- osmocom-net-gprs - lists.osmocom.org

[PATCH net-next v6 2/7] gtp: Implement GTP echo response
by Marcin Szycik 18 Feb '22

18 Feb '22

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

1 0

[PATCH net-next v6 1/7] gtp: Allow to create GTP device without FDs
by Marcin Szycik 18 Feb '22

18 Feb '22

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

1 0

[PATCH net-next v6 1/7] gtp: Allow to create GTP device without FDs
by Marcin Szycik 18 Feb '22

18 Feb '22

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

1 0

[RFC PATCH net-next v4 4/6] gtp: Implement GTP echo response
by Marcin Szycik 12 Feb '22

12 Feb '22

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

4 6

[RFC PATCH net-next v3 1/5] gtp: Allow to create GTP device without FDs
by Marcin Szycik 12 Feb '22

12 Feb '22

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

4 7

[RFC PATCH net-next v5 4/6] gtp: Add support for checking GTP device type
by Marcin Szycik 12 Feb '22

12 Feb '22

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 1

[RFC PATCH net-next v5 2/6] gtp: Implement GTP echo response
by Marcin Szycik 12 Feb '22

12 Feb '22

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 1

[RFC PATCH net-next v5 1/6] gtp: Allow to create GTP device without FDs
by Marcin Szycik 12 Feb '22

12 Feb '22

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 1

[RFC PATCH net-next v5 3/6] net/sched: Allow flower to match on GTP options
by Marcin Szycik 11 Feb '22

11 Feb '22

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

1 0

[RFC PATCH net-next v4 3/6] net/sched: Allow flower to match on GTP options
by Marcin Szycik 04 Feb '22

04 Feb '22

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

1 0

[RFC PATCH net-next v4 2/6] gtp: Add support for checking GTP device type
by Marcin Szycik 04 Feb '22

04 Feb '22

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

1 0

[RFC PATCH net-next v4 1/6] gtp: Allow to create GTP device without FDs
by Marcin Szycik 04 Feb '22

04 Feb '22

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

1 0

[RFC PATCH net-next v2 1/5] gtp: Allow to create GTP device without FDs
by Marcin Szycik 27 Jan '22

27 Jan '22

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

3 2

[RFC PATCH net-next v3 2/5] gtp: Add support for checking GTP device type
by Marcin Szycik 27 Jan '22

27 Jan '22

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

1 0

[RFC PATCH net-next v2 2/5] gtp: Add support for checking GTP device type
by Marcin Szycik 27 Jan '22

27 Jan '22

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

1 0

lists.osmocom.org migrated to mailman3
by Harald Welte 03 Jan '22

03 Jan '22

Dear Osmocom community, today our mailing list server lists.osmocom.org has finally been migrated from mailman2-on-freebsd to mailman3-on-linux. This also included a variety of changes to DNS. I'll spare you the details, but everything _should_ be up and running now. * The List-Id headers should not have changed. * all list subscriptions + user accounts have been converted. * old 'static html' archives are still available (read only) at URLs like https://lists.osmocom.org/pipermail/baseband-devel/ * old List URLs like https://lists.osmocom.org/mailman/listinfo/baseband-devel are redirected to their respective modern counterparts In case you notice any mailing list related problem, please don't hesitate to contact me. Happy hacking, 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)

1 0

None
by None 03 Jan '22

03 Jan '22

RLC/MAC layer in real networks is always used in acknowledged/reliable mode, and LLC is always only used in unreliable (UI) mode. The reason for this is that TCP cannot deal with the enormous packet loss that GPRS or similar radio links pose. If TCP would be exposed to the real packet loss, it would stall way too much. So my rationale was to use the same while creating osmo-sgsn. We use LLC in UI mode. Or are we talking about something completely else? Regards, Harald -- - Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

1 0

None
by None 03 Jan '22

03 Jan '22

* With multiple TBF assignment the tbf_by_tlli will return a list =A0 anyway. * For the multiple RACH cases we could it probably handle it more =A0 nicely. E.g. avoid this: <0005> bts.cpp:1035 Got RACH from TLLI=3D0xcf7611f6 while TBF(TFI=3D0= TLLI=3D0xcf7611f6 DIR=3DDL) still exists. Killing pending DL TBF <0002> bts.cpp:359 Got IMM.ASS confirm, but TLLI=3Dcf7611f6 does not = exit -- - Holger Freyther <<a href=3D"mailto:hfreyther@sysmocom.de">hfreyther@sy= smocom.de</a>> =A0 =A0 =A0 <a href=3D"http://www.sysmocom.de/" target=3D= "_blank">http://www.sysmocom.de/</a> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D * sysmocom - systems for mobile communications GmbH * Schivelbeiner Str. 5 * 10439 Berlin, Germany * Sitz / Registered office: Berlin, HRB 134158 B * Geschaeftsfuehrer / Managing Directors: Holger Freyther, Harald Welte </blockquote></div> </div> --20cf301d42625eb38c04ea05cbef--

1 0

None
by None 03 Jan '22

03 Jan '22

* With multiple TBF assignment the tbf_by_tlli will return a list anyway. * For the multiple RACH cases we could it probably handle it more nicely. E.g. avoid this: <0005> bts.cpp:1035 Got RACH from TLLI=0xcf7611f6 while TBF(TFI=0 TLLI=0xcf7611f6 DIR=DL) still exists. Killing pending DL TBF <0002> bts.cpp:359 Got IMM.ASS confirm, but TLLI=cf7611f6 does not exit -- - Holger Freyther <hfreyther(a)sysmocom.de> http://www.sysmocom.de/ ======================================================================= * sysmocom - systems for mobile communications GmbH * Schivelbeiner Str. 5 * 10439 Berlin, Germany * Sitz / Registered office: Berlin, HRB 134158 B * Geschaeftsfuehrer / Managing Directors: Holger Freyther, Harald Welte

1 0

None
by None 03 Jan '22

03 Jan '22

if we should use more C++ features (I am not talking STL, operator overloading)? Any opinions? holger PS: I think we should find a more nice solution to extern "C".. I spent various minutes trying to figure out which parts of the pcu have C linkage and which C++.. For things from libosmocore it is obvious.. but then we have the vty_info... -- - Holger Freyther <hfreyther(a)sysmocom.de> http://www.sysmocom.de/ ======================================================================= * sysmocom - systems for mobile communications GmbH * Schivelbeiner Str. 5 * 10439 Berlin, Germany * Sitz / Registered office: Berlin, HRB 134158 B * Geschaeftsfuehrer / Managing Directors: Holger Freyther, Harald Welte

1 0

None
by None 03 Jan '22

03 Jan '22

bug fixes. Ideally I'd love to see Andreas developing them on top of the current 'master' and submitting them to the mailing list as soon as he push them. This way Ivan (and everyone else) would be able to review them and merge as soon as possible. I hate the situation when fixes are stored in a private branch and are not merged into mainline. On Mon, Nov 26, 2012 at 5:02 PM, Ivan Kluchnikov <Ivan.Kluchnikov(a)fairwaves.ru> wrote: > Hi, all! > > I absolutely agree with Harald and I am ready to use described life-cycle > model for osmo-pcu. > Before that time I merged branch 'jolly', when the differences between > master and jolly become critical. > I didn't coordinate merges with Andreas and jolly branch wasn't rebased > before merges. I think, it was my mistake. > Now I am ready to merge jolly branch at any time, when Andreas indicates > that his branch is ready for merge. > > > 2012/11/25 Harald Welte <laforge(a)gnumonks.org> >> >> Dear all, >> >> it is a pity to see that since many weeks there has been no progress on >> merging any of Andreas' work. I think everyone involved is to be >> blamed to some extent. Andreas because the provided branches are not >> kept clean, Ivan probably for lack of time, and me for not paying more >> attention and helping this process along. >> >> The general life-cycle in any project, including osmo-pcu, should be as >> follows: >> >> * contributor starts on some improvements, creates private branch >> * master advances meanwhile >> * contributor rebases (not merges!) his changes on top of more recent >> master >> * contributor indicates that his branch is ready for merge >> * maintainer merges (some of?) the commits of the contributor branch >> * contributor re-bases remaining commits on new master >> * contributor requests merge of remaing patches >> ... >> >> If I open 'gitk' on any of the branches (master / jolly / jolly_dsp) I >> get grey hair (to say the least). One would normally expect: >> >> 1) 'jolly' to be based on some (maybe not current) master >> >> 2) 'jolly_dsp' to be based on 'jolly' >> >> Neither of the two is the case. I've tried to >> * rebase jolly on top of master >> as well as >> * rebase jolly_dsp on top of jolly >> >> and both are impossible due to the series of merges and the unclear >> ancestry / relationship of the branches. Or, just for fun, try 'git >> diff origin/jolly..origin/jolly_dsp. It shows you anything else but >> what one would expect >> >> Andreas: Please take some time to clean up the mess. Your 'jolly' >> branch should contain a series of clean per-feature commit's on top of >> current master, and 'jolly_dsp' should be a set of few patches on top of >> 'jolly'. >> >> This way there is always a clean queue of changesets that Ivan can merge >> (or cherry-pick). I can understand that if I was in Ivan's place, I >> would not really know where to even start merging some of those >> contributions. >> >> Thanks for everyone's attention. Please take some time to get this >> resolved. Let me know if you have some questions. There are plenty of >> people with lots of git experience around (Holger, Pablo, myself) who >> can help if something is unclear. >> >> Regards, >> Harald >> >> -- >> - Harald Welte <laforge(a)gnumonks.org> >> http://laforge.gnumonks.org/ >> >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D >> "Privacy in residential applications is a desirable marketing option." >> (ETSI EN 300 175-7 Ch. >> A6) >> > > > > -- > Regards, > Ivan Kluchnikov. > http://fairwaves.ru > --=20 Regards, Alexander Chemeris. CEO, Fairwaves LLC / =D0=9E=D0=9E=D0=9E =D0=A3=D0=BC=D0=A0=D0=B0=D0=B4=D0= =B8=D0=BE http://fairwaves.ru

1 0

Osmocom CNI release 202111
by Pau Espin Pedrol 19 Nov '21

19 Nov '21

Dear all, I am pleased to announce new tagged releases for Osmocom Cellular Network Infrastructure components. Find more information about the release here [1]. Osmocom "Latest" repositories in OBS [2] should already contain packages for the new versions. OpenEmbedded related meta-layers such as meta-telephony usual stable/testing branch "201705" [3] have also been updated to build recipes for new versions. Regards, Pau [1] https://osmocom.org/news/152 [2] https://osmocom.org/projects/cellular-infrastructure/wiki/Latest_Builds [3] https://git.osmocom.org/meta-telephony/log/?h=201705 -- - Pau Espin Pedrol <pespin(a)sysmocom.de> http://www.sysmocom.de/ ======================================================================= * sysmocom - systems for mobile communications GmbH * Alt-Moabit 93 * 10559 Berlin, Germany * Sitz / Registered office: Berlin, HRB 134158 B * Geschaeftsfuehrer / Managing Director: Harald Welte

2 1

TTCN-3 tests for MME
by Mirko Kovacevic 29 Oct '21

29 Oct '21

Hi, Osmonians, I don't know where to write for MME TTCN-3, so I chose to write to it closet relative - SGSN. I was trying to do slight modification to TTCN-3 code, modifing S1AP_Templates.ttcn. I embedded enodeb name for all tests with new protoc-IE = id-eNBname : ################################## template (value) S1AP_PDU ts_S1AP_SetupReq(template (value) Global_ENB_ID p_global_ENB_ID, template (value) SupportedTAs p_supportedTAs, template (value) PagingDRX p_pagingDRXs) := { initiatingMessage := { procedureCode := id_S1Setup, criticality := reject, value_ := { S1SetupRequest := { protocolIEs := { { id := S1AP_Constants.id_Global_ENB_ID, criticality := ignore, value_ := { Global_ENB_ID := p_global_ENB_ID } }, { id := S1AP_Constants.id_SupportedTAs, criticality := reject, value_ := {SupportedTAs := p_supportedTAs} }, { id := S1AP_Constants.id_eNBname,//id-eNBname criticality := ignore, value_ := {ENBname := "OsmoEnodeb"} } /* HACK: work around nextepc bug , { id := S1AP_Constants.id_pagingDRX, criticality := ignore, value_ := {PagingDRX := p_pagingDRXs} } */ } } } } } ######################## Code compiles, but PDU is not encoded well. I attached Wireshark preview. Can you help me with this issue ? -- Puno pozdrava, Mirko Kovacevic

3 6

Osmocom redmine upgrade / Let me know if something broke
by Harald Welte 27 Oct '21

27 Oct '21

Dear Osmocom community, today we finally upgraded our redmine installation from the unmaintained 3.4.x to the latest 4.2.3. The upgrade had been overdue for years, but today we (actually, Kevin) finally managed to find out how to make the openid_provider plugin to work with modern rails 5.x. In any case, I'm just informing you in case there is some unexpected fallout. I've verified that at least the following appears working: * logging into redmine * openid authentication from gerrit.osmocom.org (also after logout/relogin) * graphviz rendering * mscgen rendering There could however still be unexpected problems, particularly in features such as * e-mail notifications from redmine * updating redmine issues via e-mail * git integration ("Closes: OS#xxxx") If you run into any troubles, pleas report to https://osmocom.org/projects/osmocom-servers or if that also fails, feel free to send e-mails. 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)

1 0

[PATCH net-next] gtp: use skb_dst_update_pmtu_no_confirm() instead of direct call
by Gyeongun Kang 06 Oct '21

06 Oct '21

skb_dst_update_pmtu_no_confirm() is a just wrapper function of ->update_pmtu(). So, it doesn't change logic Signed-off-by: Gyeongun Kang <kyeongun15(a)gmail.com> --- drivers/net/gtp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 30e0a10595a1..24e5c54d06c1 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -539,7 +539,7 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, mtu = dst_mtu(&rt->dst); } - rt->dst.ops->update_pmtu(&rt->dst, NULL, skb, mtu, false); + skb_dst_update_pmtu_no_confirm(skb, mtu); if (!skb_is_gso(skb) && (iph->frag_off & htons(IP_DF)) && mtu < ntohs(iph->tot_len)) { -- 2.25.1

2 1

RFC PATCHES - Merge gtp driver with https://github.com/free5gc/gtp5g
by Richard Haines 05 Aug '21

05 Aug '21

Harald, As requested I've copied this to the osmocom-net-gprs(a)lists.osmocom.org list. I'm not sure if any of you are interested in this (or even if it's useful as the "gtp5g" module is still 'work in progress'), however if you are interested any comments ??. I've attached a tarball (gtp-patches.tar.xz) with all the patches and here are some notes: 1) This has been built and tested on Fedora 34 using linux-5.14-rc4 kernel. 2) The first two kernel patches reinstall the current gtp driver into manageable modules. You can build the kernel with just these patches and should still be able to use libgtpnl, osmo-ggsn etc. as no code changes. 3) The remaining three kernel patches install the services from https://github.com/free5gc/gtp5g to DRV_VERSION "1.0.3a", also update core services (where possible) to use common functions e.g. gtp_dev_xmit(). It is possible to implement more common functions, but that can be done if the patch series is of any use !!! 4) There are three possible Kconfig configuration options: GTP_PDP only - GTP_PDR only - GTP_PDP and GTP_PDR 5) If GTP_PDP and GTP_PDR build selected, the basic tests using osmo-ggsn or openggsn work using tests in [1] and the tests from libgtp5gnl [2] (./run.sh SimpleUPTest & ULCLTest1) will also run - provided the following tarball patch has been applied to update [2]: 0001-libgtp5gnl-Update-to-match-new-linux-gtp.h-entries.patch 6) Be aware that apps/libs should now define their role specifically (i.e. GTP_ROLE_SGSN, GTP_ROLE_RAN) and not default thinking that '0' will always be GTP_ROLE_GGSN or GTP_ROLE_UPF. Both libgtp*nl versions do this. 7) I've only limited test options so cannot guarantee all will work !! 8) The kernel and tests have been built using Fedora 34. When building the kernel with Fedora the gtp5g side will hang when testing. This happens whether I build/load as described in the https://github.com/free5gc/gtp5g README i.e. untouched by me, or with my patches. To solve this issue I've commented out some code in pdr.c pdr_context_free() with a comment: /* FIXME - Stop crash when kernel built on Fedora 34 */ I'll try and fix this at some stage. 9) The original gtp5g driver 'struct net_device_ops' used: .ndo_get_stats64 = ip_tunnel_get_stats64, However this is deprecated so set to: .ndo_get_stats64 = dev_get_tstats64, 10) Running checkpatch on the kernel patches will show plenty of warning and check messages, however all 'errors' are fixed. 11) I've patches that add the LSM/SELinux services, however have not included these as I guess still nfi. [1] https://osmocom.org/projects/linux-kernel-gtp-u/wiki/Basic_Testing [2] https://github.com/free5gc/libgtp5gnl Richard Haines (5): gtp: Delete current gtp driver gtp: Split gtp driver into modules gtp: Add support for PDR services - Part 1 gtp: Add support for PDR services - Part 2 gtp: Add support for PDR services - Part 3 drivers/net/Kconfig | 17 +- drivers/net/Makefile | 2 +- drivers/net/gtp.c | 1462 ---------------------------- drivers/net/gtp/Kconfig | 45 + drivers/net/gtp/Makefile | 8 + drivers/net/gtp/dev_core.c | 103 ++ drivers/net/gtp/encap.c | 406 ++++++++ drivers/net/gtp/far.c | 596 ++++++++++++ drivers/net/gtp/gtp_core.c | 243 +++++ drivers/net/gtp/include/dev_core.h | 59 ++ drivers/net/gtp/include/encap.h | 27 + drivers/net/gtp/include/far.h | 64 ++ drivers/net/gtp/include/gtp_core.h | 105 ++ drivers/net/gtp/include/nl_core.h | 18 + drivers/net/gtp/include/pdp.h | 53 + drivers/net/gtp/include/pdp_ipv4.h | 31 + drivers/net/gtp/include/pdr.h | 96 ++ drivers/net/gtp/include/pdr_ipv4.h | 49 + drivers/net/gtp/include/proc_dbg.h | 85 ++ drivers/net/gtp/include/qer.h | 49 + drivers/net/gtp/nl_core.c | 444 +++++++++ drivers/net/gtp/pdp.c | 578 +++++++++++ drivers/net/gtp/pdp_ipv4.c | 166 ++++ drivers/net/gtp/pdr.c | 1272 ++++++++++++++++++++++++ drivers/net/gtp/pdr_ipv4.c | 330 +++++++ drivers/net/gtp/proc_dbg.c | 384 ++++++++ drivers/net/gtp/qer.c | 463 +++++++++ include/net/gtp.h | 105 +- include/uapi/linux/gtp.h | 200 +++- include/uapi/linux/if_link.h | 5 + 30 files changed, 5981 insertions(+), 1484 deletions(-) delete mode 100644 drivers/net/gtp.c create mode 100644 drivers/net/gtp/Kconfig create mode 100644 drivers/net/gtp/Makefile create mode 100644 drivers/net/gtp/dev_core.c create mode 100644 drivers/net/gtp/encap.c create mode 100644 drivers/net/gtp/far.c create mode 100644 drivers/net/gtp/gtp_core.c create mode 100644 drivers/net/gtp/include/dev_core.h create mode 100644 drivers/net/gtp/include/encap.h create mode 100644 drivers/net/gtp/include/far.h create mode 100644 drivers/net/gtp/include/gtp_core.h create mode 100644 drivers/net/gtp/include/nl_core.h create mode 100644 drivers/net/gtp/include/pdp.h create mode 100644 drivers/net/gtp/include/pdp_ipv4.h create mode 100644 drivers/net/gtp/include/pdr.h create mode 100644 drivers/net/gtp/include/pdr_ipv4.h create mode 100644 drivers/net/gtp/include/proc_dbg.h create mode 100644 drivers/net/gtp/include/qer.h create mode 100644 drivers/net/gtp/nl_core.c create mode 100644 drivers/net/gtp/pdp.c create mode 100644 drivers/net/gtp/pdp_ipv4.c create mode 100644 drivers/net/gtp/pdr.c create mode 100644 drivers/net/gtp/pdr_ipv4.c create mode 100644 drivers/net/gtp/proc_dbg.c create mode 100644 drivers/net/gtp/qer.c -- 2.31.1

1 0

[PATCH net-next] gtp: Fix a typo
by Zheng Yongjun 01 Jun '21

01 Jun '21

Suppport ==> Support Signed-off-by: Zheng Yongjun <zhengyongjun3(a)huawei.com> --- drivers/net/gtp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 39c00f050fbd..1c9023d47e00 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -436,7 +436,7 @@ static inline void gtp1_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) gtp1->length = htons(payload_len); gtp1->tid = htonl(pctx->u.v1.o_tei); - /* TODO: Suppport for extension header, sequence number and N-PDU. + /* TODO: Support for extension header, sequence number and N-PDU. * Update the length field if any of them is available. */ } -- 2.25.1

2 1

Linux kernel GTP-U
by Encinar Sanz, Paula (ext) 25 May '21

25 May '21

Good afternoon, I am investigating linux kernel GTP-U to build a 5G Core with this kernel but I understand that it does not support 5G and only works with 4G. Could you solve this doubt? Thanks in advance, Regards This e-mail and the documents attached are confidential and intended solely for the addressee; it may also be privileged. If you receive this e-mail in error, please notify the sender immediately and destroy it. As its integrity cannot be secured on the Internet, the Atos group liability cannot be triggered for the message content. Although the sender endeavors to maintain a computer virus-free network, the sender does not warrant that this transmission is virus-free and will not be liable for any damages resulting from any virus transmitted. Este mensaje y los ficheros adjuntos pueden contener información confidencial destinada solamente a la(s) persona(s) mencionadas anteriormente y pueden estar protegidos por secreto profesional. Si usted recibe este correo electrónico por error, gracias por informar inmediatamente al remitente y destruir el mensaje. Al no estar asegurada la integridad de este mensaje sobre la red, Atos no se hace responsable por su contenido. Su contenido no constituye ningún compromiso para el grupo Atos, salvo ratificación escrita por ambas partes. Aunque se esfuerza al máximo por mantener su red libre de virus, el emisor no puede garantizar nada al respecto y no será responsable de cualesquiera daños que puedan resultar de una transmisión de virus.

