netdev - Re: [PATCH net-next 2/4] net: dsa: remove phy arg from port enable/disable

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: <87377eob5t.fsf@weeman.i-did-not-set--mail-host-address--so-tickle-me>
Date:   Fri, 22 Sep 2017 14:12:14 -0400
From:   Vivien Didelot <vivien.didelot@...oirfairelinux.com>
To:     Florian Fainelli <f.fainelli@...il.com>, netdev@...r.kernel.org
Cc:     linux-kernel@...r.kernel.org, kernel@...oirfairelinux.com,
        "David S. Miller" <davem@...emloft.net>,
        Andrew Lunn <andrew@...n.ch>
Subject: Re: [PATCH net-next 2/4] net: dsa: remove phy arg from port enable/disable

Hi Florian,

Florian Fainelli <f.fainelli@...il.com> writes:

> On 09/22/2017 09:17 AM, Vivien Didelot wrote:
>> The .port_enable and .port_disable functions are meant to deal with the
>> switch ports only, and no driver is using the phy argument anyway.
>> Remove it.
>
> I don't think this makes sense, there are perfectly legit reasons why a
> switch driver may have something to do with the PHY device attached to
> its per-port network interface, we should definitively keep that around,
> unless you think we should be accessing the PHY within the switch
> drivers by doing:
>
> struct phy_device *phydev = ds->ports[port].netdev->phydev?

bcm_sf2 is the only user for this phy argument right now. The reason I'm
doing this is because I prefer to discourage switch drivers to dig into
the phy device themselves while as you said there must be a cleaner
solution. This must be handled somehow elsewhere in the stack.

In the meantime, moving the PHY device up to the dsa_port structure is a
good solution, in order not to expose it in switch ops, but still make
it available to more complex drivers.

Do you know if netdev->phydev is usable? Why do DSA has its own copy in
dsa_slave_priv then?

I'll respin, thanks.

    Vivien