VA-API
2.22.0
|
Classes | |
struct | VARectangle |
Structure to describe rectangle. More... | |
struct | VAMotionVector |
Generic motion vector data structure. More... | |
struct | VAConfigAttrib |
union | VAConfigAttribValMaxFrameSize |
Attribute value for VAConfigAttribMaxFrameSize. More... | |
union | VAConfigAttribValEncJPEG |
Attribute value for VAConfigAttribEncJPEG. More... | |
union | VAConfigAttribValEncROI |
Attribute value for VAConfigAttribEncROI. More... | |
union | VAConfigAttribValEncRateControlExt |
Attribute value for VAConfigAttribEncRateControlExt. More... | |
union | VAConfigAttribValMultipleFrame |
Attribute value for VAConfigAttribMultipleFrame. More... | |
union | VAConfigAttribValContextPriority |
union | VAConfigAttribValEncPerBlockControl |
struct | VAGenericValue |
Generic value. More... | |
struct | VASurfaceAttrib |
Surface attribute. More... | |
union | VASurfaceAttribAlignmentStruct |
VASurfaceAttribAlignmentStruct structure for the VASurfaceAttribAlignmentSize attribute. More... | |
struct | VASurfaceAttribExternalBuffers |
VASurfaceAttribExternalBuffers structure for the VASurfaceAttribExternalBufferDescriptor attribute. More... | |
struct | VAContextParameterUpdateBuffer |
update the context parameter this structure is used to update context parameters, such as priority of the context backend driver should keep the parameter unchanged if there no new parameter updated. More... | |
struct | VAEncryptionSegmentInfo |
structure for encrypted segment info. More... | |
struct | VAEncryptionParameters |
Encryption parameters buffer for VAEncryptionParameterBufferType. More... | |
struct | VAProcessingRateParameterEnc |
struct | VAProcessingRateParameterDec |
struct | VAEncPackedHeaderParameterBuffer |
Packed header parameter. More... | |
struct | VAEncMiscParameterBuffer |
struct | VAEncMiscParameterTemporalLayerStructure |
Temporal layer Structure. More... | |
struct | VAEncMiscParameterRateControl |
Rate control parameters. More... | |
struct | VAEncMiscParameterFrameRate |
struct | VAEncMiscParameterMaxSliceSize |
struct | VAEncMiscParameterHRD |
struct | VAEncMiscParameterBufferMaxFrameSize |
Defines a maximum frame size (in bits). More... | |
struct | VAEncMiscParameterBufferMultiPassFrameSize |
Maximum frame size (in bytes) settings for multiple pass. More... | |
struct | VAEncMiscParameterBufferQualityLevel |
Encoding quality level. More... | |
struct | VAEncMiscParameterQuantization |
Quantization settings for encoding. More... | |
struct | VAEncMiscParameterSkipFrame |
Encoding skip frame. More... | |
struct | VAEncROI |
Encoding region-of-interest (ROI). More... | |
struct | VAEncMiscParameterParallelRateControl |
Attribute value for VAConfigAttribEncParallelRateControl. More... | |
struct | VAEncMiscParameterEncQuality |
struct | VAEncMiscParameterCustomRoundingControl |
Custom Encoder Rounding Offset Control. Application may use this structure to set customized rounding offset parameters for quantization. Valid when VAConfigAttribCustomRoundingControl equals 1. More... | |
struct | VAHuffmanTableBufferJPEGBaseline |
Huffman table for JPEG decoding. More... | |
struct | VAIQMatrixBufferMPEG2 |
struct | VASliceParameterBufferMPEG2 |
struct | VAMacroblockParameterBufferMPEG2 |
struct | VAPictureParameterBufferMPEG4 |
struct | VAIQMatrixBufferMPEG4 |
struct | VASliceParameterBufferMPEG4 |
struct | VAPictureParameterBufferVC1 |
struct | VASliceParameterBufferVC1 |
struct | VAPictureParameterBufferH264 |
struct | VAIQMatrixBufferH264 |
struct | VASliceParameterBufferH264 |
struct | VAEncSliceParameterBuffer |
Encode Slice Parameter Buffer. More... | |
struct | VACodedBufferSegment |
Coded buffer segment. More... | |
struct | VABufferInfo |
VA buffer information. More... | |
struct | VASurfaceDecodeMBErrors |
Macros | |
#define | VA_STATUS_SUCCESS 0x00000000 |
#define | VA_STATUS_ERROR_INVALID_VALUE 0x00000019 |
An invalid/unsupported value was supplied. More... | |
#define | VA_STATUS_ERROR_UNSUPPORTED_FILTER 0x00000020 |
An unsupported filter was supplied. | |
#define | VA_STATUS_ERROR_INVALID_FILTER_CHAIN 0x00000021 |
An invalid filter chain was supplied. | |
#define | VA_STATUS_ERROR_HW_BUSY 0x00000022 |
Indicate HW busy (e.g. run multiple encoding simultaneously). | |
#define | VA_STATUS_ERROR_UNSUPPORTED_MEMORY_TYPE 0x00000024 |
An unsupported memory type was supplied. | |
#define | VA_STATUS_ERROR_NOT_ENOUGH_BUFFER 0x00000025 |
Indicate allocated buffer size is not enough for input or output. | |
#define | VA_STATUS_ERROR_TIMEDOUT 0x00000026 |
Indicate an operation isn't completed because time-out interval elapsed. | |
#define | VA_FRAME_PICTURE 0x00000000 |
#define | VA_ENABLE_BLEND 0x00000004 /* video area blend with the constant color */ |
#define | VA_CLEAR_DRAWABLE 0x00000008 |
#define | VA_SRC_COLOR_MASK 0x000000f0 |
#define | VA_FILTER_SCALING_DEFAULT 0x00000000 |
#define | VA_FILTER_INTERPOLATION_DEFAULT 0x00000000 |
#define | VA_PADDING_LOW 4 |
#define | VA_EXEC_SYNC 0x0 |
#define | VA_EXEC_ASYNC 0x1 |
#define | VA_EXEC_MODE_DEFAULT 0x0 |
#define | VA_FEATURE_NOT_SUPPORTED 0 |
#define | VA_FEATURE_SUPPORTED 1 |
#define | VA_FEATURE_REQUIRED 2 |
#define | VA_RT_FORMAT_YUV420 0x00000001 |
YUV 4:2:0 8-bit. | |
#define | VA_RT_FORMAT_YUV422 0x00000002 |
YUV 4:2:2 8-bit. | |
#define | VA_RT_FORMAT_YUV444 0x00000004 |
YUV 4:4:4 8-bit. | |
#define | VA_RT_FORMAT_YUV411 0x00000008 |
YUV 4:1:1 8-bit. | |
#define | VA_RT_FORMAT_YUV400 0x00000010 |
Greyscale 8-bit. | |
#define | VA_RT_FORMAT_YUV420_10 0x00000100 |
YUV 4:2:0 10-bit. | |
#define | VA_RT_FORMAT_YUV422_10 0x00000200 |
YUV 4:2:2 10-bit. | |
#define | VA_RT_FORMAT_YUV444_10 0x00000400 |
YUV 4:4:4 10-bit. | |
#define | VA_RT_FORMAT_YUV420_12 0x00001000 |
YUV 4:2:0 12-bit. | |
#define | VA_RT_FORMAT_YUV422_12 0x00002000 |
YUV 4:2:2 12-bit. | |
#define | VA_RT_FORMAT_YUV444_12 0x00004000 |
YUV 4:4:4 12-bit. | |
#define | VA_RT_FORMAT_RGB16 0x00010000 |
Packed RGB, 16 bits per pixel. | |
#define | VA_RT_FORMAT_RGB32 0x00020000 |
Packed RGB, 32 bits per pixel, 8 bits per colour sample. | |
#define | VA_RT_FORMAT_RGBP 0x00100000 |
Planar RGB, 8 bits per sample. | |
#define | VA_RT_FORMAT_RGB32_10 0x00200000 |
Packed RGB, 32 bits per pixel, 10 bits per colour sample. | |
#define | VA_RT_FORMAT_RGB32_10BPP VA_RT_FORMAT_RGB32_10 |
#define | VA_RT_FORMAT_YUV420_10BPP VA_RT_FORMAT_YUV420_10 |
#define | VA_ATTRIB_NOT_SUPPORTED 0x80000000 |
#define | VA_ENCRYPTION_TYPE_FULLSAMPLE_CTR 0x00000001 /* AES CTR fullsample */ |
#define | VA_SLICE_DATA_FLAG_ALL 0x00 /* whole slice is in the buffer */ |
#define | VA_CODED_BUF_STATUS_PICTURE_AVE_QP_MASK 0xff |
#define | VA_CODED_BUF_STATUS_FRAME_SIZE_OVERFLOW 0x1000 |
The frame has exceeded the maximum requested size. More... | |
#define | VA_CODED_BUF_STATUS_BAD_BITSTREAM 0x8000 |
the bitstream is bad or corrupt. | |
#define | VA_CODED_BUF_STATUS_NUMBER_PASSES_MASK 0xf000000 |
The coded buffer segment status contains frame encoding passes number. More... | |
#define | VA_CODED_BUF_STATUS_SINGLE_NALU 0x10000000 |
The coded buffer segment contains a single NAL unit. More... | |
#define | VA_MAPBUFFER_FLAG_DEFAULT 0 |
VA_MAPBUFFER_FLAG_DEFAULT is used when there are no flag specified same as VA_MAPBUFFER_FLAG_READ | VA_MAPBUFFER_FLAG_WRITE. More... | |
#define | VA_MAPBUFFER_FLAG_READ 1 |
application will read the surface after map | |
#define | VA_MAPBUFFER_FLAG_WRITE 2 |
application will write the surface after map | |
#define | VA_TIMEOUT_INFINITE 0xFFFFFFFFFFFFFFFF |
Indicates an infinite timeout. | |
#define | VA_FOURCC(ch0, ch1, ch2, ch3) |
#define | VA_FOURCC_NV12 0x3231564E |
#define | VA_FOURCC_NV21 0x3132564E |
#define | VA_FOURCC_AI44 0x34344149 |
#define | VA_FOURCC_RGBA 0x41424752 |
#define | VA_FOURCC_RGBX 0x58424752 |
#define | VA_FOURCC_BGRA 0x41524742 |
#define | VA_FOURCC_BGRX 0x58524742 |
#define | VA_FOURCC_ARGB 0x42475241 |
#define | VA_FOURCC_XRGB 0x42475258 |
#define | VA_FOURCC_ABGR 0x52474241 |
#define | VA_FOURCC_XBGR 0x52474258 |
#define | VA_FOURCC_UYVY 0x59565955 |
#define | VA_FOURCC_YUY2 0x32595559 |
#define | VA_FOURCC_AYUV 0x56555941 |
#define | VA_FOURCC_NV11 0x3131564e |
#define | VA_FOURCC_YV12 0x32315659 |
#define | VA_FOURCC_P208 0x38303250 |
#define | VA_FOURCC_I420 0x30323449 |
#define | VA_FOURCC_YV24 0x34325659 |
#define | VA_FOURCC_YV32 0x32335659 |
#define | VA_FOURCC_Y800 0x30303859 |
#define | VA_FOURCC_IMC3 0x33434D49 |
#define | VA_FOURCC_411P 0x50313134 |
#define | VA_FOURCC_411R 0x52313134 |
#define | VA_FOURCC_422H 0x48323234 |
#define | VA_FOURCC_422V 0x56323234 |
#define | VA_FOURCC_444P 0x50343434 |
#define | VA_FOURCC_RGBP 0x50424752 |
#define | VA_FOURCC_BGRP 0x50524742 |
#define | VA_FOURCC_RGB565 0x36314752 |
#define | VA_FOURCC_BGR565 0x36314742 |
#define | VA_FOURCC_Y210 0x30313259 |
#define | VA_FOURCC_Y212 0x32313259 |
#define | VA_FOURCC_Y216 0x36313259 |
#define | VA_FOURCC_Y410 0x30313459 |
#define | VA_FOURCC_Y412 0x32313459 |
#define | VA_FOURCC_Y416 0x36313459 |
#define | VA_FOURCC_YV16 0x36315659 |
#define | VA_FOURCC_P010 0x30313050 |
#define | VA_FOURCC_P012 0x32313050 |
#define | VA_FOURCC_P016 0x36313050 |
#define | VA_FOURCC_I010 0x30313049 |
#define | VA_FOURCC_IYUV 0x56555949 |
#define | VA_FOURCC_A2R10G10B10 0x30335241 /* VA_FOURCC('A','R','3','0') */ |
#define | VA_FOURCC_A2B10G10R10 0x30334241 /* VA_FOURCC('A','B','3','0') */ |
#define | VA_FOURCC_X2R10G10B10 0x30335258 /* VA_FOURCC('X','R','3','0') */ |
#define | VA_FOURCC_X2B10G10R10 0x30334258 /* VA_FOURCC('X','B','3','0') */ |
#define | VA_FOURCC_Y8 0x20203859 |
#define | VA_FOURCC_Y16 0x20363159 |
#define | VA_FOURCC_VYUY 0x59555956 |
#define | VA_FOURCC_YVYU 0x55595659 |
#define | VA_FOURCC_ARGB64 0x34475241 |
#define | VA_FOURCC_ABGR64 0x34474241 |
#define | VA_FOURCC_XYUV 0x56555958 |
#define | VA_FOURCC_Q416 0x36313451 |
#define | VA_SUBPICTURE_CHROMA_KEYING 0x0001 |
#define | VA_ROTATION_NONE 0x00000000 |
Typedefs | |
typedef void * | VADisplay |
typedef void(* | VAMessageCallback) (void *user_context, const char *message) |
typedef void * | VANativeDisplay |
typedef unsigned int | VAGenericID |
typedef VAGenericID | VAContextID |
typedef void(* | VAGenericFunc) (void) |
Generic function type. | |
typedef VAGenericID | VABufferID |
typedef VAGenericID | VASubpictureID |
Enumerations | |
