From db3d0e0e5d4bc5ab4fe445b9f413d1b486508ca5 Mon Sep 17 00:00:00 2001 From: Xiang Mei Date: Sun, 12 Jul 2026 16:42:01 -0700 Subject: netfilter: nf_conntrack_sip: widen NAT rewrite delta to s32 in sip_help_tcp() sip_help_tcp() stores the size change of each NAT-rewritten SIP message in s16 diff and accumulates it in s16 tdiff, but a single message can grow by more than S16_MAX while the packet stays under the 65535 enlarge_skb() limit: nf_nat_sip() rewrites every matching URI, and a long Contact list expands the message by tens of kilobytes. diff then wraps, and "datalen = datalen + diff - msglen" yields a huge unsigned datalen, so the next iteration's ct_sip_get_header() reads past the linearized skb tail. Widen diff, tdiff and the seq_adjust hook to s32. Both are bounded by the 65535 byte packet limit, and the seqadj core is already s32 (nf_ct_seqadj_set() takes s32), so no previously accepted input is rejected. BUG: KASAN: use-after-free in ct_sip_get_header (net/netfilter/nf_conntrack_sip.c:464) Read of size 1 at addr ffff888010800000 by task ksoftirqd/1/25 ct_sip_get_header (net/netfilter/nf_conntrack_sip.c:464) sip_help_tcp (net/netfilter/nf_conntrack_sip.c:1694) nf_confirm (net/netfilter/nf_conntrack_proto.c:183) nf_hook_slow (net/netfilter/core.c:619) ip6_output (net/ipv6/ip6_output.c:246) ip6_forward (net/ipv6/ip6_output.c:690) ipv6_rcv (net/ipv6/ip6_input.c:351) __netif_receive_skb_one_core (net/core/dev.c:6212) process_backlog (net/core/dev.c:6676) __napi_poll (net/core/dev.c:7735) net_rx_action (net/core/dev.c:7955) handle_softirqs (kernel/softirq.c:622) run_ksoftirqd (kernel/softirq.c:1076) ... Fixes: f5b321bd37fb ("netfilter: nf_conntrack_sip: add TCP support") Reported-by: Weiming Shi Link: https://patch.msgid.link/netfilter-devel/20260712234201.3213635-1-xmei5@asu.edu Assisted-by: Claude:claude-opus-4-8 Signed-off-by: Xiang Mei Signed-off-by: Pablo Neira Ayuso --- include/linux/netfilter/nf_conntrack_sip.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/netfilter/nf_conntrack_sip.h b/include/linux/netfilter/nf_conntrack_sip.h index dbc614dfe0d5..aafa0c04f917 100644 --- a/include/linux/netfilter/nf_conntrack_sip.h +++ b/include/linux/netfilter/nf_conntrack_sip.h @@ -115,7 +115,7 @@ struct nf_nat_sip_hooks { unsigned int *datalen); void (*seq_adjust)(struct sk_buff *skb, - unsigned int protoff, s16 off); + unsigned int protoff, s32 off); unsigned int (*expect)(struct sk_buff *skb, unsigned int protoff, -- cgit From 4aa63842fc92de1bce59d4709a0d32e718890bb2 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 13 Jul 2026 00:26:04 +0200 Subject: netfilter: nf_conntrack_expect: add and use nf_ct_expect_related_pair() Add a new function to insert a pair of expectations, this is required by the SIP and H323 NAT helpers. The spinlock is held to check if there is a slot for both expectations, in such case, insert them. This removes the need for nf_ct_unexpect_related() inside the loop to find a pair of consecutive ports, otherwise inserting expectations whose dead flag is already set on can happen. Bump master_help->expecting for the expectation class after checking if the expectation fits in the master expectation list, which is needed for this new _pair() function variant to run the eviction routine including the preallocated slot for the first expectation in the pair. Fixes: b8b09dc2bf35 ("netfilter: nf_conntrack_expect: use conntrack GC to reap expectations") Reported-by: Jaeyeong Lee Link: https://patch.msgid.link/178377968720.33756.12204817361601593230@proton.me/ Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_conntrack_expect.h | 3 +++ net/ipv4/netfilter/nf_nat_h323.c | 22 ++++++------------ net/netfilter/nf_conntrack_expect.c | 35 ++++++++++++++++++++++++++++- net/netfilter/nf_nat_sip.c | 20 +++++------------ 4 files changed, 50 insertions(+), 30 deletions(-) (limited to 'include') diff --git a/include/net/netfilter/nf_conntrack_expect.h b/include/net/netfilter/nf_conntrack_expect.h index c024345c9bd8..26d6babd92fc 100644 --- a/include/net/netfilter/nf_conntrack_expect.h +++ b/include/net/netfilter/nf_conntrack_expect.h @@ -161,6 +161,9 @@ static inline int nf_ct_expect_related(struct nf_conntrack_expect *expect, return nf_ct_expect_related_report(expect, 0, 0, flags); } +int nf_ct_expect_related_pair(struct nf_conntrack_expect *expect[], + unsigned int flag); + struct nf_conn_help; void nf_ct_expectation_gc(struct nf_conn_help *master_help); diff --git a/net/ipv4/netfilter/nf_nat_h323.c b/net/ipv4/netfilter/nf_nat_h323.c index 183e8a3ff2ba..6bcd6734769b 100644 --- a/net/ipv4/netfilter/nf_nat_h323.c +++ b/net/ipv4/netfilter/nf_nat_h323.c @@ -182,6 +182,7 @@ static int nat_rtp_rtcp(struct sk_buff *skb, struct nf_conn *ct, struct nf_conntrack_expect *rtp_exp, struct nf_conntrack_expect *rtcp_exp) { + struct nf_conntrack_expect *rtp_pair[2] = { rtp_exp, rtcp_exp }; struct nf_ct_h323_master *info = nfct_help_data(ct); int dir = CTINFO2DIR(ctinfo); int i; @@ -227,22 +228,13 @@ static int nat_rtp_rtcp(struct sk_buff *skb, struct nf_conn *ct, int ret; rtp_exp->tuple.dst.u.udp.port = htons(nated_port); - ret = nf_ct_expect_related(rtp_exp, 0); + rtcp_exp->tuple.dst.u.udp.port = htons(nated_port + 1); + ret = nf_ct_expect_related_pair(rtp_pair, 0); if (ret == 0) { - rtcp_exp->tuple.dst.u.udp.port = - htons(nated_port + 1); - ret = nf_ct_expect_related(rtcp_exp, 0); - if (ret == 0) - break; - else if (ret == -EBUSY) { - nf_ct_unexpect_related(rtp_exp); - continue; - } else if (ret < 0) { - nf_ct_unexpect_related(rtp_exp); - nated_port = 0; - break; - } - } else if (ret != -EBUSY) { + break; + } else if (ret == -EBUSY) { + continue; + } else if (ret < 0) { nated_port = 0; break; } diff --git a/net/netfilter/nf_conntrack_expect.c b/net/netfilter/nf_conntrack_expect.c index 7ae68d60586a..8a3b9e33e94f 100644 --- a/net/netfilter/nf_conntrack_expect.c +++ b/net/netfilter/nf_conntrack_expect.c @@ -427,7 +427,6 @@ static