| CDAC_GPIO_DEVICE_FLAGS | |
| Common Declarations | Contains standard definitions used by various interfaces |
| ▼Deep Learning Accelerator | The NvMedia Deep Learning Accelerator (DLA) API encompasses all NvMedia functions that access the DLA hardware engine for deep learning operations |
| ▼Deep Learning Accelerator | NvMedia DLA runtime APIs for accessing the DLA hardware engine for deep learning operations |
| DLA Specific Types | Defines specific types for DLA |
| Deep Learning Accelerator Synchronization | The NvMedia DLA NvSciSync API encompasses all NvMediaDla NvSciSync handling functions |
| Error Handling | Nverror.h contains our error code enumeration and helper macros |
| ▼GPU access API | GPU management and access API |
| ▼GPU access API: Device management | Device control, device capabilities, and device memory management |
| GPU access API: Device management (safety subset) | |
| ▼GPU access API: Library | Library management and device discovery |
| GPU access API: Library (safety subset) | |
| ▼GPU access API (safety subset) | |
| GPU access API: Device management (safety subset) | |
| GPU access API: Library (safety subset) | |
| ▼Graphics APIs | Graphics APIs in this section include NVIDIA and 3rd-party APIs |
| EGL Specifications and Extensions | |
| OpenGL ES 2.x/3.x Specs and Extensions | |
| OpenGL Specifications | |
| Vulkan SC Specs and Extensions | |
| Vulkan Specs and Extensions | |
| IOCTL related info | MACRO are defined for driver supported private IOCTLs. These IOCTLs can be called using SIOCDEVPRIVATE custom ioctl command |
| MACROS | EQOS generic helper MACROS |
| MTL queue AVB algorithm mode | MTL AVB queue algorithm type |
| ▼NvDisplay | For additional information, see the "NvDisplay" chapter of the Developer Guide |
| OpenWFD Extensions | Contains extensions to allow users to create WFDSource objects from NvSciBuf objects |
| ▼NvMedia API for DRIVE OS 6.0 | NVIDIA™ NvMedia technology provides powerful processing of multimedia data that enables true hardware acceleration across NVIDIA SoCs. NvMedia is a C-based, frame level API library that provides framework agnostic, distinctive software components, to realize various multimedia use case scenarios |
| 2D Processing | The 2D Processing API encompasses all NvMedia 2D image processing related functionality |
| 2D Surface and Synchronization | The NvMedia 2D NvSci API encompasses all NvMedia 2D handling for NvSciBuf NvSciSync related functions |
| ▼Common Declarations | Defines basic types used for video and images throughout the NvMedia API |
| ►Basic NvMedia Types and Structures | Defines basic types used throughout the NvMedia API |
| Device | Manages NvMediaDevice objects, which are the root of the Nvmedia object system |
| Version Information | Provides version information for the NvMedia library |
| Encode/Decode: Common Types | Defines common types and declarations for decode and encode operations |
| ►Parser API | Provides parsing for decoder frame handling |
| Abstraction of OS-Specific Client Callbacks | |
| ▼Image Handling API | The Image Processing API encompasses all NvMedia image-related functionality |
| ►Image Decoder | |
| Data types for AV1 Decoding | |
| Data types for H.264 Decoding | |
| Data types for H.265 Decoding | |
| Data types for MPEG-1 and MPEG-2 Decoding | |
| Data types for MPEG4 Part 2 Decoding | |
| Data types for VC1 Decoding | |
| Data types for VP8 Decoding | |
| Data types for VP9 Decoding | |
| Decoder Creation Flag | Defines decoder flag bit masks for constructing the decoder |
| Image Decoder: Common Types | |
| ►Image Encoder | The NvMediaIEP object takes uncompressed image data and turns it into a codec specific bitstream. Only H.264, H.265 and VP9 encoding is supported |
| ►Data types for H.264 encoding | Types and declarations for H.264 Encoding |
| Data types for AV1 encoding | Types and declarations for AV1 Encoding |
| Data types for H.265/HEVC encoding | Types and declarations for H.265/HEVC Encoding |
| Data types for VP9 encoding | Types and declarations for VP9 Encoding |
| Image Encoder Output Extradata | Defines Image Encoder Output Extradata data types which can be used with NvMedia IEP APIs to retrieve encoder statistics/configurations. This feature is not available on the QNX Safety configuration |
| Image JPEG Decoder | The NvMediaIJPD object takes a JPEG bitstream and decompress it to image data |
| Image JPEG Encoder | The NvMediaIJPE object takes uncompressed image data and turns it into a jpeg bitstream |
| Image Optical Flow Accelerator (IOFA) | The NvMediaIofa object takes an uncompressed bufObj frame pair and turns them into optical flow / stereo disparity estimation data |
| LDC Surface and Synchronization | The NvMedia LDC NvSci API encompasses all NvMedia LDC handling for NvSciBuf NvSciSync related functions |
| Lens Distortion Correction | The LDC API encompasses all NvMedia LDC/TNR image processing related functionality |
