On Tue, Jul 19, 2016 at 08:57:53PM +1000, Benjamin Herrenschmidt wrote:
>On Tue, 2016-07-19 at 10:50 +0000, David Laight wrote:
>> > + if (!is_valid_ether_addr(mac)) {
>> > + mac[5] = (m >> 8) & 0xff;
>> > + mac[4] = m & 0xff;
>> > + mac[3] = (l >> 24) & 0xff;
>> > + mac[2] = (l >> 16) & 0xff;
>> > + mac[1] = (l >> 8) & 0xff;
>> > + mac[0] = l & 0xff;
>> > + }
>> ...
>>
>> That is horrid, not all byte reversed addresses will be invalid.
>
>Right, that's just a hack for a broken vendor uboot we had here, Gavin,
>drop that part of the patch please.
>
Sure, I'll drop it in v4 or a followup patch. v3 is being reviewed this
moment.
Thanks,
Gavin
