Re: [PATCH v3 15/23] net: ravb: Allocate bb_miiphy using bb_miiphy_alloc() and fill in callbacks

[Marek Vasut]

On 2/28/25 2:28 PM, Paul Barker wrote:

On 22/02/2025 20:33, Marek Vasut wrote:

Allocate bb_miiphy using bb_miiphy_alloc() and fill in callbacks
currently listed in bb_miiphy_buses[] array. This is a temporary
duplication of assignment to avoid breakage, which will be removed
in follow up patches. At this point, the bb_miiphy callbacks can
reach these accessors by doing container_of() on struct mii_dev.

Reviewed-by: Paul Barker <paul.barker...@bp.renesas.com>
Signed-off-by: Marek Vasut <marek.vasut+rene...@mailbox.org>
---
Cc: Christian Marangi <ansuels...@gmail.com>
Cc: Ilias Apalodimas <ilias.apalodi...@linaro.org>
Cc: Jerome Forissier <jerome.foriss...@linaro.org>
Cc: Jim Liu <jjl...@nuvoton.com>
Cc: Joe Hershberger <joe.hershber...@ni.com>
Cc: Mario Six <mario....@gdsys.cc>
Cc: Michal Simek <michal.si...@amd.com>
Cc: Nobuhiro Iwamatsu <iwama...@nigauri.org>
Cc: Paul Barker <paul.barker...@bp.renesas.com>
Cc: Ramon Fried <rfried....@gmail.com>
Cc: Simon Glass <s...@chromium.org>
Cc: Sughosh Ganu <sughosh.g...@linaro.org>
Cc: Tom Rini <tr...@konsulko.com>
Cc: u-boot@lists.denx.de
---
V2: No change
V3: Add RB from Paul
---
  drivers/net/ravb.c | 19 ++++++++++++++++---
  1 file changed, 16 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ravb.c b/drivers/net/ravb.c
index 381cf250ea2..0018b694ec1 100644
--- a/drivers/net/ravb.c
+++ b/drivers/net/ravb.c
@@ -553,6 +553,7 @@ static int ravb_probe(struct udevice *dev)
  {
        struct eth_pdata *pdata = dev_get_plat(dev);
        struct ravb_priv *eth = dev_get_priv(dev);
+       struct bb_miiphy_bus *bb_miiphy;
        struct mii_dev *mdiodev;
        void __iomem *iobase;
        int ret;
@@ -564,17 +565,29 @@ static int ravb_probe(struct udevice *dev)
        if (ret < 0)
                goto err_mdio_alloc;

- mdiodev = mdio_alloc();

-       if (!mdiodev) {
+       bb_miiphy = bb_miiphy_alloc();
+       if (!bb_miiphy) {
                ret = -ENOMEM;
                goto err_mdio_alloc;
        }

+ mdiodev = &bb_miiphy->mii;

+
        mdiodev->read = bb_miiphy_read;
        mdiodev->write = bb_miiphy_write;
        bb_miiphy_buses[0].priv = eth;
        snprintf(mdiodev->name, sizeof(mdiodev->name), dev->name);

+ /* Copy the bus accessors, name and private data */

+       bb_miiphy->mdio_active = ravb_bb_mdio_active;
+       bb_miiphy->mdio_tristate = ravb_bb_mdio_tristate;
+       bb_miiphy->set_mdio = ravb_bb_set_mdio;
+       bb_miiphy->get_mdio = ravb_bb_get_mdio;
+       bb_miiphy->set_mdc = ravb_bb_set_mdc;
+       bb_miiphy->delay = ravb_bb_delay;
+       strlcpy(bb_miiphy->name, "ravb", MDIO_NAME_LEN);
+       bb_miiphy->priv = eth;
+
        ret = mdio_register(mdiodev);
        if (ret < 0)
                goto err_mdio_register;
@@ -599,7 +612,7 @@ static int ravb_probe(struct udevice *dev)
  err_mdio_reset:
        clk_release_bulk(&eth->clks);
  err_mdio_register:
-       mdio_free(mdiodev);
+       bb_miiphy_free(bb_miiphy);
  err_mdio_alloc:
        unmap_physmem(eth->iobase, MAP_NOCACHE);
        return ret;

Marek,

I've rebased my RZ/G2L Ethernet patches on top of this series and as
part of tidying things up I spotted an issue that I'd missed -
ravb_remove() still calls mdio_free(eth->bus). The error paths in the
probe functions have been converted but the remove functions were
missed. It should instead use something like:

     struct bb_miiphy_bus *bus = container_of(eth->bus, struct bb_miiphy_bus, 
mii);
     bb_miiphy_free(bus);

The same applies to the following two patches (sh_eth.c & designware.c).

This should be fixed in

[PATCH 01/11] net: miiphybb: Split off struct bb_miiphy_bus_ops