2 1

osmo-pcu always "T3193 timeout expired, freeing TBF" when osmo-sgsn send GPRS ATTACH ACCEPT
by 海鱼 11 May '21

11 May '21

Hi All, Here is the issue when use egprs with osmo-bts-trx,osmo-pcu and osmo-sgsn. osmo-pcu debugging info as below: <000c> gprs_bssgp_pcu.cpp:168 LLC [SGSN -> PCU] = TLLI: 0xaf8ce1ff IMSI: 460000323182742 len: 26 <0008> tbf.cpp:1068 Allocating DL TBF: MS_CLASS=12/12 <0008> tbf.cpp:541 TBF(TFI=0 TLLI=0x00000000 DIR=DL STATE=NULL) Setting Control TS 6 <0008> tbf.cpp:945 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=NULL) Allocated: trx = 0, ul_slots = 40, dl_slots = 40 <0008> bts.cpp:898 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) TX: START Immediate Assignment Downlink (PCH) <0008> tbf_dl.cpp:520 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) set ass. type PACCH [prev CCCH:0, PACCH:0] <0008> tbf.cpp:1270 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) start Packet Downlink Assignment (PACCH) <0009> tbf.cpp:788 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) Scheduled DL Assignment polling on PACCH (FN=2402222, TS=6) <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402270, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402322, TS=6 <0008> tbf.cpp:808 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) poll timeout for FN=2402322, TS=6 (curr FN 2402387) <0008> tbf.cpp:879 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Timeout for polling PACKET DOWNLINK ACK. <0008> tbf.cpp:1184 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Assignment was on PACCH <0008> tbf.cpp:1190 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Downlink ACK was received <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402426, TS=6 <0008> tbf.cpp:989 Allocating UL TBF: MS_CLASS=12/12 <0008> tbf.cpp:541 TBF(TFI=0 TLLI=0x00000000 DIR=UL STATE=NULL) Setting Control TS 6 <0008> tbf.cpp:945 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=NULL) Allocated: trx = 0, ul_slots = 40, dl_slots = 00 <0008> tbf.cpp:1359 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=ASSIGN) start Packet Uplink Assignment (PACCH) <0009> tbf.cpp:782 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled UL Assignment polling on PACCH (FN=2402478, TS=6) <0007> gprs_rlcmac_meas.cpp:184 DL Bandwitdh of IMSI=460000323182742 / TLLI=0xaf8ce1ff: 66 KBits/s <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402525, TS=6 <000c> tbf_ul.cpp:404 LLC [PCU -> SGSN] TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=FLOW) len=97 <000c> gprs_bssgp_pcu.cpp:168 LLC [SGSN -> PCU] = TLLI: 0xaf8ce1ff IMSI: 460000323182742 len: 9 <0008> tbf.cpp:808 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) poll timeout for FN=2402525, TS=6 (curr FN 2402586) <0008> tbf.cpp:879 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Timeout for polling PACKET DOWNLINK ACK. <0008> tbf.cpp:1184 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Assignment was on PACCH <0008> tbf.cpp:1190 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Downlink ACK was received <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402625, TS=6 <0007> gprs_rlcmac_meas.cpp:106 UL RSSI of TLLI=0xaf8ce1ff: -31 dBm <0008> tbf.cpp:486 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=FINISHED) free <0008> tbf.cpp:989 Allocating UL TBF: MS_CLASS=12/12 <0008> tbf.cpp:541 TBF(TFI=0 TLLI=0x00000000 DIR=UL STATE=NULL) Setting Control TS 6 <0008> tbf.cpp:945 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=NULL) Allocated: trx = 0, ul_slots = 40, dl_slots = 00 <0008> tbf.cpp:1359 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=ASSIGN) start Packet Uplink Assignment (PACCH) <0009> tbf.cpp:782 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled UL Assignment polling on PACCH (FN=2402673, TS=6) <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402720, TS=6 <000c> tbf_ul.cpp:404 LLC [PCU -> SGSN] TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=FLOW) len=17 <000c> gprs_bssgp_pcu.cpp:168 LLC [SGSN -> PCU] = TLLI: 0xaf8ce1ff IMSI: 460000323182742 len: 26 <0007> gprs_rlcmac_meas.cpp:184 DL Bandwitdh of IMSI=460000323182742 / TLLI=0xaf8ce1ff: 86 KBits/s <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402768, TS=6 <0007> gprs_rlcmac_meas.cpp:106 UL RSSI of TLLI=0xaf8ce1ff: -31 dBm <0008> tbf.cpp:486 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=UL STATE=FINISHED) free <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402816, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402863, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402911, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2402959, TS=6 <0007> gprs_rlcmac_meas.cpp:184 DL Bandwitdh of IMSI=460000323182742 / TLLI=0xaf8ce1ff: 240 KBits/s <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2403006, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2403054, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2403102, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2403149, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FINISHED) Scheduled Ack/Nack polling on FN=2403197, TS=6 <0008> tbf.cpp:650 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=WAIT RELEASE) T3193 timeout expired, freeing TBF <0008> tbf.cpp:486 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=RELEASING) free <000c> gprs_bssgp_pcu.cpp:168 LLC [SGSN -> PCU] = TLLI: 0xaf8ce1ff IMSI: 460000323182742 len: 26 <0008> tbf.cpp:1068 Allocating DL TBF: MS_CLASS=12/12 <0008> tbf.cpp:541 TBF(TFI=0 TLLI=0x00000000 DIR=DL STATE=NULL) Setting Control TS 6 <0008> tbf.cpp:945 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=NULL) Allocated: trx = 0, ul_slots = 40, dl_slots = 40 <0008> bts.cpp:898 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) TX: START Immediate Assignment Downlink (PCH) <0008> tbf_dl.cpp:520 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) set ass. type PACCH [prev CCCH:0, PACCH:0] <0008> tbf.cpp:1270 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) start Packet Downlink Assignment (PACCH) <0009> tbf.cpp:788 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) Scheduled DL Assignment polling on PACCH (FN=2404263, TS=6) <0007> gprs_rlcmac_meas.cpp:184 DL Bandwitdh of IMSI=460000323182742 / TLLI=0xaf8ce1ff: 22 KBits/s <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404311, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404358, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404406, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404454, TS=6 <0008> tbf.cpp:808 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) poll timeout for FN=2404454, TS=6 (curr FN 2404519) <0008> tbf.cpp:879 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Timeout for polling PACKET DOWNLINK ACK. <0008> tbf.cpp:1184 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Assignment was on PACCH <0008> tbf.cpp:1190 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Downlink ACK was received <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404558, TS=6 <0007> gprs_rlcmac_meas.cpp:184 DL Bandwitdh of IMSI=460000323182742 / TLLI=0xaf8ce1ff: 163 KBits/s <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404605, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404653, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2404701, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FINISHED) Scheduled Ack/Nack polling on FN=2404748, TS=6 <0008> tbf.cpp:650 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=WAIT RELEASE) T3193 timeout expired, freeing TBF <0008> tbf.cpp:486 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=RELEASING) free <000c> gprs_bssgp_pcu.cpp:168 LLC [SGSN -> PCU] = TLLI: 0xaf8ce1ff IMSI: 460000323182742 len: 26 <0008> tbf.cpp:1068 Allocating DL TBF: MS_CLASS=12/12 <0008> tbf.cpp:541 TBF(TFI=0 TLLI=0x00000000 DIR=DL STATE=NULL) Setting Control TS 6 <0008> tbf.cpp:945 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=NULL) Allocated: trx = 0, ul_slots = 40, dl_slots = 40 <0008> bts.cpp:898 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) TX: START Immediate Assignment Downlink (PCH) <0008> tbf_dl.cpp:520 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) set ass. type PACCH [prev CCCH:0, PACCH:0] <0008> tbf.cpp:1270 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) start Packet Downlink Assignment (PACCH) <0009> tbf.cpp:788 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=ASSIGN) Scheduled DL Assignment polling on PACCH (FN=2405537, TS=6) <0007> gprs_rlcmac_meas.cpp:184 DL Bandwitdh of IMSI=460000323182742 / TLLI=0xaf8ce1ff: 25 KBits/s <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2405585, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2405632, TS=6 <0009> tbf_dl.cpp:880 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Scheduled Ack/Nack polling on FN=2405680, TS=6 <0008> tbf.cpp:808 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) poll timeout for FN=2405680, TS=6 (curr FN 2405741) <0008> tbf.cpp:879 TBF(TFI=0 TLLI=0xaf8ce1ff DIR=DL STATE=FLOW) Timeout for polling PACKET DOWNLINK ACK. osmo-sgsn debugging info: <0011> gprs_llc_parse.c:78 LLC SAPI=1 C   U GEA0 IOV-UI=0x000000 FCS=0xa3ad04 CMD=UI DATA <0002> gprs_gmm.c:1118 MM(---/ffffffff) -> GMM ATTACH REQUEST MI(4018987519) type="GPRS attach" <0002> fsm.c:461 GMM(gmm_fsm)[0xaaaaf63e5b50]{Deregistered}: Allocated <0002> fsm.c:461 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{Init}: Allocated <0002> fsm.c:461 MM_STATE_Gb[0xaaaaf63e5e50]{Idle}: Allocated <0002> gprs_sgsn.c:299 MM(/00000000) Allocated with GEA0 cipher. <0002> gprs_gmm.c:1100 GMM(gmm_fsm)[0xaaaaf63e5b50]{Deregistered}: Received Event E_GMM_COMMON_PROC_INIT_REQ <0002> gprs_gmm_fsm.c:25 GMM(gmm_fsm)[0xaaaaf63e5b50]{Deregistered}: state_chg to CommonProcedureInitiated <0011> gprs_llc.c:1056 LLME(ffffffff/af8ce1ff){UNASSIGNED} LLGM Assign pre (af8ce1ff => cf0ebf06) <0011> gprs_llc.c:1102 LLME(af8ce1ff/cf0ebf06){ASSIGNED} LLGM Assign post (af8ce1ff => cf0ebf06) <0002> gprs_gmm.c:1271 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{Init}: Received Event E_ATTACH_REQ_RECV <0002> gprs_gmm_attach.c:50 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{Init}: state_chg to CheckIdentity <0002> gprs_gmm.c:409 MM(/cf0ebf06) <- GPRS IDENTITY REQUEST: mi_type=IMEI <0010> gprs_bssgp.c:423 BSSGP TLLI=0xaf8ce1ff Rx UPLINK-UNITDATA <0011> gprs_llc_parse.c:78 LLC SAPI=1 C   U GEA0 IOV-UI=0x000000 FCS=0x8c64cd CMD=UI DATA <0002> gprs_gb.c:40 MM_STATE_Gb(2945245695)[0xaaaaf63e5e50]{Idle}: Received Event E_MM_PDU_RECEPTION <0002> gprs_gmm.c:1040 MM(/cf0ebf06) -> GMM IDENTITY RESPONSE: MI(IMEI)=866089038960590 <0002> gprs_gmm.c:113 MM(/cf0ebf06) Stopping MM timer 3370 but 0 is running <0002> gprs_gmm.c:1084 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{CheckIdentity}: Received Event E_IDEN_RESP_RECV <0002> gprs_gmm_attach.c:111 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{CheckIdentity}: state_chg to CheckIdentity <0002> gprs_gmm.c:409 MM(/cf0ebf06) <- GPRS IDENTITY REQUEST: mi_type=IMSI <0010> gprs_bssgp.c:423 BSSGP TLLI=0xaf8ce1ff Rx UPLINK-UNITDATA <0011> gprs_llc_parse.c:78 LLC SAPI=1 C   U GEA0 IOV-UI=0x000000 FCS=0xb71895 CMD=UI DATA <0002> gprs_gb.c:40 MM_STATE_Gb(2945245695)[0xaaaaf63e5e50]{Idle}: Received Event E_MM_PDU_RECEPTION <0002> gprs_gmm.c:1040 MM(/cf0ebf06) -> GMM IDENTITY RESPONSE: MI(IMSI)=460000323182742 <0002> gprs_gmm.c:113 MM(/cf0ebf06) Stopping MM timer 3370 but 0 is running <0002> gprs_gmm.c:1084 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{CheckIdentity}: Received Event E_IDEN_RESP_RECV <0002> gprs_gmm_attach.c:113 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{CheckIdentity}: state_chg to Authenticate <0002> gprs_gmm_attach.c:148 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{Authenticate}: state_chg to WaitAttachComplete <0002> gprs_gmm.c:289 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (new P-TMSI=0xcf0ebf06) <0002> gprs_gmm.c:305 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (ra_id=0x020xf20x50 lac:2f11 rac:2) <000e> gprs_sgsn.c:1012 Checking for inactive LLMEs, time = 1652 <000f> gprs_ns.c:779 NSEI=1234 Timer expired in mode tns-test (30 seconds) <000f> gprs_ns.c:712 NSEI=1234 Tx NS ALIVE (NSVCI=1234) <000f> gprs_ns.c:751 NSEI=1234 Starting timer in mode tns-alive (3 seconds) <000f> gprs_ns.c:725 NSEI=1234 Tx NS ALIVE_ACK (NSVCI=1234) <000f> gprs_ns.c:1763 NSEI 1234 (NS-VCI=1234) setting state [ALIVE,UNBLOCKED,UNRESET] -> [ALIVE,UNBLOCKED,UNRESET] <000f> gprs_ns.c:751 NSEI=1234 Starting timer in mode tns-test (30 seconds) <0010> gprs_bssgp.c:839 BSSGP BVCI=1234 Rx Flow Control BVC <0002> fsm.c:322 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{WaitAttachComplete}: Timeout of T3350 <0002> gprs_gmm.c:289 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (new P-TMSI=0xcf0ebf06) <0002> gprs_gmm.c:305 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (ra_id=0x020xf20x50 lac:2f11 rac:2) <0002> fsm.c:322 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{WaitAttachComplete}: Timeout of T3350 <0002> gprs_gmm.c:289 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (new P-TMSI=0xcf0ebf06) <0002> gprs_gmm.c:305 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (ra_id=0x020xf20x50 lac:2f11 rac:2) <0010> gprs_bssgp.c:423 BSSGP TLLI=0xaf8ce1ff Rx UPLINK-UNITDATA <0011> gprs_llc_parse.c:78 LLC SAPI=1 C   U GEA0 IOV-UI=0x000000 FCS=0x0d0447 CMD=UI DATA <0002> gprs_gb.c:40 MM_STATE_Gb(2945245695)[0xaaaaf63e5e50]{Idle}: Received Event E_MM_PDU_RECEPTION <0002> gprs_gmm.c:1118 MM(460000323182742/cf0ebf06) -> GMM ATTACH REQUEST MI(4018987519) type="GPRS attach" <0002> gprs_gmm.c:1100 GMM(gmm_fsm)[0xaaaaf63e5b50]{CommonProcedureInitiated}: Received Event E_GMM_COMMON_PROC_INIT_REQ <0002> gprs_gmm.c:1100 GMM(gmm_fsm)[0xaaaaf63e5b50]{CommonProcedureInitiated}: Event E_GMM_COMMON_PROC_INIT_REQ not permitted <0011> gprs_llc.c:1056 LLME(af8ce1ff/cf0ebf06){ASSIGNED} LLGM Assign pre (af8ce1ff => cf0ebf06) <0011> gprs_llc.c:1102 LLME(af8ce1ff/cf0ebf06){ASSIGNED} LLGM Assign post (af8ce1ff => cf0ebf06) <0002> gprs_gmm.c:1271 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{WaitAttachComplete}: Received Event E_ATTACH_REQ_RECV <0002> gprs_gmm_attach.c:360 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{WaitAttachComplete}: state_chg to Init <0002> gprs_gmm_attach.c:50 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{Init}: state_chg to CheckIdentity <0002> gprs_gmm.c:409 MM(460000323182742/cf0ebf06) <- GPRS IDENTITY REQUEST: mi_type=IMEI <0010> gprs_bssgp.c:839 BSSGP BVCI=1234 Rx Flow Control BVC <0010> gprs_bssgp.c:423 BSSGP TLLI=0xaf8ce1ff Rx UPLINK-UNITDATA <0011> gprs_llc_parse.c:78 LLC SAPI=1 C   U GEA0 IOV-UI=0x000000 FCS=0xea5c4f CMD=UI DATA <0002> gprs_gb.c:40 MM_STATE_Gb(2945245695)[0xaaaaf63e5e50]{Idle}: Received Event E_MM_PDU_RECEPTION <0002> gprs_gmm.c:1040 MM(460000323182742/cf0ebf06) -> GMM IDENTITY RESPONSE: MI(IMEI)=866089038960590 <0002> gprs_gmm.c:113 MM(460000323182742/cf0ebf06) Stopping MM timer 3370 but 0 is running <0002> gprs_gmm.c:1084 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{CheckIdentity}: Received Event E_IDEN_RESP_RECV <0002> gprs_gmm_attach.c:113 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{CheckIdentity}: state_chg to Authenticate <0002> gprs_gmm_attach.c:148 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{Authenticate}: state_chg to WaitAttachComplete <0002> gprs_gmm.c:289 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (new P-TMSI=0xcf0ebf06) <0002> gprs_gmm.c:305 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (ra_id=0x020xf20x50 lac:2f11 rac:2) <0002> fsm.c:322 GMM_ATTACH_REQ_FSM(gb_gmm_req)[0xaaaaf63e5c80]{WaitAttachComplete}: Timeout of T3350 <0002> gprs_gmm.c:289 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (new P-TMSI=0xcf0ebf06) <0002> gprs_gmm.c:305 MM(460000323182742/cf0ebf06) <- GPRS ATTACH ACCEPT (ra_id=0x020xf20x50 lac:2f11 rac:2) Configure information： gprs mode egprs   gprs routing area 2   gprs cell bvci 1234   gprs nsei 1234   gprs nsvc 0 nsvci 1234   gprs nsvc 0 local udp port 23001   gprs nsvc 0 remote udp port 23000   。。。    timeslot 6     phys_chan_config PDCH     hopping enabled 0    timeslot 7     phys_chan_config PDCH     hopping enabled 0 pcu  flow-control-interval 10  cs 2  cs max 4  cs threshold 10 33  cs downgrade-threshold 200  cs link-quality-ranges cs1 6 cs2 5 8 cs3 7 13 cs4 12  mcs link-quality-ranges mcs1 6 mcs2 5 8 mcs3 7 13 mcs4 12 15 mcs5 14 17 mcs6 16 18 mcs7 17 20 mcs8 19 24 mcs9 23  mcs max 9  window-size 64 0  queue idle-ack-delay 10  queue codel  alloc-algorithm dynamic  alpha 0  gamma 0  dl-tbf-idle-time 2000 sgsn  gtp local-ip 127.0.0.1  ggsn 0 remote-ip 127.0.0.6  ggsn 0 gtp-version 1  auth-policy accept-all ns  timer tns-block 3  timer tns-block-retries 3  timer tns-reset 3  timer tns-reset-retries 3  timer tns-test 30  timer tns-alive 3  timer tns-alive-retries 10  encapsulation udp local-ip 127.0.0.1  encapsulation udp local-port 23000  encapsulation framerelay-gre enabled 0 Version information: OsmoPCU version 0.8.0 OsmoSGSN version 1.6.2 I tried to debug this issue but the no any progress for this issue.The root cause is confifure or something else? I need help. Thanks a lot. Br, Perter

3 5

Osmo-ggsn MTU issue with Kernel GTP
by Keith A. Milner 13 Apr '21

13 Apr '21

I've had an interesting situation with osmo-ggsn and kernel GTP module that I wonder if someone could shed some light on. I've set up osmo-ggsn (v1.7.1) on Debian 10. There are two ethernet interfaces on the system, ens224 (for the Gn Interface) and ens256 (for the Gi Interface). Using the following ggsn configuration seems to work fine: ggsn ggsn0 description Keith's Test APN gtp state-dir /tmp gtp bind-ip <ens224 IP> gtp control-ip <ens224 IP> apn keithtest gtpu-mode tun tun-device apn0 type-support v4 ip prefix dynamic 192.168.42.0/24 ip dns 0 <redacted> ip dns 1 <redacted> ip ifconfig 192.168.42.0/24 no shutdown default-apn keithtest no shutdown ggsn I get the following tun device: 5: apn0: <POINTOPOINT,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UNKNOWN group default qlen 500 link/none inet 192.168.42.0/24 scope global apn0 valid_lft forever preferred_lft forever inet6 fe80::fada:2e69:ad71:a61b/64 scope link stable-privacy valid_lft forever preferred_lft forever Traffic routes, all is good with the world. However, if I change to using Kernel GTP, by changing the gtpu-mode to "kernel-gtp", traffic will not route (and, yes, I have done modprobe gtp). After some exploration, it appears the tun device is being created with an MTU of zero which then, cannot pass traffic: 10: apn0: <POINTOPOINT,UP,LOWER_UP> mtu 0 qdisc noqueue state UNKNOWN group default qlen 1000 link/none inet 192.168.42.0/24 scope global apn0 valid_lft forever preferred_lft forever Here are the logs: Tue Apr 13 11:02:06 2021 DGGSN NOTICE <0002> ggsn.c:830 GGSN(ggsn0): Successfully started Tue Apr 13 11:02:06 2021 DGGSN INFO <0002> ggsn.c:186 APN(keithtest): Starting Tue Apr 13 11:02:06 2021 DGGSN INFO <0002> ggsn.c:204 APN(keithtest): Opening Kernel GTP device apn0 Tue Apr 13 11:02:06 2021 DGGSN NOTICE <0002> gtp-kernel.c:79 Initialized GTP kernel mode (genl ID is 24) Tue Apr 13 11:02:06 2021 DTUN NOTICE <0001> tun.c:217 GTP kernel configured Tue Apr 13 11:02:06 2021 DGGSN INFO <0002> ggsn.c:236 APN(keithtest): Setting tun IP address 192.168.42.0/24 Tue Apr 13 11:02:06 2021 DGGSN INFO <0002> ggsn.c:294 APN(keithtest): Creating IPv4 pool 192.168.42.0/24 Tue Apr 13 11:02:06 2021 DGGSN INFO <0002> ggsn.c:168 APN(keithtest): Blacklist tun IP 192.168.42.0/24 Tue Apr 13 11:02:06 2021 DGGSN NOTICE <0002> ggsn.c:325 APN(keithtest): Successfully started Note that if I manually alter the MTU, everything starts working again: ip link set dev apn0 mtu 1500 So I can put this into a script and run it via ipup-script, and this works fine. I just get the impression I shouldn't have to do this, and the system should be setting the MTU, just as it does when I use tun mode. Any thoughts on what I'm doing wrong here? Cheers, Keith

2 2

Question about establishing stable internet link between UE and HNB connected to our developed network core.
by morteza ali Ahmadi 26 Mar '21

26 Mar '21

Hi, osmocom team members. In order to have an internet connection in UE, we have run OsmoGGSN project to set up GGSN. Also, SGSN is developed in our system. At this step, UE connects to our network core and we can query to an arbitrary site in the UE browser and we see the answer of the query without any problem and everything is OK. But after about 30 or 40 seconds, UE sends "Deactivate PDP Context Request" with SM Cause "SM Cause: Protocol error, unspecified (111)" and the connection is reset and a new "Activate PDP Context Request" is required. What is the reason and how can we solve it? -- *When there is much light, The shadow is deep...*

1 0

Automatic testing for kernel GTP tunnel driver
by Harald Welte 16 Mar '21

16 Mar '21

Hi all, in recent months - after many years of no patches at all - the kernel GTP driver has received significant interest in terms of contributions. Given the presumed few users of the GTP tunnel driver, as well as the tendency to use super ancient versions of software in the telecom world, I think that the usual "let's have the users test -rcX kernels" approach is unlikely going to work. Within Osmocom (a group of FOSS projects implementing various cellular standards, protocol stacks and network elements) we are used to rather comprehensive functional test suites that we execute automatically at least once per 24 hours. That is for the 99% of our software that is userspace code. For the kernel GTP code, it's of course not that simple, and we never had any related testing so far. In 2021, 5 years after the GTP kernel driver was merged mainline, we now finally have set up some test jobs for kernel GTP. Those jobs execute the same GGSN test suite that we run against the userspace dataplane (handling GTP-U and GTP-C in a userspace program "osmo-ggsn"). The only difference is that they configure osmo-ggsn to use the kernel GTP driver for GTP-U. Those tests are executed by jenkins, in the following jobs: * ttcn3-ggsn-test-kernel-latest-torvalds https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-kernel-l… runs GGSN test suite against latest released osmo-ggsn version with kernel-gtp of torvalds/linux.git test results at https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-kernel-l… (make sure to click "Expand All") * ttcn3-ggsn-test-kernel-latest-torvalds https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-kernel-l… runs GGSN test suite against latest released osmo-ggsn version with kernel-gtp of net-next.git test results at https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-kernel-l… (make sure to click "Expand All") * ttcn3-ggsn-test-kernel-latest https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-kernel-l… runs GGSN test suite against latest released osmo-ggsn version with kernel-gtp of Debian 10 kernel package test results at https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-kernel-l… (make sure to click "Expand All") * ttcn3-ggsn-test-latest https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-latest/ runs GGSN test suite against latest released osmo-ggsn version with userspace GTP-U, not using any kernel GTP driver test results at https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-latest/t… this serves as a base line that shows the test suite can fully pass, and any failures in the aove are either bugs or lack of features (like IPv6) Contrary to the above jenkins jobs which all run automatically once every 24 hours, there is also one jenkins job for manual execution: * ttcn3-ggsn-test-kernel-git https://jenkins.osmocom.org/jenkins/view/TTCN3/job/ttcn3-ggsn-test-kernel-g… this is the same test suite and osmo-ggsn as above, but a developer can manually trigger the job and specify * URL of the kernel git repo to build * branch of the kernel git repo to build * whether to use osmo-ggsn latest tag or master If any one is working on the kernel GTP driver and wants to get access to triggering that jenkins job in order to test your patches/branches before submission, please register an account at https://jenkins.osmocom.org/ and contact me privately. We are always happy about any contributions in terms of extending test coverage. This could be done by e.g. adding jobs using other P-GW/GGSN software than osmo-ggsn, or by extending the coverage of the test cases of the test suite. For anyone interested in more details, please see our redmine issue https://osmocom.org/issues/3208 tracks the development of the tests. The test suite source code (in TTCN-3) is at https://git.osmocom.org/osmo-ttcn3-hacks/tree/ggsn_tests with containerization + configs at https://git.osmocom.org/docker-playground/tree/ttcn3-ggsn-test I'd like to thank Oliver Smith <osmith(a)sysmocom.de> for creating the above automatic test CI integration. Development of this was funded by sysmocom (https://sysmocom.de/) If you have any questions, please feel free to reach out to Oliver and/or me. Regards, Harald -- - Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

2 1

[PATCH] Net: gtp: Fixed missing blank line after declaration
by Chinmayi Shetty 15 Mar '21

15 Mar '21

Signed-off-by: Chinmayi Shetty <chinmayishetty359(a)gmail.com> --- drivers/net/gtp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index efe5247d8c42..79998f4394e5 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -437,7 +437,7 @@ static inline void gtp1_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) gtp1->length = htons(payload_len); gtp1->tid = htonl(pctx->u.v1.o_tei); - /* TODO: Suppport for extension header, sequence number and N-PDU. + /* TODO: Support for extension header, sequence number and N-PDU. * Update the length field if any of them is available. */ } -- 2.25.1

2 1

[PATCH] Net: gtp: Fixed missing blank line after declaration
by Chinmayi Shetty 13 Mar '21

13 Mar '21

Signed-off-by: Chinmayi Shetty <chinmayishetty359(a)gmail.com> --- drivers/net/gtp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index efe5247d8c42..79998f4394e5 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -437,7 +437,7 @@ static inline void gtp1_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) gtp1->length = htons(payload_len); gtp1->tid = htonl(pctx->u.v1.o_tei); - /* TODO: Suppport for extension header, sequence number and N-PDU. + /* TODO: Support for extension header, sequence number and N-PDU. * Update the length field if any of them is available. */ } -- 2.25.1

1 0

Osmocom Mailing List re-organization
by Harald Welte 03 Mar '21

03 Mar '21

Dear Osmocom community, This topic has been past due for way too many years by now: A re-organization of our major mailing lists. I would like to propose the following changes. Pleas let me know if you have any comments or feedback. I'm aware that renaming will mean people have to update their mail filter rules, but I think we're long past the point where the names of some of our lists started to confuse users. == openbsc(a)lists.osmocom.org == * openbsc doesn't exist anymore since OsmoNITB, which is also obsolete * does already cover anything "Osmocom CNI" related * Proposed new name: osmocom-cni(a)lists.osmocom.org == osmocom-net-gprs(a)lists.osmocom.org == This date back to when GPRS was a highly experimental add-on to our GSM code base. This list should simply be merged with openbsc@ as osmocom-cni(a)lists.osmocom.org == simtrace(a)lists.osmocom.org == Historically was created to cover only the simtrace project. We should rename this to osmocom-simcard(a)lists.osmocom.org or something along those lines. I would like to suggest it covers * SIMtrace / SIMtrace2 hardware + firmware * pySim and related tools for working with SIM/USIM/UICC cards * any other information / discussion related to SIM/USIM/UICC cards, like OTA, ARA-M, ... == osmodevcon(a)lists.osmocom.org == This has been a private list for people attending OsmoDevCon I would like to open up list membership to the general public, and ensure it also covers the new OsmoDevCall. We could then have discussions regarding feedback, topics, scheduling, etc. on that list. Maybe rename it to osmocom-events(a)lists.osmocom.org instead? To differentiate: osmocom-event-orga(a)lists.osmocom.org should remain a private list related to organizational / administrative topics of those involved with organizing future events. == nextepc(a)lists.osmocom.org == Should have been renamed to open5gs(a)lists.osmocom.org quite some time ago, I simply forgot about it. My apologies. -- - 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 1

