February 2022 - osmocom-net-gprs

OsmoDevCon 2022 ?

by Harald Welte

Dear fellow Osmocom developers, as you all know, we've sadly had to skip OsmoDevCon 2020 and 2021, trying to compensate it at least to some extent with our OsmoDevCall every two weeks. The COVID-19 pandemic is far from over, and we don't know what the upcoming winter season will bring. Nevertheless, I think it would be a good idea to start a discussion of whether we should plan for an OsmoDevCon in 2022. I personally would say let's plan for the usual late April 2022 time frame, and if the pandemic situation deteriorates, we can still cancel it with something like one month lead time. I would also personally suggest to limit attendance to people who are fully vaccinated, and in addition do a self-test for all participants every morning. In terms of venue, we might also consider to move to a venue that allows better ventilation. Irrespective of the above we can also bring the air filters from the sysmocom office. So with that as an input statement, I would like to hear your opinion on the above proposals. Who would want to attend? Any complaints against the "vaccinated only plus daily self-tests in the morning" approach? Regards, Harald -- - Harald Welte <laforge(a)osmocom.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

2 years, 1 month

5
7
0 0

[PATCH net-next v9 0/7] ice: GTP support in switchdev

by Marcin Szycik

Add support for adding GTP-C and GTP-U filters in switchdev mode. To create a filter for GTP, create a GTP-type netdev with ip tool, enable hardware offload, add qdisc and add a filter in tc: ip link add $GTP0 type gtp role <sgsn/ggsn> hsize <hsize> ethtool -K $PF0 hw-tc-offload on tc qdisc add dev $GTP0 ingress tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \ action mirred egress redirect dev $VF1_PR By default, a filter for GTP-U will be added. To add a filter for GTP-C, specify enc_dst_port = 2123, e.g.: tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \ enc_dst_port 2123 action mirred egress redirect dev $VF1_PR Note: outer IPv6 offload is not supported yet. Note: GTP-U with no payload offload is not supported yet. ICE COMMS package is required to create a filter as it contains GTP profiles. Changes in iproute2 [1] are required to be able to add GTP netdev and use GTP-specific options (QFI and PDU type). [1] https://lore.kernel.org/netdev/20220211182902.11542-1-wojciech.drewek@intel… --- v2: Add more CC v3: Fix mail thread, sorry for spam v4: Add GTP echo response in gtp module v5: Change patch order v6: Add GTP echo request in gtp module v7: Fix kernel-docs in ice v8: Remove handling of GTP Echo Response v9: Add sending of multicast message on GTP Echo Response, fix GTP-C dummy packet selection Marcin Szycik (1): ice: Support GTP-U and GTP-C offload in switchdev Michal Swiatkowski (1): ice: Fix FV offset searching Wojciech Drewek (5): gtp: Allow to create GTP device without FDs gtp: Implement GTP echo response gtp: Implement GTP echo request net/sched: Allow flower to match on GTP options gtp: Add support for checking GTP device type drivers/net/ethernet/intel/ice/ice.h | 1 + .../net/ethernet/intel/ice/ice_flex_pipe.c | 52 +- .../net/ethernet/intel/ice/ice_flex_pipe.h | 2 +- .../net/ethernet/intel/ice/ice_flex_type.h | 6 +- .../ethernet/intel/ice/ice_protocol_type.h | 19 + drivers/net/ethernet/intel/ice/ice_switch.c | 643 ++++++++++++++++-- drivers/net/ethernet/intel/ice/ice_switch.h | 9 + drivers/net/ethernet/intel/ice/ice_tc_lib.c | 105 ++- drivers/net/ethernet/intel/ice/ice_tc_lib.h | 3 + drivers/net/gtp.c | 549 +++++++++++++-- include/net/gtp.h | 42 ++ include/uapi/linux/gtp.h | 1 + include/uapi/linux/if_link.h | 2 + include/uapi/linux/if_tunnel.h | 4 +- include/uapi/linux/pkt_cls.h | 15 + net/sched/cls_flower.c | 116 ++++ 16 files changed, 1456 insertions(+), 113 deletions(-) -- 2.35.1

2 years, 1 month

2
10
0 0

[PATCH net-next v7 0/7] ice: GTP support in switchdev

by Marcin Szycik

Add support for adding GTP-C and GTP-U filters in switchdev mode. To create a filter for GTP, create a GTP-type netdev with ip tool, enable hardware offload, add qdisc and add a filter in tc: ip link add $GTP0 type gtp role <sgsn/ggsn> hsize <hsize> ethtool -K $PF0 hw-tc-offload on tc qdisc add dev $GTP0 ingress tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \ action mirred egress redirect dev $VF1_PR By default, a filter for GTP-U will be added. To add a filter for GTP-C, specify enc_dst_port = 2123, e.g.: tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \ enc_dst_port 2123 action mirred egress redirect dev $VF1_PR Note: outer IPv6 offload is not supported yet. Note: GTP-U with no payload offload is not supported yet. ICE COMMS package is required to create a filter as it contains GTP profiles. Changes in iproute2 [1] are required to be able to add GTP netdev and use GTP-specific options (QFI and PDU type). [1] https://lore.kernel.org/netdev/20220211182902.11542-1-wojciech.drewek@intel… --- v2: Adding more CC v3: Fixed mail thread, sorry for spam v4: Added GTP echo response in gtp module v5: Change patch order v6: Added GTP echo request in gtp module v7: Fix kernel-docs in ice Marcin Szycik (1): ice: Support GTP-U and GTP-C offload in switchdev Michal Swiatkowski (1): ice: Fix FV offset searching Wojciech Drewek (5): gtp: Allow to create GTP device without FDs gtp: Implement GTP echo response gtp: Implement GTP echo request net/sched: Allow flower to match on GTP options gtp: Add support for checking GTP device type drivers/net/ethernet/intel/ice/ice.h | 1 + .../net/ethernet/intel/ice/ice_flex_pipe.c | 52 +- .../net/ethernet/intel/ice/ice_flex_pipe.h | 2 +- .../net/ethernet/intel/ice/ice_flex_type.h | 6 +- .../ethernet/intel/ice/ice_protocol_type.h | 19 + drivers/net/ethernet/intel/ice/ice_switch.c | 630 +++++++++++++++-- drivers/net/ethernet/intel/ice/ice_switch.h | 9 + drivers/net/ethernet/intel/ice/ice_tc_lib.c | 105 ++- drivers/net/ethernet/intel/ice/ice_tc_lib.h | 3 + drivers/net/gtp.c | 661 +++++++++++++++++- include/net/gtp.h | 42 ++ include/uapi/linux/gtp.h | 2 + include/uapi/linux/if_link.h | 2 + include/uapi/linux/if_tunnel.h | 4 +- include/uapi/linux/pkt_cls.h | 15 + net/sched/cls_flower.c | 116 +++ 16 files changed, 1565 insertions(+), 104 deletions(-) -- 2.35.1

2 years, 1 month

3
11
0 0

[PATCH net-next v8 0/7] ice: GTP support in switchdev

by Marcin Szycik

Add support for adding GTP-C and GTP-U filters in switchdev mode. To create a filter for GTP, create a GTP-type netdev with ip tool, enable hardware offload, add qdisc and add a filter in tc: ip link add $GTP0 type gtp role <sgsn/ggsn> hsize <hsize> ethtool -K $PF0 hw-tc-offload on tc qdisc add dev $GTP0 ingress tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \ action mirred egress redirect dev $VF1_PR By default, a filter for GTP-U will be added. To add a filter for GTP-C, specify enc_dst_port = 2123, e.g.: tc filter add dev $GTP0 ingress prio 1 flower enc_key_id 1337 \ enc_dst_port 2123 action mirred egress redirect dev $VF1_PR Note: outer IPv6 offload is not supported yet. Note: GTP-U with no payload offload is not supported yet. ICE COMMS package is required to create a filter as it contains GTP profiles. Changes in iproute2 [1] are required to be able to add GTP netdev and use GTP-specific options (QFI and PDU type). [1] https://lore.kernel.org/netdev/20220211182902.11542-1-wojciech.drewek@intel… --- v2: Adding more CC v3: Fixed mail thread, sorry for spam v4: Added GTP echo response in gtp module v5: Change patch order v6: Added GTP echo request in gtp module v7: Fix kernel-docs in ice v8: Remove handling of GTP Echo Response Marcin Szycik (1): ice: Support GTP-U and GTP-C offload in switchdev Michal Swiatkowski (1): ice: Fix FV offset searching Wojciech Drewek (5): gtp: Allow to create GTP device without FDs gtp: Implement GTP echo response gtp: Implement GTP echo request net/sched: Allow flower to match on GTP options gtp: Add support for checking GTP device type drivers/net/ethernet/intel/ice/ice.h | 1 + .../net/ethernet/intel/ice/ice_flex_pipe.c | 52 +- .../net/ethernet/intel/ice/ice_flex_pipe.h | 2 +- .../net/ethernet/intel/ice/ice_flex_type.h | 6 +- .../ethernet/intel/ice/ice_protocol_type.h | 19 + drivers/net/ethernet/intel/ice/ice_switch.c | 630 ++++++++++++++++-- drivers/net/ethernet/intel/ice/ice_switch.h | 9 + drivers/net/ethernet/intel/ice/ice_tc_lib.c | 105 ++- drivers/net/ethernet/intel/ice/ice_tc_lib.h | 3 + drivers/net/gtp.c | 426 +++++++++++- include/net/gtp.h | 42 ++ include/uapi/linux/gtp.h | 1 + include/uapi/linux/if_link.h | 2 + include/uapi/linux/if_tunnel.h | 4 +- include/uapi/linux/pkt_cls.h | 15 + net/sched/cls_flower.c | 116 ++++ 16 files changed, 1330 insertions(+), 103 deletions(-) -- 2.35.1

2 years, 1 month

1
7
0 0

[PATCH net-next v6 5/7] gtp: Add support for checking GTP device type

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Add a function that checks if a net device type is GTP. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> Reviewed-by: Harald Welte <laforge(a)gnumonks.org> --- include/net/gtp.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/net/gtp.h b/include/net/gtp.h index 23c2aaae8a42..c1d6169df331 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -63,6 +63,12 @@ struct gtp_pdu_session_info { /* According to 3GPP TS 38.415. */ u8 qfi; }; +static inline bool netif_is_gtp(const struct net_device *dev) +{ + return dev->rtnl_link_ops && + !strcmp(dev->rtnl_link_ops->kind, "gtp"); +} + #define GTP1_F_NPDU 0x01 #define GTP1_F_SEQ 0x02 #define GTP1_F_EXTHDR 0x04 -- 2.31.1

2 years, 2 months

1
0
0 0

[PATCH net-next v6 4/7] net/sched: Allow flower to match on GTP options

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Options are as follows: PDU_TYPE:QFI and they refernce to the fields from the PDU Session Protocol. PDU Session data is conveyed in GTP-U Extension Header. GTP-U Extension Header is described in 3GPP TS 29.281. PDU Session Protocol is described in 3GPP TS 38.415. PDU_TYPE - indicates the type of the PDU Session Information (4 bits) QFI - QoS Flow Identifier (6 bits) # ip link add gtp_dev type gtp role sgsn # tc qdisc add dev gtp_dev ingress # tc filter add dev gtp_dev protocol ip parent ffff: \ flower \ enc_key_id 11 \ gtp_opts 1:8/ff:ff \ action mirred egress redirect dev eth0 Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- include/net/gtp.h | 5 ++ include/uapi/linux/if_tunnel.h | 4 +- include/uapi/linux/pkt_cls.h | 15 +++++ net/sched/cls_flower.c | 116 +++++++++++++++++++++++++++++++++ 4 files changed, 139 insertions(+), 1 deletion(-) diff --git a/include/net/gtp.h b/include/net/gtp.h index 0e12c37f2958..23c2aaae8a42 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -58,6 +58,11 @@ struct gtp1u_packet { struct gtp_ie ie; } __packed; +struct gtp_pdu_session_info { /* According to 3GPP TS 38.415. */ + u8 pdu_type; + u8 qfi; +}; + #define GTP1_F_NPDU 0x01 #define GTP1_F_SEQ 0x02 #define GTP1_F_EXTHDR 0x04 diff --git a/include/uapi/linux/if_tunnel.h b/include/uapi/linux/if_tunnel.h index 7d9105533c7b..102119628ff5 100644 --- a/include/uapi/linux/if_tunnel.h +++ b/include/uapi/linux/if_tunnel.h @@ -176,8 +176,10 @@ enum { #define TUNNEL_VXLAN_OPT __cpu_to_be16(0x1000) #define TUNNEL_NOCACHE __cpu_to_be16(0x2000) #define TUNNEL_ERSPAN_OPT __cpu_to_be16(0x4000) +#define TUNNEL_GTP_OPT __cpu_to_be16(0x8000) #define TUNNEL_OPTIONS_PRESENT \ - (TUNNEL_GENEVE_OPT | TUNNEL_VXLAN_OPT | TUNNEL_ERSPAN_OPT) + (TUNNEL_GENEVE_OPT | TUNNEL_VXLAN_OPT | TUNNEL_ERSPAN_OPT | \ + TUNNEL_GTP_OPT) #endif /* _UAPI_IF_TUNNEL_H_ */ diff --git a/include/uapi/linux/pkt_cls.h b/include/uapi/linux/pkt_cls.h index ee38b35c3f57..404f97fb239c 100644 --- a/include/uapi/linux/pkt_cls.h +++ b/include/uapi/linux/pkt_cls.h @@ -616,6 +616,10 @@ enum { * TCA_FLOWER_KEY_ENC_OPT_ERSPAN_ * attributes */ + TCA_FLOWER_KEY_ENC_OPTS_GTP, /* Nested + * TCA_FLOWER_KEY_ENC_OPT_GTP_ + * attributes + */ __TCA_FLOWER_KEY_ENC_OPTS_MAX, }; @@ -654,6 +658,17 @@ enum { #define TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX \ (__TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX - 1) +enum { + TCA_FLOWER_KEY_ENC_OPT_GTP_UNSPEC, + TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE, /* u8 */ + TCA_FLOWER_KEY_ENC_OPT_GTP_QFI, /* u8 */ + + __TCA_FLOWER_KEY_ENC_OPT_GTP_MAX, +}; + +#define TCA_FLOWER_KEY_ENC_OPT_GTP_MAX \ + (__TCA_FLOWER_KEY_ENC_OPT_GTP_MAX - 1) + enum { TCA_FLOWER_KEY_MPLS_OPTS_UNSPEC, TCA_FLOWER_KEY_MPLS_OPTS_LSE, diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c index 1a9b1f140f9e..c80fc49c0da1 100644 --- a/net/sched/cls_flower.c +++ b/net/sched/cls_flower.c @@ -25,6 +25,7 @@ #include <net/geneve.h> #include <net/vxlan.h> #include <net/erspan.h> +#include <net/gtp.h> #include <net/dst.h> #include <net/dst_metadata.h> @@ -723,6 +724,7 @@ enc_opts_policy[TCA_FLOWER_KEY_ENC_OPTS_MAX + 1] = { [TCA_FLOWER_KEY_ENC_OPTS_GENEVE] = { .type = NLA_NESTED }, [TCA_FLOWER_KEY_ENC_OPTS_VXLAN] = { .type = NLA_NESTED }, [TCA_FLOWER_KEY_ENC_OPTS_ERSPAN] = { .type = NLA_NESTED }, + [TCA_FLOWER_KEY_ENC_OPTS_GTP] = { .type = NLA_NESTED }, }; static const struct nla_policy @@ -746,6 +748,12 @@ erspan_opt_policy[TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX + 1] = { [TCA_FLOWER_KEY_ENC_OPT_ERSPAN_HWID] = { .type = NLA_U8 }, }; +static const struct nla_policy +gtp_opt_policy[TCA_FLOWER_KEY_ENC_OPT_GTP_MAX + 1] = { + [TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE] = { .type = NLA_U8 }, + [TCA_FLOWER_KEY_ENC_OPT_GTP_QFI] = { .type = NLA_U8 }, +}; + static const struct nla_policy mpls_stack_entry_policy[TCA_FLOWER_KEY_MPLS_OPT_LSE_MAX + 1] = { [TCA_FLOWER_KEY_MPLS_OPT_LSE_DEPTH] = { .type = NLA_U8 }, @@ -1262,6 +1270,49 @@ static int fl_set_erspan_opt(const struct nlattr *nla, struct fl_flow_key *key, return sizeof(*md); } +static int fl_set_gtp_opt(const struct nlattr *nla, struct fl_flow_key *key, + int depth, int option_len, + struct netlink_ext_ack *extack) +{ + struct nlattr *tb[TCA_FLOWER_KEY_ENC_OPT_GTP_MAX + 1]; + struct gtp_pdu_session_info *sinfo; + u8 len = key->enc_opts.len; + int err; + + sinfo = (struct gtp_pdu_session_info *)&key->enc_opts.data[len]; + memset(sinfo, 0xff, option_len); + + if (!depth) + return sizeof(*sinfo); + + if (nla_type(nla) != TCA_FLOWER_KEY_ENC_OPTS_GTP) { + NL_SET_ERR_MSG_MOD(extack, "Non-gtp option type for mask"); + return -EINVAL; + } + + err = nla_parse_nested(tb, TCA_FLOWER_KEY_ENC_OPT_GTP_MAX, nla, + gtp_opt_policy, extack); + if (err < 0) + return err; + + if (!option_len && + (!tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE] || + !tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI])) { + NL_SET_ERR_MSG_MOD(extack, + "Missing tunnel key gtp option pdu type or qfi"); + return -EINVAL; + } + + if (tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE]) + sinfo->pdu_type = + nla_get_u8(tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE]); + + if (tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI]) + sinfo->qfi = nla_get_u8(tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI]); + + return sizeof(*sinfo); +} + static int fl_set_enc_opt(struct nlattr **tb, struct fl_flow_key *key, struct fl_flow_key *mask, struct netlink_ext_ack *extack) @@ -1386,6 +1437,38 @@ static int fl_set_enc_opt(struct nlattr **tb, struct fl_flow_key *key, return -EINVAL; } break; + case TCA_FLOWER_KEY_ENC_OPTS_GTP: + if (key->enc_opts.dst_opt_type) { + NL_SET_ERR_MSG_MOD(extack, + "Duplicate type for gtp options"); + return -EINVAL; + } + option_len = 0; + key->enc_opts.dst_opt_type = TUNNEL_GTP_OPT; + option_len = fl_set_gtp_opt(nla_opt_key, key, + key_depth, option_len, + extack); + if (option_len < 0) + return option_len; + + key->enc_opts.len += option_len; + /* At the same time we need to parse through the mask + * in order to verify exact and mask attribute lengths. + */ + mask->enc_opts.dst_opt_type = TUNNEL_GTP_OPT; + option_len = fl_set_gtp_opt(nla_opt_msk, mask, + msk_depth, option_len, + extack); + if (option_len < 0) + return option_len; + + mask->enc_opts.len += option_len; + if (key->enc_opts.len != mask->enc_opts.len) { + NL_SET_ERR_MSG_MOD(extack, + "Key and mask miss aligned"); + return -EINVAL; + } + break; default: NL_SET_ERR_MSG(extack, "Unknown tunnel option type"); return -EINVAL; @@ -2761,6 +2844,34 @@ static int fl_dump_key_erspan_opt(struct sk_buff *skb, return -EMSGSIZE; } +static int fl_dump_key_gtp_opt(struct sk_buff *skb, + struct flow_dissector_key_enc_opts *enc_opts) + +{ + struct gtp_pdu_session_info *session_info; + struct nlattr *nest; + + nest = nla_nest_start_noflag(skb, TCA_FLOWER_KEY_ENC_OPTS_GTP); + if (!nest) + goto nla_put_failure; + + session_info = (struct gtp_pdu_session_info *)&enc_opts->data[0]; + + if (nla_put_u8(skb, TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE, + session_info->pdu_type)) + goto nla_put_failure; + + if (nla_put_u8(skb, TCA_FLOWER_KEY_ENC_OPT_GTP_QFI, session_info->qfi)) + goto nla_put_failure; + + nla_nest_end(skb, nest); + return 0; + +nla_put_failure: + nla_nest_cancel(skb, nest); + return -EMSGSIZE; +} + static int fl_dump_key_ct(struct sk_buff *skb, struct flow_dissector_key_ct *key, struct flow_dissector_key_ct *mask) @@ -2824,6 +2935,11 @@ static int fl_dump_key_options(struct sk_buff *skb, int enc_opt_type, if (err) goto nla_put_failure; break; + case TUNNEL_GTP_OPT: + err = fl_dump_key_gtp_opt(skb, enc_opts); + if (err) + goto nla_put_failure; + break; default: goto nla_put_failure; } -- 2.31.1