void nf_ct_expect_insert(struct nf_conntrack_expect *exp, exp->timeout += helper->expect_policy[exp->class].timeout * HZ; hlist_add_head_rcu(&exp->lnode, &master_help->expectations); - master_help->expecting[exp->class]++; hlist_add_head_rcu(&exp->hnode, &nf_ct_expect_hash[h]); cnet = nf_ct_pernet(net); @@ -534,6 +533,7 @@ int nf_ct_expect_related_report(struct nf_conntrack_expect *expect, if (ret < 0) goto out; + master_help->expecting[expect->class]++; nf_ct_expect_insert(expect, master_help); nf_ct_expect_event_report(IPEXP_NEW, expect, portid, report); @@ -546,6 +546,39 @@ out: } EXPORT_SYMBOL_GPL(nf_ct_expect_related_report); +int nf_ct_expect_related_pair(struct nf_conntrack_expect *expect[], + unsigned int flags) +{ + struct nf_conn_help *master_help; + int i, ret; + + spin_lock_bh(&nf_conntrack_expect_lock); + master_help = nfct_help(expect[0]->master); + if (!master_help || master_help != nfct_help(expect[1]->master)) { + ret = -EINVAL; + goto out; + } + + for (i = 0; i < 2; i++) { + ret = __nf_ct_expect_check(expect[i], master_help, flags); + if (ret < 0) { + if (i == 1) + master_help->expecting[expect[0]->class]--; + goto out; + } + master_help->expecting[expect[i]->class]++; + } + + for (i = 0; i < 2; i++) { + nf_ct_expect_insert(expect[i], master_help); + nf_ct_expect_event_report(IPEXP_NEW, expect[i], 0, 0); + } +out: + spin_unlock_bh(&nf_conntrack_expect_lock); + return ret; +} +EXPORT_SYMBOL_GPL(nf_ct_expect_related_pair); + void nf_ct_expect_iterate_destroy(bool (*iter)(struct nf_conntrack_expect *e, void *data), void *data) { diff --git a/net/netfilter/nf_nat_sip.c b/net/netfilter/nf_nat_sip.c index a93eaf0f7d30..133bd713fe0c 100644 --- a/net/netfilter/nf_nat_sip.c +++ b/net/netfilter/nf_nat_sip.c @@ -592,6 +592,7 @@ static unsigned int nf_nat_sdp_media(struct sk_buff *skb, unsigned int protoff, unsigned int medialen, union nf_inet_addr *rtp_addr) { + struct nf_conntrack_expect *rtp_pair[2] = { rtp_exp, rtcp_exp }; enum ip_conntrack_info ctinfo; struct nf_conn *ct = nf_ct_get(skb, &ctinfo); enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); @@ -622,24 +623,15 @@ static unsigned int nf_nat_sdp_media(struct sk_buff *skb, unsigned int protoff, int ret; rtp_exp->tuple.dst.u.udp.port = htons(port); - ret = nf_ct_expect_related(rtp_exp, - NF_CT_EXP_F_SKIP_MASTER); - if (ret == -EBUSY) - continue; - else if (ret < 0) { - port = 0; - break; - } rtcp_exp->tuple.dst.u.udp.port = htons(port + 1); - ret = nf_ct_expect_related(rtcp_exp, - NF_CT_EXP_F_SKIP_MASTER); + + ret = nf_ct_expect_related_pair(rtp_pair, + NF_CT_EXP_F_SKIP_MASTER); if (ret == 0) break; - else if (ret == -EBUSY) { - nf_ct_unexpect_related(rtp_exp); + else if (ret == -EBUSY) continue; - } else if (ret < 0) { - nf_ct_unexpect_related(rtp_exp); + else if (ret < 0) { port = 0; break; } -- cgit From f4f699790590bd0896c48a71e9232a65198f92f0 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Thu, 16 Jul 2026 10:13:37 +0200 Subject: netfilter: nf_tables: make nft_object rhltable per table The nft_object rhltable is global, this allows for accessing objects that are being dismangled from lookup path by other existing netns. Given the nft_obj_destroy() releases the object inmediately, this might lead to use-after-free of these objects that are being released. Make the existing rhltable per table to address this issue to deal with with the nft_rcv_nl_event() path too. Update nft_obj_lookup() to take the table as non-const, otherwise, compiler complains when passing the objname_ht to rhltable_lookup(). Fixes: 4d44175aa5bb ("netfilter: nf_tables: handle nft_object lookups via rhltable") Suggested-by: Florian Westphal Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 4 +++- net/netfilter/nf_tables_api.c | 34 ++++++++++++++++------------------ 2 files changed, 19 insertions(+), 19 deletions(-) (limited to 'include') diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index 9d844354c4d9..3be612145c13 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -1294,6 +1294,7 @@ static inline void nft_use_inc_restore(u32 *use) * @sets: sets in the table * @objects: stateful objects in the table * @flowtables: flow tables in the table + * @objname_ht: hashtable for objects lookup by name * @hgenerator: handle generator state * @handle: table handle * @use: number of chain references to this table @@ -1313,6 +1314,7 @@ struct nft_table { struct list_head sets; struct list_head objects; struct list_head flowtables; + struct rhltable objname_ht; u64 hgenerator; u64 handle; u32 use; @@ -1400,7 +1402,7 @@ static inline void *nft_obj_data(const struct nft_object *obj) #define nft_expr_obj(expr) *((struct nft_object **)nft_expr_priv(expr)) struct nft_object *nft_obj_lookup(const struct net *net, - const struct nft_table *table, + struct nft_table *table, const struct nlattr *nla, u32 objtype, u8 genmask); diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index a9eaf9455c77..af357f6c5070 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -45,8 +45,6 @@ enum { NFT_VALIDATE_DO, }; -static struct rhltable nft_objname_ht; - static u32 nft_chain_hash(const void *data, u32 len, u32 seed); static u32 nft_chain_hash_obj(const void *data, u32 len, u32 seed); static int nft_chain_hash_cmp(struct rhashtable_compare_arg *, const void *); @@ -1635,6 +1633,10 @@ static int nf_tables_newtable(struct sk_buff *skb, const struct nfnl_info *info, if (err) goto err_chain_ht; + err = rhltable_init(&table->objname_ht, &nft_objname_ht_params); + if (err < 0) + goto err_obj_ht; + INIT_LIST_HEAD(&table->chains); INIT_LIST_HEAD(&table->sets); INIT_LIST_HEAD(&table->objects); @@ -1653,6 +1655,8 @@ static int nf_tables_newtable(struct sk_buff *skb, const struct nfnl_info *info, list_add_tail_rcu(&table->list, &nft_net->tables); return 0; err_trans: + rhltable_destroy(&table->objname_ht); +err_obj_ht: rhltable_destroy(&table->chains_ht); err_chain_ht: kfree(table->udata); @@ -1819,6 +1823,7 @@ static void nf_tables_table_destroy(struct nft_table *table) return; rhltable_destroy(&table->chains_ht); + rhltable_destroy(&table->objname_ht); kfree(table->name); kfree(table->udata); kfree(table); @@ -8086,7 +8091,7 @@ void nft_unregister_obj(struct nft_object_type *obj_type) EXPORT_SYMBOL_GPL(nft_unregister_obj); struct nft_object *nft_obj_lookup(const struct net *net, - const struct nft_table *table, + struct nft_table *table, const struct nlattr *nla, u32 objtype, u8 genmask) { @@ -8102,7 +8107,7 @@ struct nft_object *nft_obj_lookup(const struct net *net, !lockdep_commit_lock_is_held(net)); rcu_read_lock(); - list = rhltable_lookup(&nft_objname_ht, &k, nft_objname_ht_params); + list = rhltable_lookup(&table->objname_ht, &k, nft_objname_ht_params); if (!list) goto out; @@ -8382,7 +8387,7 @@ static int nf_tables_newobj(struct sk_buff *skb, const struct nfnl_info *info, if (err < 0) goto err_trans; - err = rhltable_insert(&nft_objname_ht, &obj->rhlhead, + err = rhltable_insert(&table->objname_ht, &obj->rhlhead, nft_objname_ht_params); if (err < 0) goto err_obj_ht; @@ -8567,8 +8572,8 @@ nf_tables_getobj_single(u32 portid, const struct nfnl_info *info, struct netlink_ext_ack *extack = info->extack; u8 genmask = nft_genmask_cur(info->net); u8 family = info->nfmsg->nfgen_family; - const struct nft_table *table; struct net *net = info->net; + struct nft_table *table; struct nft_object *obj; struct sk_buff *skb2; u32 objtype; @@ -10437,9 +10442,9 @@ static void nf_tables_commit_chain(struct net *net, struct nft_chain *chain) nf_tables_commit_chain_free_rules_old(g0); } -static void nft_obj_del(struct nft_object *obj) +static void nft_obj_del(struct nft_table *table, struct nft_object *obj) { - rhltable_remove(&nft_objname_ht, &obj->rhlhead, nft_objname_ht_params); + rhltable_remove(&table->objname_ht, &obj->rhlhead, nft_objname_ht_params); list_del_rcu(&obj->list); } @@ -11124,7 +11129,7 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb) break; case NFT_MSG_DELOBJ: case NFT_MSG_DESTROYOBJ: - nft_obj_del(nft_trans_obj(trans)); + nft_obj_del(table, nft_trans_obj(trans)); nf_tables_obj_notify(&ctx, nft_trans_obj(trans), trans->msg_type); break; @@ -11416,7 +11421,7 @@ static int __nf_tables_abort(struct net *net, enum nfnl_abort_action action) nft_trans_destroy(trans); } else { nft_use_dec_restore(&table->use); - nft_obj_del(nft_trans_obj(trans)); + nft_obj_del(table, nft_trans_obj(trans)); } break; case NFT_MSG_DELOBJ: @@ -12043,7 +12048,7 @@ static void __nft_release_table(struct net *net, struct nft_table *table) nft_set_destroy(&ctx, set); } list_for_each_entry_safe(obj, ne, &table->objects, list) { - nft_obj_del(obj); + nft_obj_del(table, obj); nft_use_dec(&table->use); nft_obj_destroy(&ctx, obj); } @@ -12225,10 +12230,6 @@ static int __init nf_tables_module_init(void) if (err < 0) goto err_netdev_notifier; - err = rhltable_init(&nft_objname_ht, &nft_objname_ht_params); - if (err < 0) - goto err_rht_objname; - err = nft_offload_init(); if (err < 0) goto err_offload; @@ -12251,8 +12252,6 @@ err_nfnl_subsys: err_netlink_notifier: nft_offload_exit(); err_offload: - rhltable_destroy(&nft_objname_ht); -err_rht_objname: unregister_netdevice_notifier(&nf_tables_flowtable_notifier); err_netdev_notifier: nf_tables_core_module_exit(); @@ -12274,7 +12273,6 @@ static void __exit nf_tables_module_exit(void) unregister_pernet_subsys(&nf_tables_net_ops); cancel_work_sync(&trans_gc_work); rcu_barrier(); - rhltable_destroy(&nft_objname_ht); nf_tables_core_module_exit(); } -- cgit From e876b75b9020a97bbdc79721e7fc749024891c65 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Wed, 22 Jul 2026 13:15:15 +0300 Subject: ipvs: fix the checksum validations ip_vs_in_icmp_v6() is missing checksum validation for ICMPv6 packets from clients. In fact, as for TCP/UDP we should validate the checksum for ICMP packets only when we mangle the packets on MASQ or on reply for tunnel. Also, Sashiko points out that handle_response_icmp() being common for IPv4 and IPv6 is missing the pseudo-header calculation while validating ICMPv6 messages from real servers which is a problem if checksum is not validated by the hardware. Fix the problems by creating ip_vs_checksum_common_check() helper and use it for TCP/UDP/ICMP both for IPv4 and IPv6. Rely on the nf_checksum() for validating the ICMP messages but use it also for TCP and UDP. Use correct IP offset for IP_VS_DBG_RL_PKT for TCP/UDP/SCTP. IPVS packets (TCP/UDP/SCTP/ICMP) do not need checksum validation on LOCAL_OUT (local clients or local real servers) and on FORWARD (traffic from servers on LAN). Do it only on LOCAL_IN, in case nf_checksum() is not called on PRE_ROUTING. Also, ip_vs_checksum_complete() can be marked static. Fixes: 2a3b791e6e11 ("IPVS: Add/adjust Netfilter hook functions and helpers for v6") Link: https://sashiko.dev/#/patchset/20260708180315.77413-1-ja%40ssi.bg Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- include/net/ip_vs.h | 31 ++++++++++++++++++++-- net/netfilter/ipvs/ip_vs_core.c | 20 +++++++++++--- net/netfilter/ipvs/ip_vs_proto_sctp.c | 15 ++++++----- net/netfilter/ipvs/ip_vs_proto_tcp.c | 44 ++++++------------------------ net/netfilter/ipvs/ip_vs_proto_udp.c | 50 ++++++++--------------------------- 5 files changed, 74 insertions(+), 86 deletions(-) (limited to 'include') diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 417ff51f62fc..d8f9ddb0fb38 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -25,7 +25,9 @@ #include /* for union nf_inet_addr */ #include #include /* for struct ipv6hdr */ +#include #include +#include #if IS_ENABLED(CONFIG_NF_CONNTRACK) #include #endif @@ -2066,8 +2068,6 @@ void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, struct ip_vs_conn *cp, int dir); #endif -__sum16 ip_vs_checksum_complete(struct sk_buff *skb, int offset); - static inline __wsum ip_vs_check_diff4(__be32 old, __be32 new, __wsum oldsum) { __be32 diff[2] = { ~old, new }; @@ -2093,6 +2093,33 @@ static inline __wsum ip_vs_check_diff2(__be16 old, __be16 new, __wsum oldsum) return csum_partial(diff, sizeof(diff), oldsum); } +static inline bool ip_vs_checksum_needed(struct sk_buff *skb, int af) +{ + /* Checksum unnecessary or already validated? */ + if (skb_csum_unnecessary(skb)) + return false; + /* LOCAL_OUT ? */ + if (!skb->dev || skb->dev->flags & IFF_LOOPBACK) + return false; + /* !LOCAL_IN (FORWARD) ? */ + if (af == AF_INET6) { + if (!