enum | VAProfile { VAProfileNone = -1 , VAProfileMPEG2Simple = 0 , VAProfileMPEG2Main = 1 , VAProfileMPEG4Simple = 2 , VAProfileMPEG4AdvancedSimple = 3 , VAProfileMPEG4Main = 4 , va_deprecated_enum = 5 , VAProfileH264Main = 6 , VAProfileH264High = 7 , VAProfileVC1Simple = 8 , VAProfileVC1Main = 9 , VAProfileVC1Advanced = 10 , VAProfileH263Baseline = 11 , VAProfileJPEGBaseline = 12 , VAProfileH264ConstrainedBaseline = 13 , VAProfileVP8Version0_3 = 14 , VAProfileH264MultiviewHigh = 15 , VAProfileH264StereoHigh = 16 , VAProfileHEVCMain = 17 , VAProfileHEVCMain10 = 18 , VAProfileVP9Profile0 = 19 , VAProfileVP9Profile1 = 20 , VAProfileVP9Profile2 = 21 , VAProfileVP9Profile3 = 22 , VAProfileHEVCMain12 = 23 , VAProfileHEVCMain422_10 = 24 , VAProfileHEVCMain422_12 = 25 , VAProfileHEVCMain444 = 26 , VAProfileHEVCMain444_10 = 27 , VAProfileHEVCMain444_12 = 28 , VAProfileHEVCSccMain = 29 , VAProfileHEVCSccMain10 = 30 , VAProfileHEVCSccMain444 = 31 , VAProfileAV1Profile0 = 32 , VAProfileAV1Profile1 = 33 , VAProfileHEVCSccMain444_10 = 34 , VAProfileProtected = 35 , VAProfileH264High10 = 36 , VAProfileVVCMain10 = 37 , VAProfileVVCMultilayerMain10 = 38 } |
enum | VAEntrypoint { } |
enum | VAConfigAttribType { } |
enum | VAGenericValueType { VAGenericValueTypeInteger = 1 , VAGenericValueTypeFloat , VAGenericValueTypePointer , VAGenericValueTypeFunc } |
Generic value types. More... | |
enum | VASurfaceAttribType { } |
Surface attribute types. More... | |
enum | VABufferType { } |
enum | VAEncMiscParameterType { } |
enum | VAEncPackedHeaderType { VAEncPackedHeaderSequence = 1 , VAEncPackedHeaderPicture = 2 , VAEncPackedHeaderSlice = 3 , VAEncPackedHeaderRawData = 4 , va_deprecated_enum = 5 } |
Packed header type. More... | |
enum | VAMvModeVC1 |
enum | VADisplayAttribBLEMode |
Functions | |
const char * | vaErrorStr (VAStatus error_status) |
VAMessageCallback | vaSetErrorCallback (VADisplay dpy, VAMessageCallback callback, void *user_context) |
VAMessageCallback | vaSetInfoCallback (VADisplay dpy, VAMessageCallback callback, void *user_context) |
VAStatus | vaSetDriverName (VADisplay dpy, char *driver_name) |
VAStatus | vaInitialize (VADisplay dpy, int *major_version, int *minor_version) |
VAStatus | vaTerminate (VADisplay dpy) |
const char * | vaQueryVendorString (VADisplay dpy) |
VAPrivFunc | vaGetLibFunc (VADisplay dpy, const char *func) |
int | vaMaxNumProfiles (VADisplay dpy) |
int | vaMaxNumEntrypoints (VADisplay dpy) |
int | vaMaxNumConfigAttributes (VADisplay dpy) |
VAStatus | vaQueryConfigProfiles (VADisplay dpy, VAProfile *profile_list, int *num_profiles) |
VAStatus | vaQueryConfigEntrypoints (VADisplay dpy, VAProfile profile, VAEntrypoint *entrypoint_list, int *num_entrypoints) |
VAStatus | vaGetConfigAttributes (VADisplay dpy, VAProfile profile, VAEntrypoint entrypoint, VAConfigAttrib *attrib_list, int num_attribs) |
VAStatus | vaCreateConfig (VADisplay dpy, VAProfile profile, VAEntrypoint entrypoint, VAConfigAttrib *attrib_list, int num_attribs, VAConfigID *config_id) |
VAStatus | vaDestroyConfig (VADisplay dpy, VAConfigID config_id) |
VAStatus | vaQueryConfigAttributes (VADisplay dpy, VAConfigID config_id, VAProfile *profile, VAEntrypoint *entrypoint, VAConfigAttrib *attrib_list, int *num_attribs) |
VAStatus | vaQuerySurfaceAttributes (VADisplay dpy, VAConfigID config, VASurfaceAttrib *attrib_list, unsigned int *num_attribs) |
Queries surface attributes for the supplied config. More... | |
VAStatus | vaCreateSurfaces (VADisplay dpy, unsigned int format, unsigned int width, unsigned int height, VASurfaceID *surfaces, unsigned int num_surfaces, VASurfaceAttrib *attrib_list, unsigned int num_attribs) |
Creates an array of surfaces. More... | |
VAStatus | vaDestroySurfaces (VADisplay dpy, VASurfaceID *surfaces, int num_surfaces) |
VAStatus | vaCreateContext (VADisplay dpy, VAConfigID config_id, int picture_width, int picture_height, int flag, VASurfaceID *render_targets, int num_render_targets, VAContextID *context) |
VAStatus | vaDestroyContext (VADisplay dpy, VAContextID context) |
VAStatus | vaCreateMFContext (VADisplay dpy, VAMFContextID *mf_context) |
VAStatus | vaMFAddContext (VADisplay dpy, VAMFContextID mf_context, VAContextID context) |
VAStatus | vaMFReleaseContext (VADisplay dpy, VAMFContextID mf_context, VAContextID context) |
VAStatus | vaQueryProcessingRate (VADisplay dpy, VAConfigID config, VAProcessingRateParameter *proc_buf, unsigned int *processing_rate) |
Queries processing rate for the supplied config. More... | |
VAStatus | vaCreateBuffer (VADisplay dpy, VAContextID context, VABufferType type, unsigned int size, unsigned int num_elements, void *data, VABufferID *buf_id) |
VAStatus | vaCreateBuffer2 (VADisplay dpy, VAContextID context, VABufferType type, unsigned int width, unsigned int height, unsigned int *unit_size, unsigned int *pitch, VABufferID *buf_id) |
VAStatus | vaBufferSetNumElements (VADisplay dpy, VABufferID buf_id, unsigned int num_elements) |
VAStatus | vaMapBuffer (VADisplay dpy, VABufferID buf_id, void **pbuf) |
VAStatus | vaUnmapBuffer (VADisplay dpy, VABufferID buf_id) |
VAStatus | vaDestroyBuffer (VADisplay dpy, VABufferID buffer_id) |
VAStatus | vaAcquireBufferHandle (VADisplay dpy, VABufferID buf_id, VABufferInfo *buf_info) |
Acquires buffer handle for external API usage. More... | |
VAStatus | vaReleaseBufferHandle (VADisplay dpy, VABufferID buf_id) |
Releases buffer after usage from external API. More... | |
VAStatus | vaExportSurfaceHandle (VADisplay dpy, VASurfaceID surface_id, uint32_t mem_type, uint32_t flags, void *descriptor) |
Export a handle to a surface for use with an external API. More... | |
VAStatus | vaBeginPicture (VADisplay dpy, VAContextID context, VASurfaceID render_target) |
VAStatus | vaRenderPicture (VADisplay dpy, VAContextID context, VABufferID *buffers, int num_buffers) |
VAStatus | vaEndPicture (VADisplay dpy, VAContextID context) |
VAStatus | vaMFSubmit (VADisplay dpy, VAMFContextID mf_context, VAContextID *contexts, int num_contexts) |
VAStatus | vaSyncSurface (VADisplay dpy, VASurfaceID render_target) |
VAStatus | vaSyncSurface2 (VADisplay dpy, VASurfaceID surface, uint64_t timeout_ns) |
Synchronizes pending operations associated with the supplied surface. More... | |
VAStatus | vaQuerySurfaceStatus (VADisplay dpy, VASurfaceID render_target, VASurfaceStatus *status) |
VAStatus | vaQuerySurfaceError (VADisplay dpy, VASurfaceID surface, VAStatus error_status, void **error_info) |
VAStatus | vaSyncBuffer (VADisplay dpy, VABufferID buf_id, uint64_t timeout_ns) |
Synchronizes pending operations associated with the supplied buffer. More... | |
int | vaMaxNumImageFormats (VADisplay dpy) |
VAStatus | vaQueryImageFormats (VADisplay dpy, VAImageFormat *format_list, int *num_formats) |
VAStatus | vaCreateImage (VADisplay dpy, VAImageFormat *format, int width, int height, VAImage *image) |
VAStatus | vaDestroyImage (VADisplay dpy, VAImageID image) |
VAStatus | vaGetImage (VADisplay dpy, VASurfaceID surface, int x, int y, unsigned int width, unsigned int height, VAImageID image) |
VAStatus | vaPutImage (VADisplay dpy, VASurfaceID surface, VAImageID image, int src_x, int src_y, unsigned int src_width, unsigned int src_height, int dest_x, int dest_y, unsigned int dest_width, unsigned int dest_height) |
VAStatus | vaDeriveImage (VADisplay dpy, VASurfaceID surface, VAImage *image) |
int | vaMaxNumSubpictureFormats (VADisplay dpy) |
VAStatus | vaQuerySubpictureFormats (VADisplay dpy, VAImageFormat *format_list, unsigned int *flags, unsigned int *num_formats) |
VAStatus | vaCreateSubpicture (VADisplay dpy, VAImageID image, VASubpictureID *subpicture) |
VAStatus | vaDestroySubpicture (VADisplay dpy, VASubpictureID subpicture) |
VAStatus | vaSetSubpictureImage (VADisplay dpy, VASubpictureID subpicture, VAImageID image) |
VAStatus | vaSetSubpictureChromakey (VADisplay dpy, VASubpictureID subpicture, unsigned int chromakey_min, unsigned int chromakey_max, unsigned int chromakey_mask) |
VAStatus | vaSetSubpictureGlobalAlpha (VADisplay dpy, VASubpictureID subpicture, float global_alpha) |
VAStatus | vaAssociateSubpicture (VADisplay dpy, VASubpictureID subpicture, VASurfaceID *target_surfaces, int num_surfaces, int16_t src_x, int16_t src_y, uint16_t src_width, uint16_t src_height, int16_t dest_x, int16_t dest_y, uint16_t dest_width, uint16_t dest_height, uint32_t flags) |
VAStatus | vaDeassociateSubpicture (VADisplay dpy, VASubpictureID subpicture, VASurfaceID *target_surfaces, int num_surfaces) |
Attribute values for VAConfigAttribRateControl | |
#define | VA_RC_NONE 0x00000001 |
Driver does not support any form of rate control. | |
#define | VA_RC_CBR 0x00000002 |
Constant bitrate. | |
#define | VA_RC_VBR 0x00000004 |
Variable bitrate. | |
#define | VA_RC_VCM 0x00000008 |
Video conference mode. | |
#define | VA_RC_CQP 0x00000010 |
Constant QP. | |
#define | VA_RC_VBR_CONSTRAINED 0x00000020 |
Variable bitrate with peak rate higher than average bitrate. | |
#define | VA_RC_ICQ 0x00000040 |
Intelligent Constant Quality. Provided an initial ICQ_quality_factor, adjusts QP at a frame and MB level based on motion to improve subjective quality. | |
#define | VA_RC_MB 0x00000080 |
Macroblock based rate control. Per MB control is decided internally in the encoder. It may be combined with other RC modes, except CQP. | |
#define | VA_RC_CFS 0x00000100 |
Constant Frame Size, it is used for small tolerent | |
#define | VA_RC_PARALLEL 0x00000200 |
Parallel BRC, for hierachical B. More... | |
#define | VA_RC_QVBR 0x00000400 |
Quality defined VBR Use Quality factor to determine the good enough QP for each MB such that good enough quality can be obtained without waste of bits for this BRC mode, you must set all legacy VBR parameters and reuse quality_factor in VAEncMiscParameterRateControl . | |
#define | VA_RC_AVBR 0x00000800 |
Average VBR Average variable bitrate control algorithm focuses on overall encoding quality while meeting the specified target bitrate, within the accuracy range, after a convergence period. bits_per_second in VAEncMiscParameterRateControl is target bitrate for AVBR. Convergence is specified in the unit of frame. window_size in VAEncMiscParameterRateControl is equal to convergence for AVBR. Accuracy is in the range of [1,100], 1 means one percent, and so on. target_percentage in VAEncMiscParameterRateControl is equal to accuracy for AVBR. | |
#define | VA_RC_TCBRC 0x00001000 |
Transport Controlled BRC Specific bitrate control for real time streaming. TCBRC can instantly react to channel change to remove or significantly reduce the delay. Application (transport) provides channel feedback to driver through TargetFrameSize. When channel condition is very good (almost no constraint on instant frame size), the app should set target frame size as zero. Otherwise, channel capacity divided by fps should be used. | |
Attribute values for VAConfigAttribDecSliceMode | |
