Re: [PATCH net-next] net: sfp: add quirk for ZOERAX SFP-2.5G-T

["Russell King (Oracle)" <linux@armlinux.org.uk>]

On Sat, Mar 28, 2026 at 09:02:34PM +0100, Jan Hoffmann wrote:
> On 2026-03-28 at 04:50, Jakub Kicinski wrote:
> > On Wed, 25 Mar 2026 17:35:58 +0100 Jan Hoffmann wrote:
> > > This is a 2.5G copper module which appears to be based on a Motorcomm
> > > YT8821 PHY. There doesn't seem to be a usable way to to access the PHY
> > > (I2C address 0x56 provides only read-only C22 access, and Rollball is
> > > also not working).
> > > 
> > > The module does not report the correct extended compliance code for
> > > 2.5GBase-T, and instead claims to support SONET OC-48 and Fibre Channel:
> > > 
> > >    Identifier          : 0x03 (SFP)
> > >    Extended identifier : 0x04 (GBIC/SFP defined by 2-wire interface ID)
> > >    Connector           : 0x07 (LC)
> > >    Transceiver codes   : 0x00 0x01 0x00 0x00 0x40 0x40 0x04 0x00 0x00
> > >    Transceiver type    : FC: Multimode, 50um (M5)
> > >    Encoding            : 0x05 (SONET Scrambled)
> > >    BR Nominal          : 2500MBd
> > > 
> > > Despite this, the kernel still sets 2500Base-X as interface mode based
> > > on the (incorrect) nominal signaling rate.
> > > 
> > > However, it is also necessary to disable auto-negotiation for the module
> > > to actually work. Thus, create a SFP quirk to do this.
> > > 
> > > Signed-off-by: Jan Hoffmann <jan@3e8.eu>
> > > ---
> > > Note: I'm not quite sure "sfp_quirk_disable_autoneg" is enough here, or
> > > if it would be appropriate to use the full "sfp_quirk_oem_2_5g" quirk
> > > due to the not really correct compliance code / nominal signaling rate.
> > 
> > I'm no SFP expert but just going by the naming I strongly suspect that
> > the module is the same hardware as "SFP-2.5G-T" from "FS", so keeping
> > quirks consistent makes sense?
> Based on commit e27aca3760c0 ("net: sfp: enhance quirk for Fibrestore 2.5G
> copper SFP module"), the FS module uses an RTL8221B PHY, so it is different
> hardware. And it also supports the Rollball protocol for PHY access, unlike
> this Zoerax module.
> 
> From the existing quirks, the only valid alternative I see for the Zoerax
> module would be "sfp_quirk_oem_2_5g" (which also sets the additional bits
> for 2500Base-T link mode and explicitly enables 2500Base-X interface mode).
> But as that is not strictly necessary for making this particular module
> work, I am unsure if using this quirk would be appropriate.

See commit 50e96acbe11667b6fe9d99e1348c6c224b2f11dd which added this
function. This was for an OEM SFP-2.5G-T where the PHY was similarly
inaccessible, can only run at 2500base-X without inband and offers
2500Base-T connectivity.

This is basically the same as your Zoerax module.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!