Hi Drew,

Am 2021-07-02 23:06, schrieb Drew Fustini:

On Fri, Jul 02, 2021 at 07:03:19PM +0300, Andy Shevchenko wrote:

quoted

On Thu, Jul 1, 2021 at 3:23 AM Drew Fustini [off-list ref] 
wrote:

quoted

Add GPIO driver for the StarFive JH7100 SoC [1] used on the
BeagleV Starlight JH7100 board [2].

[1] https://github.com/starfive-tech/beaglev_doc/
[2] https://github.com/beagleboard/beaglev-starlight

quoted

Signed-off-by: Emil Renner Berthing <kernel@esmil.dk>
Signed-off-by: Huan Feng <redacted>
Signed-off-by: Drew Fustini <redacted>

Seems some Co-developed-by are missing.

Thank you for suggesting this.  Huan Feng originally wrote the driver.
Emil and I have made some changes to reorganize and clean it up for
submission.

Do you think all three of us should list Co-developed-by: for our names
in addition to the SOB?

quoted

Brief look into the code brings the Q. Can't you utilize gpio-regmap
here? Why not?

Michael Walle asked about this yesterday and it was my first time
looking at regmap and gpio-regmap.  I've been reading the code and it
does look like I should try convert this driver over to using
gpio-regmap.

The open question in my mind is how to handle the interrupt type (edge
trigged on positive or negative, level triggered on high or low).
Hopefully I can find some other examples that can help me think about
how to do that correctly.

Have a look at include/linux/regmap.h, there is "struct 
regmap_irq_type".
If you're lucky, you can just supply the corresponding values that fits
your hardware. If it doesn't match your hardware at all, then you can
keep your own functions, or if its slightly different, then maybe you
can add support for your quirk in regmap-irq. You don't necessarily have
to use regmap-irq together with gpio-regmap. You can also just use
regmap-irq or gpio-regmap independently.

A quick grep for "type_rising_" lists drivers/mfd/max77650.c and
drivers/mfd/rohm-bd70528.c for example.

-michael