On Thursday, 2 August 2018 13:35:46 MSK Dmitry Osipenko wrote:

On Thursday, 2 August 2018 02:11:23 MSK Dmitry Osipenko wrote:

quoted

On Thursday, 2 August 2018 01:51:42 MSK Stefan Agner wrote:

quoted

On 01.08.2018 22:51, Dmitry Osipenko wrote:

quoted

On Thursday, 26 July 2018 18:40:25 MSK Stefan Agner wrote:

quoted

The properties have been commented out to prevent a regression a
while ago. The first regression should be resolved by
commit 44af7927316e ("spi: Map SPI OF client IRQ at probe time").

The second regression is probably addressed by
commit 494fd7b7ad10 ("PM / core: fix deferred probe breaking suspend
resume
order") and/or maybe others. Readd the gpio-ranges properties to see
whether regressions still get reported.

This reverts commit 4f1d841475e1f6e9e32496dda11215db56f4ea73
("ARM: tegra: Comment out gpio-ranges properties").

Signed-off-by: Stefan Agner <stefan@agner.ch>
---

Unfortunately the second regression hasn't been addressed yet, this
patch
still breaks wake-up using GPIO key. BTW, better to spell "readd" as
"re-add".

You do have the hardware and could test it?

Yes, I have.

quoted

Hm, that is unfortunate. I think this was the patch which should address
the issue:
https://lkml.org/lkml/2015/9/10/218

I've tested the patch and it is still actual, i.e. it fixes the wake-up.
Welcome back to 2015 :-)

Actually, the pinctrl-gpio probe order is easily fixable by reshuffling the
drivers registration order within the kernel. I'll send patches.

Linus applied the GPIO / PINCTRL drivers probe-reorder patches for-next, so 
this DT patch should be good for 4.20.

Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Tested-by: Dmitry Osipenko <digetx@gmail.com>