[RFC PATCH 4/4] i2c: at91: Move to generic GPIO bus recovery

[Wolfram Sang]

Hi Codrin,

> The pinmux driver needs to have strict set to false, otherwise the 
> switching is not available, not at this time at least. Perhaps there is 
> room for improvement here, because the I2C bus is not using the pins 
> while we are doing GPIO recovery.

Our driver doesn't use 'strict'. The thing is that I can't describe a
pinctrl state for GPIO. GPIO is the default state until another function
is requested. Back to GPIO currently means freeing the pin again, so it
defaults back to GPIO. We are currently discussing it. Geert (CCed)
isn't very happy of describing the same pins with 'function = "gpio"'
because the Kernel already knows the mapping, just needs to revert it.
Geert, please correct me if I am wrong.

> I am not sure I'll have time the next week to work on what you asked me 
> regarding sh_mobile and PXA, but I will look into it the week after that.
> Sorry about my delayed reply, I was on vacation.

Well, no need for sh_mobile, this is my todo item :) About PXA, well, I
am still happy that you volunteered to do it, so I hope you had a
relaxing vacation!

Happy hacking,

   Wolfram

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20200904/44bb649b/attachment.sig>