| Receive Parser related information | Flexible Receive Parser commands, table size and other defines |
| ▼SIPL | SIPL provides abstract and simple API to capture the output of image sensors with optional image processing |
| ▼NvSIPL Camera | Provides top-level interfaces to program external image devices and Tegra to create and manage image processing pipelines to receive outputs in NvSciBufObj Images |
| NvSIPL Capture definitions | NvSipl Cap Defines for image capture interface and input format types |
| NvSIPL Client | Provides interfaces to retrieve the output of the SIPL Pipeline Manager |
| NvSIPL DeviceBlock Information | Describes information about devices supported by SIPL Device Block |
| NvSIPL ISP Stats | NvSipl ISP Defines for ISP Stat structures |
| NvSIPL ISP Structures | NvSipl ISP Defines for ISP Structures |
| NvSIPL Pipeline Manager | Programs Video Input (VI) and Image Signal Processor (ISP) hardware blocks to create image processing pipelines for each sensor |
| NvSIPL Version | Holds the version information for NvSIPL Camera and NvSIPL Client |
| NvSIPL Query | Manages a database of information about external devices (deserializer, serializer, sensor, and EEPROM) and the camera platform configurations supported by SIPL Device Block drivers |
| SIPL Auto Control | Describes interfaces for SIPL Auto Control implementation |
| Screen Capture | The Screen Capture API is implemented by the Screen Capture wrapper library, libnvscrncapt.so |
| ▼Software Communication Interfaces | The NVIDIA® Software Communication Interfaces (NvSci) libraries serve two primary purposes: |
| ▼Inter-Process Communication | IPC and Event Service APIs |
| Event Service APIs | |
| IPC APIs | |
| ▼NvStreams | Contains the buffer allocation, synchronization, and streaming APIs |
| ►Buffer Allocation APIs | The NvSciBuf library contains the APIs for applications to allocate and exchange buffers in memory |
| NvSciBuf APIs | List of APIs to transport NvSciBuf buffers and attribute list objects across various communication boundaries that interact using NvSciIpc |
| NvSciBuf Attribute List APIs | Methods to perform operations on NvSciBuf attribute lists |
| NvSciBuf Data Structures | Specifies NvSciBuf data structures |
| NvSciBuf Datatype Definitions | Contains a list of all NvSciBuf datatypes |
| NvSciBuf Enumerations for Attribute Keys | List of all NvSciBuf enumerations for attribute keys |
| NvSciBuf Global Constants | Definitions of all NvSciBuf Global Constants/Macros |
| NvSciBuf Initialization APIs | List of APIs to initialize/de-initialize NvSciBuf module |
| NvSciBuf Object APIs | List of APIs to create/operate on NvSciBufObj |
| NvSciBuf blanket statements. | Generic statements applicable for NvSciBuf interfaces |
| ►Streaming APIs | The NvSciStream library is a layer on top of NvSciBuf and NvSciSync libraries that provides utilities for streaming sequences of data packets between multiple application modules to support a wide variety of use cases |
| NvSciStream APIs | Methods to setup and stream sequences of data packets |
| NvSciStream Blanket Statements | Generic statements applicable for NvSciStream interfaces |
| NvSciStream Data Types | Contains a list of NvSciStream datatypes |
| Synchronization APIs | |
| SCI Error Handling | Contains error code enumeration and helper macros |
| ▼Tensor Handling API | The Tensor Processing API encompasses all NvMedia tensor related functionality |
| Tensor Handling API | The Tensor Processing API encompasses all NvMedia tensor related functionality |
| Tensor Handling Buffer Allocation API | The NvMedia Tensor NvSciSync API encompasses all NvMedia Tensor Buffer Allocation processing |
| ▼Tensor Metadata API | The NvMedia Tensor Metadata API encompasses all NvMedia Tensor metadata processing |
| ►Stride and Dimension Indices | The application can use the stride and dimension indices defined below to access the strides and sizes from the corresponding arrays. For example, the stride indices are used to access dimstrides[] in the NvMediaTensorMetaData structure |
| Dimension Indices - NCHW | |
| Dimension Indices - NCxHWx | |
| Dimension Indices - NHWC | |
| Stride Indices - NCHW | For NCHW tensor, assuming p,q,r,s correspond to N,C,H,W dimensions, the offset from the base address of the tensor to access a tensor value at coordinate (p,q,r,s) is calculated using the expression below: |
| Stride Indices - NCxHWx | For NCxHWx tensor, assuming p,q,r,s correspond to N,C,H,W dimensions, the offset from the base address of the tensor to access a tensor value at coordinate (p,q,r,s) is calculated using the expression below: |
| Stride Indices - NHWC | For the NHWC tensor, assuming p,q,r,s correspond to N,H,W,C dimensions, the offset from the base address of the tensor to access a tensor value at coordinate (p,q,r,s) is calculated using the expression below: |
| queue operation mode | MTL queue operation mode options |