2 years, 2 months

1
0
0 0

[PATCH net-next v6 3/7] gtp: Implement GTP echo request

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Adding GTP device through ip link creates the situation where GTP instance is not able to send GTP echo requests. Echo requests are used to check if GTP peer is still alive. With this patch, gtp_genl_ops are extended by new cmd (GTP_CMD_ECHOREQ) which allows to send echo request in the given version of GTP protocol (v0 or v1), from the given ms address to he given peer. TID is not inclued because in all path management messages it should be equal to 0. All echo requests are stored in echo_hash table with the flag (replied) which indicates if GTP echo response was recived in response to this request. Userspace can see all echo requests using GTP_CMD_ECHOREQ dumpit callback. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> Suggested-by: Harald Welte <laforge(a)gnumonks.org> --- drivers/net/gtp.c | 402 ++++++++++++++++++++++++++++++++++++--- include/uapi/linux/gtp.h | 2 + 2 files changed, 376 insertions(+), 28 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 5ed24fa9d5b2..14e9f8053d71 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -60,6 +60,19 @@ struct pdp_ctx { struct rcu_head rcu_head; }; +struct gtp_echo { + struct hlist_node hlist; + + struct in_addr ms_addr_ip4; + struct in_addr peer_addr_ip4; + + u8 replied; + u8 version; + int ifindex; + + struct rcu_head rcu_head; +}; + /* One instance of the GTP device. */ struct gtp_dev { struct list_head list; @@ -75,6 +88,7 @@ struct gtp_dev { unsigned int hash_size; struct hlist_head *tid_hash; struct hlist_head *addr_hash; + struct hlist_head *echo_hash; u8 restart_count; }; @@ -89,6 +103,19 @@ static u32 gtp_h_initval; static void pdp_context_delete(struct pdp_ctx *pctx); +static void gtp_echo_context_free(struct rcu_head *head) +{ + struct gtp_echo *echo = container_of(head, struct gtp_echo, rcu_head); + + kfree(echo); +} + +static void gtp_echo_delete(struct gtp_echo *echo) +{ + hlist_del_rcu(&echo->hlist); + call_rcu(&echo->rcu_head, gtp_echo_context_free); +} + static inline u32 gtp0_hashfn(u64 tid) { u32 *tid32 = (u32 *) &tid; @@ -154,6 +181,24 @@ static struct pdp_ctx *ipv4_pdp_find(struct gtp_dev *gtp, __be32 ms_addr) return NULL; } +static struct gtp_echo *gtp_find_echo(struct gtp_dev *gtp, __be32 ms_addr, + __be32 peer_addr, unsigned int version) +{ + struct hlist_head *head; + struct gtp_echo *echo; + + head = &gtp->echo_hash[ipv4_hashfn(ms_addr) % gtp->hash_size]; + + hlist_for_each_entry_rcu(echo, head, hlist) { + if (echo->ms_addr_ip4.s_addr == ms_addr && + echo->peer_addr_ip4.s_addr == peer_addr && + echo->version == version) + return echo; + } + + return NULL; +} + static bool gtp_check_ms_ipv4(struct sk_buff *skb, struct pdp_ctx *pctx, unsigned int hdrlen, unsigned int role) { @@ -243,12 +288,34 @@ static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, * by the receiver * Returns true if the echo req was correct, false otherwise. */ -static bool gtp0_validate_echo_req(struct gtp0_header *gtp0) +static bool gtp0_validate_echo_hdr(struct gtp0_header *gtp0) { return !(gtp0->tid || (gtp0->flags ^ 0x1e) || gtp0->number != 0xff || gtp0->flow); } +/* msg_type has to be GTP_ECHO_REQ or GTP_ECHO_RSP */ +static void gtp0_build_echo_msg(struct gtp0_header *hdr, __u8 msg_type) +{ + hdr->flags = 0x1e; /* v0, GTP-non-prime. */ + hdr->type = msg_type; + /* GSM TS 09.60. 7.3 In all Path Management Flow Label and TID + * are not used and shall be set to 0. + */ + hdr->flow = 0; + hdr->tid = 0; + hdr->number = 0xff; + hdr->spare[0] = 0xff; + hdr->spare[1] = 0xff; + hdr->spare[2] = 0xff; + + if (msg_type == GTP_ECHO_RSP) + hdr->length = + htons(sizeof(struct gtp0_packet) - sizeof(struct gtp0_header)); + else + hdr->length = 0; +} + static int gtp0_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) { struct gtp0_packet *gtp_pkt; @@ -260,7 +327,7 @@ static int gtp0_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) gtp0 = (struct gtp0_header *)(skb->data + sizeof(struct udphdr)); - if (!gtp0_validate_echo_req(gtp0)) + if (!gtp0_validate_echo_hdr(gtp0)) return -1; seq = gtp0->seq; @@ -271,10 +338,7 @@ static int gtp0_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) gtp_pkt = skb_push(skb, sizeof(struct gtp0_packet)); memset(gtp_pkt, 0, sizeof(struct gtp0_packet)); - gtp_pkt->gtp0_h.flags = 0x1e; /* v0, GTP-non-prime. */ - gtp_pkt->gtp0_h.type = GTP_ECHO_RSP; - gtp_pkt->gtp0_h.length = - htons(sizeof(struct gtp0_packet) - sizeof(struct gtp0_header)); + gtp0_build_echo_msg(&gtp_pkt->gtp0_h, GTP_ECHO_RSP); /* GSM TS 09.60. 7.3 The Sequence Number in a signalling response * message shall be copied from the signalling request message @@ -282,16 +346,6 @@ static int gtp0_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) */ gtp_pkt->gtp0_h.seq = seq; - /* GSM TS 09.60. 7.3 In all Path Management Flow Label and TID - * are not used and shall be set to 0. - */ - gtp_pkt->gtp0_h.flow = 0; - gtp_pkt->gtp0_h.tid = 0; - gtp_pkt->gtp0_h.number = 0xff; - gtp_pkt->gtp0_h.spare[0] = 0xff; - gtp_pkt->gtp0_h.spare[1] = 0xff; - gtp_pkt->gtp0_h.spare[2] = 0xff; - gtp_pkt->ie.tag = GTPIE_RECOVERY; gtp_pkt->ie.val = gtp->restart_count; @@ -319,6 +373,31 @@ static int gtp0_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) return 0; } +static int gtp0_handle_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp0_header *gtp0; + struct gtp_echo *echo; + struct iphdr *iph; + + gtp0 = (struct gtp0_header *)(skb->data + sizeof(struct udphdr)); + + if (!gtp0_validate_echo_hdr(gtp0)) + return -1; + + iph = ip_hdr(skb); + + echo = gtp_find_echo(gtp, iph->daddr, iph->saddr, GTP_V0); + if (!echo) { + netdev_dbg(gtp->dev, "No echo request was send to %pI4, version: %u\n", + &iph->saddr, GTP_V0); + return -1; + } + + echo->replied = true; + + return 0; +} + /* 1 means pass up to the stack, -1 means drop and 0 means decapsulated. */ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { @@ -342,6 +421,9 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if (gtp0->type == GTP_ECHO_REQ && gtp->sk_created) return gtp0_send_echo_resp(gtp, skb); + if (gtp0->type == GTP_ECHO_RSP && gtp->sk_created) + return gtp0_handle_echo_resp(gtp, skb); + if (gtp0->type != GTP_TPDU) return 1; @@ -354,6 +436,27 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) return gtp_rx(pctx, skb, hdrlen, gtp->role); } +/* msg_type has to be GTP_ECHO_REQ or GTP_ECHO_RSP */ +static void gtp1u_build_echo_msg(struct gtp1_header_long *hdr, __u8 msg_type) +{ + /* S flag must be set to 1 */ + hdr->flags = 0x32; /* v1, GTP-non-prime. */ + hdr->type = msg_type; + /* 3GPP TS 29.281 5.1 - TEID has to be set to 0 */ + hdr->tid = 0; + + /* seq, npdu and next should be counted to the length of the GTP packet + * that's why szie of gtp1_header should be subtracted, + * not size of gtp1_header_long. + */ + if (msg_type == GTP_ECHO_RSP) + hdr->length = + htons(sizeof(struct gtp1u_packet) - sizeof(struct gtp1_header)); + else + hdr->length = + htons(sizeof(struct gtp1_header_long) - sizeof(struct gtp1_header)); +} + static int gtp1u_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) { struct gtp1_header_long *gtp1u; @@ -377,17 +480,7 @@ static int gtp1u_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) gtp_pkt = skb_push(skb, sizeof(struct gtp1u_packet)); memset(gtp_pkt, 0, sizeof(struct gtp1u_packet)); - /* S flag must be set to 1 */ - gtp_pkt->gtp1u_h.flags = 0x32; - gtp_pkt->gtp1u_h.type = GTP_ECHO_RSP; - /* seq, npdu and next should be counted to the length of the GTP packet - * that's why szie of gtp1_header should be subtracted, - * not why szie of gtp1_header_long. - */ - gtp_pkt->gtp1u_h.length = - htons(sizeof(struct gtp1u_packet) - sizeof(struct gtp1_header)); - /* 3GPP TS 29.281 5.1 - TEID has to be set to 0 */ - gtp_pkt->gtp1u_h.tid = 0; + gtp1u_build_echo_msg(&gtp_pkt->gtp1u_h, GTP_ECHO_RSP); /* 3GPP TS 29.281 7.7.2 - The Restart Counter value in the * Recovery information element shall not be used, i.e. it shall @@ -422,6 +515,35 @@ static int gtp1u_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) return 0; } +static int gtp1u_handle_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp1_header_long *gtp1u; + struct gtp_echo *echo; + struct iphdr *iph; + + gtp1u = (struct gtp1_header_long *)(skb->data + sizeof(struct udphdr)); + + /* 3GPP TS 29.281 5.1 - For the Echo Request, Echo Response, + * Error Indication and Supported Extension Headers Notification + * messages, the S flag shall be set to 1 and TEID shall be set to 0. + */ + if (!(gtp1u->flags & GTP1_F_SEQ) || gtp1u->tid) + return -1; + + iph = ip_hdr(skb); + + echo = gtp_find_echo(gtp, iph->daddr, iph->saddr, GTP_V1); + if (!echo) { + netdev_dbg(gtp->dev, "No echo request was send to %pI4, version: %u\n", + &iph->saddr, GTP_V1); + return -1; + } + + echo->replied = true; + + return 0; +} + static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { unsigned int hdrlen = sizeof(struct udphdr) + @@ -444,6 +566,9 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if (gtp1->type == GTP_ECHO_REQ && gtp->sk_created) return gtp1u_send_echo_resp(gtp, skb); + if (gtp1->type == GTP_ECHO_RSP && gtp->sk_created) + return gtp1u_handle_echo_resp(gtp, skb); + if (gtp1->type != GTP_TPDU) return 1; @@ -835,6 +960,7 @@ static void gtp_destructor(struct net_device *dev) kfree(gtp->addr_hash); kfree(gtp->tid_hash); + kfree(gtp->echo_hash); } static struct sock *gtp_create_sock(int type, struct gtp_dev *gtp) @@ -954,18 +1080,23 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, out_hashtable: kfree(gtp->addr_hash); kfree(gtp->tid_hash); + kfree(gtp->echo_hash); return err; } static void gtp_dellink(struct net_device *dev, struct list_head *head) { struct gtp_dev *gtp = netdev_priv(dev); + struct gtp_echo *echo; struct pdp_ctx *pctx; int i; - for (i = 0; i < gtp->hash_size; i++) + for (i = 0; i < gtp->hash_size; i++) { hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], hlist_tid) pdp_context_delete(pctx); + hlist_for_each_entry_rcu(echo, &gtp->echo_hash[i], hlist) + gtp_echo_delete(echo); + } list_del_rcu(&gtp->list); unregister_netdevice_queue(dev, head); @@ -1040,13 +1171,21 @@ static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize) if (gtp->tid_hash == NULL) goto err1; + gtp->echo_hash = kmalloc_array(hsize, sizeof(struct hlist_head), + GFP_KERNEL | __GFP_NOWARN); + if (!gtp->echo_hash) + goto err2; + gtp->hash_size = hsize; for (i = 0; i < hsize; i++) { INIT_HLIST_HEAD(&gtp->addr_hash[i]); INIT_HLIST_HEAD(&gtp->tid_hash[i]); + INIT_HLIST_HEAD(&gtp->echo_hash[i]); } return 0; +err2: + kfree(gtp->tid_hash); err1: kfree(gtp->addr_hash); return -ENOMEM; @@ -1583,6 +1722,205 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, return skb->len; } +static int gtp_add_echo(struct gtp_dev *gtp, __be32 src_ip, __be32 dst_ip, + unsigned int version) +{ + struct gtp_echo *echo; + bool found = false; + + rcu_read_lock(); + echo = gtp_find_echo(gtp, src_ip, dst_ip, version); + rcu_read_unlock(); + + if (!echo) { + echo = kmalloc(sizeof(*echo), GFP_ATOMIC); + if (!echo) + return -ENOMEM; + } else { + found = true; + } + + echo->ms_addr_ip4.s_addr = src_ip; + echo->peer_addr_ip4.s_addr = dst_ip; + echo->replied = false; + echo->version = version; + echo->ifindex = gtp->dev->ifindex; + + if (!found) { + u32 hash_ms; + + hash_ms = ipv4_hashfn(src_ip) % gtp->hash_size; + hlist_add_head_rcu(&echo->hlist, &gtp->echo_hash[hash_ms]); + } + + return 0; +} + +static int gtp_genl_send_echo_req(struct sk_buff *skb, struct genl_info *info) +{ + struct sk_buff *skb_to_send; + __be32 src_ip, dst_ip; + unsigned int version; + struct gtp_dev *gtp; + struct flowi4 fl4; + struct rtable *rt; + struct sock *sk; + __be16 port; + int len; + int err; + + if (!info->attrs[GTPA_VERSION] || + !info->attrs[GTPA_LINK] || + !info->attrs[GTPA_PEER_ADDRESS] || + !info->attrs[GTPA_MS_ADDRESS]) + return -EINVAL; + + version = nla_get_u32(info->attrs[GTPA_VERSION]); + dst_ip = nla_get_be32(info->attrs[GTPA_PEER_ADDRESS]); + src_ip = nla_get_be32(info->attrs[GTPA_MS_ADDRESS]); + + gtp = gtp_find_dev(sock_net(skb->sk), info->attrs); + if (!gtp) + return -ENODEV; + + if (!gtp->sk_created) + return -EOPNOTSUPP; + if (!(gtp->dev->flags & IFF_UP)) + return -ENETDOWN; + + if (version == GTP_V0) { + struct gtp0_header *gtp0_h; + + len = LL_RESERVED_SPACE(gtp->dev) + sizeof(struct gtp0_header) + + sizeof(struct iphdr) + sizeof(struct udphdr); + + skb_to_send = netdev_alloc_skb_ip_align(gtp->dev, len); + if (!skb_to_send) + return -ENOMEM; + + sk = gtp->sk0; + port = htons(GTP0_PORT); + + gtp0_h = skb_push(skb_to_send, sizeof(struct gtp0_header)); + memset(gtp0_h, 0, sizeof(struct gtp0_header)); + gtp0_build_echo_msg(gtp0_h, GTP_ECHO_REQ); + } else if (version == GTP_V1) { + struct gtp1_header_long *gtp1u_h; + + len = LL_RESERVED_SPACE(gtp->dev) + sizeof(struct gtp1_header_long) + + sizeof(struct iphdr) + sizeof(struct udphdr); + + skb_to_send = netdev_alloc_skb_ip_align(gtp->dev, len); + if (!skb_to_send) + return -ENOMEM; + + sk = gtp->sk1u; + port = htons(GTP1U_PORT); + + gtp1u_h = skb_push(skb_to_send, sizeof(struct gtp1_header_long)); + memset(gtp1u_h, 0, sizeof(struct gtp1_header_long)); + gtp1u_build_echo_msg(gtp1u_h, GTP_ECHO_REQ); + } else { + return -ENODEV; + } + + rt = ip4_route_output_gtp(&fl4, sk, dst_ip, src_ip); + if (IS_ERR(rt)) { + netdev_dbg(gtp->dev, "no route for echo request to %pI4\n", + &dst_ip); + kfree_skb(skb_to_send); + return -ENODEV; + } + + err = gtp_add_echo(gtp, src_ip, dst_ip, version); + if (err) + return err; + + udp_tunnel_xmit_skb(rt, sk, skb_to_send, + fl4.saddr, fl4.daddr, + fl4.flowi4_tos, + ip4_dst_hoplimit(&rt->dst), + 0, + port, port, + !net_eq(sock_net(sk), + dev_net(gtp->dev)), + false); + return 0; +} + +static int gtp_genl_fill_echo_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, + int flags, u32 type, struct gtp_echo *echo) +{ + void *genlh; + + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, + type); + if (!genlh) + goto err; + + if (nla_put_u32(skb, GTPA_VERSION, echo->version) || + nla_put_u32(skb, GTPA_LINK, echo->ifindex) || + nla_put_be32(skb, GTPA_PEER_ADDRESS, echo->peer_addr_ip4.s_addr) || + nla_put_be32(skb, GTPA_MS_ADDRESS, echo->ms_addr_ip4.s_addr) || + nla_put_u8(skb, GTPA_ECHO_REPLIED, echo->replied)) + goto err; + + genlmsg_end(skb, genlh); + return 0; + +err: + genlmsg_cancel(skb, genlh); + return -EMSGSIZE; +} + +static int gtp_genl_dump_echo(struct sk_buff *skb, + struct netlink_callback *cb) +{ + struct gtp_dev *last_gtp = (struct gtp_dev *)cb->args[2], *gtp; + int i, j, bucket = cb->args[0], skip = cb->args[1]; + struct net *net = sock_net(skb->sk); + struct gtp_echo *echo; + struct gtp_net *gn; + + gn = net_generic(net, gtp_net_id); + + if (cb->args[4]) + return 0; + + rcu_read_lock(); + list_for_each_entry_rcu(gtp, &gn->gtp_dev_list, list) { + if (last_gtp && last_gtp != gtp) + continue; + else + last_gtp = NULL; + + for (i = bucket; i < gtp->hash_size; i++) { + j = 0; + hlist_for_each_entry_rcu(echo, &gtp->echo_hash[i], + hlist) { + int ret = gtp_genl_fill_echo_info(skb, + NETLINK_CB(cb->skb).portid, + cb->nlh->nlmsg_seq, + NLM_F_MULTI, + cb->nlh->nlmsg_type, echo); + if (j >= skip && ret) { + cb->args[0] = i; + cb->args[1] = j; + cb->args[2] = (unsigned long)gtp; + goto out; + } + j++; + } + skip = 0; + } + bucket = 0; + } + cb->args[4] = 1; +out: + rcu_read_unlock(); + return skb->len; +} + static const struct nla_policy gtp_genl_policy[GTPA_MAX + 1] = { [GTPA_LINK] = { .type = NLA_U32, }, [GTPA_VERSION] = { .type = NLA_U32, }, @@ -1593,6 +1931,7 @@ static const struct nla_policy gtp_genl_policy[GTPA_MAX + 1] = { [GTPA_NET_NS_FD] = { .type = NLA_U32, }, [GTPA_I_TEI] = { .type = NLA_U32, }, [GTPA_O_TEI] = { .type = NLA_U32, }, + [GTPA_ECHO_REPLIED] = { .type = NLA_U8, }, }; static const struct genl_small_ops gtp_genl_ops[] = { @@ -1615,6 +1954,13 @@ static const struct genl_small_ops gtp_genl_ops[] = { .dumpit = gtp_genl_dump_pdp, .flags = GENL_ADMIN_PERM, }, + { + .cmd = GTP_CMD_ECHOREQ, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .doit = gtp_genl_send_echo_req, + .dumpit = gtp_genl_dump_echo, + .flags = GENL_ADMIN_PERM, + }, }; static struct genl_family gtp_genl_family __ro_after_init = { diff --git a/include/uapi/linux/gtp.h b/include/uapi/linux/gtp.h index 79f9191bbb24..63bb60f1e4e3 100644 --- a/include/uapi/linux/gtp.h +++ b/include/uapi/linux/gtp.h @@ -8,6 +8,7 @@ enum gtp_genl_cmds { GTP_CMD_NEWPDP, GTP_CMD_DELPDP, GTP_CMD_GETPDP, + GTP_CMD_ECHOREQ, GTP_CMD_MAX, }; @@ -29,6 +30,7 @@ enum gtp_attrs { GTPA_NET_NS_FD, GTPA_I_TEI, /* for GTPv1 only */ GTPA_O_TEI, /* for GTPv1 only */ + GTPA_ECHO_REPLIED, GTPA_PAD, __GTPA_MAX, }; -- 2.31.1

