Re: [PATCH net-next 0/6] net/core, bonding: dev_uc_sync fixes, bonding update

On 05/31/2013 11:56 AM, Jay Vosburgh wrote:

Shawn Bohrer [off-list ref] wrote:

quoted

On Fri, May 31, 2013 at 01:31:55AM -0700, David Miller wrote:

quoted

From: Jay Vosburgh <redacted>
Date: Thu, 30 May 2013 17:55:38 -0700

quoted

	This patch set includes 6 patches: four fixes to the dev_mc_sync /
dev_mc_unsync system; and two patches to bonding, one to utilize the sync
/ unsync functions, and another minor fix related to MAC address handling.

These look like fixes that should go into net, why target net-next?

In my oppinion 0-4 should go into net since they fix the bug I
reported in:

http://comments.gmane.org/gmane.linux.network/270477

I've tested patches 0-4 of this series so feel free to add my tested
by to those:

Tested-by: Shawn Bohrer <redacted>

From just a casual observation of patch 5-6 they do not appear to be
bug fixes which is why this was probably marked net-next.

	They're against net-next because I was working to convert
bonding to dev_sync/unsync against net-next and neglected to rebase then
before I posted.  The bonding patches (5 and 6) do fix a couple of bugs
related to MAC address handling on s390 (the lack of additional unicast
address propagation to the slaves makes qeth unhappy in some cases), so
arguably they could go either way, but I'm ok with those in net-next if
it's an issue.

	I do agree that 1-4 should go into net, once Vlad gives them a
look.

	-J

I've reviewed the patches and ran a quick test.  They look good and fix
obvious problems.  Thanks to Jay for finding and fixing them.

Reviewed-by: Vlad Yasevich <redacted>

-vlad

---
	-Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com