Re: omap2-mcspi multi mode

[Colin Foster <colin.foster@xxxxxxxxxxxxxxxx>]

Hi Thorsten,

On Fri, Jun 21, 2024 at 10:21:26AM +0200, Linux regression tracking (Thorsten Leemhuis) wrote:
> On 11.06.24 16:21, Colin Foster wrote:
> > On Fri, Jun 07, 2024 at 04:45:43PM +0100, Mark Brown wrote:
> >> On Thu, Jun 06, 2024 at 10:14:27PM -0500, Colin Foster wrote:
> >>
> >>> So I think the question I have is:
> >>
> >>> Should the CS line be de-asserted at the end of "spi_write"?
> >>
> >> Absent bodging with cs_change after any spi message the chip select
> >> should be left deasserted.
> > 
> > Do you have hardware to reproduce my results of two spi messages no
> > longer toggling the CS line and leaving the line at GND through the
> > transactions?
> 
> Hmmm, I might have missed something, but it looks like nothing happened
> since that exchange. Did this regression fall through the cracks or can
> I consider the issue resolved for some reason?
> 
> Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)

No, you haven't missed anything. This still feels like a regression to
me, but historically "regressions" I've found end up being a
misconfiguration on my part. I'm travelling this week so it won't be
until next week / weekend that I can get to anything.

I'll plan to look into a fix if it is indeed an issue.

Colin Foster