Re: [PATCH net v3 2/2] net: ethtool: let tsconfig reach a PHY-only timestamp provider

[Nicolai Buchwitz]

Hi Maxime

On 26.9.2026 20:03, Maxime Chevallier wrote:

> [...]

> > > [Severity: Medium]
> > >
> > > The removed ethnl_set_tsconfig_validate() required both 
> > > ndo_hwtstamp_set
> > > and ndo_hwtstamp_get. This new check in ethnl_set_tsconfig() only 
> > > tests
> > > ndo_hwtstamp_set. The matching check in tsconfig_prepare_data() only
> > > tests ndo_hwtstamp_get.
> > >
> > > What happens on a netdev that has ndo_hwtstamp_set but no
> > > ndo_hwtstamp_get, and has neither a default timestamping PHY nor a
> > > hwprov?
> >
> > There is no in-tree driver with ndo_hwtstamp_set but not _get. All 73 
> > files
> > have them in pairs...
> >
> > Even further, the asymmetry is on purpose. The get side tests 
> > ndo_hwtstamp_get
> > and the set side tests ndo_hwtstamp_set, each gate testing what its 
> > own operation
> > needs.
> >
> > If anyone insists on the symmetry, a v4 could add || 
> > !ndo_hwtstamp_get. But before
> > I respin again, let's hear if this is really needed.
>
> The current way LGTM :)

Thanks!

> BTW, I raised that point a while ago [1] but now that you're properly 
> handling
> the "MAC can't TS, but PHY can" case, we may be able to get rid of the
> SIOCSHWTSTAMP code in phy_mii_ioctl() ? as a followup of course :)

The code seems dead indeed. I will have a look and send a follow up then 
next days...

> [...]

Regards
Nicolai