[PATCH 0/5] simple gtp improvements
As requested are here the simple and most obvious changes from the larger GTP changeset. I'll break down the rest and send them out separately. The module alias addition and the rcu_lock removal are just small convenience changes. The other changes are needed to correctly implement one of the 3GPP GW functions, userspace needs to see invalid T-PDU's in order to generate the proper error messages, GTP fragmentation rules need the cleared DF bit. Andreas -- Andreas Schultz (5): gtp: add genl family modules alias gtp: clear DF bit on GTP packet tx gtp: fix cross netns recv on gtp socket gtp: remove unnecessary rcu_read_lock gtp: let userspace handle packets for invalid tunnels drivers/net/gtp.c | 33 ++++++++------------------------- 1 file changed, 8 insertions(+), 25 deletions(-) -- 2.10.2
Auto-load the module when userspace asks for the gtp netlink family. Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- drivers/net/gtp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 8b6810b..7580ccc 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -1376,3 +1376,4 @@ MODULE_LICENSE("GPL"); MODULE_AUTHOR("Harald Welte <hwelte@sysmocom.de>"); MODULE_DESCRIPTION("Interface driver for GTP encapsulated traffic"); MODULE_ALIAS_RTNL_LINK("gtp"); +MODULE_ALIAS_GENL_FAMILY("gtp"); -- 2.10.2
On Tue, Jan 24, 2017 at 06:23:58PM +0100, Andreas Schultz wrote:
Auto-load the module when userspace asks for the gtp netlink family.
This qualifies as fix, since autoload is broken. You may send a batch including this for David's net tree.
3GPP TS 29.281 and 3GPP TS 29.060 imply that GTP-U packets should be sent with the DF bit cleared. For example 3GPP TS 29.060, Release 8, Section 13.2.2:
Backbone router: Any router in the backbone may fragment the GTP packet if needed, according to IPv4.
Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- 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 7580ccc..1df54d6 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -612,7 +612,7 @@ static netdev_tx_t gtp_dev_xmit(struct sk_buff *skb, struct net_device *dev) pktinfo.fl4.saddr, pktinfo.fl4.daddr, pktinfo.iph->tos, ip4_dst_hoplimit(&pktinfo.rt->dst), - htons(IP_DF), + 0, pktinfo.gtph_port, pktinfo.gtph_port, true, false); break; -- 2.10.2
On Tue, Jan 24, 2017 at 06:23:59PM +0100, Andreas Schultz wrote:
3GPP TS 29.281 and 3GPP TS 29.060 imply that GTP-U packets should be sent with the DF bit cleared. For example 3GPP TS 29.060, Release 8, Section 13.2.2:
Backbone router: Any router in the backbone may fragment the GTP packet if needed, according to IPv4.
Given this is fixing a broken implementation with regards to standards, please target this to net.
The use of the passed through netlink src_net to check for a cross netns operation was wrong. Using the GTP socket and the GTP netdevice is always correct (even if the netdev has been moved to new netns after link creation). Remove the now obsolete net field from gtp_dev. Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- drivers/net/gtp.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 1df54d6..72dd1ba 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -69,7 +69,6 @@ struct gtp_dev { struct socket *sock0; struct socket *sock1u; - struct net *net; struct net_device *dev; unsigned int hash_size; @@ -316,7 +315,7 @@ static int gtp_encap_recv(struct sock *sk, struct sk_buff *skb) netdev_dbg(gtp->dev, "encap_recv sk=%p\n", sk); - xnet = !net_eq(gtp->net, dev_net(gtp->dev)); + xnet = !net_eq(sock_net(sk), dev_net(gtp->dev)); switch (udp_sk(sk)->encap_type) { case UDP_ENCAP_GTP0: @@ -658,7 +657,7 @@ 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 net_device *dev, struct gtp_dev *gtp, - int fd_gtp0, int fd_gtp1, struct net *src_net); + int fd_gtp0, int fd_gtp1); static int gtp_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[], struct nlattr *data[]) @@ -858,7 +857,6 @@ static int gtp_encap_enable(struct net_device *dev, struct gtp_dev *gtp, gtp->sock0 = sock0; gtp->sock1u = sock1u; - gtp->net = src_net; tuncfg.sk_user_data = gtp; tuncfg.encap_rcv = gtp_encap_recv; -- 2.10.2
On Tue, Jan 24, 2017 at 06:24:00PM +0100, Andreas Schultz wrote:
The use of the passed through netlink src_net to check for a cross netns operation was wrong. Using the GTP socket and the GTP netdevice is always correct (even if the netdev has been moved to new netns after link creation).
