From patchwork Wed Aug 17 13:59:59 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945915 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 97E5EC28B2B for ; Wed, 17 Aug 2022 14:01:15 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239858AbiHQOBN (ORCPT ); Wed, 17 Aug 2022 10:01:13 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48274 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239863AbiHQOA4 (ORCPT ); Wed, 17 Aug 2022 10:00:56 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id F0F979858F; Wed, 17 Aug 2022 07:00:55 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbK-0008O2-Cp; Wed, 17 Aug 2022 16:00:54 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 01/17] netfilter: nf_tables: use READ_ONCE and WRITE_ONCE for shared generation id access Date: Wed, 17 Aug 2022 15:59:59 +0200 Message-Id: <20220817140015.25843-2-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso The generation ID is bumped from the commit path while holding the mutex, however, netlink dump operations rely on RCU. This patch also adds missing cb->base_eq initialization in nf_tables_dump_set(). Fixes: 38e029f14a97 ("netfilter: nf_tables: set NLM_F_DUMP_INTR if netlink dumping is stale") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 3cc88998b879..8b084cd669ab 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -889,7 +889,7 @@ static int nf_tables_dump_tables(struct sk_buff *skb, rcu_read_lock(); nft_net = nft_pernet(net); - cb->seq = nft_net->base_seq; + cb->seq = READ_ONCE(nft_net->base_seq); list_for_each_entry_rcu(table, &nft_net->tables, list) { if (family != NFPROTO_UNSPEC && family != table->family) @@ -1705,7 +1705,7 @@ static int nf_tables_dump_chains(struct sk_buff *skb, rcu_read_lock(); nft_net = nft_pernet(net); - cb->seq = nft_net->base_seq; + cb->seq = READ_ONCE(nft_net->base_seq); list_for_each_entry_rcu(table, &nft_net->tables, list) { if (family != NFPROTO_UNSPEC && family != table->family) @@ -3149,7 +3149,7 @@ static int nf_tables_dump_rules(struct sk_buff *skb, rcu_read_lock(); nft_net = nft_pernet(net); - cb->seq = nft_net->base_seq; + cb->seq = READ_ONCE(nft_net->base_seq); list_for_each_entry_rcu(table, &nft_net->tables, list) { if (family != NFPROTO_UNSPEC && family != table->family) @@ -4133,7 +4133,7 @@ static int nf_tables_dump_sets(struct sk_buff *skb, struct netlink_callback *cb) rcu_read_lock(); nft_net = nft_pernet(net); - cb->seq = nft_net->base_seq; + cb->seq = READ_ONCE(nft_net->base_seq); list_for_each_entry_rcu(table, &nft_net->tables, list) { if (ctx->family != NFPROTO_UNSPEC && @@ -5061,6 +5061,8 @@ static int nf_tables_dump_set(struct sk_buff *skb, struct netlink_callback *cb) rcu_read_lock(); nft_net = nft_pernet(net); + cb->seq = READ_ONCE(nft_net->base_seq); + list_for_each_entry_rcu(table, &nft_net->tables, list) { if (dump_ctx->ctx.family != NFPROTO_UNSPEC && dump_ctx->ctx.family != table->family) @@ -6941,7 +6943,7 @@ static int nf_tables_dump_obj(struct sk_buff *skb, struct netlink_callback *cb) rcu_read_lock(); nft_net = nft_pernet(net); - cb->seq = nft_net->base_seq; + cb->seq = READ_ONCE(nft_net->base_seq); list_for_each_entry_rcu(table, &nft_net->tables, list) { if (family != NFPROTO_UNSPEC && family != table->family) @@ -7873,7 +7875,7 @@ static int nf_tables_dump_flowtable(struct sk_buff *skb, rcu_read_lock(); nft_net = nft_pernet(net); - cb->seq = nft_net->base_seq; + cb->seq = READ_ONCE(nft_net->base_seq); list_for_each_entry_rcu(table, &nft_net->tables, list) { if (family != NFPROTO_UNSPEC && family != table->family) @@ -8806,6 +8808,7 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb) struct nft_trans_elem *te; struct nft_chain *chain; struct nft_table *table; + unsigned int base_seq; LIST_HEAD(adl); int err; @@ -8855,9 +8858,12 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb) * Bump generation counter, invalidate any dump in progress. * Cannot fail after this point. */ - while (++nft_net->base_seq == 0) + base_seq = READ_ONCE(nft_net->base_seq); + while (++base_seq == 0) ; + WRITE_ONCE(nft_net->base_seq, base_seq); + /* step 3. Start new generation, rules_gen_X now in use. */ net->nft.gencursor = nft_gencursor_next(net); From patchwork Wed Aug 17 14:00:00 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945916 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5520CC32771 for ; Wed, 17 Aug 2022 14:01:30 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239969AbiHQOB2 (ORCPT ); Wed, 17 Aug 2022 10:01:28 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52060 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239896AbiHQOBC (ORCPT ); Wed, 17 Aug 2022 10:01:02 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4AF119750F; Wed, 17 Aug 2022 07:01:00 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbO-0008OQ-HF; Wed, 17 Aug 2022 16:00:58 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 02/17] netfilter: nf_tables: disallow NFTA_SET_ELEM_KEY_END with NFT_SET_ELEM_INTERVAL_END flag Date: Wed, 17 Aug 2022 16:00:00 +0200 Message-Id: <20220817140015.25843-3-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso These are mutually exclusive, actually NFTA_SET_ELEM_KEY_END replaces the flag notation. Fixes: 7b225d0b5c6d ("netfilter: nf_tables: add NFTA_SET_ELEM_KEY_END attribute") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 8b084cd669ab..ac549c5b88c2 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -5901,6 +5901,7 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, nla[NFTA_SET_ELEM_EXPIRATION] || nla[NFTA_SET_ELEM_USERDATA] || nla[NFTA_SET_ELEM_EXPR] || + nla[NFTA_SET_ELEM_KEY_END] || nla[NFTA_SET_ELEM_EXPRESSIONS])) return -EINVAL; From patchwork Wed Aug 17 14:00:01 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945917 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 4480CC32771 for ; Wed, 17 Aug 2022 14:01:37 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237108AbiHQOBf (ORCPT ); Wed, 17 Aug 2022 10:01:35 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47718 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239908AbiHQOBF (ORCPT ); Wed, 17 Aug 2022 10:01:05 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 2179991082; Wed, 17 Aug 2022 07:01:04 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbS-0008Oh-L4; Wed, 17 Aug 2022 16:01:02 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 03/17] netfilter: nf_tables: possible module reference underflow in error path Date: Wed, 17 Aug 2022 16:00:01 +0200 Message-Id: <20220817140015.25843-4-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso dst->ops is set on when nft_expr_clone() fails, but module refcount has not been bumped yet, therefore nft_expr_destroy() leads to module reference underflow. Fixes: 8cfd9b0f8515 ("netfilter: nftables: generalize set expressions support") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index ac549c5b88c2..989c9782ecc3 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -5601,7 +5601,7 @@ int nft_set_elem_expr_clone(const struct nft_ctx *ctx, struct nft_set *set, err = nft_expr_clone(expr, set->exprs[i]); if (err < 0) { - nft_expr_destroy(ctx, expr); + kfree(expr); goto err_expr; } expr_array[i] = expr; From patchwork Wed Aug 17 14:00:02 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945926 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9D2EBC32771 for ; Wed, 17 Aug 2022 14:01:43 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239977AbiHQOBm (ORCPT ); Wed, 17 Aug 2022 10:01:42 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47680 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239830AbiHQOBM (ORCPT ); Wed, 17 Aug 2022 10:01:12 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 8556799265; Wed, 17 Aug 2022 07:01:08 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbW-0008P3-Pi; Wed, 17 Aug 2022 16:01:06 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal Subject: [PATCH net 04/17] netfilter: nf_ct_sane: remove pseudo skb linearization Date: Wed, 17 Aug 2022 16:00:02 +0200 Message-Id: <20220817140015.25843-5-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org For historical reason this code performs pseudo linearization of skbs via skb_header_pointer and a global 64k