Re: [PATCH net-next v2 0/2] Add Marvell PHY PTP support

["Russell King (Oracle)" <linux@armlinux.org.uk>]

On Mon, Apr 07, 2025 at 06:39:14PM +0200, Kory Maincent wrote:
> On Mon, 7 Apr 2025 17:32:43 +0100
> "Russell King (Oracle)" <linux@armlinux.org.uk> wrote:
> 
> > On Mon, Apr 07, 2025 at 06:20:28PM +0200, Kory Maincent wrote:
> > > On Mon, 7 Apr 2025 17:02:28 +0100
> > > "Russell King (Oracle)" <linux@armlinux.org.uk> wrote:
> > >   
> > > > On Mon, Apr 07, 2025 at 04:02:59PM +0200, Kory Maincent wrote:  
> >  [...]  
> > > > 
> > > > Is the PTP selection stuff actually sorted now? Last time I tested it
> > > > after it having been merged into the kernel for a while, it didn't work,
> > > > and I reported that fact. You haven't told me that you now expect it to
> > > > work.  
> > > 
> > > The last part of the series, the PTP selection support wasn't merged when
> > > you tested it, although the default PTP choice that causes your regression
> > > was merged.
> > > Now it is fully merged, even the ethtool support.
> > > https://lore.kernel.org/netdev/mjn6eeo6lestvo6z3utb7aemufmfhn5alecyoaz46dt4pwjn6v@4aaaz6qpqd4b/
> > > 
> > > The only issue is the rtln warning from the phy_detach function. About it, I
> > > have already sent you the work I have done throwing ASSERT_RTNL in
> > > phy_detach. Maybe I should resend it as RFC.
> > >   
> > > > I don't want this merged until such time that we can be sure that MVPP2
> > > > platforms can continue using the MVPP2 PTP support, which to me means
> > > > that the PTP selection between a MAC and PHY needs to work.  
> > > 
> > > It should works, the default PTP will be the MAC PTP and you will be able to
> > > select the current PTP between MAC and PHY with the following command:
> > > # ethtool --set-hwtimestamp-cfg eth0 index 0 qualifier precise
> > > Time stamping configuration for eth0:
> > > Hardware timestamp provider index: 0
> > > Hardware timestamp provider qualifier: Precise (IEEE 1588 quality)
> > > Hardware Transmit Timestamp Mode:
> > > 	off
> > > Hardware Receive Filter Mode:
> > > 	none
> > > Hardware Flags: none
> > > # ethtool --set-hwtimestamp-cfg eth0 index 1 qualifier precise
> > > Time stamping configuration for eth0:
> > > Hardware timestamp provider index: 1
> > > Hardware timestamp provider qualifier: Precise (IEEE 1588 quality)
> > > Hardware Transmit Timestamp Mode:
> > > 	off
> > > Hardware Receive Filter Mode:
> > > 	none
> > > Hardware Flags: none
> > > 
> > > You can list the PTPs with the dump command:
> > > # ethtool --show-time-stamping "*"
> > > 
> > > You will need to stop phc2sys and ptp4l during these change as linuxptp may
> > > face some issues during the PTP change.  
> > 
> > I'm preferring to my emails in connection with:
> > 
> > https://lore.kernel.org/r/ZzTMhGDoi3WcY6MR@shell.armlinux.org.uk
> > 
> > when I tested your work last time, it seemed that what was merged hadn't
> > even been tested. In the last email, you said you'd look into it, but I
> > didn't hear anything further. Have the problems I reported been
> > addressed?
> 
> It wasn't merged it was 19th version and it worked and was tested, but not
> with the best development design. I have replied to you that I will do some
> change in v20 to address this.
> https://lore.kernel.org/all/20241113171443.697ac278@kmaincent-XPS-13-7390/
> 
> It gets finally merged in v21.

Okay, so I'm pleased to report that this now works on the Macchiatobin:

# ./ynl/cli.py --spec netlink/specs/ethtool.yaml --no-schema --dump tsinfo-get --json '{"header":{"dev-name":"eth2"}}' 
[{'header': {'dev-index': 5, 'dev-name': 'eth2'},
  'hwtstamp-provider': {'index': 2, 'qualifier': 0},
  'phc-index': 2,
  'rx-filters': {'bits': {'bit': [{'index': 0, 'name': 'none'},
                                  {'index': 1, 'name': 'all'}]},
                 'nomask': True,
                 'size': 16},
  'timestamping': {'bits': {'bit': [{'index': 0, 'name': 'hardware-transmit'},
                                    {'index': 1, 'name': 'software-transmit'},
                                    {'index': 2, 'name': 'hardware-receive'},
                                    {'index': 3, 'name': 'software-receive'},
                                    {'index': 4,
                                     'name': 'software-system-clock'},
                                    {'index': 6,
                                     'name': 'hardware-raw-clock'}]},
                   'nomask': True,
                   'size': 18},
  'tx-types': {'bits': {'bit': [{'index': 0, 'name': 'off'},
                                {'index': 1, 'name': 'on'},
                                {'index': 2, 'name': 'onestep-sync'},
                                {'index': 3, 'name': 'onestep-p2p'}]},
               'nomask': True,
               'size': 4}},
 {'header': {'dev-index': 5, 'dev-name': 'eth2'},
  'hwtstamp-provider': {'index': 0, 'qualifier': 0},
  'phc-index': 0,
  'rx-filters': {'bits': {'bit': [{'index': 0, 'name': 'none'},
                                  {'index': 2, 'name': 'some'}]},
                 'nomask': True,
                 'size': 16},
  'timestamping': {'bits': {'bit': [{'index': 0, 'name': 'hardware-transmit'},
                                    {'index': 2, 'name': 'hardware-receive'},
                                    {'index': 3, 'name': 'software-receive'},
                                    {'index': 4,
                                     'name': 'software-system-clock'},
                                    {'index': 6,
                                     'name': 'hardware-raw-clock'}]},
                   'nomask': True,
                   'size': 18},
  'tx-types': {'bits': {'bit': [{'index': 0, 'name': 'off'},
                                {'index': 1, 'name': 'on'}]},
               'nomask': True,
               'size': 4}}]

where phc 2 is the mvpp2 clock, and phc 0 is the PHY.

# ethtool -T eth2
Time stamping parameters for eth2:
Capabilities:
        hardware-transmit
        software-transmit
        hardware-receive
        software-receive
        software-system-clock
        hardware-raw-clock
PTP Hardware Clock: 2
Hardware Transmit Timestamp Modes:
        off
        on
        onestep-sync
        onestep-p2p
Hardware Receive Filter Modes:
        none
        all

So I guess that means that by default it's using PHC 2, and thus using
the MVPP2 PTP implementation - which is good, it means that when we add
Marvell PHY support, this won't switch to the PHY implementation.

Now, testing ethtool:

$ ./ethtool --get-hwtimestamp-cfg eth2
netlink error: Operation not supported

Using ynl:

# ./ynl/cli.py --spec netlink/specs/ethtool.yaml --no-schema --dump tsconfig-get --json '{"header":{"dev-name":"eth2"}}'
[]

So, It's better, something still isn't correct as there's no
configuration. Maybe mvpp2 needs updating first? If that's the case,
then we're not yet in a position to merge PHY PTP support.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!