Hi Geert > > > Don't you have to initialize SITMDR[123] unconditionally, as reception > > > requires transmitting dummy data on R-Car (cfr. SPI_CONTROLLER_MUST_TX)? > > > > Good catch, but I added 1 restriction for MSIOF-I2S mode. > > I have explained it on top of this driver. The restriction is > > "MSIOF-I2S doesn't work as Clock/Frame Provider Mode". > > So, dummy transmit for RX is not needed/assumed. > > I think it is one of big-diff between MSIOF-SPI ? > > IC. Being just a mortal sound-noob, I didn't know what "Clock/Frame > Provider Mode" means ;-) Oh, now I understand. I had missed > completely that you are running MSIOF in slave mode. So everything > should be fine. ASoC is now using Provider/Consumer instead of Master/Slave. Veteran engineers are not familiar with it :) > /* SITSCR */ > #define SITSCR_V(p, d) ((p << 8) + d) > > is unused, and can be removed. OK, Thanks. Will remove in v2 Thank you for your help !! Best regards --- Kuninori Morimoto
