On Mon, Sep 26, 2022 at 03:42:30PM +0200, Vincent Whitchurch wrote:
> On Mon, Sep 19, 2022 at 03:38:29PM +0100, Mark Brown wrote:

> > The driver should just set max_transfer_size and let the core
> > handle this.

> Hmm, AFAICS, the core doesn't actually do anything with
> max_transfer_size?  It's only used from spi-mem and a handful of other
> clients via spi_max_transfer_size().

Then we should fix the core to handle it properly.