[PATCH V3 2/2] net/mlx5: add support for flows targeting multicast MAC addresses

[PATCH V3 2/2] net/mlx5: add support for flows targeting multicast MAC addresses

[Gavin Li]

Rules for multicast MAC addresses are intended to filter multicast traffic
and are managed through multicast MAC add/remove APIs. In mlx5_dev_spawn
function, devices (PF, VFs, and SFs) retrieve the netdev-configured MAC
addresses via netlink and store them in the PMD device data, which
includes multicast MAC addresses.

Previously, flows for multicast MAC addresses were incorrectly disabled,
causing the multicast MAC add API to stop working. As a result, multicast
traffic directed to those multicast MAC addresses was not received.

To resolve this and update the multicast MAC address rules, create them
within mlx5_traffic_enable.

Fixes: 2d0665a7f771 ("net/mlx5: align PF and VF/SF MAC address handling")
Cc: stable@dpdk.org

Signed-off-by: Gavin Li <gavinl@nvidia.com>
---
 drivers/net/mlx5/mlx5_trigger.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/mlx5/mlx5_trigger.c b/drivers/net/mlx5/mlx5_trigger.c
index 6c6f228afd..46479fbf09 100644
--- a/drivers/net/mlx5/mlx5_trigger.c
+++ b/drivers/net/mlx5/mlx5_trigger.c
@@ -1813,7 +1813,10 @@ mlx5_traffic_enable(struct rte_eth_dev *dev)
 	for (i = 0; i != MLX5_MAX_MAC_ADDRESSES; ++i) {
 		struct rte_ether_addr *mac = &dev->data->mac_addrs[i];
 
-		if (!memcmp(mac, &cmp, sizeof(*mac)) || rte_is_multicast_ether_addr(mac))
+		/* Add flows for unicast and multicast mac addresses added by API. */
+		if (!memcmp(mac, &cmp, sizeof(*mac)) ||
+		    !BITFIELD_ISSET(priv->mac_own, i) ||
+		    (dev->data->all_multicast && rte_is_multicast_ether_addr(mac)))
 			continue;
 		memcpy(&unicast.hdr.dst_addr.addr_bytes,
 		       mac->addr_bytes,
-- 
2.34.1

Re: [PATCH V3 2/2] net/mlx5: add support for flows targeting multicast MAC addresses

[Thomas Monjalon]

25/08/2025 16:13, Gavin Li:
> Rules for multicast MAC addresses are intended to filter multicast traffic
> and are managed through multicast MAC add/remove APIs.

You should name the functions to be explicit.
Also it is not only managed via DPDK API,
as you say below we also retrieve kernel addresses.

> In mlx5_dev_spawn
> function, devices (PF, VFs, and SFs) retrieve the netdev-configured MAC
> addresses via netlink and store them in the PMD device data, which
> includes multicast MAC addresses.
> 
> Previously, flows for multicast MAC addresses were incorrectly disabled,
> causing the multicast MAC add API to stop working.

To be clear, it was blocking addresses added in DPDK,
not ones created in kernel, right?

> As a result, multicast
> traffic directed to those multicast MAC addresses was not received.
> 
> To resolve this and update the multicast MAC address rules, create them
> within mlx5_traffic_enable.

Actually you allow default rules for multicast addresses to be created.

> -		if (!memcmp(mac, &cmp, sizeof(*mac)) || rte_is_multicast_ether_addr(mac))
> +		/* Add flows for unicast and multicast mac addresses added by API. */
> +		if (!memcmp(mac, &cmp, sizeof(*mac)) ||
> +		    !BITFIELD_ISSET(priv->mac_own, i) ||
> +		    (dev->data->all_multicast && rte_is_multicast_ether_addr(mac)))
>  			continue;

PS: you forgot to use --in-reply-to to keep all versions in the same mail thread.

Re: [PATCH V3 2/2] net/mlx5: add support for flows targeting multicast MAC addresses

[Minggang(Gavin) Li]

On 8/25/2025 10:52 PM, Thomas Monjalon wrote:
> 25/08/2025 16:13, Gavin Li:
>> Rules for multicast MAC addresses are intended to filter multicast traffic
>> and are managed through multicast MAC add/remove APIs.
> You should name the functions to be explicit.
> Also it is not only managed via DPDK API,
> as you say below we also retrieve kernel addresses.
ACK
>
>> In mlx5_dev_spawn
>> function, devices (PF, VFs, and SFs) retrieve the netdev-configured MAC
>> addresses via netlink and store them in the PMD device data, which
>> includes multicast MAC addresses.
>>
>> Previously, flows for multicast MAC addresses were incorrectly disabled,
>> causing the multicast MAC add API to stop working.
> To be clear, it was blocking addresses added in DPDK,
> not ones created in kernel, right?
It was blocking all multicast mac address including those added in DPDK 
and ones created in kernel.
>
>> As a result, multicast
>> traffic directed to those multicast MAC addresses was not received.
>>
>> To resolve this and update the multicast MAC address rules, create them
>> within mlx5_traffic_enable.
> Actually you allow default rules for multicast addresses to be created.
>
>> -		if (!memcmp(mac, &cmp, sizeof(*mac)) || rte_is_multicast_ether_addr(mac))
>> +		/* Add flows for unicast and multicast mac addresses added by API. */
>> +		if (!memcmp(mac, &cmp, sizeof(*mac)) ||
>> +		    !BITFIELD_ISSET(priv->mac_own, i) ||
>> +		    (dev->data->all_multicast && rte_is_multicast_ether_addr(mac)))
>>   			continue;
> PS: you forgot to use --in-reply-to to keep all versions in the same mail thread.
Sorry for the misleading. Will add it in next version.
>
>