#define | VA_DEC_SLICE_MODE_NORMAL 0x00000001 |
Driver supports normal mode for slice decoding. | |
#define | VA_DEC_SLICE_MODE_BASE 0x00000002 |
Driver supports base mode for slice decoding. | |
Attribute values for VAConfigAttribDecProcessing | |
#define | VA_DEC_PROCESSING_NONE 0x00000000 |
No decoding + processing in a single decoding call. | |
#define | VA_DEC_PROCESSING 0x00000001 |
Decode + processing in a single decoding call. | |
Attribute values for VAConfigAttribEncPackedHeaders | |
#define | VA_ENC_PACKED_HEADER_NONE 0x00000000 |
Driver does not support any packed headers mode. | |
#define | VA_ENC_PACKED_HEADER_SEQUENCE 0x00000001 |
Driver supports packed sequence headers. e.g. SPS for H.264. More... | |
#define | VA_ENC_PACKED_HEADER_PICTURE 0x00000002 |
Driver supports packed picture headers. e.g. PPS for H.264. More... | |
#define | VA_ENC_PACKED_HEADER_SLICE 0x00000004 |
Driver supports packed slice headers. e.g. slice_header() for H.264. More... | |
#define | VA_ENC_PACKED_HEADER_MISC 0x00000008 |
Driver supports misc packed headers. e.g. SEI for H.264. More... | |
#define | VA_ENC_PACKED_HEADER_RAW_DATA 0x00000010 |
Driver supports raw packed header, see VAEncPackedHeaderRawData. | |
Attribute values for VAConfigAttribEncInterlaced | |
#define | VA_ENC_INTERLACED_NONE 0x00000000 |
Driver does not support interlaced coding. | |
#define | VA_ENC_INTERLACED_FRAME 0x00000001 |
Driver supports interlaced frame coding. | |
#define | VA_ENC_INTERLACED_FIELD 0x00000002 |
Driver supports interlaced field coding. | |
#define | VA_ENC_INTERLACED_MBAFF 0x00000004 |
Driver supports macroblock adaptive frame field coding. | |
#define | VA_ENC_INTERLACED_PAFF 0x00000008 |
Driver supports picture adaptive frame field coding. | |
Attribute values for VAConfigAttribEncSliceStructure | |
#define | VA_ENC_SLICE_STRUCTURE_POWER_OF_TWO_ROWS 0x00000001 |
Driver supports a power-of-two number of rows per slice. | |
#define | VA_ENC_SLICE_STRUCTURE_ARBITRARY_MACROBLOCKS 0x00000002 |
Driver supports an arbitrary number of macroblocks per slice. | |
#define | VA_ENC_SLICE_STRUCTURE_EQUAL_ROWS 0x00000004 |
Driver support 1 row per slice. | |
#define | VA_ENC_SLICE_STRUCTURE_MAX_SLICE_SIZE 0x00000008 |
Driver support max encoded slice size per slice. | |
#define | VA_ENC_SLICE_STRUCTURE_ARBITRARY_ROWS 0x00000010 |
Driver supports an arbitrary number of rows per slice. | |
#define | VA_ENC_SLICE_STRUCTURE_EQUAL_MULTI_ROWS 0x00000020 |
Driver supports any number of rows per slice but they must be the same for all slices except for the last one, which must be equal or smaller to the previous slices. | |
Attribute values for VAConfigAttribEncQuantization | |
#define | VA_ENC_QUANTIZATION_NONE 0x00000000 |
Driver does not support special types of quantization. | |
#define | VA_ENC_QUANTIZATION_TRELLIS_SUPPORTED 0x00000001 |
Driver supports trellis quantization. | |
Attribute values for VAConfigAttribPredictionDirection | |
#define | VA_PREDICTION_DIRECTION_PREVIOUS 0x00000001 |
Driver support forward reference frame (inter frame for vpx, P frame for H26x MPEG) can work with the VAConfigAttribEncMaxRefFrames. for example: low delay B frame of HEVC. these value can be OR'd together. typical value should be VA_PREDICTION_DIRECTION_PREVIOUS or VA_PREDICTION_DIRECTION_PREVIOUS | VA_PREDICTION_DIRECTION_FUTURE, theoretically, there are no stream only include future reference frame. | |
#define | VA_PREDICTION_DIRECTION_FUTURE 0x00000002 |
Driver support backward prediction frame/slice. | |
#define | VA_PREDICTION_DIRECTION_BI_NOT_EMPTY 0x00000004 |
Dirver require both reference list must be not empty for inter frame. | |
Attribute values for VAConfigAttribEncIntraRefresh | |
#define | VA_ENC_INTRA_REFRESH_NONE 0x00000000 |
Driver does not support intra refresh. | |
#define | VA_ENC_INTRA_REFRESH_ROLLING_COLUMN 0x00000001 |
Driver supports column based rolling intra refresh. | |
#define | VA_ENC_INTRA_REFRESH_ROLLING_ROW 0x00000002 |
Driver supports row based rolling intra refresh. | |
#define | VA_ENC_INTRA_REFRESH_ADAPTIVE 0x00000010 |
Driver supports adaptive intra refresh. | |
#define | VA_ENC_INTRA_REFRESH_CYCLIC 0x00000020 |
Driver supports cyclic intra refresh. | |
#define | VA_ENC_INTRA_REFRESH_P_FRAME 0x00010000 |
Driver supports intra refresh of P frame. | |
#define | VA_ENC_INTRA_REFRESH_B_FRAME 0x00020000 |
Driver supports intra refresh of B frame. | |
#define | VA_ENC_INTRA_REFRESH_MULTI_REF 0x00040000 |
Driver supports intra refresh of multiple reference encoder. | |
Attribute values for VAConfigAttribProtectedContentCipherAlgorithm | |
#define | VA_PC_CIPHER_AES 0x00000001 |
AES cipher. | |
Attribute values for VAConfigAttribProtectedContentCipherBlockSize | |
#define | VA_PC_BLOCK_SIZE_128 0x00000001 |
128 bits block size | |
#define | VA_PC_BLOCK_SIZE_192 0x00000002 |
192 bits block size | |
#define | VA_PC_BLOCK_SIZE_256 0x00000004 |
256 bits block size | |
Attribute values for VAConfigAttribProtectedContentCipherMode | |
#define | VA_PC_CIPHER_MODE_ECB 0x00000001 |
AES ECB. | |
#define | VA_PC_CIPHER_MODE_CBC 0x00000002 |
AES CBC. | |
#define | VA_PC_CIPHER_MODE_CTR 0x00000004 |
AES CTR. | |
Attribute values for VAConfigAttribProtectedContentCipherSampleType | |
#define | VA_PC_SAMPLE_TYPE_FULLSAMPLE 0x00000001 |
Full sample. | |
#define | VA_PC_SAMPLE_TYPE_SUBSAMPLE 0x00000002 |
Sub sample. | |
Attribute values for VAConfigAttribProtectedContentUsage | |
#define | VA_PC_USAGE_DEFAULT 0x00000000 |
Default usage. | |
#define | VA_PC_USAGE_WIDEVINE 0x00000001 |
Widevine. | |
Attribute values for VAConfigAttribProcessingRate. | |
#define | VA_PROCESSING_RATE_NONE 0x00000000 |
Driver does not support processing rate report. | |
#define | VA_PROCESSING_RATE_ENCODE 0x00000001 |
Driver supports encode processing rate report | |
#define | VA_PROCESSING_RATE_DECODE 0x00000002 |
Driver supports decode processing rate report | |
Surface attribute flags | |
#define | VA_SURFACE_ATTRIB_NOT_SUPPORTED 0x00000000 |
Surface attribute is not supported. | |
#define | VA_SURFACE_ATTRIB_GETTABLE 0x00000001 |
Surface attribute can be got through vaQuerySurfaceAttributes(). | |
#define | VA_SURFACE_ATTRIB_SETTABLE 0x00000002 |
Surface attribute can be set through vaCreateSurfaces(). | |
VASurfaceAttribMemoryType values in bit fields. | |
Bits 0:7 are reserved for generic types. Bits 31:28 are reserved for Linux DRM. Bits 23:20 are reserved for Android. Bits 19:16 are reserved for Win32. DRM, Android and Win32 specific types are defined in respective va_*.h header files. | |
#define | VA_SURFACE_ATTRIB_MEM_TYPE_VA 0x00000001 |
VA memory type (default) is supported. | |
#define | VA_SURFACE_ATTRIB_MEM_TYPE_V4L2 0x00000002 |
V4L2 buffer memory type is supported. | |
#define | VA_SURFACE_ATTRIB_MEM_TYPE_USER_PTR 0x00000004 |
User pointer memory type is supported. | |
VASurfaceAttribExternalBuffers flags | |
#define | VA_SURFACE_EXTBUF_DESC_ENABLE_TILING 0x00000001 |
Enable memory tiling. | |
#define | VA_SURFACE_EXTBUF_DESC_CACHED 0x00000002 |
Memory is cacheable. | |
#define | VA_SURFACE_EXTBUF_DESC_UNCACHED 0x00000004 |
Memory is non-cacheable. | |
#define | VA_SURFACE_EXTBUF_DESC_WC 0x00000008 |
Memory is write-combined. | |
#define | VA_SURFACE_EXTBUF_DESC_PROTECTED 0x80000000 |
Memory is protected. | |
VASurfaceAttribUsageHint attribute usage hint flags | |
#define | VA_SURFACE_ATTRIB_USAGE_HINT_GENERIC 0x00000000 |
Surface usage not indicated. | |
#define | VA_SURFACE_ATTRIB_USAGE_HINT_DECODER 0x00000001 |
Surface used by video decoder. | |
#define | VA_SURFACE_ATTRIB_USAGE_HINT_ENCODER 0x00000002 |
Surface used by video encoder. | |
#define | VA_SURFACE_ATTRIB_USAGE_HINT_VPP_READ 0x00000004 |
Surface read by video post-processing. | |
#define | VA_SURFACE_ATTRIB_USAGE_HINT_VPP_WRITE 0x00000008 |
Surface written by video post-processing. | |
#define | VA_SURFACE_ATTRIB_USAGE_HINT_DISPLAY 0x00000010 |
Surface used for display. | |
#define | VA_SURFACE_ATTRIB_USAGE_HINT_EXPORT 0x00000020 |
Surface used for export to third-party APIs, e.g. via vaExportSurfaceHandle(). | |
vaExportSurfaceHandle() flags | |
#define | VA_EXPORT_SURFACE_READ_ONLY 0x0001 |
#define | VA_EXPORT_SURFACE_WRITE_ONLY 0x0002 |
#define | VA_EXPORT_SURFACE_READ_WRITE 0x0003 |
#define | VA_EXPORT_SURFACE_SEPARATE_LAYERS 0x0004 |
#define | VA_EXPORT_SURFACE_COMPOSED_LAYERS 0x0008 |
#define VA_ATTRIB_NOT_SUPPORTED 0x80000000 |
if an attribute is not applicable for a given profile/entrypoint pair, then set the value to the following
#define VA_CLEAR_DRAWABLE 0x00000008 |
Clears the drawable with background color. for hardware overlay based implementation this flag can be used to turn off the overlay
#define VA_CODED_BUF_STATUS_FRAME_SIZE_OVERFLOW 0x1000 |
The frame has exceeded the maximum requested size.
This flag indicates that the encoded frame size exceeds the value specified through a misc parameter buffer of type VAEncMiscParameterTypeMaxFrameSize.
#define VA_CODED_BUF_STATUS_NUMBER_PASSES_MASK 0xf000000 |
The coded buffer segment status contains frame encoding passes number.
This is the mask to get the number of encoding passes from the coded buffer segment status. NUMBER_PASS(bit24~bit27): the number for encoding passes executed for the coded frame.
#define VA_CODED_BUF_STATUS_PICTURE_AVE_QP_MASK 0xff |
device independent data structure for codedbuffer
#define VA_CODED_BUF_STATUS_SINGLE_NALU 0x10000000 |
The coded buffer segment contains a single NAL unit.
This flag indicates that the coded buffer segment contains a single NAL unit. This flag might be useful to the user for processing the coded buffer.
#define VA_ENABLE_BLEND 0x00000004 /* video area blend with the constant color */ |
Enabled the positioning/cropping/blending feature: 1, specify the video playback position in the isurface 2, specify the cropping info for video playback 3, encoded video will blend with background color
#define VA_ENC_PACKED_HEADER_MISC 0x00000008 |
Driver supports misc packed headers. e.g. SEI for H.264.
VA_ENC_PACKED_HEADER_RAW_DATA
to pass the corresponding packed header data buffer to the driver #define VA_ENC_PACKED_HEADER_PICTURE 0x00000002 |
Driver supports packed picture headers. e.g. PPS for H.264.
Application must provide it to driver once this falg is returned through vaGetConfigAttributes()
#define VA_ENC_PACKED_HEADER_SEQUENCE 0x00000001 |
Driver supports packed sequence headers. e.g. SPS for H.264.
Application must provide it to driver once this flag is returned through vaGetConfigAttributes()
#define VA_ENC_PACKED_HEADER_SLICE 0x00000004 |
Driver supports packed slice headers. e.g. slice_header() for H.264.
