linux-kernel - Re: PHY reset question

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: <c6bce70b-c97a-1696-2113-61bd3ba6ae99@gmail.com>
Date:   Wed, 7 Oct 2020 08:44:53 -0700
From:   Florian Fainelli <f.fainelli@...il.com>
To:     Marco Felsch <m.felsch@...gutronix.de>,
        Marek Vasut <marex@...x.de>, Rob Herring <robh+dt@...nel.org>
Cc:     Oleksij Rempel <o.rempel@...gutronix.de>,
        Andrew Lunn <andrew@...n.ch>, netdev@...r.kernel.org,
        Russell King <linux@...linux.org.uk>,
        linux-kernel@...r.kernel.org, mkl@...gutronix.de,
        kernel@...gutronix.de, David Jander <david@...tonic.nl>
Subject: Re: PHY reset question

On 10/7/2020 3:47 AM, Marco Felsch wrote:
> Florian did you send a new version of those patches?

I did not because we had a good conversation with Rob over IRC and the 
conclusion was that the only solution that scaled across drivers, 
subsystems and type of resources (regulators, clocks, resets, etc.) was 
to have a compatible string for the given device that contains the ID. 
For Ethernet PHY or MDIO device nodes that is "ethernet-phyAAAA.BBBB".

When the bus determines the presence of such a compatible string it 
needs to bypass the dynamic identification of the device and needs to 
bind the PHY driver and the device instance directly. MDIO does that, 
and so does I2C and SPI AFAICT with the modalias/compatible (there is 
not a standardized way to runtime detect an I2C or SPI client anyway), 
while PCI and USB do not, but arguably could in the future.

For the specific use case that I had which required turning on a clock 
to the Ethernet PHY, I ended up modifying the firmware to provide that 
compatible string "ethernetAAAA.BBBB" and have the driver request the 
clock from its probe function:

https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git/commit/drivers/net/phy/bcm7xxx.c?id=ba4ee3c053659119472135231dbef8f6880ce1fb
-- 
Florian