From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH v2] phy state machine: failsafe leave invalid RUNNING state Date: Sun, 08 Jan 2017 18:16:00 -0500 (EST) Message-ID: <20170108.181600.1132579382984085546.davem@davemloft.net> References: <1483542298-9747-1-git-send-email-zefir.kurtisi@neratec.com> <1483701288-14019-1-git-send-email-zefir.kurtisi@neratec.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, f.fainelli@gmail.com, andrew@lunn.ch To: zefir.kurtisi@neratec.com Return-path: Received: from shards.monkeyblade.net ([184.105.139.130]:44238 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753094AbdAHXQC (ORCPT ); Sun, 8 Jan 2017 18:16:02 -0500 In-Reply-To: <1483701288-14019-1-git-send-email-zefir.kurtisi@neratec.com> Sender: netdev-owner@vger.kernel.org List-ID: From: Zefir Kurtisi Date: Fri, 6 Jan 2017 12:14:48 +0100 > While in RUNNING state, phy_state_machine() checks for link changes by > comparing phydev->link before and after calling phy_read_status(). > This works as long as it is guaranteed that phydev->link is never > changed outside the phy_state_machine(). > > If in some setups this happens, it causes the state machine to miss > a link loss and remain RUNNING despite phydev->link being 0. > > This has been observed running a dsa setup with a process continuously > polling the link states over ethtool each second (SNMPD RFC-1213 > agent). Disconnecting the link on a phy followed by a ETHTOOL_GSET > causes dsa_slave_get_settings() / dsa_slave_get_link_ksettings() to > call phy_read_status() and with that modify the link status - and > with that bricking the phy state machine. > > This patch adds a fail-safe check while in RUNNING, which causes to > move to CHANGELINK when the link is gone and we are still RUNNING. > > Signed-off-by: Zefir Kurtisi > --- > Changes to v1: > * fix kbuild test robot error: use phydev_err instead of dev_warn > (adapt to changed struct phy_device after 4.4.21) Florian and Andrew, please provide some feedback on this. Thank you.