On Fri, Oct 04, 2019 at 05:45:22PM +0200, Nicolas Dichtel wrote:

Le 03/10/2019 à 18:19, Guillaume Nault a écrit :
[snip]

quoted

Why not using the existing NLM_F_ECHO mechanism?

IIUC, if rtnl_net_notifyid() did pass the proper nlmsghdr and portid to
rtnl_notify(), the later would automatically notify the caller with
updated information if the original request had the NLM_F_ECHO flag.

Good point. Note that with library like libnl, the auto sequence number check
will fail (seq number is 0 instead of the previous + 1) and thus must be bypassed.

Well, it's up to the caller of rtnl_notify() to build a netlink message
with proper sequence number. Currently, rtnl_net_notifyid() always sets
it to 0. That's probably why the libnl test fails.