Re: [PATCH] net: phy: dp83822: Fix NULL pointer dereference on DP83825 devices

[Maxime Chevallier <maxime.chevallier@bootlin.com>]

Hello Tomas,

On Fri, 06 Sep 2024 10:38:40 +0200 (CEST)
"Tomas Paukrt" <tomaspaukrt@email.cz> wrote:

> The probe() function is only used for DP83822 and DP83826 models,
> leaving the private data pointer uninitialized for the DP83825 models
> which causes a NULL pointer dereference in the recently changed functions
> dp8382x_config_init() and dp83822_set_wol().
> 
> Add the dp8382x_probe() function, so all PHY models will have a valid
> private data pointer to prevent similar issues in the future.
> 
> Fixes: 9ef9ecfa9e9f ("net: phy: dp8382x: keep WOL settings across suspends")
> Signed-off-by: Tomas Paukrt <tomaspaukrt@email.cz>

The dp83825 does seem to support WoL, so allocating the private data is
indeed the way to go here.

As this is a fix, you should send the patch targetting the "net" tree,
so you need to include that in the patch subject (use the
--subject-prefix="PATCH net" option for git format-patch).

> ---
>  drivers/net/phy/dp83822.c | 21 +++++++++++++++++----
>  1 file changed, 17 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/phy/dp83822.c b/drivers/net/phy/dp83822.c
> index efeb643..58877c0 100644
> --- a/drivers/net/phy/dp83822.c
> +++ b/drivers/net/phy/dp83822.c
> @@ -271,8 +271,7 @@ static int dp83822_config_intr(struct phy_device *phydev)
>  				DP83822_ENERGY_DET_INT_EN |
>  				DP83822_LINK_QUAL_INT_EN);
>  
> -		/* Private data pointer is NULL on DP83825 */
> -		if (!dp83822 || !dp83822->fx_enabled)
> +		if (!dp83822->fx_enabled)
>  			misr_status |= DP83822_ANEG_COMPLETE_INT_EN |
>  				       DP83822_DUP_MODE_CHANGE_INT_EN |
>  				       DP83822_SPEED_CHANGED_INT_EN;
> @@ -292,8 +291,7 @@ static int dp83822_config_intr(struct phy_device *phydev)
>  				DP83822_PAGE_RX_INT_EN |
>  				DP83822_EEE_ERROR_CHANGE_INT_EN);
>  
> -		/* Private data pointer is NULL on DP83825 */
> -		if (!dp83822 || !dp83822->fx_enabled)
> +		if (!dp83822->fx_enabled)
>  			misr_status |= DP83822_ANEG_ERR_INT_EN |
>  				       DP83822_WOL_PKT_INT_EN;
>  
> @@ -731,6 +729,20 @@ static int dp83826_probe(struct phy_device *phydev)
>  	return 0;
>  }
>  
> +static int dp8382x_probe(struct phy_device *phydev)
> +{
> +	struct dp83822_private *dp83822;
> +
> +	dp83822 = devm_kzalloc(&phydev->mdio.dev, sizeof(*dp83822),
> +			       GFP_KERNEL);
> +	if (!dp83822)
> +		return -ENOMEM;
> +
> +	phydev->priv = dp83822;
> +
> +	return 0;
> +}

Now all the PHYs from that driver use a probe() function that does that
same allocation and sets the phydev->priv, maybe we could have
dp83826_probe() and dp83826_probe() call this newly introduced function,
to avoid some duplication ?

The rest looks good to me,

Thanks,

Maxime