2 years, 2 months

1
0
0 0

[PATCH net-next v6 2/7] gtp: Implement GTP echo response

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Adding GTP device through ip link creates the situation where there is no userspace daemon which would handle GTP messages (Echo Request for example). GTP-U instance which would not respond to echo requests would violate GTP specification. When GTP packet arrives with GTP_ECHO_REQ message type, GTP_ECHO_RSP is send to the sender. GTP_ECHO_RSP message should contain information element with GTPIE_RECOVERY tag and restart counter value. For GTPv1 restart counter is not used and should be equal to 0, for GTPv0 restart counter contains information provided from userspace(IFLA_GTP_RESTART_COUNT). Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> Suggested-by: Harald Welte <laforge(a)gnumonks.org> Reviewed-by: Harald Welte <laforge(a)gnumonks.org> Tested-by: Harald Welte <laforge(a)gnumonks.org> --- v6: gtp0_echo_resp renamed to gtp0_send_echo_resp and gtp1u_echo_resp renamed to gtp1u_send_echo_resp --- drivers/net/gtp.c | 212 ++++++++++++++++++++++++++++++++--- include/net/gtp.h | 31 +++++ include/uapi/linux/if_link.h | 1 + 3 files changed, 228 insertions(+), 16 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 25d8521897b3..5ed24fa9d5b2 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -75,6 +75,8 @@ struct gtp_dev { unsigned int hash_size; struct hlist_head *tid_hash; struct hlist_head *addr_hash; + + u8 restart_count; }; static unsigned int gtp_net_id __read_mostly; @@ -217,6 +219,106 @@ static int gtp_rx(struct pdp_ctx *pctx, struct sk_buff *skb, return -1; } +static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, + const struct sock *sk, + __be32 daddr, __be32 saddr) +{ + memset(fl4, 0, sizeof(*fl4)); + fl4->flowi4_oif = sk->sk_bound_dev_if; + fl4->daddr = daddr; + fl4->saddr = saddr; + fl4->flowi4_tos = RT_CONN_FLAGS(sk); + fl4->flowi4_proto = sk->sk_protocol; + + return ip_route_output_key(sock_net(sk), fl4); +} + +/* GSM TS 09.60. 7.3 + * In all Path Management messages: + * - TID: is not used and shall be set to 0. + * - Flow Label is not used and shall be set to 0 + * In signalling messages: + * - number: this field is not yet used in signalling messages. + * It shall be set to 255 by the sender and shall be ignored + * by the receiver + * Returns true if the echo req was correct, false otherwise. + */ +static bool gtp0_validate_echo_req(struct gtp0_header *gtp0) +{ + return !(gtp0->tid || (gtp0->flags ^ 0x1e) || + gtp0->number != 0xff || gtp0->flow); +} + +static int gtp0_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp0_packet *gtp_pkt; + struct gtp0_header *gtp0; + struct rtable *rt; + struct flowi4 fl4; + struct iphdr *iph; + __be16 seq; + + gtp0 = (struct gtp0_header *)(skb->data + sizeof(struct udphdr)); + + if (!gtp0_validate_echo_req(gtp0)) + return -1; + + seq = gtp0->seq; + + /* pull GTP and UDP headers */ + skb_pull_data(skb, sizeof(struct gtp0_header) + sizeof(struct udphdr)); + + gtp_pkt = skb_push(skb, sizeof(struct gtp0_packet)); + memset(gtp_pkt, 0, sizeof(struct gtp0_packet)); + + gtp_pkt->gtp0_h.flags = 0x1e; /* v0, GTP-non-prime. */ + gtp_pkt->gtp0_h.type = GTP_ECHO_RSP; + gtp_pkt->gtp0_h.length = + htons(sizeof(struct gtp0_packet) - sizeof(struct gtp0_header)); + + /* GSM TS 09.60. 7.3 The Sequence Number in a signalling response + * message shall be copied from the signalling request message + * that the GSN is replying to. + */ + gtp_pkt->gtp0_h.seq = seq; + + /* GSM TS 09.60. 7.3 In all Path Management Flow Label and TID + * are not used and shall be set to 0. + */ + gtp_pkt->gtp0_h.flow = 0; + gtp_pkt->gtp0_h.tid = 0; + gtp_pkt->gtp0_h.number = 0xff; + gtp_pkt->gtp0_h.spare[0] = 0xff; + gtp_pkt->gtp0_h.spare[1] = 0xff; + gtp_pkt->gtp0_h.spare[2] = 0xff; + + gtp_pkt->ie.tag = GTPIE_RECOVERY; + gtp_pkt->ie.val = gtp->restart_count; + + iph = ip_hdr(skb); + + /* find route to the sender, + * src address becomes dst address and vice versa. + */ + rt = ip4_route_output_gtp(&fl4, gtp->sk0, iph->saddr, iph->daddr); + if (IS_ERR(rt)) { + netdev_dbg(gtp->dev, "no route for echo response from %pI4\n", + &iph->saddr); + return -1; + } + + udp_tunnel_xmit_skb(rt, gtp->sk0, skb, + fl4.saddr, fl4.daddr, + iph->tos, + ip4_dst_hoplimit(&rt->dst), + 0, + htons(GTP0_PORT), htons(GTP0_PORT), + !net_eq(sock_net(gtp->sk1u), + dev_net(gtp->dev)), + false); + return 0; +} + /* 1 means pass up to the stack, -1 means drop and 0 means decapsulated. */ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { @@ -233,6 +335,13 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if ((gtp0->flags >> 5) != GTP_V0) return 1; + /* If the sockets were created in kernel, it means that + * there is no daemon running in userspace which would + * handle echo request. + */ + if (gtp0->type == GTP_ECHO_REQ && gtp->sk_created) + return gtp0_send_echo_resp(gtp, skb); + if (gtp0->type != GTP_TPDU) return 1; @@ -245,6 +354,74 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) return gtp_rx(pctx, skb, hdrlen, gtp->role); } +static int gtp1u_send_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp1_header_long *gtp1u; + struct gtp1u_packet *gtp_pkt; + struct rtable *rt; + struct flowi4 fl4; + struct iphdr *iph; + + gtp1u = (struct gtp1_header_long *)(skb->data + sizeof(struct udphdr)); + + /* 3GPP TS 29.281 5.1 - For the Echo Request, Echo Response, + * Error Indication and Supported Extension Headers Notification + * messages, the S flag shall be set to 1 and TEID shall be set to 0. + */ + if (!(gtp1u->flags & GTP1_F_SEQ) || gtp1u->tid) + return -1; + + /* pull GTP and UDP headers */ + skb_pull_data(skb, sizeof(struct gtp1_header_long) + sizeof(struct udphdr)); + + gtp_pkt = skb_push(skb, sizeof(struct gtp1u_packet)); + memset(gtp_pkt, 0, sizeof(struct gtp1u_packet)); + + /* S flag must be set to 1 */ + gtp_pkt->gtp1u_h.flags = 0x32; + gtp_pkt->gtp1u_h.type = GTP_ECHO_RSP; + /* seq, npdu and next should be counted to the length of the GTP packet + * that's why szie of gtp1_header should be subtracted, + * not why szie of gtp1_header_long. + */ + gtp_pkt->gtp1u_h.length = + htons(sizeof(struct gtp1u_packet) - sizeof(struct gtp1_header)); + /* 3GPP TS 29.281 5.1 - TEID has to be set to 0 */ + gtp_pkt->gtp1u_h.tid = 0; + + /* 3GPP TS 29.281 7.7.2 - The Restart Counter value in the + * Recovery information element shall not be used, i.e. it shall + * be set to zero by the sender and shall be ignored by the receiver. + * The Recovery information element is mandatory due to backwards + * compatibility reasons. + */ + gtp_pkt->ie.tag = GTPIE_RECOVERY; + gtp_pkt->ie.val = 0; + + iph = ip_hdr(skb); + + /* find route to the sender, + * src address becomes dst address and vice versa. + */ + rt = ip4_route_output_gtp(&fl4, gtp->sk1u, iph->saddr, iph->daddr); + if (IS_ERR(rt)) { + netdev_dbg(gtp->dev, "no route for echo response from %pI4\n", + &iph->saddr); + return -1; + } + + udp_tunnel_xmit_skb(rt, gtp->sk1u, skb, + fl4.saddr, fl4.daddr, + iph->tos, + ip4_dst_hoplimit(&rt->dst), + 0, + htons(GTP1U_PORT), htons(GTP1U_PORT), + !net_eq(sock_net(gtp->sk1u), + dev_net(gtp->dev)), + false); + return 0; +} + static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { unsigned int hdrlen = sizeof(struct udphdr) + @@ -260,6 +437,13 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if ((gtp1->flags >> 5) != GTP_V1) return 1; + /* If the sockets were created in kernel, it means that + * there is no daemon running in userspace which would + * handle echo request. + */ + if (gtp1->type == GTP_ECHO_REQ && gtp->sk_created) + return gtp1u_send_echo_resp(gtp, skb); + if (gtp1->type != GTP_TPDU) return 1; @@ -398,20 +582,6 @@ static void gtp_dev_uninit(struct net_device *dev) free_percpu(dev->tstats); } -static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, - const struct sock *sk, - __be32 daddr) -{ - memset(fl4, 0, sizeof(*fl4)); - fl4->flowi4_oif = sk->sk_bound_dev_if; - fl4->daddr = daddr; - fl4->saddr = inet_sk(sk)->inet_saddr; - fl4->flowi4_tos = RT_CONN_FLAGS(sk); - fl4->flowi4_proto = sk->sk_protocol; - - return ip_route_output_key(sock_net(sk), fl4); -} - static inline void gtp0_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) { int payload_len = skb->len; @@ -517,7 +687,8 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, } netdev_dbg(dev, "found PDP context %p\n", pctx); - rt = ip4_route_output_gtp(&fl4, pctx->sk, pctx->peer_addr_ip4.s_addr); + rt = ip4_route_output_gtp(&fl4, pctx->sk, pctx->peer_addr_ip4.s_addr, + inet_sk(pctx->sk)->inet_saddr); if (IS_ERR(rt)) { netdev_dbg(dev, "no route to SSGN %pI4\n", &pctx->peer_addr_ip4.s_addr); @@ -746,6 +917,11 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, } gtp->role = role; + if (!data[IFLA_GTP_RESTART_COUNT]) + gtp->restart_count = 0; + else + gtp->restart_count = nla_get_u8(data[IFLA_GTP_RESTART_COUNT]); + gtp->net = src_net; err = gtp_hashtable_new(gtp, hashsize); @@ -801,6 +977,7 @@ static const struct nla_policy gtp_policy[IFLA_GTP_MAX + 1] = { [IFLA_GTP_PDP_HASHSIZE] = { .type = NLA_U32 }, [IFLA_GTP_ROLE] = { .type = NLA_U32 }, [IFLA_GTP_CREATE_SOCKETS] = { .type = NLA_U8 }, + [IFLA_GTP_RESTART_COUNT] = { .type = NLA_U8 }, }; static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], @@ -815,7 +992,8 @@ static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], static size_t gtp_get_size(const struct net_device *dev) { return nla_total_size(sizeof(__u32)) + /* IFLA_GTP_PDP_HASHSIZE */ - nla_total_size(sizeof(__u32)); /* IFLA_GTP_ROLE */ + nla_total_size(sizeof(__u32)) + /* IFLA_GTP_ROLE */ + nla_total_size(sizeof(__u8)); /* IFLA_GTP_RESTART_COUNT */ } static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) @@ -826,6 +1004,8 @@ static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) goto nla_put_failure; if (nla_put_u32(skb, IFLA_GTP_ROLE, gtp->role)) goto nla_put_failure; + if (nla_put_u8(skb, IFLA_GTP_RESTART_COUNT, gtp->restart_count)) + goto nla_put_failure; return 0; diff --git a/include/net/gtp.h b/include/net/gtp.h index 0e16ebb2a82d..0e12c37f2958 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -7,8 +7,13 @@ #define GTP0_PORT 3386 #define GTP1U_PORT 2152 +/* GTP messages types */ +#define GTP_ECHO_REQ 1 /* Echo Request */ +#define GTP_ECHO_RSP 2 /* Echo Response */ #define GTP_TPDU 255 +#define GTPIE_RECOVERY 14 + struct gtp0_header { /* According to GSM TS 09.60. */ __u8 flags; __u8 type; @@ -27,6 +32,32 @@ struct gtp1_header { /* According to 3GPP TS 29.060. */ __be32 tid; } __attribute__ ((packed)); +struct gtp1_header_long { /* According to 3GPP TS 29.060. */ + __u8 flags; + __u8 type; + __be16 length; + __be32 tid; + __be16 seq; + __u8 npdu; + __u8 next; +} __packed; + +/* GTP Information Element */ +struct gtp_ie { + __u8 tag; + __u8 val; +} __packed; + +struct gtp0_packet { + struct gtp0_header gtp0_h; + struct gtp_ie ie; +} __packed; + +struct gtp1u_packet { + struct gtp1_header_long gtp1u_h; + struct gtp_ie ie; +} __packed; + #define GTP1_F_NPDU 0x01 #define GTP1_F_SEQ 0x02 #define GTP1_F_EXTHDR 0x04 diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 42f3fb097271..0123a8024ef2 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -823,6 +823,7 @@ enum { IFLA_GTP_PDP_HASHSIZE, IFLA_GTP_ROLE, IFLA_GTP_CREATE_SOCKETS, + IFLA_GTP_RESTART_COUNT, __IFLA_GTP_MAX, }; #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1) -- 2.31.1

