Hi,
> -----Original Message-----
> From: Christophe JAILLET <christophe.jaillet@xxxxxxxxxx>
> Sent: Saturday, September 14, 2024 6:03 PM
> Subject: [PATCH v4] phy: renesas: rcar-gen3-usb2: Fix an error handling path in
> rcar_gen3_phy_usb2_probe()
>
> If an error occurs after the reset_control_deassert(),
> reset_control_assert() must be called, as already done in the remove function.
>
> Use devm_add_action_or_reset() to add the missing call and simplify the
> .remove() function accordingly.
>
> Fixes: 4eae16375357 ("phy: renesas: rcar-gen3-usb2: Add support to initialize the bus")
> Signed-off-by: Christophe JAILLET <christophe.jaillet@xxxxxxxxxx>
> ---
> This patch changes the order of function calls when releasing the resources in the .remove function().
> Looks fine to me, but pm_ functions are sometimes tricky.
>
> Changes in v4:
> - Use the error handling path in rcar_gen3_phy_usb2_init_bus()
>
> Changes in v3:
> - Use devm_add_action_or_reset() [Biju Das]
> v3:
> https://lore.kernel.org/all/290b25827e3f0742808940719455ff0c5cb9d01d.1726329925.git.christophe.jaillet
> @wanadoo.fr/
>
> Changes in v2: (broken proposal)
> - Re-use 'error' to simplify the patch [claudiu beznea]
> - Update the commit description to explain why it is safe.
> v2:
> https://lore.kernel.org/all/4efe2d0419cbe98163e2422ebe0c7896b8a5efed.1725717505.git.christophe.jaillet
> @wanadoo.fr/
>
> v1:
> https://lore.kernel.org/all/fc9f7b444f0ca645411868992bbe16514aeccfed.1725652654.git.christophe.jaillet
> @wanadoo.fr/
> ---
> drivers/phy/renesas/phy-rcar-gen3-usb2.c | 13 ++++++++++++-
> 1 file changed, 12 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/phy/renesas/phy-rcar-gen3-usb2.c b/drivers/phy/renesas/phy-rcar-gen3-usb2.c
> index 58e123305152..f900fe42c311 100644
> --- a/drivers/phy/renesas/phy-rcar-gen3-usb2.c
> +++ b/drivers/phy/renesas/phy-rcar-gen3-usb2.c
> @@ -668,6 +668,13 @@ static enum usb_dr_mode rcar_gen3_get_dr_mode(struct device_node *np)
> return candidate;
> }
>
> +static void rcar_gen3_reset_assert(void *data) {
> + struct reset_control *rstc = data;
nit:
Maybe since it is void*, you could avoid this assignment and use data instead??
Reviewed-by: Biju Das <biju.das.jz@xxxxxxxxxxxxxx>
Cheers,
Biju
> +
> + reset_control_assert(rstc);
> +}
> +
> static int rcar_gen3_phy_usb2_init_bus(struct rcar_gen3_chan *channel) {
> struct device *dev = channel->dev;
> @@ -686,6 +693,11 @@ static int rcar_gen3_phy_usb2_init_bus(struct rcar_gen3_chan *channel)
> if (ret)
> goto rpm_put;
>
> + ret = devm_add_action_or_reset(dev, rcar_gen3_reset_assert,
> + channel->rstc);
> + if (ret)
> + goto rpm_put;
> +
> val = readl(channel->base + USB2_AHB_BUS_CTR);
> val &= ~USB2_AHB_BUS_CTR_MBL_MASK;
> val |= USB2_AHB_BUS_CTR_MBL_INCR4;
> @@ -815,7 +827,6 @@ static void rcar_gen3_phy_usb2_remove(struct platform_device *pdev)
> if (channel->is_otg_channel)
> device_remove_file(&pdev->dev, &dev_attr_role);
>
> - reset_control_assert(channel->rstc);
> pm_runtime_disable(&pdev->dev);
> };
>
> --
> 2.46.0
