Re: [PATCH bluetooth-next 1/1] drivers: mrf24j40: add support for reset and wake pin

[Stefan Schmidt <stefan@osg.samsung.com>]

Hello.

On 15/11/16 14:18, smlng wrote:

Please fix your mail from line to have your real name to show up as 
patch author.

> The mrf24j40 requires a high on the RESET pin, otherwise it will reset
> itself over and over. With this patch the driver will ensure that a GPIO
> (if connected) will be set to high during device initialization. Details:
>
>  - set reset to high on driver probe
>  - make pin configurable via device tree (overlay)
>  - similar to the at86rf233 driver
>  - update devicetree documentation

Thanks for taking the time to submit this patch! This looks good to me 
in general. I cc'ed Alan Ott as the driver maintainer now to see what he 
thinks about this patch.

What I miss here is a bit of context for the wake pin. The commit 
message only covers the reset pin handling. What do we do with the wake 
pin here?

> Signed-off-by: smlng <s@mlng.net>

Same as the from line. Please use your real name for the SOB here.

> ---
>  .../bindings/net/ieee802154/mrf24j40.txt           |  4 ++
>  drivers/net/ieee802154/mrf24j40.c                  | 59 ++++++++++++++++++++--
>  include/linux/spi/mrf24j40.h                       |  9 ++++
>  3 files changed, 69 insertions(+), 3 deletions(-)
>  create mode 100644 include/linux/spi/mrf24j40.h
>
> diff --git a/Documentation/devicetree/bindings/net/ieee802154/mrf24j40.txt b/Documentation/devicetree/bindings/net/ieee802154/mrf24j40.txt
> index a4ed2efb5b73..2fabba041e48 100644
> --- a/Documentation/devicetree/bindings/net/ieee802154/mrf24j40.txt
> +++ b/Documentation/devicetree/bindings/net/ieee802154/mrf24j40.txt
> @@ -9,6 +9,10 @@ Required properties:
>    - reg:		the chipselect index
>    - interrupts:		the interrupt generated by the device.
>
> +Optional properties:
> +  - reset-gpio:		GPIO spec for the rstn pin
> +  - wake-gpio:		GPIO spec for the wake pin
> +
>  Example:
>
>  	mrf24j40ma@0 {
> diff --git a/drivers/net/ieee802154/mrf24j40.c b/drivers/net/ieee802154/mrf24j40.c
> index 7b131f8e4093..275b54d6bad3 100644
> --- a/drivers/net/ieee802154/mrf24j40.c
> +++ b/drivers/net/ieee802154/mrf24j40.c
> @@ -16,9 +16,11 @@
>   */
>
>  #include <linux/spi/spi.h>
> +#include <linux/spi/mrf24j40.h>
>  #include <linux/interrupt.h>
>  #include <linux/module.h>
>  #include <linux/regmap.h>
> +#include <linux/of_gpio.h>
>  #include <linux/ieee802154.h>
>  #include <linux/irq.h>
>  #include <net/cfg802154.h>
> @@ -1276,16 +1278,67 @@ static void  mrf24j40_phy_setup(struct mrf24j40 *devrec)
>  	}
>  }
>
> +static int
> +mrf24j40_get_pdata(struct spi_device *spi, int *rstn, int *wake)
> +{
> +	struct mrf24j40_platform_data *pdata = spi->dev.platform_data;
> +
> +	if (!IS_ENABLED(CONFIG_OF) || !spi->dev.of_node) {
> +		if (!pdata)
> +			return -ENOENT;
> +
> +		*rstn = pdata->rstn;
> +		*wake = pdata->wake;
> +		return 0;
> +	}
> +
> +	*rstn = of_get_named_gpio(spi->dev.of_node, "reset-gpio", 0);
> +	*wake = of_get_named_gpio(spi->dev.of_node, "wake-gpio", 0);
> +	return 0;
> +}
> +
>  static int mrf24j40_probe(struct spi_device *spi)
>  {
>  	int ret = -ENOMEM, irq_type;
> +	int rc, rstn, wake;
>  	struct ieee802154_hw *hw;
>  	struct mrf24j40 *devrec;
>
> -	dev_info(&spi->dev, "probe(). IRQ: %d\n", spi->irq);

You removed this line entirely. Was that on purpose? Having the IRQ 
printed in dmesg might be handy for debugging.

> +	if (!spi->irq) {
> +		dev_err(&spi->dev, "no IRQ specified\n");
> +		return -EINVAL;
> +	}
>
> -	/* Register with the 802154 subsystem */
> +	rc = mrf24j40_get_pdata(spi, &rstn, &wake);
> +	if (rc < 0) {
> +		dev_err(&spi->dev, "failed to parse platform_data: %d\n", rc);
> +		return rc;
> +	}
> +
> +	if (gpio_is_valid(rstn)) {
> +		rc = devm_gpio_request_one(&spi->dev, rstn,
> +					   GPIOF_OUT_INIT_HIGH, "rstn");
> +		if (rc)
> +			return rc;
> +	}
>
> +	if (gpio_is_valid(wake)) {
> +		rc = devm_gpio_request_one(&spi->dev, wake,
> +					   GPIOF_OUT_INIT_LOW, "wake");
> +		if (rc)
> +			return rc;
> +	}
> +
> +	/* Reset */
> +	if (gpio_is_valid(rstn)) {
> +		udelay(1);
> +		gpio_set_value(rstn, 0);
> +		udelay(1);
> +		gpio_set_value(rstn, 1);
> +		usleep_range(120, 240);
> +	}
> +
> +	/* Register with the 802154 subsystem */
>  	hw = ieee802154_alloc_hw(sizeof(*devrec), &mrf24j40_ops);
>  	if (!hw)
>  		goto err_ret;
> @@ -1350,7 +1403,7 @@ static int mrf24j40_probe(struct spi_device *spi)
>  		goto err_register_device;
>  	}
>
> -	dev_dbg(printdev(devrec), "registered mrf24j40\n");
> +	dev_info(printdev(devrec), "registered mrf24j40\n");

Not related to the rest of this patch. The change itself is good though. 
Maybe combine this with IRQ printout from above and have this as an 
additional patch?

>  	ret = ieee802154_register_hw(devrec->hw);
>  	if (ret)
>  		goto err_register_device;
> diff --git a/include/linux/spi/mrf24j40.h b/include/linux/spi/mrf24j40.h
> new file mode 100644
> index 000000000000..60810194cae9
> --- /dev/null
> +++ b/include/linux/spi/mrf24j40.h
> @@ -0,0 +1,9 @@
> +#ifndef _SPI_MRF24J40_H
> +#define _SPI_MRF24J40_H
> +
> +struct mrf24j40_platform_data {
> +	int rstn;
> +	int wake;
> +};
> +
> +#endif /* _SPI_MRF24J40_H */
>

Just to clarify this. This header file is needed for the usage of 
platform data, right? If not I would argue that we do not create a 
header for this in include/spi but have the struct in the driver itself. 
I vaguely remember that we need this for the usage of pdata.

regards
Stefan Schmidt