2 years, 2 months

1
0
0 0

[PATCH net-next v6 1/7] gtp: Allow to create GTP device without FDs

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Currently, when the user wants to create GTP device, he has to provide file handles to the sockets created in userspace (IFLA_GTP_FD0, IFLA_GTP_FD1). This behaviour is not ideal, considering the option of adding support for GTP device creation through ip link. Ip link application is not a good place to create such sockets. This patch allows to create GTP device without providing IFLA_GTP_FD0 and IFLA_GTP_FD1 arguments. If the user sets IFLA_GTP_CREATE_SOCKETS attribute, then GTP module takes care of creating UDP sockets by itself. Sockets are created with the commonly known UDP ports used for GTP protocol (GTP0_PORT and GTP1U_PORT). In this case we don't have to provide encap_destroy because no extra deinitialization is needed, everything is covered by udp_tunnel_sock_release. Note: GTP instance created with only this change applied, does not handle GTP Echo Requests. This is implemented in the following patch. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- v4: use ntohs when creating UDP socket v5: IFLA_GTP_CREATE_SOCKETS introduced, gtp_newlink refactor v6: reordering refactor removed --- drivers/net/gtp.c | 101 +++++++++++++++++++++++++++++------ include/uapi/linux/if_link.h | 1 + 2 files changed, 85 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index bf087171bcf0..25d8521897b3 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -66,8 +66,10 @@ struct gtp_dev { struct sock *sk0; struct sock *sk1u; + u8 sk_created; struct net_device *dev; + struct net *net; unsigned int role; unsigned int hash_size; @@ -320,8 +322,16 @@ static void gtp_encap_disable_sock(struct sock *sk) static void gtp_encap_disable(struct gtp_dev *gtp) { - gtp_encap_disable_sock(gtp->sk0); - gtp_encap_disable_sock(gtp->sk1u); + if (gtp->sk_created) { + udp_tunnel_sock_release(gtp->sk0->sk_socket); + udp_tunnel_sock_release(gtp->sk1u->sk_socket); + gtp->sk_created = false; + gtp->sk0 = NULL; + gtp->sk1u = NULL; + } else { + gtp_encap_disable_sock(gtp->sk0); + gtp_encap_disable_sock(gtp->sk1u); + } } /* UDP encapsulation receive handler. See net/ipv4/udp.c. @@ -656,17 +666,69 @@ static void gtp_destructor(struct net_device *dev) kfree(gtp->tid_hash); } +static struct sock *gtp_create_sock(int type, struct gtp_dev *gtp) +{ + struct udp_tunnel_sock_cfg tuncfg = {}; + struct udp_port_cfg udp_conf = { + .local_ip.s_addr = htonl(INADDR_ANY), + .family = AF_INET, + }; + struct net *net = gtp->net; + struct socket *sock; + int err; + + if (type == UDP_ENCAP_GTP0) + udp_conf.local_udp_port = htons(GTP0_PORT); + else if (type == UDP_ENCAP_GTP1U) + udp_conf.local_udp_port = htons(GTP1U_PORT); + else + return ERR_PTR(-EINVAL); + + err = udp_sock_create(net, &udp_conf, &sock); + if (err) + return ERR_PTR(err); + + tuncfg.sk_user_data = gtp; + tuncfg.encap_type = type; + tuncfg.encap_rcv = gtp_encap_recv; + tuncfg.encap_destroy = NULL; + + setup_udp_tunnel_sock(net, sock, &tuncfg); + + return sock->sk; +} + +static int gtp_create_sockets(struct gtp_dev *gtp, struct nlattr *data[]) +{ + struct sock *sk1u = NULL; + struct sock *sk0 = NULL; + + sk0 = gtp_create_sock(UDP_ENCAP_GTP0, gtp); + if (IS_ERR(sk0)) + return PTR_ERR(sk0); + + sk1u = gtp_create_sock(UDP_ENCAP_GTP1U, gtp); + if (IS_ERR(sk1u)) { + udp_tunnel_sock_release(sk0->sk_socket); + return PTR_ERR(sk1u); + } + + gtp->sk_created = true; + gtp->sk0 = sk0; + gtp->sk1u = sk1u; + + return 0; +} + static int gtp_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) { + unsigned int role = GTP_ROLE_GGSN; struct gtp_dev *gtp; struct gtp_net *gn; int hashsize, err; - if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) - return -EINVAL; - gtp = netdev_priv(dev); if (!data[IFLA_GTP_PDP_HASHSIZE]) { @@ -677,11 +739,23 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, hashsize = 1024; } + if (data[IFLA_GTP_ROLE]) { + role = nla_get_u32(data[IFLA_GTP_ROLE]); + if (role > GTP_ROLE_SGSN) + return -EINVAL; + } + gtp->role = role; + + gtp->net = src_net; + err = gtp_hashtable_new(gtp, hashsize); if (err < 0) return err; - err = gtp_encap_enable(gtp, data); + if (data[IFLA_GTP_CREATE_SOCKETS]) + err = gtp_create_sockets(gtp, data); + else + err = gtp_encap_enable(gtp, data); if (err < 0) goto out_hashtable; @@ -726,6 +800,7 @@ static const struct nla_policy gtp_policy[IFLA_GTP_MAX + 1] = { [IFLA_GTP_FD1] = { .type = NLA_U32 }, [IFLA_GTP_PDP_HASHSIZE] = { .type = NLA_U32 }, [IFLA_GTP_ROLE] = { .type = NLA_U32 }, + [IFLA_GTP_CREATE_SOCKETS] = { .type = NLA_U8 }, }; static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], @@ -848,7 +923,9 @@ static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) { struct sock *sk1u = NULL; struct sock *sk0 = NULL; - unsigned int role = GTP_ROLE_GGSN; + + if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) + return -EINVAL; if (data[IFLA_GTP_FD0]) { u32 fd0 = nla_get_u32(data[IFLA_GTP_FD0]); @@ -868,18 +945,8 @@ static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) } } - if (data[IFLA_GTP_ROLE]) { - role = nla_get_u32(data[IFLA_GTP_ROLE]); - if (role > GTP_ROLE_SGSN) { - gtp_encap_disable_sock(sk0); - gtp_encap_disable_sock(sk1u); - return -EINVAL; - } - } - gtp->sk0 = sk0; gtp->sk1u = sk1u; - gtp->role = role; return 0; } diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 6218f93f5c1a..42f3fb097271 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -822,6 +822,7 @@ enum { IFLA_GTP_FD1, IFLA_GTP_PDP_HASHSIZE, IFLA_GTP_ROLE, + IFLA_GTP_CREATE_SOCKETS, __IFLA_GTP_MAX, }; #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1) -- 2.31.1

2 years, 2 months

1
0
0 0

[PATCH net-next v6 1/7] gtp: Allow to create GTP device without FDs

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Currently, when the user wants to create GTP device, he has to provide file handles to the sockets created in userspace (IFLA_GTP_FD0, IFLA_GTP_FD1). This behaviour is not ideal, considering the option of adding support for GTP device creation through ip link. Ip link application is not a good place to create such sockets. This patch allows to create GTP device without providing IFLA_GTP_FD0 and IFLA_GTP_FD1 arguments. If the user sets IFLA_GTP_CREATE_SOCKETS attribute, then GTP module takes care of creating UDP sockets by itself. Sockets are created with the commonly known UDP ports used for GTP protocol (GTP0_PORT and GTP1U_PORT). In this case we don't have to provide encap_destroy because no extra deinitialization is needed, everything is covered by udp_tunnel_sock_release. Note: GTP instance created with only this change applied, does not handle GTP Echo Requests. This is implemented in the following patch. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- v4: use ntohs when creating UDP socket v5: IFLA_GTP_CREATE_SOCKETS introduced, gtp_newlink refactor v6: reordering refactor removed --- drivers/net/gtp.c | 101 +++++++++++++++++++++++++++++------ include/uapi/linux/if_link.h | 1 + 2 files changed, 85 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index bf087171bcf0..25d8521897b3 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -66,8 +66,10 @@ struct gtp_dev { struct sock *sk0; struct sock *sk1u; + u8 sk_created; struct net_device *dev; + struct net *net; unsigned int role; unsigned int hash_size; @@ -320,8 +322,16 @@ static void gtp_encap_disable_sock(struct sock *sk) static void gtp_encap_disable(struct gtp_dev *gtp) { - gtp_encap_disable_sock(gtp->sk0); - gtp_encap_disable_sock(gtp->sk1u); + if (gtp->sk_created) { + udp_tunnel_sock_release(gtp->sk0->sk_socket); + udp_tunnel_sock_release(gtp->sk1u->sk_socket); + gtp->sk_created = false; + gtp->sk0 = NULL; + gtp->sk1u = NULL; + } else { + gtp_encap_disable_sock(gtp->sk0); + gtp_encap_disable_sock(gtp->sk1u); + } } /* UDP encapsulation receive handler. See net/ipv4/udp.c. @@ -656,17 +666,69 @@ static void gtp_destructor(struct net_device *dev) kfree(gtp->tid_hash); } +static struct sock *gtp_create_sock(int type, struct gtp_dev *gtp) +{ + struct udp_tunnel_sock_cfg tuncfg = {}; + struct udp_port_cfg udp_conf = { + .local_ip.s_addr = htonl(INADDR_ANY), + .family = AF_INET, + }; + struct net *net = gtp->net; + struct socket *sock; + int err; + + if (type == UDP_ENCAP_GTP0) + udp_conf.local_udp_port = htons(GTP0_PORT); + else if (type == UDP_ENCAP_GTP1U) + udp_conf.local_udp_port = htons(GTP1U_PORT); + else + return ERR_PTR(-EINVAL); + + err = udp_sock_create(net, &udp_conf, &sock); + if (err) + return ERR_PTR(err); + + tuncfg.sk_user_data = gtp; + tuncfg.encap_type = type; + tuncfg.encap_rcv = gtp_encap_recv; + tuncfg.encap_destroy = NULL; + + setup_udp_tunnel_sock(net, sock, &tuncfg); + + return sock->sk; +} + +static int gtp_create_sockets(struct gtp_dev *gtp, struct nlattr *data[]) +{ + struct sock *sk1u = NULL; + struct sock *sk0 = NULL; + + sk0 = gtp_create_sock(UDP_ENCAP_GTP0, gtp); + if (IS_ERR(sk0)) + return PTR_ERR(sk0); + + sk1u = gtp_create_sock(UDP_ENCAP_GTP1U, gtp); + if (IS_ERR(sk1u)) { + udp_tunnel_sock_release(sk0->sk_socket); + return PTR_ERR(sk1u); + } + + gtp->sk_created = true; + gtp->sk0 = sk0; + gtp->sk1u = sk1u; + + return 0; +} + static int gtp_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) { + unsigned int role = GTP_ROLE_GGSN; struct gtp_dev *gtp; struct gtp_net *gn; int hashsize, err; - if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) - return -EINVAL; - gtp = netdev_priv(dev); if (!data[IFLA_GTP_PDP_HASHSIZE]) { @@ -677,11 +739,23 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, hashsize = 1024; } + if (data[IFLA_GTP_ROLE]) { + role = nla_get_u32(data[IFLA_GTP_ROLE]); + if (role > GTP_ROLE_SGSN) + return -EINVAL; + } + gtp->role = role; + + gtp->net = src_net; + err = gtp_hashtable_new(gtp, hashsize); if (err < 0) return err; - err = gtp_encap_enable(gtp, data); + if (data[IFLA_GTP_CREATE_SOCKETS]) + err = gtp_create_sockets(gtp, data); + else + err = gtp_encap_enable(gtp, data); if (err < 0) goto out_hashtable; @@ -726,6 +800,7 @@ static const struct nla_policy gtp_policy[IFLA_GTP_MAX + 1] = { [IFLA_GTP_FD1] = { .type = NLA_U32 }, [IFLA_GTP_PDP_HASHSIZE] = { .type = NLA_U32 }, [IFLA_GTP_ROLE] = { .type = NLA_U32 }, + [IFLA_GTP_CREATE_SOCKETS] = { .type = NLA_U8 }, }; static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], @@ -848,7 +923,9 @@ static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) { struct sock *sk1u = NULL; struct sock *sk0 = NULL; - unsigned int role = GTP_ROLE_GGSN; + + if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) + return -EINVAL; if (data[IFLA_GTP_FD0]) { u32 fd0 = nla_get_u32(data[IFLA_GTP_FD0]); @@ -868,18 +945,8 @@ static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) } } - if (data[IFLA_GTP_ROLE]) { - role = nla_get_u32(data[IFLA_GTP_ROLE]); - if (role > GTP_ROLE_SGSN) { - gtp_encap_disable_sock(sk0); - gtp_encap_disable_sock(sk1u); - return -EINVAL; - } - } - gtp->sk0 = sk0; gtp->sk1u = sk1u; - gtp->role = role; return 0; } diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 6218f93f5c1a..42f3fb097271 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -822,6 +822,7 @@ enum { IFLA_GTP_FD1, IFLA_GTP_PDP_HASHSIZE, IFLA_GTP_ROLE, + IFLA_GTP_CREATE_SOCKETS, __IFLA_GTP_MAX, }; #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1) -- 2.31.1

