netdev - Re: [PATCH net-next v5] ethtool: linkstate: add a statistic for PHY down events

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 4 Nov 2022 15:48:09 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     "Russell King (Oracle)" <linux@...linux.org.uk>
Cc:     davem@...emloft.net, netdev@...r.kernel.org, edumazet@...gle.com,
        pabeni@...hat.com, Florian Fainelli <f.fainelli@...il.com>,
        Michael Chan <michael.chan@...adcom.com>,
        Andrew Lunn <andrew@...n.ch>, corbet@....net,
        hkallweit1@...il.com, huangguangbin2@...wei.com,
        chenhao288@...ilicon.com, moshet@...dia.com,
        linux@...pel-privat.de, linux-doc@...r.kernel.org
Subject: Re: [PATCH net-next v5] ethtool: linkstate: add a statistic for PHY
 down events

On Fri, 4 Nov 2022 20:18:40 +0000 Russell King (Oracle) wrote:
> I guess we'll need phylink to support this as well, so phylink using
> drivers can provide this statistic not only for phylib based PHYs, but
> also for direct SFP connections as well.
> 
> Thinking about the complexities of copper SFPs that may contain a PHY,
> it seems to me that the sensible implementation would be for phylink
> to keep the counter and not use the phylib counter (as that PHY may
> be re-plugged and thus the count can reset back to zero) which I
> suspect userspace would not be prepared for.

Makes sense, having the counter go back on a netdev could be highly
confusing for local detection.

How would you like to proceed? I can try to take a stab at a phylink
implementation but a stab it will be.