Re: [PATCH net v2] net: phy: dp83867: add w/a for packet errors seen with short cables

[Siddharth Vadapalli]

On 09-05-2023 16:10, Heiner Kallweit wrote:
> On 09.05.2023 12:27, Siddharth Vadapalli wrote:
>>
>>
>> On 09/05/23 14:29, Heiner Kallweit wrote:
>>> On 09.05.2023 07:21, Siddharth Vadapalli wrote:
>>>> From: Grygorii Strashko <grygorii.strashko@xxxxxx>
>>>>
>>>> Introduce the W/A for packet errors seen with short cables (<1m) between
>>>> two DP83867 PHYs.
>>>>
>>>> The W/A recommended by DM requires FFE Equalizer Configuration tuning by
>>>> writing value 0x0E81 to DSP_FFE_CFG register (0x012C), surrounded by hard
>>>> and soft resets as follows:
>>>>
>>>> write_reg(0x001F, 0x8000); //hard reset
>>>> write_reg(DSP_FFE_CFG, 0x0E81);
>>>> write_reg(0x001F, 0x4000); //soft reset
>>>>
>>>> Since  DP83867 PHY DM says "Changing this register to 0x0E81, will not
>>>> affect Long Cable performance.", enable the W/A by default.
>>>>
>>>> Fixes: 2a10154abcb7 ("net: phy: dp83867: Add TI dp83867 phy")
>>>> Signed-off-by: Grygorii Strashko <grygorii.strashko@xxxxxx>
>>>> Signed-off-by: Siddharth Vadapalli <s-vadapalli@xxxxxx>
>>>> ---
>>>>
>>>> V1 patch at:
>>>> https://lore.kernel.org/r/20230508070019.356548-1-s-vadapalli@xxxxxx
>>>>
>>>> Changes since v1 patch:
>>>> - Wrap the line invoking phy_write_mmd(), limiting it to 80 characters.
>>>> - Replace 0X0E81 with 0x0e81 in the call to phy_write_mmd().
>>>> - Replace 0X012C with 0x012c in the new define for DP83867_DSP_FFE_CFG.
>>>>
>>>> RFC patch at:
>>>> https://lore.kernel.org/r/20230425054429.3956535-2-s-vadapalli@xxxxxx/
>>>>
>>>> Changes since RFC patch:
>>>> - Change patch subject to PATCH net.
>>>> - Add Fixes tag.
>>>> - Check return value of phy_write_mmd().
>>>>
>>>>  drivers/net/phy/dp83867.c | 18 +++++++++++++++++-
>>>>  1 file changed, 17 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/drivers/net/phy/dp83867.c b/drivers/net/phy/dp83867.c
>>>> index d75f526a20a4..bbdcc595715d 100644
>>>> --- a/drivers/net/phy/dp83867.c
>>>> +++ b/drivers/net/phy/dp83867.c
>>>> @@ -44,6 +44,7 @@
>>>>  #define DP83867_STRAP_STS1	0x006E
>>>>  #define DP83867_STRAP_STS2	0x006f
>>>>  #define DP83867_RGMIIDCTL	0x0086
>>>> +#define DP83867_DSP_FFE_CFG	0x012c
>>>>  #define DP83867_RXFCFG		0x0134
>>>>  #define DP83867_RXFPMD1	0x0136
>>>>  #define DP83867_RXFPMD2	0x0137
>>>> @@ -941,8 +942,23 @@ static int dp83867_phy_reset(struct phy_device *phydev)
>>>>  
>>>>  	usleep_range(10, 20);
>>>>  
>>>> -	return phy_modify(phydev, MII_DP83867_PHYCTRL,
>>>> +	err = phy_modify(phydev, MII_DP83867_PHYCTRL,
>>>>  			 DP83867_PHYCR_FORCE_LINK_GOOD, 0);
>>>> +	if (err < 0)
>>>> +		return err;
>>>> +
>>>
>>> Would be good to add a comment here explaining what this magic write does.
>>
>> Sure. Is the following comment acceptable?
>>
>> "Configure the DSP Feedforward Equalizer Configuration register to improve short
>> cable (< 1 meter) performance. This will not affect long cable performance."
>>
> Sounds good. Important is just that the magic value write is explained, so that
> readers of the source code don't have to scroll through the commit history.

Thank you for letting me know. I will post the v3 patch with the comment.

> 
>>>
>>>> +	err = phy_write_mmd(phydev, DP83867_DEVADDR, DP83867_DSP_FFE_CFG,
>>>> +			    0x0e81);
>>>> +	if (err < 0)
>>>> +		return err;
>>>> +
>>>> +	err = phy_write(phydev, DP83867_CTRL, DP83867_SW_RESTART);
>>>> +	if (err < 0)
>>>> +		return err;
>>>> +
>>>> +	usleep_range(10, 20);
>>>> +
>>>> +	return 0;
>>>>  }
>>>>  
>>>>  static void dp83867_link_change_notify(struct phy_device *phydev)
>>>
>>
> 

-- 
Regards,
Siddharth.