2 years, 2 months

1
0
0 0

[RFC PATCH net-next v4 4/6] gtp: Implement GTP echo response

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Adding GTP device through ip link creates the situation where there is no userspace daemon which would handle GTP messages (Echo Request for example). GTP-U instance which would not respond to echo requests would violate GTP specification. When GTP packet arrives with GTP_ECHO_REQ message type, GTP_ECHO_RSP is send to the sender. GTP_ECHO_RSP message should contain information element with GTPIE_RECOVERY tag and restart counter value. For GTPv1 restart counter is not used and should be equal to 0, for GTPv0 restart counter contains information provided from userspace(IFLA_GTP_RESTART_COUNT). Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- drivers/net/gtp.c | 212 ++++++++++++++++++++++++++++++++--- include/net/gtp.h | 31 +++++ include/uapi/linux/if_link.h | 1 + 3 files changed, 228 insertions(+), 16 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 6fa1cfe023ef..d20bc272ff6c 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -75,6 +75,8 @@ struct gtp_dev { unsigned int hash_size; struct hlist_head *tid_hash; struct hlist_head *addr_hash; + + u8 restart_count; }; static unsigned int gtp_net_id __read_mostly; @@ -217,6 +219,106 @@ static int gtp_rx(struct pdp_ctx *pctx, struct sk_buff *skb, return -1; } +static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, + const struct sock *sk, + __be32 daddr, __be32 saddr) +{ + memset(fl4, 0, sizeof(*fl4)); + fl4->flowi4_oif = sk->sk_bound_dev_if; + fl4->daddr = daddr; + fl4->saddr = saddr; + fl4->flowi4_tos = RT_CONN_FLAGS(sk); + fl4->flowi4_proto = sk->sk_protocol; + + return ip_route_output_key(sock_net(sk), fl4); +} + +/* GSM TS 09.60. 7.3 + * In all Path Management messages: + * - TID: is not used and shall be set to 0. + * - Flow Label is not used and shall be set to 0 + * In signalling messages: + * - number: this field is not yet used in signalling messages. + * It shall be set to 255 by the sender and shall be ignored + * by the receiver + * Returns true if the echo req was correct, false otherwise. + */ +static bool gtp0_validate_echo_req(struct gtp0_header *gtp0) +{ + return !(gtp0->tid || (gtp0->flags ^ 0x1e) || + gtp0->number != 0xff || gtp0->flow); +} + +static int gtp0_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp0_packet *gtp_pkt; + struct gtp0_header *gtp0; + struct rtable *rt; + struct flowi4 fl4; + struct iphdr *iph; + __be16 seq; + + gtp0 = (struct gtp0_header *)(skb->data + sizeof(struct udphdr)); + + if (!gtp0_validate_echo_req(gtp0)) + return -1; + + seq = gtp0->seq; + + /* pull GTP and UDP headers */ + skb_pull_data(skb, sizeof(struct gtp0_header) + sizeof(struct udphdr)); + + gtp_pkt = skb_push(skb, sizeof(struct gtp0_packet)); + memset(gtp_pkt, 0, sizeof(struct gtp0_packet)); + + gtp_pkt->gtp0_h.flags = 0x1e; /* v0, GTP-non-prime. */ + gtp_pkt->gtp0_h.type = GTP_ECHO_RSP; + gtp_pkt->gtp0_h.length = + htons(sizeof(struct gtp0_packet) - sizeof(struct gtp0_header)); + + /* GSM TS 09.60. 7.3 The Sequence Number in a signalling response + * message shall be copied from the signalling request message + * that the GSN is replying to. + */ + gtp_pkt->gtp0_h.seq = seq; + + /* GSM TS 09.60. 7.3 In all Path Management Flow Label and TID + * are not used and shall be set to 0. + */ + gtp_pkt->gtp0_h.flow = 0; + gtp_pkt->gtp0_h.tid = 0; + gtp_pkt->gtp0_h.number = 0xff; + gtp_pkt->gtp0_h.spare[0] = 0xff; + gtp_pkt->gtp0_h.spare[1] = 0xff; + gtp_pkt->gtp0_h.spare[2] = 0xff; + + gtp_pkt->ie.tag = GTPIE_RECOVERY; + gtp_pkt->ie.val = gtp->restart_count; + + iph = ip_hdr(skb); + + /* find route to the sender, + * src address becomes dst address and vice versa. + */ + rt = ip4_route_output_gtp(&fl4, gtp->sk0, iph->saddr, iph->daddr); + if (IS_ERR(rt)) { + netdev_dbg(gtp->dev, "no route for echo response from %pI4\n", + &iph->saddr); + return -1; + } + + udp_tunnel_xmit_skb(rt, gtp->sk0, skb, + fl4.saddr, fl4.daddr, + iph->tos, + ip4_dst_hoplimit(&rt->dst), + 0, + htons(GTP0_PORT), htons(GTP0_PORT), + !net_eq(sock_net(gtp->sk1u), + dev_net(gtp->dev)), + false); + return 0; +} + /* 1 means pass up to the stack, -1 means drop and 0 means decapsulated. */ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { @@ -233,6 +335,13 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if ((gtp0->flags >> 5) != GTP_V0) return 1; + /* If the sockets were created in kernel, it means that + * there is no daemon running in userspace which would + * handle echo request. + */ + if (gtp0->type == GTP_ECHO_REQ && gtp->sk_created) + return gtp0_echo_resp(gtp, skb); + if (gtp0->type != GTP_TPDU) return 1; @@ -245,6 +354,74 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) return gtp_rx(pctx, skb, hdrlen, gtp->role); } +static int gtp1u_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp1_header_long *gtp1u; + struct gtp1u_packet *gtp_pkt; + struct rtable *rt; + struct flowi4 fl4; + struct iphdr *iph; + + gtp1u = (struct gtp1_header_long *)(skb->data + sizeof(struct udphdr)); + + /* 3GPP TS 29.281 5.1 - For the Echo Request, Echo Response, + * Error Indication and Supported Extension Headers Notification + * messages, the S flag shall be set to 1 and TEID shall be set to 0. + */ + if (!(gtp1u->flags & GTP1_F_SEQ) || gtp1u->tid) + return -1; + + /* pull GTP and UDP headers */ + skb_pull_data(skb, sizeof(struct gtp1_header_long) + sizeof(struct udphdr)); + + gtp_pkt = skb_push(skb, sizeof(struct gtp1u_packet)); + memset(gtp_pkt, 0, sizeof(struct gtp1u_packet)); + + /* S flag must be set to 1 */ + gtp_pkt->gtp1u_h.flags = 0x32; + gtp_pkt->gtp1u_h.type = GTP_ECHO_RSP; + /* seq, npdu and next should be counted to the length of the GTP packet + * that's why szie of gtp1_header should be subtracted, + * not why szie of gtp1_header_long. + */ + gtp_pkt->gtp1u_h.length = + htons(sizeof(struct gtp1u_packet) - sizeof(struct gtp1_header)); + /* 3GPP TS 29.281 5.1 - TEID has to be set to 0 */ + gtp_pkt->gtp1u_h.tid = 0; + + /* 3GPP TS 29.281 7.7.2 - The Restart Counter value in the + * Recovery information element shall not be used, i.e. it shall + * be set to zero by the sender and shall be ignored by the receiver. + * The Recovery information element is mandatory due to backwards + * compatibility reasons. + */ + gtp_pkt->ie.tag = GTPIE_RECOVERY; + gtp_pkt->ie.val = 0; + + iph = ip_hdr(skb); + + /* find route to the sender, + * src address becomes dst address and vice versa. + */ + rt = ip4_route_output_gtp(&fl4, gtp->sk1u, iph->saddr, iph->daddr); + if (IS_ERR(rt)) { + netdev_dbg(gtp->dev, "no route for echo response from %pI4\n", + &iph->saddr); + return -1; + } + + udp_tunnel_xmit_skb(rt, gtp->sk1u, skb, + fl4.saddr, fl4.daddr, + iph->tos, + ip4_dst_hoplimit(&rt->dst), + 0, + htons(GTP1U_PORT), htons(GTP1U_PORT), + !net_eq(sock_net(gtp->sk1u), + dev_net(gtp->dev)), + false); + return 0; +} + static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { unsigned int hdrlen = sizeof(struct udphdr) + @@ -260,6 +437,13 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if ((gtp1->flags >> 5) != GTP_V1) return 1; + /* If the sockets were created in kernel, it means that + * there is no daemon running in userspace which would + * handle echo request. + */ + if (gtp1->type == GTP_ECHO_REQ && gtp->sk_created) + return gtp1u_echo_resp(gtp, skb); + if (gtp1->type != GTP_TPDU) return 1; @@ -398,20 +582,6 @@ static void gtp_dev_uninit(struct net_device *dev) free_percpu(dev->tstats); } -static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, - const struct sock *sk, - __be32 daddr) -{ - memset(fl4, 0, sizeof(*fl4)); - fl4->flowi4_oif = sk->sk_bound_dev_if; - fl4->daddr = daddr; - fl4->saddr = inet_sk(sk)->inet_saddr; - fl4->flowi4_tos = RT_CONN_FLAGS(sk); - fl4->flowi4_proto = sk->sk_protocol; - - return ip_route_output_key(sock_net(sk), fl4); -} - static inline void gtp0_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) { int payload_len = skb->len; @@ -517,7 +687,8 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, } netdev_dbg(dev, "found PDP context %p\n", pctx); - rt = ip4_route_output_gtp(&fl4, pctx->sk, pctx->peer_addr_ip4.s_addr); + rt = ip4_route_output_gtp(&fl4, pctx->sk, pctx->peer_addr_ip4.s_addr, + inet_sk(pctx->sk)->inet_saddr); if (IS_ERR(rt)) { netdev_dbg(dev, "no route to SSGN %pI4\n", &pctx->peer_addr_ip4.s_addr); @@ -684,6 +855,11 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, hashsize = 1024; } + if (!data[IFLA_GTP_RESTART_COUNT]) + gtp->restart_count = 0; + else + gtp->restart_count = nla_get_u8(data[IFLA_GTP_RESTART_COUNT]); + gtp->net = src_net; err = gtp_hashtable_new(gtp, hashsize); @@ -735,6 +911,7 @@ static const struct nla_policy gtp_policy[IFLA_GTP_MAX + 1] = { [IFLA_GTP_FD1] = { .type = NLA_U32 }, [IFLA_GTP_PDP_HASHSIZE] = { .type = NLA_U32 }, [IFLA_GTP_ROLE] = { .type = NLA_U32 }, + [IFLA_GTP_RESTART_COUNT] = { .type = NLA_U8 }, }; static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], @@ -749,7 +926,8 @@ static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], static size_t gtp_get_size(const struct net_device *dev) { return nla_total_size(sizeof(__u32)) + /* IFLA_GTP_PDP_HASHSIZE */ - nla_total_size(sizeof(__u32)); /* IFLA_GTP_ROLE */ + nla_total_size(sizeof(__u32)) + /* IFLA_GTP_ROLE */ + nla_total_size(sizeof(__u8)); /* IFLA_GTP_RESTART_COUNT */ } static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) @@ -760,6 +938,8 @@ static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) goto nla_put_failure; if (nla_put_u32(skb, IFLA_GTP_ROLE, gtp->role)) goto nla_put_failure; + if (nla_put_u8(skb, IFLA_GTP_RESTART_COUNT, gtp->restart_count)) + goto nla_put_failure; return 0; diff --git a/include/net/gtp.h b/include/net/gtp.h index c78702e3d663..53a712f73b13 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -7,8 +7,13 @@ #define GTP0_PORT 3386 #define GTP1U_PORT 2152 +/* GTP messages types */ +#define GTP_ECHO_REQ 1 /* Echo Request */ +#define GTP_ECHO_RSP 2 /* Echo Response */ #define GTP_TPDU 255 +#define GTPIE_RECOVERY 14 + struct gtp0_header { /* According to GSM TS 09.60. */ __u8 flags; __u8 type; @@ -27,11 +32,37 @@ struct gtp1_header { /* According to 3GPP TS 29.060. */ __be32 tid; } __attribute__ ((packed)); +struct gtp1_header_long { /* According to 3GPP TS 29.060. */ + __u8 flags; + __u8 type; + __be16 length; + __be32 tid; + __be16 seq; + __u8 npdu; + __u8 next; +} __packed; + struct gtp_pdu_session_info { /* According to 3GPP TS 38.415. */ u8 pdu_type; u8 qfi; }; +/* GTP Information Element */ +struct gtp_ie { + __u8 tag; + __u8 val; +} __packed; + +struct gtp0_packet { + struct gtp0_header gtp0_h; + struct gtp_ie ie; +} __packed; + +struct gtp1u_packet { + struct gtp1_header_long gtp1u_h; + struct gtp_ie ie; +} __packed; + static inline bool netif_is_gtp(const struct net_device *dev) { return dev->rtnl_link_ops && diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 6218f93f5c1a..2fc5a0371283 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -822,6 +822,7 @@ enum { IFLA_GTP_FD1, IFLA_GTP_PDP_HASHSIZE, IFLA_GTP_ROLE, + IFLA_GTP_RESTART_COUNT, __IFLA_GTP_MAX, }; #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1) -- 2.31.1

2 years, 2 months

4
6
0 0

[RFC PATCH net-next v3 1/5] gtp: Allow to create GTP device without FDs

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Currently, when the user wants to create GTP device, he has to provide file handles to the sockets created in userspace (IFLA_GTP_FD0, IFLA_GTP_FD1). This behaviour is not ideal, considering the option of adding support for GTP device creation through ip link. Ip link application is not a good place to create such sockets. This patch allows to create GTP device without providing IFLA_GTP_FD0 and IFLA_GTP_FD1 arguments. If the user does not provide file handles to the sockets, then GTP module takes care of creating UDP sockets by itself. Sockets are created with the commonly known UDP ports used for GTP protocol (GTP0_PORT and GTP1U_PORT). In this case we don't have to provide encap_destroy because no extra deinitialization is needed, everything is covered by udp_tunnel_sock_release. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- drivers/net/gtp.c | 74 ++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 67 insertions(+), 7 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 24e5c54d06c1..a2ad0af913cb 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -66,8 +66,10 @@ struct gtp_dev { struct sock *sk0; struct sock *sk1u; + u8 sk_created; struct net_device *dev; + struct net *net; unsigned int role; unsigned int hash_size; @@ -320,8 +322,16 @@ static void gtp_encap_disable_sock(struct sock *sk) static void gtp_encap_disable(struct gtp_dev *gtp) { - gtp_encap_disable_sock(gtp->sk0); - gtp_encap_disable_sock(gtp->sk1u); + if (gtp->sk_created) { + udp_tunnel_sock_release(gtp->sk0->sk_socket); + udp_tunnel_sock_release(gtp->sk1u->sk_socket); + gtp->sk_created = false; + gtp->sk0 = NULL; + gtp->sk1u = NULL; + } else { + gtp_encap_disable_sock(gtp->sk0); + gtp_encap_disable_sock(gtp->sk1u); + } } /* UDP encapsulation receive handler. See net/ipv4/udp.c. @@ -664,9 +674,6 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, struct gtp_net *gn; int hashsize, err; - if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) - return -EINVAL; - gtp = netdev_priv(dev); if (!data[IFLA_GTP_PDP_HASHSIZE]) { @@ -677,6 +684,8 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, hashsize = 1024; } + gtp->net = src_net; + err = gtp_hashtable_new(gtp, hashsize); if (err < 0) return err; @@ -844,6 +853,38 @@ static struct sock *gtp_encap_enable_socket(int fd, int type, return sk; } +static struct sock *gtp_encap_create_sock(int type, struct gtp_dev *gtp) +{ + struct udp_tunnel_sock_cfg tuncfg = {}; + struct udp_port_cfg udp_conf = { + .local_ip.s_addr = htonl(INADDR_ANY), + .family = AF_INET, + }; + struct net *net = gtp->net; + struct socket *sock; + int err; + + if (type == UDP_ENCAP_GTP0) + udp_conf.local_udp_port = GTP0_PORT; + else if (type == UDP_ENCAP_GTP1U) + udp_conf.local_udp_port = GTP1U_PORT; + else + return ERR_PTR(-EINVAL); + + err = udp_sock_create(net, &udp_conf, &sock); + if (err) + return ERR_PTR(err); + + tuncfg.sk_user_data = gtp; + tuncfg.encap_type = type; + tuncfg.encap_rcv = gtp_encap_recv; + tuncfg.encap_destroy = NULL; + + setup_udp_tunnel_sock(net, sock, &tuncfg); + + return sock->sk; +} + static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) { struct sock *sk1u = NULL; @@ -868,11 +909,30 @@ static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) } } + if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) { + sk0 = gtp_encap_create_sock(UDP_ENCAP_GTP0, gtp); + if (IS_ERR(sk0)) + return PTR_ERR(sk0); + + sk1u = gtp_encap_create_sock(UDP_ENCAP_GTP1U, gtp); + if (IS_ERR(sk1u)) { + udp_tunnel_sock_release(sk0->sk_socket); + return PTR_ERR(sk1u); + } + gtp->sk_created = true; + } + if (data[IFLA_GTP_ROLE]) { role = nla_get_u32(data[IFLA_GTP_ROLE]); if (role > GTP_ROLE_SGSN) { - gtp_encap_disable_sock(sk0); - gtp_encap_disable_sock(sk1u); + if (gtp->sk_created) { + udp_tunnel_sock_release(sk0->sk_socket); + udp_tunnel_sock_release(sk1u->sk_socket); + gtp->sk_created = false; + } else { + gtp_encap_disable_sock(sk0); + gtp_encap_disable_sock(sk1u); + } return -EINVAL; } } -- 2.31.1

2 years, 2 months

4
7
0 0

[RFC PATCH net-next v5 4/6] gtp: Add support for checking GTP device type

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Add a function that checks if a net device type is GTP. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- include/net/gtp.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/net/gtp.h b/include/net/gtp.h index 23c2aaae8a42..c1d6169df331 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -63,6 +63,12 @@ struct gtp_pdu_session_info { /* According to 3GPP TS 38.415. */ u8 qfi; }; +static inline bool netif_is_gtp(const struct net_device *dev) +{ + return dev->rtnl_link_ops && + !strcmp(dev->rtnl_link_ops->kind, "gtp"); +} + #define GTP1_F_NPDU 0x01 #define GTP1_F_SEQ 0x02 #define GTP1_F_EXTHDR 0x04 -- 2.31.1

