On 10/13/26 12:25, Yuyang Huang wrote:
- [Low] The commit message misdescribes the precedent it copies.
Looks like a nit, will adjust it if need new revision.
- [Low] The patch adds an 8-byte 'struct net_device *mc_dev' to every
struct netdev_hw_addr_list (include/linux/netdevice.h:253), but only…
It seems to me this issue by itself doesn't deserve a revision. Will
fix it if new revision is required.
I agree this is not a blocker, but I think it would be nice if you could
follow-up on this one.
The solution proposed by the LLM LGTM.
- [Low] __hw_addr_changed() (net/core/dev_addr_lists.c:22-26) reads the
device netns with plain dev_net(list->mc_dev).
This was raised on earlier versions. Every dev->mc writer already
reads dev_net() under only the address lock, so this adds no new race.
Same here.
Also, still speaking of possible follow-up, I think it would be nice if
you could change the addr list lifecycle to RCU (no idea of how much
complex would be that).
Thanks,
Paolo
