Received: by 2002:a25:5b86:0:0:0:0:0 with SMTP id p128csp501411ybb; Thu, 28 Mar 2019 06:48:28 -0700 (PDT) X-Google-Smtp-Source: APXvYqz22b8b1zgd9NlcMC885E3bTVziO8hVwK+Z4P6CiCfLQRUsV1FM8SQKocurgl8/yQ2qMV1C X-Received: by 2002:a17:902:f24:: with SMTP id 33mr41064212ply.47.1553780908684; Thu, 28 Mar 2019 06:48:28 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1553780908; cv=none; d=google.com; s=arc-20160816; b=JM+ShISaPtXqo7lqLctpsE10F8i8000tpHaGbe/hd1Lpjwd7QHCCUpDjvnZXkGCw88 dn+1He8ERZMmp+nmXf18wlTTSBdG4i6LF5Xx0zfOPT17icsDkfwnnkRI1looknjzY4e9 ertY2eBpAerntKm+bMWw9kFY4IQsPHF58OGs31Nk8DkdXFoR1efWs632zPx85zCjCCON cLO0pyo6y95neRqklItJfCArBDVSp/y6cgCMIr4aLcK/lDfjLkbC16mmzSk4aTUroTKD b0R60+d0Edu4Gfdu/ydBLd6cdzNg0C7dUBJpQo0U1tgNJ28sXcHP5FaLJ41rBYqoN8ch TWzA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:user-agent:in-reply-to :content-disposition:mime-version:references:message-id:subject:cc :to:from:date:dkim-signature; bh=H1SONzOFrhCaVeBVVjXYzVdUnnLq9j3zRsbhDm/J90o=; b=z98PySxvMPaAjYCpYYRlH9GkWsf2lJ2EWxndG7jxZ3ihfTmhyY3zglv5KcBDND2k3g +ZTeieT0FqXRl9wGSwWNfnb/74HjGKpKIooe9MxzNoydOGbbZ1m1GqgABbX+1LK/0kVX Pd9y/gvZvWsmBUpsD3lxRSc2Dlp0MymmxC58cJ5cmJNoYJcE2UDFrZG7ZMPvJX02saSC QRG8FGZfnc0KQFtkWnk7V3OQj06D41eOexdlNDUThhXf6NruIVOwAd1Pwho+BYQ596Aj shIRkZtVZ1MI5tG6BO1j15OrD9zuC78aIrHwUBA9IzmHOykPHF9pKuG2IK5q4CDQi9YV 0iuA== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@gmail.com header.s=20161025 header.b=omVx9PYM; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=QUARANTINE dis=NONE) header.from=gmail.com Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id f3si21909860plf.300.2019.03.28.06.48.12; Thu, 28 Mar 2019 06:48:28 -0700 (PDT) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; dkim=pass header.i=@gmail.com header.s=20161025 header.b=omVx9PYM; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=QUARANTINE dis=NONE) header.from=gmail.com Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1726455AbfC1Nrf (ORCPT + 99 others); Thu, 28 Mar 2019 09:47:35 -0400 Received: from mail-pg1-f194.google.com ([209.85.215.194]:41704 "EHLO mail-pg1-f194.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726190AbfC1Nre (ORCPT ); Thu, 28 Mar 2019 09:47:34 -0400 Received: by mail-pg1-f194.google.com with SMTP id f6so7510928pgs.8; Thu, 28 Mar 2019 06:47:34 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=date:from:to:cc:subject:message-id:references:mime-version :content-disposition:in-reply-to:user-agent; bh=H1SONzOFrhCaVeBVVjXYzVdUnnLq9j3zRsbhDm/J90o=; b=omVx9PYM3taBMDNkfaGprmXJnoEiZoP859gKG1mVMbRv/nHpQzAY0xDEf7H4yJGF0W lYdVtyvs2KEUaYROR+7HRaUzPEubJaE9TjNALwsvPoUeRg891J1BSKpOF6i5sve6NItv IUe8REhSfWAafNLPjiNwHwGFuz9E0tPTIzX0u8QPFjeJQDOyNhJeuJuhDO4mJ2PNqpmV H/EsvR8cW1j6kxNA32wp2rFuvMqne+DXxLcLNShhuYm7eH7ps7ycp4K3ZW9V3LDyWrlX yf9qAZzlEtWvD2zj/QRVV0zITRj37vaOUUHcyKmBZsRRO9iqDq0j+toco2QPMq1VldjO BSlQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:date:from:to:cc:subject:message-id:references :mime-version:content-disposition:in-reply-to:user-agent; bh=H1SONzOFrhCaVeBVVjXYzVdUnnLq9j3zRsbhDm/J90o=; b=sdrUiZCFuYEotupxsuqceHZLj4ohl6q0p6vJM32H6uNUMPKlJhWH8WPTtaVp9P52PS