Comment # 1 on bug 908257 from Jiri Bohac

Is this with the ax88179_set_mac_addr patch applied?

It looks like bond_enslave() returns something >0, which propagates to the
return value of the sysfs write, misleading echo to think that only a few bytes
have been written and writing the rest of the string again.

Either it's the problem with ax88179_set_mac_addr() or a similar problem in one
of the other callbacks of ax88179.