netdev - Re: [RFC PATCH v3 net-next 09/10] net: dsa: ocelot: felix: add support for VSC75XX control over SPI

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: <20210814120211.v2qjqgi6l3slnkq2@skbuf>
Date:   Sat, 14 Aug 2021 15:02:11 +0300
From:   Vladimir Oltean <olteanv@...il.com>
To:     Colin Foster <colin.foster@...advantage.com>
Cc:     andrew@...n.ch, vivien.didelot@...il.com, f.fainelli@...il.com,
        davem@...emloft.net, kuba@...nel.org, robh+dt@...nel.org,
        claudiu.manoil@....com, alexandre.belloni@...tlin.com,
        UNGLinuxDriver@...rochip.com, hkallweit1@...il.com,
        linux@...linux.org.uk, netdev@...r.kernel.org,
        devicetree@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [RFC PATCH v3 net-next 09/10] net: dsa: ocelot: felix: add
 support for VSC75XX control over SPI

On Sat, Aug 14, 2021 at 02:43:29PM +0300, Vladimir Oltean wrote:
> The issue is that the registers for the PCS1G block look nothing like
> the MDIO clause 22 layout, so anything that tries to map the struct
> ocelot_pcs over a struct mdio_device is going to look like a horrible
> shoehorn.
> 
> For that we might need Russell's assistance.
> 
> The documentation is at:
> http://ww1.microchip.com/downloads/en/DeviceDoc/VMDS-10489.pdf
> search for "Information about the registers for this product is available in the attached file."
> and then open the PDF embedded within the PDF.

In fact I do notice now that as long as you don't use any of the
optional phylink_mii_c22_pcs_* helpers in your PCS driver, then
struct phylink_pcs has pretty much zero dependency on struct mdio_device,
which means that I'm wrong and it should be completely within reach to
write a dedicated PCS driver for this hardware.

As to how to make the common felix.c work with different implementations
of struct phylink_pcs, one thing that certainly has to change is that
struct felix should hold a struct phylink_pcs **pcs and not a
struct lynx_pcs **pcs.

Does this mean that we should refactor lynx_pcs_create() to return a
struct phylink_pcs * instead of struct lynx_pcs *, and lynx_pcs_destroy()
to receive the struct phylink_pcs *, use container_of() and free the
larger struct lynx_pcs *? Yes, probably.

If you feel uncomfortable with this, I can try to refactor lynx_pcs to
make it easier to accomodate a different PCS driver in felix.