Osmocom CNI release 202102
by Pau Espin Pedrol 26 Feb '21

26 Feb '21

Dear all, I am pleased to announce new tagged releases for Osmocom Cellular Network Infrastructure components. Find more information about the release here [1]. Osmocom "Latest" repositories in OBS [2] should already contain packages for the new versions. OpenEmbedded related meta-layers such as meta-telephony usual stable/testing branch "201705" [3] have also been updated to build recipes for new versions. Regards, Pau [1] https://osmocom.org/news/132 [2] https://osmocom.org/projects/cellular-infrastructure/wiki/Latest_Builds [3] https://git.osmocom.org/meta-telephony/log/?h=201705 -- - Pau Espin Pedrol <pespin(a)sysmocom.de> http://www.sysmocom.de/ ======================================================================= * sysmocom - systems for mobile communications GmbH * Alt-Moabit 93 * 10559 Berlin, Germany * Sitz / Registered office: Berlin, HRB 134158 B * Geschaeftsfuehrer / Managing Director: Harald Welte

1 0

Very useful wireshark fix for NS / BSSGP decoding
by Harald Welte 13 Feb '21

13 Feb '21

In case anyone has been annoyed by this before: wireshark so far always showed _only_ the BSSGP decode in the Information column of a Gb interface pcap file, even if that BSSGP was contained in the "erroneous message" IE of a NS-STATUS message. This made it particularly difficult to spot NS-STATUS, and without custom coloring rules or search filters, a protocol trace would just look fine - except that in reality it isn't. I've now found out what's the proper method to resolve this and submitted a patch to wireshark, which was merged very quickly: https://gitlab.com/wireshark/wireshark/-/merge_requests/2081 After this change, the Information column in the packet list will show NS-STATUS with the human-redaable cause value for the reject, rather than displaying BSSGP as if nothing odd happened. 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)

1 0

Question about OsmoGGSN project
by morteza ali Ahmadi 08 Feb '21

08 Feb '21

Hi, osmocom team members We have run OsmoGGSN project using the attached config file in this email and the following command: osmo-ggsn As you can see in the config file, we have set IP 127.0.0.2 for GGSN. After executing the above command, a tunnel named tun4 is created. To communicate with SGSN, we use the SGSN emulator in OsmoGGSN as follow: sgsnemu -l 127.0.0.1 -r 127.0.0.2 After executing the above command, we see that 2 packets are sent from SGSN to GGSN. These packets are "Echo request" and "Create PDP context request" and GGSN responses to SGSN with the packets "Echo response" and "Create PDP context response". To test tun4, we send a GTP <DNS> packet containing an arbitrary query to GGSN and we see that this GTP <DNS> packet is received in Loopback (lo) interface and also, the DNS packet is received in tun4 interface. But there is no DNS response to our query neither in tun4 interface nor lo interface. A screenshot of the lo interface and tun4 interface in the wireshark is attached in this email (right is lo and left is tun4). How can we receive DNS responses to our DNS query? (Does tun4 require routing or something else?) -- *When there is much light, The shadow is deep...*

2 2

Question about libgtpnl
by morteza ali Ahmadi 03 Feb '21

03 Feb '21

Hi Osmocom team We have a basic question about libgtpnl (Linux Kernel GTP-U). We receive a lot of GTP <DNS> packets which are sent from MS to our HNB and from HNB to our SGSN (a screenshot is attached in this email). As you can see in the one of these packets in the attached image, our MS IP is 10.250.0.114 and our SGSN IP is 172.60.3.154. Also, our TEID is 100 which is set in RAB-Assignment Request sending from SGSN to HNB. In order to put these GTP packets in our GTP tunnel, we use libgtpnl and the two following commands: gtp-link add gtp1 gtp-tunnel add gtp1 v1 100 1 10.250.0.114 172.60.3.154 But when we monitor the gtp1 interface in the wireshark, there is no packet on this interface. Could you please help us to know the reason for this problem? (It is worth mentioning that after receiving these GTP <DNS> packets from MS to SGSN and using netstat, we see that Recv-Q (bytes in UDP port 2152 which are ready to read) is increased to a value more than 0) Best regards -- *When there is much light, The shadow is deep...*

2 2

Question about establishing an internet connection between MS and network core
by morteza ali Ahmadi 21 Jan '21

21 Jan '21

Hi Osmocom team. We are sorry to disturb you. We have a question about establishing a stable internet connection (IuPS_PS) between MS and our developed network core system using our home nodeB (HNB). We set *Primary DNS Address: 172.60.3.154 *(core IP) and *Secondary DNS Address: 8.8.8.8* in *Activate PDP Context Accept.* At this step, we can see the MS packets in the interface between HNB and core using wireshark. The protocol is *GTP <DNS>* and the stacked layers are *DNS / UDP / IP / GTP / UDP / IP* which I attached a screenshot of the wireshark log. As you can see in the attached image, our query in the MS is *server.hnb* *(http://server.hnb <http://server.hnb>)*. In order to show our default page to the user query, we set up a DNS server and HTTP server (Apache) in 172.60.3.154 and *BIND9* is configured as the primary server for the domain *server.hnb*. The command *dig @localhost server.hnb* outputs the following: ; <<>> DiG 9.16.1-Ubuntu <<>> @localhost server.hnb ; (1 server found) ;; global options: +cmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 49532 ;; flags: qr aa rd ra; QUERY: 1, ANSWER: 1, AUTHORITY: 0, ADDITIONAL: 1 ;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags:; udp: 4096 ; COOKIE: 15821f64047895870100000060070fb921a17af91fd22f9f (good) ;; QUESTION SECTION: ;server.hnb. IN A ;; ANSWER SECTION: server.hnb. 604800 IN A 172.60.3.154 ;; Query time: 0 msec ;; SERVER: 127.0.0.1#53(127.0.0.1) ;; WHEN: Jan 19 20:28:33 +0330 2021 ;; MSG SIZE rcvd: 83 and also when we query *http://server.hnb <http://server.hnb> *in the browser of the core system (172.60.3.154), we see the * Apache2 Ubuntu Default Page* but this default page is not shown in the MS. Furthermore, we think that a GTP tunnel may be required and we set up a GTP tunnel using *libgtpnl* and the two following commands: gtp-link add gtp1 gtp-tunnel add gtp1 v1 1 100 10.250.0.114 172.60.3.154 where 1 is the gTP-TEI obtained in *RAB-Assignment Response *and 100 is the similar field in *RAB-Assignment Request. *Also, 10.250.0.114 is the MS IP and 172.60.3.154 is the core IP. But there is no captured package in this tunnel (using wireshark). Could you please help us to solve this problem which is mentioned in the following questions? Is the above procedure performed in the correct way? We want to show our default page in MS when the user query is* http://server.hnb <http://server.hnb>*, but we think that a GTP manager is required to split the (( DNS / UDP / IP )) layers from (( DNS / UDP / IP / GTP / UDP / IP )) to send to DNS server. How can it be solved, if this assumption is correct? Thank you again for spending time with us. Best regards, Morteza Ali Ahmadi -- *When there is much light, The shadow is deep...*

1 0

Re: [PATCH net-next v4] GTP: add support for flow based tunneling API
by Jakub Kicinski 18 Jan '21

18 Jan '21

On Tue, 5 Jan 2021 20:45:05 -0800 Pravin B Shelar wrote: > Following patch add support for LWT flow based tunneling API > to send and recv GTP tunnel packet over tunnel metadata API. > This would allow this device integration with OVS or eBPF using > flow based tunneling APIs. > > Signed-off-by: Pravin B Shelar <pbshelar(a)fb.com> GTP folks please review. > v3-v4: > - add check for non-zero dst port > v2-v3: > - Fixed coding style > - changed IFLA_GTP_FD1 to optional param for LWT dev. > v1-v2: > - Fixed according to comments from Jonas Bonn > > drivers/net/gtp.c | 528 +++++++++++++++++++++-------- > include/uapi/linux/gtp.h | 12 + > include/uapi/linux/if_link.h | 1 + > include/uapi/linux/if_tunnel.h | 1 + > tools/include/uapi/linux/if_link.h | 1 + > 5 files changed, 399 insertions(+), 144 deletions(-) > > diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c > index 4c04e271f184..fc9d843df736 100644 > --- a/drivers/net/gtp.c > +++ b/drivers/net/gtp.c > @@ -21,6 +21,7 @@ > #include <linux/file.h> > #include <linux/gtp.h> > > +#include <net/dst_metadata.h> > #include <net/net_namespace.h> > #include <net/protocol.h> > #include <net/ip.h> > @@ -73,6 +74,9 @@ struct gtp_dev { > unsigned int hash_size; > struct hlist_head *tid_hash; > struct hlist_head *addr_hash; > + /* Used by LWT tunnel. */ > + bool collect_md; > + struct socket *collect_md_sock; > }; > > static unsigned int gtp_net_id __read_mostly; > @@ -179,33 +183,121 @@ static bool gtp_check_ms(struct sk_buff *skb, struct pdp_ctx *pctx, > return false; > } > > -static int gtp_rx(struct pdp_ctx *pctx, struct sk_buff *skb, > - unsigned int hdrlen, unsigned int role) > +static int gtp_set_tun_dst(struct gtp_dev *gtp, struct sk_buff *skb, > + unsigned int hdrlen, u8 gtp_version, > + __be64 tid, u8 flags) > { > - if (!gtp_check_ms(skb, pctx, hdrlen, role)) { > - netdev_dbg(pctx->dev, "No PDP ctx for this MS\n"); > - return 1; > + struct metadata_dst *tun_dst; > + int opts_len = 0; > + > + if (unlikely(flags & GTP1_F_MASK)) > + opts_len = sizeof(struct gtpu_metadata); > + > + tun_dst = udp_tun_rx_dst(skb, gtp->sk1u->sk_family, TUNNEL_KEY, tid, opts_len); > + if (!tun_dst) { > + netdev_dbg(gtp->dev, "Failed to allocate tun_dst"); > + goto err; > } > > + netdev_dbg(gtp->dev, "attaching metadata_dst to skb, gtp ver %d hdrlen %d\n", > + gtp_version, hdrlen); > + if (unlikely(opts_len)) { > + struct gtpu_metadata *opts; > + struct gtp1_header *gtp1; > + > + opts = ip_tunnel_info_opts(&tun_dst->u.tun_info); > + gtp1 = (struct gtp1_header *)(skb->data + sizeof(struct udphdr)); > + opts->ver = GTP_METADATA_V1; > + opts->flags = gtp1->flags; > + opts->type = gtp1->type; > + netdev_dbg(gtp->dev, "recved control pkt: flag %x type: %d\n", > + opts->flags, opts->type); > + tun_dst->u.tun_info.key.tun_flags |= TUNNEL_GTPU_OPT; > + tun_dst->u.tun_info.options_len = opts_len; > + skb->protocol = htons(0xffff); /* Unknown */ > + } > /* Get rid of the GTP + UDP headers. */ > if (iptunnel_pull_header(skb, hdrlen, skb->protocol, > - !net_eq(sock_net(pctx->sk), dev_net(pctx->dev)))) > - return -1; > + !net_eq(sock_net(gtp->sk1u), dev_net(gtp->dev)))) { > + gtp->dev->stats.rx_length_errors++; > + goto err; > + } > + > + skb_dst_set(skb, &tun_dst->dst); > + return 0; > +err: > + return -1; > +} > + > +static int gtp_rx(struct gtp_dev *gtp, struct sk_buff *skb, > + unsigned int hdrlen, u8 gtp_version, unsigned int role, > + __be64 tid, u8 flags, u8 type) > +{ > + if (ip_tunnel_collect_metadata() || gtp->collect_md) { > + int err; > + > + err = gtp_set_tun_dst(gtp, skb, hdrlen, gtp_version, tid, flags); > + if (err) > + goto err; > + } else { > + struct pdp_ctx *pctx; > + > + if (flags & GTP1_F_MASK) > + hdrlen += 4; > > - netdev_dbg(pctx->dev, "forwarding packet from GGSN to uplink\n"); > + if (type != GTP_TPDU) > + return 1; > + > + if (gtp_version == GTP_V0) > + pctx = gtp0_pdp_find(gtp, be64_to_cpu(tid)); > + else > + pctx = gtp1_pdp_find(gtp, be64_to_cpu(tid)); > + if (!pctx) { > + netdev_dbg(gtp->dev, "No PDP ctx to decap skb=%p\n", skb); > + return 1; > + } > + > + if (!gtp_check_ms(skb, pctx, hdrlen, role)) { > + netdev_dbg(pctx->dev, "No PDP ctx for this MS\n"); > + return 1; > + } > + /* Get rid of the GTP + UDP headers. */ > + if (iptunnel_pull_header(skb, hdrlen, skb->protocol, > + !net_eq(sock_net(pctx->sk), dev_net(gtp->dev)))) { > + gtp->dev->stats.rx_length_errors++; > + goto err; > + } > + } > + netdev_dbg(gtp->dev, "forwarding packet from GGSN to uplink\n"); > > /* Now that the UDP and the GTP header have been removed, set up the > * new network header. This is required by the upper layer to > * calculate the transport header. > */ > skb_reset_network_header(skb); > + if (pskb_may_pull(skb, sizeof(struct iphdr))) { > + struct iphdr *iph; > + > + iph = ip_hdr(skb); > + if (iph->version == 4) { > + netdev_dbg(gtp->dev, "inner pkt: ipv4"); > + skb->protocol = htons(ETH_P_IP); > + } else if (iph->version == 6) { > + netdev_dbg(gtp->dev, "inner pkt: ipv6"); > + skb->protocol = htons(ETH_P_IPV6); > + } else { > + netdev_dbg(gtp->dev, "inner pkt error: Unknown type"); > + } > + } > > - skb->dev = pctx->dev; > - > - dev_sw_netstats_rx_add(pctx->dev, skb->len); > - > + skb->dev = gtp->dev; > + dev_sw_netstats_rx_add(gtp->dev, skb->len); > netif_rx(skb); > return 0; > + > +err: > + gtp->dev->stats.rx_dropped++; > + return -1; > } > > /* 1 means pass up to the stack, -1 means drop and 0 means decapsulated. */ > @@ -214,7 +306,6 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) > unsigned int hdrlen = sizeof(struct udphdr) + > sizeof(struct gtp0_header); > struct gtp0_header *gtp0; > - struct pdp_ctx *pctx; > > if (!pskb_may_pull(skb, hdrlen)) > return -1; > @@ -224,16 +315,7 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) > if ((gtp0->flags >> 5) != GTP_V0) > return 1; > > - if (gtp0->type != GTP_TPDU) > - return 1; > - > - pctx = gtp0_pdp_find(gtp, be64_to_cpu(gtp0->tid)); > - if (!pctx) { > - netdev_dbg(gtp->dev, "No PDP ctx to decap skb=%p\n", skb); > - return 1; > - } > - > - return gtp_rx(pctx, skb, hdrlen, gtp->role); > + return gtp_rx(gtp, skb, hdrlen, GTP_V0, gtp->role, gtp0->tid, gtp0->flags, gtp0->type); > } > > static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) > @@ -241,41 +323,30 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb) > unsigned int hdrlen = sizeof(struct udphdr) + > sizeof(struct gtp1_header); > struct gtp1_header *gtp1; > - struct pdp_ctx *pctx; > > if (!pskb_may_pull(skb, hdrlen)) > return -1; > > gtp1 = (struct gtp1_header *)(skb->data + sizeof(struct udphdr)); > > + netdev_dbg(gtp->dev, "GTPv1 recv: flags %x\n", gtp1->flags); > if ((gtp1->flags >> 5) != GTP_V1) > return 1; > > - if (gtp1->type != GTP_TPDU) > - return 1; > - > /* From 29.060: "This field shall be present if and only if any one or > * more of the S, PN and E flags are set.". > * > * If any of the bit is set, then the remaining ones also have to be > * set. > */ > - if (gtp1->flags & GTP1_F_MASK) > - hdrlen += 4; > - > /* Make sure the header is larger enough, including extensions. */ > if (!pskb_may_pull(skb, hdrlen)) > return -1; > > gtp1 = (struct gtp1_header *)(skb->data + sizeof(struct udphdr)); > > - pctx = gtp1_pdp_find(gtp, ntohl(gtp1->tid)); > - if (!pctx) { > - netdev_dbg(gtp->dev, "No PDP ctx to decap skb=%p\n", skb); > - return 1; > - } > - > - return gtp_rx(pctx, skb, hdrlen, gtp->role); > + return gtp_rx(gtp, skb, hdrlen, GTP_V1, gtp->role, > + key32_to_tunnel_id(gtp1->tid), gtp1->flags, gtp1->type); > } > > static void __gtp_encap_destroy(struct sock *sk) > @@ -315,6 +386,11 @@ static void gtp_encap_disable(struct gtp_dev *gtp) > { > gtp_encap_disable_sock(gtp->sk0); > gtp_encap_disable_sock(gtp->sk1u); > + if (gtp->collect_md_sock) { > + udp_tunnel_sock_release(gtp->collect_md_sock); > + gtp->collect_md_sock = NULL; > + netdev_dbg(gtp->dev, "GTP socket released.\n"); > + } > } > > /* UDP encapsulation receive handler. See net/ipv4/udp.c. > @@ -329,7 +405,8 @@ static int gtp_encap_recv(struct sock *sk, struct sk_buff *skb) > if (!gtp) > return 1; > > - netdev_dbg(gtp->dev, "encap_recv sk=%p\n", sk); > + netdev_dbg(gtp->dev, "encap_recv sk=%p type %d\n", > + sk, udp_sk(sk)->encap_type); > > switch (udp_sk(sk)->encap_type) { > case UDP_ENCAP_GTP0: > @@ -383,12 +460,13 @@ static void gtp_dev_uninit(struct net_device *dev) > > static struct rtable *ip4_route_output_gtp(struct flowi4 *fl4, > const struct sock *sk, > - __be32 daddr) > + __be32 daddr, > + __be32 saddr) > { > memset(fl4, 0, sizeof(*fl4)); > fl4->flowi4_oif = sk->sk_bound_dev_if; > fl4->daddr = daddr; > - fl4->saddr = inet_sk(sk)->inet_saddr; > + fl4->saddr = saddr; > fl4->flowi4_tos = RT_CONN_FLAGS(sk); > fl4->flowi4_proto = sk->sk_protocol; > > @@ -412,7 +490,7 @@ static inline void gtp0_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) > gtp0->tid = cpu_to_be64(pctx->u.v0.tid); > } > > -static inline void gtp1_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) > +static inline void gtp1_push_header(struct sk_buff *skb, __be32 tid) > { > int payload_len = skb->len; > struct gtp1_header *gtp1; > @@ -428,46 +506,63 @@ static inline void gtp1_push_header(struct sk_buff *skb, struct pdp_ctx *pctx) > gtp1->flags = 0x30; /* v1, GTP-non-prime. */ > gtp1->type = GTP_TPDU; > gtp1->length = htons(payload_len); > - gtp1->tid = htonl(pctx->u.v1.o_tei); > + gtp1->tid = tid; > > /* TODO: Suppport for extension header, sequence number and N-PDU. > * Update the length field if any of them is available. > */ > } > > -struct gtp_pktinfo { > - struct sock *sk; > - struct iphdr *iph; > - struct flowi4 fl4; > - struct rtable *rt; > - struct pdp_ctx *pctx; > - struct net_device *dev; > - __be16 gtph_port; > -}; > - > -static void gtp_push_header(struct sk_buff *skb, struct gtp_pktinfo *pktinfo) > +static inline int gtp1_push_control_header(struct sk_buff *skb, > + __be32 tid, > + struct gtpu_metadata *opts, > + struct net_device *dev) > { > - switch (pktinfo->pctx->gtp_version) { > - case GTP_V0: > - pktinfo->gtph_port = htons(GTP0_PORT); > - gtp0_push_header(skb, pktinfo->pctx); > - break; > - case GTP_V1: > - pktinfo->gtph_port = htons(GTP1U_PORT); > - gtp1_push_header(skb, pktinfo->pctx); > - break; > + struct gtp1_header *gtp1c; > + int payload_len; > + > + if (opts->ver != GTP_METADATA_V1) > + return -ENOENT; > + > + if (opts->type == 0xFE) { > + /* for end marker ignore skb data. */ > + netdev_dbg(dev, "xmit pkt with null data"); > + pskb_trim(skb, 0); > } > + if (skb_cow_head(skb, sizeof(*gtp1c)) < 0) > + return -ENOMEM; > + > + payload_len = skb->len; > + > + gtp1c = skb_push(skb, sizeof(*gtp1c)); > + > + gtp1c->flags = opts->flags; > + gtp1c->type = opts->type; > + gtp1c->length = htons(payload_len); > + gtp1c->tid = tid; > + netdev_dbg(dev, "xmit control pkt: ver %d flags %x type %x pkt len %d tid %x", > + opts->ver, opts->flags, opts->type, skb->len, tid); > + return 0; > } > > +struct gtp_pktinfo { > + struct sock *sk; > + __u8 tos; > + struct flowi4 fl4; > + struct rtable *rt; > + struct net_device *dev; > + __be16 gtph_port; > +}; > + > static inline void gtp_set_pktinfo_ipv4(struct gtp_pktinfo *pktinfo, > - struct sock *sk, struct iphdr *iph, > - struct pdp_ctx *pctx, struct rtable *rt, > + struct sock *sk, > + __u8 tos, > + struct rtable *rt, > struct flowi4 *fl4, > struct net_device *dev) > { > pktinfo->sk = sk; > - pktinfo->iph = iph; > - pktinfo->pctx = pctx; > + pktinfo->tos = tos; > pktinfo->rt = rt; > pktinfo->fl4 = *fl4; > pktinfo->dev = dev; > @@ -477,40 +572,99 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, > struct gtp_pktinfo *pktinfo) > { > struct gtp_dev *gtp = netdev_priv(dev); > + struct gtpu_metadata *opts = NULL; > struct pdp_ctx *pctx; > struct rtable *rt; > struct flowi4 fl4; > - struct iphdr *iph; > - __be16 df; > + struct sock *sk = NULL; > + __be32 tun_id; > + __be32 daddr; > + __be32 saddr; > + u8 gtp_version; > int mtu; > + __u8 tos; > + __be16 df = 0; > > - /* Read the IP destination address and resolve the PDP context. > - * Prepend PDP header with TEI/TID from PDP ctx. > - */ > - iph = ip_hdr(skb); > - if (gtp->role == GTP_ROLE_SGSN) > - pctx = ipv4_pdp_find(gtp, iph->saddr); > - else > - pctx = ipv4_pdp_find(gtp, iph->daddr); > + if (gtp->collect_md) { > + /* LWT GTP1U encap */ > + struct ip_tunnel_info *info = NULL; > > - if (!pctx) { > - netdev_dbg(dev, "no PDP ctx found for %pI4, skip\n", > - &iph->daddr); > - return -ENOENT; > + info = skb_tunnel_info(skb); > + if (!info) { > + netdev_dbg(dev, "missing tunnel info"); > + return -ENOENT; > + } > + if (info->key.tp_dst && ntohs(info->key.tp_dst) != GTP1U_PORT) { > + netdev_dbg(dev, "unexpected GTP dst port: %d", ntohs(info->key.tp_dst)); > + return -EOPNOTSUPP; > + } > + pctx = NULL; > + gtp_version = GTP_V1; > + tun_id = tunnel_id_to_key32(info->key.tun_id); > + daddr = info->key.u.ipv4.dst; > + saddr = info->key.u.ipv4.src; > + sk = gtp->sk1u; > + if (!sk) { > + netdev_dbg(dev, "missing tunnel sock"); > + return -EOPNOTSUPP; > + } > + tos = info->key.tos; > + if (info->key.tun_flags & TUNNEL_DONT_FRAGMENT) > + df = htons(IP_DF); > + > + if (info->options_len != 0) { > + if (info->key.tun_flags & TUNNEL_GTPU_OPT) { > + opts = ip_tunnel_info_opts(info); > + } else { > + netdev_dbg(dev, "missing tunnel metadata for control pkt"); > + return -EOPNOTSUPP; > + } > + } > + netdev_dbg(dev, "flow-based GTP1U encap: tunnel id %d\n", > + be32_to_cpu(tun_id)); > + } else { > + struct iphdr *iph; > + > + if (ntohs(skb->protocol) != ETH_P_IP) > + return -EOPNOTSUPP; > + > + iph = ip_hdr(skb); > + > + /* Read the IP destination address and resolve the PDP context. > + * Prepend PDP header with TEI/TID from PDP ctx. > + */ > + if (gtp->role == GTP_ROLE_SGSN) > + pctx = ipv4_pdp_find(gtp, iph->saddr); > + else > + pctx = ipv4_pdp_find(gtp, iph->daddr); > + > + if (!pctx) { > + netdev_dbg(dev, "no PDP ctx found for %pI4, skip\n", > + &iph->daddr); > + return -ENOENT; > + } > + sk = pctx->sk; > + netdev_dbg(dev, "found PDP context %p\n", pctx); > + > + gtp_version = pctx->gtp_version; > + tun_id = htonl(pctx->u.v1.o_tei); > + daddr = pctx->peer_addr_ip4.s_addr; > + saddr = inet_sk(sk)->inet_saddr; > + tos = iph->tos; > + df = iph->frag_off; > + netdev_dbg(dev, "gtp -> IP src: %pI4 dst: %pI4\n", > + &iph->saddr, &iph->daddr); > } > - 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, sk, daddr, saddr); > if (IS_ERR(rt)) { > - netdev_dbg(dev, "no route to SSGN %pI4\n", > - &pctx->peer_addr_ip4.s_addr); > + netdev_dbg(dev, "no route to SSGN %pI4\n", &daddr); > dev->stats.tx_carrier_errors++; > goto err; > } > > if (rt->dst.dev == dev) { > - netdev_dbg(dev, "circular route to SSGN %pI4\n", > - &pctx->peer_addr_ip4.s_addr); > + netdev_dbg(dev, "circular route to SSGN %pI4\n", &daddr); > dev->stats.collisions++; > goto err_rt; > } > @@ -518,11 +672,10 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, > skb_dst_drop(skb); > > /* This is similar to tnl_update_pmtu(). */ > - df = iph->frag_off; > if (df) { > mtu = dst_mtu(&rt->dst) - dev->hard_header_len - > sizeof(struct iphdr) - sizeof(struct udphdr); > - switch (pctx->gtp_version) { > + switch (gtp_version) { > case GTP_V0: > mtu -= sizeof(struct gtp0_header); > break; > @@ -536,17 +689,38 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, > > rt->dst.ops->update_pmtu(&rt->dst, NULL, skb, mtu, false); > > - if (!skb_is_gso(skb) && (iph->frag_off & htons(IP_DF)) && > - mtu < ntohs(iph->tot_len)) { > - netdev_dbg(dev, "packet too big, fragmentation needed\n"); > + if (!skb_is_gso(skb) && (df & htons(IP_DF)) && mtu < skb->len) { > + netdev_dbg(dev, "packet too big, fragmentation needed"); > memset(IPCB(skb), 0, sizeof(*IPCB(skb))); > icmp_ndo_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, > htonl(mtu)); > goto err_rt; > } > > - gtp_set_pktinfo_ipv4(pktinfo, pctx->sk, iph, pctx, rt, &fl4, dev); > - gtp_push_header(skb, pktinfo); > + gtp_set_pktinfo_ipv4(pktinfo, sk, tos, rt, &fl4, dev); > + > + if (unlikely(opts)) { > + int err; > + > + pktinfo->gtph_port = htons(GTP1U_PORT); > + err = gtp1_push_control_header(skb, tun_id, opts, dev); > + if (err) { > + netdev_info(dev, "cntr pkt error %d", err); > + goto err_rt; > + } > + return 0; > + } > + > + switch (gtp_version) { > + case GTP_V0: > + pktinfo->gtph_port = htons(GTP0_PORT); > + gtp0_push_header(skb, pctx); > + break; > + case GTP_V1: > + pktinfo->gtph_port = htons(GTP1U_PORT); > + gtp1_push_header(skb, tun_id); > + break; > + } > > return 0; > err_rt: > @@ -557,7 +731,6 @@ static int gtp_build_skb_ip4(struct sk_buff *skb, struct net_device *dev, > > static netdev_tx_t gtp_dev_xmit(struct sk_buff *skb, struct net_device *dev) > { > - unsigned int proto = ntohs(skb->protocol); > struct gtp_pktinfo pktinfo; > int err; > > @@ -569,32 +742,22 @@ static netdev_tx_t gtp_dev_xmit(struct sk_buff *skb, struct net_device *dev) > > /* PDP context lookups in gtp_build_skb_*() need rcu read-side lock. */ > rcu_read_lock(); > - switch (proto) { > - case ETH_P_IP: > - err = gtp_build_skb_ip4(skb, dev, &pktinfo); > - break; > - default: > - err = -EOPNOTSUPP; > - break; > - } > + err = gtp_build_skb_ip4(skb, dev, &pktinfo); > rcu_read_unlock(); > > if (err < 0) > goto tx_err; > > - switch (proto) { > - case ETH_P_IP: > - netdev_dbg(pktinfo.dev, "gtp -> IP src: %pI4 dst: %pI4\n", > - &pktinfo.iph->saddr, &pktinfo.iph->daddr); > - udp_tunnel_xmit_skb(pktinfo.rt, pktinfo.sk, skb, > - pktinfo.fl4.saddr, pktinfo.fl4.daddr, > - pktinfo.iph->tos, > - ip4_dst_hoplimit(&pktinfo.rt->dst), > - 0, > - pktinfo.gtph_port, pktinfo.gtph_port, > - true, false); > - break; > - } > + udp_tunnel_xmit_skb(pktinfo.rt, pktinfo.sk, skb, > + pktinfo.fl4.saddr, > + pktinfo.fl4.daddr, > + pktinfo.tos, > + ip4_dst_hoplimit(&pktinfo.rt->dst), > + 0, > + pktinfo.gtph_port, > + pktinfo.gtph_port, > + true, > + false); > > return NETDEV_TX_OK; > tx_err: > @@ -610,6 +773,19 @@ static const struct net_device_ops gtp_netdev_ops = { > .ndo_get_stats64 = dev_get_tstats64, > }; > > +static struct gtp_dev *gtp_find_flow_based_dev(struct net *net) > +{ > + struct gtp_net *gn = net_generic(net, gtp_net_id); > + struct gtp_dev *gtp; > + > + list_for_each_entry(gtp, &gn->gtp_dev_list, list) { > + if (gtp->collect_md) > + return gtp; > + } > + > + return NULL; > +} > + > static void gtp_link_setup(struct net_device *dev) > { > dev->netdev_ops = &gtp_netdev_ops; > @@ -634,7 +810,7 @@ static void gtp_link_setup(struct net_device *dev) > } > > 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_encap_enable(struct gtp_dev *gtp, struct net_device *dev, struct nlattr *data[]); > > static void gtp_destructor(struct net_device *dev) > { > @@ -652,11 +828,24 @@ 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]) > + if (!data[IFLA_GTP_FD0] && !data[IFLA_GTP_FD1] && > + !data[IFLA_GTP_COLLECT_METADATA]) > return -EINVAL; > > gtp = netdev_priv(dev); > > + if (data[IFLA_GTP_COLLECT_METADATA]) { > + if (data[IFLA_GTP_FD0]) { > + netdev_dbg(dev, "LWT device does not support setting v0 socket"); > + return -EINVAL; > + } > + if (gtp_find_flow_based_dev(src_net)) { > + netdev_dbg(dev, "LWT device already exist"); > + return -EBUSY; > + } > + gtp->collect_md = true; > + } > + > if (!data[IFLA_GTP_PDP_HASHSIZE]) { > hashsize = 1024; > } else { > @@ -669,7 +858,7 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, > if (err < 0) > return err; > > - err = gtp_encap_enable(gtp, data); > + err = gtp_encap_enable(gtp, dev, data); > if (err < 0) > goto out_hashtable; > > @@ -683,7 +872,7 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, > list_add_rcu(&gtp->list, &gn->gtp_dev_list); > dev->priv_destructor = gtp_destructor; > > - netdev_dbg(dev, "registered new GTP interface\n"); > + netdev_dbg(dev, "registered new GTP interface %s\n", dev->name); > > return 0; > > @@ -706,6 +895,7 @@ static void gtp_dellink(struct net_device *dev, struct list_head *head) > pdp_context_delete(pctx); > > list_del_rcu(&gtp->list); > + > unregister_netdevice_queue(dev, head); > } > > @@ -714,6 +904,7 @@ 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_COLLECT_METADATA] = { .type = NLA_FLAG }, > }; > > static int gtp_validate(struct nlattr *tb[], struct nlattr *data[], > @@ -737,6 +928,9 @@ static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) > if (nla_put_u32(skb, IFLA_GTP_PDP_HASHSIZE, gtp->hash_size)) > goto nla_put_failure; > > + if (gtp->collect_md && nla_put_flag(skb, IFLA_GTP_COLLECT_METADATA)) > + goto nla_put_failure; > + > return 0; > > nla_put_failure: > @@ -782,35 +976,24 @@ static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize) > return -ENOMEM; > } > > -static struct sock *gtp_encap_enable_socket(int fd, int type, > - struct gtp_dev *gtp) > +static int __gtp_encap_enable_socket(struct socket *sock, 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; > + pr_debug("socket not UDP\n"); > + return -EINVAL; > } > > lock_sock(sk); > if (sk->sk_user_data) { > - sk = ERR_PTR(-EBUSY); > - goto out_rel_sock; > + release_sock(sock->sk); > + return -EBUSY; > } > > sock_hold(sk); > @@ -821,15 +1004,58 @@ static struct sock *gtp_encap_enable_socket(int fd, int type, > 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: > + return 0; > +} > + > +static struct sock *gtp_encap_enable_socket(int fd, int type, > + struct gtp_dev *gtp) > +{ > + struct socket *sock; > + 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; > + } > + err = __gtp_encap_enable_socket(sock, type, gtp); > sockfd_put(sock); > - return sk; > + if (err) > + return ERR_PTR(err); > + > + return sock->sk; > } > > -static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) > +static struct socket *gtp_create_gtp_socket(struct gtp_dev *gtp, struct net_device *dev) > +{ > + struct udp_port_cfg udp_conf; > + struct socket *sock; > + int err; > + > + memset(&udp_conf, 0, sizeof(udp_conf)); > + udp_conf.family = AF_INET; > + udp_conf.local_ip.s_addr = htonl(INADDR_ANY); > + udp_conf.local_udp_port = htons(GTP1U_PORT); > + > + err = udp_sock_create(dev_net(dev), &udp_conf, &sock); > + if (err < 0) { > + pr_debug("create gtp sock failed: %d\n", err); > + return ERR_PTR(err); > + } > + err = __gtp_encap_enable_socket(sock, UDP_ENCAP_GTP1U, gtp); > + if (err) { > + pr_debug("enable gtp sock encap failed: %d\n", err); > + udp_tunnel_sock_release(sock); > + return ERR_PTR(err); > + } > + pr_debug("create gtp sock done\n"); > + return sock; > +} > + > +static int gtp_encap_enable(struct gtp_dev *gtp, struct net_device *dev, struct nlattr *data[]) > { > struct sock *sk1u = NULL; > struct sock *sk0 = NULL; > @@ -853,11 +1079,25 @@ static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]) > } > } > > + if (data[IFLA_GTP_COLLECT_METADATA]) { > + struct socket *sock; > + > + if (!sk1u) { > + sock = gtp_create_gtp_socket(gtp, dev); > + if (IS_ERR(sock)) > + return PTR_ERR(sock); > + > + gtp->collect_md_sock = sock; > + sk1u = sock->sk; > + } else { > + gtp->collect_md_sock = NULL; > + } > + } > + > 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); > + gtp_encap_disable(gtp); > return -EINVAL; > } > } > @@ -1416,7 +1656,7 @@ static int __init gtp_init(void) > if (err < 0) > goto unreg_genl_family; > > - pr_info("GTP module loaded (pdp ctx size %zd bytes)\n", > + pr_info("GTP module loaded (pdp ctx size %zd bytes) with tnl-md support\n", > sizeof(struct pdp_ctx)); > return 0; > > diff --git a/include/uapi/linux/gtp.h b/include/uapi/linux/gtp.h > index 79f9191bbb24..62aff78b7c56 100644 > --- a/include/uapi/linux/gtp.h > +++ b/include/uapi/linux/gtp.h > @@ -2,6 +2,8 @@ > #ifndef _UAPI_LINUX_GTP_H_ > #define _UAPI_LINUX_GTP_H_ > > +#include <linux/types.h> > + > #define GTP_GENL_MCGRP_NAME "gtp" > > enum gtp_genl_cmds { > @@ -34,4 +36,14 @@ enum gtp_attrs { > }; > #define GTPA_MAX (__GTPA_MAX + 1) > > +enum { > + GTP_METADATA_V1 > +}; > + > +struct gtpu_metadata { > + __u8 ver; > + __u8 flags; > + __u8 type; > +}; > + > #endif /* _UAPI_LINUX_GTP_H_ */ > diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h > index 874cc12a34d9..15117fead4b1 100644 > --- a/include/uapi/linux/if_link.h > +++ b/include/uapi/linux/if_link.h > @@ -808,6 +808,7 @@ enum { > IFLA_GTP_FD1, > IFLA_GTP_PDP_HASHSIZE, > IFLA_GTP_ROLE, > + IFLA_GTP_COLLECT_METADATA, > __IFLA_GTP_MAX, > }; > #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1) > diff --git a/include/uapi/linux/if_tunnel.h b/include/uapi/linux/if_tunnel.h > index 7d9105533c7b..802da679fab1 100644 > --- a/include/uapi/linux/if_tunnel.h > +++ b/include/uapi/linux/if_tunnel.h > @@ -176,6 +176,7 @@ 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_GTPU_OPT __cpu_to_be16(0x8000) > > #define TUNNEL_OPTIONS_PRESENT \ > (TUNNEL_GENEVE_OPT | TUNNEL_VXLAN_OPT | TUNNEL_ERSPAN_OPT) > diff --git a/tools/include/uapi/linux/if_link.h b/tools/include/uapi/linux/if_link.h > index d208b2af697f..28d649bda686 100644 > --- a/tools/include/uapi/linux/if_link.h > +++ b/tools/include/uapi/linux/if_link.h > @@ -617,6 +617,7 @@ enum { > IFLA_GTP_FD1, > IFLA_GTP_PDP_HASHSIZE, > IFLA_GTP_ROLE, > + IFLA_GTP_COLLECT_METADATA, > __IFLA_GTP_MAX, > }; > #define IFLA_GTP_MAX (__IFLA_GTP_MAX - 1)

