On Fri, Mar 18, 2022 at 9:56 PM AngeloGioacchino Del Regno <angelogioacchino.delregno@xxxxxxxxxxxxx> wrote: > > Il 18/03/22 14:51, Krzysztof Kozlowski ha scritto: > > On 18/03/2022 12:12, AngeloGioacchino Del Regno wrote: > >> Il 18/03/22 10:54, Jianjun Wang ha scritto: > >>> Add YAML schema documentation for PCIe PHY on MediaTek chipsets. > >>> > >>> Signed-off-by: Jianjun Wang <jianjun.wang@xxxxxxxxxxxx> > >>> --- > >>> .../bindings/phy/mediatek,pcie-phy.yaml | 75 +++++++++++++++++++ > >>> 1 file changed, 75 insertions(+) > >>> create mode 100644 Documentation/devicetree/bindings/phy/mediatek,pcie-phy.yaml > >>> > >>> diff --git a/Documentation/devicetree/bindings/phy/mediatek,pcie-phy.yaml b/Documentation/devicetree/bindings/phy/mediatek,pcie-phy.yaml > >>> new file mode 100644 > >>> index 000000000000..868bf976568b > >>> --- /dev/null > >>> +++ b/Documentation/devicetree/bindings/phy/mediatek,pcie-phy.yaml > >>> @@ -0,0 +1,75 @@ > >>> +# SPDX-License-Identifier: (GPL-2.0 OR BSD-2-Clause) > >>> +%YAML 1.2 > >>> +--- > >>> +$id: http://devicetree.org/schemas/phy/mediatek,pcie-phy.yaml# > >>> +$schema: http://devicetree.org/meta-schemas/core.yaml# > >>> + > >>> +title: MediaTek PCIe PHY > >>> + > >>> +maintainers: > >>> + - Jianjun Wang <jianjun.wang@xxxxxxxxxxxx> > >>> + > >>> +description: | > >>> + The PCIe PHY supports physical layer functionality for PCIe Gen3 port. > >>> + > >>> +properties: > >>> + compatible: > >>> + const: mediatek,mt8195-pcie-phy > >> > >> Since I don't expect this driver to be only for MT8195, but to be extended to > >> support some more future MediaTek SoCs and, depending on the number of differences > >> in the possible future Gen4 PHYs, even different gen's, I propose to add a generic > >> compatible as const. > >> > >> So you'll have something like: > >> > >> - enum: > >> - mediatek,mt8195-pcie-phy > >> - const: mediatek,pcie-gen3-phy > > > > I am not sure if this is a good idea. How sure are you that there will > > be no different PCIe Gen3 PHY not compatible with this one? > > > > > > Thanks for pointing that out, I have underestimated this option. > > Perhaps Jianjun may be more informed about whether my proposal is valid or not. Just FYI, for Allwinner and I believe Rockchip as well, the compatible strings always list the first SoC the hardware block was seen on known at the time of driver/binding submission. No generic compatible strings are ever used. Not sure if that's the general rule or not. ChenYu