Remove the now obsolete net field from gtp_dev.
The net tree can take fixes anytime, so if you target this patch to [PATCH net] this speeds up integration into mainline kernels. Note, as soon as this patch hits Linus tree, we can request -stable submission so older -stable kernels can get this. If this follows the net-next path, then this fix is going to take several weeks (sometimes months) to show in mainline kernels. So please add [PATCH net] or [PATCH net-next] so it's clear to everyone what is your target, David usually requests this. BTW, probably you can target this small fix to net, then you can request David to pull net into net-next so the fix propagates onwards. Sorry for this bureaucratic stuff, but given the workload we deal with, you will really helps us if you deal with these nitpicks. Thanks Andreas!
Hi Andreas, [auto build test ERROR on net-next/master] [also build test ERROR on v4.10-rc5 next-20170124] [if your patch is applied to the wrong git tree, please drop us a note to help improve the system] url: https://github.com/0day-ci/linux/commits/Andreas-Schultz/simple-gtp-improvem... config: i386-allmodconfig (attached as .config) compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901 reproduce: # save the attached .config to linux build tree make ARCH=i386 All error/warnings (new ones prefixed by >>): drivers/net/gtp.c: In function 'gtp_newlink':
drivers/net/gtp.c:677:8: error: too many arguments to function 'gtp_encap_enable' err = gtp_encap_enable(dev, gtp, fd0, fd1, src_net); ^~~~~~~~~~~~~~~~ drivers/net/gtp.c:659:12: note: declared here static int gtp_encap_enable(struct net_device *dev, struct gtp_dev *gtp, ^~~~~~~~~~~~~~~~ drivers/net/gtp.c: At top level: drivers/net/gtp.c:822:12: error: conflicting types for 'gtp_encap_enable' static int gtp_encap_enable(struct net_device *dev, struct gtp_dev *gtp, ^~~~~~~~~~~~~~~~ drivers/net/gtp.c:659:12: note: previous declaration of 'gtp_encap_enable' was here static int gtp_encap_enable(struct net_device *dev, struct gtp_dev *gtp, ^~~~~~~~~~~~~~~~ drivers/net/gtp.c:659:12: warning: 'gtp_encap_enable' used but never defined drivers/net/gtp.c:822:12: warning: 'gtp_encap_enable' defined but not used [-Wunused-function] static int gtp_encap_enable(struct net_device *dev, struct gtp_dev *gtp, ^~~~~~~~~~~~~~~~
vim +/gtp_encap_enable +677 drivers/net/gtp.c 459aa660 Pablo Neira 2016-05-09 653 sizeof(struct udphdr) + 459aa660 Pablo Neira 2016-05-09 654 sizeof(struct gtp0_header); 459aa660 Pablo Neira 2016-05-09 655 } 459aa660 Pablo Neira 2016-05-09 656 459aa660 Pablo Neira 2016-05-09 657 static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize); 459aa660 Pablo Neira 2016-05-09 658 static void gtp_hashtable_free(struct gtp_dev *gtp); 459aa660 Pablo Neira 2016-05-09 @659 static int gtp_encap_enable(struct net_device *dev, struct gtp_dev *gtp, 0e9ed139 Andreas Schultz 2017-01-24 660 int fd_gtp0, int fd_gtp1); 459aa660 Pablo Neira 2016-05-09 661 459aa660 Pablo Neira 2016-05-09 662 