| Message ID | 20230113033347.264135-1-xiaoning.wang@nxp.com |
|---|---|
| Headers | show |
| Series | Add eqos and fec support for imx93 | expand |
On 1/13/2023 11:33 AM, Clark Wang wrote: > Add imx93 platform support for dwmac-imx driver. > > Signed-off-by: Clark Wang <xiaoning.wang@nxp.com> Reviewed-by: Peng Fan <peng.fan@nxp.com> > --- > V2 change: > - change pr_debug() to dev_dbg() > --- > .../net/ethernet/stmicro/stmmac/dwmac-imx.c | 55 +++++++++++++++++-- > 1 file changed, 50 insertions(+), 5 deletions(-) > > diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c > index bd52fb7cf486..a7ea69d81c11 100644 > --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c > +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-imx.c > @@ -31,6 +31,12 @@ > #define GPR_ENET_QOS_CLK_TX_CLK_SEL (0x1 << 20) > #define GPR_ENET_QOS_RGMII_EN (0x1 << 21) > > +#define MX93_GPR_ENET_QOS_INTF_MODE_MASK GENMASK(3, 0) > +#define MX93_GPR_ENET_QOS_INTF_SEL_MII (0x0 << 1) > +#define MX93_GPR_ENET_QOS_INTF_SEL_RMII (0x4 << 1) > +#define MX93_GPR_ENET_QOS_INTF_SEL_RGMII (0x1 << 1) > +#define MX93_GPR_ENET_QOS_CLK_GEN_EN (0x1 << 0) > + > struct imx_dwmac_ops { > u32 addr_width; > bool mac_rgmii_txclk_auto_adj; > @@ -90,6 +96,35 @@ imx8dxl_set_intf_mode(struct plat_stmmacenet_data *plat_dat) > return ret; > } > > +static int imx93_set_intf_mode(struct plat_stmmacenet_data *plat_dat) > +{ > + struct imx_priv_data *dwmac = plat_dat->bsp_priv; > + int val; > + > + switch (plat_dat->interface) { > + case PHY_INTERFACE_MODE_MII: > + val = MX93_GPR_ENET_QOS_INTF_SEL_MII; > + break; > + case PHY_INTERFACE_MODE_RMII: > + val = MX93_GPR_ENET_QOS_INTF_SEL_RMII; > + break; > + case PHY_INTERFACE_MODE_RGMII: > + case PHY_INTERFACE_MODE_RGMII_ID: > + case PHY_INTERFACE_MODE_RGMII_RXID: > + case PHY_INTERFACE_MODE_RGMII_TXID: > + val = MX93_GPR_ENET_QOS_INTF_SEL_RGMII; > + break; > + default: > + dev_dbg(dwmac->dev, "imx dwmac doesn't support %d interface\n", > + plat_dat->interface); > + return -EINVAL; > + } > + > + val |= MX93_GPR_ENET_QOS_CLK_GEN_EN; > + return regmap_update_bits(dwmac->intf_regmap, dwmac->intf_reg_off, > + MX93_GPR_ENET_QOS_INTF_MODE_MASK, val); > +}; > + > static int imx_dwmac_clks_config(void *priv, bool enabled) > { > struct imx_priv_data *dwmac = priv; > @@ -188,7 +223,9 @@ imx_dwmac_parse_dt(struct imx_priv_data *dwmac, struct device *dev) > } > > dwmac->clk_mem = NULL; > - if (of_machine_is_compatible("fsl,imx8dxl")) { > + > + if (of_machine_is_compatible("fsl,imx8dxl") || > + of_machine_is_compatible("fsl,imx93")) { > dwmac->clk_mem = devm_clk_get(dev, "mem"); > if (IS_ERR(dwmac->clk_mem)) { > dev_err(dev, "failed to get mem clock\n"); > @@ -196,10 +233,11 @@ imx_dwmac_parse_dt(struct imx_priv_data *dwmac, struct device *dev) > } > } > > - if (of_machine_is_compatible("fsl,imx8mp")) { > - /* Binding doc describes the property: > - is required by i.MX8MP. > - is optional for i.MX8DXL. > + if (of_machine_is_compatible("fsl,imx8mp") || > + of_machine_is_compatible("fsl,imx93")) { > + /* Binding doc describes the propety: > + * is required by