[PATCH net-next 6/7] net: pch_gbe: convert to use ndo_hwtstamp callbacks

[Vadim Fedorenko]

On 31/10/2025 18:38, Jakub Kicinski wrote:
> On Fri, 31 Oct 2025 18:35:25 +0100 Kory Maincent wrote:
>>>   	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
>>> -		adapter->hwts_rx_en = 0;
>>> +		adapter->hwts_rx_en = cfg->rx_filter;
>>
>> It seems there is a functional change here.
>>
>>> 		pch_ch_control_write(pdev, SLAVE_MODE | CAP_MODE0);
> 
> Good catch. Looks like SLAVE | MODE0 translates to 0 | 0
> so presumably the device doesn't actually support timestamping
> of V1 L4 SYNC messages? Unclear what to do about this.
> Maybe let's leave it be? keep the hwts_rx_en = 0; ?
> Not strictly correct but at least we won't break anything.

Ok, it looked like typo as the function actually does the register
manipulation while it's not doing anything for disabling RX timestamping
which probably means it's enabled by default.
But as the code was added back in 2012, and nobody complained, we should 
probably keep it as is. I'll send v2 shortly.