Received: from mail.netfilter.org (mail.netfilter.org [217.70.190.124]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 3F4233A5E68 for ; Thu, 30 Apr 2026 20:25:54 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=217.70.190.124 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1777580756; cv=none; b=ZyceG3nyHEz+y5QQW7z+RACftJ+8RKW5cWxJGJTPDX6zYkUKWRzXSRdEW7Xuv8kyZtbZd211Kg7umQ1we1jTWAT4lRVqhkGgexGqNJ4CfDe092zHVrhmRbRt5802gQ8NpUa98Ne2cwtUW0KXgG++Y0ORnB3d8bwVO6JPW2H8exc= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1777580756; c=relaxed/simple; bh=hajEqtkiwmwGsxnvIqijxcJe9bz3B6WO/8JmrnsLx9o=; h=From:To:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=gO/6tqdZeEkwNIigxJVR72yr5uarChnEGPvA/ByRsoFXE9ShxgGOUhU812+olyVW4wI7Zw03MSlZvsJcpAjB6dv/HZVf15HxVKSoNxBJ3NOVZQFYT3ngb3lg2NlWLJBuiJAB2kx2tUCvTg+efpAyAqr360QSH53M7dPEsz4q/4M= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=netfilter.org; spf=pass smtp.mailfrom=netfilter.org; dkim=pass (2048-bit key) header.d=netfilter.org header.i=@netfilter.org header.b=Yx+7Uo4T; arc=none smtp.client-ip=217.70.190.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=netfilter.org Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=netfilter.org Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=netfilter.org header.i=@netfilter.org header.b="Yx+7Uo4T" Received: from localhost.localdomain (mail-agni [217.70.190.124]) by mail.netfilter.org (Postfix) with ESMTPSA id 4F65260181 for ; Thu, 30 Apr 2026 22:25:52 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=netfilter.org; s=2025; t=1777580752; bh=Ub4edRqRByLPOLIXcNYyFw/lQPHmk2pXs3IlT8bHcdA=; h=From:To:Subject:Date:In-Reply-To:References:From; b=Yx+7Uo4TmFS7aM/qmJHU1t7fVk5LvFoLSnUmWlATWgfSZ3yw7iL2TkpJXXDkE5CNA 7Vtjf6xNfILxT+Avp7zITRWnS5V9S/lRXdNh9/UNXWXgrR2NIrcZrf4qb5PRs7jjZi lX1L/Y0OYQKiGiIMFfA9m3EwwPa9c72qJFKEMUv743ceTzhp1rVCI7+eK5/RrUcRaP QQcS3AfWzWmBcTlCwKx0bydYIjgEGyVJWzZH8Bh7G9XlwRHVV1D2/b3gfu6U1d0P96 0qwDdPM5eEeFLgMCJU14/6MIJnhb7u/mUSnmVEW0Be3P/0E2R5o+HHa5F0UcRxLUgw gsCvsNOTgjnbg== From: Pablo Neira Ayuso To: netfilter-devel@vger.kernel.org Subject: [PATCH nf,v6 2/4] netfilter: flowtable: fix inline vlan encapsulation in xmit path Date: Thu, 30 Apr 2026 22:25:45 +0200 Message-ID: <20260430202547.274256-2-pablo@netfilter.org> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20260430202547.274256-1-pablo@netfilter.org> References: <20260430202547.274256-1-pablo@netfilter.org> Precedence: bulk X-Mailing-List: netfilter-devel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Several issues in the inline vlan support: - The layer 2 encapsulation representation in the tuple takes encap[0] as the outer header and encap[1] as the inner header as seen from the ingress path. Reverse the encap loop to push first the inner then the outer vlan header. - Postpone pushing the layer 2 header once destination device is known. This allows to calculate the needed hearoom via LL_RESERVED_SPACE to accommodate the layer 2 headers. - Add and use nf_flow_vlan_push() as suggested by Eric Woudstra, this is a simplified version of skb_vlan_push() for egress path only. - skb_postpush_rcsum() Fixes: c653d5a78f34 ("netfilter: flowtable: inline vlan encapsulation in xmit path") Signed-off-by: Pablo Neira Ayuso --- v6: add skb_postpush_rcsum() as suggested by AI reviewer. net/netfilter/nf_flow_table_ip.c | 110 ++++++++++++++++++++----------- 1 file changed, 73 insertions(+), 37 deletions(-) diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index 8d5fb7e940a1..0ce3c209050c 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -462,32 +462,6 @@ static void nf_flow_encap_pop(struct nf_flowtable_ctx *ctx, nf_flow_ip_tunnel_pop(ctx, skb); } -struct nf_flow_xmit { - const void *dest; - const void *source; - struct net_device *outdev; -}; - -static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, - struct nf_flow_xmit *xmit) -{ - struct net_device *dev = xmit->outdev; - unsigned int hh_len = LL_RESERVED_SPACE(dev); - - if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { - skb = skb_expand_head(skb, hh_len); - if (!skb) - return NF_STOLEN; - } - - skb->dev = dev; - dev_hard_header(skb, dev, ntohs(skb->protocol), - xmit->dest, xmit->source, skb->len); - dev_queue_xmit(skb); - - return NF_STOLEN; -} - static struct flow_offload_tuple_rhash * nf_flow_offload_lookup(struct nf_flowtable_ctx *ctx, struct nf_flowtable *flow_table, struct sk_buff *skb) @@ -553,6 +527,32 @@ static int nf_flow_offload_forward(struct nf_flowtable_ctx *ctx, return 1; } +/* Similar to skb_vlan_push. */ +static int nf_flow_vlan_push(struct sk_buff *skb, __be16 proto, u16 id, + u32 needed_headroom) +{ + if (skb_vlan_tag_present(skb)) { + struct vlan_hdr *vhdr; + + if (skb_cow_head(skb, needed_headroom + VLAN_HLEN)) + return -1; + + __skb_push(skb, VLAN_HLEN); + if (skb_mac_header_was_set(skb)) + skb->mac_header -= VLAN_HLEN; + + vhdr = (struct vlan_hdr *)skb->data; + skb->network_header -= VLAN_HLEN; + vhdr->h_vlan_TCI = htons(skb_vlan_tag_get(skb)); + vhdr->h_vlan_encapsulated_proto = skb->protocol; + skb->protocol = skb->vlan_proto; + skb_postpush_rcsum(skb, skb->data, VLAN_HLEN); + } + __vlan_hwaccel_put_tag(skb, proto, id); + + return 0; +} + static int nf_flow_pppoe_push(struct sk_buff *skb, u16 id) { int data_len = skb->len + sizeof(__be16); @@ -739,17 +739,19 @@ static int nf_flow_tunnel_v6_push(struct net *net, struct sk_buff *skb, } static int nf_flow_encap_push(struct sk_buff *skb, - struct flow_offload_tuple *tuple) + struct flow_offload_tuple *tuple, + struct net_device *outdev) { + u32 needed_headroom = LL_RESERVED_SPACE(outdev); int i; - for (i = 0; i < tuple->encap_num; i++) { + for (i = tuple->encap_num - 1; i >= 0; i--) { switch (tuple->encap[i].proto) { case htons(ETH_P_8021Q): case htons(ETH_P_8021AD): - skb_reset_mac_header(skb); - if (skb_vlan_push(skb, tuple->encap[i].proto, - tuple->encap[i].id) < 0) + if (nf_flow_vlan_push(skb, tuple->encap[i].proto, + tuple->encap[i].id, + needed_headroom) < 0) return -1; break; case htons(ETH_P_PPP_SES): @@ -762,6 +764,44 @@ static int nf_flow_encap_push(struct sk_buff *skb, return 0; } +struct nf_flow_xmit { + const void *dest; + const void *source; + struct net_device *outdev; + struct flow_offload_tuple *tuple; +}; + +static void __nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, + struct nf_flow_xmit *xmit) +{ + struct net_device *dev = xmit->outdev; + unsigned int hh_len = LL_RESERVED_SPACE(dev); + + if (unlikely(skb_headroom(skb) < hh_len && dev->header_ops)) { + skb = skb_expand_head(skb, hh_len); + if (!skb) + return; + } + + skb->dev = dev; + dev_hard_header(skb, dev, ntohs(skb->protocol), + xmit->dest, xmit->source, skb->len); + dev_queue_xmit(skb); +} + +static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, + struct nf_flow_xmit *xmit) +{ + if (xmit->tuple->encap_num) { + if (nf_flow_encap_push(skb, xmit->tuple, xmit->outdev) < 0) + return NF_DROP; + } + + __nf_flow_queue_xmit(net, skb, xmit); + + return NF_STOLEN; +} + unsigned int nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, const struct nf_hook_state *state) @@ -806,9 +846,6 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, if (nf_flow_tunnel_v4_push(state->net, skb, other_tuple, &ip_daddr) < 0) return NF_DROP; - if (nf_flow_encap_push(skb, other_tuple) < 0) - return NF_DROP; - switch (tuplehash->tuple.xmit_type) { case FLOW_OFFLOAD_XMIT_NEIGH: rt = dst_rtable(tuplehash->tuple.dst_cache); @@ -838,6 +875,7 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, WARN_ON_ONCE(1); return NF_DROP; } + xmit.tuple = other_tuple; return nf_flow_queue_xmit(state->net, skb, &xmit); } @@ -1128,9 +1166,6 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, &ip6_daddr, encap_limit) < 0) return NF_DROP; - if (nf_flow_encap_push(skb, other_tuple) < 0) - return NF_DROP; - switch (tuplehash->tuple.xmit_type) { case FLOW_OFFLOAD_XMIT_NEIGH: rt = dst_rt6_info(tuplehash->tuple.dst_cache); @@ -1160,6 +1195,7 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, WARN_ON_ONCE(1); return NF_DROP; } + xmit.tuple = other_tuple; return nf_flow_queue_xmit(state->net, skb, &xmit); } -- 2.47.3