2 years, 2 months

2
1
0 0

[RFC PATCH net-next v5 2/6] gtp: Implement GTP echo response

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Adding GTP device through ip link creates the situation where there is no userspace daemon which would handle GTP messages (Echo Request for example). GTP-U instance which would not respond to echo requests would violate GTP specification. When GTP packet arrives with GTP_ECHO_REQ message type, GTP_ECHO_RSP is send to the sender. GTP_ECHO_RSP message should contain information element with GTPIE_RECOVERY tag and restart counter value. For GTPv1 restart counter is not used and should be equal to 0, for GTPv0 restart counter contains information provided from userspace(IFLA_GTP_RESTART_COUNT). Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- drivers/net/gtp.c | 212 ++++++++++++++++++++++++++++++++--- include/net/gtp.h | 31 +++++ include/uapi/linux/if_link.h | 1 + 3 files changed, 228 insertions(+), 16 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 29978c5e37e8..a51782868d23 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -75,6 +75,8 @@ struct gtp_dev { unsigned int hash_size; struct hlist_head *tid_hash; struct hlist_head *addr_hash; + + u8 restart_count; }; static unsigned int gtp_net_id __read_mostly; @@ -217,6 +219,106 @@ static int gtp_rx(struct pdp_ctx *pctx, struct sk_buff *skb, return -1; } +static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, + const struct sock *sk, + __be32 daddr, __be32 saddr) +{ + memset(fl4, 0, sizeof(*fl4)); + fl4->flowi4_oif = sk->sk_bound_dev_if; + fl4->daddr = daddr; + fl4->saddr = saddr; + fl4->flowi4_tos = RT_CONN_FLAGS(sk); + fl4->flowi4_proto = sk->sk_protocol; + + return ip_route_output_key(sock_net(sk), fl4); +} + +/* GSM TS 09.60. 7.3 + * In all Path Management messages: + * - TID: is not used and shall be set to 0. + * - Flow Label is not used and shall be set to 0 + * In signalling messages: + * - number: this field is not yet used in signalling messages. + * It shall be set to 255 by the sender and shall be ignored + * by the receiver + * Returns true if the echo req was correct, false otherwise. + */ +static bool gtp0_validate_echo_req(struct gtp0_header *gtp0) +{ + return !(gtp0->tid || (gtp0->flags ^ 0x1e) || + gtp0->number != 0xff || gtp0->flow); +} + +static int gtp0_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp0_packet *gtp_pkt; + struct gtp0_header *gtp0; + struct rtable *rt; + struct flowi4 fl4; + struct iphdr *iph; + __be16 seq; + + gtp0 = (struct gtp0_header *)(skb->data + sizeof(struct udphdr)); + + if (!gtp0_validate_echo_req(gtp0)) + return -1; + + seq = gtp0->seq; + + /* pull GTP and UDP headers */ + skb_pull_data(skb, sizeof(struct gtp0_header) + sizeof(struct udphdr)); + + gtp_pkt = skb_push(skb, sizeof(struct gtp0_packet)); + memset(gtp_pkt, 0, sizeof(struct gtp0_packet)); + + gtp_pkt->gtp0_h.flags = 0x1e; /* v0, GTP-non-prime. */ + gtp_pkt->gtp0_h.type = GTP_ECHO_RSP; + gtp_pkt->gtp0_h.length = + htons(sizeof(struct gtp0_packet) - sizeof(struct gtp0_header)); + + /* GSM TS 09.60. 7.3 The Sequence Number in a signalling response + * message shall be copied from the signalling request message + * that the GSN is replying to. + */ + gtp_pkt->gtp0_h.seq = seq; + + /* GSM TS 09.60. 7.3 In all Path Management Flow Label and TID + * are not used and shall be set to 0. + */ + gtp_pkt->gtp0_h.flow = 0; + gtp_pkt->gtp0_h.tid = 0; + gtp_pkt->gtp0_h.number = 0xff; + gtp_pkt->gtp0_h.spare[0] = 0xff; + gtp_pkt->gtp0_h.spare[1] = 0xff; + gtp_pkt->gtp0_h.spare[2] = 0xff; + + gtp_pkt->ie.tag = GTPIE_RECOVERY; + gtp_pkt->ie.val = gtp->restart_count; + + iph = ip_hdr(skb); + + /* find route to the sender, + * src address becomes dst address and vice versa. + */ + rt = ip4_route_output_gtp(&fl4, gtp->sk0, iph->saddr, iph->daddr); + if (IS_ERR(rt)) { + netdev_dbg(gtp->dev, "no route for echo response from %pI4\n", + &iph->saddr); + return -1; + } + + udp_tunnel_xmit_skb(rt, gtp->sk0, skb, + fl4.saddr, fl4.daddr, + iph->tos, + ip4_dst_hoplimit(&rt->dst), + 0, + htons(GTP0_PORT), htons(GTP0_PORT), + !net_eq(sock_net(gtp->sk1u), + dev_net(gtp->dev)), + false); + return 0; +} + /* 1 means pass up to the stack, -1 means drop and 0 means decapsulated. */ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { @@ -233,6 +335,13 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if ((gtp0->flags >> 5) != GTP_V0) return 1; + /* If the sockets were created in kernel, it means that + * there is no daemon running in userspace which would + * handle echo request. + */ + if (gtp0->type == GTP_ECHO_REQ && gtp->sk_created) + return gtp0_echo_resp(gtp, skb); + if (gtp0->type != GTP_TPDU) return 1; @@ -245,6 +354,74 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) return gtp_rx(pctx, skb, hdrlen, gtp->role); } +static int gtp1u_echo_resp(struct gtp_dev *gtp, struct sk_buff *skb) +{ + struct gtp1_header_long *gtp1u; + struct gtp1u_packet *gtp_pkt; + struct rtable *rt; + struct flowi4 fl4; + struct iphdr *iph; + + gtp1u = (struct gtp1_header_long *)(skb->data + sizeof(struct udphdr)); + + /* 3GPP TS 29.281 5.1 - For the Echo Request, Echo Response, + * Error Indication and Supported Extension Headers Notification + * messages, the S flag shall be set to 1 and TEID shall be set to 0. + */ + if (!(gtp1u->flags & GTP1_F_SEQ) || gtp1u->tid) + return -1; + + /* pull GTP and UDP headers */ + skb_pull_data(skb, sizeof(struct gtp1_header_long) + sizeof(struct udphdr)); + + gtp_pkt = skb_push(skb, sizeof(struct gtp1u_packet)); + memset(gtp_pkt, 0, sizeof(struct gtp1u_packet)); + + /* S flag must be set to 1 */ + gtp_pkt->gtp1u_h.flags = 0x32; + gtp_pkt->gtp1u_h.type = GTP_ECHO_RSP; + /* seq, npdu and next should be counted to the length of the GTP packet + * that's why szie of gtp1_header should be subtracted, + * not why szie of gtp1_header_long. + */ + gtp_pkt->gtp1u_h.length = + htons(sizeof(struct gtp1u_packet) - sizeof(struct gtp1_header)); + /* 3GPP TS 29.281 5.1 - TEID has to be set to 0 */ + gtp_pkt->gtp1u_h.tid = 0; + + /* 3GPP TS 29.281 7.7.2 - The Restart Counter value in the + * Recovery information element shall not be used, i.e. it shall + * be set to zero by the sender and shall be ignored by the receiver. + * The Recovery information element is mandatory due to backwards + * compatibility reasons. + */ + gtp_pkt->ie.tag = GTPIE_RECOVERY; + gtp_pkt->ie.val = 0; + + iph = ip_hdr(skb); + + /* find route to the sender, + * src address becomes dst address and vice versa. + */ + rt = ip4_route_output_gtp(&fl4, gtp->sk1u, iph->saddr, iph->daddr); + if (IS_ERR(rt)) { + netdev_dbg(gtp->dev, "no route for echo response from %pI4\n", + &iph->saddr); + return -1; + } + + udp_tunnel_xmit_skb(rt, gtp->sk1u, skb, + fl4.saddr, fl4.daddr, + iph->tos, + ip4_dst_hoplimit(&rt->dst), + 0, + htons(GTP1U_PORT), htons(GTP1U_PORT), + !net_eq(sock_net(gtp->sk1u), + dev_net(gtp->dev)), + false); + return 0; +} + static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) { unsigned int hdrlen = sizeof(struct udphdr) + @@ -260,6 +437,13 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) if ((gtp1->flags >> 5) != GTP_V1) return 1; + /* If the sockets were created in kernel, it means that + * there is no daemon running in userspace which would + * handle echo request. + */ + if (gtp1->type == GTP_ECHO_REQ && gtp->sk_created) + return gtp1u_echo_resp(gtp, skb); + if (gtp1->type != GTP_TPDU) return 1; @@ -398,20 +582,6 @@ static void gtp_dev_uninit(struct net_device *dev) free_percpu(dev->tstats); } -static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, - const struct sock *sk, - __be32 daddr) -{ - memset(fl4, 0, sizeof(*fl4)); - fl4->flowi4_oif = sk->sk_bound_dev_if; - fl4->daddr = daddr; - fl4->saddr = inet_sk(sk)->inet_saddr; - fl4->flowi4_tos = RT_CONN_FLAGS(sk); - fl4->flowi4_proto = sk->sk_protocol; - - return ip_route_output_key(sock_net(sk), fl4); -} - static inline void gtp0_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) { int payload_len = skb->len; @@ -517,7 +687,8 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, } netdev_dbg(dev, "found PDP context %p\n", pctx); - rt = ip4_route_output_gtp(&fl4, pctx->sk, pctx->peer_addr_ip4.s_addr); + rt = ip4_route_output_gtp(&fl4, pctx->sk, pctx->peer_addr_ip4.s_addr, + inet_sk(pctx->sk)->inet_saddr); if (IS_ERR(rt)) { netdev_dbg(dev, "no route to SSGN %pI4\n", &pctx->peer_addr_ip4.s_addr); @@ -848,6 +1019,11 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, } gtp->role = role; + if (!data[IFLA_GTP_RESTART_COUNT]) + gtp->restart_count = 0; + else + gtp->restart_count = nla_get_u8(data[IFLA_GTP_RESTART_COUNT]); + gtp->net = src_net; err = gtp_hashtable_new(gtp, hashsize); @@ -903,6 +1079,7 @@ static const struct nla_policy gtp_policy[IFLA_GTP_MAX + 1] = { [IFLA_GTP_PDP_HASHSIZE] = { .type = NLA_U32 }, [IFLA_GTP_ROLE] = { .type = NLA_U32 }, [IFLA_GTP_CREATE_SOCKETS] = { .type = NLA_U8 }, + [IFLA_GTP_RESTART_COUNT] = { .type = NLA_U8 }, }; static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], @@ -917,7 +1094,8 @@ static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], static size_t gtp_get_size(const struct net_device *dev) { return nla_total_size(sizeof(__u32)) + /* IFLA_GTP_PDP_HASHSIZE */ - nla_total_size(sizeof(__u32)); /* IFLA_GTP_ROLE */ + nla_total_size(sizeof(__u32)) + /* IFLA_GTP_ROLE */ + nla_total_size(sizeof(__u8)); /* IFLA_GTP_RESTART_COUNT */ } static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) @@ -928,6 +1106,8 @@ static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) goto nla_put_failure; if (nla_put_u32(skb, IFLA_GTP_ROLE, gtp->role)) goto nla_put_failure; + if (nla_put_u8(skb, IFLA_GTP_RESTART_COUNT, gtp->restart_count)) + goto nla_put_failure; return 0; diff --git a/include/net/gtp.h b/include/net/gtp.h index 0e16ebb2a82d..0e12c37f2958 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -7,8 +7,13 @@ #define GTP0_PORT 3386 #define GTP1U_PORT 2152 +/* GTP messages types */ +#define GTP_ECHO_REQ 1 /* Echo Request */ +#define GTP_ECHO_RSP 2 /* Echo Response */ #define GTP_TPDU 255 +#define GTPIE_RECOVERY 14 + struct gtp0_header { /* According to GSM TS 09.60. */ __u8 flags; __u8 type; @@ -27,6 +32,32 @@ struct gtp1_header { /* According to 3GPP TS 29.060. */ __be32 tid; } __attribute__ ((packed)); +struct gtp1_header_long { /* According to 3GPP TS 29.060. */ + __u8 flags; + __u8 type; + __be16 length; + __be32 tid; + __be16 seq; + __u8 npdu; + __u8 next; +} __packed; + +/* GTP Information Element */ +struct gtp_ie { + __u8 tag; + __u8 val; +} __packed; + +struct gtp0_packet { + struct gtp0_header gtp0_h; + struct gtp_ie ie; +} __packed; + +struct gtp1u_packet { + struct gtp1_header_long gtp1u_h; + struct gtp_ie ie; +} __packed; + #define GTP1_F_NPDU 0x01 #define GTP1_F_SEQ 0x02 #define GTP1_F_EXTHDR 0x04 diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 42f3fb097271..0123a8024ef2 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -823,6 +823,7 @@ enum { IFLA_GTP_PDP_HASHSIZE, IFLA_GTP_ROLE, IFLA_GTP_CREATE_SOCKETS, + IFLA_GTP_RESTART_COUNT, __IFLA_GTP_MAX, }; #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1) -- 2.31.1

