On Thu, 15 Jul 2021 09:56:59 +0200, Ujfalusi, Peter wrote: > > > > On 7/15/2021 10:51, Péter Ujfalusi wrote: > > > > > > On 15/07/2021 06:39, Vinod Koul wrote: > >> On 14-07-21, 18:24, Takashi Iwai wrote: > >>> Hi, > >>> > >>> this series contains a couple of patches for compress-offload API, one > >>> for a cleanup of unused API functions and another for the proper mutex > >>> initialization. I stumbled on the second while doing the first. > >> > >> Thanks for the cleanup. We never have a compress only device and it is a > >> mix of PCM and compress FEs mostly :) > > > > Actually I'm working on a series where a card would only contain nothing > > else but a compr, no PCMs... > > and I should have replied from here ;) > > I'm preparing to split up the SOF into IPC clients [1] and the probes > support (which is using compress API) is going to be moved as a separate > client, creating it's own sound card with the single comprCxD0 (and the > controlCx). > > Is this cleanup going to prevent this? Unless you use snd_compress_register() function, nothing can break. And if you were to use the function together with ASoC, it must be very wrong :) Takashi