RE: [net-next, v5, 08/11] net: sock: extend SO_TIMESTAMPING for PHC binding

["Y.b. Lu" <yangbo.lu@nxp.com>]

Hi Richard,

> -----Original Message-----
> From: Richard Cochran <richardcochran@gmail.com>
> Sent: 2021年7月4日 21:34
> To: Y.b. Lu <yangbo.lu@nxp.com>
> Cc: netdev@vger.kernel.org; linux-kernel@vger.kernel.org;
> linux-kselftest@vger.kernel.org; mptcp@lists.linux.dev; David S . Miller
> <davem@davemloft.net>; Jakub Kicinski <kuba@kernel.org>; Mat Martineau
> <mathew.j.martineau@linux.intel.com>; Matthieu Baerts
> <matthieu.baerts@tessares.net>; Shuah Khan <shuah@kernel.org>; Michal
> Kubecek <mkubecek@suse.cz>; Florian Fainelli <f.fainelli@gmail.com>;
> Andrew Lunn <andrew@lunn.ch>; Rui Sousa <rui.sousa@nxp.com>; Sebastien
> Laveze <sebastien.laveze@nxp.com>
> Subject: Re: [net-next, v5, 08/11] net: sock: extend SO_TIMESTAMPING for
> PHC binding
> 
> On Wed, Jun 30, 2021 at 04:11:59PM +0800, Yangbo Lu wrote:
> > Since PTP virtual clock support is added, there can be several PTP
> > virtual clocks based on one PTP physical clock for timestamping.
> >
> > This patch is to extend SO_TIMESTAMPING API to support PHC (PTP
> > Hardware Clock) binding by adding a new flag
> > SOF_TIMESTAMPING_BIND_PHC. When PTP virtual clocks are in use, user
> > space can configure to bind one for timestamping, but PTP physical
> > clock is not supported and not needed to bind.
> 
> Would it not be better to simply bind automatically?
> 
> Like this pseudo code:
> 
> 	if (hw_timestamping_requested() && interface_is_vclock()) {
> 		bind_vclock();
> 	}
> 
> It would be great to avoid forcing user space to use a new option.
> 
> Especially because NOT setting the option makes no sense.  Or maybe there is
> a use case for omitting the option?
> 
> 
> Thoughts?

When several ptp virtual clocks are created, the ptp physical clock is guaranteed for free running.

What I think is, for timestamping, if no flag SOF_TIMESTAMPING_BIND_PHC, the timestamping keeps using ptp physical clock.
If application wants to bind one ptp virtual clock for timestamping, the flag SOF_TIMESTAMPING_BIND_PHC should be set and clock index should be provided.
After all, several ptp virtual clocks created are likely for different timescale/use case. There should be a method for any of applications to select the right one to use.

Does it make sense?
Thank you.

> 
> Thanks,
> Richard