On Wed, Oct 16, 2013 at 03:41:51PM -0600, Stephen Warren wrote:
> On 10/16/2013 03:00 PM, Mark Brown wrote:
> > -static int tegra_slink_transfer_one_message(struct spi_master *master,
> > - struct spi_message *msg)
> ...
> > - msg->actual_length += xfer->len;
> > - if (xfer->cs_change && xfer->delay_usecs) {
> > - tegra_slink_writel(tspi, tspi->def_command_reg,
> > - SLINK_COMMAND);
> > - udelay(xfer->delay_usecs);
> > - }
> That chunk didn't seem to end up anywhere else. Is that expected?
It should be handled by framework code and it's buggy anyway, the two
fields are not connected and cs_change looks awfully like it'd at best
only work the first time it's used.
Attachment:
signature.asc
Description: Digital signature