UMNd+Gg2HHD2WeitRf45gTLdfgK5UOF4gfkQ4qu2HJSI9lLN7rFBClKNunIMRRhx+h4O /gbgBytZOKs62vFc/vZIPDWwSk4PnNJWF02PPgwHkQ4DZdaJeMJ8tr8HlIXMfoJC4QKx vXi15EmC23y9ZkqGxdckDhqWe+PAWcnfGqPH7YXjmFCgfCke0KrTJX5GFOSP98MxkJ7i XQpurllvrLt8WHkQemgpzAhng1dan25aI5RR0F5DvTyfFTJAhBwi8ft87u6zU8IFnqpT 7XEw== X-Gm-Message-State: APjAAAVY4+f/NwP9FUl6DutCog4CsTB/LNBXDth10oS8LVIYrjhT19It /Y4++NlgR459U11v9J+/WQ== X-Received: by 2002:a63:c104:: with SMTP id w4mr12818601pgf.409.1553780853665; Thu, 28 Mar 2019 06:47:33 -0700 (PDT) Received: from localhost ([2001:19f0:7001:54c5:5400:1ff:fec8:7fc2]) by smtp.gmail.com with ESMTPSA id d130sm51950299pfg.49.2019.03.28.06.47.30 (version=TLS1_2 cipher=ECDHE-RSA-CHACHA20-POLY1305 bits=256/256); Thu, 28 Mar 2019 06:47:32 -0700 (PDT) Date: Thu, 28 Mar 2019 21:47:26 +0800 From: Jacky Hu To: Simon Horman Cc: jacky.hu@walmart.com, jason.niesz@walmart.com, "David S. Miller" , Alexey Kuznetsov , Hideaki YOSHIFUJI , Wensong Zhang , Julian Anastasov , Pablo Neira Ayuso , Jozsef Kadlecsik , Florian Westphal , netdev@vger.kernel.org, lvs-devel@vger.kernel.org, linux-kernel@vger.kernel.org, netfilter-devel@vger.kernel.org, coreteam@netfilter.org Subject: Re: [PATCH v7] ipvs: allow tunneling with gue encapsulation Message-ID: <20190328134726.GA2246@i716> References: <20190326103121.31083-1-hengqing.hu@gmail.com> <20190328095950.66xhna76xndokdp6@verge.net.au> <20190328114708.GA10198@i716> <20190328123909.5uqvokkcxh2sjvfl@verge.net.au> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <20190328123909.5uqvokkcxh2sjvfl@verge.net.au> User-Agent: Mutt/1.11.4 (2019-03-13) Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Mar 28, 2019 at 01:39:09PM +0100, Simon Horman wrote: > On Thu, Mar 28, 2019 at 07:47:08PM +0800, Jacky Hu wrote: > > Hi Simon, > > On Thu, Mar 28, 2019 at 10:59:50AM +0100, Simon Horman wrote: > > > Hi Jacky, > > > > > > On Tue, Mar 26, 2019 at 06:31:21PM +0800, Jacky Hu wrote: > > > > ipip packets are blocked in some public cloud environments, this patch > > > > allows gue encapsulation with the tunneling method, which would make > > > > tunneling working in those environments. > > > > > > > > Signed-off-by: Jacky Hu > > > > > > It would help a lot if you provided a short summary > > > of what changed between versions of this patch. > > > > > > > Here is the summary: > > > > v7->v6: > > 1) pass proper gso type mask to gso_inner_segment for gue tunnel > > > > v6->v5: > > 1) simply using an if statement for tun_type discrimination > > > > v5->v4: > > 1) use __be16 for tun_port and __u16 for tun_type and avoid any > > atomic_t usage > > 2) fixed sparse warnings > > > > v4->v3: > > 1) removed changes to setsockopt interface > > 2) use correct nla_get/put function for tun_port > > 3) moved cp->dest null check to __ip_vs_get_out_rt() and > > __ip_vs_get_out_rt_v6() > > 4) account the added header when calculating mtu > > 5) use net instead of dev_net(tdev) > > 6) set udest->tun_type to IP_VS_CONN_F_TUNNEL_TYPE_IPIP in > > ip_vs_copy_udest_compat() > > 7) renamed __build_gue_header