![](https://secure.gravatar.com/avatar/86272ea437f62fe3f10a16762513af3d.jpg?s=120&d=mm&r=g)
On Wed, 2010-10-06 at 15:30 +0200, Takashi Iwai wrote:
At Wed, 06 Oct 2010 10:25:40 +0100, Dimitris Papastamos wrote:
On Wed, 2010-10-06 at 08:31 +0100, Mark Brown wrote: On Wed, Oct 06, 2010 at 09:10:23AM +0200, Takashi Iwai wrote:
If the above difference is intentional, it should be commented somewhere.
Meh, yes. Dimitris, please fix or add a comment as appropriate.
In snd_soc_7_9_spi_write we prepare the tx buffer to be register followed by data packed into 16 bits. In snd_soc_4_12_spi_write the tx buffer is swapped. I'd expect this to be consistent between the two transfers.
Sorry, I don't understand your statement clearly. So, the byte-swap behavior in snd_soc_4_12_spi_write() is designed?
I meant to say that snd_soc_4_12_spi_write looks suspicious and that I'd expect it to behave similarly to snd_soc_7_9_spi_write. I don't see why the byte swapping is needed.
Thanks, Dimitrios