Messages in this thread Patch in this message | | | From | Linus Walleij <> | Date | Thu, 29 Dec 2022 01:23:18 +0100 | Subject | Re: [PATCH v3 1/2] mmc: atmel-mci: Convert to gpio descriptors |
| |
Hi Balamanikandan,
thanks for your patch!
On Mon, Dec 26, 2022 at 8:39 AM Balamanikandan Gunasundar <balamanikandan.gunasundar@microchip.com> wrote:
> Replace the legacy GPIO APIs with gpio descriptor consumer interface. > > To maintain backward compatibility, we rely on the "cd-inverted" > property to manage the invertion flag instead of GPIO property. > > Signed-off-by: Balamanikandan Gunasundar <balamanikandan.gunasundar@microchip.com>
Overall this looks really nice! I thought about converting this driver but was a bit afraid of doing mistakes since it was a bit elaborate. Nice that you use fwnode accessors!
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Improvement to consider (can also be a separate patch):
> + if (slot->detect_pin) { > + present = !(gpiod_get_raw_value(slot->detect_pin) ^ > slot->detect_is_active_high);
Normally we contain all this active low/high mess in mmc_of_parse(), but I guess this doesn't work here because the of node is different from the device :P
Since the code says *explicitly* "active high" not "inverted" as the core code parses it, could you try just push this to gpiolib like we usually do, by deleting the detect_is_active_high stuff and just apply a patch like this?
diff --git a/drivers/gpio/gpiolib-of.c b/drivers/gpio/gpiolib-of.c index 4fff7258ee41..5979c9a75cf9 100644 --- a/drivers/gpio/gpiolib-of.c +++ b/drivers/gpio/gpiolib-of.c @@ -199,7 +199,9 @@ static void of_gpio_set_polarity_by_property(const struct device_node *np, { "fsl,imx8mm-pcie", "reset-gpio", "reset-gpio-active-high" }, { "fsl,imx8mp-pcie", "reset-gpio", "reset-gpio-active-high" }, #endif - +#if IS_ENABLED(CONFIG_MMC_ATMELMCI) + { "atmel,hsmci", "cd-gpios", "cd-inverted" }, +#endif /* * The regulator GPIO handles are specified such that the * presence or absence of "enable-active-high" solely controls (It's fine to include in your MMC patch if some GPIO maintainer like me ACK it.)
Yours, Linus Walleij
| |