<div dir="ltr"><div>Acked-by: Zachary Tahenakos <<a href="mailto:zachary.tahenakos@canonical.com">zachary.tahenakos@canonical.com</a>></div> <div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Apr 5, 2022 at 11:39 AM Bodong Wang <<a href="mailto:bodong@nvidia.com">bodong@nvidia.com</a>> wrote: </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">From: Paul Blakey <<a href="mailto:paulb@nvidia.com" target="_blank">paulb@nvidia.com</a>> BugLink: <a href="https://bugs.launchpad.net/bugs/1967892" rel="noreferrer" target="_blank">https://bugs.launchpad.net/bugs/1967892</a> After cited commit optimizted hw insertion, flow table entries are populated with ifindex information which was intended to only be used for HW offload. This tuple ifindex is hashed in the flow table key, so it must be filled for lookup to be successful. But tuple ifindex is only relevant for the netfilter flowtables (nft), so it's not filled in act_ct flow table lookup, resulting in lookup failure, and no SW offload and no offload teardown for TCP connection FIN/RST packets. To fix this, add new tc ifindex field to tuple, which will only be used for offloading, not for lookup, as it will not be part of the tuple hash. Fixes: 9795ded7f924 ("net/sched: act_ct: Fill offloading tuple iifidx") Signed-off-by: Paul Blakey <<a href="mailto:paulb@nvidia.com" target="_blank">paulb@nvidia.com</a>> Signed-off-by: Pablo Neira Ayuso <<a href="mailto:pablo@netfilter.org" target="_blank">pablo@netfilter.org</a>> (backported from commit db6140e5e35a48405e669353bd54042c1d4c3841) [Oz: Add missing enum ] Signed-off-by: Oz Shlomo <<a href="mailto:ozsh@nvidia.com" target="_blank">ozsh@nvidia.com</a>> Signed-off-by: Bodong Wang <<a href="mailto:bodong@nvidia.com" target="_blank">bodong@nvidia.com</a>> --- include/net/netfilter/nf_flow_table.h | 16 ++++++++++++++++ net/netfilter/nf_flow_table_offload.c | 6 +++++- net/sched/act_ct.c | 13 +++++++++---- 3 files changed, 30 insertions(+), 5 deletions(-) diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index b40772f..a0c11bc 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -88,6 +88,14 @@ enum flow_offload_tuple_dir { FLOW_OFFLOAD_DIR_MAX = IP_CT_DIR_MAX }; +enum flow_offload_xmit_type { + FLOW_OFFLOAD_XMIT_UNSPEC = 0, + FLOW_OFFLOAD_XMIT_NEIGH, + FLOW_OFFLOAD_XMIT_XFRM, + FLOW_OFFLOAD_XMIT_DIRECT, + FLOW_OFFLOAD_XMIT_TC, +}; + struct flow_offload_tuple { union { struct in_addr src_v4; @@ -111,6 +119,14 @@ struct flow_offload_tuple { u16 mtu; struct dst_entry *dst_cache; + + /* fix conflicting upstream commit db6140e5e35a48405e669353bd54042c1d4c3841 */ + u8 xmit_type; + union { + struct { + u32 iifidx; + } tc; + }; }; struct flow_offload_tuple_rhash { diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index b6421a8..e41b5c5 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -103,7 +103,11 @@ static int nf_flow_rule_match(struct nf_flow_match *match, nf_flow_rule_lwt_match(match, tun_info); } - key->meta.ingress_ifindex = tuple->iifidx; + if (tuple->xmit_type == FLOW_OFFLOAD_XMIT_TC) + key->meta.ingress_ifindex = tuple->tc.iifidx; + else + key->meta.ingress_ifindex = tuple->iifidx; + mask->meta.ingress_ifindex = 0xffffffff; switch (tuple->l3proto) { diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index a54ba2e..ed310be 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -356,6 +356,13 @@ static void tcf_ct_flow_table_put(struct tcf_ct_params *params) } } +static void tcf_ct_flow_tc_ifidx(struct flow_offload *entry, + struct nf_conn_act_ct_ext *act_ct_ext, u8 dir) +{ + entry->tuplehash[dir].tuple.xmit_type = FLOW_OFFLOAD_XMIT_TC; + entry->tuplehash[dir].tuple.tc.iifidx = act_ct_ext->ifindex[dir]; +} + static void tcf_ct_flow_table_add(struct tcf_ct_flow_table *ct_ft, struct nf_conn *ct, bool tcp) @@ -380,10 +387,8 @@ static void tcf_ct_flow_table_add(struct tcf_ct_flow_table *ct_ft, act_ct_ext = nf_conn_act_ct_ext_find(ct); if (act_ct_ext) { - entry->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.iifidx = - act_ct_ext->ifindex[IP_CT_DIR_ORIGINAL]; - entry->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.iifidx = - act_ct_ext->ifindex[IP_CT_DIR_REPLY]; + tcf_ct_flow_tc_ifidx(entry, act_ct_ext, FLOW_OFFLOAD_DIR_ORIGINAL); + tcf_ct_flow_tc_ifidx(entry, act_ct_ext, FLOW_OFFLOAD_DIR_REPLY); } err = flow_offload_add(&ct_ft->nf_ft, entry); -- 1.8.3.1 -- kernel-team mailing list <a href="mailto:kernel-team@lists.ubuntu.com" target="_blank">kernel-team@lists.ubuntu.com</a> <a href="https://lists.ubuntu.com/mailman/listinfo/kernel-team" rel="noreferrer" target="_blank">https://lists.ubuntu.com/mailman/listinfo/kernel-team</a> </blockquote></div></div>