i.MX8MP, i.MX93. > + * is optinoal for i.MX8DXL. > */ > dwmac->intf_regmap = syscon_regmap_lookup_by_phandle(np, "intf_mode"); > if (IS_ERR(dwmac->intf_regmap)) > @@ -296,9 +334,16 @@ static struct imx_dwmac_ops imx8dxl_dwmac_data = { > .set_intf_mode = imx8dxl_set_intf_mode, > }; > > +static struct imx_dwmac_ops imx93_dwmac_data = { > + .addr_width = 32, > + .mac_rgmii_txclk_auto_adj = true, > + .set_intf_mode = imx93_set_intf_mode, > +}; > + > static const struct of_device_id imx_dwmac_match[] = { > { .compatible = "nxp,imx8mp-dwmac-eqos", .data = &imx8mp_dwmac_data }, > { .compatible = "nxp,imx8dxl-dwmac-eqos", .data = &imx8dxl_dwmac_data }, > + { .compatible = "nxp,imx93-dwmac-eqos", .data = &imx93_dwmac_data }, > { } > }; > MODULE_DEVICE_TABLE(of, imx_dwmac_match);
On 1/13/2023 11:33 AM, Clark Wang wrote: > Add EQoS node for imx93 platform. > > Signed-off-by: Clark Wang <xiaoning.wang@nxp.com> Reviewed-by: Peng Fan <peng.fan@nxp.com> > --- > New patch added in V2, split dtsi and dts changes into separate patches > --- > arch/arm64/boot/dts/freescale/imx93.dtsi | 22 ++++++++++++++++++++++ > 1 file changed, 22 insertions(+) > > diff --git a/arch/arm64/boot/dts/freescale/imx93.dtsi b/arch/arm64/boot/dts/freescale/imx93.dtsi > index 6808321ed809..ff2253cb7d4a 100644 > --- a/arch/arm64/boot/dts/freescale/imx93.dtsi > +++ b/arch/arm64/boot/dts/freescale/imx93.dtsi > @@ -564,6 +564,28 @@ usdhc2: mmc@42860000 { > status = "disabled"; > }; > > + eqos: ethernet@428a0000 { > + compatible = "nxp,imx93-dwmac-eqos", "snps,dwmac-5.10a"; > + reg = <0x428a0000 0x10000>; > + interrupts = <GIC_SPI 183 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 184 IRQ_TYPE_LEVEL_HIGH>; > + interrupt-names = "eth_wake_irq", "macirq"; > + clocks = <&clk IMX93_CLK_ENET_QOS_GATE>, > + <&clk IMX93_CLK_ENET_QOS_GATE>, > + <&clk IMX93_CLK_ENET_TIMER2>, > + <&clk IMX93_CLK_ENET>, > + <&clk IMX93_CLK_ENET_QOS_GATE>; > + clock-names = "stmmaceth", "pclk", "ptp_ref", "tx", "mem"; > + assigned-clocks = <&clk IMX93_CLK_ENET_TIMER2>, > + <&clk IMX93_CLK_ENET>; > + assigned-clock-parents = <&clk IMX93_CLK_SYS_PLL_PFD1_DIV2>, > + <&clk IMX93_CLK_SYS_PLL_PFD0_DIV2>; > + assigned-clock-rates = <100000000>, <250000000>; > + intf_mode = <&wakeupmix_gpr 0x28>; > + clk_csr = <0>; > + status = "disabled"; > + }; > + > usdhc3: mmc@428b0000 { > compatible = "fsl,imx93-usdhc", "fsl,imx8mm-usdhc"; > reg = <0x428b0000 0x10000>;
On 1/13/2023 11:33 AM, Clark Wang wrote: > Add FEC node for imx93 platform. > > Signed-off-by: Clark Wang <xiaoning.wang@nxp.com> Reviewed-by: Peng Fan <peng.fan@nxp.com> > --- > New patch added in V2 for FEC > --- > arch/arm64/boot/dts/freescale/imx93.dtsi | 26 ++++++++++++++++++++++++ > 1 file changed, 26 insertions(+) > > diff --git a/arch/arm64/boot/dts/freescale/imx93.dtsi b/arch/arm64/boot/dts/freescale/imx93.dtsi > index ff2253cb7d4a..12e0350ad35a 100644 > --- a/arch/arm64/boot/dts/freescale/imx93.dtsi > +++ b/arch/arm64/boot/dts/freescale/imx93.dtsi > @@ -586,6 +586,32 @@ eqos: ethernet@428a0000 { > status = "disabled"; > }; > > + fec: ethernet@42890000 { > + compatible = "fsl,imx93-fec", "fsl,imx8mq-fec", "fsl,imx6sx-fec"; > + reg = <0x42890000 0x10000>; > + interrupts = <GIC_SPI 179 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 180 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 181 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 182 IRQ_TYPE_LEVEL_HIGH>; > + clocks = <&clk IMX93_CLK_ENET1_GATE>, > + <&clk IMX93_CLK_ENET1_GATE>, > + <&clk IMX93_CLK_ENET_TIMER1>, > + <&clk IMX93_CLK_ENET_REF>, > + <&clk IMX93_CLK_ENET_REF_PHY>; > + clock-names = "ipg", "ahb", "ptp", > + "enet_clk_ref", "enet_out"; > + assigned-clocks = <&clk IMX93_CLK_ENET_TIMER1>, > + <&clk IMX93_CLK_ENET_REF>, > + <&clk IMX93_CLK_ENET_REF_PHY>; > + assigned-clock-parents = <&clk IMX93_CLK_SYS_PLL_PFD1_DIV2>, > + <&clk IMX93_CLK_SYS_PLL_PFD0_DIV2>, > + <&clk IMX93_CLK_SYS_PLL_PFD1_DIV2>; > + assigned-clock-rates = <100000000>, <250000000>, <50000000>; > + fsl,num-tx-queues = <3>; > + fsl,num-rx-queues = <3>; > + status = "disabled"; > + }; > + > usdhc3: mmc@428b0000 { > compatible = "fsl,imx93-usdhc", "fsl,imx8mm-usdhc"; > reg = <0x428b0000 0x10000>;
On 1/13/2023 11:33 AM, Clark Wang wrote: > Enable EQoS function for imx93-11x11-evk board. > > Signed-off-by: Clark Wang <xiaoning.wang@nxp.com> Reviewed-by: Peng Fan <peng.fan@nxp.com> > --- > New patch added in V2, split dtsi and dts changes into separate patches > --- > .../boot/dts/freescale/imx93-11x11-evk.dts | 39 +++++++++++++++++++ > 1 file changed, 39 insertions(+) > > diff --git a/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts b/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts > index 27f9a9f33134..6f7f1974cbb7 100644 > --- a/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts > +++ b/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts > @@ -35,6 +35,26 @@ &mu2 { > status = "okay"; > }; > > +&eqos { > + pinctrl-names = "default"; > + pinctrl-0 = <&pinctrl_eqos>; > + phy-mode = "rgmii-id"; > + phy-handle = <ðphy1>; > + status = "okay"; > + > + mdio { > + compatible = "snps,dwmac-mdio"; > + #address-cells = <1>; > + #size-cells = <0>; > + clock-frequency = <5000000>; > + > + ethphy1: ethernet-phy@1 { > + reg = <1>; > + eee-broken-1000t; > + }; > + }; > +}; > + > &lpuart1 { /* console */ > pinctrl-names = "default"; > pinctrl-0 = <&pinctrl_uart1>; > @@ -65,6 +85,25 @@ &usdhc2 { > }; > > &iomuxc { > + pinctrl_eqos: eqosgrp { > + fsl,pins = < > + MX93_PAD_ENET1_MDC__ENET_QOS_MDC 0x57e > + MX93_PAD_ENET1_MDIO__ENET_QOS_MDIO 0x57e > + MX93_PAD_ENET1_RD0__ENET_QOS_RGMII_RD0 0x57e > + MX93_PAD_ENET1_RD1__ENET_QOS_RGMII_RD1 0x57e > + MX93_PAD_ENET1_RD2__ENET_QOS_RGMII_RD2 0x57e > + MX93_PAD_ENET1_RD3__ENET_QOS_RGMII_RD3 0x57e > + MX93_PAD_ENET1_RXC__CCM_ENET_QOS_CLOCK_GENERATE_RX_CLK 0x5fe > + MX93_PAD_ENET1_RX_CTL__ENET_QOS_RGMII_RX_CTL 0x57e > + MX93_PAD_ENET1_TD0__ENET_QOS_RGMII_TD0 0x57e > + MX93_PAD_ENET1_TD1__ENET_QOS_RGMII_TD1 0x57e > + MX93_PAD_ENET1_TD2__ENET_QOS_RGMII_TD2 0x57e > + MX93_PAD_ENET1_TD3__ENET_QOS_RGMII_TD3 0x57e > + MX93_PAD_ENET1_TXC__CCM_ENET_QOS_CLOCK_GENERATE_TX_CLK 0x5fe > + MX93_PAD_ENET1_TX_CTL__ENET_QOS_RGMII_TX_CTL 0x57e > + >; > + }; > + > pinctrl_uart1: uart1grp { > fsl,pins = < > MX93_PAD_UART1_RXD__LPUART1_RX 0x31e