2 1

GTP kernel module patchset discussion on netdev
by Harald Welte 12 Dec '20

12 Dec '20

In case anyone is interested, there's two larger patches/patchsets for the Linux kernel GTP module currently submitted and under discussion on the netdev mailing list. I'm not sure why this list was not copied, despite it being listed in the MAINTAINERS file. See https://lore.kernel.org/netdev/20201211122612.869225-1-jonas@norrbonn.se/T/… and https://lore.kernel.org/netdev/20201212044017.55865-1-pbshelar@fb.com/T/#u Any feedback / comments / review or even testing is appreciated. -- - Harald Welte <laforge(a)gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)

1 0

[PATCH net-next v3 00/10] net: add and use dev_get_tstats64
by Heiner Kallweit 10 Nov '20

10 Nov '20

It's a frequent pattern to use netdev->stats for the less frequently accessed counters and per-cpu counters for the frequently accessed counters (rx/tx bytes/packets). Add a default ndo_get_stats64() implementation for this use case. Subsequently switch more drivers to use this pattern. v2: - add patches for replacing ip_tunnel_get_stats64 Requested additional migrations will come in a separate series. v3: - add atomic_long_t member rx_frame_errors in patch 3 for making counter updates atomic Heiner Kallweit (10): net: core: add dev_get_tstats64 as a ndo_get_stats64 implementation net: dsa: use net core stats64 handling tun: switch to net core provided statistics counters ip6_tunnel: use ip_tunnel_get_stats64 as ndo_get_stats64 callback net: switch to dev_get_tstats64 gtp: switch to dev_get_tstats64 wireguard: switch to dev_get_tstats64 vti: switch to dev_get_tstats64 ipv4/ipv6: switch to dev_get_tstats64 net: remove ip_tunnel_get_stats64 drivers/net/bareudp.c | 2 +- drivers/net/geneve.c | 2 +- drivers/net/gtp.c | 2 +- drivers/net/tun.c | 121 +++++++++------------------------ drivers/net/vxlan.c | 4 +- drivers/net/wireguard/device.c | 2 +- include/linux/netdevice.h | 1 + include/net/ip_tunnels.h | 2 - net/core/dev.c | 15 ++++ net/dsa/dsa.c | 7 +- net/dsa/dsa_priv.h | 2 - net/dsa/slave.c | 29 ++------ net/ipv4/ip_gre.c | 6 +- net/ipv4/ip_tunnel_core.c | 9 --- net/ipv4/ip_vti.c | 2 +- net/ipv4/ipip.c | 2 +- net/ipv6/ip6_gre.c | 6 +- net/ipv6/ip6_tunnel.c | 32 +-------- net/ipv6/ip6_vti.c | 2 +- net/ipv6/sit.c | 2 +- 20 files changed, 75 insertions(+), 175 deletions(-) -- 2.29.2

2 11

[PATCH net-next v2 00/10] net: add and use dev_get_tstats64
by Heiner Kallweit 06 Nov '20

06 Nov '20

It's a frequent pattern to use netdev->stats for the less frequently accessed counters and per-cpu counters for the frequently accessed counters (rx/tx bytes/packets). Add a default ndo_get_stats64() implementation for this use case. Subsequently switch more drivers to use this pattern. v2: - add patches for replacing ip_tunnel_get_stats64 Requested additional migrations will come in a separate series. Heiner Kallweit (10): net: core: add dev_get_tstats64 as a ndo_get_stats64 implementation net: dsa: use net core stats64 handling tun: switch to net core provided statistics counters ip6_tunnel: switch to dev_get_tstats64 net: switch to dev_get_tstats64 gtp: switch to dev_get_tstats64 wireguard: switch to dev_get_tstats64 vti: switch to dev_get_tstats64 ipv4/ipv6: switch to dev_get_tstats64 net: remove ip_tunnel_get_stats64 drivers/net/bareudp.c | 2 +- drivers/net/geneve.c | 2 +- drivers/net/gtp.c | 2 +- drivers/net/tun.c | 127 ++++++++------------------------- drivers/net/vxlan.c | 4 +- drivers/net/wireguard/device.c | 2 +- include/linux/netdevice.h | 1 + include/net/ip_tunnels.h | 2 - net/core/dev.c | 15 ++++ net/dsa/dsa.c | 7 +- net/dsa/dsa_priv.h | 2 - net/dsa/slave.c | 29 ++------ net/ipv4/ip_gre.c | 6 +- net/ipv4/ip_tunnel_core.c | 9 --- net/ipv4/ip_vti.c | 2 +- net/ipv4/ipip.c | 2 +- net/ipv6/ip6_gre.c | 6 +- net/ipv6/ip6_tunnel.c | 32 +-------- net/ipv6/ip6_vti.c | 2 +- net/ipv6/sit.c | 2 +- 20 files changed, 72 insertions(+), 184 deletions(-) -- 2.29.2

4 16

[PATCH net] gtp: fix an use-before-init in gtp_newlink()
by Masahiro Fujiwara 29 Oct '20

29 Oct '20

*_pdp_find() from gtp_encap_recv() would trigger a crash when a peer sends GTP packets while creating new GTP device. RIP: 0010:gtp1_pdp_find.isra.0+0x68/0x90 [gtp] <SNIP> Call Trace: <IRQ> gtp_encap_recv+0xc2/0x2e0 [gtp] ? gtp1_pdp_find.isra.0+0x90/0x90 [gtp] udp_queue_rcv_one_skb+0x1fe/0x530 udp_queue_rcv_skb+0x40/0x1b0 udp_unicast_rcv_skb.isra.0+0x78/0x90 __udp4_lib_rcv+0x5af/0xc70 udp_rcv+0x1a/0x20 ip_protocol_deliver_rcu+0xc5/0x1b0 ip_local_deliver_finish+0x48/0x50 ip_local_deliver+0xe5/0xf0 ? ip_protocol_deliver_rcu+0x1b0/0x1b0 gtp_encap_enable() should be called after gtp_hastable_new() otherwise *_pdp_find() will access the uninitialized hash table. Fixes: 1e3a3abd8 ("gtp: make GTP sockets in gtp_newlink optional") Signed-off-by: Masahiro Fujiwara <fujiwara.masahiro(a)gmail.com> --- drivers/net/gtp.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 8e47d0112e5d..6c56337b02a3 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -663,10 +663,6 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, gtp = netdev_priv(dev); - err = gtp_encap_enable(gtp, data); - if (err < 0) - return err; - if (!data[IFLA_GTP_PDP_HASHSIZE]) { hashsize = 1024; } else { @@ -676,13 +672,18 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, } err = gtp_hashtable_new(gtp, hashsize); + if (err < 0) { + return err; + } + + err = gtp_encap_enable(gtp, data); if (err < 0) goto out_encap; err = register_netdevice(dev); if (err < 0) { netdev_dbg(dev, "failed to register new netdev %d\n", err); - goto out_hashtable; + goto out_encap; } gn = net_generic(dev_net(dev), gtp_net_id); @@ -693,11 +694,10 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, return 0; -out_hashtable: - kfree(gtp->addr_hash); - kfree(gtp->tid_hash); out_encap: gtp_encap_disable(gtp); + kfree(gtp->addr_hash); + kfree(gtp->tid_hash); return err; } -- 2.24.3

2 7

IPv6 PDP type support
by Jean-Marc Katembwe 21 Oct '20

21 Oct '20

Hi, I would like to know if the current OsmoSGSN version can support IPv6 PDP type. Thanks, Jean-Marc

2 1

[RFC PATCH 3/3] selinux: Add SELinux GTP support
by Richard Haines 17 Oct '20

17 Oct '20

The SELinux GTP implementation is explained in: Documentation/security/GTP.rst Signed-off-by: Richard Haines <richard_c_haines(a)btinternet.com> --- Documentation/security/GTP.rst | 59 ++++++++++++++++++++++++++ security/selinux/hooks.c | 64 +++++++++++++++++++++++++++++ security/selinux/include/classmap.h | 2 + security/selinux/include/objsec.h | 4 ++ 4 files changed, 129 insertions(+) diff --git a/Documentation/security/GTP.rst b/Documentation/security/GTP.rst index e307d0b59..55e41ecd0 100644 --- a/Documentation/security/GTP.rst +++ b/Documentation/security/GTP.rst @@ -15,6 +15,9 @@ For security module support, three GTP specific hooks have been implemented:: security_gtp_dev_del() security_gtp_dev_cmd() +The usage of these hooks are described below with the SELinux implementation +described in the `GTP SELinux Support`_ chapter. + security_gtp_dev_add() ~~~~~~~~~~~~~~~~~~~~~~ @@ -37,3 +40,59 @@ zero on success, negative values on failure. The commands are based on values from ``include/uapi/linux/gtp.h`` as follows:: ``enum gtp_genl_cmds { GTP_CMD_NEWPDP, GTP_CMD_DELPDP, GTP_CMD_GETPDP };`` + + +GTP SELinux Support +=================== + +Policy Statements +----------------- +The following class and permissions to support GTP are available within the +kernel:: + + class gtp { add del get } + +The permissions are described in the sections that follow. + +Security Hooks +-------------- + +The `GTP LSM Support`_ chapter above describes the following GTP security +hooks with the SELinux specifics expanded below:: + + security_gtp_dev_add(&gtp->security) + security_gtp_dev_del(gtp->security) + security_gtp_dev_cmd(gtp->security, cmd) + + +security_gtp_dev_add() +~~~~~~~~~~~~~~~~~~~~~~ +Allocates a security structure for a GTP device provided the caller has the +``gtp { add }`` permission. Can return errors ``-ENOMEM`` or ``-EACCES``. +Returns zero if the security structure has been allocated. + + +security_gtp_dev_del() +~~~~~~~~~~~~~~~~~~~~~~ +Frees a security structure for a GTP device provided the caller has the +``gtp { del }`` permission. Can return error ``-EACCES``. Returns zero if the +security structure has been freed. + + +security_gtp_dev_cmd() +~~~~~~~~~~~~~~~~~~~~~~ +Validate if the caller (current SID) and the GTP device SID have the required +permission to perform the operation. The GTP/SELinux permission map as follow:: + + GTP_CMD_NEWPDP = gtp { add } + GTP_CMD_DELPDP = gtp { del } + GTP_CMD_GETPDP = gtp { get } + +Returns ``-EACCES`` if denied or zero if allowed. + +NOTES: + 1) If the GTP device has the ``{ add }`` permission it can add device and + also add PDP's. + + 2) If the GTP device has the ``{ del }`` permission it can delete a device + and also delete PDP's. diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index d6b182c11..74b2bea87 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -5520,6 +5520,67 @@ static int selinux_tun_dev_open(void *security) return 0; } +static int selinux_gtp_dev_add(void **security) +{ + struct gtp_security_struct *gtpsec = NULL; + u32 sid = current_sid(); + int err; + + err = avc_has_perm(&selinux_state, sid, sid, + SECCLASS_GTP, GTP__ADD, NULL); + if (err < 0) + return err; + + gtpsec = kzalloc(sizeof(*gtpsec), GFP_KERNEL); + if (!gtpsec) + return -ENOMEM; + + gtpsec->sid = sid; + *security = gtpsec; + + return 0; +} + +static int selinux_gtp_dev_del(void *security) +{ + struct gtp_security_struct *gtpsec = security; + u32 sid = current_sid(); + int err; + + err = avc_has_perm(&selinux_state, sid, gtpsec->sid, + SECCLASS_GTP, GTP__DEL, NULL); + if (err < 0) + return err; + + kfree(security); + + return 0; +} + +static int selinux_gtp_dev_cmd(void *security, enum gtp_genl_cmds cmd) +{ + struct gtp_security_struct *gtpsec = security; + u32 perm, sid = current_sid(); + + switch (cmd) { + case GTP_CMD_NEWPDP: + perm = GTP__ADD; + break; + case GTP_CMD_DELPDP: + perm = GTP__DEL; + break; + case GTP_CMD_GETPDP: + perm = GTP__GET; + break; + default: + WARN_ON(1); + return -EPERM; + } + + return avc_has_perm(&selinux_state, sid, gtpsec->sid, + SECCLASS_GTP, perm, NULL); +} + #ifdef CONFIG_NETFILTER static unsigned int selinux_ip_forward(struct sk_buff *skb, @@ -7130,6 +7191,9 @@ static struct security_hook_list selinux_hooks[] __lsm_ro_after_init = { LSM_HOOK_INIT(tun_dev_attach_queue, selinux_tun_dev_attach_queue), LSM_HOOK_INIT(tun_dev_attach, selinux_tun_dev_attach), LSM_HOOK_INIT(tun_dev_open, selinux_tun_dev_open), + LSM_HOOK_INIT(gtp_dev_add, selinux_gtp_dev_add), + LSM_HOOK_INIT(gtp_dev_del, selinux_gtp_dev_del), + LSM_HOOK_INIT(gtp_dev_cmd, selinux_gtp_dev_cmd), #ifdef CONFIG_SECURITY_INFINIBAND LSM_HOOK_INIT(ib_pkey_access, selinux_ib_pkey_access), LSM_HOOK_INIT(ib_endport_manage_subnet, diff --git a/security/selinux/include/classmap.h b/security/selinux/include/classmap.h index 40cebde62..3865a4549 100644 --- a/security/selinux/include/classmap.h +++ b/security/selinux/include/classmap.h @@ -249,6 +249,8 @@ struct security_class_mapping secclass_map[] = { {"open", "cpu", "kernel", "tracepoint", "read", "write"} }, { "lockdown", { "integrity", "confidentiality", NULL } }, + { "gtp", + { "add", "del", "get", NULL } }, { NULL } }; diff --git a/security/selinux/include/objsec.h b/security/selinux/include/objsec.h index 330b7b6d4..311ffb6ea 100644 --- a/security/selinux/include/objsec.h +++ b/security/selinux/include/objsec.h @@ -148,6 +148,10 @@ struct perf_event_security_struct { u32 sid; /* SID of perf_event obj creator */ }; +struct gtp_security_struct { + u32 sid; /* SID of gtp obj creator */ +}; + extern struct lsm_blob_sizes selinux_blob_sizes; static inline struct task_security_struct *selinux_cred(const struct cred *cred) { -- 2.26.2

