[PATCH v4 2/4] PCI: st: Add Device Tree bindings for sti pcie

[Gabriel Fernandez]

Hi Rob,

Thanks for the review.

Best regards

Gabriel

On 28 August 2015 at 02:06, Rob Herring <robherring2 at gmail.com> wrote:
> On Thu, Aug 27, 2015 at 7:34 AM, Gabriel Fernandez
> <gabriel.fernandez at linaro.org> wrote:
>> sti pcie is built around a Synopsis Designware PCIe IP.
>>
>> Signed-off-by: Fabrice Gasnier <fabrice.gasnier at st.com>
>> Signed-off-by: Gabriel Fernandez <gabriel.fernandez at linaro.org>
>> ---
>>  Documentation/devicetree/bindings/pci/st-pcie.txt | 53 +++++++++++++++++++++++
>>  1 file changed, 53 insertions(+)
>>  create mode 100644 Documentation/devicetree/bindings/pci/st-pcie.txt
>>
>> diff --git a/Documentation/devicetree/bindings/pci/st-pcie.txt b/Documentation/devicetree/bindings/pci/st-pcie.txt
>> new file mode 100644
>> index 0000000..25fcab3
>> --- /dev/null
>> +++ b/Documentation/devicetree/bindings/pci/st-pcie.txt
>> @@ -0,0 +1,53 @@
>> +STMicroelectronics STi PCIe controller
>> +
>> +This PCIe host controller is based on the Synopsis Designware PCIe IP
>> +and thus inherits all the common properties defined in designware-pcie.txt.
>> +
>> +Required properties:
>> + - compatible: "st,stih407-pcie"
>
> What about "snps,dw-pcie" as well?
>
You are right.

>> + - reg: base address and length of the pcie controller, mem-window address
>> +   and length available to the controller.
>
> What is mem-window? Seems rather large and perhaps should be under ranges.
>
No the purpose is to specify the physical memory available to the controller.
reg property is more appropriate.

>> + - interrupts: A list of interrupt outputs of the controller. Must contain an
>> +   entry for each entry in the interrupt-names property.
>
> Define how many interrupts.
>
ok i will fix it.

>> + - interrupt-names: Should be "msi". STi interrupt that is asserted when an
>> +   MSI is received.
>
> Kind of pointless with a single interrupt.
>
ok

>> + - st,syscfg : should be a phandle of the syscfg node. Also contains syscfg
>> +   offset for IP configuration.
>> + - resets, reset-names: the power-down and soft-reset lines of PCIe IP.
>> +   Associated names must be "powerdown" and "softreset".
>> + - phys, phy-names: the phandle for the PHY device.
>> +   Associated name must be "pcie"
>
> What does this mean?
>
i will reformulate this paragraph.

>> +
>> +Optional properties:
>> + - reset-gpio: a GPIO spec to define which pin is connected to the bus reset.
>> +
>> +Example:
>> +
>> +pcie0: pcie at 9b00000 {
>> +       compatible = "st,pcie", "snps,dw-pcie";
>> +       device_type = "pci";
>> +       reg = <0x09b00000 0x4000>,      /* dbi cntrl registers */
>> +             <0x2fff0000 0x00010000>,  /* configuration space */
>> +             <0x40000000 0x80000000>;  /* lmi mem window */
>> +       reg-names = "dbi", "config", "mem-window";
>> +       st,syscfg = <&syscfg_core 0xd8 0xe0>;
>> +       #address-cells = <3>;
>> +       #size-cells = <2>;
>> +       ranges = <0x82000000 0 0x20000000 0x20000000 0 0x0FFF0000>; /* non-prefetchable memory */
>
> No i/o support?
>
Exactly there is no i/o support.

>> +       num-lanes = <1>;
>> +       interrupts = <GIC_SPI 164 IRQ_TYPE_LEVEL_HIGH>;
>> +       interrupt-names = "msi";
>> +       #interrupt-cells = <1>;
>> +       interrupt-map-mask = <0 0 0 7>;
>> +       interrupt-map = <0 0 0 1 &intc GIC_SPI 160 IRQ_TYPE_LEVEL_HIGH>, /* INT A */
>> +                       <0 0 0 2 &intc GIC_SPI 161 IRQ_TYPE_LEVEL_HIGH>, /* INT B */
>> +                       <0 0 0 3 &intc GIC_SPI 162 IRQ_TYPE_LEVEL_HIGH>, /* INT C */
>> +                       <0 0 0 4 &intc GIC_SPI 163 IRQ_TYPE_LEVEL_HIGH>; /* INT D */
>> +
>> +       resets = <&powerdown STIH407_PCIE0_POWERDOWN>,
>> +                <&softreset STIH407_PCIE0_SOFTRESET>;
>> +       reset-names = "powerdown",
>> +                     "softreset";
>> +       phys = <&phy_port0 PHY_TYPE_PCIE>;
>> +       phy-names = "pcie";
>> +};
>> --
>> 1.9.1
>>