| Message ID | 20220829094412.554018-3-wojciech.drewek@intel.com (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | ice: L2TPv3 offload support | expand |
On Mon, Aug 29, 2022 at 11:44:09AM +0200, Wojciech Drewek wrote: > Allow to dissect L2TPv3 specific field which is: > - session ID (32 bits) > > L2TPv3 might be transported over IP or over UDP, > this ipmplementation is only about L2TPv3 over IP. > IP protocold carries L2TPv3 when ip_proto is s/protocold/protocol/ > +static void __skb_flow_dissect_l2tpv3(const struct sk_buff *skb, > + struct flow_dissector *flow_dissector, > + void *target_container, const void *data, > + int nhoff, int hlen) > +{ > + struct flow_dissector_key_l2tpv3 *key_l2tpv3; > + struct { > + __be32 session_id; > + } *hdr, _hdr; > + > + hdr = __skb_header_pointer(skb, nhoff, sizeof(_hdr), data, hlen, &_hdr); > + if (!hdr) > + return; > + > + if (!dissector_uses_key(flow_dissector, FLOW_DISSECTOR_KEY_L2TPV3)) > + return; I'd find it more logical to test !dissector_uses_key() first, then call __skb_header_pointer(). But that probably just a personnal preference. In any case the code looks good to me. Acked-by: Guillaume Nault <gnault@redhat.com>
Hi Guillaume, Thanks for the review! > -----Original Message----- > From: Guillaume Nault <gnault@redhat.com> > Sent: środa, 31 sierpnia 2022 12:07 > To: Drewek, Wojciech <wojciech.drewek@intel.com> > Cc: netdev@vger.kernel.org; Lobakin, Alexandr <alexandr.lobakin@intel.com>; Brandeburg, Jesse <jesse.brandeburg@intel.com>; > Nguyen, Anthony L <anthony.l.nguyen@intel.com>; davem@davemloft.net; edumazet@google.com; kuba@kernel.org; > pabeni@redhat.com; jhs@mojatatu.com; xiyou.wangcong@gmail.com; jiri@resnulli.us; marcin.szycik@linux.intel.com; > michal.swiatkowski@linux.intel.com; kurt@linutronix.de; boris.sukholitko@broadcom.com; vladbu@nvidia.com; > komachi.yoshiki@gmail.com; paulb@nvidia.com; baowen.zheng@corigine.com; louis.peens@corigine.com; > simon.horman@corigine.com; pablo@netfilter.org; maksym.glubokiy@plvision.eu; intel-wired-lan@lists.osuosl.org; > jchapman@katalix.com > Subject: Re: [RFC PATCH net-next v2 2/5] flow_dissector: Add L2TPv3 dissectors > > On Mon, Aug 29, 2022 at 11:44:09AM +0200, Wojciech Drewek wrote: > > Allow to dissect L2TPv3 specific field which is: > > - session ID (32 bits) > > > > L2TPv3 might be transported over IP or over UDP, > > this ipmplementation is only about L2TPv3 over IP. > > IP protocold carries L2TPv3 when ip_proto is > > s/protocold/protocol/ > > > +static void __skb_flow_dissect_l2tpv3(const struct sk_buff *skb, > > + struct flow_dissector *flow_dissector, > > + void *target_container, const void *data, > > + int nhoff, int hlen) > > +{ > > + struct flow_dissector_key_l2tpv3 *key_l2tpv3; > > + struct { > > + __be32 session_id; > > + } *hdr, _hdr; > > + > > + hdr = __skb_header_pointer(skb, nhoff, sizeof(_hdr), data, hlen, &_hdr); > > + if (!hdr) > > + return; > > + > > + if (!dissector_uses_key(flow_dissector, FLOW_DISSECTOR_KEY_L2TPV3)) > > + return; > > I'd find it more logical to test !dissector_uses_key() first, then call > __skb_header_pointer(). But that probably just a personnal preference. Makes sense to me, I'll change that in the next version. > > In any case the code looks good to me. > > Acked-by: Guillaume Nault <gnault@redhat.com>
diff --git a/include/net/flow_dissector.h b/include/net/flow_dissector.h index 6c74812d64b2..5ccf52ef8809 100644 --- a/include/net/flow_dissector.h +++ b/include/net/flow_dissector.h @@ -289,6 +289,14 @@ struct flow_dissector_key_pppoe { __be16 type; }; +/** + * struct flow_dissector_key_l2tpv3: + * @session_id: identifier for a l2tp session + */ +struct flow_dissector_key_l2tpv3 { + __be32 session_id; +}; + enum flow_dissector_key_id { FLOW_DISSECTOR_KEY_CONTROL, /* struct flow_dissector_key_control */ FLOW_DISSECTOR_KEY_BASIC, /* struct flow_dissector_key_basic */ @@ -320,6 +328,7 @@ enum flow_dissector_key_id { FLOW_DISSECTOR_KEY_HASH, /* struct flow_dissector_key_hash */ FLOW_DISSECTOR_KEY_NUM_OF_VLANS, /* struct flow_dissector_key_num_of_vlans */ FLOW_DISSECTOR_KEY_PPPOE, /* struct flow_dissector_key_pppoe */ + FLOW_DISSECTOR_KEY_L2TPV3, /* struct flow_dissector_key_l2tpv3 */ FLOW_DISSECTOR_KEY_MAX, }; diff --git a/net/core/flow_dissector.c b/net/core/flow_dissector.c index 764c4cb3fe8f..dbd944de4129 100644 --- a/net/core/flow_dissector.c +++ b/net/core/flow_dissector.c @@ -204,6 +204,30 @@ static void __skb_flow_dissect_icmp(const struct sk_buff *skb, skb_flow_get_icmp_tci(skb, key_icmp, data, thoff, hlen); } +static void __skb_flow_dissect_l2tpv3(const struct sk_buff *skb, + struct flow_dissector *flow_dissector, + void *target_container, const void *data, + int nhoff, int hlen) +{ + struct flow_dissector_key_l2tpv3 *key_l2tpv3; + struct { + __be32 session_id; + } *hdr, _hdr; + + hdr = __skb_header_pointer(skb, nhoff, sizeof(_hdr), data, hlen, &_hdr); + if (!hdr) + return; + + if (!dissector_uses_key(flow_dissector, FLOW_DISSECTOR_KEY_L2TPV3)) + return; + + key_l2tpv3 = skb_flow_dissector_target(flow_dissector, + FLOW_DISSECTOR_KEY_L2TPV3, + target_container); + + key_l2tpv3->session_id = hdr->session_id; +} + void skb_flow_dissect_meta(const struct sk_buff *skb, struct flow_dissector *flow_dissector, void *target_container) @@ -1497,6 +1521,10 @@ bool __skb_flow_dissect(const struct net *net, __skb_flow_dissect_icmp(skb, flow_dissector, target_container, data, nhoff, hlen); break; + case IPPROTO_L2TP: + __skb_flow_dissect_l2tpv3(skb, flow_dissector, target_container, + data, nhoff, hlen); + break; default: break;
Allow to dissect L2TPv3 specific field which is: - session ID (32 bits) L2TPv3 might be transported over IP or over UDP, this ipmplementation is only about L2TPv3 over IP. IP protocold carries L2TPv3 when ip_proto is IPPROTO_L2TP (115). Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com> --- include/net/flow_dissector.h | 9 +++++++++ net/core/flow_dissector.c | 28 ++++++++++++++++++++++++++++ 2 files changed, 37 insertions(+)