Re: spi: spi-synquacer: fix set_cs handling

[Jassi Brar <jaswinder.singh@xxxxxxxxxx>]

On Mon, 25 Jan 2021 at 06:37, Mark Brown <broonie@xxxxxxxxxx> wrote:
>
> On Sun, Jan 24, 2021 at 04:17:55PM -0600, jassisinghbrar@xxxxxxxxx wrote:
>
> > Respect the set_cs() request by actually flushing the FIFOs
> > and start/stop the SPI instance.
>
> set_cs() is a request to set the chip select not flush the FIFOs or
> restart the hardware - what's the actual issue here?  Transfers should
> happen in the transfer callback, the driver shouldn't be assuming there
> is anything going on with chip select when completing transfers.
>
The controller has one block for each slave-select, and we need to
actually stop the block to deassert the CS. At the minimum we need to
set the DMSTOP_STOP bit.
I will revise the patch to be much easier on the eyes.

thanks.