> Could the changelog drop or reword that claim? Okay, I'll drop the claim in v4.
> This isn't a bug introduced by this patch, but can this copy expose uninitialized bytes from ha->addr? Not related to this patch, will send a follow up. > Can this positional resume cursor silently skip dev->mc entries? In theory yes, but this follows the existing dumps: there is no generation counter for dev->mc, and in_dev_dump_ifmcaddr() has the same window. I'd rather keep it consistent than add a counter just for this one. If the issue is really worth fixing, I will send a follow up to fix both places.
