On Wed, Oct 3, 2012 at 5:24 PM, Florian Fainelli <florian@xxxxxxxxxxx> wrote:
> On Wednesday 03 October 2012 17:21:37 Manuel Lauss wrote:
>> On Wed, Oct 3, 2012 at 5:03 PM, Florian Fainelli <florian@xxxxxxxxxxx> wrote:
>> > This also greatly simplifies the power_{on,off} callbacks and make them
>> > work on platform device id instead of checking the OHCI controller base
>> > address like what was done in ohci-au1xxx.c.
>>
>> That was by design -- the base address is far more reliable in identifying
> the
>> correct controller instance than the platform device id. There are systems
>> in the field which don't use the alchemy/common/platform.c file at all.
>
> Fair enough, but the way it was done previously was very error-prone if the
> base address changed for any reason in the platform code, and you did not
> notice it had to be changed in the OHCI driver too, then it simply did not
Since the Alchemy line is dead this point is moot.
> work. By systems in the field you mean out of tree users? If so, I'd say that
> it's up to you to get them maintained or merged.
I'm not against the patch at all, quite the contrary.
Manuel
