[PATCH net-next] net: enetc: kill PHY-less mode for PFs

[PATCH net-next] net: enetc: kill PHY-less mode for PFs

[Vladimir Oltean]

Right now, a PHY-less port (no phy-mode, no fixed-link, no phy-handle)
doesn't register with phylink, but calls netif_carrier_on() from
enetc_start().

This makes sense for a VF, but for a PF, this is braindead, because we
never call enetc_mac_enable() so the MAC is left inoperational.
Furthermore, commit 71b77a7a27a3 ("enetc: Migrate to PHYLINK and
PCS_LYNX") put the nail in the coffin because it removed the initial
netif_carrier_off() call done right after register_netdev().

Without that call, netif_carrier_on() does not call
linkwatch_fire_event(), so the operstate remains IF_OPER_UNKNOWN.

Just deny the broken configuration by requiring that a phy-mode is
present, and always register a PF with phylink.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
---
 .../net/ethernet/freescale/enetc/enetc_pf.c   | 24 +++++++++++--------
 1 file changed, 14 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/freescale/enetc/enetc_pf.c b/drivers/net/ethernet/freescale/enetc/enetc_pf.c
index 7cccdf54359f..5cc2847ad05e 100644
--- a/drivers/net/ethernet/freescale/enetc/enetc_pf.c
+++ b/drivers/net/ethernet/freescale/enetc/enetc_pf.c
@@ -1105,8 +1105,7 @@ static int enetc_phylink_create(struct enetc_ndev_priv *priv,
 
 static void enetc_phylink_destroy(struct enetc_ndev_priv *priv)
 {
-	if (priv->phylink)
-		phylink_destroy(priv->phylink);
+	phylink_destroy(priv->phylink);
 }
 
 /* Initialize the entire shared memory for the flow steering entries
@@ -1273,16 +1272,20 @@ static int enetc_pf_probe(struct pci_dev *pdev,
 		goto err_alloc_msix;
 	}
 
-	if (!of_get_phy_mode(node, &pf->if_mode)) {
-		err = enetc_mdiobus_create(pf, node);
-		if (err)
-			goto err_mdiobus_create;
-
-		err = enetc_phylink_create(priv, node);
-		if (err)
-			goto err_phylink_create;
+	err = of_get_phy_mode(node, &pf->if_mode);
+	if (err) {
+		dev_err(&pdev->dev, "Failed to read PHY mode\n");
+		goto err_phy_mode;
 	}
 
+	err = enetc_mdiobus_create(pf, node);
+	if (err)
+		goto err_mdiobus_create;
+
+	err = enetc_phylink_create(priv, node);
+	if (err)
+		goto err_phylink_create;
+
 	err = register_netdev(ndev);
 	if (err)
 		goto err_reg_netdev;
@@ -1296,6 +1299,7 @@ static int enetc_pf_probe(struct pci_dev *pdev,
 err_mdiobus_create:
 	enetc_free_msix(priv);
 err_config_si:
+err_phy_mode:
 err_alloc_msix:
 	enetc_free_si_resources(priv);
 err_alloc_si_res:
-- 
2.25.1

RE: [PATCH net-next] net: enetc: kill PHY-less mode for PFs

[Claudiu Manoil]

> -----Original Message-----
> From: Vladimir Oltean <vladimir.oltean@nxp.com>
> Sent: Tuesday, May 10, 2022 7:40 PM
[...]
> Subject: [PATCH net-next] net: enetc: kill PHY-less mode for PFs
> 
[...]
> +	err = of_get_phy_mode(node, &pf->if_mode);
> +	if (err) {
> +		dev_err(&pdev->dev, "Failed to read PHY mode\n");
> +		goto err_phy_mode;
>  	}
> 
> +	err = enetc_mdiobus_create(pf, node);
> +	if (err)
> +		goto err_mdiobus_create;
> +
> +	err = enetc_phylink_create(priv, node);
> +	if (err)
> +		goto err_phylink_create;
> +
>  	err = register_netdev(ndev);
>  	if (err)
>  		goto err_reg_netdev;
> @@ -1296,6 +1299,7 @@ static int enetc_pf_probe(struct pci_dev *pdev,
>  err_mdiobus_create:
>  	enetc_free_msix(priv);
>  err_config_si:
> +err_phy_mode:

err_phy_mode should bail out at enetc_free_msix()

>  err_alloc_msix:
>  	enetc_free_si_resources(priv);
>  err_alloc_si_res:
> --
> 2.25.1

Re: [PATCH net-next] net: enetc: kill PHY-less mode for PFs

[Vladimir Oltean]

On Wed, May 11, 2022 at 07:28:36AM +0000, Claudiu Manoil wrote:
> err_phy_mode should bail out at enetc_free_msix()

thanks for noticing.