(dst_rt6_info(skb_dst(skb))->rt6i_flags & RTF_LOCAL)) + return false; + } else { + if (!(skb_rtable(skb)->rt_flags & RTCF_LOCAL)) + return false; + } + return true; +} + +static inline bool ip_vs_checksum_common_check(struct sk_buff *skb, + int offset, int proto, int af) +{ + if (!ip_vs_checksum_needed(skb, af)) + return true; + return !nf_checksum(skb, NF_INET_LOCAL_IN, offset, proto, af); +} + /* Forget current conntrack (unconfirmed) and attach notrack entry */ static inline void ip_vs_notrack(struct sk_buff *skb) { diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index bafab93451d0..c8b512725e6e 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -867,7 +867,7 @@ static int sysctl_nat_icmp_send(struct netns_ipvs *ipvs) { return 0; } #endif -__sum16 ip_vs_checksum_complete(struct sk_buff *skb, int offset) +static __sum16 ip_vs_checksum_complete(struct sk_buff *skb, int offset) { return csum_fold(skb_checksum(skb, offset, skb->len - offset, 0)); } @@ -1038,13 +1038,14 @@ static int handle_response_icmp(int af, struct sk_buff *skb, unsigned int offset, unsigned int ihl, unsigned int hooknum) { + int iproto = af == AF_INET6 ? IPPROTO_ICMPV6 : IPPROTO_ICMP; unsigned int verdict = NF_DROP; if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) goto after_nat; /* Ensure the checksum is correct */ - if (!skb_csum_unnecessary(skb) && ip_vs_checksum_complete(skb, ihl)) { + if (!ip_vs_checksum_common_check(skb, ihl, iproto, af)) { /* Failed checksum! */ IP_VS_DBG_BUF(1, "Forward ICMP: failed checksum from %s!\n", IP_VS_DBG_ADDR(af, snet)); @@ -1898,7 +1899,8 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, verdict = NF_DROP; /* Ensure the checksum is correct */ - if (!skb_csum_unnecessary(skb) && ip_vs_checksum_complete(skb, ihl)) { + if ((IP_VS_FWD_METHOD(cp) == IP_VS_CONN_F_MASQ || tunnel) && + !ip_vs_checksum_common_check(skb, ihl, IPPROTO_ICMP, AF_INET)) { /* Failed checksum! */ IP_VS_DBG(1, "Incoming ICMP: failed checksum from %pI4!\n", &iph->saddr); @@ -2064,6 +2066,18 @@ static int ip_vs_in_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, goto out; } + verdict = NF_DROP; + + /* Ensure the checksum is correct */ + if (IP_VS_FWD_METHOD(cp) == IP_VS_CONN_F_MASQ && + !ip_vs_checksum_common_check(skb, iph->len, IPPROTO_ICMPV6, + AF_INET6)) { + /* Failed checksum! */ + IP_VS_DBG(1, "Incoming ICMPv6: failed checksum from %pI6c!\n", + &iph->saddr); + goto out; + } + /* do the statistics and put it back */ ip_vs_in_stats(cp, skb); diff --git a/net/netfilter/ipvs/ip_vs_proto_sctp.c b/net/netfilter/ipvs/ip_vs_proto_sctp.c index c67317be17df..f6f732b7dfa8 100644 --- a/net/netfilter/ipvs/ip_vs_proto_sctp.c +++ b/net/netfilter/ipvs/ip_vs_proto_sctp.c @@ -11,7 +11,7 @@ static int sctp_csum_check(int af, struct sk_buff *skb, struct ip_vs_protocol *pp, - unsigned int sctphoff); + struct ip_vs_iphdr *iph); static int sctp_conn_schedule(struct netns_ipvs *ipvs, int af, struct sk_buff *skb, @@ -109,7 +109,7 @@ sctp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, int ret; /* Some checks before mangling */ - if (!sctp_csum_check(cp->af, skb, pp, sctphoff)) + if (!sctp_csum_check(cp->af, skb, pp, iph)) return 0; /* Call application helper if needed */ @@ -157,7 +157,7 @@ sctp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, int ret; /* Some checks before mangling */ - if (!sctp_csum_check(cp->af, skb, pp, sctphoff)) + if (!sctp_csum_check(cp->af, skb, pp, iph)) return 0; /* Call application helper if needed */ @@ -187,19 +187,22 @@ sctp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, static int sctp_csum_check(int af, struct sk_buff *skb, struct ip_vs_protocol *pp, - unsigned int sctphoff) + struct ip_vs_iphdr *iph) { + unsigned int sctphoff = iph->len; struct sctphdr *sh; __le32 cmp, val; + if (!ip_vs_checksum_needed(skb, af)) + return 1; sh = (struct sctphdr *)(skb->data + sctphoff); cmp = sh->checksum; val = sctp_compute_cksum(skb, sctphoff); if (val != cmp) { /* CRC failure, dump it. */ - IP_VS_DBG_RL_PKT(0, af, pp, skb, 0, - "Failed checksum for"); + IP_VS_DBG_RL_PKT(0, af, pp, skb, iph->off, + "Failed checksum for"); return 0; } return 1; diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c index f86b763efcc4..533fce3e5e4e 100644 --- a/net/netfilter/ipvs/ip_vs_proto_tcp.c +++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c @@ -29,7 +29,7 @@ static int tcp_csum_check(int af, struct sk_buff *skb, struct ip_vs_protocol *pp, - unsigned int tcphoff); + struct ip_vs_iphdr *iph); static int tcp_conn_schedule(struct netns_ipvs *ipvs, int af, struct sk_buff *skb, @@ -166,7 +166,7 @@ tcp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, int ret; /* Some checks before mangling */ - if (!tcp_csum_check(cp->af, skb, pp, tcphoff)) + if (!tcp_csum_check(cp->af, skb, pp, iph)) return 0; /* Call application helper if needed */ @@ -244,7 +244,7 @@ tcp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, int ret; /* Some checks before mangling */ - if (!tcp_csum_check(cp->af, skb, pp, tcphoff)) + if (!tcp_csum_check(cp->af, skb, pp, iph)) return 0; /* @@ -302,41 +302,13 @@ tcp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, static int tcp_csum_check(int af, struct sk_buff *skb, struct ip_vs_protocol *pp, - unsigned int tcphoff) + struct ip_vs_iphdr *iph) { - switch (skb->ip_summed) { - case CHECKSUM_NONE: - skb->csum = skb_checksum(skb, tcphoff, skb->len - tcphoff, 0); - fallthrough; - case CHECKSUM_COMPLETE: -#ifdef CONFIG_IP_VS_IPV6 - if (af == AF_INET6) { - if (csum_ipv6_magic(&ipv6_hdr(skb)->saddr, - &ipv6_hdr(skb)->daddr, - skb->len - tcphoff, - IPPROTO_TCP, - skb->csum)) { - IP_VS_DBG_RL_PKT(0, af, pp, skb, 0, - "Failed checksum for"); - return 0; - } - } else -#endif - if (csum_tcpudp_magic(ip_hdr(skb)->saddr, - ip_hdr(skb)->daddr, - skb->len - tcphoff, - ip_hdr(skb)->protocol, - skb->csum)) { - IP_VS_DBG_RL_PKT(0, af, pp, skb, 0, - "Failed checksum for"); - return 0; - } - break; - default: - /* No need to checksum. */ - break; + if (!ip_vs_checksum_common_check(skb, iph->len, IPPROTO_TCP, af)) { + IP_VS_DBG_RL_PKT(0, af, pp, skb, iph->off, + "Failed checksum for"); + return 0; } - return 1; } diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c index 58f9e255927e..de3597347542 100644 --- a/net/netfilter/ipvs/ip_vs_proto_udp.c +++ b/net/netfilter/ipvs/ip_vs_proto_udp.c @@ -25,7 +25,7 @@ static int udp_csum_check(int af, struct sk_buff *skb, struct ip_vs_protocol *pp, - unsigned int udphoff); + struct ip_vs_iphdr *iph); static int udp_conn_schedule(struct netns_ipvs *ipvs, int af, struct sk_buff *skb, @@ -155,7 +155,7 @@ udp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, int ret; /* Some checks before mangling */ - if (!udp_csum_check(cp->af, skb, pp, udphoff)) + if (!udp_csum_check(cp->af, skb, pp, iph)) return 0; /* @@ -238,7 +238,7 @@ udp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, int ret; /* Some checks before mangling */ - if (!udp_csum_check(cp->af, skb, pp, udphoff)) + if (!udp_csum_check(cp->af, skb, pp, iph)) return 0; /* @@ -298,48 +298,20 @@ udp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, static int udp_csum_check(int af, struct sk_buff *skb, struct ip_vs_protocol *pp, - unsigned int udphoff) + struct ip_vs_iphdr *iph) { struct udphdr _udph, *uh; - uh = skb_header_pointer(skb, udphoff, sizeof(_udph), &_udph); + uh = skb_header_pointer(skb, iph->len, sizeof(_udph), &_udph); if (uh == NULL) return 0; - if (uh->check != 0) { - switch (skb->ip_summed) { - case CHECKSUM_NONE: - skb->csum = skb_checksum(skb, udphoff, - skb->len - udphoff, 0); - fallthrough; - case CHECKSUM_COMPLETE: -#ifdef CONFIG_IP_VS_IPV6 - if (af == AF_INET6) { - if (csum_ipv6_magic(&ipv6_hdr(skb)->saddr, - &ipv6_hdr(skb)->daddr, - skb->len - udphoff, - IPPROTO_UDP, - skb->csum)) { - IP_VS_DBG_RL_PKT(0, af, pp, skb, 0, - "Failed checksum for"); - return 0; - } - } else -#endif - if (csum_tcpudp_magic(ip_hdr(skb)->saddr, - ip_hdr(skb)->daddr, - skb->len - udphoff, - ip_hdr(skb)->protocol, - skb->csum)) { - IP_VS_DBG_RL_PKT(0, af, pp, skb, 0, - "Failed checksum for"); - return 0; - } - break; - default: - /* No need to checksum. */ - break; - } + if (!uh->check) + return 1; + if (!ip_vs_checksum_common_check(skb, iph->len, IPPROTO_UDP, af)) { + IP_VS_DBG_RL_PKT(0, af, pp, skb, iph->off, + "Failed checksum for"); + return 0; } return 1; } -- cgit From 15cab31a3730e05f0767b922a7450e5d784b2607 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Wed, 22 Jul 2026 13:15:16 +0300 Subject: ipvs: fix places with wrong packet offsets The offsets we use to packet headers and payloads should be based on skb->data. We even already respect non-zero network offset in ip_vs_fill_iph_skb() but some places do it wrongly and support only zero offset which is expected for the IP layer where IPVS has hooks. Change all places that instead of skb->data use offsets based on the network header (skb_network_header, ip_hdr, etc) because this doubles the network offset as noted by Sashiko. For ip_vs_nat_icmp_v6() we can even rely on the IPv6 header parsing done by the caller. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Link: https://sashiko.dev/#/patchset/20260710143733.29741-2-fw%40strlen.de Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- include/net/ip_vs.h | 15 ++-- net/netfilter/ipvs/ip_vs_app.c | 4 +- net/netfilter/ipvs/ip_vs_core.c | 133 +++++++++++++++++----------------- net/netfilter/ipvs/ip_vs_proto_sctp.c | 4 +- net/netfilter/ipvs/ip_vs_proto_tcp.c | 4 +- net/netfilter/ipvs/ip_vs_proto_udp.c | 4 +- net/netfilter/ipvs/ip_vs_xmit.c | 26 ++++--- 7 files changed, 97 insertions(+), 93 deletions(-) (limited to 'include') diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index d8f9ddb0fb38..4a10a01d6e2f 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -1974,8 +1974,9 @@ int ip_vs_tunnel_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, int ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph); int ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, - struct ip_vs_protocol *pp, int offset, - unsigned int hooknum, struct ip_vs_iphdr *iph); + struct ip_vs_protocol *pp, unsigned int toff, + unsigned int wlen, unsigned int hooknum, + struct ip_vs_iphdr *ciph); void ip_vs_dest_dst_rcu_free(struct rcu_head *head); #ifdef CONFIG_IP_VS_IPV6 @@ -1988,8 +1989,9 @@ int ip_vs_tunnel_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, int ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph); int ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, - struct ip_vs_protocol *pp, int offset, - unsigned int hooknum, struct ip_vs_iphdr *iph); + struct ip_vs_protocol *pp, unsigned int toff, + unsigned int wlen, unsigned int hooknum, + struct ip_vs_iphdr *ciph); #endif #ifdef CONFIG_SYSCTL @@ -2061,11 +2063,12 @@ static inline bool ip_vs_conn_use_hash2(struct ip_vs_conn *cp) } void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp, - struct ip_vs_conn *cp, int dir); + struct ip_vs_conn *cp, int dir, unsigned int toff); #ifdef CONFIG_IP_VS_IPV6 void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, - struct ip_vs_conn *cp, int dir); + struct ip_vs_conn *cp, int dir, unsigned int toff, + struct ip_vs_iphdr *ciph); #endif static inline __wsum ip_vs_check_diff4(__be32 old, __be32 new, __wsum oldsum) diff --git a/net/netfilter/ipvs/ip_vs_app.c b/net/netfilter/ipvs/ip_vs_app.c index b0e00be85cb1..11cbdbaf561d 100644 --- a/net/netfilter/ipvs/ip_vs_app.c +++ b/net/netfilter/ipvs/ip_vs_app.c @@ -367,7 +367,7 @@ static inline int app_tcp_pkt_out(struct ip_vs_conn *cp, struct sk_buff *skb, if (skb_ensure_writable(skb, ipvsh->len + sizeof(*th))) return 0; - th = (struct tcphdr *)(skb_network_header(skb) + ipvsh->len); + th = (struct tcphdr *)(skb->data + ipvsh->len); /* * Remember seq number in case this pkt gets resized @@ -443,7 +443,7 @@ static inline int app_tcp_pkt_in(struct ip_vs_conn *cp, struct sk_buff *skb, if (skb_ensure_writable(skb, ipvsh->len + sizeof(*th))) return 0; - th = (struct tcphdr *)(skb_network_header(skb) + ipvsh->len); + th = (struct tcphdr *)(skb->data + ipvsh->len); /* * Remember seq number in case this pkt gets resized diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index c8b512725e6e..cd5eb71543ec 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -924,13 +924,12 @@ static int ip_vs_route_me_harder(struct netns_ipvs *ipvs, int af, * - inout: 1=in->out, 0=out->in */ void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp, - struct ip_vs_conn *cp, int inout) + struct ip_vs_conn *cp, int inout, unsigned int toff) { struct iphdr *iph = ip_hdr(skb); - unsigned int icmp_offset = iph->ihl*4; - struct icmphdr *icmph = (struct icmphdr *)(skb_network_header(skb) + - icmp_offset); + struct icmphdr *icmph = (struct icmphdr *)(skb->data + toff); struct iphdr *ciph = (struct iphdr *)(icmph + 1); + unsigned int coff __maybe_unused = toff + sizeof(struct icmphdr); if (inout) { iph->saddr = cp->vaddr.ip; @@ -957,48 +956,45 @@ void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp, /* And finally the ICMP checksum */ icmph->checksum = 0; - icmph->checksum = ip_vs_checksum_complete(skb, icmp_offset); + icmph->checksum = ip_vs_checksum_complete(skb, toff); skb->ip_summed = CHECKSUM_UNNECESSARY; if (inout) - IP_VS_DBG_PKT(11, AF_INET, pp, skb, (void *)ciph - (void *)iph, - "Forwarding altered outgoing ICMP"); + IP_VS_DBG_PKT(11, AF_INET, pp, skb, coff, + "Forwarding altered outgoing ICMP"); else - IP_VS_DBG_PKT(11, AF_INET, pp, skb, (void *)ciph - (void *)iph, - "Forwarding altered incoming ICMP"); + IP_VS_DBG_PKT(11, AF_INET, pp, skb, coff, + "Forwarding altered incoming ICMP"); } #ifdef CONFIG_IP_VS_IPV6 void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, - struct ip_vs_conn *cp, int inout) + struct ip_vs_conn *cp, int inout, unsigned int toff, + struct ip_vs_iphdr *ciph) { struct ipv6hdr *iph = ipv6_hdr(skb); - unsigned int icmp_offset = 0; - unsigned int offs = 0; /* header offset*/ int protocol; struct icmp6hdr *icmph; - struct ipv6hdr *ciph; - unsigned short fragoffs; + struct ipv6hdr *cih; - ipv6_find_hdr(skb, &icmp_offset, IPPROTO_ICMPV6, &fragoffs, NULL); - icmph = (struct icmp6hdr *)(skb_network_header(skb) + icmp_offset); - offs = icmp_offset + sizeof(struct icmp6hdr); - ciph = (struct ipv6hdr *)(skb_network_header(skb) + offs); + icmph = (struct icmp6hdr *)(skb->data + toff); + cih = (struct ipv6hdr *)(skb->data + ciph->off); - protocol = ipv6_find_hdr(skb, &offs, -1, &fragoffs, NULL); + protocol = ciph->protocol; if (inout) { iph->saddr = cp->vaddr.in6; - ciph->daddr = cp->vaddr.in6; + cih->daddr = cp->vaddr.in6; } else { iph->daddr = cp->daddr.in6; - ciph->saddr = cp->daddr.in6; + cih->saddr = cp->daddr.in6; } /* the TCP/UDP/SCTP port */ - if (!fragoffs && (IPPROTO_TCP == protocol || IPPROTO_UDP == protocol || - IPPROTO_SCTP == protocol)) { - __be16 *ports = (void *)(skb_network_header(skb) + offs); + if (!ciph->fragoffs && + (protocol == IPPROTO_TCP || protocol == IPPROTO_UDP || + protocol == IPPROTO_SCTP)) { + __be16 *ports = (void *)(skb->data + ciph->len); IP_VS_DBG(11, "%s() changed port %d to %d\n", __func__, ntohs(inout ? ports[1] : ports[0]), @@ -1011,19 +1007,17 @@ void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, /* And finally the ICMP checksum */ icmph->icmp6_cksum = ~csum_ipv6_magic(&iph->saddr, &iph->daddr, - skb->len - icmp_offset, + skb->len - toff, IPPROTO_ICMPV6, 0); - skb->csum_start = skb_network_header(skb) - skb->head + icmp_offset; + skb->csum_start = skb_headroom(skb) + toff; skb->csum_offset = offsetof(struct icmp6hdr, icmp6_cksum); skb->ip_summed = CHECKSUM_PARTIAL; if (inout) - IP_VS_DBG_PKT(11, AF_INET6, pp, skb, - (void *)ciph - (void *)iph, + IP_VS_DBG_PKT(11, AF_INET6, pp, skb, ciph->off, "Forwarding altered outgoing ICMPv6"); else - IP_VS_DBG_PKT(11, AF_INET6, pp, skb, - (void *)ciph - (void *)iph, + IP_VS_DBG_PKT(11, AF_INET6, pp, skb, ciph->off, "Forwarding altered incoming ICMPv6"); } #endif @@ -1033,37 +1027,38 @@ void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, */ static int handle_response_icmp(int af, struct sk_buff *skb, union nf_inet_addr *snet, - __u8 protocol, struct ip_vs_conn *cp, + struct ip_vs_conn *cp, struct ip_vs_protocol *pp, - unsigned int offset, unsigned int ihl, - unsigned int hooknum) + struct ip_vs_iphdr *ciph, + unsigned int toff, unsigned int hooknum) { int iproto = af == AF_INET6 ? IPPROTO_ICMPV6 : IPPROTO_ICMP; unsigned int verdict = NF_DROP; + unsigned int ctoff = ciph->len; if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) goto after_nat; /* Ensure the checksum is correct */ - if (!ip_vs_checksum_common_check(skb, ihl, iproto, af)) { + if (!ip_vs_checksum_common_check(skb, toff, iproto, af)) { /* Failed checksum! */ IP_VS_DBG_BUF(1, "Forward ICMP: failed checksum from %s!