netdev - Re: [RFC PATCH net-next 2/8] ethtool: Add ability to reset transceiver modules

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <YRLMSt2JAqXr8zko@lunn.ch>
Date:   Tue, 10 Aug 2021 20:58:18 +0200
From:   Andrew Lunn <andrew@...n.ch>
To:     Ido Schimmel <idosch@...sch.org>
Cc:     Jakub Kicinski <kuba@...nel.org>, netdev@...r.kernel.org,
        davem@...emloft.net, mkubecek@...e.cz, pali@...nel.org,
        vadimp@...dia.com, mlxsw@...dia.com,
        Ido Schimmel <idosch@...dia.com>
Subject: Re: [RFC PATCH net-next 2/8] ethtool: Add ability to reset
 transceiver modules

> I thought there are already RTNL-lock-less ethtool ops, but maybe I
> imagined it. Given that we also want to support firmware update on
> modules and that user space might want to update several modules
> simultaneously, do you have a suggestion on how to handle it from
> locking perspective?

I had a similar problem for Ethernet cable testing. It takes a few
seconds to perform a cable test, and you don't want to hold the RTNL
for that, if you can help it. I made changes to the PHY state machine,
so it has an additional state, indicating a cable test is in
operation. The ethtool netlink op simply starts the cable test and
then returns. Once the cable test is complete, it async reports the
results to user space.

So for module upgrade, you probably need to add a per module state
machine. Changes to the state need to hold RTNL, but you can release
it between state changes.

   Andrew