Re: [PATCH net-next] ptp: add vclock timestamp conversion IOCTL

[Sebastien Laveze]

On Mon, 2021-09-27 at 07:59 -0700, Richard Cochran wrote:
> I'm not wild about having yet another ioctl for functionality that
> already exists.

I was expecting some pushback :)

> > This binding works well if the application requires all timestamps in the
> > same domain but is not convenient when multiple domains need to be
> > supported using a single socket.
> 
> Opening multiple sockets is not rocket science.

I agree but you end-up handling or filtering the same traffic for each
socket. Not rocket science, but not "ideal".

> > Typically, IEEE 802.1AS-2020 can be implemented using a single socket,
> > the CMLDS layer using raw PHC timestamps and the domain specific
> > timestamps converted in the appropriate gPTP domain using this IOCTL.
> 
> You say "typically", but how many applications actually do this?  I
> can't think of any at all.

The "typically" was more a reference to this possible implementation of
AS-2020 using a common CMLDS layer and several domains using a single
socket.

So, without this IOCTL the design would be 1 socket for CMLDS layer
and 1 socket for each domain plus some specific filtering for each
socket to avoid processing the unwanted traffic.

With this IOCTL, the design would be 1 socket and 1 conversion for the
sync messages in the appropriate domain.

This also brings a finer granularity for per-domain timestamps which
may be useful for other applications.

> 
> Thanks,
> Richard

Thanks,
Seb