On 10/24/25 01:59, Hal Feng wrote:
> Get and enable a optional power gpio. This feature is ported
> from the jh7110 pcie driver in Linux. VisionFive 2 Lite needs
> this gpio to enable the PCI bus device (M.2 M-Key) power.
>
That's alright. Good.
> Signed-off-by: Hal Feng <[email protected]>
> ---
> drivers/pci/pcie_starfive_jh7110.c | 8 ++++++++
> 1 file changed, 8 insertions(+)
>
> diff --git a/drivers/pci/pcie_starfive_jh7110.c
> b/drivers/pci/pcie_starfive_jh7110.c
> index 0908ae16b67..04088b48ddc 100644
> --- a/drivers/pci/pcie_starfive_jh7110.c
> +++ b/drivers/pci/pcie_starfive_jh7110.c
> @@ -45,6 +45,7 @@ struct starfive_pcie {
> struct pcie_plda plda;
> struct clk_bulk clks;
> struct reset_ctl_bulk rsts;
> + struct gpio_desc power_gpio;
> struct gpio_desc reset_gpio;
> struct regmap *regmap;
> unsigned int stg_pcie_base;
> @@ -184,6 +185,10 @@ static int starfive_pcie_parse_dt(struct udevice *dev)
> dev_err(dev, "reset-gpio is not valid\n");
> return -EINVAL;
> }
> +
> + gpio_request_by_name(dev, "enable-gpios", 0, &priv->power_gpio,
> + GPIOD_IS_OUT);
> +
Are we missing some error handling here? I don't know, everything else
above in the function is verbose with the dev_err() reporting.
> return 0;
> }
>
> @@ -205,6 +210,9 @@ static int starfive_pcie_init_port(struct udevice *dev)
> goto err_deassert_clk;
> }
>
> + if (dm_gpio_is_valid(&priv->power_gpio))
> + dm_gpio_set_value(&priv->power_gpio, 1);
> +
> dm_gpio_set_value(&priv->reset_gpio, 1);
> /* Disable physical functions except #0 */
> for (i = 1; i < PLDA_FUNC_NUM; i++) {
With that,
Reviewed-by: E Shattow <[email protected]>
