From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH net] Revert "net: ethernet: bcmgenet: use phydev from struct net_device" Date: Sun, 25 Sep 2016 05:35:06 -0400 (EDT) Message-ID: <20160925.053506.2140883650540997409.davem@davemloft.net> References: <1474747110-6496-1-git-send-email-f.fainelli@gmail.com> <20160924.201054.1736819083455070537.davem@davemloft.net> <9fa471d4-a179-519b-005f-5b52f31c5903@gmail.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, tremyfr@gmail.com, jaedon.shin@gmail.com To: f.fainelli@gmail.com Return-path: Received: from shards.monkeyblade.net ([184.105.139.130]:46354 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753897AbcIYJfK (ORCPT ); Sun, 25 Sep 2016 05:35:10 -0400 In-Reply-To: <9fa471d4-a179-519b-005f-5b52f31c5903@gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: From: Florian Fainelli Date: Sat, 24 Sep 2016 19:51:49 -0700 > > > On 09/24/2016 05:10 PM, David Miller wrote: >> From: Florian Fainelli >> Date: Sat, 24 Sep 2016 12:58:30 -0700 >> >>> There is already a commit: >>> >>> Revert "net: ethernet: bcmgenet: use phy_ethtool_{get|set}_link_ksettings" >>> >>> which should make this apply cleanly to "net" now. >> >> But look at net-next, it got re-added there. >> >> This is going to be a bit of a merge hassle, and this is why I pushed >> back on the other attempt to revert this thing. > > OK, so how about this: > > - this patch applies to net which should be okay for now > - to avoid future conflicts when you merge net into net-next, I submit a > revert of "net: ethernet: bcmgenet: use > phy_ethtool_{get|set}_link_ksettings" against net-next > > Does that work for you? The ksettings patch will still be in net-next and thus the driver will be broken in a different way than the problem being fixed by the revert. If you revert this phydev patch you must also revert the ksettings patch. The ksettings patch depends upon it. This is what I've been trying to explain all along. What I really hoped someone would try, would be to make this phydev change work properly instead of just straight up reverting it. In that way all of these merge hassles would be eliminated entirely.