2 years, 2 months

2
1
0 0

[RFC PATCH net-next v5 1/6] gtp: Allow to create GTP device without FDs

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Currently, when the user wants to create GTP device, he has to provide file handles to the sockets created in userspace (IFLA_GTP_FD0, IFLA_GTP_FD1). This behaviour is not ideal, considering the option of adding support for GTP device creation through ip link. Ip link application is not a good place to create such sockets. This patch allows to create GTP device without providing IFLA_GTP_FD0 and IFLA_GTP_FD1 arguments. If the user sets IFLA_GTP_CREATE_SOCKETS attribute, then GTP module takes care of creating UDP sockets by itself. Sockets are created with the commonly known UDP ports used for GTP protocol (GTP0_PORT and GTP1U_PORT). In this case we don't have to provide encap_destroy because no extra deinitialization is needed, everything is covered by udp_tunnel_sock_release. Note: GTP instance created with only this change applied, does not handle GTP Echo Requests. This is implemented in the following patch. Besides that, small refactor was done, functions used in gtp_newlink were moved above it. Handling of IFLA_GTP_ROLE attribute was moved to gtp_newlink. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- v4: use ntohs when creating UDP socket v5: IFLA_GTP_CREATE_SOCKETS introduced, gtp_newlink refactor --- drivers/net/gtp.c | 306 +++++++++++++++++++++-------------- include/uapi/linux/if_link.h | 1 + 2 files changed, 186 insertions(+), 121 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 24e5c54d06c1..29978c5e37e8 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -66,8 +66,10 @@ struct gtp_dev { struct sock *sk0; struct sock *sk1u; + u8 sk_created; struct net_device *dev; + struct net *net; unsigned int role; unsigned int hash_size; @@ -320,8 +322,16 @@ static void gtp_encap_disable_sock(struct sock *sk) static void gtp_encap_disable(struct gtp_dev *gtp) { - gtp_encap_disable_sock(gtp->sk0); - gtp_encap_disable_sock(gtp->sk1u); + if (gtp->sk_created) { + udp_tunnel_sock_release(gtp->sk0->sk_socket); + udp_tunnel_sock_release(gtp->sk1u->sk_socket); + gtp->sk_created = false; + gtp->sk0 = NULL; + gtp->sk1u = NULL; + } else { + gtp_encap_disable_sock(gtp->sk0); + gtp_encap_disable_sock(gtp->sk1u); + } } /* UDP encapsulation receive handler. See net/ipv4/udp.c. @@ -645,8 +655,164 @@ static void gtp_link_setup(struct net_device *dev) dev->needed_headroom = LL_MAX_HEADER + max_gtp_header_len; } -static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize); -static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]); +static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize) +{ + int i; + + gtp->addr_hash = kmalloc_array(hsize, sizeof(struct hlist_head), + GFP_KERNEL | __GFP_NOWARN); + if (gtp->addr_hash == NULL) + return -ENOMEM; + + gtp->tid_hash = kmalloc_array(hsize, sizeof(struct hlist_head), + GFP_KERNEL | __GFP_NOWARN); + if (gtp->tid_hash == NULL) + goto err1; + + gtp->hash_size = hsize; + + for (i = 0; i < hsize; i++) { + INIT_HLIST_HEAD(&gtp->addr_hash[i]); + INIT_HLIST_HEAD(&gtp->tid_hash[i]); + } + return 0; +err1: + kfree(gtp->addr_hash); + return -ENOMEM; +} + +static struct sock *gtp_encap_enable_socket(int fd, int type, + struct gtp_dev *gtp) +{ + struct udp_tunnel_sock_cfg tuncfg = {NULL}; + struct socket *sock; + struct sock *sk; + int err; + + pr_debug("enable gtp on %d, %d\n", fd, type); + + sock = sockfd_lookup(fd, &err); + if (!sock) { + pr_debug("gtp socket fd=%d not found\n", fd); + return NULL; + } + + sk = sock->sk; + if (sk->sk_protocol != IPPROTO_UDP || + sk->sk_type != SOCK_DGRAM || + (sk->sk_family != AF_INET && sk->sk_family != AF_INET6)) { + pr_debug("socket fd=%d not UDP\n", fd); + sk = ERR_PTR(-EINVAL); + goto out_sock; + } + + lock_sock(sk); + if (sk->sk_user_data) { + sk = ERR_PTR(-EBUSY); + goto out_rel_sock; + } + + sock_hold(sk); + + tuncfg.sk_user_data = gtp; + tuncfg.encap_type = type; + tuncfg.encap_rcv = gtp_encap_recv; + tuncfg.encap_destroy = gtp_encap_destroy; + + setup_udp_tunnel_sock(sock_net(sock->sk), sock, &tuncfg); + +out_rel_sock: + release_sock(sock->sk); +out_sock: + sockfd_put(sock); + return sk; +} + +static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) +{ + struct sock *sk1u = NULL; + struct sock *sk0 = NULL; + + if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) + return -EINVAL; + + if (data[IFLA_GTP_FD0]) { + u32 fd0 = nla_get_u32(data[IFLA_GTP_FD0]); + + sk0 = gtp_encap_enable_socket(fd0, UDP_ENCAP_GTP0, gtp); + if (IS_ERR(sk0)) + return PTR_ERR(sk0); + } + + if (data[IFLA_GTP_FD1]) { + u32 fd1 = nla_get_u32(data[IFLA_GTP_FD1]); + + sk1u = gtp_encap_enable_socket(fd1, UDP_ENCAP_GTP1U, gtp); + if (IS_ERR(sk1u)) { + gtp_encap_disable_sock(sk0); + return PTR_ERR(sk1u); + } + } + + gtp->sk0 = sk0; + gtp->sk1u = sk1u; + + return 0; +} + +static struct sock *gtp_create_sock(int type, struct gtp_dev *gtp) +{ + struct udp_tunnel_sock_cfg tuncfg = {}; + struct udp_port_cfg udp_conf = { + .local_ip.s_addr = htonl(INADDR_ANY), + .family = AF_INET, + }; + struct net *net = gtp->net; + struct socket *sock; + int err; + + if (type == UDP_ENCAP_GTP0) + udp_conf.local_udp_port = htons(GTP0_PORT); + else if (type == UDP_ENCAP_GTP1U) + udp_conf.local_udp_port = htons(GTP1U_PORT); + else + return ERR_PTR(-EINVAL); + + err = udp_sock_create(net, &udp_conf, &sock); + if (err) + return ERR_PTR(err); + + tuncfg.sk_user_data = gtp; + tuncfg.encap_type = type; + tuncfg.encap_rcv = gtp_encap_recv; + tuncfg.encap_destroy = NULL; + + setup_udp_tunnel_sock(net, sock, &tuncfg); + + return sock->sk; +} + +static int gtp_create_sockets(struct gtp_dev *gtp, struct nlattr *data[]) +{ + struct sock *sk1u = NULL; + struct sock *sk0 = NULL; + + sk0 = gtp_create_sock(UDP_ENCAP_GTP0, gtp); + if (IS_ERR(sk0)) + return PTR_ERR(sk0); + + sk1u = gtp_create_sock(UDP_ENCAP_GTP1U, gtp); + if (IS_ERR(sk1u)) { + udp_tunnel_sock_release(sk0->sk_socket); + return PTR_ERR(sk1u); + } + + gtp->sk_created = true; + gtp->sk0 = sk0; + gtp->sk1u = sk1u; + + return 0; +} static void gtp_destructor(struct net_device *dev) { @@ -660,13 +826,11 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) { + unsigned int role = GTP_ROLE_GGSN; struct gtp_dev *gtp; struct gtp_net *gn; int hashsize, err; - if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) - return -EINVAL; - gtp = netdev_priv(dev); if (!data[IFLA_GTP_PDP_HASHSIZE]) { @@ -677,11 +841,23 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, hashsize = 1024; } + if (data[IFLA_GTP_ROLE]) { + role = nla_get_u32(data[IFLA_GTP_ROLE]); + if (role > GTP_ROLE_SGSN) + return -EINVAL; + } + gtp->role = role; + + gtp->net = src_net; + err = gtp_hashtable_new(gtp, hashsize); if (err < 0) return err; - err = gtp_encap_enable(gtp, data); + if (data[IFLA_GTP_CREATE_SOCKETS]) + err = gtp_create_sockets(gtp, data); + else + err = gtp_encap_enable(gtp, data); if (err < 0) goto out_hashtable; @@ -726,6 +902,7 @@ static const struct nla_policy gtp_policy[IFLA_GTP_MAX + 1] = { [IFLA_GTP_FD1] = { .type = NLA_U32 }, [IFLA_GTP_PDP_HASHSIZE] = { .type = NLA_U32 }, [IFLA_GTP_ROLE] = { .type = NLA_U32 }, + [IFLA_GTP_CREATE_SOCKETS] = { .type = NLA_U8 }, }; static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], @@ -771,119 +948,6 @@ static struct rtnl_link_ops gtp_link_ops __read_mostly = { .fill_info = gtp_fill_info, }; -static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize) -{ - int i; - - gtp->addr_hash = kmalloc_array(hsize, sizeof(struct hlist_head), - GFP_KERNEL | __GFP_NOWARN); - if (gtp->addr_hash == NULL) - return -ENOMEM; - - gtp->tid_hash = kmalloc_array(hsize, sizeof(struct hlist_head), - GFP_KERNEL | __GFP_NOWARN); - if (gtp->tid_hash == NULL) - goto err1; - - gtp->hash_size = hsize; - - for (i = 0; i < hsize; i++) { - INIT_HLIST_HEAD(&gtp->addr_hash[i]); - INIT_HLIST_HEAD(&gtp->tid_hash[i]); - } - return 0; -err1: - kfree(gtp->addr_hash); - return -ENOMEM; -} - -static struct sock *gtp_encap_enable_socket(int fd, int type, - struct gtp_dev *gtp) -{ - struct udp_tunnel_sock_cfg tuncfg = {NULL}; - struct socket *sock; - struct sock *sk; - int err; - - pr_debug("enable gtp on %d, %d\n", fd, type); - - sock = sockfd_lookup(fd, &err); - if (!sock) { - pr_debug("gtp socket fd=%d not found\n", fd); - return NULL; - } - - sk = sock->sk; - if (sk->sk_protocol != IPPROTO_UDP || - sk->sk_type != SOCK_DGRAM || - (sk->sk_family != AF_INET && sk->sk_family != AF_INET6)) { - pr_debug("socket fd=%d not UDP\n", fd); - sk = ERR_PTR(-EINVAL); - goto out_sock; - } - - lock_sock(sk); - if (sk->sk_user_data) { - sk = ERR_PTR(-EBUSY); - goto out_rel_sock; - } - - sock_hold(sk); - - tuncfg.sk_user_data = gtp; - tuncfg.encap_type = type; - tuncfg.encap_rcv = gtp_encap_recv; - tuncfg.encap_destroy = gtp_encap_destroy; - - setup_udp_tunnel_sock(sock_net(sock->sk), sock, &tuncfg); - -out_rel_sock: - release_sock(sock->sk); -out_sock: - sockfd_put(sock); - return sk; -} - -static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) -{ - struct sock *sk1u = NULL; - struct sock *sk0 = NULL; - unsigned int role = GTP_ROLE_GGSN; - - if (data[IFLA_GTP_FD0]) { - u32 fd0 = nla_get_u32(data[IFLA_GTP_FD0]); - - sk0 = gtp_encap_enable_socket(fd0, UDP_ENCAP_GTP0, gtp); - if (IS_ERR(sk0)) - return PTR_ERR(sk0); - } - - if (data[IFLA_GTP_FD1]) { - u32 fd1 = nla_get_u32(data[IFLA_GTP_FD1]); - - sk1u = gtp_encap_enable_socket(fd1, UDP_ENCAP_GTP1U, gtp); - if (IS_ERR(sk1u)) { - gtp_encap_disable_sock(sk0); - return PTR_ERR(sk1u); - } - } - - if (data[IFLA_GTP_ROLE]) { - role = nla_get_u32(data[IFLA_GTP_ROLE]); - if (role > GTP_ROLE_SGSN) { - gtp_encap_disable_sock(sk0); - gtp_encap_disable_sock(sk1u); - return -EINVAL; - } - } - - gtp->sk0 = sk0; - gtp->sk1u = sk1u; - gtp->role = role; - - return 0; -} - static struct gtp_dev *gtp_find_dev(struct net *src_net, struct nlattr *nla[]) { struct gtp_dev *gtp = NULL; diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 6218f93f5c1a..42f3fb097271 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -822,6 +822,7 @@ enum { IFLA_GTP_FD1, IFLA_GTP_PDP_HASHSIZE, IFLA_GTP_ROLE, + IFLA_GTP_CREATE_SOCKETS, __IFLA_GTP_MAX, }; #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1) -- 2.31.1

2 years, 2 months

2
1
0 0

[RFC PATCH net-next v5 3/6] net/sched: Allow flower to match on GTP options

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Options are as follows: PDU_TYPE:QFI and they refernce to the fields from the PDU Session Protocol. PDU Session data is conveyed in GTP-U Extension Header. GTP-U Extension Header is described in 3GPP TS 29.281. PDU Session Protocol is described in 3GPP TS 38.415. PDU_TYPE - indicates the type of the PDU Session Information (4 bits) QFI - QoS Flow Identifier (6 bits) # ip link add gtp_dev type gtp role sgsn # tc qdisc add dev gtp_dev ingress # tc filter add dev gtp_dev protocol ip parent ffff: \ flower \ enc_key_id 11 \ gtp_opts 1:8/ff:ff \ action mirred egress redirect dev eth0 Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- include/net/gtp.h | 5 ++ include/uapi/linux/if_tunnel.h | 4 +- include/uapi/linux/pkt_cls.h | 15 +++++ net/sched/cls_flower.c | 116 +++++++++++++++++++++++++++++++++ 4 files changed, 139 insertions(+), 1 deletion(-) diff --git a/include/net/gtp.h b/include/net/gtp.h index 0e12c37f2958..23c2aaae8a42 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -58,6 +58,11 @@ struct gtp1u_packet { struct gtp_ie ie; } __packed; +struct gtp_pdu_session_info { /* According to 3GPP TS 38.415. */ + u8 pdu_type; + u8 qfi; +}; + #define GTP1_F_NPDU 0x01 #define GTP1_F_SEQ 0x02 #define GTP1_F_EXTHDR 0x04 diff --git a/include/uapi/linux/if_tunnel.h b/include/uapi/linux/if_tunnel.h index 7d9105533c7b..102119628ff5 100644 --- a/include/uapi/linux/if_tunnel.h +++ b/include/uapi/linux/if_tunnel.h @@ -176,8 +176,10 @@ enum { #define TUNNEL_VXLAN_OPT __cpu_to_be16(0x1000) #define TUNNEL_NOCACHE __cpu_to_be16(0x2000) #define TUNNEL_ERSPAN_OPT __cpu_to_be16(0x4000) +#define TUNNEL_GTP_OPT __cpu_to_be16(0x8000) #define TUNNEL_OPTIONS_PRESENT \ - (TUNNEL_GENEVE_OPT | TUNNEL_VXLAN_OPT | TUNNEL_ERSPAN_OPT) + (TUNNEL_GENEVE_OPT | TUNNEL_VXLAN_OPT | TUNNEL_ERSPAN_OPT | \ + TUNNEL_GTP_OPT) #endif /* _UAPI_IF_TUNNEL_H_ */ diff --git a/include/uapi/linux/pkt_cls.h b/include/uapi/linux/pkt_cls.h index ee38b35c3f57..404f97fb239c 100644 --- a/include/uapi/linux/pkt_cls.h +++ b/include/uapi/linux/pkt_cls.h @@ -616,6 +616,10 @@ enum { * TCA_FLOWER_KEY_ENC_OPT_ERSPAN_ * attributes */ + TCA_FLOWER_KEY_ENC_OPTS_GTP, /* Nested + * TCA_FLOWER_KEY_ENC_OPT_GTP_ + * attributes + */ __TCA_FLOWER_KEY_ENC_OPTS_MAX, }; @@ -654,6 +658,17 @@ enum { #define TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX \ (__TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX - 1) +enum { + TCA_FLOWER_KEY_ENC_OPT_GTP_UNSPEC, + TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE, /* u8 */ + TCA_FLOWER_KEY_ENC_OPT_GTP_QFI, /* u8 */ + + __TCA_FLOWER_KEY_ENC_OPT_GTP_MAX, +}; + +#define TCA_FLOWER_KEY_ENC_OPT_GTP_MAX \ + (__TCA_FLOWER_KEY_ENC_OPT_GTP_MAX - 1) + enum { TCA_FLOWER_KEY_MPLS_OPTS_UNSPEC, TCA_FLOWER_KEY_MPLS_OPTS_LSE, diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c index 1a9b1f140f9e..c80fc49c0da1 100644 --- a/net/sched/cls_flower.c +++ b/net/sched/cls_flower.c @@ -25,6 +25,7 @@ #include <net/geneve.h> #include <net/vxlan.h> #include <net/erspan.h> +#include <net/gtp.h> #include <net/dst.h> #include <net/dst_metadata.h> @@ -723,6 +724,7 @@ enc_opts_policy[TCA_FLOWER_KEY_ENC_OPTS_MAX + 1] = { [TCA_FLOWER_KEY_ENC_OPTS_GENEVE] = { .type = NLA_NESTED }, [TCA_FLOWER_KEY_ENC_OPTS_VXLAN] = { .type = NLA_NESTED }, [TCA_FLOWER_KEY_ENC_OPTS_ERSPAN] = { .type = NLA_NESTED }, + [TCA_FLOWER_KEY_ENC_OPTS_GTP] = { .type = NLA_NESTED }, }; static const struct nla_policy @@ -746,6 +748,12 @@ erspan_opt_policy[TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX + 1] = { [TCA_FLOWER_KEY_ENC_OPT_ERSPAN_HWID] = { .type = NLA_U8 }, }; +static const struct nla_policy +gtp_opt_policy[TCA_FLOWER_KEY_ENC_OPT_GTP_MAX + 1] = { + [TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE] = { .type = NLA_U8 }, + [TCA_FLOWER_KEY_ENC_OPT_GTP_QFI] = { .type = NLA_U8 }, +}; + static const struct nla_policy mpls_stack_entry_policy[TCA_FLOWER_KEY_MPLS_OPT_LSE_MAX + 1] = { [TCA_FLOWER_KEY_MPLS_OPT_LSE_DEPTH] = { .type = NLA_U8 }, @@ -1262,6 +1270,49 @@ static int fl_set_erspan_opt(const struct nlattr *nla, struct fl_flow_key *key, return sizeof(*md); } +static int fl_set_gtp_opt(const struct nlattr *nla, struct fl_flow_key *key, + int depth, int option_len, + struct netlink_ext_ack *extack) +{ + struct nlattr *tb[TCA_FLOWER_KEY_ENC_OPT_GTP_MAX + 1]; + struct gtp_pdu_session_info *sinfo; + u8 len = key->enc_opts.len; + int err; + + sinfo = (struct gtp_pdu_session_info *)&key->enc_opts.data[len]; + memset(sinfo, 0xff, option_len); + + if (!depth) + return sizeof(*sinfo); + + if (nla_type(nla) != TCA_FLOWER_KEY_ENC_OPTS_GTP) { + NL_SET_ERR_MSG_MOD(extack, "Non-gtp option type for mask"); + return -EINVAL; + } + + err = nla_parse_nested(tb, TCA_FLOWER_KEY_ENC_OPT_GTP_MAX, nla, + gtp_opt_policy, extack); + if (err < 0) + return err; + + if (!option_len && + (!tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE] || + !tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI])) { + NL_SET_ERR_MSG_MOD(extack, + "Missing tunnel key gtp option pdu type or qfi"); + return -EINVAL; + } + + if (tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE]) + sinfo->pdu_type = + nla_get_u8(tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE]); + + if (tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI]) + sinfo->qfi = nla_get_u8(tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI]); + + return sizeof(*sinfo); +} + static int fl_set_enc_opt(struct nlattr **tb, struct fl_flow_key *key, struct fl_flow_key *mask, struct netlink_ext_ack *extack) @@ -1386,6 +1437,38 @@ static int fl_set_enc_opt(struct nlattr **tb, struct fl_flow_key *key, return -EINVAL; } break; + case TCA_FLOWER_KEY_ENC_OPTS_GTP: + if (key->enc_opts.dst_opt_type) { + NL_SET_ERR_MSG_MOD(extack, + "Duplicate type for gtp options"); + return -EINVAL; + } + option_len = 0; + key->enc_opts.dst_opt_type = TUNNEL_GTP_OPT; + option_len = fl_set_gtp_opt(nla_opt_key, key, + key_depth, option_len, + extack); + if (option_len < 0) + return option_len; + + key->enc_opts.len += option_len; + /* At the same time we need to parse through the mask + * in order to verify exact and mask attribute lengths. + */ + mask->enc_opts.dst_opt_type = TUNNEL_GTP_OPT; + option_len = fl_set_gtp_opt(nla_opt_msk, mask, + msk_depth, option_len, + extack); + if (option_len < 0) + return option_len; + + mask->enc_opts.len += option_len; + if (key->enc_opts.len != mask->enc_opts.len) { + NL_SET_ERR_MSG_MOD(extack, + "Key and mask miss aligned"); + return -EINVAL; + } + break; default: NL_SET_ERR_MSG(extack, "Unknown tunnel option type"); return -EINVAL; @@ -2761,6 +2844,34 @@ static int fl_dump_key_erspan_opt(struct sk_buff *skb, return -EMSGSIZE; } +static int fl_dump_key_gtp_opt(struct sk_buff *skb, + struct flow_dissector_key_enc_opts *enc_opts) + +{ + struct gtp_pdu_session_info *session_info; + struct nlattr *nest; + + nest = nla_nest_start_noflag(skb, TCA_FLOWER_KEY_ENC_OPTS_GTP); + if (!nest) + goto nla_put_failure; + + session_info = (struct gtp_pdu_session_info *)&enc_opts->data[0]; + + if (nla_put_u8(skb, TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE, + session_info->pdu_type)) + goto nla_put_failure; + + if (nla_put_u8(skb, TCA_FLOWER_KEY_ENC_OPT_GTP_QFI, session_info->qfi)) + goto nla_put_failure; + + nla_nest_end(skb, nest); + return 0; + +nla_put_failure: + nla_nest_cancel(skb, nest); + return -EMSGSIZE; +} + static int fl_dump_key_ct(struct sk_buff *skb, struct flow_dissector_key_ct *key, struct flow_dissector_key_ct *mask) @@ -2824,6 +2935,11 @@ static int fl_dump_key_options(struct sk_buff *skb, int enc_opt_type, if (err) goto nla_put_failure; break; + case TUNNEL_GTP_OPT: + err = fl_dump_key_gtp_opt(skb, enc_opts); + if (err) + goto nla_put_failure; + break; default: goto nla_put_failure; } -- 2.31.1

