Hi Lorenzo,
On Mon, 19 Nov 2018 11:40:36 +0000 <lorenzo.pieral...@arm.com> wrote:
> On Tue, Oct 16, 2018 at 02:27:20PM +0900, Kunihiko Hayashi wrote:
> > Add DT bindings for PCIe controller implemented in UniPhier SoCs when
> > configured in Root Complex (host) mode. This controller is based on
> > the DesignWare PCIe core.
> >
> > Signed-off-by: Kunihiko Hayashi <hayashi.kunih...@socionext.com>
> > Reviewed-by: Rob Herring <r...@kernel.org>
> > ---
> > .../devicetree/bindings/pci/uniphier-pcie.txt | 81
> > ++++++++++++++++++++++
> > 1 file changed, 81 insertions(+)
> > create mode 100644 Documentation/devicetree/bindings/pci/uniphier-pcie.txt
> >
> > diff --git a/Documentation/devicetree/bindings/pci/uniphier-pcie.txt
> > b/Documentation/devicetree/bindings/pci/uniphier-pcie.txt
> > new file mode 100644
> > index 0000000..46a2754
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/pci/uniphier-pcie.txt
> > @@ -0,0 +1,81 @@
> > +Socionext UniPhier PCIe host controller bindings
> > +
> > +This describes the devicetree bindings for PCIe host controller implemented
> > +on Socionext UniPhier SoCs.
> > +
> > +UniPhier PCIe host controller is based on the Synopsys DesignWare PCI core.
> > +It shares common functions with the PCIe DesignWare core driver and
> > inherits
> > +common properties defined in
> > +Documentation/devicetree/bindings/pci/designware-pcie.txt.
> > +
> > +Required properties:
> > +- compatible: Should be "socionext,uniphier-pcie".
> > +- reg: Specifies offset and length of the register set for the device.
> > + According to the reg-names, appropriate register sets are required.
> > +- reg-names: Must include the following entries:
> > + "dbi" - controller configuration registers
> > + "link" - SoC-specific glue layer registers
> > + "config" - PCIe configuration space
> > +- clocks: A phandle to the clock gate for PCIe glue layer including
> > + the host controller.
> > +- resets: A phandle to the reset line for PCIe glue layer including
> > + the host controller.
> > +- interrupts: A list of interrupt specifiers. According to the
> > + interrupt-names, appropriate interrupts are required.
> > +- interrupt-names: Must include the following entries:
> > + "dma" - DMA interrupt
> > + "msi" - MSI interrupt
> > +
> > +Optional properties:
> > +- phys: A phandle to generic PCIe PHY. According to the phy-names,
> > appropriate
> > + phys are required.
> > +- phy-names: Must be "pcie-phy".
> > +
> > +Required sub-node:
> > +- legacy-interrupt-controller: Specifies interrupt controller for legacy
> > PCI
> > + interrupts.
> > +
> > +Required properties for legacy-interrupt-controller:
> > +- interrupt-controller: identifies the node as an interrupt controller.
> > +- #interrupt-cells: specifies the number of cells needed to encode an
> > + interrupt source. The value must be 1.
> > +- interrupt-parent: Phandle to the parent interrupt controller.
> > +- interrupts: An interrupt specifier for legacy interrupt.
> > +
> > +Example:
> > +
> > + pcie: pcie@66000000 {
> > + compatible = "socionext,uniphier-pcie", "snps,dw-pcie";
> > + status = "disabled";
> > + reg-names = "dbi", "link", "config";
> > + reg = <0x66000000 0x1000>, <0x66010000 0x10000>,
> > + <0x2fff0000 0x10000>;
> > + #address-cells = <3>;
> > + #size-cells = <2>;
> > + clocks = <&sys_clk 24>;
> > + resets = <&sys_rst 24>;
> > + num-lanes = <1>;
> > + num-viewport = <1>;
> > + bus-range = <0x0 0xff>;
> > + device_type = "pci";
> > + ranges =
> > + /* downstream I/O */
> > + <0x81000000 0 0x00000000 0x2ffe0000 0 0x00010000
> > + /* non-prefetchable memory */
> > + 0x82000000 0 0x00000000 0x20000000 0 0x0ffe0000>;
> > + #interrupt-cells = <1>;
> > + interrupt-names = "dma", "msi";
> > + interrupts = <0 224 4>, <0 225 4>;
> > + interrupt-map-mask = <0 0 0 7>;
> > + interrupt-map = <0 0 0 1 &pcie_intc 1>, /* INTA */
> > + <0 0 0 2 &pcie_intc 2>, /* INTB */
> > + <0 0 0 3 &pcie_intc 3>, /* INTC */
> > + <0 0 0 4 &pcie_intc 4>; /* INTD */
>
> This is not correct and we must fix this concept for all DWC based host
> bridges bindings. We are mapping INTX to inputs [0,1,2,3] of the host
> bridge interrupt controllers and the binding should reflect that.
>
> This should be the correct mapping:
>
> interrupt-map = <0 0 0 1 &pcie_intc 0>, /* INTA */
> <0 0 0 2 &pcie_intc 1>, /* INTB */
> <0 0 0 3 &pcie_intc 2>, /* INTC */
> <0 0 0 4 &pcie_intc 3>; /* INTD */
I see. Although this numbering has been affected by pci_irqd_intx_xlate(),
I understand it's wrong for DWC based host.
I'll fix the numbering next.
Thank you,
---
Best Regards,
Kunihiko Hayashi