static int gtp_newlink(struct net *src_net, struct net_device *dev, 459aa660 Pablo Neira 2016-05-09 663 struct nlattr *tb[], struct nlattr *data[]) 459aa660 Pablo Neira 2016-05-09 664 { 459aa660 Pablo Neira 2016-05-09 665 int hashsize, err, fd0, fd1; 459aa660 Pablo Neira 2016-05-09 666 struct gtp_dev *gtp; 459aa660 Pablo Neira 2016-05-09 667 struct gtp_net *gn; 459aa660 Pablo Neira 2016-05-09 668 459aa660 Pablo Neira 2016-05-09 669 if (!data[IFLA_GTP_FD0] || !data[IFLA_GTP_FD1]) 459aa660 Pablo Neira 2016-05-09 670 return -EINVAL; 459aa660 Pablo Neira 2016-05-09 671 459aa660 Pablo Neira 2016-05-09 672 gtp = netdev_priv(dev); 459aa660 Pablo Neira 2016-05-09 673 459aa660 Pablo Neira 2016-05-09 674 fd0 = nla_get_u32(data[IFLA_GTP_FD0]); 459aa660 Pablo Neira 2016-05-09 675 fd1 = nla_get_u32(data[IFLA_GTP_FD1]); 459aa660 Pablo Neira 2016-05-09 676 459aa660 Pablo Neira 2016-05-09 @677 err = gtp_encap_enable(dev, gtp, fd0, fd1, src_net); 459aa660 Pablo Neira 2016-05-09 678 if (err < 0) 459aa660 Pablo Neira 2016-05-09 679 goto out_err; 459aa660 Pablo Neira 2016-05-09 680 459aa660 Pablo Neira 2016-05-09 681 if (!data[IFLA_GTP_PDP_HASHSIZE]) 459aa660 Pablo Neira 2016-05-09 682 hashsize = 1024; 459aa660 Pablo Neira 2016-05-09 683 else 459aa660 Pablo Neira 2016-05-09 684 hashsize = nla_get_u32(data[IFLA_GTP_PDP_HASHSIZE]); 459aa660 Pablo Neira 2016-05-09 685 459aa660 Pablo Neira 2016-05-09 686 err = gtp_hashtable_new(gtp, hashsize); 459aa660 Pablo Neira 2016-05-09 687 if (err < 0) 459aa660 Pablo Neira 2016-05-09 688 goto out_encap; 459aa660 Pablo Neira 2016-05-09 689 459aa660 Pablo Neira 2016-05-09 690 err = register_netdevice(dev); 459aa660 Pablo Neira 2016-05-09 691 if (err < 0) { 459aa660 Pablo Neira 2016-05-09 692 netdev_dbg(dev, "failed to register new netdev %d\n", err); 459aa660 Pablo Neira 2016-05-09 693 goto out_hashtable; 459aa660 Pablo Neira 2016-05-09 694 } 459aa660 Pablo Neira 2016-05-09 695 459aa660 Pablo Neira 2016-05-09 696 gn = net_generic(dev_net(dev), gtp_net_id); 459aa660 Pablo Neira 2016-05-09 697 list_add_rcu(>p->list, &gn->gtp_dev_list); 459aa660 Pablo Neira 2016-05-09 698 459aa660 Pablo Neira 2016-05-09 699 netdev_dbg(dev, "registered new GTP interface\n"); 459aa660 Pablo Neira 2016-05-09 700 459aa660 Pablo Neira 2016-05-09 701 return 0; 459aa660 Pablo Neira 2016-05-09 702 459aa660 Pablo Neira 2016-05-09 703 out_hashtable: 459aa660 Pablo Neira 2016-05-09 704 gtp_hashtable_free(gtp); 459aa660 Pablo Neira 2016-05-09 705 out_encap: 459aa660 Pablo Neira 2016-05-09 706 gtp_encap_disable(gtp); 459aa660 Pablo Neira 2016-05-09 707 out_err: 459aa660 Pablo Neira 2016-05-09 708 return err; 459aa660 Pablo Neira 2016-05-09 709 } 459aa660 Pablo Neira 2016-05-09 710 459aa660 Pablo Neira 2016-05-09 711 static void gtp_dellink(struct net_device *dev, struct list_head *head) 459aa660 Pablo Neira 2016-05-09 712 { 459aa660 Pablo Neira 2016-05-09 713 struct gtp_dev *gtp = netdev_priv(dev); 459aa660 Pablo Neira 2016-05-09 714 459aa660 Pablo Neira 2016-05-09 