Re: [PATCH] net: phy: marvell-88q2xxx: Add support for PHY LEDs on 88q2xxx

[Stefan Eichenberger <eichest@gmail.com>]

Hi Dimitri ,

On Fri, Jan 10, 2025 at 04:10:04PM +0100, Dimitri Fedrau wrote:
> Marvell 88Q2XXX devices support up to two configurable Light Emitting
> Diode (LED). Add minimal LED controller driver supporting the most common
> uses with the 'netdev' trigger.
> 
> Signed-off-by: Dimitri Fedrau <dima.fedrau@gmail.com>
> ---
>  drivers/net/phy/marvell-88q2xxx.c | 161 ++++++++++++++++++++++++++++++++++++++
>  1 file changed, 161 insertions(+)
> 
> diff --git a/drivers/net/phy/marvell-88q2xxx.c b/drivers/net/phy/marvell-88q2xxx.c
> index 5107f58338aff4ed6cfea4d91e37282d9bb60ba5..bef3357b9d279fca5d1f86ff0eaa0d45a699e3f9 100644
> --- a/drivers/net/phy/marvell-88q2xxx.c
> +++ b/drivers/net/phy/marvell-88q2xxx.c
> @@ -8,6 +8,7 @@
>   */
>  #include <linux/ethtool_netlink.h>
>  #include <linux/marvell_phy.h>
> +#include <linux/of.h>
>  #include <linux/phy.h>
>  #include <linux/hwmon.h>
>  
> @@ -27,6 +28,9 @@
>  #define MDIO_MMD_AN_MV_STAT2_100BT1		0x2000
>  #define MDIO_MMD_AN_MV_STAT2_1000BT1		0x4000
>  
> +#define MDIO_MMD_PCS_MV_RESET_CTRL		32768
> +#define MDIO_MMD_PCS_MV_RESET_CTRL_TX_DISABLE	0x8
> +
>  #define MDIO_MMD_PCS_MV_INT_EN			32784
>  #define MDIO_MMD_PCS_MV_INT_EN_LINK_UP		0x0040
>  #define MDIO_MMD_PCS_MV_INT_EN_LINK_DOWN	0x0080
> @@ -40,6 +44,15 @@
>  #define MDIO_MMD_PCS_MV_GPIO_INT_CTRL			32787
>  #define MDIO_MMD_PCS_MV_GPIO_INT_CTRL_TRI_DIS		0x0800
>  
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL			32790
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL_GPIO_MASK		GENMASK(7, 4)
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL_TX_EN_MASK	GENMASK(3, 0)
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL_LINK		0x0 /* Link established */
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL_LINK_RX_TX	0x1 /* Link established, blink for rx or tx activity */
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL_RX_TX		0x4 /* Receive or Transmit activity */
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL_TX		0x5 /* Transmit activity */
> +#define MDIO_MMD_PCS_MV_LED_FUNC_CTRL_LINK_1000BT1	0x7 /* 1000BT1 link established */
> +
>  #define MDIO_MMD_PCS_MV_TEMP_SENSOR1			32833
>  #define MDIO_MMD_PCS_MV_TEMP_SENSOR1_RAW_INT		0x0001
>  #define MDIO_MMD_PCS_MV_TEMP_SENSOR1_INT		0x0040
> @@ -95,6 +108,9 @@
>  
>  #define MDIO_MMD_PCS_MV_TDR_OFF_CUTOFF			65246
>  
> +#define MV88Q2XXX_LED_INDEX_TX_ENABLE	0
> +#define MV88Q2XXX_LED_INDEX_GPIO	1

Not sure if I understand this. TX_ENABLE would be LED0 and GPIO would be
LED1? In my datasheet the 88Q222x only has a GPIO pin (which is also
TX_ENABLE), is this a problem? Would we need a led_count variable per
chip? 

In the 88Q2110 I can see that there is a TX_ENABLE (0) and a GPIO (1)
pin. In the register description they just call it LED [0] Control and
LED [1] Control. Maybe calling it LED_0 and LED_1 would be easier to
understand? Same for MDIO_MMD_PCS_MV_LED_FUNC_CTRL_GPIO_MASK and
MDIO_MMD_PCS_MV_LED_FUNC_CTRL_TX_EN_MASK.

> +
>  struct mmd_val {
>  	int devad;
>  	u32 regnum;
> @@ -741,8 +757,58 @@ static int mv88q2xxx_hwmon_probe(struct phy_device *phydev)
>  }
>  #endif
>  
> +#if IS_ENABLED(CONFIG_OF_MDIO)
> +static int mv88q2xxx_leds_probe(struct phy_device *phydev)
> +{
> +	struct device_node *node = phydev->mdio.dev.of_node;
> +	struct device_node *leds;
> +	int ret = 0;
> +	u32 index;
> +
> +	if (!node)
> +		return 0;
> +
> +	leds = of_get_child_by_name(node, "leds");
> +	if (!leds)
> +		return 0;
> +
> +	for_each_available_child_of_node_scoped(leds, led) {
> +		ret = of_property_read_u32(led, "reg", &index);
> +		if (ret)
> +			goto exit;
> +
> +		if (index > MV88Q2XXX_LED_INDEX_GPIO) {
> +			ret = -EINVAL;
> +			goto exit;
> +		}
> +
> +		if (index == MV88Q2XXX_LED_INDEX_TX_ENABLE)
> +			ret = phy_clear_bits_mmd(phydev, MDIO_MMD_PCS,
> +						 MDIO_MMD_PCS_MV_RESET_CTRL,
> +						 MDIO_MMD_PCS_MV_RESET_CTRL_TX_DISABLE);

If I understand it correctly, this switches the function of the pin from
TX_DISABLE to GPIO? Can you maybe add a comment here?

Regards,
Stefan