Re: [RFC v2 2/3] net: dsa: initial MT7628 tagging driver

[Daniel Golle <daniel@makrotopia.org>]

On Sat, Mar 14, 2026 at 05:08:44PM +0200, Joris Vaisvila wrote:
> Add support for the MT7628 embedded switch's tag.
> 
> The MT7628 tag is merged with the VLAN TPID field when a VLAN is
> appended by the switch hardware. It is not installed if the VLAN tag is
> already there on ingress. Due to this hardware quirk the tag cannot be
> trusted for port 0 if we don't know that the VLAN was added by the
> hardware. As a workaround for this the switch is configured to always
> append the port PVID tag even if the incoming packet is already tagged.
> The tagging driver can then trust that the tag is always accurate and
> the whole VLAN tag can be removed on ingress as it's only metadata for
> the tagger.
> 
> On egress the MT7628 tag allows precise TX, but the correct VLAN tag
> from tag_8021q is still appended or the switch will not forward the
> packet.
> 
> Signed-off-by: Joris Vaisvila <joey@tinyisr.com>
> ---
>  include/net/dsa.h    |  2 +
>  net/dsa/Kconfig      |  6 +++
>  net/dsa/Makefile     |  1 +
>  net/dsa/tag_mt7628.c | 93 ++++++++++++++++++++++++++++++++++++++++++++
>  4 files changed, 102 insertions(+)
>  create mode 100644 net/dsa/tag_mt7628.c
> 
> [...]
> +#define MT7628_TAG_TX_PORT_BIT_MASK GENMASK(5, 0)
> +#define MT7628_TAG_RX_PORT_MASK GENMASK(2, 0)
> +#define MT7628_TAG_LEN 4
> +
> +static struct sk_buff *mt7628_tag_xmit(struct sk_buff *skb,
> +				       struct net_device *dev)
> +{
> +	struct dsa_port *dp;
> +	u16 xmit_vlan;
> +	u8 *tag;
> +
> +	dp = dsa_user_to_port(dev);
> +	xmit_vlan = dsa_tag_8021q_standalone_vid(dp);
> +
> +	skb_push(skb, MT7628_TAG_LEN);
> +	dsa_alloc_etype_header(skb, MT7628_TAG_LEN);
> +
> +	tag = dsa_etype_header_pos_tx(skb);
> +
> +	tag[0] = ETH_P_8021Q >> 8;
> +	tag[1] = FIELD_PREP(MT7628_TAG_TX_PORT_BIT_MASK,
> +			    dsa_xmit_port_mask(skb, dev));
> +
> +	tag[2] = xmit_vlan >> 8;
> +	tag[3] = xmit_vlan & 0xff;

So there are basically two 16-bit fields which need endian swapping
from host-byte order (little endian on MT7628) to network byte-order.
Imho it would be better expressed by using something along the lines
of

	__be16 *tag;
...
	tag = dsa_etype_header_pos_tx(skb);
	tag[0] = htons(ETH_P_8021Q |
		       FIELD_PREP(MT7628_TAG_TX_PORT_BIT_MASK,
				  dsa_xmit_port_mask(skb, dev)));
	tag[1] = htons(xmit_vlan);


> +	return skb;
> +}
> +
> +static struct sk_buff *mt7628_tag_rcv(struct sk_buff *skb,
> +				      struct net_device *dev)
> +{
> +	int src_port;
> +	__be16 *phdr;
> +	u16 tpid;
> +
> +	if (unlikely(!pskb_may_pull(skb, MT7628_TAG_LEN)))
> +		return NULL;
> +
> +	phdr = dsa_etype_header_pos_rx(skb);
> +	tpid = ntohs(*phdr);
> +	skb_pull_rcsum(skb, MT7628_TAG_LEN);
> +	dsa_strip_etype_header(skb, MT7628_TAG_LEN);
> +
> +	src_port = tpid & MT7628_TAG_RX_PORT_MASK;
> +
> +	skb->dev = dsa_conduit_find_user(dev, 0, src_port);
> +	if (!skb->dev)
> +		return NULL;
> +	dsa_default_offload_fwd_mark(skb);
> +	return skb;

This looks like it is semantically identical to the RX path in tag_mtk.c.

Maybe the TX-side can be added to the existing mtk_tag.c driver, so the
RX-side can be shared? tag_brcm.c also defines multiple similar protocols
in the same C source file.