Application must provide it to driver once this flag is returned through vaGetConfigAttributes()
#define VA_ENCRYPTION_TYPE_FULLSAMPLE_CTR 0x00000001 /* AES CTR fullsample */ |
These ENCRYPTION_TYPEs are used for the attribute values for VAConfigAttribEncryption
and for encryption_type in VAEncryptionParameters.
When used for VAConfigAttribEncryption
, it be used via vaQueryConfigEntrypoints to check which type are supported for specific profile or not.
When used for encryption_type in VAEncryptionParameters, it tells driver the parameters in VAEncryptionParameters are used for which encryption type.
#define VA_EXEC_ASYNC 0x1 |
asynchronization,application should call additonal sync operation to access output
#define VA_EXEC_MODE_DEFAULT 0x0 |
operation mode
#define VA_EXEC_SYNC 0x0 |
operation options synchronization, block call, output should be ready after execution function return
#define VA_EXPORT_SURFACE_COMPOSED_LAYERS 0x0008 |
Export surface with composed layers.
For example, NV12 surfaces should be exported as a single NV12 composed object.
#define VA_EXPORT_SURFACE_READ_ONLY 0x0001 |
Export surface to be read by external API.
#define VA_EXPORT_SURFACE_READ_WRITE 0x0003 |
Export surface to be both read and written by external API.
#define VA_EXPORT_SURFACE_SEPARATE_LAYERS 0x0004 |
Export surface with separate layers.
For example, NV12 surfaces should be exported as two separate planes for luma and chroma.
#define VA_EXPORT_SURFACE_WRITE_ONLY 0x0002 |
Export surface to be written by external API.
#define VA_FEATURE_NOT_SUPPORTED 0 |
The feature is not supported by the device.
Any corresponding feature flag must not be set.
#define VA_FEATURE_REQUIRED 2 |
The feature is required by the device.
The device does not support not enabling this feature, so any corresponding feature flag must be set and any additional configuration needed by the feature must be supplied.
#define VA_FEATURE_SUPPORTED 1 |
The feature is supported by the device.
The user may decide whether or not to use this feature.
Note that support for a feature only indicates that the hardware is able to use it; whether it is actually a positive change to enable it in a given situation will depend on other factors including the input provided by the user.
#define VA_FILTER_INTERPOLATION_DEFAULT 0x00000000 |
Interpolation method for scaling
#define VA_FILTER_SCALING_DEFAULT 0x00000000 |
Scaling flags for vaPutSurface()
#define VA_FOURCC | ( | ch0, | |
ch1, | |||
ch2, | |||
ch3 | |||
) |
Notes about synchronization interfaces: vaSyncSurface:
vaSyncSurface2:
vaSyncBuffer:
#define VA_FOURCC_411P 0x50313134 |
411P: three-plane 8-bit YUV 4:1:1.
The three planes contain Y, U and V respectively.
#define VA_FOURCC_411R 0x52313134 |
411R: three-plane 8-bit YUV.
The subsampling is the transpose of 4:1:1 - full chroma appears on every fourth line. The three planes contain Y, U and V respectively.
#define VA_FOURCC_422H 0x48323234 |
422H: three-plane 8-bit YUV 4:2:2.
The three planes contain Y, U and V respectively.
#define VA_FOURCC_422V 0x56323234 |
422V: three-plane 8-bit YUV 4:4:0.
The three planes contain Y, U and V respectively.
#define VA_FOURCC_444P 0x50343434 |
444P: three-plane 8-bit YUV 4:4:4.
The three planes contain Y, U and V respectively.
#define VA_FOURCC_A2B10G10R10 0x30334241 /* VA_FOURCC('A','B','3','0') */ |
10-bit Pixel BGR formats.
#define VA_FOURCC_A2R10G10B10 0x30335241 /* VA_FOURCC('A','R','3','0') */ |
10-bit Pixel RGB formats.
#define VA_FOURCC_ABGR 0x52474241 |
ABGR: packed 8-bit RGBA.
Four bytes per pixel: alpha, blue, green, red.
#define VA_FOURCC_ABGR64 0x34474241 |
ABGR64: three-plane 16-bit ABGR 16:16:16:16
The four planes contain: alpha, blue, green, red respectively.
#define VA_FOURCC_AI44 0x34344149 |
AI44: packed 4-bit YA.
The bottom half of each byte contains luma, the top half contains alpha.
#define VA_FOURCC_ARGB 0x42475241 |
ARGB: packed 8-bit RGBA.
Four bytes per pixel: alpha, red, green, blue.
#define VA_FOURCC_ARGB64 0x34475241 |
AGRB64: three-plane 16-bit ARGB 16:16:16:16
The four planes contain: alpha, red, green, blue respectively.
#define VA_FOURCC_AYUV 0x56555941 |
AYUV: packed 8-bit YUVA 4:4:4.
Four bytes per pixel: A, Y, U, V.
#define VA_FOURCC_BGR565 0x36314742 |
BG16: packed 5/6-bit RGB.
Each pixel is a two-byte little-endian value. Blue, green and red are found in bits 15:11, 10:5, 4:0 respectively.
#define VA_FOURCC_BGRA 0x41524742 |
BGRA: packed 8-bit RGBA.
Four bytes per pixel: blue, green, red, alpha.
#define VA_FOURCC_BGRP 0x50524742 |
BGRP: three-plane 8-bit RGB.
The three planes contain blue, green and red respectively.
#define VA_FOURCC_BGRX 0x58524742 |
BGRX: packed 8-bit RGB.
Four bytes per pixel: blue, green, red, unspecified.
#define VA_FOURCC_I010 0x30313049 |
I010: three-plane 10-bit YUV 4:2:0.
Each sample is a two-byte little-endian value with the top six bits ignored. The three planes contain Y, V and U respectively.
#define VA_FOURCC_I420 0x30323449 |
I420: three-plane 8-bit YUV 4:2:0.
The three planes contain Y, U and V respectively.
#define VA_FOURCC_IMC3 0x33434D49 |
IMC3: three-plane 8-bit YUV 4:2:0.
Equivalent to YV12, but with the additional constraint that the pitch of all three planes must be the same.
#define VA_FOURCC_IYUV 0x56555949 |
IYUV: three-plane 8-bit YUV 4:2:0.
#define VA_FOURCC_NV11 0x3131564e |
NV11: two-plane 8-bit YUV 4:1:1.
The first plane contains Y, the second plane contains U and V in pairs of bytes.
#define VA_FOURCC_NV12 0x3231564E |
NV12: two-plane 8-bit YUV 4:2:0. The first plane contains Y, the second plane contains U and V in pairs of bytes.
#define VA_FOURCC_NV21 0x3132564E |
NV21: two-plane 8-bit YUV 4:2:0. Same as NV12, but with U and V swapped.
#define VA_FOURCC_P010 0x30313050 |
P010: two-plane 10-bit YUV 4:2:0.
Each sample is a two-byte little-endian value with the bottom six bits ignored. The first plane contains Y, the second plane contains U and V in pairs of samples.
#define VA_FOURCC_P012 0x32313050 |
P012: two-plane 12-bit YUV 4:2:0.
Each sample is a two-byte little-endian value with the bottom four bits ignored. The first plane contains Y, the second plane contains U and V in pairs of samples.
#define VA_FOURCC_P016 0x36313050 |
P016: two-plane 16-bit YUV 4:2:0.
Each sample is a two-byte little-endian value. The first plane contains Y, the second plane contains U and V in pairs of samples.
#define VA_FOURCC_P208 0x38303250 |
P208: two-plane 8-bit YUV 4:2:2.
The first plane contains Y, the second plane contains U and V in pairs of bytes.
#define VA_FOURCC_Q416 0x36313451 |
Q416: three-plane 16-bit YUV 4:4:4.
The three planes contain Y, U and V respectively.
#define VA_FOURCC_RGB565 0x36314752 |
RG16: packed 5/6-bit RGB.
Each pixel is a two-byte little-endian value. Red, green and blue are found in bits 15:11, 10:5, 4:0 respectively.
#define VA_FOURCC_RGBA 0x41424752 |
RGBA: packed 8-bit RGBA.
Four bytes per pixel: red, green, blue, alpha.
#define VA_FOURCC_RGBP 0x50424752 |
RGBP: three-plane 8-bit RGB.
The three planes contain red, green and blue respectively.
#define VA_FOURCC_RGBX 0x58424752 |
RGBX: packed 8-bit RGB.
Four bytes per pixel: red, green, blue, unspecified.
#define VA_FOURCC_UYVY 0x59565955 |
UYUV: packed 8-bit YUV 4:2:2.
Four bytes per pair of pixels: U, Y, U, V.
#define VA_FOURCC_VYUY 0x59555956 |
VYUV: packed 8-bit YUV 4:2:2.
Four bytes per pair of pixels: V, Y, U, V.
#define VA_FOURCC_X2B10G10R10 0x30334258 /* VA_FOURCC('X','B','3','0') */ |
10-bit Pixel BGR formats without alpha.
#define VA_FOURCC_X2R10G10B10 0x30335258 /* VA_FOURCC('X','R','3','0') */ |
10-bit Pixel RGB formats without alpha.
#define VA_FOURCC_XBGR 0x52474258 |
XBGR: packed 8-bit RGB.
Four bytes per pixel: unspecified, blue, green, red.
#define VA_FOURCC_XRGB 0x42475258 |
XRGB: packed 8-bit RGB.
Four bytes per pixel: unspecified, red, green, blue.
#define VA_FOURCC_XYUV 0x56555958 |
XYUV: packed 8-bit YUVX 4:4:4.
Four bytes per pixel: X, Y, U, V.
#define VA_FOURCC_Y16 0x20363159 |
Y16: 16-bit greyscale.
Only a single sample, 16 bit Y plane for monochrome images
#define VA_FOURCC_Y210 0x30313259 |
Y210: packed 10-bit YUV 4:2:2.
Eight bytes represent a pair of pixels. Each sample is a two-byte little-endian value, with the bottom six bits ignored. The samples are in the order Y, U, Y, V.
#define VA_FOURCC_Y212 0x32313259 |
Y212: packed 12-bit YUV 4:2:2.
Eight bytes represent a pair of pixels. Each sample is a two-byte little-endian value. The samples are in the order Y, U, Y, V.
#define VA_FOURCC_Y216 0x36313259 |
Y216: packed 16-bit YUV 4:2:2.
Eight bytes represent a pair of pixels. Each sample is a two-byte little-endian value. The samples are in the order Y, U, Y, V.
#define VA_FOURCC_Y410 0x30313459 |
Y410: packed 10-bit YUVA 4:4:4.
Each pixel is a four-byte little-endian value. A, V, Y, U are found in bits 31:30, 29:20, 19:10, 9:0 respectively.
#define VA_FOURCC_Y412 0x32313459 |
Y412 packed 12-bit YUVA 4:4:4.
Each pixel is a set of four samples, each of which is a two-byte little-endian value. The samples are in the order A, V, Y, U.
#define VA_FOURCC_Y416 0x36313459 |
Y416: packed 16-bit YUVA 4:4:4.
Each pixel is a set of four samples, each of which is a two-byte little-endian value. The samples are in the order A, V, Y, U.
#define VA_FOURCC_Y8 0x20203859 |
Y8: 8-bit greyscale.
Only a single sample, 8 bit Y plane for monochrome images
#define VA_FOURCC_Y800 0x30303859 |
Y800: 8-bit greyscale.
#define VA_FOURCC_YUY2 0x32595559 |
YUY2: packed 8-bit YUV 4:2:2.
Four bytes per pair of pixels: Y, U, Y, V.
#define VA_FOURCC_YV12 0x32315659 |
YV12: three-plane 8-bit YUV 4:2:0.
The three planes contain Y, V and U respectively.
#define VA_FOURCC_YV16 0x36315659 |
YV16: three-plane 8-bit YUV 4:2:2.
The three planes contain Y, V and U respectively.
#define VA_FOURCC_YV24 0x34325659 |
YV24: three-plane 8-bit YUV 4:4:4.
The three planes contain Y, V and U respectively.
#define VA_FOURCC_YV32 0x32335659 |
YV32: four-plane 8-bit YUVA 4:4:4
The four planes contain Y, V, U and A respectively.
#define VA_FOURCC_YVYU 0x55595659 |
YVYU: packed 8-bit YUV 4:2:2.
Four bytes per pair of pixels: Y, V, Y, U.
#define VA_FRAME_PICTURE 0x00000000 |
#define VA_MAPBUFFER_FLAG_DEFAULT 0 |
VA_MAPBUFFER_FLAG_DEFAULT is used when there are no flag specified same as VA_MAPBUFFER_FLAG_READ | VA_MAPBUFFER_FLAG_WRITE.
Map data store of the buffer into the client's address space this interface could be used to convey the operation hint backend driver could use these hint to optimize the implementations
#define VA_PADDING_LOW 4 |
Padding size in 4-bytes
#define VA_RC_PARALLEL 0x00000200 |
Parallel BRC, for hierachical B.
For hierachical B, B frames can be refered by other B frames. Currently three layers of hierachy are defined: B0 - regular B, no reference to other B frames. B1 - reference to only I, P and regular B0 frames. B2 - reference to any other frames, including B1. In Hierachical B structure, B frames on the same layer can be processed simultaneously. And BRC would adjust accordingly. This is so called Parallel BRC.
