[+cc devicetree list; I'm devicetree-illiterate]

On Tue, Dec 10, 2013 at 12:53 PM, Rajat Jain <[email protected]> wrote:
> Hello,
>
> I am trying to understand and create device tree for a new powerpc platform. 
> (Please feel free to point out any problems with my device tree below.) My 
> questions:
>
> 1) For a PCI-PCI bridge that I'm representing like this:
>
> pcie@2,3 {
>         vendor-id = <0x10b5>;
>       device-id = <0x8618>;
>       class-code = <0x060400>;
>       #size-cells = <2>;
>       #address-cells = <3>;
>       device_type = "pci";
>       bus-range = <0x20 0x2f>;
>       reg = <0x00021800 0x00000000 0x000000000        /* Cfg:2.3.0: */
>              0x00000000 0x00000000>;
>       ranges = <0x02000000 0x00000000 0xA0000000      /* Mem behind bridge: 
> 0xA0000000 - 0xA03FFFFF */
>                 0x02000000 0x00000000 0xA0000000
>                 0x00000000 0x04000000>;
> };
>
> Is the above is only to _notify_ the kernel that this is how the firmware 
> shall be setting the PCI bridge (Thus firmware has obligation of setting it 
> like this before passing control to the kernel), and hence the kernel will 
> merely take note of it and not really configure the bridge registers?
>
> Or is the above telling the kernel that this is how the kernel should _setup_ 
> the device (thus firmware does not set up the device, but rather the kernel 
> use this info to setup the device)?
>
> 2) For an PCI endpoint that I am representing like this:
>
>
> xyz@20,0 {
>         vendor-id = ...
>       device-id = ...
>       class-code = <0x0880>;
>       compatible = ...
>         ...
>       reg = <0x00200000 0x00000000 0x00000000 /* Cfg:20.0.0: */
>              0x00000000 0x00000000
>              0x02200010 0x00000000 0xA0000000 /* Mem:20.0.0:(Bar0) 
> 64MB@0xA0000000 */
>              0x00000000 0x04000000>;
>         ...
> };
>
> Does the above tell the kernel to _program_ the BAR0 register using the above 
> address? Or does it cause the kernel to assume that the BAR0 has been 
> programmed with this address?
>
> 3) Lastly, what is the difference in behavior when I select PCI_PROBE_NORMAL 
> over PCI_PROBE_DEVTREE?
>
> Thanks,
>
> Rajat
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pci" in
> the body of a message to [email protected]
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to