buffer. With arrival of BIG TCP, packets generated by TCP stack can exceed 64kb. Rewrite this to only extract the needed header data. This also allows to get rid of the locking. Fixes: 7c4e983c4f3c ("net: allow gso_max_size to exceed 65536") Fixes: 0fe79f28bfaf ("net: allow gro_max_size to exceed 65536") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_sane.c | 68 ++++++++++++++----------------- 1 file changed, 30 insertions(+), 38 deletions(-) diff --git a/net/netfilter/nf_conntrack_sane.c b/net/netfilter/nf_conntrack_sane.c index fcb33b1d5456..13dc421fc4f5 100644 --- a/net/netfilter/nf_conntrack_sane.c +++ b/net/netfilter/nf_conntrack_sane.c @@ -34,10 +34,6 @@ MODULE_AUTHOR("Michal Schmidt "); MODULE_DESCRIPTION("SANE connection tracking helper"); MODULE_ALIAS_NFCT_HELPER(HELPER_NAME); -static char *sane_buffer; - -static DEFINE_SPINLOCK(nf_sane_lock); - #define MAX_PORTS 8 static u_int16_t ports[MAX_PORTS]; static unsigned int ports_c; @@ -67,14 +63,16 @@ static int help(struct sk_buff *skb, unsigned int dataoff, datalen; const struct tcphdr *th; struct tcphdr _tcph; - void *sb_ptr; int ret = NF_ACCEPT; int dir = CTINFO2DIR(ctinfo); struct nf_ct_sane_master *ct_sane_info = nfct_help_data(ct); struct nf_conntrack_expect *exp; struct nf_conntrack_tuple *tuple; - struct sane_request *req; struct sane_reply_net_start *reply; + union { + struct sane_request req; + struct sane_reply_net_start repl; + } buf; /* Until there's been traffic both ways, don't look in packets. */ if (ctinfo != IP_CT_ESTABLISHED && @@ -92,59 +90,62 @@ static int help(struct sk_buff *skb, return NF_ACCEPT; datalen = skb->len - dataoff; - - spin_lock_bh(&nf_sane_lock); - sb_ptr = skb_header_pointer(skb, dataoff, datalen, sane_buffer); - if (!sb_ptr) { - spin_unlock_bh(&nf_sane_lock); - return NF_ACCEPT; - } - if (dir == IP_CT_DIR_ORIGINAL) { + const struct sane_request *req; + if (datalen != sizeof(struct sane_request)) - goto out; + return NF_ACCEPT; + + req = skb_header_pointer(skb, dataoff, datalen, &buf.req); + if (!req) + return NF_ACCEPT; - req = sb_ptr; if (req->RPC_code != htonl(SANE_NET_START)) { /* Not an interesting command */ - ct_sane_info->state = SANE_STATE_NORMAL; - goto out; + WRITE_ONCE(ct_sane_info->state, SANE_STATE_NORMAL); + return NF_ACCEPT; } /* We're interested in the next reply */ - ct_sane_info->state = SANE_STATE_START_REQUESTED; - goto out; + WRITE_ONCE(ct_sane_info->state, SANE_STATE_START_REQUESTED); + return NF_ACCEPT; } + /* IP_CT_DIR_REPLY */ + /* Is it a reply to an uninteresting command? */ - if (ct_sane_info->state != SANE_STATE_START_REQUESTED) - goto out; + if (READ_ONCE(ct_sane_info->state) != SANE_STATE_START_REQUESTED) + return NF_ACCEPT; /* It's a reply to SANE_NET_START. */ - ct_sane_info->state = SANE_STATE_NORMAL; + WRITE_ONCE(ct_sane_info->state, SANE_STATE_NORMAL); if (datalen < sizeof(struct sane_reply_net_start)) { pr_debug("NET_START reply too short\n"); - goto out; + return NF_ACCEPT; } - reply = sb_ptr; + datalen = sizeof(struct sane_reply_net_start); + + reply = skb_header_pointer(skb, dataoff, datalen, &buf.repl); + if (!reply) + return NF_ACCEPT; + if (reply->status != htonl(SANE_STATUS_SUCCESS)) { /* saned refused the command */ pr_debug("unsuccessful SANE_STATUS = %u\n", ntohl(reply->status)); - goto out; + return NF_ACCEPT; } /* Invalid saned reply? Ignore it. */ if (reply->zero != 0) - goto out; + return NF_ACCEPT; exp = nf_ct_expect_alloc(ct); if (exp == NULL) { nf_ct_helper_log(skb, ct, "cannot alloc expectation"); - ret = NF_DROP; - goto out; + return NF_DROP; } tuple = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple; @@ -162,9 +163,6 @@ static int help(struct sk_buff *skb, } nf_ct_expect_put(exp); - -out: - spin_unlock_bh(&nf_sane_lock); return ret; } @@ -178,7 +176,6 @@ static const struct nf_conntrack_expect_policy sane_exp_policy = { static void __exit nf_conntrack_sane_fini(void) { nf_conntrack_helpers_unregister(sane, ports_c * 2); - kfree(sane_buffer); } static int __init nf_conntrack_sane_init(void) @@ -187,10 +184,6 @@ static int __init nf_conntrack_sane_init(void) NF_CT_HELPER_BUILD_BUG_ON(sizeof(struct nf_ct_sane_master)); - sane_buffer = kmalloc(65536, GFP_KERNEL); - if (!sane_buffer) - return -ENOMEM; - if (ports_c == 0) ports[ports_c++] = SANE_PORT; @@ -210,7 +203,6 @@ static int __init nf_conntrack_sane_init(void) ret = nf_conntrack_helpers_register(sane, ports_c * 2); if (ret < 0) { pr_err("failed to register helpers\n"); - kfree(sane_buffer); return ret; } From patchwork Wed Aug 17 14:00:03 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945927 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 0C145C32771 for ; Wed, 17 Aug 2022 14:01:55 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239964AbiHQOBx (ORCPT ); Wed, 17 Aug 2022 10:01:53 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52568 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239952AbiHQOBR (ORCPT ); Wed, 17 Aug 2022 10:01:17 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 0737B97D44; Wed, 17 Aug 2022 07:01:12 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJba-0008PI-Tk; Wed, 17 Aug 2022 16:01:10 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal Subject: [PATCH net 05/17] netfilter: nf_ct_h323: cap packet size at 64k Date: Wed, 17 Aug 2022 16:00:03 +0200 Message-Id: <20220817140015.25843-6-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org With BIG TCP, packets generated by tcp stack may exceed 64kb. Cap datalen at 64kb. The internal message format uses 16bit fields, so no embedded message can exceed 64k size. Multiple h323 messages in a single superpacket may now result in a message to get treated as incomplete/truncated, but thats better than scribbling past h323_buffer. Another alternative suitable for net tree would be a switch to skb_linearize(). Fixes: 7c4e983c4f3c ("net: allow gso_max_size to exceed 65536") Fixes: 0fe79f28bfaf ("net: allow gro_max_size to exceed 65536") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_h323_main.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nf_conntrack_h323_main.c b/net/netfilter/nf_conntrack_h323_main.c index bb76305bb7ff..5a9bce24f3c3 100644 --- a/net/netfilter/nf_conntrack_h323_main.c +++ b/net/netfilter/nf_conntrack_h323_main.c @@ -34,6 +34,8 @@ #include #include +#define H323_MAX_SIZE 65535 + /* Parameters */ static unsigned int default_rrq_ttl __read_mostly = 300; module_param(default_rrq_ttl, uint, 0600); @@ -86,6 +88,9 @@ static int get_tpkt_data(struct sk_buff *skb, unsigned int protoff, if (tcpdatalen <= 0) /* No TCP data */ goto clear_out; + if (tcpdatalen > H323_MAX_SIZE) + tcpdatalen = H323_MAX_SIZE; + if (*data == NULL) { /* first TPKT */ /* Get first TPKT pointer */ tpkt = skb_header_pointer(skb, tcpdataoff, tcpdatalen, @@ -1169,6 +1174,9 @@ static unsigned char *get_udp_data(struct sk_buff *skb, unsigned int protoff, if (dataoff >= skb->len) return NULL; *datalen = skb->len - dataoff; + if (*datalen > H323_MAX_SIZE) + *datalen = H323_MAX_SIZE; + return skb_header_pointer(skb, dataoff, *datalen, h323_buffer); } @@ -1770,7 +1778,7 @@ static int __init nf_conntrack_h323_init(void) NF_CT_HELPER_BUILD_BUG_ON(sizeof(struct nf_ct_h323_master)); - h323_buffer = kmalloc(65536, GFP_KERNEL); + h323_buffer = kmalloc(H323_MAX_SIZE + 1, GFP_KERNEL); if (!h323_buffer) return -ENOMEM; ret = h323_helper_init(); From patchwork Wed Aug 17 14:00:04 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945928 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1D0B4C28B2B for ; Wed, 17 Aug 2022 14:02:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236051AbiHQOCI (ORCPT ); Wed, 17 Aug 2022 10:02:08 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48926 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239974AbiHQOBa (ORCPT ); Wed, 17 Aug 2022 10:01:30 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id BF1E598D2E; Wed, 17 Aug 2022 07:01:16 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbf-0008Pi-29; Wed, 17 Aug 2022 16:01:15 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal Subject: [PATCH net 06/17] netfilter: nf_ct_ftp: prefer skb_linearize Date: Wed, 17 Aug 2022 16:00:04 +0200 Message-Id: <20220817140015.25843-7-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org This uses a pseudo-linearization scheme with a 64k global buffer, but BIG TCP arrival means IPv6 TCP stack can generate skbs that exceed this size. Use skb_linearize. It should be possible to rewrite this to properly deal with segmented skbs (i.e., only do small chunk-wise accesses), but this is going to be a lot more intrusive than this because every helper function needs to get the sk_buff instead of a pointer to a raw data buffer. In practice, provided we're really looking at FTP control channel packets, there should never be a case where we deal with huge packets. Fixes: 7c4e983c4f3c ("net: allow gso_max_size to exceed 65536") Fixes: 0fe79f28bfaf ("net: allow gro_max_size to exceed 65536") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_ftp.c | 24 ++++++------------------ 1 file changed, 6 insertions(+), 18 deletions(-) diff --git a/net/netfilter/nf_conntrack_ftp.c b/net/netfilter/nf_conntrack_ftp.c index a414274338cf..0d9332e9cf71 100644 --- a/net/netfilter/nf_conntrack_ftp.c +++ b/net/netfilter/nf_conntrack_ftp.c @@ -34,11 +34,6 @@ MODULE_DESCRIPTION("ftp connection tracking helper"); MODULE_ALIAS("ip_conntrack_ftp"); MODULE_ALIAS_NFCT_HELPER(HELPER_NAME); -/* This is slow, but it's simple. --RR */ -static char *ftp_buffer; - -static DEFINE_SPINLOCK(nf_ftp_lock); - #define MAX_PORTS 8 static u_int16_t ports[MAX_PORTS]; static unsigned int ports_c; @@ -398,6 +393,9 @@ static int help(struct sk_buff *skb, return NF_ACCEPT; } + if (unlikely(skb_linearize(skb))) + return NF_DROP; + th = skb_header_pointer(skb, protoff, sizeof(_tcph), &_tcph); if (th == NULL) return NF_ACCEPT; @@ -411,12 +409,8 @@ static int help(struct sk_buff *skb, } datalen = skb->len - dataoff; - spin_lock_bh(&nf_ftp_lock); - fb_ptr = skb_header_pointer(skb, dataoff, datalen, ftp_buffer); - if (!fb_ptr) { - spin_unlock_bh(&nf_ftp_lock); - return NF_ACCEPT; - } + spin_lock_bh(&ct->lock); + fb_ptr = skb->data + dataoff; ends_in_nl = (fb_ptr[datalen - 1] == '\n'); seq = ntohl(th->seq) + datalen; @@ -544,7 +538,7 @@ static int help(struct sk_buff *skb, if (ends_in_nl) update_nl_seq(ct, seq, ct_ftp_info, dir, skb); out: - spin_unlock_bh(&nf_ftp_lock); + spin_unlock_bh(&ct->lock); return ret; } @@ -571,7 +565,6 @@ static const struct nf_conntrack_expect_policy ftp_exp_policy = { static void __exit nf_conntrack_ftp_fini(void) { nf_conntrack_helpers_unregister(ftp, ports_c * 2); - kfree(ftp_buffer); } static int __init nf_conntrack_ftp_init(void) @@ -580,10 +573,6 @@ static int __init nf_conntrack_ftp_init(void) NF_CT_HELPER_BUILD_BUG_ON(sizeof(struct nf_ct_ftp_master)); - ftp_buffer = kmalloc(65536, GFP_KERNEL); - if (!ftp_buffer) - return -ENOMEM; - if (ports_c == 0) ports[ports_c++] = FTP_PORT; @@ -603,7 +592,6 @@ static int __init nf_conntrack_ftp_init(void) ret = nf_conntrack_helpers_register(ftp, ports_c * 2); if (ret < 0) { pr_err("failed to register helpers\n"); - kfree(ftp_buffer); return ret; } From patchwork Wed Aug 17 14:00:05 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945929 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 28BADC28B2B for ; Wed, 17 Aug 2022 14:02:18 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236340AbiHQOCP (ORCPT ); Wed, 17 Aug 2022 10:02:15 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:49032 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236372AbiHQOBf (ORCPT ); Wed, 17 Aug 2022 10:01:35 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DE00799270; Wed, 17 Aug 2022 07:01:20 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbj-0008Q1-6m; Wed, 17 Aug 2022 16:01:19 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal Subject: [PATCH net 07/17] netfilter: nf_ct_irc: cap packet search space to 4k Date: Wed, 17 Aug 2022 16:00:05 +0200 Message-Id: <20220817140015.25843-8-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org This uses a pseudo-linearization scheme with a 64k global buffer, but BIG TCP arrival means IPv6 TCP stack can generate skbs that exceed this size. In practice, IRC commands are not expected to exceed 512 bytes, plus this is interactive protocol, so we should not see large packets in practice. Given most IRC connections nowadays use TLS so this helper could also be removed in the near future. Fixes: 7c4e983c4f3c ("net: allow gso_max_size to exceed 65536") Fixes: 0fe79f28bfaf ("net: allow gro_max_size to exceed 65536") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_conntrack_irc.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nf_conntrack_irc.c b/net/netfilter/nf_conntrack_irc.c index 08ee4e760a3d..1796c456ac98 100644 --- a/net/netfilter/nf_conntrack_irc.c +++ b/net/netfilter/nf_conntrack_irc.c @@ -39,6 +39,7 @@ unsigned int (*nf_nat_irc_hook)(struct sk_buff *skb, EXPORT_SYMBOL_GPL(nf_nat_irc_hook); #define HELPER_NAME "irc" +#define MAX_SEARCH_SIZE 4095 MODULE_AUTHOR("Harald Welte "); MODULE_DESCRIPTION("IRC (DCC) connection tracking helper"); @@ -121,6 +122,7 @@ static int help(struct sk_buff *skb, unsigned int protoff, int i, ret = NF_ACCEPT; char *addr_beg_p, *addr_end_p; typeof(nf_nat_irc_hook) nf_nat_irc; + unsigned int datalen; /* If packet is coming from IRC server */ if (dir == IP_CT_DIR_REPLY) @@ -140,8 +142,12 @@ static int help(struct sk_buff *skb, unsigned int protoff, if (dataoff >= skb->len) return NF_ACCEPT; + datalen = skb->len - dataoff; + if (datalen > MAX_SEARCH_SIZE) + datalen = MAX_SEARCH_SIZE; + spin_lock_bh(&irc_buffer_lock); - ib_ptr = skb_header_pointer(skb, dataoff, skb->len - dataoff, + ib_ptr = skb_header_pointer(skb, dataoff, datalen, irc_buffer); if (!ib_ptr) { spin_unlock_bh(&irc_buffer_lock); @@ -149,7 +155,7 @@ static int help(struct sk_buff *skb, unsigned int protoff, } data = ib_ptr; - data_limit = ib_ptr + skb->len - dataoff; + data_limit = ib_ptr + datalen; /* strlen("\1DCC SENT t AAAAAAAA P\1\n")=24 * 5+MINMATCHLEN+strlen("t AAAAAAAA P\1\n")=14 */ @@ -251,7 +257,7 @@ static int __init nf_conntrack_irc_init(void) irc_exp_policy.max_expected = max_dcc_channels; irc_exp_policy.timeout = dcc_timeout; - irc_buffer = kmalloc(65536, GFP_KERNEL); + irc_buffer = kmalloc(MAX_SEARCH_SIZE + 1, GFP_KERNEL); if (!irc_buffer) return -ENOMEM; From patchwork Wed Aug 17 14:00:06 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945930 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id D809EC28B2B for ; Wed, 17 Aug 2022 14:02:25 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239908AbiHQOCY (ORCPT ); Wed, 17 Aug 2022 10:02:24 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52386 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239852AbiHQOBm (ORCPT ); Wed, 17 Aug 2022 10:01:42 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 278209A948; Wed, 17 Aug 2022 07:01:25 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbn-0008QK-AE; Wed, 17 Aug 2022 16:01:23 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal Subject: [PATCH net 08/17] netfilter: nf_tables: fix scheduling-while-atomic splat Date: Wed, 17 Aug 2022 16:00:06 +0200 Message-Id: <20220817140015.25843-9-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org nf_tables_check_loops() can be called from rhashtable list walk so cond_resched() cannot be used here. Fixes: 81ea01066741 ("netfilter: nf_tables: add rescheduling points during loop detection walks") Signed-off-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 989c9782ecc3..d90434eadc1b 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -9426,13 +9426,9 @@ static int nf_tables_check_loops(const struct nft_ctx *ctx, break; } } - - cond_resched(); } list_for_each_entry(set, &ctx->table->sets, list) { - cond_resched(); - if (!nft_is_active_next(ctx->net, set)) continue; if (!