[Intel-wired-lan] [PATCH v1 net-next 1/2] igb: add PHY support for Broadcom 5461S

[Tim Harvey]

On Fri, Apr 17, 2015 at 1:23 PM, Jonathan Toppins
<jtoppins at cumulusnetworks.com> wrote:
> From: Alan Liebthal <alanl at cumulusnetworks.com>
>
> The Quanta LY8 Ethernet management port uses a Broadcom 5461S chip for
> the PHY layer. This adds support for this PHY to the Intel igb driver.
>
> Signed-off-by: Alan Liebthal <alanl at cumulusnetworks.com>
> Signed-off-by: Jonathan Toppins <jtoppins at cumulusnetworks.com>
> ---
<snip>
> --- a/drivers/net/ethernet/intel/igb/e1000_phy.c
> +++ b/drivers/net/ethernet/intel/igb/e1000_phy.c
> @@ -148,6 +148,13 @@ s32 igb_read_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 *data)
>          * Control register.  The MAC will take care of interfacing with the
>          * PHY to retrieve the desired data.
>          */
> +       if (phy->type == e1000_phy_bcm5461s) {
> +               mdic = rd32(E1000_MDICNFG);
> +               mdic &= ~E1000_MDICNFG_PHY_MASK;
> +               mdic |= (phy->addr << E1000_MDICNFG_PHY_SHIFT);
> +               wr32(E1000_MDICNFG, mdic);
> +       }
> +
>         mdic = ((offset << E1000_MDIC_REG_SHIFT) |
>                 (phy->addr << E1000_MDIC_PHY_SHIFT) |
>                 (E1000_MDIC_OP_READ));
> @@ -204,6 +211,13 @@ s32 igb_write_phy_reg_mdic(struct e1000_hw *hw, u32 offset, u16 data)
>          * Control register.  The MAC will take care of interfacing with the
>          * PHY to retrieve the desired data.
>          */
> +       if (phy->type == e1000_phy_bcm5461s) {
> +               mdic = rd32(E1000_MDICNFG);
> +               mdic &= ~E1000_MDICNFG_PHY_MASK;
> +               mdic |= (phy->addr << E1000_MDICNFG_PHY_SHIFT);
> +               wr32(E1000_MDICNFG, mdic);
> +       }
> +
>         mdic = (((u32)data) |
>                 (offset << E1000_MDIC_REG_SHIFT) |
>                 (phy->addr << E1000_MDIC_PHY_SHIFT) |
> @@ -2509,3 +2523,68 @@ static s32 igb_set_master_slave_mode(struct e1000_hw *hw)
>
>         return hw->phy.ops.write_reg(hw, PHY_1000T_CTRL, phy_data);
>  }

Jonathan,

Is this bcm5461s attached to an i210/i211? These changes look a lot
like some changes I'm trying to upstream to add support for i210/i211
which require the phy address in the MDICNFG register. If this is the
case, then I think the right approach is to check for hw->mac.type =
e1000_i210/e1000_i211 and I can submit my patch for review.

Regards,

Tim