to ipvs_gue_encap > > 8) make ip_vs_add_dest() and ip_vs_edit_dest() return EINVAL when > > GUE comes with port 0. > > 9) feed AF_INET6 to __tun_gso_type_mask in ip_vs_tunnel_xmit_v6 > > > > v3->v2: > > 1) added missing break statements to a few switch cases > > > > v2->v1: > > 1) pass tun_type and tun_port as new optional parameters > > instead of a few bits in existing conn_flags parameters > > > > > > --- > > > > include/net/ip_vs.h | 5 ++ > > > > include/uapi/linux/ip_vs.h | 11 +++++ > > > > net/netfilter/ipvs/ip_vs_ctl.c | 35 +++++++++++++- > > > > net/netfilter/ipvs/ip_vs_xmit.c | 84 +++++++++++++++++++++++++++++++-- > > > > 4 files changed, 130 insertions(+), 5 deletions(-) > > > > > > > > diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h > > > > index 047f9a5ccaad..2ac40135b576 100644 > > > > --- a/include/net/ip_vs.h > > > > +++ b/include/net/ip_vs.h > > > > @@ -600,6 +600,9 @@ struct ip_vs_dest_user_kern { > > > > > > > > /* Address family of addr */ > > > > u16 af; > > > > + > > > > + u16 tun_type; /* tunnel type */ > > > > + __be16 tun_port; /* tunnel port */ > > > > }; > > > > > > > > > > > > @@ -660,6 +663,8 @@ struct ip_vs_dest { > > > > atomic_t conn_flags; /* flags to copy to conn */ > > > > atomic_t weight; /* server weight */ > > > > atomic_t last_weight; /* server latest weight */ > > > > + __u16 tun_type; /* tunnel type */ > > > > + __be16 tun_port; /* tunnel port */ > > > > > > > > refcount_t refcnt; /* reference counter */ > > > > struct ip_vs_stats stats; /* statistics */ > > > > diff --git a/include/uapi/linux/ip_vs.h b/include/uapi/linux/ip_vs.h > > > > index 1c916b2f89dc..e34f436fc79d 100644 > > > > --- a/include/uapi/linux/ip_vs.h > > > > +++ b/include/uapi/linux/ip_vs.h > > > > @@ -124,6 +124,13 @@ > > > > > > > > #define IP_VS_PEDATA_MAXLEN 255 > > > > > > > > +/* Tunnel types */ > > > > +enum { > > > > + IP_VS_CONN_F_TUNNEL_TYPE_IPIP = 0, /* IPIP */ > > > > + IP_VS_CONN_F_TUNNEL_TYPE_GUE, /* GUE */ > > > > + IP_VS_CONN_F_TUNNEL_TYPE_MAX, > > > > +}; > > > > + > > > > /* > > > > * The struct ip_vs_service_user and struct ip_vs_dest_user are > > > > * used to set IPVS rules through setsockopt. > > > > @@ -392,6 +399,10 @@ enum { > > > > > > > > IPVS_DEST_ATTR_STATS64, /* nested attribute for dest stats */ > > > > > > > > + IPVS_DEST_ATTR_TUN_TYPE, /* tunnel type */ > > > > + > > > > + IPVS_DEST_ATTR_TUN_PORT, /* tunnel port */ > > > > + > > > > __IPVS_DEST_ATTR_MAX, > > > > }; > > > > > > > > diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c > > > > index 053cd96b9c76..328683452229 100644 > > > > --- a/net/netfilter/ipvs/ip_vs_ctl.c > > > > +++ b/net/netfilter/ipvs/ip_vs_ctl.c > > > > @@ -831,6 +831,10 @@ __ip_vs_update_dest(struct ip_vs_service *svc, struct ip_vs_dest *dest, > > > > conn_flags = udest->conn_flags & IP_VS_CONN_F_DEST_MASK; > > > > conn_flags |= IP_VS_CONN_F_INACTIVE; > > > > > > > > + /* set the tunnel info */ > > > > + dest->tun_type = udest->tun_type; > > > > + dest->tun_port = udest->tun_port; > > > > + > > > > /* set the IP_VS_CONN_F_NOOUTPUT flag if not