ALSA Topology Framework – Vinod Koul, Intel
ALSA Topology Framework – Vinod Koul, Intel
Many modern audio systems employ DSPs for audio, voice and speech processing. The processing modules within the DSP are typically implemented in firmware and are configurable by software. Typically a graph of the processing chain in the DSP is analogous to an ASoC DAPM graph and can be modelled using the ASoC DAPM model. But the graph needs to be created by the driver and each product & model may have different processing blocks in the graph. This puts the onus on ALSA driver and subsystem to provide a mechanism which allows DSP graphs to be scaled.
The new upstreamed ASoC topology framework allows us to represent the DSP processing chain as a graph/topology and load it from userland during boot resulting in configurable and customizable Linux audio systems. The talk explores this topology framework and its usage in the audio driver on Intel’s Skylake chipsets.
About Vinod Koul
Vinod works in Linux Audio group for Intel. He is involved in Audio driver development and upstreaming for Intel platforms. He also wrote and maintains the ALSA compressed audio framework. Vinod is the maintainer of Linux dmaengine subsystem.
by The Linux Foundation
linux foundation