#define VA_ROTATION_NONE 0x00000000 |
attribute value for VADisplayAttribRotation
#define VA_RT_FORMAT_RGB32_10BPP VA_RT_FORMAT_RGB32_10 |
#define VA_RT_FORMAT_YUV420_10BPP VA_RT_FORMAT_YUV420_10 |
#define VA_SLICE_DATA_FLAG_ALL 0x00 /* whole slice is in the buffer */ |
There will be cases where the bitstream buffer will not have enough room to hold the data for the entire slice, and the following flags will be used in the slice parameter to signal to the server for the possible cases. If a slice parameter buffer and slice data buffer pair is sent to the server with the slice data partially in the slice data buffer (BEGIN and MIDDLE cases below), then a slice parameter and data buffer needs to be sent again to complete this slice.
#define VA_SRC_COLOR_MASK 0x000000f0 |
Color space conversion flags for vaPutSurface()
#define VA_STATUS_ERROR_INVALID_VALUE 0x00000019 |
An invalid/unsupported value was supplied.
This is a catch-all error code for invalid or unsupported values. e.g. value exceeding the valid range, invalid type in the context of generic attribute values.
#define VA_STATUS_SUCCESS 0x00000000 |
Return status type from functions Values for the return status
#define VA_SUBPICTURE_CHROMA_KEYING 0x0001 |
flags for subpictures
typedef VAGenericID VABufferID |
Buffers Buffers are used to pass various types of data from the client to the server. The server maintains a data store for each buffer created, and the client idenfies a buffer through a unique buffer id assigned by the server.
typedef VAGenericID VAContextID |
Contexts and Surfaces
Context represents a "virtual" video decode, encode or video processing pipeline. Surfaces are render targets for a given context. The data in the surfaces are not accessible to the client except if derived image is supported and the internal data format of the surface is implementation specific.
Surfaces are provided as a hint of what surfaces will be used when the context is created through vaCreateContext(). A surface may be used by different contexts at the same time as soon as application can make sure the operations are synchronized between different contexts, e.g. a surface is used as the output of a decode context and the input of a video process context. Surfaces can only be destroyed after all contexts using these surfaces have been destroyed.
Both contexts and surfaces are identified by unique IDs and its implementation specific internals are kept opaque to the clients
typedef void* VADisplay |
Overview
The VA API is intended to provide an interface between a video decode/encode/processing application (client) and a hardware accelerator (server), to off-load video decode/encode/processing operations from the host to the hardware accelerator at various entry-points.
The basic operation steps are:
Initialization & Configuration Management
typedef unsigned int VAGenericID |
Generic ID type, can be re-typed for specific implementation
typedef void(* VAMessageCallback) (void *user_context, const char *message) |
Type of a message callback, used for both error and info log.
typedef void* VANativeDisplay |
Initialization: A display must be obtained by calling vaGetDisplay() before calling vaInitialize() and other functions. This connects the API to the native window system. For X Windows, native_dpy would be from XOpenDisplay()
typedef VAGenericID VASubpictureID |
Subpictures Subpicture is a special type of image that can be blended with a surface during vaPutSurface(). Subpicture can be used to render DVD sub-titles or closed captioning text etc.
enum VABufferType |
Enumerator | |
---|---|
VAEncQPBufferType | Encoding QP buffer. This buffer contains QP per MB for encoding. Currently VAEncQPBufferH264 is defined for H.264 encoding, see #VAEncQPBufferH264 for details |
VAProcPipelineParameterBufferType | Video processing pipeline parameter buffer. This buffer describes the video processing pipeline. See VAProcPipelineParameterBuffer for details. |
VAProcFilterParameterBufferType | Video filter parameter buffer. This buffer describes the video filter parameters. All buffers inherit from VAProcFilterParameterBufferBase, thus including a unique filter buffer type. The default buffer used by most filters is VAProcFilterParameterBuffer. Filters requiring advanced parameters include, but are not limited to, deinterlacing (VAProcFilterParameterBufferDeinterlacing), color balance (VAProcFilterParameterBufferColorBalance), etc. |
VAEncFEIMVBufferType | FEI specific buffer types. |
VAStatsStatisticsBufferType | Statistics output for VAEntrypointStats progressive and top field of interlaced case. |
VAStatsStatisticsBottomFieldBufferType | Statistics output for VAEntrypointStats bottom field of interlaced case. |
VAEncMacroblockDisableSkipMapBufferType | Force MB's to be non skip for encode.it's per-mb control buffer, The width of the MB map Surface is (width of the Picture in MB unit) * 1 byte, multiple of 64 bytes. The height is (height of the picture in MB unit). The picture is either frame or non-interleaved top or bottom field. If the application provides this surface, it will override the "skipCheckDisable" setting in VAEncMiscParameterEncQuality. |
VAEncFEICTBCmdBufferType | HEVC FEI CTB level cmd buffer it is CTB level information for future usage. |
VAEncFEICURecordBufferType | HEVC FEI CU level data buffer it's CTB level information for future usage. |
VADecodeStreamoutBufferType | decode stream out buffer, intermedia data of decode, it may include MV, MB mode etc. it can be used to detect motion and analyze the frame contain |
VASubsetsParameterBufferType | HEVC Decoding Subset Parameter buffer type. The subsets parameter buffer is concatenation with one or multiple subset entry point offsets. All the offset values are layed out one by one according to slice order with first slice segment first, second slice segment second, etc... The entry number is indicated by parameter num_entry_point_offsets. And the first entry position of the entry point offsets for any slice segment is indicated by parameter entry_offset_to_subset_array in VAPictureParameterBufferHEVC data structure. |
VAContextParameterUpdateBufferType | adjust context parameters dynamically this parameter is used to update context parameters, detail parameter is in |
VAProtectedSessionExecuteBufferType | Protected session execution buffer type. It's for TEE execution usage (vaProtectedSessionExecute()). The buffer structure is in |
VAEncryptionParameterBufferType | Encryption parameters buffer for protected content session. Refer to |
VAEncDeltaQpPerBlockBufferType | Encoding delta QP per block buffer. This buffer only could be created and accepted when |
VAAlfBufferType | VVC ALF data buffer. Refer to |
VALmcsBufferType | VVC LMCS data buffer. Refer to |
VASubPicBufferType | VVC SubPic data buffer. Refer to |
VATileBufferType | VVC Tile Dimension data buffer. Data buffer of tile widths and heights, with each element formatted as uint16_t |
VASliceStructBufferType | VVC Slice Structure data buffer. Refer to |
enum VAConfigAttribType |
Currently defined configuration attribute types
Enumerator | |
---|---|
VAConfigAttribDecSliceMode | Slice Decoding mode. Read/write. This attribute determines what mode the driver supports for slice decoding, through vaGetConfigAttributes(); and what mode the user will be providing to the driver, through vaCreateConfig(), if the driver supports those. If this attribute is not set by the user then it is assumed that VA_DEC_SLICE_MODE_NORMAL mode is used. See |
VAConfigAttribDecJPEG | JPEG decoding attribute. Read-only. This attribute exposes a number of capabilities of the underlying JPEG implementation. The attribute value is partitioned into fields as defined in the VAConfigAttribValDecJPEG union. |
VAConfigAttribDecProcessing | Decode processing support. Read/write. This attribute determines if the driver supports video processing with decoding using the decoding context in a single call, through vaGetConfigAttributes(); and if the user may use this feature, through vaCreateConfig(), if the driver supports the user scenario. The user will essentially create a regular decode VAContext. Therefore, the parameters of vaCreateContext() such as picture_width, picture_height and render_targets are in relation to the decode output parameters (not processing output parameters) as normal. If this attribute is not set by the user then it is assumed that no extra processing is done after decoding for this decode context. Since essentially the application is creating a decoder config and context, all function calls that take in the config (e.g. vaQuerySurfaceAttributes()) or context are in relation to the decoder, except those video processing function specified in the next paragraph. Once the decode config and context are created, the user must further query the supported processing filters using vaQueryVideoProcFilters(), vaQueryVideoProcFilterCaps(), vaQueryVideoProcPipelineCaps() by specifying the created decode context. The user must provide processing information and extra processing output surfaces as "additional_outputs" to the driver through VAProcPipelineParameterBufferType. The render_target specified at vaBeginPicture() time refers to the decode output surface. The target surface for the output of processing needs to be a different surface since the decode process requires the original reconstructed buffer. The "surface" member of VAProcPipelineParameterBuffer should be set to the same as "render_target" set in vaBeginPicture(), but the driver may choose to ignore this parameter. |
VAConfigAttribEncPackedHeaders | Packed headers mode. Read/write. This attribute determines what packed headers the driver supports, through vaGetConfigAttributes(); and what packed headers the user will be providing to the driver, through vaCreateConfig(), if the driver supports those. See |
VAConfigAttribEncInterlaced | Interlaced mode. Read/write. This attribute determines what kind of interlaced encoding mode the driver supports. See |
VAConfigAttribEncMaxRefFrames | Maximum number of reference frames. Read-only. This attribute determines the maximum number of reference frames supported for encoding. Note: for H.264 encoding, the value represents the maximum number of reference frames for both the reference picture list 0 (bottom 16 bits) and the reference picture list 1 (top 16 bits). |
VAConfigAttribEncMaxSlices | Maximum number of slices per frame. Read-only. This attribute determines the maximum number of slices the driver can support to encode a single frame. |
VAConfigAttribEncSliceStructure | Slice structure. Read-only. This attribute determines slice structures supported by the driver for encoding. This attribute is a hint to the user so that he can choose a suitable surface size and how to arrange the encoding process of multiple slices per frame. More specifically, for H.264 encoding, this attribute determines the range of accepted values to VAEncSliceParameterBufferH264::macroblock_address and VAEncSliceParameterBufferH264::num_macroblocks. See |
VAConfigAttribEncMacroblockInfo | Macroblock information. Read-only. This attribute determines whether the driver supports extra encoding information per-macroblock. e.g. QP. More specifically, for H.264 encoding, if the driver returns a non-zero value for this attribute, this means the application can create additional VAEncMacroblockParameterBufferH264 buffers referenced through VAEncSliceParameterBufferH264::macroblock_info. |
VAConfigAttribMaxPictureWidth | Maximum picture width. Read-only. This attribute determines the maximum picture width the driver supports for a given configuration. |
VAConfigAttribMaxPictureHeight | Maximum picture height. Read-only. This attribute determines the maximum picture height the driver supports for a given configuration. |
VAConfigAttribEncJPEG | JPEG encoding attribute. Read-only. This attribute exposes a number of capabilities of the underlying JPEG implementation. The attribute value is partitioned into fields as defined in the VAConfigAttribValEncJPEG union. |
VAConfigAttribEncQualityRange | Encoding quality range attribute. Read-only. This attribute conveys whether the driver supports different quality level settings for encoding. A value less than or equal to 1 means that the encoder only has a single quality setting, and a value greater than 1 represents the number of quality levels that can be configured. e.g. a value of 2 means there are two distinct quality levels. |
VAConfigAttribEncQuantization | Encoding quantization attribute. Read-only. This attribute conveys whether the driver supports certain types of quantization methods for encoding (e.g. trellis). See |
VAConfigAttribEncIntraRefresh | Encoding intra refresh attribute. Read-only. This attribute conveys whether the driver supports certain types of intra refresh methods for encoding (e.g. adaptive intra refresh or rolling intra refresh). See |
VAConfigAttribEncSkipFrame | Encoding skip frame attribute. Read-only. This attribute conveys whether the driver supports sending skip frame parameters (VAEncMiscParameterTypeSkipFrame) to the encoder's rate control, when the user has externally skipped frames. |
VAConfigAttribEncROI | Encoding region-of-interest (ROI) attribute. Read-only. This attribute conveys whether the driver supports region-of-interest (ROI) encoding, based on user provided ROI rectangles. The attribute value is partitioned into fields as defined in the VAConfigAttribValEncROI union. If ROI encoding is supported, the ROI information is passed to the driver using VAEncMiscParameterTypeROI. |
VAConfigAttribEncRateControlExt | Encoding extended rate control attribute. Read-only. This attribute conveys whether the driver supports any extended rate control features The attribute value is partitioned into fields as defined in the VAConfigAttribValEncRateControlExt union. |
VAConfigAttribProcessingRate | Processing rate reporting attribute. Read-only. This attribute conveys whether the driver supports reporting of encode/decode processing rate based on certain set of parameters (i.e. levels, I frame internvals) for a given configuration. If this is supported, vaQueryProcessingRate() can be used to get encode or decode processing rate. See |
VAConfigAttribEncDirtyRect | Encoding dirty rectangle. Read-only. This attribute conveys whether the driver supports dirty rectangle. encoding, based on user provided ROI rectangles which indicate the rectangular areas where the content has changed as compared to the previous picture. The regions of the picture that are not covered by dirty rect rectangles are assumed to have not changed compared to the previous picture. The encoder may do some optimizations based on this information. The attribute value returned indicates the number of regions that are supported. e.g. A value of 0 means dirty rect encoding is not supported. If dirty rect encoding is supported, the ROI information is passed to the driver using VAEncMiscParameterTypeDirtyRect. |
