| Message ID | 1679470532-163226-1-git-send-email-paulb@nvidia.com |
|---|---|
| State | Accepted, archived |
| Headers | show |
| Series | [nf-next,1/1] netfilter: ctnetlink: Support offloaded conntrack entry deletion | expand |
On Wed, Mar 22, 2023 at 09:35:32AM +0200, Paul Blakey wrote: > Currently, offloaded conntrack entries (flows) can only be deleted > after they are removed from offload, which is either by timeout, > tcp state change or tc ct rule deletion. This can cause issues for > users wishing to manually delete or flush existing entries. > > Support deletion of offloaded conntrack entries. > > Example usage: > # Delete all offloaded (and non offloaded) conntrack entries > # whose source address is 1.2.3.4 > $ conntrack -D -s 1.2.3.4 > # Delete all entries > $ conntrack -F > > Signed-off-by: Paul Blakey <paulb@nvidia.com> Reviewed-by: Simon Horman <simon.horman@corigine.com>
Hi Paul, Apologies for lagging to catch up with this. On Wed, Mar 22, 2023 at 09:35:32AM +0200, Paul Blakey wrote: > Currently, offloaded conntrack entries (flows) can only be deleted > after they are removed from offload, which is either by timeout, > tcp state change or tc ct rule deletion. This can cause issues for > users wishing to manually delete or flush existing entries. > > Support deletion of offloaded conntrack entries. > > Example usage: > # Delete all offloaded (and non offloaded) conntrack entries > # whose source address is 1.2.3.4 > $ conntrack -D -s 1.2.3.4 > # Delete all entries > $ conntrack -F This fine with me. I think probably it much be documented somewhere that in case of hardware offload, deletion is asynchronous.
diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index bfc3aaa2c872..fbc47e4b7bc3 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -1554,9 +1554,6 @@ static const struct nla_policy ct_nla_policy[CTA_MAX+1] = { static int ctnetlink_flush_iterate(struct nf_conn *ct, void *data) { - if (test_bit(IPS_OFFLOAD_BIT, &ct->status)) - return 0; - return ctnetlink_filter_match(ct, data); } @@ -1626,11 +1623,6 @@ static int ctnetlink_del_conntrack(struct sk_buff *skb, ct = nf_ct_tuplehash_to_ctrack(h); - if (test_bit(IPS_OFFLOAD_BIT, &ct->status)) { - nf_ct_put(ct); - return -EBUSY; - } - if (cda[CTA_ID]) { __be32 id = nla_get_be32(cda[CTA_ID]);
Currently, offloaded conntrack entries (flows) can only be deleted after they are removed from offload, which is either by timeout, tcp state change or tc ct rule deletion. This can cause issues for users wishing to manually delete or flush existing entries. Support deletion of offloaded conntrack entries. Example usage: # Delete all offloaded (and non offloaded) conntrack entries # whose source address is 1.2.3.4 $ conntrack -D -s 1.2.3.4 # Delete all entries $ conntrack -F Signed-off-by: Paul Blakey <paulb@nvidia.com> --- net/netfilter/nf_conntrack_netlink.c | 8 -------- 1 file changed, 8 deletions(-)