On 7/25/19 4:59 PM, Andrew Lunn wrote:
> On Thu, Jul 25, 2019 at 04:56:37PM +0200, Marek Vasut wrote:
>> On 7/25/19 4:03 PM, Andrew Lunn wrote:
>>> On Wed, Jul 24, 2019 at 03:40:48PM +0200, Marek Vasut wrote:
>>>> From: Tristram Ha <tristram...@microchip.com>
>>>> +static void ksz8795_phy_setup(struct ksz_device *dev, int port,
>>>> + struct phy_device *phy)
>>>> +{
>>>> + if (port < dev->phy_port_cnt) {
>>>> + /*
>>>> + * SUPPORTED_Asym_Pause and SUPPORTED_Pause can be removed to
>>>> + * disable flow control when rate limiting is used.
>>>> + */
>>>> + linkmode_copy(phy->advertising, phy->supported);
>>>> + }
>>>> +}
>>>
>>> Hi Marek
>>>
>>> Do you know why this is needed?
>>
>> Unfortunately, no.
>>
>> It seems it copies supported features of the PHY to advertised features
>> of the PHY for ports which are downstream (i.e. not the CPU port).
>
> Hi Marek
>
> Could you test it without this copy? Do you get sensible values from
> ethtool? Does the pause configuration look sensible?
They do look OK even without the code.