1 0

[RFC PATCH 2/3] gtp: Add LSM hooks to GPRS Tunneling Protocol (GTP)
by Richard Haines 17 Oct '20

17 Oct '20

Add security hooks to allow security modules to exercise access control over GTP. Signed-off-by: Richard Haines <richard_c_haines(a)btinternet.com> --- drivers/net/gtp.c | 49 ++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 48 insertions(+), 1 deletion(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 21640a035..ee00b12ab 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -73,6 +73,8 @@ struct gtp_dev { unsigned int hash_size; struct hlist_head *tid_hash; struct hlist_head *addr_hash; + + void *security; }; static unsigned int gtp_net_id __read_mostly; @@ -663,6 +665,12 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, gtp = netdev_priv(dev); + err = security_gtp_dev_add(&gtp->security); + pr_debug("security_gtp_dev_add() ptr: %p err: %d\n", + gtp->security, err); + if (err < 0) + return err; + err = gtp_encap_enable(gtp, data); if (err < 0) return err; @@ -705,7 +713,15 @@ static void gtp_dellink(struct net_device *dev, struct list_head *head) { struct gtp_dev *gtp = netdev_priv(dev); struct pdp_ctx *pctx; - int i; + int i, err; + + err = security_gtp_dev_del(gtp->security); + pr_debug("security_gtp_dev_del() ptr: %p err: %d\n", + gtp->security, err); + if (err < 0) { + pr_err("Failed security_gtp_dev_del_link() err: %d\n", err); + return; + } for (i = 0; i < gtp->hash_size; i++) hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], hlist_tid) @@ -1076,6 +1092,12 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) goto out_unlock; } + err = security_gtp_dev_cmd(gtp->security, GTP_CMD_NEWPDP); + pr_debug("security_gtp_dev_cmd(GTP_CMD_NEWPDP) ptr: %p err: %d\n", + gtp->security, err); + if (err < 0) + goto out_unlock; + if (version == GTP_V0) sk = gtp->sk0; else if (version == GTP_V1) @@ -1139,6 +1161,7 @@ static struct pdp_ctx *gtp_find_pdp(struct net *net, struct nlattr *nla[]) static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) { struct pdp_ctx *pctx; + struct gtp_dev *gtp; int err = 0; if (!info->attrs[GTPA_VERSION]) @@ -1152,6 +1175,13 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) goto out_unlock; } + gtp = netdev_priv(pctx->dev); + err = security_gtp_dev_cmd(gtp->security, GTP_CMD_DELPDP); + pr_debug("security_gtp_dev_cmd(GTP_CMD_DELPDP) ptr: %p err: %d\n", + gtp->security, err); + if (err < 0) + goto out_unlock; + if (pctx->gtp_version == GTP_V0) netdev_dbg(pctx->dev, "GTPv0-U: deleting tunnel id = %llx (pdp %p)\n", pctx->u.v0.tid, pctx); @@ -1208,6 +1238,7 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) { struct pdp_ctx *pctx = NULL; struct sk_buff *skb2; + struct gtp_dev *gtp; int err; if (!info->attrs[GTPA_VERSION]) @@ -1221,6 +1252,13 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } + gtp = netdev_priv(pctx->dev); + err = security_gtp_dev_cmd(gtp->security, GTP_CMD_GETPDP); + pr_debug("security_gtp_dev_cmd(GTP_CMD_GETPDP) ptr: %p err: %d\n", + gtp->security, err); + if (err < 0) + goto err_unlock; + skb2 = genlmsg_new(NLMSG_GOODSIZE, GFP_ATOMIC); if (skb2 == NULL) { err = -ENOMEM; @@ -1250,6 +1288,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, struct net *net = sock_net(skb->sk); struct pdp_ctx *pctx; struct gtp_net *gn; + int err; gn = net_generic(net, gtp_net_id); @@ -1263,6 +1302,14 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, else last_gtp = NULL; + err = security_gtp_dev_cmd(gtp->security, GTP_CMD_GETPDP); + pr_debug("security_gtp_dev_cmd(GTP_CMD_GETPDP) ptr: %p err: %d\n", + gtp->security, err); + if (err < 0) { + rcu_read_unlock(); + return err; + } + for (i = bucket; i < gtp->hash_size; i++) { j = 0; hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], -- 2.26.2

1 0

[RFC PATCH 1/3] security: Add GPRS Tunneling Protocol (GTP) security hooks
by Richard Haines 16 Oct '20

16 Oct '20

The GTP security hooks are explained in: Documentation/security/GTP.rst Signed-off-by: Richard Haines <richard_c_haines(a)btinternet.com> --- Documentation/security/GTP.rst | 39 ++++++++++++++++++++++++++++++++ Documentation/security/index.rst | 1 + include/linux/lsm_hook_defs.h | 3 +++ include/linux/lsm_hooks.h | 16 +++++++++++++ include/linux/security.h | 19 ++++++++++++++++ security/security.c | 18 +++++++++++++++ 6 files changed, 96 insertions(+) create mode 100644 Documentation/security/GTP.rst diff --git a/Documentation/security/GTP.rst b/Documentation/security/GTP.rst new file mode 100644 index 000000000..e307d0b59 --- /dev/null +++ b/Documentation/security/GTP.rst @@ -0,0 +1,39 @@ +.. SPDX-License-Identifier: GPL-2.0 + +============================= +GPRS Tunneling Protocol (GTP) +============================= + +GTP LSM Support +=============== + +Security Hooks +-------------- +For security module support, three GTP specific hooks have been implemented:: + + security_gtp_dev_add() + security_gtp_dev_del() + security_gtp_dev_cmd() + + +security_gtp_dev_add() +~~~~~~~~~~~~~~~~~~~~~~ +Allows a module to allocate a security structure for a GTP device. Returns a +zero on success, negative values on failure. +If successful the GTP device ``struct gtp_dev`` will hold the allocated +pointer in ``void *security;``. + + +security_gtp_dev_del() +~~~~~~~~~~~~~~~~~~~~~~ +Allows a module to free the security structure for a GTP device. Returns a +zero on success, negative values on failure. + + +security_gtp_dev_cmd() +~~~~~~~~~~~~~~~~~~~~~~ +Allows a module to validate a command for the selected GTP device. Returns a +zero on success, negative values on failure. The commands are based on values +from ``include/uapi/linux/gtp.h`` as follows:: + +``enum gtp_genl_cmds { GTP_CMD_NEWPDP, GTP_CMD_DELPDP, GTP_CMD_GETPDP };`` diff --git a/Documentation/security/index.rst b/Documentation/security/index.rst index 8129405eb..cdbdaa83b 100644 --- a/Documentation/security/index.rst +++ b/Documentation/security/index.rst @@ -16,3 +16,4 @@ Security Documentation siphash tpm/index digsig + GTP diff --git a/include/linux/lsm_hook_defs.h b/include/linux/lsm_hook_defs.h index 2a8c74d99..a994417fb 100644 --- a/include/linux/lsm_hook_defs.h +++ b/include/linux/lsm_hook_defs.h @@ -322,6 +322,9 @@ LSM_HOOK(int, 0, sctp_bind_connect, struct sock *sk, int optname, struct sockaddr *address, int addrlen) LSM_HOOK(void, LSM_RET_VOID, sctp_sk_clone, struct sctp_endpoint *ep, struct sock *sk, struct sock *newsk) +LSM_HOOK(int, 0, gtp_dev_add, void **security) +LSM_HOOK(int, 0, gtp_dev_del, void *security) +LSM_HOOK(int, 0, gtp_dev_cmd, void *security, enum gtp_genl_cmds cmd) #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_INFINIBAND diff --git a/include/linux/lsm_hooks.h b/include/linux/lsm_hooks.h index 9e2e3e637..3d6888d51 100644 --- a/include/linux/lsm_hooks.h +++ b/include/linux/lsm_hooks.h @@ -982,6 +982,22 @@ * This hook can be used by the module to update any security state * associated with the TUN device's security structure. * @security pointer to the TUN devices's security structure. + * @gtp_dev_add: + * This hook allows a module to allocate a security structure for a GTP + * device. + * @security pointer to a security structure pointer. + * Returns a zero on success, negative values on failure. + * @gtp_dev_del: + * This hook allows a module to free the security structure for a GTP + * device. + * @security pointer to the GTP device's security structure. + * Returns a zero on success, negative values on failure. + * @gtp_dev_cmd: + * This hook allows a module to free the security structure for a GTP + * device. + * @security pointer to the GTP device's security structure. + * @cmd contains the GTP command. + * Returns a zero on success, negative values on failure. * * Security hooks for SCTP * diff --git a/include/linux/security.h b/include/linux/security.h index 0a0a03b36..67ff43afa 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -30,6 +30,7 @@ #include <linux/err.h> #include <linux/string.h> #include <linux/mm.h> +#include <linux/gtp.h> struct linux_binprm; struct cred; @@ -1365,6 +1366,9 @@ int security_sctp_bind_connect(struct sock *sk, int optname, struct sockaddr *address, int addrlen); void security_sctp_sk_clone(struct sctp_endpoint *ep, struct sock *sk, struct sock *newsk); +int security_gtp_dev_add(void **security); +int security_gtp_dev_del(void *security); +int security_gtp_dev_cmd(void *security, enum gtp_genl_cmds cmd); #else /* CONFIG_SECURITY_NETWORK */ static inline int security_unix_stream_connect(struct sock *sock, @@ -1582,6 +1586,21 @@ static inline void security_sctp_sk_clone(struct sctp_endpoint *ep, struct sock *newsk) { } + +static inline int security_gtp_dev_add(void **security) +{ + return 0; +} + +static inline int security_gtp_dev_del(void *security) +{ + return 0; +} + +static inline int security_gtp_dev_cmd(void *security, enum gtp_genl_cmds cmd) +{ + return 0; +} #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_INFINIBAND diff --git a/security/security.c b/security/security.c index 70a7ad357..63b656848 100644 --- a/security/security.c +++ b/security/security.c @@ -2304,6 +2304,24 @@ void security_sctp_sk_clone(struct sctp_endpoint *ep, struct sock *sk, } EXPORT_SYMBOL(security_sctp_sk_clone); +int security_gtp_dev_add(void **security) +{ + return call_int_hook(gtp_dev_add, 0, security); +} +EXPORT_SYMBOL(security_gtp_dev_add); + +int security_gtp_dev_del(void *security) +{ + return call_int_hook(gtp_dev_del, 0, security); +} +EXPORT_SYMBOL(security_gtp_dev_del); + +int security_gtp_dev_cmd(void *security, enum gtp_genl_cmds cmd) +{ + return call_int_hook(gtp_dev_cmd, 0, security, cmd); +} +EXPORT_SYMBOL(security_gtp_dev_cmd); + #endif /* CONFIG_SECURITY_NETWORK */ #ifdef CONFIG_SECURITY_INFINIBAND -- 2.26.2

2 1

[PATCH 0/3] Add LSM/SELinux support for GPRS Tunneling Protocol (GTP)
by Richard Haines 16 Oct '20

16 Oct '20

These patches came about after looking at 5G open source in particular the updated 5G GTP driver at [1]. As this driver is still under development, added the LSM/SELinux hooks to the current stable GTP version in kernel selinux-next [2]. Similar hooks have also been implemented in [1] as it uses the same base code as the current 3G version (except that it handles different packet types). To test the 3G GTP driver there is an RFC patch for the selinux-testsuite at [3]. To enable the selinux-testsuite GTP tests, the libgtpnl [4] library and tools needed to be modified to: Return ERRNO on error to detect EACCES, Add gtp_match_tunnel function, Allow gtp-link to specify port numbers for multiple instances to run in the same namespace. A patch for libgtpnl is supplied in the selinux-testsuite patch as well as setup/test instructions (libgtpnl is not packaged by Fedora) These patches were tested on Fedora 32 with kernel [2] using the 'targeted' policy. Also ran the Linux Kernel GTP-U basic tests [5]. [1] https://github.com/PrinzOwO/gtp5g [2] https://github.com/SELinuxProject/selinux-kernel [3] https://lore.kernel.org/selinux/20200924085314.6120-1-richard_c_haines@btin… [4] git://git.osmocom.org/libgtpnl.git [5] https://osmocom.org/projects/linux-kernel-gtp-u/wiki/Basic_Testing Changes from RFC patches: 1) Rework the LSM hook names and contents as suggested by Casey Schaufler. This resulted in moving the gtp_dev struct from gtp.c to include/net/gtp.h so that it is visible to the LSM modules for storing the security blob. 2) Remove pr_debug's from gtp.c security_* calls. 3) Minor GTP.rst updates. 4) Added netdev to distribution list. Richard Haines (3): security: Add GPRS Tunneling Protocol (GTP) security hooks gtp: Add LSM hooks to GPRS Tunneling Protocol (GTP) selinux: Add SELinux GTP support Documentation/security/GTP.rst | 100 ++++++++++++++++++++++++++++ Documentation/security/index.rst | 1 + drivers/net/gtp.c | 50 +++++++++----- include/linux/lsm_hook_defs.h | 3 + include/linux/lsm_hooks.h | 13 ++++ include/linux/security.h | 22 ++++++ include/net/gtp.h | 21 ++++++ security/security.c | 18 +++++ security/selinux/hooks.c | 66 ++++++++++++++++++ security/selinux/include/classmap.h | 2 + security/selinux/include/objsec.h | 4 ++ 11 files changed, 284 insertions(+), 16 deletions(-) create mode 100644 Documentation/security/GTP.rst -- 2.26.2

5 15

[RFC] NS VTY interface
by Alexander 'lynxis' Couzens 16 Oct '20

16 Oct '20

Hi, with the new NS2 code [1] the vty keep compatible. However the new NS2 code is more flexible and support more use cases except the old vty doesn't allow it. I've created an osmocom ticket [2] to experiment and write a draft of the new VTY commands. I would be nice if you can give me some feedback here. Best, lynxis [1] NS <> NS2. it's still the same protocol, just a second implementation of NS. [2] https://osmocom.org/issues/4770 new vty example: 1.1. sgsn with udp bind - ipa style 1.2. sgsn with udp bind - with persistent nsvc 1.3. sgsn with sns listen on all interfaces 1.4. sgsn with sns listen on specific interface 1.5. sgsn with fr 2.1. TODO: pcu with binds to use when receiving info ind 2.2. TODO: pcu with static configuration <pre> ;; 1.1. sgsn with udp bind - dynamic with ipa style ns bind udp some listen 192.168.0.2 23000 allow-block-reset-nsvci </pre> <pre> ;; 1.2. sgsn with udp bind - no dynamic connection - only vty configured NSE (persistent) ns bind udp some listen 192.168.0.2 23000 nsei 1001 nsvc udp some 10.0.1.2 23000 </pre> <pre> ;; 1.2. sgsn with udp bind - no dynamic connection - only vty configured NSE (persistent) ns bind udp some listen 192.168.0.2 23000 bind udp other listen 192.168.1.2 23000 nsei 1001 ;; will use all bindings nsvc udp * 10.0.1.2 23000 nsei 1002 nsvc udp other 10.0.1.3 23000 nsvc udp some 10.0.1.3 23000 </pre> <pre> ;; 1.3. sgsn with sns listen on all interfaces ns bind udp some listen 192.168.0.2 23000 allow-sns group all bind udp other listen 192.168.1.2 23000 allow-sns group all </pre> <pre> ;; 1.4. sgsn with sns listen on specific interface ns bind udp some listen 192.168.0.2 23000 allow-sns group alice bind udp other listen 192.168.1.2 23000 allow-sns group bob </pre> <pre> ;; 1.5. sgsn with fr fr link hdlc1 link hdlc2 ns nsei 1002 nsvc fr hdlc1 dlci 10 nsvc fr hdlc2 dlci 10 </pre> -- Alexander Couzens mail: lynxis(a)fe80.eu jabber: lynxis(a)fe80.eu gpg: 390D CF78 8BF9 AA50 4F8F F1E2 C29E 9DA6 A0DF 8604

1 0

[RFC PATCH 0/1] selinux-testsuite: Add GTP tests
by Richard Haines 16 Oct '20

16 Oct '20

These tests were performed on Fedora 32 Workstation with the 'targeted' policy using an updated kernel with the GTP RFC kernel patches. See the tests/gtp/GTP-README for detailed setup/testing. The libgtpnl patch is at: tests/gtp/Update-libgtpnl-for-SELinux-testsuite-GTP-tests.patch The basic process is: 1) Build kernel with RFC LSM/GTP/SELinux patches (use the selinux-testsuite/defconfig file to set correct config if required). 2) Setup the selinux-testsuite and add the GTP test patch. 3) Build/install the libgtpnl library using a patch from the selinux-testsuite as this will allow errors to be returned by the library functions. 4) Build/run selinux-testsuite for all SELinux tests (sets up the environment). 5) Run the GTP tests locally to see more info. 6) Remove the selinux-testsuite policy. Richard Haines (1): selinux-testsuite: Add GTP tests defconfig | 4 + policy/Makefile | 5 + policy/test_gtp.cil | 18 + policy/test_gtp.te | 96 +++++ tests/Makefile | 4 + tests/gtp/.gitignore | 2 + tests/gtp/GTP-README | 123 ++++++ tests/gtp/Makefile | 8 + ...tpnl-for-SELinux-testsuite-GTP-tests.patch | 366 ++++++++++++++++++ tests/gtp/gtp-link.c | 134 +++++++ tests/gtp/gtp-tunnel.c | 263 +++++++++++++ tests/gtp/test | 169 ++++++++ 12 files changed, 1192 insertions(+) create mode 100644 policy/test_gtp.cil create mode 100644 policy/test_gtp.te create mode 100644 tests/gtp/.gitignore create mode 100644 tests/gtp/GTP-README create mode 100644 tests/gtp/Makefile create mode 100644 tests/gtp/Update-libgtpnl-for-SELinux-testsuite-GTP-tests.patch create mode 100644 tests/gtp/gtp-link.c create mode 100644 tests/gtp/gtp-tunnel.c create mode 100755 tests/gtp/test -- 2.26.2

1 1

ns2 thoughts
by Harald Welte 21 Sep '20

21 Sep '20

Hi List and Lynxis, I've been reading through the ns2 code more thoroughly and had some thoughts for improvement. As we have no users yet, and the code is unreleased, we can still make changes now. == unused argument in ns2_recv_vc int ns2_recv_vc(struct gprs_ns2_inst *nsi, struct gprs_ns2_vc *nsvc, struct msgb *msg) The 'nsi' is not used and should be redundant, as the nsvc has a back-pointer anyway, right? == consider using an osmo_ prefix to all symbols / types The fact that the old code doesn't have that is a tribute to its age, and not something we need to keep. The current code has quite a bit of 'gprs_ns2' prefixing for types, but not for the symbols/functions. At least that inconsistency should be resolved, so all have the same prefix, even if it is without osmo. == const-ify input arguments If arguments pointing to data are used read-only, let's express that by the const qualifier. The easy ones are the likes of "gprs_ns2_is_frgre_bind(struct gprs_ns2_vc_bind *bind)", but there are definitely many more. == use of msgb->dst In several other osmocom implementations we use msgb->dst to point to the underlying element receiving or transmitting a message. So I could imagine ns2_recv_vc() not only without the 'nsi' argument, but also without the 'nsvc' argument, if we introduce the convention that msgb->dst would always point to the ns-vc. Not sure here, it has pros and cons. Just brainstorming. == output arguments vs. return value There are functions like gprs_ns2_find_vc_by_sockaddr() where the result is not returned, but rather a **pointer output argument is used. What is the rationale here? I don't understand what benefit the extra 0/1 return value gives. The only other return value is -EINVAL if the **ptr was NULL - an error situation that wouldn't exist if we simply returned the pointer from the function. 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)

3 2

GSM, GPRS
by Сергей Егоров 20 Sep '20

20 Sep '20

Hi! I with my friends-students got a task to investigate ciphering algorithm GEA3, using in GPRS/EDGE. While communication MS and Osmocom base station (BS) we've got the LLC-message: LLC MESSAGE (ciphered): 01C00F1840623297A594EE196714A2653D23D5A2AC52792F0434 with the following parameters: Cipher Algorithm: GEA3 Kc_GPRS: A8FC3A996A80D000 SAPI: 1 NU: 3 IOV_UI: 0 OC: 0 After applying algorithm GEA3 we've got deciphered LLC-message: LLC MESSAGE (deciphered): 01C00F0802012A0452F00300010017161805F4D7BF04CA 26176B We get, calculated FCS (6B1726) using deciphered LLC-message and contained in LLC-message FCS (6B1726) are equal. While communication MS and the real BS we've got the following LLC-message: LLC MESSAGE (ciphered) 41C00B9299A1EB51A1AD1FE71633786B23CBD8E6D41C9F658C89C9544AF2BAAC35 with the following parameters: Cipher Algorithm: GEA3 KC_GPRS: 8f94a69c3d9bdf48 SAPI: 1 NU: 2 IOV_UI: 0x10000000 (got from XID) OC: from 0 to 8388608 So we tried to apply the OC parameter from 0 to 8388608 to decipher the message (other parameters were not been changed). In every step, we calculated FCS and compared it with contained in LLC-message FCS and had no success. Finally the question: Can the value of IOV_UI (Osmocom BS: 0, real BS: 0x10000000) affect the deciphering, and if yes then how?? With regards, students of the telecommunication department. Thank you for your attention! Best regards, Sergei ----------------------------------------------------------------------

2 1

GSM, GPRS
by Сергей Егоров 18 Sep '20

18 Sep '20

Good afternoon! I with my friends-students got a task to investigate ciphering algorithm GEA3, using in GPRS/EDGE. While communication MS and Osmocom base station (BS) we've got the LLC-message: LLC MESSAGE (ciphered): 01C00F1840623297A594EE196714A2653D23D5A2AC52792F0434 with the following parameters: Cipher Algorithm: GEA3 Kc_GPRS: A8FC3A996A80D000 SAPI: 1 NU: 3 IOV_UI: 0 OC: 0 After applying algorithm GEA3 we've got deciphered LLC-message: LLC MESSAGE (deciphered): 01C00F0802012A0452F00300010017161805F4D7BF04CA 26176B We get, calculated FCS (6B1726) using deciphered LLC-message and contained in LLC-message FCS (6B1726) are equal. While communication MS and the real BS we've got the following LLC-message: LLC MESSAGE (ciphered) 41C00B9299A1EB51A1AD1FE71633786B23CBD8E6D41C9F658C89C9544AF2BAAC35 with the following parameters: Cipher Algorithm: GEA3 KC_GPRS: 8f94a69c3d9bdf48 SAPI: 1 NU: 2 IOV_UI: 0x10000000 (got from XID) OC: from 0 to 8388608 So we tried to apply the OC parameter from 0 to 8388608 to decipher the message (other parameters were not been changed). In every step, we calculated FCS and compared it with contained in LLC-message FCS and had no success. Finally the question: Can the value of IOV_UI (Osmocom BS: 0, real BS: 0x10000000) affect the deciphering, and if yes then how?? With regards, students of the telecommunication department. Thank you for your attention! Best regards, Sergei ----------------------------------------------------------------------

1 0

OsmoDevCon / COVID-19
by Harald Welte 04 Sep '20

04 Sep '20

Dear fellow Osmocom developers, as you all know, we've sadly had to postpone OsmoDevCon 2020 back in April this year. At the time, we discussed to re-visit the situation in October 2020. While legally it is no problem at all to host an event with ~ 20 participants in Berlin/Germany (specific regulations really only start from 50+ participants) - I'm not entirely convinced it would be the smartest move. Legality and public health regulations are only one part of the equation - common sense and profound care for the key members of our community for sure are more relevant considerations to me. I'm not 100% in favour and not 100% against. Hence, I would like to get your input. Should we a) try to get an event organized on-site in Berlin? We'd have to move to a larger venue than IN-Berlin with proper ventilation and sufficient space so we can keep physical distance, but I think that's manageable for sysmocom as organizer. b) simply postpone to 2021? I'm convinced the situation will not change significantly (in a positive way) until late April 2021, so it's not really a "solution" as it will likely mean we have to think of late 2021 or 2022. c) plan some kind of online conference? To be honest, I think this model works fine for events where a single speaker wants to give lectures to hundreds or thousands of participants. But OsmoDevCon is much more interactive. We could record or live-stream some talks or screencasts from home, sure. But that only captures one part of the event. We could also try to set a date for a collaborative mumble, or the like - for the "hallway track". What are your thoughts? Let's avoid cross-posting the discussion to all of the mailing lists and simply have it on openbsc(a)lists.osmocom.org. 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)

1 0

[PATCH net-next 0/2] gtp: minor enhancements
by Nicolas Dichtel 31 Aug '20

31 Aug '20

The first patch removes a useless rcu lock and the second relax alloc constraints when a PDP context is added. drivers/net/gtp.c | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) Comments are welcomed, Nicolas

