linux-kernel - Re: [PATCH net-next] net: phy: mscc: Add support for PHY LEDs on VSC8541

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <ee6a79ae-4857-44e4-b8e9-29cdd80d828f@lunn.ch>
Date: Thu, 6 Nov 2025 21:45:26 +0100
From: Andrew Lunn <andrew@...n.ch>
To: Prabhakar <prabhakar.csengg@...il.com>
Cc: Heiner Kallweit <hkallweit1@...il.com>,
	Russell King <linux@...linux.org.uk>,
	"David S. Miller" <davem@...emloft.net>,
	Eric Dumazet <edumazet@...gle.com>,
	Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
	Horatiu Vultur <horatiu.vultur@...rochip.com>,
	Geert Uytterhoeven <geert+renesas@...der.be>,
	Vladimir Oltean <vladimir.oltean@....com>,
	Vadim Fedorenko <vadim.fedorenko@...ux.dev>,
	Maxime Chevallier <maxime.chevallier@...tlin.com>,
	netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-renesas-soc@...r.kernel.org,
	Biju Das <biju.das.jz@...renesas.com>,
	Fabrizio Castro <fabrizio.castro.jz@...esas.com>,
	Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>
Subject: Re: [PATCH net-next] net: phy: mscc: Add support for PHY LEDs on
 VSC8541

> +static int vsc85xx_led_cntl_set_lock_unlock(struct phy_device *phydev,
> +					    u8 led_num,
> +					    u8 mode, bool lock)
>  {
>  	int rc;
>  	u16 reg_val;
>  
> -	mutex_lock(&phydev->lock);
> +	if (lock)
> +		mutex_lock(&phydev->lock);
>  	reg_val = phy_read(phydev, MSCC_PHY_LED_MODE_SEL);
>  	reg_val &= ~LED_MODE_SEL_MASK(led_num);
>  	reg_val |= LED_MODE_SEL(led_num, (u16)mode);
>  	rc = phy_write(phydev, MSCC_PHY_LED_MODE_SEL, reg_val);
> -	mutex_unlock(&phydev->lock);
> +	if (lock)
> +		mutex_unlock(&phydev->lock);
>  
>  	return rc;
>  }

The normal way to do this is have _vsc85xx_led_cntl_set() manipulate
the hardware, no locking. And have vsc85xx_led_cntl_set() take the
lock, call _vsc85xx_led_cntl_set(), and then release the lock. You can
then call _vsc85xx_led_cntl_set() if needed.

> +static int vsc8541_led_combine_disable_set(struct phy_device *phydev, u8 led_num,
> +					   bool combine_disable)
> +{
> +	u16 reg_val;
> +
> +	reg_val = phy_read(phydev, MSCC_PHY_LED_BEHAVIOR);

phy_read() can return a negative value. You should not assign that to
a u16.

Also, BEHAVIOUR.

> +	reg_val &= ~LED_COMBINE_DIS_MASK(led_num);
> +	reg_val |= LED_COMBINE_DIS(led_num, combine_disable);
> +
> +	return phy_write(phydev, MSCC_PHY_LED_BEHAVIOR, reg_val);

You can probably use phy_modify() here.

> +static int vsc8541_led_hw_is_supported(struct phy_device *phydev, u8 index,
> +				       unsigned long rules)
> +{
> +	struct vsc8531_private *vsc8531 = phydev->priv;
> +	static const unsigned long supported = BIT(TRIGGER_NETDEV_LINK) |
> +					       BIT(TRIGGER_NETDEV_LINK_1000) |
> +					       BIT(TRIGGER_NETDEV_LINK_100) |
> +					       BIT(TRIGGER_NETDEV_LINK_10) |
> +					       BIT(TRIGGER_NETDEV_RX) |
> +					       BIT(TRIGGER_NETDEV_TX);
> +

Reverse Christmas tree. The lines should be sorted, longest first,
shortest last.

> +static int vsc8541_led_hw_control_get(struct phy_device *phydev, u8 index,
> +				      unsigned long *rules)
> +{
> +	struct vsc8531_private *vsc8531 = phydev->priv;
> +	u16 reg;
> +
> +	if (index >= vsc8531->nleds)
> +		return -EINVAL;
> +
> +	reg = phy_read(phydev, MSCC_PHY_LED_MODE_SEL) & LED_MODE_SEL_MASK(index);

Another cause of u16, when int should be used. Please check all
instances of phy_read().

> +	reg >>= LED_MODE_SEL_POS(index);
> +	switch (reg) {
> +	case VSC8531_LINK_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;
> +
> +	case VSC8531_LINK_1000_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_LINK_1000) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;
> +
> +	case VSC8531_LINK_100_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_LINK_100) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;
> +
> +	case VSC8531_LINK_10_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_LINK_10) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;
> +
> +	case VSC8531_LINK_100_1000_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_LINK_100) |
> +			 BIT(TRIGGER_NETDEV_LINK_1000) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;
> +
> +	case VSC8531_LINK_10_1000_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_LINK_10) |
> +			 BIT(TRIGGER_NETDEV_LINK_1000) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;
> +
> +	case VSC8531_LINK_10_100_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_LINK_10) |
> +			 BIT(TRIGGER_NETDEV_LINK_100) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;
> +
> +	case VSC8531_ACTIVITY:
> +		*rules = BIT(TRIGGER_NETDEV_LINK) |
> +			 BIT(TRIGGER_NETDEV_RX) |
> +			 BIT(TRIGGER_NETDEV_TX);
> +		break;

Should the combine bit be taken into account here?

> @@ -2343,6 +2532,26 @@ static int vsc85xx_probe(struct phy_device *phydev)
>  	if (!vsc8531->stats)
>  		return -ENOMEM;
>  
> +	phy_id = phydev->drv->phy_id & phydev->drv->phy_id_mask;
> +	if (phy_id == PHY_ID_VSC8541) {

The VSC8541 has its own probe function, vsc8514_probe(). Why is this
needed?

    Andrew

---
pw-bot: cr