Re: [PATCH] 8139cp: allow to set mac address on running device

[Ivan Vecera <ivecera@redhat.com>]

Jiri Pirko wrote:
> Thu, Mar 12, 2009 at 06:11:21PM CET, mschmidt@redhat.com wrote:
>> On Thu, 12 Mar 2009 17:27:31 +0100
>> Jiri Pirko <jpirko@redhat.com> wrote:
>>
>>> +	cpw32_f(MAC0 + 0, le32_to_cpu (*(__le32 *) (dev->dev_addr +
>>> 0)));
>>> +	cpw32_f(MAC0 + 4, le32_to_cpu (*(__le32 *) (dev->dev_addr +
>>> 4)));
>> You're writing to the card, so using *_to_cpu looks suspicious.
> Well, I'm using the same approach as it is already done in function
> cp_init_hw(). Quote:
> 
>  /* Restore our idea of the MAC address. */
>  cpw32_f (MAC0 + 0, le32_to_cpu (*(__le32 *) (dev->dev_addr + 0)));
>  cpw32_f (MAC0 + 4, le32_to_cpu (*(__le32 *) (dev->dev_addr + 4)));
> 
Yes, that's right but I would use more cleaner approach:
===
u32 low, high;
low  = addr[0] | (addr[1] << 8) | (addr[2] << 16) | (addr[3] << 24);
high = addr[4] | (addr[5] << 8);
cpw32_f(MAC0 + 0, low);
cpw32_f(MAC0 + 4, high);
===

Ivan