Trait cros_codecs::decoders::vp9::backends::StatelessDecoderBackend

pub(crate) trait StatelessDecoderBackend: VideoDecoderBackend {
    fn new_sequence(&mut self, header: &Header) -> Result<()>;
    fn submit_picture(
        &mut self,
        picture: &Header,
        reference_frames: &[Option<Self::Handle>; 8],
        bitstream: &[u8],
        timestamp: u64,
        segmentation: &[Segmentation; 8],
        block: BlockingMode
    ) -> Result<Self::Handle>;
}

Trait for stateless decoder backends. The decoder will call into the backend to request decode operations. The backend can operate in blocking mode, where it will wait until the current decode finishes, or in non-blocking mode, where it should return immediately with any previously decoded frames that happen to be ready.

Required Methods

fn new_sequence(&mut self, header: &Header) -> Result<()>

Called when new stream parameters are found.

fn submit_picture(
    &mut self,
    picture: &Header,
    reference_frames: &[Option<Self::Handle>; 8],
    bitstream: &[u8],
    timestamp: u64,
    segmentation: &[Segmentation; 8],
    block: BlockingMode
) -> Result<Self::Handle>

Called when the decoder wants the backend to finish the decoding operations for picture. The argument block dictates whether this call should wait until the current decode finishes, or whether it should return immediately.

This call will assign the ownership of the BackendHandle to the Picture and then assign the ownership of the Picture to the Handle.

Implementors

impl StatelessDecoderBackend for Backend