VAConfigAttribEncParallelRateControl | Parallel Rate Control (hierachical B) attribute. Read-only. This attribute conveys whether the encoder supports parallel rate control. It is a integer value 0 - unsupported, > 0 - maximum layer supported. This is the way when hireachical B frames are encoded, multiple independent B frames on the same layer may be processed at same time. If supported, app may enable it by setting enable_parallel_brc in VAEncMiscParameterRateControl,and the number of B frames per layer per GOP will be passed to driver through VAEncMiscParameterParallelRateControl structure.Currently three layers are defined. |
VAConfigAttribEncDynamicScaling | Dynamic Scaling Attribute. Read-only. This attribute conveys whether encoder is capable to determine dynamic frame resolutions adaptive to bandwidth utilization and processing power, etc. It is a boolean value 0 - unsupported, 1 - supported. If it is supported,for VP9, suggested frame resolution can be retrieved from VACodedBufferVP9Status. |
VAConfigAttribFrameSizeToleranceSupport | frame size tolerance support it indicates the tolerance of frame size |
VAConfigAttribFEIFunctionType | Encode function type for FEI. This attribute conveys whether the driver supports different function types for encode. It can be VA_FEI_FUNCTION_ENC, VA_FEI_FUNCTION_PAK, or VA_FEI_FUNCTION_ENC_PAK. Currently it is for FEI entry point only. Default is VA_FEI_FUNCTION_ENC_PAK. |
VAConfigAttribFEIMVPredictors | Maximum number of FEI MV predictors. Read-only. This attribute determines the maximum number of MV predictors the driver can support to encode a single frame. 0 means no MV predictor is supported. Currently it is for FEI entry point only. |
VAConfigAttribStats | Statistics attribute. Read-only. This attribute exposes a number of capabilities of the VAEntrypointStats entry point. The attribute value is partitioned into fields as defined in the VAConfigAttribValStats union. Currently it is for VAEntrypointStats only. |
VAConfigAttribEncTileSupport | Tile Support Attribute. Read-only. This attribute conveys whether encoder is capable to support tiles. If not supported, the tile related parameters sent to encoder, such as tiling structure, should be ignored. 0 - unsupported, 1 - supported. |
VAConfigAttribCustomRoundingControl | whether accept rouding setting from application. Read-only. This attribute is for encode quality, if it is report, application can change the rounding setting by VAEncMiscParameterTypeCustomRoundingControl |
VAConfigAttribQPBlockSize | Encoding QP info block size attribute. Read-only. This attribute conveys the block sizes that underlying driver support for QP info for buffer #VAEncQpBuffer. |
VAConfigAttribMaxFrameSize | encode max frame size attribute. Read-only attribute value |
VAConfigAttribPredictionDirection | inter frame prediction directrion attribute. Read-only. this attribute conveys the prediction direction (backward or forword) for specific config the value could be VA_PREDICTION_DIRECTION_XXXX. it can be combined with VAConfigAttribEncMaxRefFrames to describe reference list , and the prediction direction. if this attrib is not present,both direction should be supported, no restriction. for example: normal HEVC encoding , maximum reference frame number in reflist 0 and reflist 1 is deduced by VAConfigAttribEncMaxRefFrames. so there are typical P frame, B frame, if VAConfigAttribPredictionDirection is also present. it will stipulate prediction direction in both reference list. if only one prediction direction present(such as PREVIOUS),all reference frame should be previous frame (PoC < current). |
VAConfigAttribMultipleFrame | combined submission of multiple frames from different streams, it is optimization for different HW implementation, multiple frames encode/decode can improve HW concurrency |
VAConfigAttribContextPriority | priority setting for the context. Read-Write attribute value is |
VAConfigAttribDecAV1Features | AV1 decoding features. Read-only. This attribute describes the supported features of an AV1 decoder configuration. The value returned uses the VAConfigAttribValDecAV1Features type. |
VAConfigAttribTEEType | TEE could be any HW secure device. Read-only. |
VAConfigAttribTEETypeClient | TEE type client is a specific module supporting specific functions in TEE. Read-only. |
VAConfigAttribProtectedContentCipherAlgorithm | Cipher algorithm of the protected content session. This attribute specifies the cipher algorithm of the protected content session. It could be |
VAConfigAttribProtectedContentCipherBlockSize | Cipher block size of the protected content session. This attribute specifies the block size of the protected content session. It could be |
VAConfigAttribProtectedContentCipherMode | Cipher mode of the protected content session. This attribute specifies the cipher mode of the protected content session. It could be |
VAConfigAttribProtectedContentCipherSampleType | Decryption sample type of the protected content session. This attribute specifies the decryption sample type of the protected content session. It could be |
VAConfigAttribProtectedContentUsage | Special usage attribute of the protected session. The attribute specifies the flow for the protected session could be used. For example, it could be |
VAConfigAttribEncHEVCFeatures | HEVC/H.265 encoding features. Read-only. This attribute describes the supported features of an HEVC/H.265 encoder configuration. The value returned uses the VAConfigAttribValEncHEVCFeatures type. If this attribute is supported by a driver then it must also support the VAConfigAttribEncHEVCBlockSizes attribute. |
VAConfigAttribEncHEVCBlockSizes | HEVC/H.265 encoding block sizes. Read-only. This attribute describes the supported coding tree and transform block sizes of an HEVC/H.265 encoder configuration. The value returned uses the VAConfigAttribValEncHEVCBlockSizes type. If this attribute is supported by a driver then it must also support the VAConfigAttribEncHEVCFeatures attribute. |
VAConfigAttribEncAV1 | AV1 encoding attribute. Read-only. This attribute exposes a number of capabilities of the underlying AV1 implementation. The attribute value is partitioned into fields as defined in the VAConfigAttribValEncAV1 union. |
VAConfigAttribEncAV1Ext1 | AV1 encoding attribute extend1. Read-only. This attribute exposes a number of capabilities of the underlying AV1 implementation. The attribute value is partitioned into fields as defined in the VAConfigAttribValEncAV1Ext1 union. |
VAConfigAttribEncAV1Ext2 | AV1 encoding attribute extend2. Read-only. This attribute exposes a number of capabilities of the underlying AV1 implementation. The attribute value is partitioned into fields as defined in the VAConfigAttribValEncAV1Ext2 union. |
VAConfigAttribEncPerBlockControl | Settings per block attribute for Encoding. Read-only. This attribute describes whether to support delta qp per block, the supported size of delta qp block and the size of delta QP in bytes. The value returned uses the VAConfigAttribValEncPerBlockControl type. |
VAConfigAttribEncMaxTileRows | Maximum number of tile rows. Read-only. This attribute determines the maximum number of tile rows supported for encoding with tile support. |
VAConfigAttribEncMaxTileCols | Maximum number of tile cols. Read-only. This attribute determines the maximum number of tile columns supported for encoding with tile support. |
Display attributes Display attributes are used to control things such as contrast, hue, saturation, brightness etc. in the rendering process. The application can query what attributes are supported by the driver, and then set the appropriate attributes before calling vaPutSurface()
Display attributes can also be used to query/set platform or display adaptor (vaDisplay) related information. These attributes do not depend on vaConfig, and could not be used for vaPutSurface. Application can use vaQueryDisplayAttributes/vaGetDisplayAttributes at anytime after vaInitialize, but (for settable attributes) vaSetDisplayAttributes should be called after vaInitialize and before any other function call.
To distinguish these two types of display attributes, display adaptor related attributes should be marked as "HW attribute" in the description.
Packed header type.
Enumerator | |
---|---|
VAEncPackedHeaderSequence | Packed sequence header. |
VAEncPackedHeaderPicture | Packed picture header. |
VAEncPackedHeaderSlice | Packed slice header. |
VAEncPackedHeaderRawData | Packed raw header. Packed raw data header can be used by the client to insert a header into the bitstream data buffer at the point it is passed, the driver will handle the raw packed header based on "has_emulation_bytes" field in the packed header parameter structure. |
va_deprecated_enum | Misc packed header. See codec-specific definitions.
|
enum VAEntrypoint |
Currently defined entrypoints
Enumerator | |
---|---|
VAEntrypointVideoProc | Video pre/post-processing. |
VAEntrypointFEI | VAEntrypointFEI. The purpose of FEI (Flexible Encoding Infrastructure) is to allow applications to have more controls and trade off quality for speed with their own IPs. The application can optionally provide input to ENC for extra encode control and get the output from ENC. Application can chose to modify the ENC output/PAK input during encoding, but the performance impact is significant. On top of the existing buffers for normal encode, there will be one extra input buffer (VAEncMiscParameterFEIFrameControl) and three extra output buffers (VAEncFEIMVBufferType, VAEncFEIMBModeBufferType and VAEncFEIDistortionBufferType) for VAEntrypointFEI entry function. If separate PAK is set, two extra input buffers (VAEncFEIMVBufferType, VAEncFEIMBModeBufferType) are needed for PAK input. |
VAEntrypointStats | VAEntrypointStats. A pre-processing function for getting some statistics and motion vectors is added, and some extra controls for Encode pipeline are provided. The application can optionally call the statistics function to get motion vectors and statistics like variances, distortions before calling Encode function via this entry point. Checking whether Statistics is supported can be performed with vaQueryConfigEntrypoints(). If Statistics entry point is supported, then the list of returned entry-points will include VAEntrypointStats. Supported pixel format, maximum resolution and statistics specific attributes can be obtained via normal attribute query. One input buffer (VAStatsStatisticsParameterBufferType) and one or two output buffers (VAStatsStatisticsBufferType, VAStatsStatisticsBottomFieldBufferType (for interlace only) and VAStatsMVBufferType) are needed for this entry point. |
VAEntrypointProtectedTEEComm | VAEntrypointProtectedTEEComm. A function for communicating with TEE (Trusted Execution Environment). |
VAEntrypointProtectedContent | VAEntrypointProtectedContent. A function for protected content to decrypt encrypted content. |
enum VAGenericValueType |
enum VAMvModeVC1 |
VC-1 data structures
enum VAProfile |
Currently defined profiles
Enumerator | |
---|---|
VAProfileNone | Profile ID used for video processing. |
va_deprecated_enum | Misc packed header. See codec-specific definitions.
|
VAProfileProtected | Profile ID used for protected video playback. |
enum VASurfaceAttribType |
Surface attribute types.
Enumerator | |
---|---|
VASurfaceAttribPixelFormat | Pixel format as a FOURCC (int, read/write). When vaQuerySurfaceAttributes() is called, the driver will return one PixelFormat attribute per supported pixel format. When provided as an input to vaCreateSurfaces(), the driver will allocate a surface with the provided pixel format. |
VASurfaceAttribMinWidth | Minimal width in pixels (int, read-only). |
VASurfaceAttribMaxWidth | Maximal width in pixels (int, read-only). |
VASurfaceAttribMinHeight | Minimal height in pixels (int, read-only). |
VASurfaceAttribMaxHeight | Maximal height in pixels (int, read-only). |
VASurfaceAttribMemoryType | Surface memory type expressed in bit fields (int, read/write). |
VASurfaceAttribExternalBufferDescriptor | External buffer descriptor (pointer, write). Refer to the documentation for the memory type being created to determine what descriptor structure to pass here. If not otherwise stated, the common VASurfaceAttribExternalBuffers should be used. |
VASurfaceAttribUsageHint | Surface usage hint, gives the driver a hint of intended usage to optimize allocation (e.g. tiling) (int, read/write). |
VASurfaceAttribDRMFormatModifiers | List of possible DRM format modifiers (pointer, write). The value must be a pointer to a VADRMFormatModifierList. This can only be used when allocating a new buffer, it's invalid to use this attribute when importing an existing buffer. |
VASurfaceAttribAlignmentSize | width and height log2 aligment in pixels (int, read-only) For special HW requirement used in some codecs, if VASurfaceAttribAlignmentSize is not implemented in the driver, then the surface_width and surface_height should keep the original logic without any modification, this is an add-on requirement to surface_width and surface_height. |
VASurfaceAttribCount | Number of surface attributes. |
VAStatus vaAcquireBufferHandle | ( | VADisplay | dpy, |
VABufferID | buf_id, | ||
VABufferInfo * | buf_info | ||
) |
Acquires buffer handle for external API usage.
Locks the VA buffer object buf_id for external API usage like EGL or OpenCL (OCL). This function is a synchronization point. This means that any pending operation is guaranteed to be completed prior to returning from the function.
If the referenced VA buffer object is the backing store of a VA surface, then this function acts as if vaSyncSurface() on the parent surface was called first.
The VABufferInfo argument shall be zero'ed on input. On successful output, the data structure is filled in with all the necessary buffer level implementation details like handle, type, memory type and memory size.
Note: the external API implementation, or the application, can express the memory types it is interested in by filling in the mem_type field accordingly. On successful output, the memory type that fits best the request and that was used is updated in the VABufferInfo data structure. If none of the supplied memory types is supported, then a VA_STATUS_ERROR_UNSUPPORTED_MEMORY_TYPE error is returned.
The VABufferInfo data is valid until vaReleaseBufferHandle() is called. Besides, no additional operation is allowed on any of the buffer parent object until vaReleaseBufferHandle() is called. e.g. decoding into a VA surface backed with the supplied VA buffer object buf_id would fail with a VA_STATUS_ERROR_SURFACE_BUSY error.
Possible errors:
[in] | dpy | the VA display |
[in] | buf_id | the VA buffer |
[in,out] | buf_info | the associated VA buffer information |
VAStatus vaAssociateSubpicture | ( | VADisplay | dpy, |
VASubpictureID | subpicture, | ||
VASurfaceID * | target_surfaces, | ||
int | num_surfaces, | ||
int16_t | src_x, | ||
int16_t | src_y, | ||
uint16_t | src_width, | ||
uint16_t | src_height, | ||
int16_t | dest_x, | ||
int16_t | dest_y, | ||
uint16_t | dest_width, | ||
uint16_t | dest_height, | ||
uint32_t | flags | ||
) |
vaAssociateSubpicture associates the subpicture with target_surfaces. It defines the region mapping between the subpicture and the target surfaces through source and destination rectangles (with the same width and height). Both will be displayed at the next call to vaPutSurface. Additional associations before the call to vaPutSurface simply overrides the association.