(set->flags & NFT_SET_MAP) || From patchwork Wed Aug 17 14:00:07 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945931 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id B5D80C32771 for ; Wed, 17 Aug 2022 14:02:39 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239952AbiHQOCh (ORCPT ); Wed, 17 Aug 2022 10:02:37 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52586 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239954AbiHQOBx (ORCPT ); Wed, 17 Aug 2022 10:01:53 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id E64A09A973; Wed, 17 Aug 2022 07:01:29 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbr-0008Qh-H8; Wed, 17 Aug 2022 16:01:27 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal , Yi Chen Subject: [PATCH net 09/17] netfilter: nfnetlink: re-enable conntrack expectation events Date: Wed, 17 Aug 2022 16:00:07 +0200 Message-Id: <20220817140015.25843-10-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org To avoid allocation of the conntrack extension area when possible, the default behaviour was changed to only allocate the event extension if a userspace program is subscribed to a notification group. Problem is that while 'conntrack -E' does enable the event allocation behind the scenes, 'conntrack -E expect' does not: no expectation events are delivered unless user sets "net.netfilter.nf_conntrack_events" back to 1 (always on). Fix the autodetection to also consider EXP type group. We need to track the 6 event groups (3+3, new/update/destroy for events and for expectations each) independently, else we'd disable events again if an expectation group becomes empty while there is still an active event group. Fixes: 2794cdb0b97b ("netfilter: nfnetlink: allow to detect if ctnetlink listeners exist") Reported-by: Yi Chen Signed-off-by: Florian Westphal --- include/net/netns/conntrack.h | 2 +- net/netfilter/nfnetlink.c | 83 ++++++++++++++++++++++++++++++----- 2 files changed, 72 insertions(+), 13 deletions(-) diff --git a/include/net/netns/conntrack.h b/include/net/netns/conntrack.h index 0677cd3de034..c396a3862e80 100644 --- a/include/net/netns/conntrack.h +++ b/include/net/netns/conntrack.h @@ -95,7 +95,7 @@ struct nf_ip_net { struct netns_ct { #ifdef CONFIG_NF_CONNTRACK_EVENTS - bool ctnetlink_has_listener; + u8 ctnetlink_has_listener; bool ecache_dwork_pending; #endif u8 sysctl_log_invalid; /* Log invalid packets */ diff --git a/net/netfilter/nfnetlink.c b/net/netfilter/nfnetlink.c index c24b1240908f..9c44518cb70f 100644 --- a/net/netfilter/nfnetlink.c +++ b/net/netfilter/nfnetlink.c @@ -44,6 +44,10 @@ MODULE_DESCRIPTION("Netfilter messages via netlink socket"); static unsigned int nfnetlink_pernet_id __read_mostly; +#ifdef CONFIG_NF_CONNTRACK_EVENTS +static DEFINE_SPINLOCK(nfnl_grp_active_lock); +#endif + struct nfnl_net { struct sock *nfnl; }; @@ -654,6 +658,44 @@ static void nfnetlink_rcv(struct sk_buff *skb) netlink_rcv_skb(skb, nfnetlink_rcv_msg); } +static void nfnetlink_bind_event(struct net *net, unsigned int group) +{ +#ifdef CONFIG_NF_CONNTRACK_EVENTS + int type, group_bit; + u8 v; + + /* All NFNLGRP_CONNTRACK_* group bits fit into u8. + * The other groups are not relevant and can be ignored. + */ + if (group >= 8) + return; + + type = nfnl_group2type[group]; + + switch (type) { + case NFNL_SUBSYS_CTNETLINK: + break; + case NFNL_SUBSYS_CTNETLINK_EXP: + break; + default: + return; + } + + group_bit = (1 << group); + + spin_lock(&nfnl_grp_active_lock); + v = READ_ONCE(net->ct.ctnetlink_has_listener); + if ((v & group_bit) == 0) { + v |= group_bit; + + /* read concurrently without nfnl_grp_active_lock held. */ + WRITE_ONCE(net->ct.ctnetlink_has_listener, v); + } + + spin_unlock(&nfnl_grp_active_lock); +#endif +} + static int nfnetlink_bind(struct net *net, int group) { const struct nfnetlink_subsystem *ss; @@ -670,28 +712,45 @@ static int nfnetlink_bind(struct net *net, int group) if (!ss) request_module_nowait("nfnetlink-subsys-%d", type); -#ifdef CONFIG_NF_CONNTRACK_EVENTS - if (type == NFNL_SUBSYS_CTNETLINK) { - nfnl_lock(NFNL_SUBSYS_CTNETLINK); - WRITE_ONCE(net->ct.ctnetlink_has_listener, true); - nfnl_unlock(NFNL_SUBSYS_CTNETLINK); - } -#endif + nfnetlink_bind_event(net, group); return 0; } static void nfnetlink_unbind(struct net *net, int group) { #ifdef CONFIG_NF_CONNTRACK_EVENTS + int type, group_bit; + if (group <= NFNLGRP_NONE || group > NFNLGRP_MAX) return; - if (nfnl_group2type[group] == NFNL_SUBSYS_CTNETLINK) { - nfnl_lock(NFNL_SUBSYS_CTNETLINK); - if (!nfnetlink_has_listeners(net, group)) - WRITE_ONCE(net->ct.ctnetlink_has_listener, false); - nfnl_unlock(NFNL_SUBSYS_CTNETLINK); + type = nfnl_group2type[group]; + + switch (type) { + case NFNL_SUBSYS_CTNETLINK: + break; + case NFNL_SUBSYS_CTNETLINK_EXP: + break; + default: + return; + } + + /* ctnetlink_has_listener is u8 */ + if (group >= 8) + return; + + group_bit = (1 << group); + + spin_lock(&nfnl_grp_active_lock); + if (!nfnetlink_has_listeners(net, group)) { + u8 v = READ_ONCE(net->ct.ctnetlink_has_listener); + + v &= ~group_bit; + + /* read concurrently without nfnl_grp_active_lock held. */ + WRITE_ONCE(net->ct.ctnetlink_has_listener, v); } + spin_unlock(&nfnl_grp_active_lock); #endif } From patchwork Wed Aug 17 14:00:08 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945932 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id B820AC3F6B0 for ; Wed, 17 Aug 2022 14:02:41 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239939AbiHQOCj (ORCPT ); Wed, 17 Aug 2022 10:02:39 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48750 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237185AbiHQOCC (ORCPT ); Wed, 17 Aug 2022 10:02:02 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4AECE9A9A0; Wed, 17 Aug 2022 07:01:33 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbv-0008R2-LY; Wed, 17 Aug 2022 16:01:31 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 10/17] netfilter: nf_tables: really skip inactive sets when allocating name Date: Wed, 17 Aug 2022 16:00:08 +0200 Message-Id: <20220817140015.25843-11-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso While looping to build the bitmap of used anonymous set names, check the current set in the iteration, instead of the one that is being created. Fixes: 37a9cc525525 ("netfilter: nf_tables: add generation mask to sets") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index d90434eadc1b..1b9459a364ba 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -3907,7 +3907,7 @@ static int nf_tables_set_alloc_name(struct nft_ctx *ctx, struct nft_set *set, list_for_each_entry(i, &ctx->table->sets, list) { int tmp; - if (!nft_is_active_next(ctx->net, set)) + if (!nft_is_active_next(ctx->net, i)) continue; if (!sscanf(i->name, name, &tmp)) continue; From patchwork Wed Aug 17 14:00:09 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945935 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id D2E88C32771 for ; Wed, 17 Aug 2022 14:02:49 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239967AbiHQOCr (ORCPT ); Wed, 17 Aug 2022 10:02:47 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:46458 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239886AbiHQOCG (ORCPT ); Wed, 17 Aug 2022 10:02:06 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3B97198588; Wed, 17 Aug 2022 07:01:37 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJbz-0008RP-PH; Wed, 17 Aug 2022 16:01:35 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 11/17] netfilter: nf_tables: validate NFTA_SET_ELEM_OBJREF based on NFT_SET_OBJECT flag Date: Wed, 17 Aug 2022 16:00:09 +0200 Message-Id: <20220817140015.25843-12-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso If the NFTA_SET_ELEM_OBJREF netlink attribute is present and NFT_SET_OBJECT flag is set on, report EINVAL. Move existing sanity check earlier to validate that NFT_SET_OBJECT requires NFTA_SET_ELEM_OBJREF. Fixes: 8aeff920dcc9 ("netfilter: nf_tables: add stateful object reference to set elements") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 1b9459a364ba..bcfe8120e014 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -5894,6 +5894,15 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, return -EINVAL; } + if (set->flags & NFT_SET_OBJECT) { + if (!nla[NFTA_SET_ELEM_OBJREF] && + !(flags & NFT_SET_ELEM_INTERVAL_END)) + return -EINVAL; + } else { + if (nla[NFTA_SET_ELEM_OBJREF]) + return -EINVAL; + } + if ((flags & NFT_SET_ELEM_INTERVAL_END) && (nla[NFTA_SET_ELEM_DATA] || nla[NFTA_SET_ELEM_OBJREF] || @@ -6032,10 +6041,6 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, } if (nla[NFTA_SET_ELEM_OBJREF] != NULL) { - if (!