On 11/29/2017 09:02 AM, Tony Lindgren wrote:

* Linus Walleij [off-list ref] [171129 13:03]:

quoted

On Fri, Nov 3, 2017 at 12:15 AM, Florian Fainelli [off-list ref] wrote:

quoted

Some platforms (e.g: Broadcom STB: BMIPS_GENERIC/ARCH_BRCMSTB) will lose
their register contents when entering their lower power state. In such a
case, the pinctrl-single driver that is used will not be able to restore
the power states without telling the core about it and having
pinctrl_select_state() check for that.

This patch adds a new optional boolean property that Device Tree can
define in order to obtain exactly that and having the core pinctrl code
take that into account.

Signed-off-by: Florian Fainelli <redacted>

Florian, I'm really sorry for losing track of this patch set, it's
important stuff and I see why systems are dependent on something
like this.

Tony: can you look at this from a pinctrl-single point of view?
This is the intended consumer: pinctrl-single users that lose the
hardware state over suspend/resume.

How do you see this working with other pinctrl-single users?

Hmm well typically a device driver that loses it's context just does
save and restore of the registers in runtime PM suspend/resume
as needed. In this case it would mean duplicating the state for
potentially for hundreds of registers.. So using the existing
state in the pinctrl subsystem totally makes sense for the pins.

Florian do you have other reasons why this should be done in the
pinctrl framework instead of the driver? Might be worth describing
the reasoning in the patch descriptions :)

The pinctrl provider driver that I am using is pinctrl-single, which has
proper suspend/resume callbacks but those are not causing any HW
programming to happen because of the (p->state == state) check, hence
this patch series.

So as long as the pinctrl framework state is used to restore the
state by the pinctrl driver instead of the pinctrl consumer drivers,
I don't have issues with this patchset. So probably just improving
the patch messages a bit should do it.

FYI, on omaps, the PRCM hardware saves and restores the pinctrl
state so this has not been so far an issue.

Regards,

Tony

-- 
Florian
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html