715 gtp_encap_disable(gtp); 459aa660 Pablo Neira 2016-05-09 716 gtp_hashtable_free(gtp); 459aa660 Pablo Neira 2016-05-09 717 list_del_rcu(>p->list); 459aa660 Pablo Neira 2016-05-09 718 unregister_netdevice_queue(dev, head); 459aa660 Pablo Neira 2016-05-09 719 } 459aa660 Pablo Neira 2016-05-09 720 459aa660 Pablo Neira 2016-05-09 721 static const struct nla_policy gtp_policy[IFLA_GTP_MAX + 1] = { 459aa660 Pablo Neira 2016-05-09 722 [IFLA_GTP_FD0] = { .type = NLA_U32 }, 459aa660 Pablo Neira 2016-05-09 723 [IFLA_GTP_FD1] = { .type = NLA_U32 }, 459aa660 Pablo Neira 2016-05-09 724 [IFLA_GTP_PDP_HASHSIZE] = { .type = NLA_U32 }, 459aa660 Pablo Neira 2016-05-09 725 }; 459aa660 Pablo Neira 2016-05-09 726 459aa660 Pablo Neira 2016-05-09 727 static int gtp_validate(struct nlattr *tb[], struct nlattr *data[]) 459aa660 Pablo Neira 2016-05-09 728 { 459aa660 Pablo Neira 2016-05-09 729 if (!data) 459aa660 Pablo Neira 2016-05-09 730 return -EINVAL; 459aa660 Pablo Neira 2016-05-09 731 459aa660 Pablo Neira 2016-05-09 732 return 0; 459aa660 Pablo Neira 2016-05-09 733 } 459aa660 Pablo Neira 2016-05-09 734 459aa660 Pablo Neira 2016-05-09 735 static size_t gtp_get_size(const struct net_device *dev) 459aa660 Pablo Neira 2016-05-09 736 { 459aa660 Pablo Neira 2016-05-09 737 return nla_total_size(sizeof(__u32)); /* IFLA_GTP_PDP_HASHSIZE */ 459aa660 Pablo Neira 2016-05-09 738 } 459aa660 Pablo Neira 2016-05-09 739 459aa660 Pablo Neira 2016-05-09 740 static int gtp_fill_info(struct sk_buff *skb, const struct net_device *dev) 459aa660 Pablo Neira 2016-05-09 741 { 459aa660 Pablo Neira 2016-05-09 742 struct gtp_dev *gtp = netdev_priv(dev); 459aa660 Pablo Neira 2016-05-09 743 459aa660 Pablo Neira 2016-05-09 744 if (nla_put_u32(skb, IFLA_GTP_PDP_HASHSIZE, gtp->hash_size)) 459aa660 Pablo Neira 2016-05-09 745 goto nla_put_failure; 459aa660 Pablo Neira 2016-05-09 746 459aa660 Pablo Neira 2016-05-09 747 return 0; 459aa660 Pablo Neira 2016-05-09 748 459aa660 Pablo Neira 2016-05-09 749 nla_put_failure: 459aa660 Pablo Neira 2016-05-09 750 return -EMSGSIZE; 459aa660 Pablo Neira 2016-05-09 751 } 459aa660 Pablo Neira 2016-05-09 752 459aa660 Pablo Neira 2016-05-09 753 static struct rtnl_link_ops gtp_link_ops __read_mostly = { 459aa660 Pablo Neira 2016-05-09 754 .kind = "gtp", 459aa660 Pablo Neira 2016-05-09 755 .maxtype = IFLA_GTP_MAX, 459aa660 Pablo Neira 2016-05-09 756 .policy = gtp_policy, 459aa660 Pablo Neira 2016-05-09 757 .priv_size = sizeof(struct gtp_dev), 459aa660 Pablo Neira 2016-05-09 758 .setup = gtp_link_setup, 459aa660 Pablo Neira 2016-05-09 759 .validate = gtp_validate, 459aa660 Pablo Neira 2016-05-09 760 .newlink = gtp_newlink, 459aa660 Pablo Neira 2016-05-09 761 .dellink = gtp_dellink, 459aa660 Pablo Neira 2016-05-09 762 .get_size = gtp_get_size, 459aa660 Pablo Neira 2016-05-09 763 .fill_info = gtp_fill_info, 459aa660 Pablo Neira 2016-05-09 764 }; 459aa660 Pablo Neira 2016-05-09 765 459aa660 Pablo Neira 2016-05-09 766 static struct net *gtp_genl_get_net(struct net *src_net, struct nlattr *tb[]) 459aa660 Pablo Neira 2016-05-09 767 { 459aa660 Pablo Neira 2016-05-09 768 struct net *net; 459aa660 Pablo Neira 2016-05-09 769 459aa660 Pablo Neira 2016-05-09 770 /* Examine the link attributes and figure out which network namespace 459aa660 Pablo Neira 2016-05-09 771 * we are talking about. 