(set->flags & NFT_SET_OBJECT)) { - err = -EINVAL; - goto err_parse_key_end; - } obj = nft_obj_lookup(ctx->net, ctx->table, nla[NFTA_SET_ELEM_OBJREF], set->objtype, genmask); From patchwork Wed Aug 17 14:00:10 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945933 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id BFD1BC28B2B for ; Wed, 17 Aug 2022 14:02:51 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239899AbiHQOCs (ORCPT ); Wed, 17 Aug 2022 10:02:48 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47152 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239897AbiHQOCI (ORCPT ); Wed, 17 Aug 2022 10:02:08 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 56F257E333; Wed, 17 Aug 2022 07:01:41 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJc3-0008Rk-TC; Wed, 17 Aug 2022 16:01:39 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 12/17] netfilter: nf_tables: NFTA_SET_ELEM_KEY_END requires concat and interval flags Date: Wed, 17 Aug 2022 16:00:10 +0200 Message-Id: <20220817140015.25843-13-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso If the NFT_SET_CONCAT|NFT_SET_INTERVAL flags are set on, then the netlink attribute NFTA_SET_ELEM_KEY_END must be specified. Otherwise, NFTA_SET_ELEM_KEY_END should not be present. For catch-all element, NFTA_SET_ELEM_KEY_END should not be present. The NFT_SET_ELEM_INTERVAL_END is never used with this set flags combination. Fixes: 7b225d0b5c6d ("netfilter: nf_tables: add NFTA_SET_ELEM_KEY_END attribute") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index bcfe8120e014..1d14d694f654 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -5844,6 +5844,24 @@ static void nft_setelem_remove(const struct net *net, set->ops->remove(net, set, elem); } +static bool nft_setelem_valid_key_end(const struct nft_set *set, + struct nlattr **nla, u32 flags) +{ + if ((set->flags & (NFT_SET_CONCAT | NFT_SET_INTERVAL)) == + (NFT_SET_CONCAT | NFT_SET_INTERVAL)) { + if (flags & NFT_SET_ELEM_INTERVAL_END) + return false; + if (!nla[NFTA_SET_ELEM_KEY_END] && + !(flags & NFT_SET_ELEM_CATCHALL)) + return false; + } else { + if (nla[NFTA_SET_ELEM_KEY_END]) + return false; + } + + return true; +} + static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, const struct nlattr *attr, u32 nlmsg_flags) { @@ -5903,6 +5921,9 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, return -EINVAL; } + if (!nft_setelem_valid_key_end(set, nla, flags)) + return -EINVAL; + if ((flags & NFT_SET_ELEM_INTERVAL_END) && (nla[NFTA_SET_ELEM_DATA] || nla[NFTA_SET_ELEM_OBJREF] || @@ -6333,6 +6354,9 @@ static int nft_del_setelem(struct nft_ctx *ctx, struct nft_set *set, if (!nla[NFTA_SET_ELEM_KEY] && !(flags & NFT_SET_ELEM_CATCHALL)) return -EINVAL; + if (!nft_setelem_valid_key_end(set, nla, flags)) + return -EINVAL; + nft_set_ext_prepare(&tmpl); if (flags != 0) { From patchwork Wed Aug 17 14:00:11 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945934 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id AB6ABC32792 for ; Wed, 17 Aug 2022 14:02:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239196AbiHQOCv (ORCPT ); Wed, 17 Aug 2022 10:02:51 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47718 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239981AbiHQOCI (ORCPT ); Wed, 17 Aug 2022 10:02:08 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 709C19858F; Wed, 17 Aug 2022 07:01:45 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJc8-0008S7-0A; Wed, 17 Aug 2022 16:01:44 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 13/17] netfilter: nf_tables: disallow NFT_SET_ELEM_CATCHALL and NFT_SET_ELEM_INTERVAL_END Date: Wed, 17 Aug 2022 16:00:11 +0200 Message-Id: <20220817140015.25843-14-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso These flags are mutually exclusive, report EINVAL in this case. Fixes: aaa31047a6d2 ("netfilter: nftables: add catch-all set element support") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 1d14d694f654..b1b12e083abb 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -5198,6 +5198,9 @@ static int nft_setelem_parse_flags(const struct nft_set *set, if (!(set->flags & NFT_SET_INTERVAL) && *flags & NFT_SET_ELEM_INTERVAL_END) return -EINVAL; + if ((*flags & (NFT_SET_ELEM_INTERVAL_END | NFT_SET_ELEM_CATCHALL)) == + (NFT_SET_ELEM_INTERVAL_END | NFT_SET_ELEM_CATCHALL)) + return -EINVAL; return 0; } From patchwork Wed Aug 17 14:00:12 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945936 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3E984C25B08 for ; Wed, 17 Aug 2022 14:02:54 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S240003AbiHQOCw (ORCPT ); Wed, 17 Aug 2022 10:02:52 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47720 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239973AbiHQOCP (ORCPT ); Wed, 17 Aug 2022 10:02:15 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id B096899259; Wed, 17 Aug 2022 07:01:49 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJcC-0008ST-3p; Wed, 17 Aug 2022 16:01:48 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski Subject: [PATCH net 14/17] netfilter: nf_tables: check NFT_SET_CONCAT flag if field_count is specified Date: Wed, 17 Aug 2022 16:00:12 +0200 Message-Id: <20220817140015.25843-15-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Pablo Neira Ayuso Since f3a2181e16f1 ("netfilter: nf_tables: Support for sets with multiple ranged fields"), it possible to combine intervals and concatenations. Later on, ef516e8625dd ("netfilter: nf_tables: reintroduce the NFT_SET_CONCAT flag") provides the NFT_SET_CONCAT flag for userspace to report that the set stores a concatenation. Make sure NFT_SET_CONCAT is set on if field_count is specified for consistency. Otherwise, if NFT_SET_CONCAT is specified with no field_count, bail out with EINVAL. Fixes: ef516e8625dd ("netfilter: nf_tables: reintroduce the NFT_SET_CONCAT flag") Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index b1b12e083abb..62cfb0e31c40 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -4451,6 +4451,11 @@ static int nf_tables_newset(struct sk_buff *skb, const struct nfnl_info *info, err = nf_tables_set_desc_parse(&desc, nla[NFTA_SET_DESC]); if (err < 0) return err; + + if (desc.field_count > 1 && !(flags & NFT_SET_CONCAT)) + return -EINVAL; + } else if (flags & NFT_SET_CONCAT) { + return -EINVAL; } if (nla[NFTA_SET_EXPR] || nla[NFTA_SET_EXPRESSIONS]) From patchwork Wed Aug 17 14:00:13 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945937 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5C402C28B2B for ; Wed, 17 Aug 2022 14:02:55 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S240004AbiHQOCx (ORCPT ); Wed, 17 Aug 2022 10:02:53 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48612 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236372AbiHQOCQ (ORCPT ); Wed, 17 Aug 2022 10:02:16 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DFE22979D2; Wed, 17 Aug 2022 07:01:53 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJcG-0008Sn-79; Wed, 17 Aug 2022 16:01:52 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Geert Uytterhoeven , Florian Westphal Subject: [PATCH net 15/17] netfilter: conntrack: NF_CONNTRACK_PROCFS should no longer default to y Date: Wed, 17 Aug 2022 16:00:13 +0200 Message-Id: <20220817140015.25843-16-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org From: Geert Uytterhoeven NF_CONNTRACK_PROCFS was marked obsolete in commit 54b07dca68557b09 ("netfilter: provide config option to disable ancient procfs parts") in v3.3. Signed-off-by: Geert Uytterhoeven Signed-off-by: Florian Westphal --- net/netfilter/Kconfig | 1 - 1 file changed, 1 deletion(-) diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig index 22f15ebf6045..4b8d04640ff3 100644 --- a/net/netfilter/Kconfig +++ b/net/netfilter/Kconfig @@ -144,7 +144,6 @@ config NF_CONNTRACK_ZONES config NF_CONNTRACK_PROCFS bool "Supply CT list in procfs (OBSOLETE)" - default y depends on PROC_FS help This option enables for the list of known conntrack entries From patchwork Wed Aug 17 14:00:14 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945938 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id DAC2BC25B08 for ; Wed, 17 Aug 2022 14:03:08 