masquerading/NAT */ > > > > if ((conn_flags & IP_VS_CONN_F_FWD_MASK) != IP_VS_CONN_F_MASQ) { > > > > conn_flags |= IP_VS_CONN_F_NOOUTPUT; > > > > @@ -987,6 +991,13 @@ ip_vs_add_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) > > > > return -ERANGE; > > > > } > > > > > > > > + if (udest->tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) { > > > > + if (udest->tun_port == 0) { > > > > + pr_err("%s(): tunnel port is zero\n", __func__); > > > > + return -EINVAL; > > > > + } > > > > + } > > > > + > > > > ip_vs_addr_copy(udest->af, &daddr, &udest->addr); > > > > > > > > /* We use function that requires RCU lock */ > > > > @@ -1051,6 +1062,13 @@ ip_vs_edit_dest(struct ip_vs_service *svc, struct ip_vs_dest_user_kern *udest) > > > > return -ERANGE; > > > > } > > > > > > > > + if (udest->tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) { > > > > + if (udest->tun_port == 0) { > > > > + pr_err("%s(): tunnel port is zero\n", __func__); > > > > + return -EINVAL; > > > > + } > > > > + } > > > > + > > > > ip_vs_addr_copy(udest->af, &daddr, &udest->addr); > > > > > > > > /* We use function that requires RCU lock */ > > > > @@ -2333,6 +2351,7 @@ static void ip_vs_copy_udest_compat(struct ip_vs_dest_user_kern *udest, > > > > udest->u_threshold = udest_compat->u_threshold; > > > > udest->l_threshold = udest_compat->l_threshold; > > > > udest->af = AF_INET; > > > > + udest->tun_type = IP_VS_CONN_F_TUNNEL_TYPE_IPIP; > > > > } > > > > > > > > static int > > > > @@ -2890,6 +2909,8 @@ static const struct nla_policy ip_vs_dest_policy[IPVS_DEST_ATTR_MAX + 1] = { > > > > [IPVS_DEST_ATTR_PERSIST_CONNS] = { .type = NLA_U32 }, > > > > [IPVS_DEST_ATTR_STATS] = { .type = NLA_NESTED }, > > > > [IPVS_DEST_ATTR_ADDR_FAMILY] = { .type = NLA_U16 }, > > > > + [IPVS_DEST_ATTR_TUN_TYPE] = { .type = NLA_U8 }, > > > > + [IPVS_DEST_ATTR_TUN_PORT] = { .type = NLA_U16 }, > > > > }; > > > > > > > > static int ip_vs_genl_fill_stats(struct sk_buff *skb, int container_type, > > > > @@ -3193,6 +3214,10 @@ static int ip_vs_genl_fill_dest(struct sk_buff *skb, struct ip_vs_dest *dest) > > > > IP_VS_CONN_F_FWD_MASK)) || > > > > nla_put_u32(skb, IPVS_DEST_ATTR_WEIGHT, > > > > atomic_read(&dest->weight)) || > > > > + nla_put_u8(skb, IPVS_DEST_ATTR_TUN_TYPE, > > > > + dest->tun_type) || > > > > + nla_put_be16(skb, IPVS_DEST_ATTR_TUN_PORT, > > > > + dest->tun_port) || > > > > nla_put_u32(skb, IPVS_DEST_ATTR_U_THRESH, dest->u_threshold) || > > > > nla_put_u32(skb, IPVS_DEST_ATTR_L_THRESH, dest->l_threshold) || > > > > nla_put_u32(skb, IPVS_DEST_ATTR_ACTIVE_CONNS, > > > > @@ -3315,12 +3340,14 @@ static int ip_vs_genl_parse_dest(struct ip_vs_dest_user_kern *udest, > > > > /* If a full entry was requested, check for the additional fields */ > > > > if (full_entry) { > > > > struct nlattr *nla_fwd, *nla_weight, *nla_u_thresh, > > > > - *nla_l_thresh; > > > > + *nla_l_thresh, *nla_tun_type, *nla_tun_port; > > > > > > > > nla_fwd = attrs[IPVS_DEST_ATTR_FWD_METHOD]; > > > > nla_weight = attrs[IPVS_DEST_ATTR_WEIGHT]; > > > > nla_u_thresh = attrs[IPVS_DEST_ATTR_U_THRESH]; > > > > nla_l_thresh = attrs[IPVS_DEST_ATTR_L_THRESH]; > > > > + nla_tun_type = attrs[IPVS_DEST_ATTR_TUN_TYPE]; > > > > + nla_tun_port = attrs[IPVS_DEST_ATTR_TUN_PORT]; > > > > > > > > if (!