459aa660 Pablo Neira 2016-05-09 772 */ 459aa660 Pablo Neira 2016-05-09 773 if (tb[GTPA_NET_NS_FD]) 459aa660 Pablo Neira 2016-05-09 774 net = get_net_ns_by_fd(nla_get_u32(tb[GTPA_NET_NS_FD])); 459aa660 Pablo Neira 2016-05-09 775 else 459aa660 Pablo Neira 2016-05-09 776 net = get_net(src_net); 459aa660 Pablo Neira 2016-05-09 777 459aa660 Pablo Neira 2016-05-09 778 return net; 459aa660 Pablo Neira 2016-05-09 779 } 459aa660 Pablo Neira 2016-05-09 780 459aa660 Pablo Neira 2016-05-09 781 static int gtp_hashtable_new(struct gtp_dev *gtp, int hsize) 459aa660 Pablo Neira 2016-05-09 782 { 459aa660 Pablo Neira 2016-05-09 783 int i; 459aa660 Pablo Neira 2016-05-09 784 459aa660 Pablo Neira 2016-05-09 785 gtp->addr_hash = kmalloc(sizeof(struct hlist_head) * hsize, GFP_KERNEL); 459aa660 Pablo Neira 2016-05-09 786 if (gtp->addr_hash == NULL) 459aa660 Pablo Neira 2016-05-09 787 return -ENOMEM; 459aa660 Pablo Neira 2016-05-09 788 459aa660 Pablo Neira 2016-05-09 789 gtp->tid_hash = kmalloc(sizeof(struct hlist_head) * hsize, GFP_KERNEL); 459aa660 Pablo Neira 2016-05-09 790 if (gtp->tid_hash == NULL) 459aa660 Pablo Neira 2016-05-09 791 goto err1; 459aa660 Pablo Neira 2016-05-09 792 459aa660 Pablo Neira 2016-05-09 793 gtp->hash_size = hsize; 459aa660 Pablo Neira 2016-05-09 794 459aa660 Pablo Neira 2016-05-09 795 for (i = 0; i < hsize; i++) { 459aa660 Pablo Neira 2016-05-09 796 INIT_HLIST_HEAD(>p->addr_hash[i]); 459aa660 Pablo Neira 2016-05-09 797 INIT_HLIST_HEAD(>p->tid_hash[i]); 459aa660 Pablo Neira 2016-05-09 798 } 459aa660 Pablo Neira 2016-05-09 799 return 0; 459aa660 Pablo Neira 2016-05-09 800 err1: 459aa660 Pablo Neira 2016-05-09 801 kfree(gtp->addr_hash); 459aa660 Pablo Neira 2016-05-09 802 return -ENOMEM; 459aa660 Pablo Neira 2016-05-09 803 } 459aa660 Pablo Neira 2016-05-09 804 459aa660 Pablo Neira 2016-05-09 805 static void gtp_hashtable_free(struct gtp_dev *gtp) 459aa660 Pablo Neira 2016-05-09 806 { 459aa660 Pablo Neira 2016-05-09 807 struct pdp_ctx *pctx; 459aa660 Pablo Neira 2016-05-09 808 int i; 459aa660 Pablo Neira 2016-05-09 809 459aa660 Pablo Neira 2016-05-09 810 for (i = 0; i < gtp->hash_size; i++) { 459aa660 Pablo Neira 2016-05-09 811 hlist_for_each_entry_rcu(pctx, >p->tid_hash[i], hlist_tid) { 459aa660 Pablo Neira 2016-05-09 812 hlist_del_rcu(&pctx->hlist_tid); 459aa660 Pablo Neira 2016-05-09 813 hlist_del_rcu(&pctx->hlist_addr); 459aa660 Pablo Neira 2016-05-09 814 kfree_rcu(pctx, rcu_head); 459aa660 Pablo Neira 2016-05-09 815 } 459aa660 Pablo Neira 2016-05-09 816 } 459aa660 Pablo Neira 2016-05-09 817 synchronize_rcu(); 459aa660 Pablo Neira 2016-05-09 818 kfree(gtp->addr_hash); 459aa660 Pablo Neira 2016-05-09 819 kfree(gtp->tid_hash); 