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239543AbiHQODH (ORCPT ); Wed, 17 Aug 2022 10:03:07 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48274 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239536AbiHQOCW (ORCPT ); Wed, 17 Aug 2022 10:02:22 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id CABDE97521; Wed, 17 Aug 2022 07:01:57 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJcK-0008T7-Bc; Wed, 17 Aug 2022 16:01:56 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal Subject: [PATCH net 16/17] testing: selftests: nft_flowtable.sh: use random netns names Date: Wed, 17 Aug 2022 16:00:14 +0200 Message-Id: <20220817140015.25843-17-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org "ns1" is a too generic name, use a random suffix to avoid errors when such a netns exists. Also allows to run multiple instances of the script in parallel. Signed-off-by: Florian Westphal --- .../selftests/netfilter/nft_flowtable.sh | 246 +++++++++--------- 1 file changed, 128 insertions(+), 118 deletions(-) diff --git a/tools/testing/selftests/netfilter/nft_flowtable.sh b/tools/testing/selftests/netfilter/nft_flowtable.sh index d4ffebb989f8..c336e6c148d1 100755 --- a/tools/testing/selftests/netfilter/nft_flowtable.sh +++ b/tools/testing/selftests/netfilter/nft_flowtable.sh @@ -14,6 +14,11 @@ # nft_flowtable.sh -o8000 -l1500 -r2000 # +sfx=$(mktemp -u "XXXXXXXX") +ns1="ns1-$sfx" +ns2="ns2-$sfx" +nsr1="nsr1-$sfx" +nsr2="nsr2-$sfx" # Kselftest framework requirement - SKIP code is 4. ksft_skip=4 @@ -36,18 +41,17 @@ checktool (){ checktool "nft --version" "run test without nft tool" checktool "ip -Version" "run test without ip tool" checktool "which nc" "run test without nc (netcat)" -checktool "ip netns add nsr1" "create net namespace" +checktool "ip netns add $nsr1" "create net namespace $nsr1" -ip netns add ns1 -ip netns add ns2 - -ip netns add nsr2 +ip netns add $ns1 +ip netns add $ns2 +ip netns add $nsr2 cleanup() { - for i in 1 2; do - ip netns del ns$i - ip netns del nsr$i - done + ip netns del $ns1 + ip netns del $ns2 + ip netns del $nsr1 + ip netns del $nsr2 rm -f "$ns1in" "$ns1out" rm -f "$ns2in" "$ns2out" @@ -59,22 +63,21 @@ trap cleanup EXIT sysctl -q net.netfilter.nf_log_all_netns=1 -ip link add veth0 netns nsr1 type veth peer name eth0 netns ns1 -ip link add veth1 netns nsr1 type veth peer name veth0 netns nsr2 +ip link add veth0 netns $nsr1 type veth peer name eth0 netns $ns1 +ip link add veth1 netns $nsr1 type veth peer name veth0 netns $nsr2 -ip link add veth1 netns nsr2 type veth peer name eth0 netns ns2 +ip link add veth1 netns $nsr2 type veth peer name eth0 netns $ns2 for dev in lo veth0 veth1; do - for i in 1 2; do - ip -net nsr$i link set $dev up - done + ip -net $nsr1 link set $dev up + ip -net $nsr2 link set $dev up done -ip -net nsr1 addr add 10.0.1.1/24 dev veth0 -ip -net nsr1 addr add dead:1::1/64 dev veth0 +ip -net $nsr1 addr add 10.0.1.1/24 dev veth0 +ip -net $nsr1 addr add dead:1::1/64 dev veth0 -ip -net nsr2 addr add 10.0.2.1/24 dev veth1 -ip -net nsr2 addr add dead:2::1/64 dev veth1 +ip -net $nsr2 addr add 10.0.2.1/24 dev veth1 +ip -net $nsr2 addr add dead:2::1/64 dev veth1 # set different MTUs so we need to push packets coming from ns1 (large MTU) # to ns2 (smaller MTU) to stack either to perform fragmentation (ip_no_pmtu_disc=1), @@ -106,49 +109,56 @@ do esac done -if ! ip -net nsr1 link set veth0 mtu $omtu; then +if ! ip -net $nsr1 link set veth0 mtu $omtu; then exit 1 fi -ip -net ns1 link set eth0 mtu $omtu +ip -net $ns1 link set eth0 mtu $omtu -if ! ip -net nsr2 link set veth1 mtu $rmtu; then +if ! ip -net $nsr2 link set veth1 mtu $rmtu; then exit 1 fi -ip -net ns2 link set eth0 mtu $rmtu +ip -net $ns2 link set eth0 mtu $rmtu # transfer-net between nsr1 and nsr2. # these addresses are not used for connections. -ip -net nsr1 addr add 192.168.10.1/24 dev veth1 -ip -net nsr1 addr add fee1:2::1/64 dev veth1 - -ip -net nsr2 addr add 192.168.10.2/24 dev veth0 -ip -net nsr2 addr add fee1:2::2/64 dev veth0 - -for i in 1 2; do - ip netns exec nsr$i sysctl net.ipv4.conf.veth0.forwarding=1 > /dev/null - ip netns exec nsr$i sysctl net.ipv4.conf.veth1.forwarding=1 > /dev/null - - ip -net ns$i link set lo up - ip -net ns$i link set eth0 up - ip -net ns$i addr add 10.0.$i.99/24 dev eth0 - ip -net ns$i route add default via 10.0.$i.1 - ip -net ns$i addr add dead:$i::99/64 dev eth0 - ip -net ns$i route add default via dead:$i::1 - if ! ip netns exec ns$i sysctl net.ipv4.tcp_no_metrics_save=1 > /dev/null; then +ip -net $nsr1 addr add 192.168.10.1/24 dev veth1 +ip -net $nsr1 addr add fee1:2::1/64 dev veth1 + +ip -net $nsr2 addr add 192.168.10.2/24 dev veth0 +ip -net $nsr2 addr add fee1:2::2/64 dev veth0 + +for i in 0 1; do + ip netns exec $nsr1 sysctl net.ipv4.conf.veth$i.forwarding=1 > /dev/null + ip netns exec $nsr2 sysctl net.ipv4.conf.veth$i.forwarding=1 > /dev/null +done + +for ns in $ns1 $ns2;do + ip -net $ns link set lo up + ip -net $ns link set eth0 up + + if ! ip netns exec $ns sysctl net.ipv4.tcp_no_metrics_save=1 > /dev/null; then echo "ERROR: Check Originator/Responder values (problem during address addition)" exit 1 fi - # don't set ip DF bit for first two tests - ip netns exec ns$i sysctl net.ipv4.ip_no_pmtu_disc=1 > /dev/null + ip netns exec $ns sysctl net.ipv4.ip_no_pmtu_disc=1 > /dev/null done -ip -net nsr1 route add default via 192.168.10.2 -ip -net nsr2 route add default via 192.168.10.1 +ip -net $ns1 addr add 10.0.1.99/24 dev eth0 +ip -net $ns2 addr add 10.0.2.99/24 dev eth0 +ip -net $ns1 route add default via 10.0.1.1 +ip -net $ns2 route add default via 10.0.2.1 +ip -net $ns1 addr add dead:1::99/64 dev eth0 +ip -net $ns2 addr add dead:2::99/64 dev eth0 +ip -net $ns1 route add default via dead:1::1 +ip -net $ns2 route add default via dead:2::1 + +ip -net $nsr1 route add default via 192.168.10.2 +ip -net $nsr2 route add default via 192.168.10.1 -ip netns exec nsr1 nft -f - < /dev/null; then - echo "ERROR: ns1 cannot reach ns2" 1>&2 +if ! ip netns exec $ns1 ping -c 1 -q 10.0.2.99 > /dev/null; then + echo "ERROR: $ns1 cannot reach ns2" 1>&2 exit 1 fi -if ! ip netns exec ns2 ping -c 1 -q 10.0.1.99 > /dev/null; then - echo "ERROR: ns2 cannot reach ns1" 1>&2 +if ! ip netns exec $ns2 ping -c 1 -q 10.0.1.99 > /dev/null; then + echo "ERROR: $ns2 cannot reach $ns1" 1>&2 exit 1 fi if [ $ret -eq 0 ];then - echo "PASS: netns routing/connectivity: ns1 can reach ns2" + echo "PASS: netns routing/connectivity: $ns1 can reach $ns2" fi ns1in=$(mktemp) @@ -312,24 +322,24 @@ make_file "$ns2in" # First test: # No PMTU discovery, nsr1 is expected to fragment packets from ns1 to ns2 as needed. -if test_tcp_forwarding ns1 ns2; then +if test_tcp_forwarding $ns1 $ns2; then echo "PASS: flow offloaded for ns1/ns2" else echo "FAIL: flow offload for ns1/ns2:" 1>&2 - ip netns exec nsr1 nft list ruleset + ip netns exec $nsr1 nft list ruleset ret=1 fi # delete default route, i.e. ns2 won't be able to reach ns1 and # will depend on ns1 being masqueraded in nsr1. # expect ns1 has nsr1 address. -ip -net ns2 route del default via 10.0.2.1 -ip -net ns2 route del default via dead:2::1 -ip -net ns2 route add 192.168.10.1 via 10.0.2.1 +ip -net $ns2 route del default via 10.0.2.1 +ip -net $ns2 route del default via dead:2::1 +ip -net $ns2 route add 192.168.10.1 via 10.0.2.1 # Second test: # Same, but with NAT enabled. -ip netns exec nsr1 nft -f - <&2 - ip netns exec nsr1 nft list ruleset + ip netns exec $nsr1 nft list ruleset ret=1 fi # Third test: # Same as second test, but with PMTU discovery enabled. -handle=$(ip netns exec nsr1 nft -a list table inet filter | grep something-to-grep-for | cut -d \# -f 2) +handle=$(ip netns exec $nsr1 nft -a list table inet filter | grep something-to-grep-for | cut -d \# -f 2) -if ! ip netns exec nsr1 nft delete rule inet filter forward $handle; then +if ! ip netns exec $nsr1 nft delete rule inet filter forward $handle; then echo "FAIL: Could not delete large-packet accept rule" exit 1 fi -ip netns exec ns1 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null -ip netns exec ns2 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null +ip netns exec $ns1 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null +ip