On 9/18/19 5:48 AM, Mark Brown wrote:
On Wed, Sep 18, 2019 at 11:25:39AM +0100, Jon Hunter wrote:
Why 2? From looking at various codecs that support dsp-a/b modes, it is more common for the f-sync to be 1 regardless of the number of slots.
In DSP modes only one edge really matters anyway so it's not super important how long the pulse is.
There are exceptions to the rule. In the early days of SOF, we had to provide support for amplifiers that did require a pulse larger than a bit. In the SOF IPC we added an 'frame_pulse_width' field to pass the configuration all the way from topology to the firmware and Intel SSP driver. The other quirk we added is the ability to control zero-padding per slot instead of at the end of the frame, e.g. 1 bit of padding after 24 bits when using 4 slots w/ 25 bits in a 100-bit frame.