459aa660 Pablo Neira 2016-05-09 820 } 459aa660 Pablo Neira 2016-05-09 821 459aa660 Pablo Neira 2016-05-09 @822 static int gtp_encap_enable(struct net_device *dev, struct gtp_dev *gtp, 459aa660 Pablo Neira 2016-05-09 823 int fd_gtp0, int fd_gtp1, struct net *src_net) 459aa660 Pablo Neira 2016-05-09 824 { 459aa660 Pablo Neira 2016-05-09 825 struct udp_tunnel_sock_cfg tuncfg = {NULL}; :::::: The code at line 677 was first introduced by commit :::::: 459aa660eb1d8ce67080da1983bb81d716aa5a69 gtp: add initial driver for datapath of GPRS Tunneling Protocol (GTP-U) :::::: TO: Pablo Neira <pablo@netfilter.org> :::::: CC: David S. Miller <davem@davemloft.net> --- 0-DAY kernel test infrastructure Open Source Technology Center https://lists.01.org/pipermail/kbuild-all Intel Corporation
The rcu read lock is hold by default in the ip input path. There is no need to hold it twice in the socket recv decapsulate code path. Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- drivers/net/gtp.c | 24 ++++-------------------- 1 file changed, 4 insertions(+), 20 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 72dd1ba..912721e 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -183,7 +183,6 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, sizeof(struct gtp0_header); struct gtp0_header *gtp0; struct pdp_ctx *pctx; - int ret = 0; if (!pskb_may_pull(skb, hdrlen)) return -1; @@ -196,26 +195,19 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, if (gtp0->type != GTP_TPDU) return 1; - rcu_read_lock(); 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); - ret = -1; - goto out_rcu; + return -1; } if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - ret = -1; - goto out_rcu; + return -1; } - rcu_read_unlock(); /* Get rid of the GTP + UDP headers. */ return iptunnel_pull_header(skb, hdrlen, skb->protocol, xnet); -out_rcu: - rcu_read_unlock(); - return ret; } static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, @@ -225,7 +217,6 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, sizeof(struct gtp1_header); struct gtp1_header *gtp1; struct pdp_ctx *pctx; - int ret = 0; if (!pskb_may_pull(skb, hdrlen)) return -1; @@ -253,26 +244,19 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, gtp1 = (struct gtp1_header *)(skb->data + sizeof(struct udphdr)); - rcu_read_lock(); pctx = gtp1_pdp_find(gtp, ntohl(gtp1->tid)); if (!pctx) { netdev_dbg(gtp->dev, "No PDP ctx to decap skb=%p\n", skb); - ret = -1; - goto out_rcu; + return -1; } if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - ret = -1; - goto out_rcu; + return -1; } - rcu_read_unlock(); /* Get rid of the GTP + UDP headers. */ return iptunnel_pull_header(skb, hdrlen, skb->protocol, xnet); -out_rcu: - rcu_read_unlock(); - return ret; } static void gtp_encap_disable(struct gtp_dev *gtp) -- 2.10.2
On Tue, Jan 24, 2017 at 06:24:01PM +0100, Andreas Schultz wrote:
The rcu read lock is hold by default in the ip input path. There is no need to hold it twice in the socket recv decapsulate code path.
