Re: [PATCH] of: Custom printk format specifier for device node

[Rob Herring]

On Tue, Jan 20, 2015 at 10:05 AM, MÃns RullgÃrd <mans@xxxxxxxxx> wrote:
> Pantelis Antoniou <pantelis.antoniou@xxxxxxxxxxxx> writes:
>
>> Hi Geert,
>>
>>> On Jan 20, 2015, at 17:24 , Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
>>>
>>> On Tue, Jan 20, 2015 at 3:47 PM, Rob Herring <robherring2@xxxxxxxxx> wrote:
>>>>> +       Examples:
>>>>> +
>>>>> +       %pO     /foo/bar@0              - Node full name
>>>>> +       %pO0    /foo/bar@0              - Same as above
>>>>> +       %pO1    /foo/bar@0[10]          - Node full name + phandle
>>>>> +       %pO2    /foo/bar@0[10:DdPB]     - Node full name + phandle + node flags
>>>>> +                                        D - dynamic
>>>>> +                                        d - detached
>>>>> +                                        P - Populated
>>>>> +                                        B - Populated bus
>>>>
>>>> We should think about what else we want to print for a node. Perhaps
>>>> 'On' for name, 'Oc' for compatible, etc.
>>>
>>> I was just going to say "The least verbose variant is name, not full_nameâ.
>>>
>>
>> Unfortunately in the context of device tree nodes âname' is usually
>> not what you want to print to identify the node in question. ânameâ is
>> usually not unique.
>
> Name and address without the full path is usually a good compromise
> between uniqueness (it is usually unique for memory-mapped things) and
> verbosity.

How much of the address is in the name depends on how the address
translation is done. I don't think we really need to do full address
translations here.

%pOn     /foo/bar@0              - Node full name
%pOn0   bar@0              - Node name and unit address
%pOn1    /foo/bar@0[10]          - Node full name + phandle
%pOn2    /foo/bar@0[10:DdPB]     - Node full name + phandle + node flags
%pOc      vendor,foo-bar          - Most significant compatible string

We could do phandle and/or node flags as separate specifiers such as
%pOf for flags.

I'm not proposing implementing all these now, but just want to make
sure we have a structure to do so later.

Rob
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/