2 3

[PATCH net-next] gtp: add notification mechnism
by Nicolas Dichtel 27 Aug '20

27 Aug '20

Like all other network functions, let's notify gtp context on creation and deletion. Signed-off-by: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Tested-by: Gabriel Ganne <gabriel.ganne(a)6wind.com> --- drivers/net/gtp.c | 58 +++++++++++++++++++++++++++++++++------- include/uapi/linux/gtp.h | 2 ++ 2 files changed, 51 insertions(+), 9 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 8e47d0112e5d..360c1dc9381e 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -928,8 +928,8 @@ static void ipv4_pdp_fill(struct pdp_ctx *pctx, struct genl_info *info) } } -static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, - struct genl_info *info) +static struct pdp_ctx *gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, + struct genl_info *info) { struct pdp_ctx *pctx, *pctx_tid = NULL; struct net_device *dev = gtp->dev; @@ -956,12 +956,12 @@ static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, if (found) { if (info->nlhdr->nlmsg_flags & NLM_F_EXCL) - return -EEXIST; + return ERR_PTR(-EEXIST); if (info->nlhdr->nlmsg_flags & NLM_F_REPLACE) - return -EOPNOTSUPP; + return ERR_PTR(-EOPNOTSUPP); if (pctx && pctx_tid) - return -EEXIST; + return ERR_PTR(-EEXIST); if (!pctx) pctx = pctx_tid; @@ -974,13 +974,13 @@ static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, netdev_dbg(dev, "GTPv1-U: update tunnel id = %x/%x (pdp %p)\n", pctx->u.v1.i_tei, pctx->u.v1.o_tei, pctx); - return 0; + return pctx; } pctx = kmalloc(sizeof(*pctx), GFP_ATOMIC); if (pctx == NULL) - return -ENOMEM; + return ERR_PTR(-ENOMEM); sock_hold(sk); pctx->sk = sk; @@ -1018,7 +1018,7 @@ static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, break; } - return 0; + return pctx; } static void pdp_context_free(struct rcu_head *head) @@ -1036,9 +1036,12 @@ static void pdp_context_delete(struct pdp_ctx *pctx) call_rcu(&pctx->rcu_head, pdp_context_free); } +static int gtp_tunnel_notify(struct pdp_ctx *pctx, u8 cmd); + static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) { unsigned int version; + struct pdp_ctx *pctx; struct gtp_dev *gtp; struct sock *sk; int err; @@ -1088,7 +1091,13 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) goto out_unlock; } - err = gtp_pdp_add(gtp, sk, info); + pctx = gtp_pdp_add(gtp, sk, info); + if (IS_ERR(pctx)) { + err = PTR_ERR(pctx); + } else { + gtp_tunnel_notify(pctx, GTP_CMD_NEWPDP); + err = 0; + } out_unlock: rcu_read_unlock(); @@ -1159,6 +1168,7 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) netdev_dbg(pctx->dev, "GTPv1-U: deleting tunnel id = %x/%x (pdp %p)\n", pctx->u.v1.i_tei, pctx->u.v1.o_tei, pctx); + gtp_tunnel_notify(pctx, GTP_CMD_DELPDP); pdp_context_delete(pctx); out_unlock: @@ -1168,6 +1178,14 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; +enum gtp_multicast_groups { + GTP_GENL_MCGRP, +}; + +static const struct genl_multicast_group gtp_genl_mcgrps[] = { + [GTP_GENL_MCGRP] = { .name = GTP_GENL_MCGRP_NAME }, +}; + static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, int flags, u32 type, struct pdp_ctx *pctx) { @@ -1205,6 +1223,26 @@ static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, return -EMSGSIZE; } +static int gtp_tunnel_notify(struct pdp_ctx *pctx, u8 cmd) +{ + struct sk_buff *msg; + int ret; + + msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_ATOMIC); + if (!msg) + return -ENOMEM; + + ret = gtp_genl_fill_info(msg, 0, 0, 0, cmd, pctx); + if (ret < 0) { + nlmsg_free(msg); + return ret; + } + + ret = genlmsg_multicast_netns(&gtp_genl_family, dev_net(pctx->dev), msg, + 0, GTP_GENL_MCGRP, GFP_ATOMIC); + return ret; +} + static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) { struct pdp_ctx *pctx = NULL; @@ -1335,6 +1373,8 @@ static struct genl_family gtp_genl_family __ro_after_init = { .module = THIS_MODULE, .ops = gtp_genl_ops, .n_ops = ARRAY_SIZE(gtp_genl_ops), + .mcgrps = gtp_genl_mcgrps, + .n_mcgrps = ARRAY_SIZE(gtp_genl_mcgrps), }; static int __net_init gtp_net_init(struct net *net) diff --git a/include/uapi/linux/gtp.h b/include/uapi/linux/gtp.h index c7d66755d212..79f9191bbb24 100644 --- a/include/uapi/linux/gtp.h +++ b/include/uapi/linux/gtp.h @@ -2,6 +2,8 @@ #ifndef _UAPI_LINUX_GTP_H_ #define _UAPI_LINUX_GTP_H_ +#define GTP_GENL_MCGRP_NAME "gtp" + enum gtp_genl_cmds { GTP_CMD_NEWPDP, GTP_CMD_DELPDP, -- 2.26.2

3 12

[PATCH net] gtp: add GTPA_LINK info to msg sent to userspace
by Nicolas Dichtel 25 Aug '20

25 Aug '20

During a dump, this attribute is essential, it enables the userspace to know on which interface the context is linked to. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Nicolas Dichtel <nicolas.dichtel(a)6wind.com> Tested-by: Gabriel Ganne <gabriel.ganne(a)6wind.com> --- I target this to net, because I think this is a bug fix. The dump result cannot be used if there is more than one gtp interface on the system. drivers/net/gtp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 21640a035d7d..8e47d0112e5d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1179,6 +1179,7 @@ static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, goto nlmsg_failure; if (nla_put_u32(skb, GTPA_VERSION, pctx->gtp_version) || + nla_put_u32(skb, GTPA_LINK, pctx->dev->ifindex) || nla_put_be32(skb, GTPA_PEER_ADDRESS, pctx->peer_addr_ip4.s_addr) || nla_put_be32(skb, GTPA_MS_ADDRESS, pctx->ms_addr_ip4.s_addr)) goto nla_put_failure; -- 2.26.2

2 1

GTP tunnel seems to exist, but No encapsulation when using libgtpnl
by Papa Tana 23 Aug '20

23 Aug '20

Hi All, This is my first post, I have a similar problem to the topic "Network is unreachable error for GTP interface". I followed all instructions, installed all needed dependencies, upgrade my kernel version to 4.9.0-6 as stated in https://osmocom.org/projects/openggsn/wiki/Kernel_GTP Unfortunately, no GTP T-PDU encapsulation for my packets. ## Tunnel listing is OK root@routeurA:/home/bob/libgtpnl/tools# ./gtp-tunnel list version 1 tei 200/100 ms_addr 172.23.10.163 sgsn_addr 10.11.12.14 ## I have upgraded my Kernel version to 4.9.0-6 as stated in https://osmocom.org/projects/openggsn/wiki/Kernel_GTP At the time of writing (2018-04-26) of this wiki, below listed distributions have support of GTP kernels : Debian Debian 9 "stretch" (kernel 4.9.0-6) root@routeurA:/home/bob/libgtpnl/tools# uname -r 4.9.0-6-amd64 ## GTP module root@routeurA:/home/bob/libgtpnl/tools# lsmod | grep gtp gtp 28672 0 udp_tunnel 16384 1 gtp ## ping remote ms_addr is not ok root@routeurA:/home/bob/libgtpnl/tools# ping 172.23.10.163 PING 172.23.10.163 (172.23.10.163) 56(84) bytes of data. ^C --- 172.23.10.163 ping statistics --- 1 packets transmitted, 0 received, 100% packet loss, time 0ms ## remove the route using "gtpa" device root@routeurA:/home/bob/libgtpnl/tools# ip route del 172.23.10.163/32 dev gtpa ## add new route using normal interface root@routeurA:/home/bob/libgtpnl/tools# ip route add 172.23.10.163/32 via 10.11.12.14 ## ping is OK root@routeurA:/home/bob/libgtpnl/tools# ping 172.23.10.163 PING 172.23.10.163 (172.23.10.163) 56(84) bytes of data. 64 bytes from 172.23.10.163: icmp_seq=1 ttl=64 time=0.592 ms 64 bytes from 172.23.10.163: icmp_seq=2 ttl=64 time=0.713 ms ## remove again the route root@routeurA:/home/bob/libgtpnl/tools# ip route del 172.23.10.163/32 ## switch it to "gtpa" device root@routeurA:/home/bob/libgtpnl/tools# ip route add 172.23.10.163/32 dev gtpa root@routeurA:/home/bob/libgtpnl/tools# ping 172.23.10.163 PING 172.23.10.163 (172.23.10.163) 56(84) bytes of data. ^C ## tcpdump shows ICMP between the 2 ms_addr, no encapsulation at all Am I missing something somewhere? FYI, I'm not using openggsn or ergw, I have developped my small userspace GTP-C ready, but I'm stuck at GTP-U side. Thanks in advance, Best Regards,

3 11

Getting "Operation Not Permitted" When Running gtp-link
by Donal Crowley 21 Jul '20

21 Jul '20

Hi all, I am running an opensource 5G Core however I cannot get my UPF up and running ads the GTP links and tunnels I am trying to create are failing with “Operation Not Permitted”. I’ll recompile again to run through GDB but would really like to hear back if you guys have seen this before on Ubuntu 18.04 (Kernel > 5) Any help would be appreciated Donal Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10

1 0

Osmocom scheduled outage on *2020-07-19 9am CEST*
by Harald Welte 19 Jul '20

19 Jul '20

There is an urgent need to migrate our most important public infrastructure to a new server, and I will be doing that on *Sunday, July 19 2020*, starting about 9am CEST. The migration involves redmine (main osmocom.org website), jenkins, gerrit, git, and cgit. In theory, the migration should be quick. I would expect (significantly) less than one hour of downtime. However, we all know Murphys law. Services not affected are mail (including mailman lists), ftp, dns. So in case of doubt, we can still use mailing lists to communicate. In case anyone urgently needs osmocom source code on Sunday morning during the downtime: There are public mirrors available on github. 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)

1 1

Network is unreachable error for GTP interface
by avatli＠itu.edu.tr 18 Jul '20

18 Jul '20

Hi I'm trying to build a setup by using GTP module and libgtpnl tools on Centos 7 but I haven't been successful yet. Worse, I don't know how to debug the problem. I also stopped firewall and iptables. How can I debug/solve, I will be very glad if you help. dmesg or system messages show show nothing. Why is GTP interface (link) is unreachable. Thanks in advance - Volkan $ modinfo gtp filename: /lib/modules/5.7.7-1.el7.elrepo.x86_64/kernel/drivers/net/gtp.ko alias: net-pf-16-proto-16-family-gtp alias: rtnl-link-gtp description: Interface driver for GTP encapsulated traffic author: Harald Welte <hwelte(a)sysmocom.de> license: GPL srcversion: 191407DA5399304D93D62C7 depends: udp_tunnel retpoline: Y intree: Y name: gtp vermagic: 5.7.7-1.el7.elrepo.x86_64 SMP mod_unload modversions $ modinfo udp_tunnel filename: /lib/modules/5.7.7-1.el7.elrepo.x86_64/kernel/net/ipv4/udp_tunnel.ko license: GPL srcversion: 0A315BA6124B0664F4D23FB depends: retpoline: Y intree: Y name: udp_tunnel vermagic: 5.7.7-1.el7.elrepo.x86_64 SMP mod_unload modversions $ ip addr add 172.0.0.1/24 dev enp9s0 $ ip addr add 172.99.0.1/32 dev lo $ ./gtp-link add gtp1 WARNING: attaching dummy socket descriptors. Keep this process running for testing purposes. $ ./gtp-tunnel add gtp1 v1 200 100 172.99.0.2 172.0.0.2 $ ip route add 172.99.0.2/32 dev gtp1 $ ./gtp-tunnel list version 1 tei 200/100 ms_addr 172.99.0.2 sgsn_addr 172.0.0.2 $ ip a 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet 172.99.0.1/32 scope global lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 7: enp9s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000 link/ether 08:35:71:ab:54:5f brd ff:ff:ff:ff:ff:ff inet 172.0.0.1/24 scope global enp9s0 valid_lft forever preferred_lft forever 8: gtp1: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 0 qdisc noqueue state UNKNOWN group default qlen 1000 link/none $ ip route default via 192.168.1.1 dev enp2s0 proto static metric 100 172.0.0.0/24 dev enp9s0 proto kernel scope link src 172.0.0.1 172.99.0.2 dev gtp1 scope link $ ping 172.99.0.2 PING 172.99.0.2 (172.99.0.2) 56(84) bytes of data. ping: sendmsg: Network is unreachable ping: sendmsg: Network is unreachable ping: sendmsg: Network is unreachable

3 3

Basic testing for Kernel GTP-U is not up to date
by avatli＠itu.edu.tr 17 Jul '20

17 Jul '20

Hello I tried to follow the steps in the basic testing below but it is not up-to-date https://osmocom.org/projects/linux-kernel-gtp-u/wiki/Basic_Testing To make the setup work, I execute the GGSN on the host by using osmo-ggsn/doc/examples/osmo-ggsn.cfg. I changed the "gtp bind-ip" to 172.31.1.1, "ip prefix dynamic" to 192.168.71.0/24, "ip ifconfig" to 192.168.71.0/24 and then execute the emulated SGSN inside the sgsn namespace (ip netns exec sgsn sgsnemu -d -r 172.31.1.1 -l 172.31.1.2 --defaultroute --createif). In this case; GGSN can create PDP Context successful but there is not any GTP tunnel. How can I get the up-to-date version of this document? or how can I run the basic testing steps? Thanks in advance. - Volkan

2 2

[PATCH AUTOSEL 5.6 121/606] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Sasha Levin 08 Jun '20

08 Jun '20

From: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> [ Upstream commit 846c68f7f1ac82c797a2f1db3344a2966c0fe2e1 ] In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 672cd2caf2fb..21640a035d7d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1169,11 +1169,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1227,8 +1227,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1271,6 +1271,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.25.1

1 0

Compilation of TTCN3 tests for SGSN failed
by Mirko Kovacevic 21 May '20

21 May '20

*Hi, Osmocom-SGSN* I have downloaded latest master branch sources for TTCN3 tests. I followed instructions for compilations and dependencies. In file : *regen_makefile.sh * If line with ../regen-makefile.sh SGSN_Tests.ttcn $FILES is replaced with ../regen-makefile.sh $FILES then this warning is not present File `SGSN_Tests.ttcn' was given more than once for the Makefile. Also, after command *make compile * *(same with * *make sgsn * *from upper dir.)* I got from output: .... GSM_RR_Types.ttcn:404.2-408.2: In type definition `SecondPartAssign': GSM_RR_Types.ttcn:409.18-20: error: in variant attribute, at or before token `CSN': syntax error, unexpected XIdentifier, expecting $end GSM_RR_Types.ttcn:510.2-512.2: In type definition `IaRestOctLL': GSM_RR_Types.ttcn:513.42-44: error: in variant attribute, at or before token `CSN': syntax error, unexpected XIdentifier, expecting $end GSM_RR_Types.ttcn:588.2-594.2: In type definition `IaRestOctets': GSM_RR_Types.ttcn:595.23-25: error: in variant attribute, at or before token `CSN': syntax error, unexpected XIdentifier, expecting $end ........ Notify: Errors found in the input modules. Code will not be generated. make: *** [Makefile:206: compile] Error 1 Last time in April, I played with TTCN3 for SGSN, everything was good. Can you help me with this ?

2 4

[PATCH AUTOSEL 4.9 20/27] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Sasha Levin 14 May '20

14 May '20

From: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> [ Upstream commit 846c68f7f1ac82c797a2f1db3344a2966c0fe2e1 ] In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index a9e8a7356c41d..fe844888e0ed8 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1108,11 +1108,11 @@ static struct genl_family gtp_genl_family = { }; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1208,8 +1208,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1252,6 +1252,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.20.1

1 0

[PATCH AUTOSEL 4.14 30/39] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Sasha Levin 14 May '20

14 May '20

From: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> [ Upstream commit 846c68f7f1ac82c797a2f1db3344a2966c0fe2e1 ] In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 92e4e5d53053f..090607e725a24 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1177,11 +1177,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1235,8 +1235,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1279,6 +1279,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.20.1

1 0

[PATCH AUTOSEL 4.19 21/31] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Sasha Levin 14 May '20

14 May '20

From: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> [ Upstream commit 846c68f7f1ac82c797a2f1db3344a2966c0fe2e1 ] In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index eab9984f73a88..d73850ebb671f 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1177,11 +1177,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1235,8 +1235,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1279,6 +1279,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.20.1

1 0

[PATCH AUTOSEL 5.4 28/49] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Sasha Levin 14 May '20

14 May '20

From: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> [ Upstream commit 846c68f7f1ac82c797a2f1db3344a2966c0fe2e1 ] In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 3a53d222bfcc1..d89ec99abcd63 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1172,11 +1172,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1230,8 +1230,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1274,6 +1274,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.20.1

1 0

[PATCH AUTOSEL 5.6 31/62] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Sasha Levin 14 May '20

14 May '20

From: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> [ Upstream commit 846c68f7f1ac82c797a2f1db3344a2966c0fe2e1 ] In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 672cd2caf2fbe..21640a035d7df 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1169,11 +1169,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1227,8 +1227,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1271,6 +1271,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.20.1

1 0

[PATCH net] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Yoshiyuki Kurauchi 01 May '20

01 May '20

In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 672cd2caf2fb..21640a035d7d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1169,11 +1169,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1227,8 +1227,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1271,6 +1271,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.17.1

2 1

[PATCH] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Yoshiyuki Kurauchi 29 Apr '20

29 Apr '20

In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> --- drivers/net/gtp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 672cd2caf2fb..21640a035d7d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1169,11 +1169,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1227,8 +1227,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + 0, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1271,6 +1271,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.17.1

2 1

[PATCH] gtp: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Yoshiyuki Kurauchi 27 Apr '20

27 Apr '20

In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> --- drivers/net/gtp.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 672cd2caf2fb..10fa731bae6d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1097,7 +1097,7 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) } static struct pdp_ctx *gtp_find_pdp_by_link(struct net *net, - struct nlattr *nla[]) + struct nlattr *nla[]) { struct gtp_dev *gtp; @@ -1169,11 +1169,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1227,8 +1227,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + info->nlhdr->nlmsg_flags, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1271,6 +1271,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.17.1

2 3

TTCN3 3G/2G Authentication using SIM card/SIM card reader
by Mirko Kovacevic 02 Apr '20

02 Apr '20

HI, Osmo Packet Core Guys, I am working from home nowadays, testing commercial packet core equipment (Ericsson). I have already listen/read about using TTCN3 in all your famous products. But never get a try. I am interesting to deploy similar tests for production systems like yours for OsmoSGSN, due to lack of test radio equipment from home and also to make some sort of automation. I ran your TTCN3 tests for OsmoSGSN and get loved on a first sight. You did amazing job and made countless possibilities. Thank for that. I know you have C code support in libosmocore library for 3g/2g auth. *Can we use SIM card reader in TTCN3 as an option to evaluate authentication ?!* *Did you ever tried?!* Best regards, Mirko K.

2 1

[PATCH] net: set NLM_F_MULTI flag in gtp_genl_dump_pdp()
by Yoshiyuki Kurauchi 24 Feb '20

24 Feb '20

In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers. Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> --- drivers/net/gtp.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 672cd2caf2fb..10fa731bae6d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1097,7 +1097,7 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) } static struct pdp_ctx *gtp_find_pdp_by_link(struct net *net, - struct nlattr *nla[]) + struct nlattr *nla[]) { struct gtp_dev *gtp; @@ -1169,11 +1169,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1227,8 +1227,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + info->nlhdr->nlmsg_flags, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1271,6 +1271,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.17.1

1 0

[PATCH] net: set NLM_F_MULTI flag in gtp_genl_dump_pdp() In drivers/net/gtp.c, gtp_genl_dump_pdp() should set NLM_F_MULTI flag since it returns multipart message. This patch adds a new arg "flags" in gtp_genl_fill_info() so that flags can be set by the callers.
by Yoshiyuki Kurauchi 24 Feb '20

24 Feb '20

Signed-off-by: Yoshiyuki Kurauchi <ahochauwaaaaa(a)gmail.com> --- drivers/net/gtp.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 672cd2caf2fb..10fa731bae6d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1097,7 +1097,7 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) } static struct pdp_ctx *gtp_find_pdp_by_link(struct net *net, - struct nlattr *nla[]) + struct nlattr *nla[]) { struct gtp_dev *gtp; @@ -1169,11 +1169,11 @@ static int gtp_genl_del_pdp(struct sk_buff *skb, struct genl_info *info) static struct genl_family gtp_genl_family; static int gtp_genl_fill_info(struct sk_buff *skb, u32 snd_portid, u32 snd_seq, - u32 type, struct pdp_ctx *pctx) + int flags, u32 type, struct pdp_ctx *pctx) { void *genlh; - genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, 0, + genlh = genlmsg_put(skb, snd_portid, snd_seq, &gtp_genl_family, flags, type); if (genlh == NULL) goto nlmsg_failure; @@ -1227,8 +1227,8 @@ static int gtp_genl_get_pdp(struct sk_buff *skb, struct genl_info *info) goto err_unlock; } - err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, - info->snd_seq, info->nlhdr->nlmsg_type, pctx); + err = gtp_genl_fill_info(skb2, NETLINK_CB(skb).portid, info->snd_seq, + info->nlhdr->nlmsg_flags, info->nlhdr->nlmsg_type, pctx); if (err < 0) goto err_unlock_free; @@ -1271,6 +1271,7 @@ static int gtp_genl_dump_pdp(struct sk_buff *skb, gtp_genl_fill_info(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + NLM_F_MULTI, cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; cb->args[1] = j; -- 2.17.1

1 0

[PATCH net] gtp: use __GFP_NOWARN to avoid memalloc warning
by Taehee Yoo 04 Feb '20

04 Feb '20

gtp hashtable size is received by user-space. So, this hashtable size could be too large. If so, kmalloc will internally print a warning message. This warning message is actually not necessary for the gtp module. So, this patch adds __GFP_NOWARN to avoid this message. Splat looks like: [ 2171.200049][ T1860] WARNING: CPU: 1 PID: 1860 at mm/page_alloc.c:4713 __alloc_pages_nodemask+0x2f3/0x740 [ 2171.238885][ T1860] Modules linked in: gtp veth openvswitch nsh nf_conncount nf_nat nf_conntrack nf_defrag_ipv] [ 2171.262680][ T1860] CPU: 1 PID: 1860 Comm: gtp-link Not tainted 5.5.0+ #321 [ 2171.263567][ T1860] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006 [ 2171.264681][ T1860] RIP: 0010:__alloc_pages_nodemask+0x2f3/0x740 [ 2171.265332][ T1860] Code: 64 fe ff ff 65 48 8b 04 25 c0 0f 02 00 48 05 f0 12 00 00 41 be 01 00 00 00 49 89 47 0 [ 2171.267301][ T1860] RSP: 0018:ffff8880b51af1f0 EFLAGS: 00010246 [ 2171.268320][ T1860] RAX: ffffed1016a35e43 RBX: 0000000000000000 RCX: 0000000000000000 [ 2171.269517][ T1860] RDX: 0000000000000000 RSI: 000000000000000b RDI: 0000000000000000 [ 2171.270305][ T1860] RBP: 0000000000040cc0 R08: ffffed1018893109 R09: dffffc0000000000 [ 2171.275973][ T1860] R10: 0000000000000001 R11: ffffed1018893108 R12: 1ffff11016a35e43 [ 2171.291039][ T1860] R13: 000000000000000b R14: 000000000000000b R15: 00000000000f4240 [ 2171.292328][ T1860] FS: 00007f53cbc83740(0000) GS:ffff8880da000000(0000) knlGS:0000000000000000 [ 2171.293409][ T1860] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 2171.294586][ T1860] CR2: 000055f540014508 CR3: 00000000b49f2004 CR4: 00000000000606e0 [ 2171.295424][ T1860] Call Trace: [ 2171.295756][ T1860] ? mark_held_locks+0xa5/0xe0 [ 2171.296659][ T1860] ? __alloc_pages_slowpath+0x21b0/0x21b0 [ 2171.298283][ T1860] ? gtp_encap_enable_socket+0x13e/0x400 [gtp] [ 2171.298962][ T1860] ? alloc_pages_current+0xc1/0x1a0 [ 2171.299475][ T1860] kmalloc_order+0x22/0x80 [ 2171.299936][ T1860] kmalloc_order_trace+0x1d/0x140 [ 2171.300437][ T1860] __kmalloc+0x302/0x3a0 [ 2171.300896][ T1860] gtp_newlink+0x293/0xba0 [gtp] [ ... ] Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> --- drivers/net/gtp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 7032a2405e1a..af07ea760b35 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -767,12 +767,12 @@ 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_KERNEL | __GFP_NOWARN); if (gtp->addr_hash == NULL) return -ENOMEM; gtp->tid_hash = kmalloc_array(hsize, sizeof(struct hlist_head), - GFP_KERNEL); + GFP_KERNEL | __GFP_NOWARN); if (gtp->tid_hash == NULL) goto err1; -- 2.17.1

2 1

Re: [PATCH] gtp: simplify error handling code in 'gtp_encap_enable()'
by Simon Horman 06 Jan '20

06 Jan '20

On Sun, Jan 05, 2020 at 06:36:07PM +0100, Christophe JAILLET wrote: > 'gtp_encap_disable_sock(sk)' handles the case where sk is NULL, so there > is no need to test it before calling the function. > > This saves a few line of code. > > Signed-off-by: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> Reviewed-by: Simon Horman <simon.horman(a)netronome.com>

2 1

[PATCH AUTOSEL 4.9 23/38] gtp: avoid zero size hashtable
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 6a902c0f31993ab02e1b6ea7085002b9c9083b6a ] GTP default hashtable size is 1024 and userspace could set specific hashtable size with IFLA_GTP_PDP_HASHSIZE. If hashtable size is set to 0 from userspace, hashtable will not work and panic will occur. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 1d2fc81dc714..c2898718b593 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -677,10 +677,13 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, if (err < 0) goto out_err; - if (!data[IFLA_GTP_PDP_HASHSIZE]) + if (!data[IFLA_GTP_PDP_HASHSIZE]) { hashsize = 1024; - else + } else { hashsize = nla_get_u32(data[IFLA_GTP_PDP_HASHSIZE]); + if (!hashsize) + hashsize = 1024; + } err = gtp_hashtable_new(gtp, hashsize); if (err < 0) -- 2.20.1

