[PATCHv2 net-next 0/2] net: thunder_bgx: convert fwnode to OF

[PATCHv2 net-next 0/2] net: thunder_bgx: convert fwnode to OF

[Rosen Penev]

NVMEM support addition and fwnode removals in favor of OF

v2: split up patch

Rosen Penev (2):
  net: thunder_bgx: check for MAC probe defer
  net: thunder_bgx: use OF loop instead of fwnode

 .../net/ethernet/cavium/thunder/thunder_bgx.c | 24 ++++++++-----------
 1 file changed, 10 insertions(+), 14 deletions(-)

-- 
2.51.0

[PATCHv2 net-next 1/2] net: thunder_bgx: check for MAC probe defer

[Rosen Penev]

of_get_mac_address supports NVMEM, which can load after the driver.
Check for it and defer in such a case.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
---
 drivers/net/ethernet/cavium/thunder/thunder_bgx.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
index 9efb60842ad1..a68dccb7c2da 100644
--- a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
+++ b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
@@ -1471,6 +1471,7 @@ static int bgx_init_of_phy(struct bgx *bgx)
 	struct fwnode_handle *fwn;
 	struct device_node *node = NULL;
 	u8 lmac = 0;
+	int err;
 
 	device_for_each_child_node(&bgx->pdev->dev, fwn) {
 		struct phy_device *pd;
@@ -1483,7 +1484,9 @@ static int bgx_init_of_phy(struct bgx *bgx)
 		if (!node)
 			break;
 
-		of_get_mac_address(node, bgx->lmac[lmac].mac);
+		err = of_get_mac_address(node, bgx->lmac[lmac].mac);
+		if (err == -EPROBE_DEFER)
+			goto defer;
 
 		SET_NETDEV_DEV(bgx->lmac[lmac].netdev, &bgx->pdev->dev);
 		bgx->lmac[lmac].lmacid = lmac;
-- 
2.51.0

[PATCHv2 net-next 2/2] net: thunder_bgx: use OF loop instead of fwnode

[Rosen Penev]

This is implemented under CONFIG_OF_MDIO, which already assumes an OF
node. Also the loop already checks to see if of_node is NULL before
proceeding.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
---
 .../net/ethernet/cavium/thunder/thunder_bgx.c | 23 +++++++------------
 1 file changed, 8 insertions(+), 15 deletions(-)

diff --git a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
index a68dccb7c2da..06ca305de1b4 100644
--- a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
+++ b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
@@ -1468,30 +1468,23 @@ static int bgx_init_acpi_phy(struct bgx *bgx)
 
 static int bgx_init_of_phy(struct bgx *bgx)
 {
-	struct fwnode_handle *fwn;
-	struct device_node *node = NULL;
+	struct device_node *node = bgx->pdev->dev.of_node;
+	struct device_node *child;
 	u8 lmac = 0;
-	int err;
 
-	device_for_each_child_node(&bgx->pdev->dev, fwn) {
+	for_each_child_of_node(node, child) {
 		struct phy_device *pd;
 		struct device_node *phy_np;
+		int err;
 
-		/* Should always be an OF node.  But if it is not, we
-		 * cannot handle it, so exit the loop.
-		 */
-		node = to_of_node(fwn);
-		if (!node)
-			break;
-
-		err = of_get_mac_address(node, bgx->lmac[lmac].mac);
+		err = of_get_mac_address(child, bgx->lmac[lmac].mac);
 		if (err == -EPROBE_DEFER)
 			goto defer;
 
 		SET_NETDEV_DEV(bgx->lmac[lmac].netdev, &bgx->pdev->dev);
 		bgx->lmac[lmac].lmacid = lmac;
 
-		phy_np = of_parse_phandle(node, "phy-handle", 0);
+		phy_np = of_parse_phandle(child, "phy-handle", 0);
 		/* If there is no phy or defective firmware presents
 		 * this cortina phy, for which there is no driver
 		 * support, ignore it.
@@ -1511,7 +1504,7 @@ static int bgx_init_of_phy(struct bgx *bgx)
 
 		lmac++;
 		if (lmac == bgx->max_lmac) {
-			of_node_put(node);
+			of_node_put(child);
 			break;
 		}
 	}
@@ -1528,7 +1521,7 @@ static int bgx_init_of_phy(struct bgx *bgx)
 			bgx->lmac[lmac].phydev = NULL;
 		}
 	}
-	of_node_put(node);
+	of_node_put(child);
 	return -EPROBE_DEFER;
 }
 
-- 
2.51.0

Re: [PATCHv2 net-next 1/2] net: thunder_bgx: check for MAC probe defer

[Andrew Lunn]

On Thu, Sep 04, 2025 at 02:32:27PM -0700, Rosen Penev wrote:
> of_get_mac_address supports NVMEM, which can load after the driver.
> Check for it and defer in such a case.
> 
> Signed-off-by: Rosen Penev <rosenp@gmail.com>

Reviewed-by: Andrew Lunn <andrew@lunn.ch>

    Andrew

Re: [PATCHv2 net-next 2/2] net: thunder_bgx: use OF loop instead of fwnode

[Andrew Lunn]

On Thu, Sep 04, 2025 at 02:32:28PM -0700, Rosen Penev wrote:
> This is implemented under CONFIG_OF_MDIO, which already assumes an OF
> node.

I'm not sure that is correct.

bgx_probe() -> bgx_init_phy() -> bgx_init_of_phy()

static int bgx_init_phy(struct bgx *bgx)
{
	if (!acpi_disabled)
		return bgx_init_acpi_phy(bgx);

	return bgx_init_of_phy(bgx);
}

There is no check made to see if this is an ACPI node and so
bgx_init_acpi_phy() is called. So it might be an ACPI node is passed
to bgx_init_of_phy().

How have you tested this code? On an ACPI machine, or only a DT
machine?

	Andrew