VAStatus vaBeginPicture | ( | VADisplay | dpy, |
VAContextID | context, | ||
VASurfaceID | render_target | ||
) |
Render (Video Decode/Encode/Processing) Pictures
A picture represents either a frame or a field.
The Begin/Render/End sequence sends the video decode/encode/processing buffers to the server Get ready for a video pipeline
VAStatus vaBufferSetNumElements | ( | VADisplay | dpy, |
VABufferID | buf_id, | ||
unsigned int | num_elements | ||
) |
Convey to the server how many valid elements are in the buffer. e.g. if multiple slice parameters are being held in a single buffer, this will communicate to the server the number of slice parameters that are valid in the buffer.
VAStatus vaCreateBuffer | ( | VADisplay | dpy, |
VAContextID | context, | ||
VABufferType | type, | ||
unsigned int | size, | ||
unsigned int | num_elements, | ||
void * | data, | ||
VABufferID * | buf_id | ||
) |
Buffer functions Creates a buffer for "num_elements" elements of "size" bytes and initalize with "data". if "data" is null, then the contents of the buffer data store are undefined. Basically there are two ways to get buffer data to the server side. One is to call vaCreateBuffer() with a non-null "data", which results the data being copied to the data store on the server side. A different method that eliminates this copy is to pass null as "data" when calling vaCreateBuffer(), and then use vaMapBuffer() to map the data store from the server side to the client address space for access. The user must call vaDestroyBuffer() to destroy a buffer. Note: image buffers are created by the library, not the client. Please see vaCreateImage on how image buffers are managed.
VAStatus vaCreateBuffer2 | ( | VADisplay | dpy, |
VAContextID | context, | ||
VABufferType | type, | ||
unsigned int | width, | ||
unsigned int | height, | ||
unsigned int * | unit_size, | ||
unsigned int * | pitch, | ||
VABufferID * | buf_id | ||
) |
Create a buffer for given width & height get unit_size, pitch, buf_id for 2D buffer for permb qp buffer, it will return unit_size for one MB or LCU and the pitch for alignments can call vaMapBuffer with this Buffer ID to get virtual address. e.g. AVC 1080P encode, 1920x1088, the size in MB is 120x68,but inside driver, maybe it should align with 256, and one byte present one Qp.so, call the function. then get unit_size = 1, pitch = 256. call vaMapBuffer to get the virtual address (pBuf). then read write the memory like 2D. the size is 256x68, application can only use 120x68 pBuf + 256 is the start of next line. different driver implementation maybe return different unit_size and pitch
VAStatus vaCreateConfig | ( | VADisplay | dpy, |
VAProfile | profile, | ||
VAEntrypoint | entrypoint, | ||
VAConfigAttrib * | attrib_list, | ||
int | num_attribs, | ||
VAConfigID * | config_id | ||
) |
Create a configuration for the video decode/encode/processing pipeline it passes in the attribute list that specifies the attributes it cares about, with the rest taking default values.
VAStatus vaCreateContext | ( | VADisplay | dpy, |
VAConfigID | config_id, | ||
int | picture_width, | ||
int | picture_height, | ||
int | flag, | ||
VASurfaceID * | render_targets, | ||
int | num_render_targets, | ||
VAContextID * | context | ||
) |
vaCreateContext - Create a context dpy: display config_id: configuration for the context picture_width: coded picture width picture_height: coded picture height flag: any combination of the following: VA_PROGRESSIVE (only progressive frame pictures in the sequence when set) render_targets: a hint for render targets (surfaces) tied to the context num_render_targets: number of render targets in the above array context: created context id upon return
VAStatus vaCreateImage | ( | VADisplay | dpy, |
VAImageFormat * | format, | ||
int | width, | ||
int | height, | ||
VAImage * | image | ||
) |
Create a VAImage structure The width and height fields returned in the VAImage structure may get enlarged for some YUV formats. Upon return from this function, image->buf has been created and proper storage allocated by the library. The client can access the image through the Map/Unmap calls.
VAStatus vaCreateMFContext | ( | VADisplay | dpy, |
VAMFContextID * | mf_context | ||
) |
vaCreateMFContext - Create a multi-frame context interface encapsulating common for all streams memory objects and structures required for single GPU task submission from several VAContextID's. Allocation: This call only creates an instance, doesn't allocate any additional memory. Support identification: Application can identify multi-frame feature support by ability to create multi-frame context. If driver supports multi-frame - call successful, mf_context != NULL and VAStatus = VA_STATUS_SUCCESS, otherwise if multi-frame processing not supported driver returns VA_STATUS_ERROR_UNIMPLEMENTED and mf_context = NULL. return values: VA_STATUS_SUCCESS - operation successful. VA_STATUS_ERROR_UNIMPLEMENTED - no support for multi-frame. dpy: display adapter. mf_context: Multi-Frame context encapsulating all associated context for multi-frame submission.
VAStatus vaCreateSubpicture | ( | VADisplay | dpy, |
VAImageID | image, | ||
VASubpictureID * | subpicture | ||
) |
Subpictures are created with an image associated.
VAStatus vaCreateSurfaces | ( | VADisplay | dpy, |
unsigned int | format, | ||
unsigned int | width, | ||
unsigned int | height, | ||
VASurfaceID * | surfaces, | ||
unsigned int | num_surfaces, | ||
VASurfaceAttrib * | attrib_list, | ||
unsigned int | num_attribs | ||
) |
Creates an array of surfaces.
Creates an array of surfaces. The optional list of attributes shall be constructed based on what the underlying hardware could expose through vaQuerySurfaceAttributes().
[in] | dpy | the VA display |
[in] | format | the desired surface format. See VA_RT_FORMAT_* |
[in] | width | the surface width |
[in] | height | the surface height |
[out] | surfaces | the array of newly created surfaces |
[in] | num_surfaces | the number of surfaces to create |
[in] | attrib_list | the list of (optional) attributes, or NULL |
[in] | num_attribs | the number of attributes supplied in attrib_list , or zero |
VAStatus vaDeassociateSubpicture | ( | VADisplay | dpy, |
VASubpictureID | subpicture, | ||
VASurfaceID * | target_surfaces, | ||
int | num_surfaces | ||
) |
vaDeassociateSubpicture removes the association of the subpicture with target_surfaces.
VAStatus vaDeriveImage | ( | VADisplay | dpy, |
VASurfaceID | surface, | ||
VAImage * | image | ||
) |
Derive an VAImage from an existing surface. This interface will derive a VAImage and corresponding image buffer from an existing VA Surface. The image buffer can then be mapped/unmapped for direct CPU access. This operation is only possible on implementations with direct rendering capabilities and internal surface formats that can be represented with a VAImage. When the operation is not possible this interface will return VA_STATUS_ERROR_OPERATION_FAILED. Clients should then fall back to using vaCreateImage + vaPutImage to accomplish the same task in an indirect manner.
Implementations should only return success when the resulting image buffer would be useable with vaMap/Unmap.
When directly accessing a surface special care must be taken to insure proper synchronization with the graphics hardware. Clients should call vaQuerySurfaceStatus to insure that a surface is not the target of concurrent rendering or currently being displayed by an overlay.
Additionally nothing about the contents of a surface should be assumed following a vaPutSurface. Implementations are free to modify the surface for scaling or subpicture blending within a call to vaPutImage.
Calls to vaPutImage or vaGetImage using the same surface from which the image has been derived will return VA_STATUS_ERROR_SURFACE_BUSY. vaPutImage or vaGetImage with other surfaces is supported.
An image created with vaDeriveImage should be freed with vaDestroyImage. The image and image buffer structures will be destroyed; however, the underlying surface will remain unchanged until freed with vaDestroySurfaces.
VAStatus vaDestroyBuffer | ( | VADisplay | dpy, |
VABufferID | buffer_id | ||
) |
After this call, the buffer is deleted and this buffer_id is no longer valid
A buffer can be re-used and sent to the server by another Begin/Render/End sequence if vaDestroyBuffer() is not called with this buffer.
Note re-using a shared buffer (e.g. a slice data buffer) between the host and the hardware accelerator can result in performance dropping.
VAStatus vaDestroyConfig | ( | VADisplay | dpy, |
VAConfigID | config_id | ||
) |
Free resources associdated with a given config
VAStatus vaDestroyContext | ( | VADisplay | dpy, |
VAContextID | context | ||
) |
vaDestroyContext - Destroy a context dpy: display context: context to be destroyed
VAStatus vaDestroyImage | ( | VADisplay | dpy, |
VAImageID | image | ||
) |
Should call DestroyImage before destroying the surface it is bound to
VAStatus vaDestroySubpicture | ( | VADisplay | dpy, |
VASubpictureID | subpicture | ||
) |
Destroy the subpicture before destroying the image it is assocated to
VAStatus vaDestroySurfaces | ( | VADisplay | dpy, |
VASurfaceID * | surfaces, | ||
int | num_surfaces | ||
) |
vaDestroySurfaces - Destroy resources associated with surfaces. Surfaces can only be destroyed after all contexts using these surfaces have been destroyed. dpy: display surfaces: array of surfaces to destroy num_surfaces: number of surfaces in the array to be destroyed.
VAStatus vaEndPicture | ( | VADisplay | dpy, |
VAContextID | context | ||
) |
Make the end of rendering for a picture. The server should start processing all pending operations for this surface. This call is non-blocking. The client can start another Begin/Render/End sequence on a different render target. if VAContextID used in this function previously successfully passed vaMFAddContext call, real processing will be started during vaMFSubmit
const char* vaErrorStr | ( | VAStatus | error_status | ) |
Returns a short english description of error_status
VAStatus vaExportSurfaceHandle | ( | VADisplay | dpy, |
VASurfaceID | surface_id, | ||
uint32_t | mem_type, | ||
uint32_t | flags, | ||
void * | descriptor | ||
) |
Export a handle to a surface for use with an external API.
The exported handles are owned by the caller, and the caller is responsible for freeing them when no longer needed (e.g. by closing DRM PRIME file descriptors).
This does not perform any synchronisation. If the contents of the surface will be read, vaSyncSurface() must be called before doing so. If the contents of the surface are written, then all operations must be completed externally before using the surface again by via VA-API functions.
[in] | dpy | VA display. |
[in] | surface_id | Surface to export. |
[in] | mem_type | Memory type to export to. |
[in] | flags | Combination of flags to apply (VA_EXPORT_SURFACE_*). |
[out] | descriptor | Pointer to the descriptor structure to fill with the handle details. The type of this structure depends on the value of mem_type. |
VAStatus vaGetConfigAttributes | ( | VADisplay | dpy, |
VAProfile | profile, | ||
VAEntrypoint | entrypoint, | ||
VAConfigAttrib * | attrib_list, | ||
int | num_attribs | ||
) |
Get attributes for a given profile/entrypoint pair The caller must provide an "attrib_list" with all attributes to be retrieved. Upon return, the attributes in "attrib_list" have been updated with their value. Unknown attributes or attributes that are not supported for the given profile/entrypoint pair will have their value set to VA_ATTRIB_NOT_SUPPORTED
VAStatus vaGetImage | ( | VADisplay | dpy, |
VASurfaceID | surface, | ||
int | x, | ||
int | y, | ||
unsigned int | width, | ||
unsigned int | height, | ||
VAImageID | image | ||
) |
Retrive surface data into a VAImage Image must be in a format supported by the implementation
VAPrivFunc vaGetLibFunc | ( | VADisplay | dpy, |
const char * | func | ||
) |
Return a function pointer given a function name in the library. This allows private interfaces into the library
VAStatus vaInitialize | ( | VADisplay | dpy, |
int * | major_version, | ||
int * | minor_version | ||
) |
Initialize the library
VAStatus vaMapBuffer | ( | VADisplay | dpy, |
VABufferID | buf_id, | ||
void ** | pbuf | ||
) |
Map data store of the buffer into the client's address space vaCreateBuffer() needs to be called with "data" set to NULL before calling vaMapBuffer()
if buffer type is VAEncCodedBufferType, pbuf points to link-list of VACodedBufferSegment, and the list is terminated if "next" is NULL
int vaMaxNumConfigAttributes | ( | VADisplay | dpy | ) |
Get maximum number of attributs supported by the implementation
int vaMaxNumEntrypoints | ( | VADisplay | dpy | ) |
Get maximum number of entrypoints supported by the implementation
int vaMaxNumImageFormats | ( | VADisplay | dpy | ) |
Get maximum number of image formats supported by the implementation
int vaMaxNumProfiles | ( | VADisplay | dpy | ) |
Get maximum number of profiles supported by the implementation
int vaMaxNumSubpictureFormats | ( | VADisplay | dpy | ) |
Get maximum number of subpicture formats supported by the implementation
VAStatus vaMFAddContext | ( | VADisplay | dpy, |
VAMFContextID | mf_context, | ||
VAContextID | context | ||
) |
vaMFAddContext - Provide ability to associate each context used for Multi-Frame submission and common Multi-Frame context. Try to add context to understand if it is supported. Allocation: this call allocates and/or reallocates all memory objects common for all contexts associated with particular Multi-Frame context. All memory required for each context(pixel buffers, internal driver buffers required for processing) allocated during standard vaCreateContext call for each context. Runtime dependency - if current implementation doesn't allow to run different entry points/profile, first context added will set entry point/profile for whole Multi-Frame context, all other entry points and profiles can be rejected to be added. Return values: VA_STATUS_SUCCESS - operation successful, context was added. VA_STATUS_ERROR_OPERATION_FAILED - something unexpected happened - application have to close current mf_context and associated contexts and start working with new ones. VA_STATUS_ERROR_INVALID_CONTEXT - ContextID is invalid, means: 1 - mf_context is not valid context or 2 - driver can't suport different VAEntrypoint or VAProfile simultaneosly and current context contradicts with previously added, application can continue with current mf_context and other contexts passed this call, rejected context can continue work in stand-alone mode or other mf_context. VA_STATUS_ERROR_UNSUPPORTED_ENTRYPOINT - particular context being added was created with with unsupported VAEntrypoint. Application can continue with current mf_context and other contexts passed this call, rejected context can continue work in stand-alone mode. VA_STATUS_ERROR_UNSUPPORTED_PROFILE - Current context with Particular VAEntrypoint is supported but VAProfile is not supported. Application can continue with current mf_context and other contexts passed this call, rejected context can continue work in stand-alone mode. dpy: display adapter. context: context being associated with Multi-Frame context. mf_context: - multi-frame context used to associate contexts for multi-frame submission.
