On Thu, 18 May 2023 09:57:51 +0200, Oswald Buddenhagen wrote: > > On Wed, May 17, 2023 at 10:25:00PM +0200, Takashi Iwai wrote: > > On Wed, 17 May 2023 19:42:53 +0200, > > Oswald Buddenhagen wrote: > >> > >> The period_bytes_min parameter made no sense at all, as it didn't > >> correlate with any hardware limitation. > > > > Does the device really work with less than 64 bytes period size? > > I meant not in theory but in practice. > > > somewhat predictably, not. > > > Without any value set, > > dumb applications may try to set 4 bytes for period size, > > > the "try to" is key here. it will fail, because the frame-based > constraint will prevent it from doing so. Ah OK, this should be commented that the lower bound is set in a different way. Takashi
