Re: [PATCH regression] net: phy: fix initialization (config_init) for Marvel 88E1116R PHYs

[Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>]

Hello.

On 02-04-2014 3:55, Alexander Holler wrote:

> Commit 7cd1463664c2a15721ff4ccfb61d4d970815cb3d (introduced with 3.14)
> changed the initialization of the mv643xx_eth driver to use phy_init_hw()
> to reset the PHY. Unfortunately the initialization for the 88E1116R PHY
> was broken such, that it used mdelay() instead of really waiting for a
> reset to finish.

> The effect was that the ethernet on my Kirkwood 88F6281 based device didn't
> come up anymore (no carrier).

> Fix this by waiting for a reset to finish before proceeding further.

> Signed-off-by: Alexander Holler <holler@ahsoftware.de>

> Cc: Michal Simek <michal.simek@xilinx.com>
> Cc: Florian Fainelli <f.fainelli@gmail.com>
> Cc: <stable@vger.kernel.org>
> ---
>   drivers/net/phy/marvell.c | 8 ++++++--
>   1 file changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/net/phy/marvell.c b/drivers/net/phy/marvell.c
> index bd37e45..5b84808 100644
> --- a/drivers/net/phy/marvell.c
> +++ b/drivers/net/phy/marvell.c
> @@ -396,7 +396,9 @@ static int m88e1116r_config_init(struct phy_device *phydev)
>   	if (err < 0)
>   		return err;
>
> -	mdelay(500);
> +	do
> +		temp = phy_read(phydev, MII_BMCR);
> +	while (temp & BMCR_RESET);

    I don't understand what's the point of this BMCR reset if phy_init_hw() 
will have already done it before calling phy_init_hw().

> @@ -429,7 +431,9 @@ static int m88e1116r_config_init(struct phy_device *phydev)
>   	if (err < 0)
>   		return err;
>
> -	mdelay(500);
> +	do
> +		temp = phy_read(phydev, MII_BMCR);
> +	while (temp & BMCR_RESET);

    Not clear why it's necessary to reset one more time... Also, tight loop 
without a timeout (0.5 sec, specified by IEEE 802.3) doesn't look good. The 
comment above phy_poll_reset() suggests that it could be used in the PHY 
drivers as well. Florian?

WBR, Sergei