On 1/13/2023 11:33 AM, Clark Wang wrote: > Enable FEC function for imx93-11x11-evk board. > > Signed-off-by: Clark Wang <xiaoning.wang@nxp.com> Reviewed-by: Peng Fan <peng.fan@nxp.com> > --- > New patch added in V2 for FEC > --- > .../boot/dts/freescale/imx93-11x11-evk.dts | 39 +++++++++++++++++++ > 1 file changed, 39 insertions(+) > > diff --git a/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts b/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts > index 6f7f1974cbb7..cdcc5093c763 100644 > --- a/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts > +++ b/arch/arm64/boot/dts/freescale/imx93-11x11-evk.dts > @@ -55,6 +55,26 @@ ethphy1: ethernet-phy@1 { > }; > }; > > +&fec { > + pinctrl-names = "default"; > + pinctrl-0 = <&pinctrl_fec>; > + phy-mode = "rgmii-id"; > + phy-handle = <ðphy2>; > + fsl,magic-packet; > + status = "okay"; > + > + mdio { > + #address-cells = <1>; > + #size-cells = <0>; > + clock-frequency = <5000000>; > + > + ethphy2: ethernet-phy@2 { > + reg = <2>; > + eee-broken-1000t; > + }; > + }; > +}; > + > &lpuart1 { /* console */ > pinctrl-names = "default"; > pinctrl-0 = <&pinctrl_uart1>; > @@ -104,6 +124,25 @@ MX93_PAD_ENET1_TX_CTL__ENET_QOS_RGMII_TX_CTL 0x57e > >; > }; > > + pinctrl_fec: fecgrp { > + fsl,pins = < > + MX93_PAD_ENET2_MDC__ENET1_MDC 0x57e > + MX93_PAD_ENET2_MDIO__ENET1_MDIO 0x57e > + MX93_PAD_ENET2_RD0__ENET1_RGMII_RD0 0x57e > + MX93_PAD_ENET2_RD1__ENET1_RGMII_RD1 0x57e > + MX93_PAD_ENET2_RD2__ENET1_RGMII_RD2 0x57e > + MX93_PAD_ENET2_RD3__ENET1_RGMII_RD3 0x57e > + MX93_PAD_ENET2_RXC__ENET1_RGMII_RXC 0x5fe > + MX93_PAD_ENET2_RX_CTL__ENET1_RGMII_RX_CTL 0x57e > + MX93_PAD_ENET2_TD0__ENET1_RGMII_TD0 0x57e > + MX93_PAD_ENET2_TD1__ENET1_RGMII_TD1 0x57e > + MX93_PAD_ENET2_TD2__ENET1_RGMII_TD2 0x57e > + MX93_PAD_ENET2_TD3__ENET1_RGMII_TD3 0x57e > + MX93_PAD_ENET2_TXC__ENET1_RGMII_TXC 0x5fe > + MX93_PAD_ENET2_TX_CTL__ENET1_RGMII_TX_CTL 0x57e > + >; > + }; > + > pinctrl_uart1: uart1grp { > fsl,pins = < > MX93_PAD_UART1_RXD__LPUART1_RX 0x31e
Hi, On Fri, 2023-01-13 at 11:33 +0800, Clark Wang wrote: > This patchset add imx93 support for dwmac-imx glue driver. > There are some changes of GPR implement. > And add fec and eqos nodes for imx93 dts. > > Clark Wang (7): > net: stmmac: add imx93 platform support > dt-bindings: add mx93 description > dt-bindings: net: fec: add mx93 description > arm64: dts: imx93: add eqos support > arm64: dts: imx93: add FEC support > arm64: dts: imx93-11x11-evk: enable eqos > arm64: dts: imx93-11x11-evk: enable fec function > > .../devicetree/bindings/net/fsl,fec.yaml | 1 + > .../bindings/net/nxp,dwmac-imx.yaml | 4 +- > .../boot/dts/freescale/imx93-11x11-evk.dts | 78 +++++++++++++++++++ > arch/arm64/boot/dts/freescale/imx93.dtsi | 48 ++++++++++++ > .../net/ethernet/stmicro/stmmac/dwmac-imx.c | 55 +++++++++++-- > 5 files changed, 180 insertions(+), 6 deletions(-) It's not clear to me if the whole series should go via netdev. I think/fear such option could cause later conflicts for Linus. Does it make sense to split this in 2 chunks, and have only the first 3 patches merged via netdev? Thanks! Paolo