1 0

[PATCH AUTOSEL 4.9 22/38] gtp: fix wrong condition in gtp_genl_dump_pdp()
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 94a6d9fb88df43f92d943c32b84ce398d50bf49f ] gtp_genl_dump_pdp() is ->dumpit() callback of GTP module and it is used to dump pdp contexts. it would be re-executed because of dump packet size. If dump packet size is too big, it saves current dump pointer (gtp interface pointer, bucket, TID value) then it restarts dump from last pointer. Current GTP code allows adding zero TID pdp context but dump code ignores zero TID value. So, last dump pointer will not be found. In addition, this patch adds missing rcu_read_lock() in gtp_genl_dump_pdp(). Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 36 +++++++++++++++++++----------------- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 7e1df403a37d..1d2fc81dc714 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -42,7 +42,6 @@ struct pdp_ctx { struct hlist_node hlist_addr; union { - u64 tid; struct { u64 tid; u16 flow; @@ -1221,43 +1220,46 @@ static int gtp_genl_dump_pdp(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_net *gn = net_generic(net, gtp_net_id); - unsigned long tid = cb->args[1]; - int i, k = cb->args[0], ret; struct pdp_ctx *pctx; + 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 = k; i < gtp->hash_size; i++) { - hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], hlist_tid) { - if (tid && tid != pctx->u.tid) - continue; - else - tid = 0; - - ret = gtp_genl_fill_info(skb, - NETLINK_CB(cb->skb).portid, - cb->nlh->nlmsg_seq, - cb->nlh->nlmsg_type, pctx); - if (ret < 0) { + for (i = bucket; i < gtp->hash_size; i++) { + j = 0; + hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], + hlist_tid) { + if (j >= skip && + gtp_genl_fill_info(skb, + NETLINK_CB(cb->skb).portid, + cb->nlh->nlmsg_seq, + cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; - cb->args[1] = pctx->u.tid; + 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; } -- 2.20.1

1 0

[PATCH AUTOSEL 4.14 36/57] gtp: avoid zero size hashtable
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 6a902c0f31993ab02e1b6ea7085002b9c9083b6a ] GTP default hashtable size is 1024 and userspace could set specific hashtable size with IFLA_GTP_PDP_HASHSIZE. If hashtable size is set to 0 from userspace, hashtable will not work and panic will occur. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 52628a9986b1..b2f5f7f3d6b4 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -664,10 +664,13 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, if (err < 0) return err; - if (!data[IFLA_GTP_PDP_HASHSIZE]) + if (!data[IFLA_GTP_PDP_HASHSIZE]) { hashsize = 1024; - else + } else { hashsize = nla_get_u32(data[IFLA_GTP_PDP_HASHSIZE]); + if (!hashsize) + hashsize = 1024; + } err = gtp_hashtable_new(gtp, hashsize); if (err < 0) -- 2.20.1

1 0

[PATCH AUTOSEL 4.14 35/57] gtp: fix wrong condition in gtp_genl_dump_pdp()
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 94a6d9fb88df43f92d943c32b84ce398d50bf49f ] gtp_genl_dump_pdp() is ->dumpit() callback of GTP module and it is used to dump pdp contexts. it would be re-executed because of dump packet size. If dump packet size is too big, it saves current dump pointer (gtp interface pointer, bucket, TID value) then it restarts dump from last pointer. Current GTP code allows adding zero TID pdp context but dump code ignores zero TID value. So, last dump pointer will not be found. In addition, this patch adds missing rcu_read_lock() in gtp_genl_dump_pdp(). Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 36 +++++++++++++++++++----------------- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 0522cbdce44a..52628a9986b1 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -42,7 +42,6 @@ struct pdp_ctx { struct hlist_node hlist_addr; union { - u64 tid; struct { u64 tid; u16 flow; @@ -1247,43 +1246,46 @@ static int gtp_genl_dump_pdp(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_net *gn = net_generic(net, gtp_net_id); - unsigned long tid = cb->args[1]; - int i, k = cb->args[0], ret; struct pdp_ctx *pctx; + 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 = k; i < gtp->hash_size; i++) { - hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], hlist_tid) { - if (tid && tid != pctx->u.tid) - continue; - else - tid = 0; - - ret = gtp_genl_fill_info(skb, - NETLINK_CB(cb->skb).portid, - cb->nlh->nlmsg_seq, - cb->nlh->nlmsg_type, pctx); - if (ret < 0) { + for (i = bucket; i < gtp->hash_size; i++) { + j = 0; + hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], + hlist_tid) { + if (j >= skip && + gtp_genl_fill_info(skb, + NETLINK_CB(cb->skb).portid, + cb->nlh->nlmsg_seq, + cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; - cb->args[1] = pctx->u.tid; + 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; } -- 2.20.1

1 0

[PATCH AUTOSEL 4.14 34/57] gtp: do not allow adding duplicate tid and ms_addr pdp context
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 6b01b1d9b2d38dc84ac398bfe9f00baff06a31e5 ] GTP RX packet path lookups pdp context with TID. If duplicate TID pdp contexts are existing in the list, it couldn't select correct pdp context. So, TID value should be unique. GTP TX packet path lookups pdp context with ms_addr. If duplicate ms_addr pdp contexts are existing in the list, it couldn't select correct pdp context. So, ms_addr value should be unique. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 32 ++++++++++++++++++++++---------- 1 file changed, 22 insertions(+), 10 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 5de4053774b8..0522cbdce44a 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -929,24 +929,31 @@ static void ipv4_pdp_fill(struct pdp_ctx *pctx, struct genl_info *info) } } -static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, - struct genl_info *info) +static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, + struct genl_info *info) { + struct pdp_ctx *pctx, *pctx_tid = NULL; struct net_device *dev = gtp->dev; u32 hash_ms, hash_tid = 0; - struct pdp_ctx *pctx; + unsigned int version; bool found = false; __be32 ms_addr; ms_addr = nla_get_be32(info->attrs[GTPA_MS_ADDRESS]); hash_ms = ipv4_hashfn(ms_addr) % gtp->hash_size; + version = nla_get_u32(info->attrs[GTPA_VERSION]); - hlist_for_each_entry_rcu(pctx, &gtp->addr_hash[hash_ms], hlist_addr) { - if (pctx->ms_addr_ip4.s_addr == ms_addr) { - found = true; - break; - } - } + pctx = ipv4_pdp_find(gtp, ms_addr); + if (pctx) + found = true; + if (version == GTP_V0) + pctx_tid = gtp0_pdp_find(gtp, + nla_get_u64(info->attrs[GTPA_TID])); + else if (version == GTP_V1) + pctx_tid = gtp1_pdp_find(gtp, + nla_get_u32(info->attrs[GTPA_I_TEI])); + if (pctx_tid) + found = true; if (found) { if (info->nlhdr->nlmsg_flags & NLM_F_EXCL) @@ -954,6 +961,11 @@ static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, if (info->nlhdr->nlmsg_flags & NLM_F_REPLACE) return -EOPNOTSUPP; + if (pctx && pctx_tid) + return -EEXIST; + if (!pctx) + pctx = pctx_tid; + ipv4_pdp_fill(pctx, info); if (pctx->gtp_version == GTP_V0) @@ -1077,7 +1089,7 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) goto out_unlock; } - err = ipv4_pdp_add(gtp, sk, info); + err = gtp_pdp_add(gtp, sk, info); out_unlock: rcu_read_unlock(); -- 2.20.1

1 0

[PATCH AUTOSEL 4.19 54/84] gtp: avoid zero size hashtable
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 6a902c0f31993ab02e1b6ea7085002b9c9083b6a ] GTP default hashtable size is 1024 and userspace could set specific hashtable size with IFLA_GTP_PDP_HASHSIZE. If hashtable size is set to 0 from userspace, hashtable will not work and panic will occur. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index d64ebf0f9677..eeae8617f7ec 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -664,10 +664,13 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, if (err < 0) return err; - if (!data[IFLA_GTP_PDP_HASHSIZE]) + if (!data[IFLA_GTP_PDP_HASHSIZE]) { hashsize = 1024; - else + } else { hashsize = nla_get_u32(data[IFLA_GTP_PDP_HASHSIZE]); + if (!hashsize) + hashsize = 1024; + } err = gtp_hashtable_new(gtp, hashsize); if (err < 0) -- 2.20.1

1 0

[PATCH AUTOSEL 4.19 53/84] gtp: fix wrong condition in gtp_genl_dump_pdp()
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 94a6d9fb88df43f92d943c32b84ce398d50bf49f ] gtp_genl_dump_pdp() is ->dumpit() callback of GTP module and it is used to dump pdp contexts. it would be re-executed because of dump packet size. If dump packet size is too big, it saves current dump pointer (gtp interface pointer, bucket, TID value) then it restarts dump from last pointer. Current GTP code allows adding zero TID pdp context but dump code ignores zero TID value. So, last dump pointer will not be found. In addition, this patch adds missing rcu_read_lock() in gtp_genl_dump_pdp(). Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 36 +++++++++++++++++++----------------- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 7ca96ba00f20..d64ebf0f9677 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -42,7 +42,6 @@ struct pdp_ctx { struct hlist_node hlist_addr; union { - u64 tid; struct { u64 tid; u16 flow; @@ -1249,43 +1248,46 @@ static int gtp_genl_dump_pdp(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_net *gn = net_generic(net, gtp_net_id); - unsigned long tid = cb->args[1]; - int i, k = cb->args[0], ret; struct pdp_ctx *pctx; + 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 = k; i < gtp->hash_size; i++) { - hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], hlist_tid) { - if (tid && tid != pctx->u.tid) - continue; - else - tid = 0; - - ret = gtp_genl_fill_info(skb, - NETLINK_CB(cb->skb).portid, - cb->nlh->nlmsg_seq, - cb->nlh->nlmsg_type, pctx); - if (ret < 0) { + for (i = bucket; i < gtp->hash_size; i++) { + j = 0; + hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], + hlist_tid) { + if (j >= skip && + gtp_genl_fill_info(skb, + NETLINK_CB(cb->skb).portid, + cb->nlh->nlmsg_seq, + cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; - cb->args[1] = pctx->u.tid; + 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; } -- 2.20.1

1 0

[PATCH AUTOSEL 4.19 52/84] gtp: do not allow adding duplicate tid and ms_addr pdp context
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 6b01b1d9b2d38dc84ac398bfe9f00baff06a31e5 ] GTP RX packet path lookups pdp context with TID. If duplicate TID pdp contexts are existing in the list, it couldn't select correct pdp context. So, TID value should be unique. GTP TX packet path lookups pdp context with ms_addr. If duplicate ms_addr pdp contexts are existing in the list, it couldn't select correct pdp context. So, ms_addr value should be unique. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 32 ++++++++++++++++++++++---------- 1 file changed, 22 insertions(+), 10 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index d178d5bad7e4..7ca96ba00f20 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -931,24 +931,31 @@ static void ipv4_pdp_fill(struct pdp_ctx *pctx, struct genl_info *info) } } -static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, - struct genl_info *info) +static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, + struct genl_info *info) { + struct pdp_ctx *pctx, *pctx_tid = NULL; struct net_device *dev = gtp->dev; u32 hash_ms, hash_tid = 0; - struct pdp_ctx *pctx; + unsigned int version; bool found = false; __be32 ms_addr; ms_addr = nla_get_be32(info->attrs[GTPA_MS_ADDRESS]); hash_ms = ipv4_hashfn(ms_addr) % gtp->hash_size; + version = nla_get_u32(info->attrs[GTPA_VERSION]); - hlist_for_each_entry_rcu(pctx, &gtp->addr_hash[hash_ms], hlist_addr) { - if (pctx->ms_addr_ip4.s_addr == ms_addr) { - found = true; - break; - } - } + pctx = ipv4_pdp_find(gtp, ms_addr); + if (pctx) + found = true; + if (version == GTP_V0) + pctx_tid = gtp0_pdp_find(gtp, + nla_get_u64(info->attrs[GTPA_TID])); + else if (version == GTP_V1) + pctx_tid = gtp1_pdp_find(gtp, + nla_get_u32(info->attrs[GTPA_I_TEI])); + if (pctx_tid) + found = true; if (found) { if (info->nlhdr->nlmsg_flags & NLM_F_EXCL) @@ -956,6 +963,11 @@ static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, if (info->nlhdr->nlmsg_flags & NLM_F_REPLACE) return -EOPNOTSUPP; + if (pctx && pctx_tid) + return -EEXIST; + if (!pctx) + pctx = pctx_tid; + ipv4_pdp_fill(pctx, info); if (pctx->gtp_version == GTP_V0) @@ -1079,7 +1091,7 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) goto out_unlock; } - err = ipv4_pdp_add(gtp, sk, info); + err = gtp_pdp_add(gtp, sk, info); out_unlock: rcu_read_unlock(); -- 2.20.1

1 0

[PATCH AUTOSEL 5.4 115/187] gtp: avoid zero size hashtable
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 6a902c0f31993ab02e1b6ea7085002b9c9083b6a ] GTP default hashtable size is 1024 and userspace could set specific hashtable size with IFLA_GTP_PDP_HASHSIZE. If hashtable size is set to 0 from userspace, hashtable will not work and panic will occur. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 5450b1099c6d..e5b7d6d2286e 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -667,10 +667,13 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, if (err < 0) return err; - if (!data[IFLA_GTP_PDP_HASHSIZE]) + if (!data[IFLA_GTP_PDP_HASHSIZE]) { hashsize = 1024; - else + } else { hashsize = nla_get_u32(data[IFLA_GTP_PDP_HASHSIZE]); + if (!hashsize) + hashsize = 1024; + } err = gtp_hashtable_new(gtp, hashsize); if (err < 0) -- 2.20.1

1 0

[PATCH AUTOSEL 5.4 114/187] gtp: fix an use-after-free in ipv4_pdp_find()
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 94dc550a5062030569d4aa76e10e50c8fc001930 ] ipv4_pdp_find() is called in TX packet path of GTP. ipv4_pdp_find() internally uses gtp->tid_hash to lookup pdp context. In the current code, gtp->tid_hash and gtp->addr_hash are freed by ->dellink(), which is gtp_dellink(). But gtp_dellink() would be called while packets are processing. So, gtp_dellink() should not free gtp->tid_hash and gtp->addr_hash. Instead, dev->priv_destructor() would be used because this callback is called after all packet processing safely. Test commands: ip link add veth1 type veth peer name veth2 ip a a 172.0.0.1/24 dev veth1 ip link set veth1 up ip a a 172.99.0.1/32 dev lo gtp-link add gtp1 & gtp-tunnel add gtp1 v1 200 100 172.99.0.2 172.0.0.2 ip r a 172.99.0.2/32 dev gtp1 ip link set gtp1 mtu 1500 ip netns add ns2 ip link set veth2 netns ns2 ip netns exec ns2 ip a a 172.0.0.2/24 dev veth2 ip netns exec ns2 ip link set veth2 up ip netns exec ns2 ip a a 172.99.0.2/32 dev lo ip netns exec ns2 ip link set lo up ip netns exec ns2 gtp-link add gtp2 & ip netns exec ns2 gtp-tunnel add gtp2 v1 100 200 172.99.0.1 172.0.0.1 ip netns exec ns2 ip r a 172.99.0.1/32 dev gtp2 ip netns exec ns2 ip link set gtp2 mtu 1500 hping3 172.99.0.2 -2 --flood & ip link del gtp1 Splat looks like: [ 72.568081][ T1195] BUG: KASAN: use-after-free in ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.568916][ T1195] Read of size 8 at addr ffff8880b9a35d28 by task hping3/1195 [ 72.569631][ T1195] [ 72.569861][ T1195] CPU: 2 PID: 1195 Comm: hping3 Not tainted 5.5.0-rc1 #199 [ 72.570547][ T1195] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006 [ 72.571438][ T1195] Call Trace: [ 72.571764][ T1195] dump_stack+0x96/0xdb [ 72.572171][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.572761][ T1195] print_address_description.constprop.5+0x1be/0x360 [ 72.573400][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.573971][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.574544][ T1195] __kasan_report+0x12a/0x16f [ 72.575014][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.575593][ T1195] kasan_report+0xe/0x20 [ 72.576004][ T1195] ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.576577][ T1195] gtp_build_skb_ip4+0x199/0x1420 [gtp] [ ... ] [ 72.647671][ T1195] BUG: unable to handle page fault for address: ffff8880b9a35d28 [ 72.648512][ T1195] #PF: supervisor read access in kernel mode [ 72.649158][ T1195] #PF: error_code(0x0000) - not-present page [ 72.649849][ T1195] PGD a6c01067 P4D a6c01067 PUD 11fb07067 PMD 11f939067 PTE 800fffff465ca060 [ 72.652958][ T1195] Oops: 0000 [#1] SMP DEBUG_PAGEALLOC KASAN PTI [ 72.653834][ T1195] CPU: 2 PID: 1195 Comm: hping3 Tainted: G B 5.5.0-rc1 #199 [ 72.668062][ T1195] RIP: 0010:ipv4_pdp_find.isra.12+0x86/0x170 [gtp] [ ... ] [ 72.679168][ T1195] Call Trace: [ 72.679603][ T1195] gtp_build_skb_ip4+0x199/0x1420 [gtp] [ 72.681915][ T1195] ? ipv4_pdp_find.isra.12+0x170/0x170 [gtp] [ 72.682513][ T1195] ? lock_acquire+0x164/0x3b0 [ 72.682966][ T1195] ? gtp_dev_xmit+0x35e/0x890 [gtp] [ 72.683481][ T1195] gtp_dev_xmit+0x3c2/0x890 [gtp] [ ... ] Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 34 +++++++++++++++++----------------- 1 file changed, 17 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index a010e0a11c33..5450b1099c6d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -640,9 +640,16 @@ static void gtp_link_setup(struct net_device *dev) } static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize); -static void gtp_hashtable_free(struct gtp_dev *gtp); static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]); +static void gtp_destructor(struct net_device *dev) +{ + struct gtp_dev *gtp = netdev_priv(dev); + + kfree(gtp->addr_hash); + kfree(gtp->tid_hash); +} + static int gtp_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) @@ -677,13 +684,15 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, gn = net_generic(dev_net(dev), gtp_net_id); list_add_rcu(&gtp->list, &gn->gtp_dev_list); + dev->priv_destructor = gtp_destructor; netdev_dbg(dev, "registered new GTP interface\n"); return 0; out_hashtable: - gtp_hashtable_free(gtp); + kfree(gtp->addr_hash); + kfree(gtp->tid_hash); out_encap: gtp_encap_disable(gtp); return err; @@ -692,8 +701,13 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, static void gtp_dellink(struct net_device *dev, struct list_head *head) { struct gtp_dev *gtp = netdev_priv(dev); + struct pdp_ctx *pctx; + int 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); - gtp_hashtable_free(gtp); list_del_rcu(&gtp->list); unregister_netdevice_queue(dev, head); } @@ -771,20 +785,6 @@ static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize) return -ENOMEM; } -static void gtp_hashtable_free(struct gtp_dev *gtp) -{ - struct pdp_ctx *pctx; - int 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); - - synchronize_rcu(); - kfree(gtp->addr_hash); - kfree(gtp->tid_hash); -} - static struct sock *gtp_encap_enable_socket(int fd, int type, struct gtp_dev *gtp) { -- 2.20.1

1 0

[PATCH AUTOSEL 5.4 113/187] gtp: fix wrong condition in gtp_genl_dump_pdp()
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 94a6d9fb88df43f92d943c32b84ce398d50bf49f ] gtp_genl_dump_pdp() is ->dumpit() callback of GTP module and it is used to dump pdp contexts. it would be re-executed because of dump packet size. If dump packet size is too big, it saves current dump pointer (gtp interface pointer, bucket, TID value) then it restarts dump from last pointer. Current GTP code allows adding zero TID pdp context but dump code ignores zero TID value. So, last dump pointer will not be found. In addition, this patch adds missing rcu_read_lock() in gtp_genl_dump_pdp(). Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 36 +++++++++++++++++++----------------- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 8b742edf793d..a010e0a11c33 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -38,7 +38,6 @@ struct pdp_ctx { struct hlist_node hlist_addr; union { - u64 tid; struct { u64 tid; u16 flow; @@ -1244,43 +1243,46 @@ static int gtp_genl_dump_pdp(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_net *gn = net_generic(net, gtp_net_id); - unsigned long tid = cb->args[1]; - int i, k = cb->args[0], ret; struct pdp_ctx *pctx; + 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 = k; i < gtp->hash_size; i++) { - hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], hlist_tid) { - if (tid && tid != pctx->u.tid) - continue; - else - tid = 0; - - ret = gtp_genl_fill_info(skb, - NETLINK_CB(cb->skb).portid, - cb->nlh->nlmsg_seq, - cb->nlh->nlmsg_type, pctx); - if (ret < 0) { + for (i = bucket; i < gtp->hash_size; i++) { + j = 0; + hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], + hlist_tid) { + if (j >= skip && + gtp_genl_fill_info(skb, + NETLINK_CB(cb->skb).portid, + cb->nlh->nlmsg_seq, + cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; - cb->args[1] = pctx->u.tid; + 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; } -- 2.20.1

1 0

[PATCH AUTOSEL 5.4 112/187] gtp: do not allow adding duplicate tid and ms_addr pdp context
by Sasha Levin 27 Dec '19

27 Dec '19

From: Taehee Yoo <ap420073(a)gmail.com> [ Upstream commit 6b01b1d9b2d38dc84ac398bfe9f00baff06a31e5 ] GTP RX packet path lookups pdp context with TID. If duplicate TID pdp contexts are existing in the list, it couldn't select correct pdp context. So, TID value should be unique. GTP TX packet path lookups pdp context with ms_addr. If duplicate ms_addr pdp contexts are existing in the list, it couldn't select correct pdp context. So, ms_addr value should be unique. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> Signed-off-by: Jakub Kicinski <jakub.kicinski(a)netronome.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- drivers/net/gtp.c | 32 ++++++++++++++++++++++---------- 1 file changed, 22 insertions(+), 10 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index ecfe26215935..8b742edf793d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -926,24 +926,31 @@ static void ipv4_pdp_fill(struct pdp_ctx *pctx, struct genl_info *info) } } -static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, - struct genl_info *info) +static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, + struct genl_info *info) { + struct pdp_ctx *pctx, *pctx_tid = NULL; struct net_device *dev = gtp->dev; u32 hash_ms, hash_tid = 0; - struct pdp_ctx *pctx; + unsigned int version; bool found = false; __be32 ms_addr; ms_addr = nla_get_be32(info->attrs[GTPA_MS_ADDRESS]); hash_ms = ipv4_hashfn(ms_addr) % gtp->hash_size; + version = nla_get_u32(info->attrs[GTPA_VERSION]); - hlist_for_each_entry_rcu(pctx, &gtp->addr_hash[hash_ms], hlist_addr) { - if (pctx->ms_addr_ip4.s_addr == ms_addr) { - found = true; - break; - } - } + pctx = ipv4_pdp_find(gtp, ms_addr); + if (pctx) + found = true; + if (version == GTP_V0) + pctx_tid = gtp0_pdp_find(gtp, + nla_get_u64(info->attrs[GTPA_TID])); + else if (version == GTP_V1) + pctx_tid = gtp1_pdp_find(gtp, + nla_get_u32(info->attrs[GTPA_I_TEI])); + if (pctx_tid) + found = true; if (found) { if (info->nlhdr->nlmsg_flags & NLM_F_EXCL) @@ -951,6 +958,11 @@ static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, if (info->nlhdr->nlmsg_flags & NLM_F_REPLACE) return -EOPNOTSUPP; + if (pctx && pctx_tid) + return -EEXIST; + if (!pctx) + pctx = pctx_tid; + ipv4_pdp_fill(pctx, info); if (pctx->gtp_version == GTP_V0) @@ -1074,7 +1086,7 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) goto out_unlock; } - err = ipv4_pdp_add(gtp, sk, info); + err = gtp_pdp_add(gtp, sk, info); out_unlock: rcu_read_unlock(); -- 2.20.1

1 0

[PATCH net 0/4] gtp: fix several bugs in gtp module
by Taehee Yoo 15 Dec '19

15 Dec '19

This patchset fixes several bugs in the GTP module. 1. Do not allow adding duplicate TID and ms_addr pdp context. In the current code, duplicate TID and ms_addr pdp context could be added. So, RX and TX path could find correct pdp context. 2. Fix wrong condition in ->dumpit() callback. ->dumpit() callback is re-called if dump packet size is too big. So, before return, it saves last position and then restart from last dump position. TID value is used to find last dump position. GTP module allows adding zero TID value. But ->dumpit() callback ignores zero TID value. So, dump would not work correctly if dump packet size too big. 3. Fix use-after-free in ipv4_pdp_find(). RX and TX patch always uses gtp->tid_hash and gtp->addr_hash. but while packet processing, these hash pointer would be freed. So, use-after-free would occur. 4. Fix panic because of zero size hashtable GTP hashtable size could be set by user-space. If hashsize is set to 0, hashtable will not work and panic will occur. Taehee Yoo (4): gtp: do not allow adding duplicate tid and ms_addr pdp context gtp: fix wrong condition in gtp_genl_dump_pdp() gtp: fix an use-after-free in ipv4_pdp_find() gtp: avoid zero size hashtable drivers/net/gtp.c | 109 +++++++++++++++++++++++++++------------------- 1 file changed, 63 insertions(+), 46 deletions(-) -- 2.17.1

2 1

[PATCH net 4/4] gtp: avoid zero size hashtable
by Taehee Yoo 11 Dec '19

11 Dec '19

GTP default hashtable size is 1024 and userspace could set specific hashtable size with IFLA_GTP_PDP_HASHSIZE. If hashtable size is set to 0 from userspace, hashtable will not work and panic will occur. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> --- drivers/net/gtp.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 5450b1099c6d..e5b7d6d2286e 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -667,10 +667,13 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, if (err < 0) return err; - if (!data[IFLA_GTP_PDP_HASHSIZE]) + if (!data[IFLA_GTP_PDP_HASHSIZE]) { hashsize = 1024; - else + } else { hashsize = nla_get_u32(data[IFLA_GTP_PDP_HASHSIZE]); + if (!hashsize) + hashsize = 1024; + } err = gtp_hashtable_new(gtp, hashsize); if (err < 0) -- 2.17.1

