On 02/24/2015 05:29 PM, Rob Herring wrote:
> On Tue, Feb 24, 2015 at 8:58 AM, Michal Simek <michal.simek at xilinx.com> wrote:
>> On 02/24/2015 03:42 PM, Rob Herring wrote:
>>> On Tue, Feb 24, 2015 at 1:56 AM, Michal Simek <michal.simek at xilinx.com> wrote:
>>>> Initial version of device tree for Xilinx ZynqMP SoC.
>>>>
>>>> Signed-off-by: Michal Simek <michal.simek at xilinx.com>
>>>> Acked-by: Sören Brinkmann <soren.brinkmann at xilinx.com>
>>>> ---
>>>
>>> [...]
>>>
>>>> + gic: interrupt-controller at f9010000 {
>>>> + compatible = "arm,cortex-a15-gic", "arm,cortex-a9-gic";
>>>
>>> gic-400, right?
>>
>> yep
>>
>>>
>>>> + #interrupt-cells = <3>;
>>>> + reg = <0x0 0xf9010000 0x10000>,
>>>> + <0x0 0xf9020000 0x20000>,
>>>> + <0x0 0xf9040000 0x20000>,
>>>> + <0x0 0xf9060000 0x20000>;
>>>
>>> These addresses are wrong if you are doing address swizzling to do 64K
>>> offsets. We don't really have an answer yet as to what is the right
>>> way. See the XGene GIC discussion[1].
>>
>> Is this better for GICC?
>> <0x0 0xf902f000 0x2000>
>
> Yes, and the VCPU interface needs this as well. As far as sizes, we're
> still discussing that.
ok. Will test and fix. What about the rest of DT description?
Thanks,
Michal