      VAProfileNone                   : VAEntrypointVideoProc
      VAProfileNone                   : <unknown entrypoint>
      VAProfileMPEG2Simple            : VAEntrypointEncSlice
      VAProfileMPEG2Simple            : VAEntrypointVLD
      VAProfileMPEG2Main              : VAEntrypointEncSlice
      VAProfileMPEG2Main              : VAEntrypointVLD
      VAProfileMPEG4Simple            : VAEntrypointVLD
      VAProfileMPEG4AdvancedSimple    : VAEntrypointVLD
      VAProfileMPEG4Main              : VAEntrypointVLD
      VAProfileH264Baseline           : VAEntrypointEncSlice
      VAProfileH264Baseline           : <unknown entrypoint>
      VAProfileH264Baseline           : <unknown entrypoint>
      VAProfileH264Main               : VAEntrypointVLD
      VAProfileH264Main               : VAEntrypointEncSlice
      VAProfileH264Main               : <unknown entrypoint>
      VAProfileH264Main               : <unknown entrypoint>
      VAProfileH264High               : VAEntrypointVLD
      VAProfileH264High               : VAEntrypointEncSlice
      VAProfileH264High               : <unknown entrypoint>
      VAProfileH264High               : <unknown entrypoint>
      VAProfileVC1Simple              : VAEntrypointVLD
      VAProfileVC1Main                : VAEntrypointVLD
      VAProfileVC1Advanced            : VAEntrypointVLD
      VAProfileJPEGBaseline           : VAEntrypointVLD
      VAProfileJPEGBaseline           : VAEntrypointEncPicture
      VAProfileH264ConstrainedBaseline: VAEntrypointVLD
      VAProfileH264ConstrainedBaseline: VAEntrypointEncSlice
      VAProfileH264ConstrainedBaseline: <unknown entrypoint>
      VAProfileH264ConstrainedBaseline: <unknown entrypoint>
      VAProfileVP8Version0_3          : VAEntrypointEncSlice
      VAProfileVP8Version0_3          : VAEntrypointVLD
      VAProfileVP8Version0_3          : <unknown entrypoint>

$ env

LIBVA_DRIVERS_PATH=/opt/intel/mediasdk/lib64
LIBVA_DRIVER_NAME=iHD
MFX_HOME=/opt/intel/mediasdk