2 years, 2 months

1
0
0 0

[RFC PATCH net-next v4 3/6] net/sched: Allow flower to match on GTP options

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Options are as follows: PDU_TYPE:QFI and they refernce to the fields from the PDU Session Protocol. PDU Session data is conveyed in GTP-U Extension Header. GTP-U Extension Header is described in 3GPP TS 29.281. PDU Session Protocol is described in 3GPP TS 38.415. PDU_TYPE - indicates the type of the PDU Session Information (4 bits) QFI - QoS Flow Identifier (6 bits) # ip link add gtp_dev type gtp role sgsn # tc qdisc add dev gtp_dev ingress # tc filter add dev gtp_dev protocol ip parent ffff: \ flower \ enc_key_id 11 \ gtp_opts 1:8/ff:ff \ action mirred egress redirect dev eth0 Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- include/net/gtp.h | 5 ++ include/uapi/linux/if_tunnel.h | 4 +- include/uapi/linux/pkt_cls.h | 15 +++++ net/sched/cls_flower.c | 116 +++++++++++++++++++++++++++++++++ 4 files changed, 139 insertions(+), 1 deletion(-) diff --git a/include/net/gtp.h b/include/net/gtp.h index ae915dd33d20..c78702e3d663 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -27,6 +27,11 @@ struct gtp1_header { /* According to 3GPP TS 29.060. */ __be32 tid; } __attribute__ ((packed)); +struct gtp_pdu_session_info { /* According to 3GPP TS 38.415. */ + u8 pdu_type; + u8 qfi; +}; + static inline bool netif_is_gtp(const struct net_device *dev) { return dev->rtnl_link_ops && diff --git a/include/uapi/linux/if_tunnel.h b/include/uapi/linux/if_tunnel.h index 7d9105533c7b..102119628ff5 100644 --- a/include/uapi/linux/if_tunnel.h +++ b/include/uapi/linux/if_tunnel.h @@ -176,8 +176,10 @@ enum { #define TUNNEL_VXLAN_OPT __cpu_to_be16(0x1000) #define TUNNEL_NOCACHE __cpu_to_be16(0x2000) #define TUNNEL_ERSPAN_OPT __cpu_to_be16(0x4000) +#define TUNNEL_GTP_OPT __cpu_to_be16(0x8000) #define TUNNEL_OPTIONS_PRESENT \ - (TUNNEL_GENEVE_OPT | TUNNEL_VXLAN_OPT | TUNNEL_ERSPAN_OPT) + (TUNNEL_GENEVE_OPT | TUNNEL_VXLAN_OPT | TUNNEL_ERSPAN_OPT | \ + TUNNEL_GTP_OPT) #endif /* _UAPI_IF_TUNNEL_H_ */ diff --git a/include/uapi/linux/pkt_cls.h b/include/uapi/linux/pkt_cls.h index ee38b35c3f57..404f97fb239c 100644 --- a/include/uapi/linux/pkt_cls.h +++ b/include/uapi/linux/pkt_cls.h @@ -616,6 +616,10 @@ enum { * TCA_FLOWER_KEY_ENC_OPT_ERSPAN_ * attributes */ + TCA_FLOWER_KEY_ENC_OPTS_GTP, /* Nested + * TCA_FLOWER_KEY_ENC_OPT_GTP_ + * attributes + */ __TCA_FLOWER_KEY_ENC_OPTS_MAX, }; @@ -654,6 +658,17 @@ enum { #define TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX \ (__TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX - 1) +enum { + TCA_FLOWER_KEY_ENC_OPT_GTP_UNSPEC, + TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE, /* u8 */ + TCA_FLOWER_KEY_ENC_OPT_GTP_QFI, /* u8 */ + + __TCA_FLOWER_KEY_ENC_OPT_GTP_MAX, +}; + +#define TCA_FLOWER_KEY_ENC_OPT_GTP_MAX \ + (__TCA_FLOWER_KEY_ENC_OPT_GTP_MAX - 1) + enum { TCA_FLOWER_KEY_MPLS_OPTS_UNSPEC, TCA_FLOWER_KEY_MPLS_OPTS_LSE, diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c index 1a9b1f140f9e..c80fc49c0da1 100644 --- a/net/sched/cls_flower.c +++ b/net/sched/cls_flower.c @@ -25,6 +25,7 @@ #include <net/geneve.h> #include <net/vxlan.h> #include <net/erspan.h> +#include <net/gtp.h> #include <net/dst.h> #include <net/dst_metadata.h> @@ -723,6 +724,7 @@ enc_opts_policy[TCA_FLOWER_KEY_ENC_OPTS_MAX + 1] = { [TCA_FLOWER_KEY_ENC_OPTS_GENEVE] = { .type = NLA_NESTED }, [TCA_FLOWER_KEY_ENC_OPTS_VXLAN] = { .type = NLA_NESTED }, [TCA_FLOWER_KEY_ENC_OPTS_ERSPAN] = { .type = NLA_NESTED }, + [TCA_FLOWER_KEY_ENC_OPTS_GTP] = { .type = NLA_NESTED }, }; static const struct nla_policy @@ -746,6 +748,12 @@ erspan_opt_policy[TCA_FLOWER_KEY_ENC_OPT_ERSPAN_MAX + 1] = { [TCA_FLOWER_KEY_ENC_OPT_ERSPAN_HWID] = { .type = NLA_U8 }, }; +static const struct nla_policy +gtp_opt_policy[TCA_FLOWER_KEY_ENC_OPT_GTP_MAX + 1] = { + [TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE] = { .type = NLA_U8 }, + [TCA_FLOWER_KEY_ENC_OPT_GTP_QFI] = { .type = NLA_U8 }, +}; + static const struct nla_policy mpls_stack_entry_policy[TCA_FLOWER_KEY_MPLS_OPT_LSE_MAX + 1] = { [TCA_FLOWER_KEY_MPLS_OPT_LSE_DEPTH] = { .type = NLA_U8 }, @@ -1262,6 +1270,49 @@ static int fl_set_erspan_opt(const struct nlattr *nla, struct fl_flow_key *key, return sizeof(*md); } +static int fl_set_gtp_opt(const struct nlattr *nla, struct fl_flow_key *key, + int depth, int option_len, + struct netlink_ext_ack *extack) +{ + struct nlattr *tb[TCA_FLOWER_KEY_ENC_OPT_GTP_MAX + 1]; + struct gtp_pdu_session_info *sinfo; + u8 len = key->enc_opts.len; + int err; + + sinfo = (struct gtp_pdu_session_info *)&key->enc_opts.data[len]; + memset(sinfo, 0xff, option_len); + + if (!depth) + return sizeof(*sinfo); + + if (nla_type(nla) != TCA_FLOWER_KEY_ENC_OPTS_GTP) { + NL_SET_ERR_MSG_MOD(extack, "Non-gtp option type for mask"); + return -EINVAL; + } + + err = nla_parse_nested(tb, TCA_FLOWER_KEY_ENC_OPT_GTP_MAX, nla, + gtp_opt_policy, extack); + if (err < 0) + return err; + + if (!option_len && + (!tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE] || + !tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI])) { + NL_SET_ERR_MSG_MOD(extack, + "Missing tunnel key gtp option pdu type or qfi"); + return -EINVAL; + } + + if (tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE]) + sinfo->pdu_type = + nla_get_u8(tb[TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE]); + + if (tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI]) + sinfo->qfi = nla_get_u8(tb[TCA_FLOWER_KEY_ENC_OPT_GTP_QFI]); + + return sizeof(*sinfo); +} + static int fl_set_enc_opt(struct nlattr **tb, struct fl_flow_key *key, struct fl_flow_key *mask, struct netlink_ext_ack *extack) @@ -1386,6 +1437,38 @@ static int fl_set_enc_opt(struct nlattr **tb, struct fl_flow_key *key, return -EINVAL; } break; + case TCA_FLOWER_KEY_ENC_OPTS_GTP: + if (key->enc_opts.dst_opt_type) { + NL_SET_ERR_MSG_MOD(extack, + "Duplicate type for gtp options"); + return -EINVAL; + } + option_len = 0; + key->enc_opts.dst_opt_type = TUNNEL_GTP_OPT; + option_len = fl_set_gtp_opt(nla_opt_key, key, + key_depth, option_len, + extack); + if (option_len < 0) + return option_len; + + key->enc_opts.len += option_len; + /* At the same time we need to parse through the mask + * in order to verify exact and mask attribute lengths. + */ + mask->enc_opts.dst_opt_type = TUNNEL_GTP_OPT; + option_len = fl_set_gtp_opt(nla_opt_msk, mask, + msk_depth, option_len, + extack); + if (option_len < 0) + return option_len; + + mask->enc_opts.len += option_len; + if (key->enc_opts.len != mask->enc_opts.len) { + NL_SET_ERR_MSG_MOD(extack, + "Key and mask miss aligned"); + return -EINVAL; + } + break; default: NL_SET_ERR_MSG(extack, "Unknown tunnel option type"); return -EINVAL; @@ -2761,6 +2844,34 @@ static int fl_dump_key_erspan_opt(struct sk_buff *skb, return -EMSGSIZE; } +static int fl_dump_key_gtp_opt(struct sk_buff *skb, + struct flow_dissector_key_enc_opts *enc_opts) + +{ + struct gtp_pdu_session_info *session_info; + struct nlattr *nest; + + nest = nla_nest_start_noflag(skb, TCA_FLOWER_KEY_ENC_OPTS_GTP); + if (!nest) + goto nla_put_failure; + + session_info = (struct gtp_pdu_session_info *)&enc_opts->data[0]; + + if (nla_put_u8(skb, TCA_FLOWER_KEY_ENC_OPT_GTP_PDU_TYPE, + session_info->pdu_type)) + goto nla_put_failure; + + if (nla_put_u8(skb, TCA_FLOWER_KEY_ENC_OPT_GTP_QFI, session_info->qfi)) + goto nla_put_failure; + + nla_nest_end(skb, nest); + return 0; + +nla_put_failure: + nla_nest_cancel(skb, nest); + return -EMSGSIZE; +} + static int fl_dump_key_ct(struct sk_buff *skb, struct flow_dissector_key_ct *key, struct flow_dissector_key_ct *mask) @@ -2824,6 +2935,11 @@ static int fl_dump_key_options(struct sk_buff *skb, int enc_opt_type, if (err) goto nla_put_failure; break; + case TUNNEL_GTP_OPT: + err = fl_dump_key_gtp_opt(skb, enc_opts); + if (err) + goto nla_put_failure; + break; default: goto nla_put_failure; } -- 2.31.1

2 years, 2 months

1
0
0 0

[RFC PATCH net-next v4 2/6] gtp: Add support for checking GTP device type

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Add a function that checks if a net device type is GTP. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- include/net/gtp.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/net/gtp.h b/include/net/gtp.h index 0e16ebb2a82d..ae915dd33d20 100644 --- a/include/net/gtp.h +++ b/include/net/gtp.h @@ -27,6 +27,12 @@ struct gtp1_header { /* According to 3GPP TS 29.060. */ __be32 tid; } __attribute__ ((packed)); +static inline bool netif_is_gtp(const struct net_device *dev) +{ + return dev->rtnl_link_ops && + !strcmp(dev->rtnl_link_ops->kind, "gtp"); +} + #define GTP1_F_NPDU 0x01 #define GTP1_F_SEQ 0x02 #define GTP1_F_EXTHDR 0x04 -- 2.31.1

2 years, 2 months

1
0
0 0

[RFC PATCH net-next v4 1/6] gtp: Allow to create GTP device without FDs

by Marcin Szycik

From: Wojciech Drewek <wojciech.drewek(a)intel.com> Currently, when the user wants to create GTP device, he has to provide file handles to the sockets created in userspace (IFLA_GTP_FD0, IFLA_GTP_FD1). This behaviour is not ideal, considering the option of adding support for GTP device creation through ip link. Ip link application is not a good place to create such sockets. This patch allows to create GTP device without providing IFLA_GTP_FD0 and IFLA_GTP_FD1 arguments. If the user does not provide file handles to the sockets, then GTP module takes care of creating UDP sockets by itself. Sockets are created with the commonly known UDP ports used for GTP protocol (GTP0_PORT and GTP1U_PORT). In this case we don't have to provide encap_destroy because no extra deinitialization is needed, everything is covered by udp_tunnel_sock_release. Signed-off-by: Wojciech Drewek <wojciech.drewek(a)intel.com> --- v4: use ntohs when creating UDP socket --- drivers/net/gtp.c | 74 ++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 67 insertions(+), 7 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 24e5c54d06c1..6fa1cfe023ef 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -66,8 +66,10 @@ struct gtp_dev { struct sock *sk0; struct sock *sk1u; + u8 sk_created; struct net_device *dev; + struct net *net; unsigned int role; unsigned int hash_size; @@ -320,8 +322,16 @@ static void gtp_encap_disable_sock(struct sock *sk) static void gtp_encap_disable(struct gtp_dev *gtp) { - gtp_encap_disable_sock(gtp->sk0); - gtp_encap_disable_sock(gtp->sk1u); + if (gtp->sk_created) { + udp_tunnel_sock_release(gtp->sk0->sk_socket); + udp_tunnel_sock_release(gtp->sk1u->sk_socket); + gtp->sk_created = false; + gtp->sk0 = NULL; + gtp->sk1u = NULL; + } else { + gtp_encap_disable_sock(gtp->sk0); + gtp_encap_disable_sock(gtp->sk1u); + } } /* UDP encapsulation receive handler. See net/ipv4/udp.c. @@ -664,9 +674,6 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, struct gtp_net *gn; int hashsize, err; - if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) - return -EINVAL; - gtp = netdev_priv(dev); if (!data[IFLA_GTP_PDP_HASHSIZE]) { @@ -677,6 +684,8 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, hashsize = 1024; } + gtp->net = src_net; + err = gtp_hashtable_new(gtp, hashsize); if (err < 0) return err; @@ -844,6 +853,38 @@ static struct sock *gtp_encap_enable_socket(int fd, int type, return sk; } +static struct sock *gtp_encap_create_sock(int type, struct gtp_dev *gtp) +{ + struct udp_tunnel_sock_cfg tuncfg = {}; + struct udp_port_cfg udp_conf = { + .local_ip.s_addr = htonl(INADDR_ANY), + .family = AF_INET, + }; + struct net *net = gtp->net; + struct socket *sock; + int err; + + if (type == UDP_ENCAP_GTP0) + udp_conf.local_udp_port = ntohs(GTP0_PORT); + else if (type == UDP_ENCAP_GTP1U) + udp_conf.local_udp_port = ntohs(GTP1U_PORT); + else + return ERR_PTR(-EINVAL); + + err = udp_sock_create(net, &udp_conf, &sock); + if (err) + return ERR_PTR(err); + + tuncfg.sk_user_data = gtp; + tuncfg.encap_type = type; + tuncfg.encap_rcv = gtp_encap_recv; + tuncfg.encap_destroy = NULL; + + setup_udp_tunnel_sock(net, sock, &tuncfg); + + return sock->sk; +} + static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) { struct sock *sk1u = NULL; @@ -868,11 +909,30 @@ static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) } } + if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1]) { + sk0 = gtp_encap_create_sock(UDP_ENCAP_GTP0, gtp); + if (IS_ERR(sk0)) + return PTR_ERR(sk0); + + sk1u = gtp_encap_create_sock(UDP_ENCAP_GTP1U, gtp); + if (IS_ERR(sk1u)) { + udp_tunnel_sock_release(sk0->sk_socket); + return PTR_ERR(sk1u); + } + gtp->sk_created = true; + } + if (data[IFLA_GTP_ROLE]) { role = nla_get_u32(data[IFLA_GTP_ROLE]); if (role > GTP_ROLE_SGSN) { - gtp_encap_disable_sock(sk0); - gtp_encap_disable_sock(sk1u); + if (gtp->sk_created) { + udp_tunnel_sock_release(sk0->sk_socket); + udp_tunnel_sock_release(sk1u->sk_socket); + gtp->sk_created = false; + } else { + gtp_encap_disable_sock(sk0); + gtp_encap_disable_sock(sk1u); + } return -EINVAL; } } -- 2.31.1

2 years, 2 months

1
0
0 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

osmocom-net-gprs February 2022