Re: [RFC 1/1] net: move IFF_LIVE_ADDR_CHANGE to public flag

[Jakub Kicinski <kuba@kernel.org>]

On Thu,  4 Aug 2022 10:43:07 -0700 James Prestwood wrote:
> By exposing IFF_LIVE_ADDR_CHANGE to userspace it at least gives an
> indication that we can successfully randomize the address and
> connect. In the worst case address randomization can be avoided
> ahead of time. A secondary win is also time, since userspace can
> avoid a power down unless its required which saves some time.

It's not a generic thing tho, it's most of an implicit argument 
to eth_mac_addr(). Not all netdevs are Ethernet.

The semantics in wireless are also a little stretched because normally
if the flag is not set the netdev will _refuse_ (-EBUSY) to change the
address while running, not do some crazy fw reset.

Perhaps we should wait for Johannes to return form vacation but my
immediate reaction would be to add a knob (in wireless?) that controls
whether the reset dance is allowed.