[EXT] Re: [PATCH v1 2/5] net: phy: marvell: extend 88E2110 to use both 2.5GHz modes

[Marek Behun]

On Thu, 8 Apr 2021 10:24:22 +0200
Stefan Roese <sr at denx.de> wrote:

> Hi Stefan,
> Hi Marek,
> 
> On 25.03.21 13:59, Stefan Chulski wrote:
> >> Could you please ask internally at Marvell?
> >> We are trying to get to the bottom of this because we are stuck in
> >> development of code for Amethyst. We need to get 2500base-x to work, we
> >> need to know whether it does or does not support AN, or maybe does but
> >> only for some devices. Otherwise it may happen that some SFPs won't link
> >> with our hardware.
> >>
> >> Thank you.
> >>
> >> Marek  
> > 
> > To avoid confusions, I suggest we take this issue directly with Marvell
> > SoHo switch FAE. I'm willing to start another thread to discuss this(I
> > will check who can assist you).
> > 
> > Regarding to this patch series. Let's drop PHY_INTERFACE_MODE_SGMII_2500,
> > from PHY_INTERFACE enum and we would handle SGMII mode in 2.5G speed
> > differently in PPv2 driver.  
> 
> I'm just now getting back to this patch.
> 
> JFYI: PHY_INTERFACE_MODE_SGMII_2500 is used in NXP ethernet drivers as
> well:
> 
> $ git grep PHY_INTERFACE_MODE_SGMII_2500
> board/freescale/ls1012aqds/eth.c: 
>   PHY_INTERFACE_MODE_SGMII_2500);
> board/freescale/ls1012aqds/eth.c: 
>   PHY_INTERFACE_MODE_SGMII_2500);
> board/freescale/ls1012ardb/eth.c: 
>           PHY_INTERFACE_MODE_SGMII_2500);
> board/freescale/ls1012ardb/eth.c: 
>           PHY_INTERFACE_MODE_SGMII_2500);
> board/freescale/ls1043aqds/eth.c: 
> PHY_INTERFACE_MODE_SGMII_2500) {
> board/freescale/ls1043aqds/eth.c:               case 
> PHY_INTERFACE_MODE_SGMII_2500:
> board/freescale/ls1043aqds/eth.c:                       } else if 
> (interface == PHY_INTERFACE_MODE_SGMII_2500) {
> board/freescale/ls1046aqds/eth.c:       } else if 
> (fm_info_get_enet_if(port) == PHY_INTERFACE_MODE_SGMII_2500) {
> board/freescale/t102xrdb/eth_t102xrdb.c:                case 
> PHY_INTERFACE_MODE_SGMII_2500:
> board/freescale/t102xrdb/eth_t102xrdb.c:        if 
> (((fm_info_get_enet_if(port) == PHY_INTERFACE_MODE_SGMII_2500) ||
> drivers/net/fm/eth.c:                   PHY_INTERFACE_MODE_SGMII_2500) ? 
> true : false;
> drivers/net/fm/eth.c:       fm_eth->enet_if == 
> PHY_INTERFACE_MODE_SGMII_2500)
> drivers/net/fm/eth.c:        (fm_eth->enet_if == 
> PHY_INTERFACE_MODE_SGMII_2500) ||
> drivers/net/fm/eth.c:   if (fm_eth->enet_if == 
> PHY_INTERFACE_MODE_SGMII_2500)
> drivers/net/fm/eth.c:   case PHY_INTERFACE_MODE_SGMII_2500:
> drivers/net/fm/ls1043.c:                        return 
> PHY_INTERFACE_MODE_SGMII_2500;
> drivers/net/fm/ls1043.c:                        return 
> PHY_INTERFACE_MODE_SGMII_2500;
> drivers/net/fm/ls1046.c:                        return 
> PHY_INTERFACE_MODE_SGMII_2500;
> drivers/net/fm/memac.c: case PHY_INTERFACE_MODE_SGMII_2500:
> drivers/net/fm/t1024.c:                 return 
> PHY_INTERFACE_MODE_SGMII_2500;
> drivers/net/fsl_enetc.c:        if (priv->if_type == 
> PHY_INTERFACE_MODE_SGMII_2500)
> drivers/net/fsl_enetc.c:        case PHY_INTERFACE_MODE_SGMII_2500:
> drivers/net/mscc_eswitch/felix_switch.c:            phy->interface == 
> PHY_INTERFACE_MODE_SGMII_2500)
> drivers/net/mscc_eswitch/felix_switch.c:        case 
> PHY_INTERFACE_MODE_SGMII_2500:
> ...
> 
> Stefan, you suggest to drop this define from PHY_INTERFACE enum which
> we can't easily do with other drivers (like NXP) also referencing this
> macro.
> 
> How to continue then?

I think NXP also uses this macro incorrectly, they should instead use
2500BASEX. SGMII_2500 does not exist in Linux. We can look at the
corresponding NXP driver in Linux (if it exists) to see if the
corresponging mode is 2500BASEX.

Marek