(nla_fwd && nla_weight && nla_u_thresh && nla_l_thresh)) > > > > return -EINVAL; > > > > @@ -3330,6 +3357,12 @@ static int ip_vs_genl_parse_dest(struct ip_vs_dest_user_kern *udest, > > > > udest->weight = nla_get_u32(nla_weight); > > > > udest->u_threshold = nla_get_u32(nla_u_thresh); > > > > udest->l_threshold = nla_get_u32(nla_l_thresh); > > > > + > > > > + if (nla_tun_type) > > > > + udest->tun_type = nla_get_u8(nla_tun_type); > > > > + > > > > + if (nla_tun_port) > > > > + udest->tun_port = nla_get_be16(nla_tun_port); > > > > } > > > > > > > > return 0; > > > > diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c > > > > index 175349fcf91f..8d6f94b67772 100644 > > > > --- a/net/netfilter/ipvs/ip_vs_xmit.c > > > > +++ b/net/netfilter/ipvs/ip_vs_xmit.c > > > > @@ -32,6 +32,7 @@ > > > > #include > > > > #include /* for tcphdr */ > > > > #include > > > > +#include > > > > #include /* for csum_tcpudp_magic */ > > > > #include > > > > #include /* for icmp_send */ > > > > @@ -382,6 +383,10 @@ __ip_vs_get_out_rt(struct netns_ipvs *ipvs, int skb_af, struct sk_buff *skb, > > > > mtu = dst_mtu(&rt->dst); > > > > } else { > > > > mtu = dst_mtu(&rt->dst) - sizeof(struct iphdr); > > > > + if (!dest) > > > > + goto err_put; > > > > + if (dest->tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + mtu -= sizeof(struct udphdr) + sizeof(struct guehdr); > > > > if (mtu < 68) { > > > > IP_VS_DBG_RL("%s(): mtu less than 68\n", __func__); > > > > goto err_put; > > > > @@ -533,6 +538,10 @@ __ip_vs_get_out_rt_v6(struct netns_ipvs *ipvs, int skb_af, struct sk_buff *skb, > > > > mtu = dst_mtu(&rt->dst); > > > > else { > > > > mtu = dst_mtu(&rt->dst) - sizeof(struct ipv6hdr); > > > > + if (!dest) > > > > + goto err_put; > > > > + if (dest->tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + mtu -= sizeof(struct udphdr) + sizeof(struct guehdr); > > > > if (mtu < IPV6_MIN_MTU) { > > > > IP_VS_DBG_RL("%s(): mtu less than %d\n", __func__, > > > > IPV6_MIN_MTU); > > > > @@ -989,6 +998,41 @@ static inline int __tun_gso_type_mask(int encaps_af, int orig_af) > > > > } > > > > } > > > > > > > > +static int > > > > +ipvs_gue_encap(struct net *net, struct sk_buff *skb, > > > > + struct ip_vs_conn *cp, __u8 *next_protocol) > > > > +{ > > > > + __be16 dport; > > > > + __be16 sport = udp_flow_src_port(net, skb, 0, 0, false); > > > > + struct udphdr *udph; /* Our new UDP header */ > > > > + struct guehdr *gueh; /* Our new GUE header */ > > > > + > > > > + skb_push(skb, sizeof(struct guehdr)); > > > > + > > > > + gueh = (struct guehdr *)skb->data; > > > > + > > > > + gueh->control = 0; > > > > + gueh->version = 0; > > > > + gueh->hlen = 0; > > > > + gueh->flags = 0; > > > > + gueh->proto_ctype = *next_protocol; > > > > + > > > > + skb_push(skb, sizeof(struct udphdr)); > > > > + skb_reset_transport_header(skb); > > > > + > > > > + udph = udp_hdr(skb); > > > > + > > > > + dport = cp->dest->tun_port; > > > > + udph->dest = dport; > > > > + udph->source = sport; > > > > + udph->len = htons(skb->len); > > > > + udph->check = 0; > > > > + > > > > + *next_protocol = IPPROTO_UDP; > > > > + > > > > + return 0; > > > > +} > > > > + > > > > /* > > > > * IP Tunneling transmitter > > > > * > > > > @@ -1025,6 +1069,7 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, > > > > struct iphdr *iph; /* Our new IP header */ > > > > unsigned int max_headroom; /* The extra header space needed */ > > > > int ret, local; > > > > + int tun_type, gso_type; > > > > > > > > EnterFunction(10); > > > > > > > > @@ -1046,6 +1091,11 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, > > > > */ > > > > max_headroom = LL_RESERVED_SPACE(tdev) + sizeof(struct iphdr); > > > > > > > > + tun_type = cp->dest->tun_type; > > > > + > > > > + if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + max_headroom += sizeof(struct udphdr) + sizeof(struct guehdr); > > > > + > > > > /* We only care about the df field if sysctl_pmtu_disc(ipvs) is set */ > > > > dfp = sysctl_pmtu_disc(ipvs) ? &df : NULL; > > > > skb = ip_vs_prepare_tunneled_skb(skb, cp->af, max_headroom, > > > > @@ -1054,11 +1104,20 @@ ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, > > > > if (IS_ERR(skb)) > > > > goto tx_error; > > > > > > > > - if (iptunnel_handle_offloads(skb, __tun_gso_type_mask(AF_INET, cp->af))) > > > > + gso_type = __tun_gso_type_mask(AF_INET, cp->af); > > > > + if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + gso_type |= SKB_GSO_UDP_TUNNEL; > > > > > > Should the gso_type really be __tun_gso_type_mask() | SKB_GSO_UDP_TUNNEL. > > > It seems tome that SKB_GSO_UDP_TUNNEL would be more appropriate. > > > > > Do you mean you want me to change the code above to: > > gso_type = __tun_gso_type_mask(AF_INET, cp->af); > > if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > gso_type = __tun_gso_type_mask(AF_INET, cp->af) | SKB_GSO_UDP_TUNNEL; > > I am more thinking of: > > if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > gso_type = SKB_GSO_UDP_TUNNEL; > else > gso_type = __tun_gso_type_mask(AF_INET, cp->af); > > I think you intentionally changed this since an earlier version > of this patch but I'm not sure why. > That's because of the following new commit introduced some new checks: https://github.com/torvalds/linux/commit/418e897e0716b238ea4252ed22a73ca37d3cbbc1#diff-9783279e0dd62b9f996300a8127ec964R1391 Which was pointed out by Julia in the review of v6 of the patch. > > > Also, should there be a mechanism to enable cums in the outer header, > > > and thus use SKB_GSO_UDP_TUNNEL_CSUM? > > > > > Yep, that we can definitely do it in a follow up patch later, by adding > > some new parameters for check sum control. > > Thanks, I think that would be a good follow-up. > > > > > + > > > > + if (iptunnel_handle_offloads(skb, gso_type)) > > > > goto tx_error; > > > > > > > > skb->transport_header = skb->network_header; > > > > > > > > + skb_set_inner_ipproto(skb, next_protocol); > > > > + > > > > + if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + ipvs_gue_encap(net, skb, cp, &next_protocol); > > > > + > > > > skb_push(skb, sizeof(struct iphdr)); > > > > skb_reset_network_header(skb); > > > > memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt)); > > > > @@ -1102,6 +1161,8 @@ int > > > > ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, > > > > struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh) > > > > { > > > > + struct netns_ipvs *ipvs = cp->ipvs; > > > > + struct net *net = ipvs->net; > > > > struct rt6_info *rt; /* Route to the other host */ > > > > struct in6_addr saddr; /* Source for tunnel */ > > > > struct net_device *tdev; /* Device to other host */ > > > > @@ -1112,10 +1173,11 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, > > > > struct ipv6hdr *iph; /* Our new IP header */ > > > > unsigned int max_headroom; /* The extra header space needed */ > > > > int ret, local; > > > > + int tun_type, gso_type; > > > > > > > > EnterFunction(10); > > > > > > > > - local = __ip_vs_get_out_rt_v6(cp->ipvs, cp->af, skb, cp->dest, > > > > + local = __ip_vs_get_out_rt_v6(ipvs, cp->af, skb, cp->dest, > > > > &cp->daddr.in6, > > > > &saddr, ipvsh, 1, > > > > IP_VS_RT_MODE_LOCAL | > > > > @@ -1134,17 +1196,31 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, > > > > */ > > > > max_headroom = LL_RESERVED_SPACE(tdev) + sizeof(struct ipv6hdr); > > > > > > > > + tun_type = cp->dest->tun_type; > > > > + > > > > + if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + max_headroom += sizeof(struct udphdr) + sizeof(struct guehdr); > > > > + > > > > skb = ip_vs_prepare_tunneled_skb(skb, cp->af, max_headroom, > > > > &next_protocol, &payload_len, > > > > &dsfield, &ttl, NULL); > > > > if (IS_ERR(skb)) > > > > goto tx_error; > > > > > > > > - if (iptunnel_handle_offloads(skb, __tun_gso_type_mask(AF_INET6, cp->af))) > > > > + gso_type = __tun_gso_type_mask(AF_INET6, cp->af); > > > > + if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + gso_type |= SKB_GSO_UDP_TUNNEL; > > > > + > > > > + if (iptunnel_handle_offloads(skb, gso_type)) > > > > goto tx_error; > > > > > > > > skb->transport_header = skb->network_header; > > > > > > > > + skb_set_inner_ipproto(skb, next_protocol); > > > > > > Is the above needed in the non-GUE case? > > Looking at implementation of ipip6_tunnel_xmit and ipip_tunnel_xmit, > > this seems to be needed also. > > > > > > > + > > > > + if (tun_type == IP_VS_CONN_F_TUNNEL_TYPE_GUE) > > > > + ipvs_gue_encap(net, skb, cp, &next_protocol); > > > > + > > > > skb_push(skb, sizeof(struct ipv6hdr)); > > > > skb_reset_network_header(skb); > > > > memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt)); > > > > @@ -1167,7 +1243,7 @@ ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, > > > > > > > > ret = ip_vs_tunnel_xmit_prepare(skb, cp); > > > > if (ret == NF_ACCEPT) > > > > - ip6_local_out(cp->ipvs->net, skb->sk, skb); > > > > + ip6_local_out(net, skb->sk, skb); > > > > else if (ret == NF_DROP) > > > > kfree_skb(skb); > > > > > > > > -- > > > > 2.21.0 > > > > > > Thanks and Best Regards, > > Jacky > >