Re: [PATCH] spi: Fix incorrect cs_setup delay handling

[Hector Martin]

On 11/12/2021 01.21, Mark Brown wrote:
> On Sat, Dec 11, 2021 at 12:36:34AM +0900, Hector Martin wrote:
>
> > We need to wait *after* asserting CS and before the transfer, not before
> > asserting CS which isn't very useful.
>
> This needs a better changelog, there's multiple delays being handled
> here and it's not clear from this which are affected here or what the
> problem is.

cs_setup is affected, I thought at least that was clear from the patch 
summary :-)

From spi.h:

* @cs_setup: delay to be introduced by the controller after CS is asserted

So clearly that delay has to happen at the end of spi_set_cs, *after* 
the assertion, not at the beginning before it.

If you prefer, I can resend it with a reference to the spi.h comment.

--
Hector Martin (marcan@xxxxxxxxx)
Public Key: https://mrcn.st/pub