Re: [PATCH net-next v7 10/10] net: airoha: Support multiple LAN/WAN interfaces for hw MAC address configuration

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Device probe fails when multiple LAN/WAN interfaces are assigned randomly generated MAC addresses.
--

commit 7861c44e7e1da4203e80f67d02418deac7e491b4
Author: Lorenzo Bianconi <lorenzo@kernel.org>

net: airoha: Support multiple LAN/WAN interfaces for hw MAC address configuration

This commit extends the hardware MAC address configuration logic in the airoha
driver to support multiple LAN or WAN interfaces by deriving minimum and
maximum values for the lower MAC bytes, while expecting the upper MAC bytes to
be shared.

> diff --git a/drivers/net/ethernet/airoha/airoha_eth.c b/drivers/net/ethernet/airoha/airoha_eth.c
> --- a/drivers/net/ethernet/airoha/airoha_eth.c
> +++ b/drivers/net/ethernet/airoha/airoha_eth.c
> @@ -71,20 +71,69 @@ static void airoha_qdma_irq_disable(struct airoha_irq_bank *irq_bank,
>  	airoha_qdma_set_irqmask(irq_bank, index, mask, 0);
>  }
>  
> -static void airoha_set_macaddr(struct airoha_gdm_dev *dev, const u8 *addr)
> +static int airoha_set_macaddr(struct airoha_gdm_dev *dev, const u8 *addr)
>  {
> +	u8 ref_addr[ETH_ALEN] __aligned(2);
>  	struct airoha_eth *eth = dev->eth;

[ ... ]

> +			netdev = iter_dev->dev;
> +			if (netdev->reg_state != NETREG_REGISTERED)
> +				continue;
> +
> +			ether_addr_copy(ref_addr, netdev->dev_addr);
> +			val = (netdev->dev_addr[3] << 16) |
> +			      (netdev->dev_addr[4] << 8) | netdev->dev_addr[5];
> +			if (val < lmin)
> +				lmin = val;
> +			if (val > lmax)
> +				lmax = val;
> +		}
> +	}
> +
> +	if (!is_zero_ether_addr(ref_addr) && memcmp(ref_addr, addr, 3)) {

If the device tree lacks a MAC address for an interface and the driver assigns
a random one via eth_hw_addr_random(), will this comparison fail during device
probe?

Since eth_hw_addr_random() generates independent random bytes for each
interface, the upper 3 bytes will likely mismatch. This returns -EINVAL and
propagates up to register_netdev(), which causes the driver probe to abort.

Prior to this patch, the hardware MAC register was overwritten with the newest
interface's MAC, but the probe continued successfully. Can this be adjusted to
handle randomly generated MAC addresses without aborting the probe?

> +		/* According to the HW design, hw mac address MS bits
> +		 * must be the same for each net_device with the same
> +		 * LAN/WAN configuration.
> +		 */
> +		netdev_err(dev->dev, "wrong mac addr for shared GDM port\n");
> +		return -EINVAL;
> +	}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260516-airoha-eth-multi-serdes-v7-0-99e0093303e2@kernel.org?part=10