netns exec $ns2 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null -if test_tcp_forwarding_nat ns1 ns2; then +if test_tcp_forwarding_nat $ns1 $ns2; then echo "PASS: flow offloaded for ns1/ns2 with NAT and pmtu discovery" else echo "FAIL: flow offload for ns1/ns2 with NAT and pmtu discovery" 1>&2 - ip netns exec nsr1 nft list ruleset + ip netns exec $nsr1 nft list ruleset fi # Another test: # Add bridge interface br0 to Router1, with NAT enabled. -ip -net nsr1 link add name br0 type bridge -ip -net nsr1 addr flush dev veth0 -ip -net nsr1 link set up dev veth0 -ip -net nsr1 link set veth0 master br0 -ip -net nsr1 addr add 10.0.1.1/24 dev br0 -ip -net nsr1 addr add dead:1::1/64 dev br0 -ip -net nsr1 link set up dev br0 +ip -net $nsr1 link add name br0 type bridge +ip -net $nsr1 addr flush dev veth0 +ip -net $nsr1 link set up dev veth0 +ip -net $nsr1 link set veth0 master br0 +ip -net $nsr1 addr add 10.0.1.1/24 dev br0 +ip -net $nsr1 addr add dead:1::1/64 dev br0 +ip -net $nsr1 link set up dev br0 -ip netns exec nsr1 sysctl net.ipv4.conf.br0.forwarding=1 > /dev/null +ip netns exec $nsr1 sysctl net.ipv4.conf.br0.forwarding=1 > /dev/null # br0 with NAT enabled. -ip netns exec nsr1 nft -f - <&2 - ip netns exec nsr1 nft list ruleset + ip netns exec $nsr1 nft list ruleset ret=1 fi # Another test: # Add bridge interface br0 to Router1, with NAT and VLAN. -ip -net nsr1 link set veth0 nomaster -ip -net nsr1 link set down dev veth0 -ip -net nsr1 link add link veth0 name veth0.10 type vlan id 10 -ip -net nsr1 link set up dev veth0 -ip -net nsr1 link set up dev veth0.10 -ip -net nsr1 link set veth0.10 master br0 - -ip -net ns1 addr flush dev eth0 -ip -net ns1 link add link eth0 name eth0.10 type vlan id 10 -ip -net ns1 link set eth0 up -ip -net ns1 link set eth0.10 up -ip -net ns1 addr add 10.0.1.99/24 dev eth0.10 -ip -net ns1 route add default via 10.0.1.1 -ip -net ns1 addr add dead:1::99/64 dev eth0.10 - -if test_tcp_forwarding_nat ns1 ns2; then +ip -net $nsr1 link set veth0 nomaster +ip -net $nsr1 link set down dev veth0 +ip -net $nsr1 link add link veth0 name veth0.10 type vlan id 10 +ip -net $nsr1 link set up dev veth0 +ip -net $nsr1 link set up dev veth0.10 +ip -net $nsr1 link set veth0.10 master br0 + +ip -net $ns1 addr flush dev eth0 +ip -net $ns1 link add link eth0 name eth0.10 type vlan id 10 +ip -net $ns1 link set eth0 up +ip -net $ns1 link set eth0.10 up +ip -net $ns1 addr add 10.0.1.99/24 dev eth0.10 +ip -net $ns1 route add default via 10.0.1.1 +ip -net $ns1 addr add dead:1::99/64 dev eth0.10 + +if test_tcp_forwarding_nat $ns1 $ns2; then echo "PASS: flow offloaded for ns1/ns2 with bridge NAT and VLAN" else echo "FAIL: flow offload for ns1/ns2 with bridge NAT and VLAN" 1>&2 - ip netns exec nsr1 nft list ruleset + ip netns exec $nsr1 nft list ruleset ret=1 fi # restore test topology (remove bridge and VLAN) -ip -net nsr1 link set veth0 nomaster -ip -net nsr1 link set veth0 down -ip -net nsr1 link set veth0.10 down -ip -net nsr1 link delete veth0.10 type vlan -ip -net nsr1 link delete br0 type bridge -ip -net ns1 addr flush dev eth0.10 -ip -net ns1 link set eth0.10 down -ip -net ns1 link set eth0 down -ip -net ns1 link delete eth0.10 type vlan +ip -net $nsr1 link set veth0 nomaster +ip -net $nsr1 link set veth0 down +ip -net $nsr1 link set veth0.10 down +ip -net $nsr1 link delete veth0.10 type vlan +ip -net $nsr1 link delete br0 type bridge +ip -net $ns1 addr flush dev eth0.10 +ip -net $ns1 link set eth0.10 down +ip -net $ns1 link set eth0 down +ip -net $ns1 link delete eth0.10 type vlan # restore address in ns1 and nsr1 -ip -net ns1 link set eth0 up -ip -net ns1 addr add 10.0.1.99/24 dev eth0 -ip -net ns1 route add default via 10.0.1.1 -ip -net ns1 addr add dead:1::99/64 dev eth0 -ip -net ns1 route add default via dead:1::1 -ip -net nsr1 addr add 10.0.1.1/24 dev veth0 -ip -net nsr1 addr add dead:1::1/64 dev veth0 -ip -net nsr1 link set up dev veth0 +ip -net $ns1 link set eth0 up +ip -net $ns1 addr add 10.0.1.99/24 dev eth0 +ip -net $ns1 route add default via 10.0.1.1 +ip -net $ns1 addr add dead:1::99/64 dev eth0 +ip -net $ns1 route add default via dead:1::1 +ip -net $nsr1 addr add 10.0.1.1/24 dev veth0 +ip -net $nsr1 addr add dead:1::1/64 dev veth0 +ip -net $nsr1 link set up dev veth0 KEY_SHA="0x"$(ps -xaf | sha1sum | cut -d " " -f 1) KEY_AES="0x"$(ps -xaf | md5sum | cut -d " " -f 1) @@ -480,23 +490,23 @@ do_esp() { } -do_esp nsr1 192.168.10.1 192.168.10.2 10.0.1.0/24 10.0.2.0/24 $SPI1 $SPI2 +do_esp $nsr1 192.168.10.1 192.168.10.2 10.0.1.0/24 10.0.2.0/24 $SPI1 $SPI2 -do_esp nsr2 192.168.10.2 192.168.10.1 10.0.2.0/24 10.0.1.0/24 $SPI2 $SPI1 +do_esp $nsr2 192.168.10.2 192.168.10.1 10.0.2.0/24 10.0.1.0/24 $SPI2 $SPI1 -ip netns exec nsr1 nft delete table ip nat +ip netns exec $nsr1 nft delete table ip nat # restore default routes -ip -net ns2 route del 192.168.10.1 via 10.0.2.1 -ip -net ns2 route add default via 10.0.2.1 -ip -net ns2 route add default via dead:2::1 +ip -net $ns2 route del 192.168.10.1 via 10.0.2.1 +ip -net $ns2 route add default via 10.0.2.1 +ip -net $ns2 route add default via dead:2::1 -if test_tcp_forwarding ns1 ns2; then +if test_tcp_forwarding $ns1 $ns2; then echo "PASS: ipsec tunnel mode for ns1/ns2" else echo "FAIL: ipsec tunnel mode for ns1/ns2" - ip netns exec nsr1 nft list ruleset 1>&2 - ip netns exec nsr1 cat /proc/net/xfrm_stat 1>&2 + ip netns exec $nsr1 nft list ruleset 1>&2 + ip netns exec $nsr1 cat /proc/net/xfrm_stat 1>&2 fi exit $ret From patchwork Wed Aug 17 14:00:15 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Florian Westphal X-Patchwork-Id: 12945939 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 55CC0C25B08 for ; Wed, 17 Aug 2022 14:03:17 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S239991AbiHQODP (ORCPT ); Wed, 17 Aug 2022 10:03:15 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52546 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239863AbiHQOCZ (ORCPT ); Wed, 17 Aug 2022 10:02:25 -0400 Received: from Chamillionaire.breakpoint.cc (Chamillionaire.breakpoint.cc [IPv6:2a0a:51c0:0:12e:520::1]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 705029A9B8; Wed, 17 Aug 2022 07:02:01 -0700 (PDT) Received: from fw by Chamillionaire.breakpoint.cc with local (Exim 4.92) (envelope-from ) id 1oOJcO-0008TT-FS; Wed, 17 Aug 2022 16:02:00 +0200 From: Florian Westphal To: netdev@vger.kernel.org Cc: , Pablo Neira Ayuso , Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Florian Westphal Subject: [PATCH net 17/17] testing: selftests: nft_flowtable.sh: rework test to detect offload failure Date: Wed, 17 Aug 2022 16:00:15 +0200 Message-Id: <20220817140015.25843-18-fw@strlen.de> X-Mailer: git-send-email 2.35.1 In-Reply-To: <20220817140015.25843-1-fw@strlen.de> References: <20220817140015.25843-1-fw@strlen.de> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org This test fails on current kernel releases because the flotwable path now calls dst_check from packet path and will then remove the offload. Test script has two purposes: 1. check that file (random content) can be sent to other netns (and vv) 2. check that the flow is offloaded (rather than handled by classic forwarding path). Since dst_check is in place, 2) fails because the nftables ruleset in router namespace 1 intentionally blocks traffic under the assumption that packets are not passed via classic path at all. Rework this: Instead of blocking traffic, create two named counters, one for original and one for reverse direction. The first three test cases are handled by classic forwarding path (path mtu discovery is disabled and packets exceed MTU). But all other tests enable PMTUD, so the originator and responder are expected to lower packet size and flowtable is expected to do the packet forwarding. For those tests, check that the packet counters (which are only incremented for packets that are passed up to classic forward path) are significantly lower than the file size transferred. I've tested that the counter-checks fail as expected when the 'flow add' statement is removed from the ruleset. Signed-off-by: Florian Westphal --- .../selftests/netfilter/nft_flowtable.sh | 141 +++++++++++------- 1 file changed, 84 insertions(+), 57 