VAStatus vaMFReleaseContext | ( | VADisplay | dpy, |
VAMFContextID | mf_context, | ||
VAContextID | context | ||
) |
vaMFReleaseContext - Removes context from multi-frame and association with multi-frame context. After association removed vaEndPicture will submit tasks, but not vaMFSubmit. Return values: VA_STATUS_SUCCESS - operation successful, context was removed. VA_STATUS_ERROR_OPERATION_FAILED - something unexpected happened. application need to destroy this VAMFContextID and all assotiated VAContextID dpy: display mf_context: VAMFContextID where context is added context: VAContextID to be added
VAStatus vaMFSubmit | ( | VADisplay | dpy, |
VAMFContextID | mf_context, | ||
VAContextID * | contexts, | ||
int | num_contexts | ||
) |
Make the end of rendering for a pictures in contexts passed with submission. The server should start processing all pending operations for contexts. All contexts passed should be associated through vaMFAddContext and call sequence Begin/Render/End performed. This call is non-blocking. The client can start another Begin/Render/End/vaMFSubmit sequence on a different render targets. Return values: VA_STATUS_SUCCESS - operation successful, context was removed. VA_STATUS_ERROR_INVALID_CONTEXT - mf_context or one of contexts are invalid due to mf_context not created or one of contexts not assotiated with mf_context through vaAddContext. VA_STATUS_ERROR_INVALID_PARAMETER - one of context has not submitted it's frame through vaBeginPicture vaRenderPicture vaEndPicture call sequence. dpy: display mf_context: Multi-Frame context contexts: list of contexts submitting their tasks for multi-frame operation. num_contexts: number of passed contexts.
VAStatus vaPutImage | ( | VADisplay | dpy, |
VASurfaceID | surface, | ||
VAImageID | image, | ||
int | src_x, | ||
int | src_y, | ||
unsigned int | src_width, | ||
unsigned int | src_height, | ||
int | dest_x, | ||
int | dest_y, | ||
unsigned int | dest_width, | ||
unsigned int | dest_height | ||
) |
Copy data from a VAImage to a surface Image must be in a format supported by the implementation Returns a VA_STATUS_ERROR_SURFACE_BUSY if the surface shouldn't be rendered into when this is called
VAStatus vaQueryConfigAttributes | ( | VADisplay | dpy, |
VAConfigID | config_id, | ||
VAProfile * | profile, | ||
VAEntrypoint * | entrypoint, | ||
VAConfigAttrib * | attrib_list, | ||
int * | num_attribs | ||
) |
Query all attributes for a given configuration The profile of the configuration is returned in "profile" The entrypoint of the configuration is returned in "entrypoint" The caller must provide an "attrib_list" array that can hold at least vaMaxNumConfigAttributes() entries. The actual number of attributes returned in "attrib_list" is returned in "num_attribs"
VAStatus vaQueryConfigEntrypoints | ( | VADisplay | dpy, |
VAProfile | profile, | ||
VAEntrypoint * | entrypoint_list, | ||
int * | num_entrypoints | ||
) |
Query supported entrypoints for a given profile The caller must provide an "entrypoint_list" array that can hold at least vaMaxNumEntrypoints() entries. The actual number of entrypoints returned in "entrypoint_list" is returned in "num_entrypoints".
Query supported profiles The caller must provide a "profile_list" array that can hold at least vaMaxNumProfile() entries. The actual number of profiles returned in "profile_list" is returned in "num_profile".
VAStatus vaQueryImageFormats | ( | VADisplay | dpy, |
VAImageFormat * | format_list, | ||
int * | num_formats | ||
) |
Query supported image formats The caller must provide a "format_list" array that can hold at least vaMaxNumImageFormats() entries. The actual number of formats returned in "format_list" is returned in "num_formats".
VAStatus vaQueryProcessingRate | ( | VADisplay | dpy, |
VAConfigID | config, | ||
VAProcessingRateParameter * | proc_buf, | ||
unsigned int * | processing_rate | ||
) |
Queries processing rate for the supplied config.
This function queries the processing rate based on parameters in proc_buf
for the given config
. Upon successful return, the processing rate value will be stored in processing_rate
. Processing rate is specified as the number of macroblocks/CTU per second.
If NULL is passed to the proc_buf
, the default processing rate for the given configuration will be returned.
[in] | dpy | the VA display |
[in] | config | the config identifying a codec or a video processing pipeline |
[in] | proc_buf | the buffer that contains the parameters for either the encode or decode processing rate |
[out] | processing_rate | processing rate in number of macroblocks per second constrained by parameters specified in proc_buf |
VAStatus vaQuerySubpictureFormats | ( | VADisplay | dpy, |
VAImageFormat * | format_list, | ||
unsigned int * | flags, | ||
unsigned int * | num_formats | ||
) |
Query supported subpicture formats The caller must provide a "format_list" array that can hold at least vaMaxNumSubpictureFormats() entries. The flags arrary holds the flag for each format to indicate additional capabilities for that format. The actual number of formats returned in "format_list" is returned in "num_formats". flags: returned value to indicate addtional capabilities VA_SUBPICTURE_CHROMA_KEYING - supports chroma-keying VA_SUBPICTURE_GLOBAL_ALPHA - supports global alpha VA_SUBPICTURE_DESTINATION_IS_SCREEN_COORD - supports unscaled screen relative subpictures for On Screen Display
VAStatus vaQuerySurfaceAttributes | ( | VADisplay | dpy, |
VAConfigID | config, | ||
VASurfaceAttrib * | attrib_list, | ||
unsigned int * | num_attribs | ||
) |
Queries surface attributes for the supplied config.
This function queries for all supported attributes for the supplied VA config
. In particular, if the underlying hardware supports the creation of VA surfaces in various formats, then this function will enumerate all pixel formats that are supported.
The attrib_list
array is allocated by the user and num_attribs
shall be initialized to the number of allocated elements in that array. Upon successful return, the actual number of attributes will be overwritten into num_attribs
. Otherwise, VA_STATUS_ERROR_MAX_NUM_EXCEEDED
is returned and num_attribs
is adjusted to the number of elements that would be returned if enough space was available.
Note: it is perfectly valid to pass NULL to the attrib_list
argument when vaQuerySurfaceAttributes() is used to determine the actual number of elements that need to be allocated.
[in] | dpy | the VA display |
[in] | config | the config identifying a codec or a video processing pipeline |
[out] | attrib_list | the output array of VASurfaceAttrib elements |
[in,out] | num_attribs | the number of elements allocated on input, the number of elements actually filled in output |
VAStatus vaQuerySurfaceError | ( | VADisplay | dpy, |
VASurfaceID | surface, | ||
VAStatus | error_status, | ||
void ** | error_info | ||
) |
After the application gets VA_STATUS_ERROR_DECODING_ERROR after calling vaSyncSurface(), it can call vaQuerySurfaceError to find out further details on the particular error. VA_STATUS_ERROR_DECODING_ERROR should be passed in as "error_status".
After the applications get VA_STATUS_HW_BUSY or VA_STATUS_SUCCESSFULL from vaSyncSurface(), it still can call vaQuerySurfaceError to find out further details to know if has real hw reset happened on this surface since umd and kmd could recover the context from reset with success in sometimes. VA_STATUS_HW_BUSY or VA_STATUS_SUCCESSFULL also could be passed in as "error_status".
Upon the return, error_info will point to an array of _VASurfaceDecodeMBErrors structure, which is allocated and filled by libVA with detailed information on the VADecodeErrorType. The array is terminated if "status==-1" is detected.
VAStatus vaQuerySurfaceStatus | ( | VADisplay | dpy, |
VASurfaceID | render_target, | ||
VASurfaceStatus * | status | ||
) |
Find out any pending ops on the render target
const char* vaQueryVendorString | ( | VADisplay | dpy | ) |
vaQueryVendorString returns a pointer to a zero-terminated string describing some aspects of the VA implemenation on a specific hardware accelerator. The format of the returned string is vendor specific and at the discretion of the implementer. e.g. for the Intel GMA500 implementation, an example would be: "Intel GMA500 - 2.0.0.32L.0005"
VAStatus vaReleaseBufferHandle | ( | VADisplay | dpy, |
VABufferID | buf_id | ||
) |
Releases buffer after usage from external API.
Unlocks the VA buffer object buf_id from external API usage like EGL or OpenCL (OCL). This function is a synchronization point. This means that any pending operation is guaranteed to be completed prior to returning from the function.
The VABufferInfo argument shall point to the original data structure that was obtained from vaAcquireBufferHandle(), unaltered. This is necessary so that the VA driver implementation could deallocate any resources that were needed.
In any case, returning from this function invalidates any contents in VABufferInfo. i.e. the underlyng buffer handle is no longer valid. Therefore, VA driver implementations are free to reset this data structure to safe defaults.
Possible errors:
[in] | dpy | the VA display |
[in] | buf_id | the VA buffer |
VAStatus vaRenderPicture | ( | VADisplay | dpy, |
VAContextID | context, | ||
VABufferID * | buffers, | ||
int | num_buffers | ||
) |
Send video decode, encode or processing buffers to the server.
VAStatus vaSetDriverName | ( | VADisplay | dpy, |
char * | driver_name | ||
) |
Set the override driver name instead of queried driver driver.
VAMessageCallback vaSetErrorCallback | ( | VADisplay | dpy, |
VAMessageCallback | callback, | ||
void * | user_context | ||
) |
Set the callback for error messages, or NULL for no logging. Returns the previous one, or NULL if it was disabled.
VAMessageCallback vaSetInfoCallback | ( | VADisplay | dpy, |
VAMessageCallback | callback, | ||
void * | user_context | ||
) |
Set the callback for info messages, or NULL for no logging. Returns the previous one, or NULL if it was disabled.
VAStatus vaSetSubpictureChromakey | ( | VADisplay | dpy, |
VASubpictureID | subpicture, | ||
unsigned int | chromakey_min, | ||
unsigned int | chromakey_max, | ||
unsigned int | chromakey_mask | ||
) |
If chromakey is enabled, then the area where the source value falls within the chromakey [min, max] range is transparent The chromakey component format is the following: For RGB: [0:7] Red [8:15] Blue [16:23] Green For YUV: [0:7] V [8:15] U [16:23] Y The chromakey mask can be used to mask out certain components for chromakey comparision
VAStatus vaSetSubpictureGlobalAlpha | ( | VADisplay | dpy, |
VASubpictureID | subpicture, | ||
float | global_alpha | ||
) |
Global alpha value is between 0 and 1. A value of 1 means fully opaque and a value of 0 means fully transparent. If per-pixel alpha is also specified then the overall alpha is per-pixel alpha multiplied by the global alpha
VAStatus vaSetSubpictureImage | ( | VADisplay | dpy, |
VASubpictureID | subpicture, | ||
VAImageID | image | ||
) |
Bind an image to the subpicture. This image will now be associated with the subpicture instead of the one at creation.
VAStatus vaSyncBuffer | ( | VADisplay | dpy, |
VABufferID | buf_id, | ||
uint64_t | timeout_ns | ||
) |
Synchronizes pending operations associated with the supplied buffer.
This function blocks during specified timeout (in nanoseconds) until all pending operations on the supplied buffer have been completed. If timeout is zero, the function returns immediately.
Possible errors:
[in] | dpy | the VA display |
[in] | buf_id | the buffer for which synchronization is performed |
[in] | timeout_ns | the timeout in nanoseconds |
VAStatus vaSyncSurface | ( | VADisplay | dpy, |
VASurfaceID | render_target | ||
) |
This function blocks until all pending operations on the render target have been completed. Upon return it is safe to use the render target for a different picture.
VAStatus vaSyncSurface2 | ( | VADisplay | dpy, |
VASurfaceID | surface, | ||
uint64_t | timeout_ns | ||
) |
Synchronizes pending operations associated with the supplied surface.
This function blocks during specified timeout (in nanoseconds) until all pending operations on the render target have been completed. If timeout is zero, the function returns immediately.
Possible errors:
[in] | dpy | the VA display |
[in] | surface | the surface for which synchronization is performed |
[in] | timeout_ns | the timeout in nanoseconds |
VAStatus vaTerminate | ( | VADisplay | dpy | ) |
After this call, all library internal resources will be cleaned up
VAStatus vaUnmapBuffer | ( | VADisplay | dpy, |
VABufferID | buf_id | ||
) |
After client making changes to a mapped data store, it needs to "Unmap" it to let the server know that the data is ready to be consumed by the server