Re: [PATCH net-next 4/4] net: mscc: ocelot: set up traps for PTP packets

[Kurt Kanzenbach <kurt@linutronix.de>]

[-- Attachment #1: Type: text/plain, Size: 1159 bytes --]

Hi Richard,

On Fri Nov 26 2021, Richard Cochran wrote:
> On Fri, Nov 26, 2021 at 05:01:13PM +0000, Vladimir Oltean wrote:
>
>> This, to me, sounds more like the bridge trapping the packets on br0
>> instead of letting them flow on the port netdevices, which is solved by
>> some netfilter rules? Or is it really a driver/hardware issue?
>> 
>> https://lore.kernel.org/netdev/20211116102138.26vkpeh23el6akya@skbuf/
>
> Yeah, thanks for the link.  I had seen it, but alas it came too late
> for me to try on actual working HW.  Maybe it fixes the issue.
>
> If someone out there has a Marvell switch, please try it and let us
> know...

I did some testing on a Marvell Topaz switch (mv88e6341). The Boundary
Clock over UDPv4/UDPv6 using Vladimir's netfilter rules doesn't
work. Even though I've successfully tested these rules on TI CPSW
(switchdev variant) and hellcreek.

However, BC over Layer 2 transport is also not really working well, as
the switch forwards PTP packets (instead of trapping?). Furthermore,
ptp4l reports loosing of timestamps when running on multiple switch
ports.

Tested with Linux v5.16.

Thanks,
Kurt

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 861 bytes --]