Hello: This series was applied to netdev/net-next.git (master) by David S. Miller <davem@davemloft.net>: On Fri, 13 Jan 2023 11:33:40 +0800 you wrote: > Hi, > > This patchset add imx93 support for dwmac-imx glue driver. > There are some changes of GPR implement. > And add fec and eqos nodes for imx93 dts. > > Clark Wang (7): > net: stmmac: add imx93 platform support > dt-bindings: add mx93 description > dt-bindings: net: fec: add mx93 description > arm64: dts: imx93: add eqos support > arm64: dts: imx93: add FEC support > arm64: dts: imx93-11x11-evk: enable eqos > arm64: dts: imx93-11x11-evk: enable fec function > > [...] Here is the summary with links: - [V2,1/7] net: stmmac: add imx93 platform support https://git.kernel.org/netdev/net-next/c/e5bf35ca4547 - [V2,2/7] dt-bindings: add mx93 description https://git.kernel.org/netdev/net-next/c/b2274ffe90be - [V2,3/7] dt-bindings: net: fec: add mx93 description https://git.kernel.org/netdev/net-next/c/f743e7664dca - [V2,4/7] arm64: dts: imx93: add eqos support https://git.kernel.org/netdev/net-next/c/1f4263ea6a4b - [V2,5/7] arm64: dts: imx93: add FEC support https://git.kernel.org/netdev/net-next/c/eaaf47108540 - [V2,6/7] arm64: dts: imx93-11x11-evk: enable eqos https://git.kernel.org/netdev/net-next/c/1b110dd678d9 - [V2,7/7] arm64: dts: imx93-11x11-evk: enable fec function https://git.kernel.org/netdev/net-next/c/c897dc7f3a8d You are awesome, thank you!
On Tue, Jan 17, 2023 at 09:27:56AM +0100, Paolo Abeni wrote: > Hi, > > On Fri, 2023-01-13 at 11:33 +0800, Clark Wang wrote: > > This patchset add imx93 support for dwmac-imx glue driver. > > There are some changes of GPR implement. > > And add fec and eqos nodes for imx93 dts. > > > > Clark Wang (7): > > net: stmmac: add imx93 platform support > > dt-bindings: add mx93 description > > dt-bindings: net: fec: add mx93 description > > arm64: dts: imx93: add eqos support > > arm64: dts: imx93: add FEC support > > arm64: dts: imx93-11x11-evk: enable eqos > > arm64: dts: imx93-11x11-evk: enable fec function > > > > .../devicetree/bindings/net/fsl,fec.yaml | 1 + > > .../bindings/net/nxp,dwmac-imx.yaml | 4 +- > > .../boot/dts/freescale/imx93-11x11-evk.dts | 78 +++++++++++++++++++ > > arch/arm64/boot/dts/freescale/imx93.dtsi | 48 ++++++++++++ > > .../net/ethernet/stmicro/stmmac/dwmac-imx.c | 55 +++++++++++-- > > 5 files changed, 180 insertions(+), 6 deletions(-) > > It's not clear to me if the whole series should go via netdev. I > think/fear such option could cause later conflicts for Linus. Does it > make sense to split this in 2 chunks, and have only the first 3 patches > merged via netdev? I share the same concern here. David, Could you *not* apply DTS patches in the future? People often include driver changes and corresponding DTS ones in a single series to ease cross reviewing and testing. If picking selected patches from a series could a problem for your applying robot, we should probably start asking people to split change-sets. Shawn