I think this is net-next material since it is not essencial.
enable userspace to send error replies for invalid tunnels Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- drivers/net/gtp.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 912721e..c607333 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -198,12 +198,12 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, 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 1; } if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - return -1; + return 1; } /* Get rid of the GTP + UDP headers. */ @@ -247,12 +247,12 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, 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 1; } if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - return -1; + return 1; } /* Get rid of the GTP + UDP headers. */ -- 2.10.2
Hi Andreas, On Tue, Jan 24, 2017 at 06:24:02PM +0100, Andreas Schultz wrote:
enable userspace to send error replies for invalid tunnels
Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- drivers/net/gtp.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 912721e..c607333 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -198,12 +198,12 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, 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 1; }
if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - return -1; + return 1;
So userspace gets the packet that we cannot forward. I guess your userspace codebase performs this sanity checks again so you can send the appropriate error reply?
}
/* Get rid of the GTP + UDP headers. */ @@ -247,12 +247,12 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, 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 1; }
if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - return -1; + return 1; }
/* Get rid of the GTP + UDP headers. */ -- 2.10.2
Hi Pablo, ----- On Jan 24, 2017, at 8:03 PM, pablo pablo@netfilter.org wrote:
Hi Andreas,
On Tue, Jan 24, 2017 at 06:24:02PM +0100, Andreas Schultz wrote:
enable userspace to send error replies for invalid tunnels
Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- drivers/net/gtp.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 912721e..c607333 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -198,12 +198,12 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, 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 1; }
if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - return -1; + return 1;
So userspace gets the packet that we cannot forward. I guess your userspace codebase performs this sanity checks again so you can send the appropriate error reply?
For TEID /= 0, the only reply is a T-PDU of type error indication. There is no cause specified. So I don't actually have to repeat the check. TEID == 0 is more interesting, this tells userspace that it tried to send on an invalid tunnel and should tear it down. If you like, you can have a look at the userspace code. The relevant piece is at https://github.com/travelping/gtp_u_kmod/blob/master/src/gtp_u_kmod_port.erl... But be warned, it's written in Erlang ;-) Andreas
}
/* Get rid of the GTP + UDP headers. */ @@ -247,12 +247,12 @@ static int gtp1u_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, 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 1; }
if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - return -1; + return 1; }
/* Get rid of the GTP + UDP headers. */ -- 2.10.2
On Tue, Jan 24, 2017 at 09:02:31PM +0100, Andreas Schultz wrote:
Hi Pablo,
----- On Jan 24, 2017, at 8:03 PM, pablo pablo@netfilter.org wrote:
Hi Andreas,
On Tue, Jan 24, 2017 at 06:24:02PM +0100, Andreas Schultz wrote:
enable userspace to send error replies for invalid tunnels
Signed-off-by: Andreas Schultz <aschultz@tpip.net> --- drivers/net/gtp.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index 912721e..c607333 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -198,12 +198,12 @@ static int gtp0_udp_encap_recv(struct gtp_dev *gtp, struct sk_buff *skb, 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 1; }
if (!gtp_check_src_ms(skb, pctx, hdrlen)) { netdev_dbg(gtp->dev, "No PDP ctx for this MS\n"); - return -1; + return 1;
So userspace gets the packet that we cannot forward. I guess your userspace codebase performs this sanity checks again so you can send the appropriate error reply?
For TEID /= 0, the only reply is a T-PDU of type error indication. There is no cause specified. So I don't actually have to repeat the check. TEID == 0 is more interesting, this tells userspace that it tried to send on an invalid tunnel and should tear it down.
Ah, I see, thanks for explaining, I guess this is good to debug misconfigurations.
If you like, you can have a look at the userspace code. The relevant piece is at https://github.com/travelping/gtp_u_kmod/blob/master/src/gtp_u_kmod_port.erl...
But be warned, it's written in Erlang ;-)
Will have a look, but you taking me away from my usual domains :).
Hi Andreas, I agree with your changes (particularly those related to 3GPP specs) like 2/5 and 5/5. Also, 1/5 is of course obvious. For kernel topics like 3/5 and 4/5 I trust Pablo and the general netdev crew to have better judgement than me. -- - Harald Welte <laforge@netfilter.org> http://netfilter.org/ ============================================================================ "Fragmentation is like classful addressing -- an interesting early architectural error that shows how much experimentation was going on while IP was being designed." -- Paul Vixie
participants (4)
-
Andreas Schultz -
Harald Welte -
kbuild test robot -
Pablo Neira Ayuso