1 0

[PATCH net 3/4] gtp: fix an use-after-free in ipv4_pdp_find()
by Taehee Yoo 11 Dec '19

11 Dec '19

ipv4_pdp_find() is called in TX packet path of GTP. ipv4_pdp_find() internally uses gtp->tid_hash to lookup pdp context. In the current code, gtp->tid_hash and gtp->addr_hash are freed by ->dellink(), which is gtp_dellink(). But gtp_dellink() would be called while packets are processing. So, gtp_dellink() should not free gtp->tid_hash and gtp->addr_hash. Instead, dev->priv_destructor() would be used because this callback is called after all packet processing safely. Test commands: ip link add veth1 type veth peer name veth2 ip a a 172.0.0.1/24 dev veth1 ip link set veth1 up ip a a 172.99.0.1/32 dev lo gtp-link add gtp1 & gtp-tunnel add gtp1 v1 200 100 172.99.0.2 172.0.0.2 ip r a 172.99.0.2/32 dev gtp1 ip link set gtp1 mtu 1500 ip netns add ns2 ip link set veth2 netns ns2 ip netns exec ns2 ip a a 172.0.0.2/24 dev veth2 ip netns exec ns2 ip link set veth2 up ip netns exec ns2 ip a a 172.99.0.2/32 dev lo ip netns exec ns2 ip link set lo up ip netns exec ns2 gtp-link add gtp2 & ip netns exec ns2 gtp-tunnel add gtp2 v1 100 200 172.99.0.1 172.0.0.1 ip netns exec ns2 ip r a 172.99.0.1/32 dev gtp2 ip netns exec ns2 ip link set gtp2 mtu 1500 hping3 172.99.0.2 -2 --flood & ip link del gtp1 Splat looks like: [ 72.568081][ T1195] BUG: KASAN: use-after-free in ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.568916][ T1195] Read of size 8 at addr ffff8880b9a35d28 by task hping3/1195 [ 72.569631][ T1195] [ 72.569861][ T1195] CPU: 2 PID: 1195 Comm: hping3 Not tainted 5.5.0-rc1 #199 [ 72.570547][ T1195] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006 [ 72.571438][ T1195] Call Trace: [ 72.571764][ T1195] dump_stack+0x96/0xdb [ 72.572171][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.572761][ T1195] print_address_description.constprop.5+0x1be/0x360 [ 72.573400][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.573971][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.574544][ T1195] __kasan_report+0x12a/0x16f [ 72.575014][ T1195] ? ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.575593][ T1195] kasan_report+0xe/0x20 [ 72.576004][ T1195] ipv4_pdp_find.isra.12+0x130/0x170 [gtp] [ 72.576577][ T1195] gtp_build_skb_ip4+0x199/0x1420 [gtp] [ ... ] [ 72.647671][ T1195] BUG: unable to handle page fault for address: ffff8880b9a35d28 [ 72.648512][ T1195] #PF: supervisor read access in kernel mode [ 72.649158][ T1195] #PF: error_code(0x0000) - not-present page [ 72.649849][ T1195] PGD a6c01067 P4D a6c01067 PUD 11fb07067 PMD 11f939067 PTE 800fffff465ca060 [ 72.652958][ T1195] Oops: 0000 [#1] SMP DEBUG_PAGEALLOC KASAN PTI [ 72.653834][ T1195] CPU: 2 PID: 1195 Comm: hping3 Tainted: G B 5.5.0-rc1 #199 [ 72.668062][ T1195] RIP: 0010:ipv4_pdp_find.isra.12+0x86/0x170 [gtp] [ ... ] [ 72.679168][ T1195] Call Trace: [ 72.679603][ T1195] gtp_build_skb_ip4+0x199/0x1420 [gtp] [ 72.681915][ T1195] ? ipv4_pdp_find.isra.12+0x170/0x170 [gtp] [ 72.682513][ T1195] ? lock_acquire+0x164/0x3b0 [ 72.682966][ T1195] ? gtp_dev_xmit+0x35e/0x890 [gtp] [ 72.683481][ T1195] gtp_dev_xmit+0x3c2/0x890 [gtp] [ ... ] Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> --- drivers/net/gtp.c | 34 +++++++++++++++++----------------- 1 file changed, 17 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index a010e0a11c33..5450b1099c6d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -640,9 +640,16 @@ static void gtp_link_setup(struct net_device *dev) } static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize); -static void gtp_hashtable_free(struct gtp_dev *gtp); static int gtp_encap_enable(struct gtp_dev *gtp, struct nlattr *data[]); +static void gtp_destructor(struct net_device *dev) +{ + struct gtp_dev *gtp = netdev_priv(dev); + + kfree(gtp->addr_hash); + kfree(gtp->tid_hash); +} + static int gtp_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) @@ -677,13 +684,15 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, gn = net_generic(dev_net(dev), gtp_net_id); list_add_rcu(&gtp->list, &gn->gtp_dev_list); + dev->priv_destructor = gtp_destructor; netdev_dbg(dev, "registered new GTP interface\n"); return 0; out_hashtable: - gtp_hashtable_free(gtp); + kfree(gtp->addr_hash); + kfree(gtp->tid_hash); out_encap: gtp_encap_disable(gtp); return err; @@ -692,8 +701,13 @@ static int gtp_newlink(struct net *src_net, struct net_device *dev, static void gtp_dellink(struct net_device *dev, struct list_head *head) { struct gtp_dev *gtp = netdev_priv(dev); + struct pdp_ctx *pctx; + int 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); - gtp_hashtable_free(gtp); list_del_rcu(&gtp->list); unregister_netdevice_queue(dev, head); } @@ -771,20 +785,6 @@ static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize) return -ENOMEM; } -static void gtp_hashtable_free(struct gtp_dev *gtp) -{ - struct pdp_ctx *pctx; - int 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); - - synchronize_rcu(); - kfree(gtp->addr_hash); - kfree(gtp->tid_hash); -} - static struct sock *gtp_encap_enable_socket(int fd, int type, struct gtp_dev *gtp) { -- 2.17.1

1 0

[PATCH net 2/4] gtp: fix wrong condition in gtp_genl_dump_pdp()
by Taehee Yoo 11 Dec '19

11 Dec '19

gtp_genl_dump_pdp() is ->dumpit() callback of GTP module and it is used to dump pdp contexts. it would be re-executed because of dump packet size. If dump packet size is too big, it saves current dump pointer (gtp interface pointer, bucket, TID value) then it restarts dump from last pointer. Current GTP code allows adding zero TID pdp context but dump code ignores zero TID value. So, last dump pointer will not be found. In addition, this patch adds missing rcu_read_lock() in gtp_genl_dump_pdp(). Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> --- drivers/net/gtp.c | 36 +++++++++++++++++++----------------- 1 file changed, 19 insertions(+), 17 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 8b742edf793d..a010e0a11c33 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -38,7 +38,6 @@ struct pdp_ctx { struct hlist_node hlist_addr; union { - u64 tid; struct { u64 tid; u16 flow; @@ -1244,43 +1243,46 @@ static int gtp_genl_dump_pdp(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_net *gn = net_generic(net, gtp_net_id); - unsigned long tid = cb->args[1]; - int i, k = cb->args[0], ret; struct pdp_ctx *pctx; + 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 = k; i < gtp->hash_size; i++) { - hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], hlist_tid) { - if (tid && tid != pctx->u.tid) - continue; - else - tid = 0; - - ret = gtp_genl_fill_info(skb, - NETLINK_CB(cb->skb).portid, - cb->nlh->nlmsg_seq, - cb->nlh->nlmsg_type, pctx); - if (ret < 0) { + for (i = bucket; i < gtp->hash_size; i++) { + j = 0; + hlist_for_each_entry_rcu(pctx, &gtp->tid_hash[i], + hlist_tid) { + if (j >= skip && + gtp_genl_fill_info(skb, + NETLINK_CB(cb->skb).portid, + cb->nlh->nlmsg_seq, + cb->nlh->nlmsg_type, pctx)) { cb->args[0] = i; - cb->args[1] = pctx->u.tid; + 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; } -- 2.17.1

1 0

[PATCH net 1/4] gtp: do not allow adding duplicate tid and ms_addr pdp context
by Taehee Yoo 11 Dec '19

11 Dec '19

GTP RX packet path lookups pdp context with TID. If duplicate TID pdp contexts are existing in the list, it couldn't select correct pdp context. So, TID value should be unique. GTP TX packet path lookups pdp context with ms_addr. If duplicate ms_addr pdp contexts are existing in the list, it couldn't select correct pdp context. So, ms_addr value should be unique. Fixes: 459aa660eb1d ("gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U)") Signed-off-by: Taehee Yoo <ap420073(a)gmail.com> --- drivers/net/gtp.c | 32 ++++++++++++++++++++++---------- 1 file changed, 22 insertions(+), 10 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index ecfe26215935..8b742edf793d 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -926,24 +926,31 @@ static void ipv4_pdp_fill(struct pdp_ctx *pctx, struct genl_info *info) } } -static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, - struct genl_info *info) +static int gtp_pdp_add(struct gtp_dev *gtp, struct sock *sk, + struct genl_info *info) { + struct pdp_ctx *pctx, *pctx_tid = NULL; struct net_device *dev = gtp->dev; u32 hash_ms, hash_tid = 0; - struct pdp_ctx *pctx; + unsigned int version; bool found = false; __be32 ms_addr; ms_addr = nla_get_be32(info->attrs[GTPA_MS_ADDRESS]); hash_ms = ipv4_hashfn(ms_addr) % gtp->hash_size; + version = nla_get_u32(info->attrs[GTPA_VERSION]); - hlist_for_each_entry_rcu(pctx, &gtp->addr_hash[hash_ms], hlist_addr) { - if (pctx->ms_addr_ip4.s_addr == ms_addr) { - found = true; - break; - } - } + pctx = ipv4_pdp_find(gtp, ms_addr); + if (pctx) + found = true; + if (version == GTP_V0) + pctx_tid = gtp0_pdp_find(gtp, + nla_get_u64(info->attrs[GTPA_TID])); + else if (version == GTP_V1) + pctx_tid = gtp1_pdp_find(gtp, + nla_get_u32(info->attrs[GTPA_I_TEI])); + if (pctx_tid) + found = true; if (found) { if (info->nlhdr->nlmsg_flags & NLM_F_EXCL) @@ -951,6 +958,11 @@ static int ipv4_pdp_add(struct gtp_dev *gtp, struct sock *sk, if (info->nlhdr->nlmsg_flags & NLM_F_REPLACE) return -EOPNOTSUPP; + if (pctx && pctx_tid) + return -EEXIST; + if (!pctx) + pctx = pctx_tid; + ipv4_pdp_fill(pctx, info); if (pctx->gtp_version == GTP_V0) @@ -1074,7 +1086,7 @@ static int gtp_genl_new_pdp(struct sk_buff *skb, struct genl_info *info) goto out_unlock; } - err = ipv4_pdp_add(gtp, sk, info); + err = gtp_pdp_add(gtp, sk, info); out_unlock: rcu_read_unlock(); -- 2.17.1

1 0

OsmoDevCon 2020 registration
by Harald Welte 01 Dec '19

01 Dec '19

Dear fellow Osmocom developers, I would like to invite all developers and contributors to Osmocom [sub]projects to register for OsmoDevCon 2020 (held on April 24th-27th, 2020 in Berlin). For details known so far, please check http://osmocom.org/projects/osmo-dev-con/wiki/OsmoDevCon2020 Please enter your name at https://osmocom.org/projects/osmo-dev-con/wiki/OsmoDevCon2020#Requested in case you would like to attend. Registering early allows proper planning. Thanks! Looking forward to meeting old and new Osmocom developers in April 2020. 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)

1 0

Re: high ksoftirqd while using module gtp?
by Tony Clark 22 Nov '19

22 Nov '19

Firstly I would like to say great thanks Firat for the reply, it certainly put me on a different investigation path. And apologies for not replying sooner. I wanted to make sure it was the correct path before I replied back to the group with the findings and associated solution. If the GTP-U connection is connecting to the PG-W with a single IP at (src/dst) each side, and UDP flow has isn't been enabled on the network card of the host using gtp.ko in the kernel all the associated network traffic will be received on a single queue on the network, which is then serviced by a single ksoftirq thread. at somepoint the system will be receiving more traffic than there is available thread space to service the request, your ksoftirq will burn at 100%. That means all your traffic will be bound to a single network queue, bound to a single irq thread, and limit your overall throughput, no matter how big your network pipe is. This is because the network card hashed the packet via SRC_IP:SRC_PORT:DEST_IP:DEST_PORT:PROTO to a single queue. # take note on the discussions about udp-flow-hash udp4 using ethtool https://home.regit.org/tag/performance/ https://www.joyent.com/blog/virtualizing-nics https://www.serializing.me/2015/04/25/rxtx-buffers-rss-others-on-boot/ You can check if your card supports adjustable parameters by using "ethtool -k DEV | egrep -v fixed". As firat eludes to (below) udp flow hashing should be supported. If you enabled UDP flow hash then it will spread the hash over multiple queues. The default number of queues on the network card can vary, depending on your hardware firmware driver, and any additional associated kernel parameters. Would recommend having the latest firmware driver for your network card, and latest kernel driver for the network card if possible. Alas the network cards used by my hardware didn't support flow hash, it had intel flow director, which wasn't granular enough and worked with TCP, so to work around this limitation having multiple SRC_IPs in different name spaces with the same GTP UDP PORT numbers resolved the problem. Of course if you are sending GTP-U to a single destination from multiple sources (say 6 IP's), via 6 different kernel name spaces, you spread the load over 6 queues, which is better than nothing on a limited feature network card. Time to upgrade the 10G network card.... This took the system from 100% ksoftirq on a single cpu running at throughput 1G, to around 7 to 8GIG throughput at 90% ksoftirq over multiple cpu;s... There is still massive room for improvement. For performance some things to investigate/consider... Which I had different levels of success... Here are my ramblings..... on the linux host... Assuming your traffic is now spread across multiple queues (above) - or at least spread as best as can be... Kernel sysctl tweaking is always of benefit, if your using out of the box kernel config... Example udp buffers, queue sizes, paging and virtual memory settings... There is a application called "tuned", which allows you to adjust profiles for the kernel sysctl... My performance profile which suited the testing best was "throuhput-performance" if your looking for straight performance, disable audit processing like "auditd". Question use of SELINUX, enforcing/permissive or disabled. can bring results on performance, if you doing testing or load testing... ofcourse its a security consideration.. If you don't need to use ipfilters/firewall in my case can increase the throughput by a 3rd by disabling (cleaning the filter tables and unloading the modules). Black listing the modules so they dont get loaded at kernel time. Note you can stop modules getting loaded with kernel.modules_disabled=1, but be careful if your also messing with initramfs rebuilds, because you don't get any modules once your set that parameter, i learnt the hard way :) Investigate smp_affinity and affinity_hint, along with irqbalane using the --hintpolicy=exact. understand which irq's service the network cards, and how many queues you have.. /proc/interrupts will guide you (grep 'CPU|rxtx' /proc/interrupts)... understand the smp_affinity numbers.. "for ((irq=START_IRQ; irq<=END_IRQ; irq++)); do cat /proc/irq/$irq/smp_affinity; done | sort –u", as you can adjust which queue goes to which ksoftirq to manually balance the queues if you so desire. brilliant document on irq debugging.... https://events.static.linuxfound.org/sites/events/files/slides/LinuxConJapa… you can monitor what calls are been executed on cpu's by using... I found this most useful to understand that ipfilter was eating a significant amount of CPU cycles, and also what other calls are eating up cycles inside ksoftirq. https://github.com/brendangregg/FlameGraph Investigate additional memory management using numactl (numa daemon). remember if you are using virtualisation you might want to pin guests to specific sockets, along with numa pinning on the vmhost.. Also look at reserved memory allocation in the vmhost for the guest... This will make your guest perform better. enable sysstat (sar) as it will aid your investigation if you havent already (sar -u ALL -P ALL 1). This will show which softirqs are eating most cpu and to which cpu they are bound, this also translates directly to the network queue that the traffic is coming in on.. Ie, network card queue 6, talks to cpu/6 talking to irq/6 and so on... Using flamegraph will help you understand what syscalls and chewing the CPU.. If your using virtualisation then the number of default queues that vxnet (vmware in this example) presents to the guest might be less than the number of network card queues the vmhost sees (so watch out for that). You can adjust the number of queues to the guest by params in the vmware network driver... investigate VDMQ / netqueue, to increase the number of available hardware queues from the vmhost to the guest. depending which quest driver your using vxnet3, or others some drivers dont support NAPI (see further down). * VMDQ: array of int* * Number of Virtual Machine Device Queues: 0/1 = disable, 2-16 enable (default=8)* * RSS: array of int* * Number of Receive-Side Scaling Descriptor Queues, default 1=number of cpus* * MQ: array of int* * Disable or enable Multiple Queues, default 1* Node: array of int set the starting node to allocate memory on, default -1 * IntMode: array of int* * Change Interrupt Mode (0=Legacy, 1=MSI, 2=MSI-X), default 2* * InterruptType: array of int* * Change Interrupt Mode (0=Legacy, 1=MSI, 2=MSI-X), default IntMode (deprecated)* Make sure your virtual switch (vmware) if used has Pass-through (Direct-path I/O) enabled. NIC teaming policy should be validated depending on your requirement, example Policy "route based on IP hash" can be of benefit. Check the network card is MSI-X, and the linux driver supports NAPI (most should these days, but you never know), also check your vmhost driver supports napi, if not get a NAPI supported kvm driver, or vmware driver (vib update). Upgrade your kernel, to a later release 4.x.. even consider using a later distro of linux... I tried fedora 29. I also compiled latest osmocom from source, with compile options for "optimisation -O3 and other such". "bmon -b" was a good tool understand throughput loads, along with loading through qdisc/fq_dodel mq's.... Understand qdisc via ip link or ifconfig ( http://tldp.org/HOWTO/Traffic-Control-HOWTO/components.html), adjusting the queues has some traction, but if unsure leave as default. TSO/UFo?GSO/LRO/GRO - understand your network card with respects to these, this can improve performance if you haven't already enabled (or adversely disabled options, since sometimes it doesn't actually help). You can get the your card options using ethool TCP Segmentation Offload (TSO) Uses the TCP protocol to send large packets. Uses the NIC to handle segmentation, and then adds the TCP, IP and data link layer protocol headers to each segment. UDP Fragmentation Offload (UFO) Uses the UDP protocol to send large packets. Uses the NIC to handle IP fragmentation into MTU sized packets for large UDP datagrams. Generic Segmentation Offload (GSO) Uses the TCP or UDP protocol to send large packets. If the NIC cannot handle segmentation/fragmentation, GSO performs the same operations, bypassing the NIC hardware. This is achieved by delaying segmentation until as late as possible, for example, when the packet is processed by the device driver. Large Receive Offload (LRO) Uses the TCP protocol. All incoming packets are re-segmented as they are received, reducing the number of segments the system has to process. They can be merged either in the driver or using the NIC. A problem with LRO is that it tends to resegment all incoming packets, often ignoring differences in headers and other information which can cause errors. It is generally not possible to use LRO when IP forwarding is enabled. LRO in combination with IP forwarding can lead to checksum errors. Forwarding is enabled if /proc/sys/net/ipv4/ip_forward is set to 1. Generic Receive Offload (GRO) Uses either the TCP or UDP protocols. GRO is more rigorous than LRO when resegmenting packets. For example it checks the MAC headers of each packet, which must match, only a limited number of TCP or IP headers can be different, and the TCP timestamps must match. Resegmenting can be handled by either the NIC or the GSO code. Traffic steering was on by default with the version of linux i was using, but worth checking if your using older versions. https://www.kernel.org/doc/Documentation/networking/scaling.txt (from the txt link) note: Some advanced NICs allow steering packets to queues based on programmable filters. For example, webserver bound TCP port 80 packets can be directed to their own receive queue. Such â€œn-tupleâ€ filters can be configured from ethtool (--config-ntuple). Interestingly investigate your network card, for its hashing algorithms, how it distributes the traffic over its ring buffers, you can on some cards adjust the RSS hash function. Alas the card i was using stuck to "toeplitz" for hits hashing, which others were disabled and unavailable / xor and crc32. The indirection table can be adjusted based on the tuplets "ethtool -X" but didn't really assist too much on this. ethtool -x <dev> RX flow hash indirection table for ens192 with 8 RX ring(s): 0: 0 1 2 3 4 5 6 7 8: 0 1 2 3 4 5 6 7 16: 0 1 2 3 4 5 6 7 24: 0 1 2 3 4 5 6 7 RSS hash key: Operation not supported RSS hash function: toeplitz: on xor: off crc32: off Check the default size of the rx/tx ring buffers, they maybe suboptimal. ethtool -g ens192 Ring parameters for ens192: Pre-set maximums: RX: 4096 RX Mini: 0 RX Jumbo: 4096 TX: 4096 Current hardware settings: RX: 1024 RX Mini: 0 RX Jumbo: 256 TX: 512 If your using port channels, make sure you have the correct hashing policy enabled at the switch end... I haven't investigated this option yet but some switches also do scaling, to assist (certainly with virtualisation)... Maybe one day i will get around to this... Additionally CISCO describe that you should have VM-FEX optimisation https://www.cisco.com/c/en/us/solutions/collateral/data-center-virtualizati… note: *table 4.* Scaling of Dynamic vNIC with VMDirectPath, Virtual Machines Running on Linux Guest with VMXNET3 Emulated Driver and Multi-Queue Enabled *Table 5.* Scaling of Dynamic vNIC with VMDirectPath, Virtual Machines Running on Linux Guest with VMXNET3 Emulated Driver and Multi-Queue Disabled Another thing to consider/investigate - openvswitch/bridging... If your using eth pairs to send your traffic down name spaces... you can have some varied results with performance by trying openvswitch/brctl I really enjoyed the investigation path, again thanks to Firat for the pointer, otherwise it would have taken longer to get the answer... Tony On Fri, Jun 21, 2019 at 6:50 AM fırat sönmez <firatssonmez(a)gmail.com> wrote: > Hi, > > It has been over 2 years that I have worked with gtp and I kind of had the > same problem that time, we had a 10gbit cable and tried to see how much udp > flow we could get. I think we used iperf to test it and when we list all > the processes, the ksoftirq was using all the resource. Then I found this > page: https://blog.cloudflare.com/how-to-receive-a-million-packets/. I do > not remember the exact solution, but I guess when you configure your out > ethernet interface with the command below, it must work then. To my > understanding all the packets are processed in the same core in your > situation, because the port number is always the same. So, for example, if > you add another network with gtp-u tunnel on another port (different than > 3386) then again your packets will be processed on the other core, too. But > with the below command, the interface will be configured in a way that it > wont check the port to process on which core it should be processed, but it > will use the hash from the packet to distribute over the cores. > ethtool -n (your_out_eth_interface) rx-flow-hash udp4 > > Hope it will work you. > > Fırat > > Tony Clark <chiefy.padua(a)gmail.com>, 19 Haz 2019 Çar, 15:07 tarihinde > şunu yazdı: > >> Dear All, >> >> I've been using the GTP-U kernel module to communicate with a P-GW. >> >> Running Fedora 29, kernel 4.18.16-300.fc29.x86_64. >> >> At high traffic levels through the GTP-U tunnel I see the performance >> degrade as 100% CPU is consumed by a single ksoftirqd process. >> >> It is running on a multi-cpu machine and as far as I can tell the load is >> evenly spread across the cpus (ie either manually via smp_affinity, or even >> irqbalance, checking /proc/interrupts so forth.). >> >> Has anyone else experienced this? >> >> Is there any particular area you could recommend I investigate to find >> the root cause of this bottleneck, as i'm starting to scratch my head where >> to look next... >> >> Thanks in advance >> Tony >> >> ---- FYI >> >> modinfo gtp >> filename: >> /lib/modules/4.18.16-300.fc29.x86_64/kernel/drivers/net/gtp.ko.xz >> alias: net-pf-16-proto-16-family-gtp >> alias: rtnl-link-gtp >> description: Interface driver for GTP encapsulated traffic >> author: Harald Welte <hwelte(a)sysmocom.de> >> license: GPL >> depends: udp_tunnel >> retpoline: Y >> intree: Y >> name: gtp >> vermagic: 4.18.16-300.fc29.x86_64 SMP mod_unload >> >> modinfo udp_tunnel >> filename: >> /lib/modules/4.18.16-300.fc29.x86_64/kernel/net/ipv4/udp_tunnel.ko.xz >> license: GPL >> depends: >> retpoline: Y >> intree: Y >> name: udp_tunnel >> vermagic: 4.18.16-300.fc29.x86_64 SMP mod_unload >> >

2 1

EDGE MCS not stable
by ghiturbe＠tutanota.com 06 Nov '19

06 Nov '19

Dear all, Please do not consider the last message. First I want to introduce myslef, I'm Gael and few months ago I bought a Lime-Mini to use as BTS with EDGE. I have spent some time working in that and it is doing it but I have noticed that the UP LINK MCS is not stable, it changes very much with the mobile near the board. I have read that this issue was solved in the ticket 1833 but it is not working for me, I hope that this not about my board. Do you have any suggest for this behavior? To solve that issue, I force as I read in another post, setting the minimum an maximum values to zero till the MCS to be used. It works fine setting the pcu-cfg file like: ... mcs7 0 35 mcs8 35 35. It works but it does not look well and if I change by VTY to use mcs 9 it works too but is not posible to start the system with this configuration, so I have to start it using mcs7. With this cfg I can reach 130kbps using a mobile class 12 ( 4TS RX + TX), anyway I have 6 TS configured for PDCH. I guess this mobile should reach almost 200 kbps (4TS * 54Kbps). What is the maximum speed reached? Thanks in advance for your support. Best regards. -- Securely sent with Tutanota. Get your own encrypted, ad-free mailbox: https://tutanota.com

1 0

Correo confidencial de ghiturbe
by ghiturbe＠tutanota.com 06 Nov '19

06 Nov '19

1 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

osmocom-net-gprs