deletions(-) diff --git a/tools/testing/selftests/netfilter/nft_flowtable.sh b/tools/testing/selftests/netfilter/nft_flowtable.sh index c336e6c148d1..7060bae04ec8 100755 --- a/tools/testing/selftests/netfilter/nft_flowtable.sh +++ b/tools/testing/selftests/netfilter/nft_flowtable.sh @@ -24,8 +24,7 @@ nsr2="nsr2-$sfx" ksft_skip=4 ret=0 -ns1in="" -ns2in="" +nsin="" ns1out="" ns2out="" @@ -53,8 +52,7 @@ cleanup() { ip netns del $nsr1 ip netns del $nsr2 - rm -f "$ns1in" "$ns1out" - rm -f "$ns2in" "$ns2out" + rm -f "$nsin" "$ns1out" "$ns2out" [ $log_netns -eq 0 ] && sysctl -q net.netfilter.nf_log_all_netns=$log_netns } @@ -165,36 +163,20 @@ table inet filter { devices = { veth0, veth1 } } + counter routed_orig { } + counter routed_repl { } + chain forward { type filter hook forward priority 0; policy drop; # flow offloaded? Tag ct with mark 1, so we can detect when it fails. - meta oif "veth1" tcp dport 12345 flow offload @f1 counter - - # use packet size to trigger 'should be offloaded by now'. - # otherwise, if 'flow offload' expression never offloads, the - # test will pass. - tcp dport 12345 meta length gt 200 ct mark set 1 counter - - # this turns off flow offloading internally, so expect packets again - tcp flags fin,rst ct mark set 0 accept - - # this allows large packets from responder, we need this as long - # as PMTUd is off. - # This rule is deleted for the last test, when we expect PMTUd - # to kick in and ensure all packets meet mtu requirements. - meta length gt $lmtu accept comment something-to-grep-for + meta oif "veth1" tcp dport 12345 ct mark set 1 flow add @f1 counter name routed_orig accept - # next line blocks connection w.o. working offload. - # we only do this for reverse dir, because we expect packets to - # enter slow path due to MTU mismatch of veth0 and veth1. - tcp sport 12345 ct mark 1 counter log prefix "mark failure " drop + # count packets supposedly offloaded as per direction. + ct mark 1 counter name ct direction map { original : routed_orig, reply : routed_repl } accept ct state established,related accept - # for packets that we can't offload yet, i.e. SYN (any ct that is not confirmed) - meta length lt 200 oif "veth1" tcp dport 12345 counter accept - meta nfproto ipv4 meta l4proto icmp accept meta nfproto ipv6 meta l4proto icmpv6 accept } @@ -221,16 +203,16 @@ if [ $ret -eq 0 ];then echo "PASS: netns routing/connectivity: $ns1 can reach $ns2" fi -ns1in=$(mktemp) +nsin=$(mktemp) ns1out=$(mktemp) -ns2in=$(mktemp) ns2out=$(mktemp) make_file() { name=$1 - SIZE=$((RANDOM % (1024 * 8))) + SIZE=$((RANDOM % (1024 * 128))) + SIZE=$((SIZE + (1024 * 8))) TSIZE=$((SIZE * 1024)) dd if=/dev/urandom of="$name" bs=1024 count=$SIZE 2> /dev/null @@ -241,6 +223,38 @@ make_file() dd if=/dev/urandom conf=notrunc of="$name" bs=1 count=$SIZE 2> /dev/null } +check_counters() +{ + local what=$1 + local ok=1 + + local orig=$(ip netns exec $nsr1 nft reset counter inet filter routed_orig | grep packets) + local repl=$(ip netns exec $nsr1 nft reset counter inet filter routed_repl | grep packets) + + local orig_cnt=${orig#*bytes} + local repl_cnt=${repl#*bytes} + + local fs=$(du -sb $nsin) + local max_orig=${fs%%/*} + local max_repl=$((max_orig/4)) + + if [ $orig_cnt -gt $max_orig ];then + echo "FAIL: $what: original counter $orig_cnt exceeds expected value $max_orig" 1>&2 + ret=1 + ok=0 + fi + + if [ $repl_cnt -gt $max_repl ];then + echo "FAIL: $what: reply counter $repl_cnt exceeds expected value $max_repl" 1>&2 + ret=1 + ok=0 + fi + + if [ $ok -eq 1 ]; then + echo "PASS: $what" + fi +} + check_transfer() { in=$1 @@ -265,11 +279,11 @@ test_tcp_forwarding_ip() local dstport=$4 local lret=0 - ip netns exec $nsb nc -w 5 -l -p 12345 < "$ns2in" > "$ns2out" & + ip netns exec $nsb nc -w 5 -l -p 12345 < "$nsin" > "$ns2out" & lpid=$! sleep 1 - ip netns exec $nsa nc -w 4 "$dstip" "$dstport" < "$ns1in" > "$ns1out" & + ip netns exec $nsa nc -w 4 "$dstip" "$dstport" < "$nsin" > "$ns1out" & cpid=$! sleep 3 @@ -284,11 +298,11 @@ test_tcp_forwarding_ip() wait - if ! check_transfer "$ns1in" "$ns2out" "ns1 -> ns2"; then + if ! check_transfer "$nsin" "$ns2out" "ns1 -> ns2"; then lret=1 fi - if ! check_transfer "$ns2in" "$ns1out" "ns1 <- ns2"; then + if ! check_transfer "$nsin" "$ns1out" "ns1 <- ns2"; then lret=1 fi @@ -305,23 +319,40 @@ test_tcp_forwarding() test_tcp_forwarding_nat() { local lret + local pmtu test_tcp_forwarding_ip "$1" "$2" 10.0.2.99 12345 lret=$? + pmtu=$3 + what=$4 + if [ $lret -eq 0 ] ; then + if [ $pmtu -eq 1 ] ;then + check_counters "flow offload for ns1/ns2 with masquerade and pmtu discovery $what" + else + echo "PASS: flow offload for ns1/ns2 with masquerade $what" + fi + test_tcp_forwarding_ip "$1" "$2" 10.6.6.6 1666 lret=$? + if [ $pmtu -eq 1 ] ;then + check_counters "flow offload for ns1/ns2 with dnat and pmtu discovery $what" + elif [ $lret -eq 0 ] ; then + echo "PASS: flow offload for ns1/ns2 with dnat $what" + fi fi return $lret } -make_file "$ns1in" -make_file "$ns2in" +make_file "$nsin" # First test: # No PMTU discovery, nsr1 is expected to fragment packets from ns1 to ns2 as needed. +# Due to MTU mismatch in both directions, all packets (except small packets like pure +# acks) have to be handled by normal forwarding path. Therefore, packet counters +# are not checked. if test_tcp_forwarding $ns1 $ns2; then echo "PASS: flow offloaded for ns1/ns2" else @@ -338,7 +369,8 @@ ip -net $ns2 route del default via dead:2::1 ip -net $ns2 route add 192.168.10.1 via 10.0.2.1 # Second test: -# Same, but with NAT enabled. +# Same, but with NAT enabled. Same as in first test: we expect normal forward path +# to handle most packets. ip netns exec $nsr1 nft -f - <&2 ip netns exec $nsr1 nft list ruleset ret=1 fi # Third test: -# Same as second test, but with PMTU discovery enabled. -handle=$(ip netns exec $nsr1 nft -a list table inet filter | grep something-to-grep-for | cut -d \# -f 2) - -if ! ip netns exec $nsr1 nft delete rule inet filter forward $handle; then - echo "FAIL: Could not delete large-packet accept rule" - exit 1 -fi - +# Same as second test, but with PMTU discovery enabled. This +# means that we expect the fastpath to handle packets as soon +# as the endpoints adjust the packet size. ip netns exec $ns1 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null ip netns exec $ns2 sysctl net.ipv4.ip_no_pmtu_disc=0 > /dev/null -if test_tcp_forwarding_nat $ns1 $ns2; then - echo "PASS: flow offloaded for ns1/ns2 with NAT and pmtu discovery" -else +# reset counters. +# With pmtu in-place we'll also check that nft counters +# are lower than file size and packets were forwarded via flowtable layer. +# For earlier tests (large mtus), packets cannot be handled via flowtable +# (except pure acks and other small packets). +ip netns exec $nsr1 nft reset counters table inet filter >/dev/null + +if ! test_tcp_forwarding_nat $ns1 $ns2 1 ""; then echo "FAIL: flow offload for ns1/ns2 with NAT and pmtu discovery" 1>&2 ip netns exec $nsr1 nft list ruleset fi @@ -408,14 +438,13 @@ table ip nat { } EOF -if test_tcp_forwarding_nat $ns1 $ns2; then - echo "PASS: flow offloaded for ns1/ns2 with bridge NAT" -else +if ! test_tcp_forwarding_nat $ns1 $ns2 1 "on bridge"; then echo "FAIL: flow offload for ns1/ns2 with bridge NAT" 1>&2 ip netns exec $nsr1 nft list ruleset ret=1 fi + # Another test: # Add bridge interface br0 to Router1, with NAT and VLAN. ip -net $nsr1 link set veth0 nomaster @@ -433,9 +462,7 @@ ip -net $ns1 addr add 10.0.1.99/24 dev eth0.10 ip -net $ns1 route add default via 10.0.1.1 ip -net $ns1 addr add dead:1::99/64 dev eth0.10 -if test_tcp_forwarding_nat $ns1 $ns2; then - echo "PASS: flow offloaded for ns1/ns2 with bridge NAT and VLAN" -else +if ! test_tcp_forwarding_nat $ns1 $ns2 1 "bridge and VLAN"; then echo "FAIL: flow offload for ns1/ns2 with bridge NAT and VLAN" 1>&2 ip netns exec $nsr1 nft list ruleset ret=1 @@ -502,7 +529,7 @@ ip -net $ns2 route add default via 10.0.2.1 ip -net $ns2 route add default via dead:2::1 if test_tcp_forwarding $ns1 $ns2; then - echo "PASS: ipsec tunnel mode for ns1/ns2" + check_counters "ipsec tunnel mode for ns1/ns2" else echo "FAIL: ipsec tunnel mode for ns1/ns2" ip netns exec $nsr1 nft list ruleset 1>&2