Re: [PATCH v2 4/6] net: sun8i-emac: add v3s variant

[Andre Przywara]

On Sun, 23 May 2021 01:22:38 +0200
Andreas Rehn <rehn.an...@gmail.com> wrote:

Hi,

> Add variant V3S_EMAC.
> Handle pinmux compile time error by skipping goio setup, because
> V3s uses internal phy and don't expose pins.
>
> Signed-off-by: Andreas Rehn <rehn.an...@gmail.com>

Reviewed-by: Andre Przywara <andre.p...@arm.com>

Cheers,
Andre

> ---
> Changes in v2:
> - skip pinmux and add proper description
> - Add V3S variant add it to compatible list
> - Skip (R)GMII flags and handle sun8i_handle_internal_phy
>
> drivers/net/sun8i_emac.c | 15 ++++++++++++---
> 1 file changed, 12 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/sun8i_emac.c b/drivers/net/sun8i_emac.c
> index 5a1b38bf80..ab9f61994c 100644
> --- a/drivers/net/sun8i_emac.c
> +++ b/drivers/net/sun8i_emac.c
> @@ -145,6 +145,7 @@ enum emac_variant {
> A64_EMAC,
> R40_GMAC,
> H6_EMAC,
> + V3S_EMAC,
> };
>
> struct emac_dma_desc {
> @@ -303,7 +304,7 @@ static void sun8i_adjust_link(struct emac_eth_dev *priv,
> static u32 sun8i_emac_set_syscon_ephy(struct emac_eth_dev *priv, u32 reg)
> {
> if (priv->use_internal_phy) {
> - /* H3 based SoC's that has an Internal 100MBit PHY
> + /* H3 and V3s based SoC's that has an Internal 100MBit PHY
> * needs to be configured and powered up before use
> */
> reg &= ~H3_EPHY_DEFAULT_MASK;
> @@ -354,7 +355,8 @@ static int sun8i_emac_set_syscon(struct sun8i_eth_pdata *pdata,
> case PHY_INTERFACE_MODE_RGMII_ID:
> case PHY_INTERFACE_MODE_RGMII_RXID:
> case PHY_INTERFACE_MODE_RGMII_TXID:
> - reg |= SC_EPIT | SC_ETCS_INT_GMII;
> + if (priv->variant != V3S_EMAC)
> + reg |= SC_EPIT | SC_ETCS_INT_GMII;
> break;
> case PHY_INTERFACE_MODE_RMII:
> if (priv->variant == H3_EMAC ||
> @@ -566,6 +568,10 @@ static int parse_phy_pins(struct udevice *dev)
> iomux = SUN8I_IOMUX;
> else if (IS_ENABLED(CONFIG_MACH_SUN50I))
> iomux = SUN8I_IOMUX;
> + else if (IS_ENABLED(CONFIG_MACH_SUN8I_V3S))
> + // V3s does not expose any MAC pins,
> + // but case is required to handle BUILD_BUG_ON_MSG.
> + return 0;
> else
> BUILD_BUG_ON_MSG(1, "missing pinmux value for Ethernet pins");
>
> @@ -956,7 +962,8 @@ static int sun8i_emac_eth_of_to_plat(struct udevice *dev)
> return -EINVAL;
> }
>
> - if (priv->variant == H3_EMAC) {
> + if (priv->variant == H3_EMAC ||
> + priv->variant == V3S_EMAC) {
> ret = sun8i_handle_internal_phy(dev, priv);
> if (ret)
> return ret;
> @@ -1009,6 +1016,8 @@ static const struct udevice_id sun8i_emac_eth_ids[] = {
> .data = (uintptr_t)R40_GMAC },
> {.compatible = "allwinner,sun50i-h6-emac",
> .data = (uintptr_t)H6_EMAC },
> + {.compatible = "allwinner,sun8i-v3s-emac",
> + .data = (uintptr_t)V3S_EMAC },
> { }
> };
>

[Jagan Teki]

On Sun, May 23, 2021 at 4:52 AM Andreas Rehn <rehn.an...@gmail.com> wrote:
>
> Add variant V3S_EMAC.
> Handle pinmux compile time error by skipping goio setup, because
> V3s uses internal phy and don't expose pins.
>
> Signed-off-by: Andreas Rehn <rehn.an...@gmail.com>

Wrong multi-line comment. please fix it?

Jagan.