Re: [PATCH net-next v6 0/8] hsr: Add additional info to send/ receive skbs

[Sebastian Andrzej Siewior <bigeasy@linutronix.de>]

On 2026-09-23 14:14:01 [+0200], To netdev@vger.kernel.org wrote:
> I am trying to extend linuxptp to support PTP over a HSR network.
…

sashiko:
- #2
| Does this manual offset adjustment corrupt the hardware checksum offload offset?

Nothing sets ->ip_summed here. But the bot seems to be right here.

| Does this overly strict pull check drop valid small Ethernet frames?

after revisiting I tend to replace it with
		 if (!pskb_may_pull(skb, HSR_HLEN + movelen))
with
	         if (!pskb_may_pull(skb, movelen))

since the memmove() below only moves movelen bytes, so there is no need
to check for the HSR_HLEN which are in the headroom.

| Does this hardcoded offset corrupt the network_header for VLAN frames?
okay, will add the VLAN header.

- #3 
| Does this code break UAPI backward compatibility for standard userspace
| applications?

complains about the in-line header. Intended.

- #4
| Will packets arriving from a Single Attached Node (SAN) via the interlink
| port bypass the new PTP drop logic?

Yes. I guess it is okay to drop the PTP packets from SAN port, too.

| Does this check miss PTP packets encapsulated in VLANs or UDP/IP?

Yes. The specification does not mention UDP/IP for PTP packets. I will
ignore VLAN filtering for now.

- #5
| Does this code unconditionally drop standard L2 PTP traffic?

Complains about the in-line header again.

Let me fix that.

Sebastian