On Thursday, December 18, 2014 at 07:53:56 AM, Joe Hershberger wrote:
Hi Marek,
On Wed, Dec 17, 2014 at 10:00 AM, Marek Vasut marex@denx.de wrote:
On Wednesday, December 17, 2014 at 01:33:57 PM, René Griessl wrote:
Is there a reason you can't implement write_hwaddr() so that it will behave like other NICs (u-boot is the highest priority source of the MAC address).
The AX88179 uses an external flash to store the MAC address. So for my application I do not want the user to be able to change it, since it is a unique ID inside the server cluster. I could add it as compiler option. Maybe it is even better to implement my behavior as option, since it differs from standard. I think it will be finished mid of january, so if it is OK for you in a separate patch.
Joe, can you just pick the patch as is and wait for subsequent patch to
improve
the driver? My understanding is that the driver is operating correctly and pulling the MAC from an integrated EEPROM, which is fine. If you want to
allow
overriding this MAC using the $ethaddr, this is only an improvement and
can be
added later. This would improve our hardware support and avoid blocking
the
driver further.
What do you think please ?
I agree... You may have noticed I already Acked it and agreed that the improvement may come later.
All righty, thanks!
Would you pick it and submit for current release or do you want me to do it?
Best regards, Marek Vasut