Re: [PATCH net-next] ipv6: recreate ipv6 link-local addresses when increasing MTU over IPV6_MIN_MTU

On 10/26/2015 01:45 PM, Hannes Frederic Sowa wrote:

Hi,

On Mon, Oct 26, 2015, at 20:16, Jay Vosburgh wrote:

quoted

Hannes Frederic Sowa [off-list ref] wrote:

quoted

Hello Alex,

On Mon, Oct 26, 2015, at 16:52, Alexander Duyck wrote:

quoted

Seems like this code isn't quite correct.  You are calling ipv6_add_dev
for slave devices, and if I understand things correctly I don't believe
that was happening before and may be an unintended side effect.

Ah, btw., autoconf and ipv6 operation on IFF_SLAVE devices is actually
desired nowadays and don't think we can change this. See also:
<https://patchwork.ozlabs.org/patch/531196/>

	IPv6 addrconf on IFF_SLAVE devices was disabled for bonding
slaves in commit c2edacf80e15 because it caused issues with snooping
switches.

	This is also referenced in

https://bugzilla.redhat.com/show_bug.cgi?id=236750

	Won't re-enabling autoconf on IFF_SLAVE devices cause that issue
to return?

Both patches don't enable autoconf on IFF_SLAVE devices. Sorry for being
imprecise. The referred patch was changing the behavior to whether the
device had a master device.

Yes, the IFF_SLAVE comment on my part was an error in interpretation of 
the code.

@Alex, I will take your patch and submit it with the necessary guards to
not enable ipv6 again if we forcefully disable ipv6 and later on shrink
and increase the MTU again. I will do so in your name. Thanks again for
the patch!

No problem.  If you want to you can take over authorship of the patch 
and just leave my signed-off-by on there.  I'm good either way.

- Alex