\n", IP_VS_DBG_ADDR(af, snet)); goto out; } - if (IPPROTO_TCP == protocol || IPPROTO_UDP == protocol || - IPPROTO_SCTP == protocol) - offset += 2 * sizeof(__u16); - if (skb_ensure_writable(skb, offset)) + if (ciph->protocol == IPPROTO_TCP || ciph->protocol == IPPROTO_UDP || + ciph->protocol == IPPROTO_SCTP) + ctoff += 2 * sizeof(__u16); + if (skb_ensure_writable(skb, ctoff)) goto out; #ifdef CONFIG_IP_VS_IPV6 if (af == AF_INET6) - ip_vs_nat_icmp_v6(skb, pp, cp, 1); + ip_vs_nat_icmp_v6(skb, pp, cp, 1, toff, ciph); else #endif - ip_vs_nat_icmp(skb, pp, cp, 1); + ip_vs_nat_icmp(skb, pp, cp, 1, toff); if (ip_vs_route_me_harder(cp->ipvs, af, skb, hooknum)) goto out; @@ -1091,9 +1086,9 @@ out: * Currently handles error types - unreachable, quench, ttl exceeded. */ static int ip_vs_out_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, - int *related, unsigned int hooknum) + int *related, unsigned int hooknum, + struct ip_vs_iphdr *ipvsh) { - struct iphdr *iph; struct icmphdr _icmph, *ic; struct iphdr _ciph, *cih; /* The ip header contained within the ICMP */ struct ip_vs_iphdr ciph; @@ -1108,17 +1103,19 @@ static int ip_vs_out_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, if (ip_is_fragment(ip_hdr(skb))) { if (ip_vs_gather_frags(ipvs, skb, ip_vs_defrag_user(hooknum))) return NF_STOLEN; + if (!ip_vs_fill_iph_skb(AF_INET, skb, false, ipvsh)) + return NF_ACCEPT; } - iph = ip_hdr(skb); - offset = ihl = iph->ihl * 4; + ihl = ipvsh->len; + offset = ipvsh->len; ic = skb_header_pointer(skb, offset, sizeof(_icmph), &_icmph); if (ic == NULL) return NF_DROP; IP_VS_DBG(12, "Outgoing ICMP (%d,%d) %pI4->%pI4\n", ic->type, ntohs(icmp_id(ic)), - &iph->saddr, &iph->daddr); + &ipvsh->saddr.ip, &ipvsh->daddr.ip); /* * Work through seeing if this is for us. @@ -1137,7 +1134,7 @@ static int ip_vs_out_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, /* Now find the contained IP header */ offset += sizeof(_icmph); cih = skb_header_pointer(skb, offset, sizeof(_ciph), &_ciph); - if (cih == NULL) + if (!(cih && cih->version == 4 && cih->ihl >= 5)) return NF_ACCEPT; /* The packet looks wrong, ignore */ pp = ip_vs_proto_get(cih->protocol); @@ -1160,9 +1157,9 @@ static int ip_vs_out_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, if (!cp) return NF_ACCEPT; - snet.ip = iph->saddr; - return handle_response_icmp(AF_INET, skb, &snet, cih->protocol, cp, - pp, ciph.len, ihl, hooknum); + snet.ip = ipvsh->saddr.ip; + return handle_response_icmp(AF_INET, skb, &snet, cp, pp, &ciph, ihl, + hooknum); } #ifdef CONFIG_IP_VS_IPV6 @@ -1175,7 +1172,6 @@ static int ip_vs_out_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, struct ip_vs_conn *cp; struct ip_vs_protocol *pp; union nf_inet_addr snet; - unsigned int offset; *related = 1; ic = frag_safe_skb_hp(skb, ipvsh->len, sizeof(_icmph), &_icmph); @@ -1218,9 +1214,8 @@ static int ip_vs_out_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, return NF_ACCEPT; snet.in6 = ciph.saddr.in6; - offset = ciph.len; - return handle_response_icmp(AF_INET6, skb, &snet, ciph.protocol, cp, - pp, offset, ipvsh->len, hooknum); + return handle_response_icmp(AF_INET6, skb, &snet, cp, pp, &ciph, + ipvsh->len, hooknum); } #endif @@ -1546,7 +1541,8 @@ ip_vs_out_hook(void *priv, struct sk_buff *skb, const struct nf_hook_state *stat #endif if (unlikely(iph.protocol == IPPROTO_ICMP)) { int related; - int verdict = ip_vs_out_icmp(ipvs, skb, &related, hooknum); + int verdict = ip_vs_out_icmp(ipvs, skb, &related, + hooknum, &iph); if (related) return verdict; @@ -1754,9 +1750,8 @@ unk: */ static int ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, - unsigned int hooknum) + unsigned int hooknum, struct ip_vs_iphdr *iph) { - struct iphdr *iph; struct icmphdr _icmph, *ic; struct iphdr _ciph, *cih; /* The ip header contained within the ICMP */ struct ip_vs_iphdr ciph; @@ -1766,7 +1761,7 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, unsigned int offset, offset2, ihl, verdict; bool tunnel, new_cp = false; union nf_inet_addr *raddr; - char *outer_proto = "IPIP"; + char *outer_proto __maybe_unused = "IPIP"; unsigned int hlen_ipip; int ulen = 0; @@ -1776,17 +1771,19 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, if (ip_is_fragment(ip_hdr(skb))) { if (ip_vs_gather_frags(ipvs, skb, ip_vs_defrag_user(hooknum))) return NF_STOLEN; + if (!ip_vs_fill_iph_skb(AF_INET, skb, false, iph)) + return NF_ACCEPT; } - iph = ip_hdr(skb); - offset = ihl = iph->ihl * 4; + ihl = iph->len; + offset = iph->len; ic = skb_header_pointer(skb, offset, sizeof(_icmph), &_icmph); if (ic == NULL) return NF_DROP; IP_VS_DBG(12, "Incoming ICMP (%d,%d) %pI4->%pI4\n", ic->type, ntohs(icmp_id(ic)), - &iph->saddr, &iph->daddr); + &iph->saddr.ip, &iph->daddr.ip); /* * Work through seeing if this is for us. @@ -1903,7 +1900,7 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, !ip_vs_checksum_common_check(skb, ihl, IPPROTO_ICMP, AF_INET)) { /* Failed checksum! */ IP_VS_DBG(1, "Incoming ICMP: failed checksum from %pI4!\n", - &iph->saddr); + &iph->saddr.ip); goto out; } @@ -1974,7 +1971,8 @@ ignore_tunnel: if (IPPROTO_TCP == cih->protocol || IPPROTO_UDP == cih->protocol || IPPROTO_SCTP == cih->protocol) offset += 2 * sizeof(__u16); - verdict = ip_vs_icmp_xmit(skb, cp, pp, offset, hooknum, &ciph); + verdict = ip_vs_icmp_xmit(skb, cp, pp, iph->len, offset, hooknum, + &ciph); out: if (likely(!new_cp)) @@ -2087,7 +2085,8 @@ static int ip_vs_in_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, IPPROTO_SCTP == ciph.protocol) offset += 2 * sizeof(__u16); /* Also mangle ports */ - verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, offset, hooknum, &ciph); + verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, iph->len, offset, hooknum, + &ciph); out: if (likely(!new_cp)) @@ -2166,7 +2165,7 @@ ip_vs_in_hook(void *priv, struct sk_buff *skb, const struct nf_hook_state *state if (unlikely(iph.protocol == IPPROTO_ICMP)) { int related; int verdict = ip_vs_in_icmp(ipvs, skb, &related, - hooknum); + hooknum, &iph); if (related) return verdict; @@ -2302,6 +2301,7 @@ ip_vs_forward_icmp(void *priv, struct sk_buff *skb, const struct nf_hook_state *state) { struct netns_ipvs *ipvs = net_ipvs(state->net); + struct ip_vs_iphdr iphdr; int r; /* ipvs enabled in this netns ? */ @@ -2311,10 +2311,9 @@ ip_vs_forward_icmp(void *priv, struct sk_buff *skb, if (state->pf == NFPROTO_IPV4) { if (ip_hdr(skb)->protocol != IPPROTO_ICMP) return NF_ACCEPT; + ip_vs_fill_iph_skb(AF_INET, skb, false, &iphdr); #ifdef CONFIG_IP_VS_IPV6 } else { - struct ip_vs_iphdr iphdr; - ip_vs_fill_iph_skb(AF_INET6, skb, false, &iphdr); if (iphdr.protocol != IPPROTO_ICMPV6) @@ -2324,7 +2323,7 @@ ip_vs_forward_icmp(void *priv, struct sk_buff *skb, #endif } - return ip_vs_in_icmp(ipvs, skb, &r, state->hook); + return ip_vs_in_icmp(ipvs, skb, &r, state->hook, &iphdr); } static const struct nf_hook_ops ip_vs_ops4[] = { diff --git a/net/netfilter/ipvs/ip_vs_proto_sctp.c b/net/netfilter/ipvs/ip_vs_proto_sctp.c index f6f732b7dfa8..3dbd3096e163 100644 --- a/net/netfilter/ipvs/ip_vs_proto_sctp.c +++ b/net/netfilter/ipvs/ip_vs_proto_sctp.c @@ -121,7 +121,7 @@ sctp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, payload_csum = true; } - sctph = (void *) skb_network_header(skb) + sctphoff; + sctph = (void *)skb->data + sctphoff; /* Only update csum if we really have to */ if (sctph->source != cp->vport || payload_csum || @@ -169,7 +169,7 @@ sctp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, payload_csum = true; } - sctph = (void *) skb_network_header(skb) + sctphoff; + sctph = (void *)skb->data + sctphoff; /* Only update csum if we really have to */ if (sctph->dest != cp->dport || payload_csum || diff --git a/net/netfilter/ipvs/ip_vs_proto_tcp.c b/net/netfilter/ipvs/ip_vs_proto_tcp.c index 533fce3e5e4e..99a286fdc90c 100644 --- a/net/netfilter/ipvs/ip_vs_proto_tcp.c +++ b/net/netfilter/ipvs/ip_vs_proto_tcp.c @@ -179,7 +179,7 @@ tcp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, payload_csum = true; } - tcph = (void *)skb_network_header(skb) + tcphoff; + tcph = (void *)skb->data + tcphoff; tcph->source = cp->vport; /* Adjust TCP checksums */ @@ -260,7 +260,7 @@ tcp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, payload_csum = true; } - tcph = (void *)skb_network_header(skb) + tcphoff; + tcph = (void *)skb->data + tcphoff; tcph->dest = cp->dport; /* diff --git a/net/netfilter/ipvs/ip_vs_proto_udp.c b/net/netfilter/ipvs/ip_vs_proto_udp.c index de3597347542..f32785682402 100644 --- a/net/netfilter/ipvs/ip_vs_proto_udp.c +++ b/net/netfilter/ipvs/ip_vs_proto_udp.c @@ -170,7 +170,7 @@ udp_snat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, payload_csum = true; } - udph = (void *)skb_network_header(skb) + udphoff; + udph = (void *)skb->data + udphoff; udph->source = cp->vport; /* @@ -254,7 +254,7 @@ udp_dnat_handler(struct sk_buff *skb, struct ip_vs_protocol *pp, payload_csum = true; } - udph = (void *)skb_network_header(skb) + udphoff; + udph = (void *)skb->data + udphoff; udph->dest = cp->dport; /* diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c index 9fef4335da13..c23401c789de 100644 --- a/net/netfilter/ipvs/ip_vs_xmit.c +++ b/net/netfilter/ipvs/ip_vs_xmit.c @@ -1502,8 +1502,9 @@ tx_error: */ int ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, - struct ip_vs_protocol *pp, int offset, unsigned int hooknum, - struct ip_vs_iphdr *iph) + struct ip_vs_protocol *pp, unsigned int toff, + unsigned int wlen, unsigned int hooknum, + struct ip_vs_iphdr *ciph) { struct rtable *rt; /* Route to the other host */ int rc; @@ -1515,7 +1516,7 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, translate address/port back */ if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) { if (cp->packet_xmit) - rc = cp->packet_xmit(skb, cp, pp, iph); + rc = cp->packet_xmit(skb, cp, pp, ciph); else rc = NF_ACCEPT; /* do not touch skb anymore */ @@ -1533,7 +1534,7 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, IP_VS_RT_MODE_LOCAL | IP_VS_RT_MODE_NON_LOCAL | IP_VS_RT_MODE_RDR : IP_VS_RT_MODE_NON_LOCAL; local = __ip_vs_get_out_rt(cp->ipvs, cp->af, skb, cp->dest, cp->daddr.ip, rt_mode, - NULL, iph); + NULL, ciph); if (local < 0) goto tx_error; rt = skb_rtable(skb); @@ -1565,13 +1566,13 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, } /* copy-on-write the packet before mangling it */ - if (skb_ensure_writable(skb, offset)) + if (skb_ensure_writable(skb, wlen)) goto tx_error; if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error; - ip_vs_nat_icmp(skb, pp, cp, 0); + ip_vs_nat_icmp(skb, pp, cp, 0, toff); /* Another hack: avoid icmp_send in ip_fragment */ skb->ignore_df = 1; @@ -1587,8 +1588,9 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, #ifdef CONFIG_IP_VS_IPV6 int ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, - struct ip_vs_protocol *pp, int offset, unsigned int hooknum, - struct ip_vs_iphdr *ipvsh) + struct ip_vs_protocol *pp, unsigned int toff, + unsigned int wlen, unsigned int hooknum, + struct ip_vs_iphdr *ciph) { struct rt6_info *rt; /* Route to the other host */ int rc; @@ -1600,7 +1602,7 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, translate address/port back */ if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) { if (cp->packet_xmit) - rc = cp->packet_xmit(skb, cp, pp, ipvsh); + rc = cp->packet_xmit(skb, cp, pp, ciph); else rc = NF_ACCEPT; /* do not touch skb anymore */ @@ -1617,7 +1619,7 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, IP_VS_RT_MODE_LOCAL | IP_VS_RT_MODE_NON_LOCAL | IP_VS_RT_MODE_RDR : IP_VS_RT_MODE_NON_LOCAL; local = __ip_vs_get_out_rt_v6(cp->ipvs, cp->af, skb, cp->dest, - &cp->daddr.in6, NULL, ipvsh, 0, rt_mode); + &cp->daddr.in6, NULL, ciph, 0, rt_mode); if (local < 0) goto tx_error; rt = dst_rt6_info(skb_dst(skb)); @@ -1649,13 +1651,13 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, } /* copy-on-write the packet before mangling it */ - if (skb_ensure_writable(skb, offset)) + if (skb_ensure_writable(skb, wlen)) goto tx_error; if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error; - ip_vs_nat_icmp_v6(skb, pp, cp, 0); + ip_vs_nat_icmp_v6(skb, pp, cp, 0, toff, ciph); /* Another hack: avoid icmp_send in ip_fragment */ skb->ignore_df = 1; -- cgit From 342e24a339b90e8e339a0f8c151ca479b8565661 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Wed, 22 Jul 2026 13:15:17 +0300 Subject: ipvs: do not mangle ICMP replies for non-first fragments Sashiko warns that ip_vs_nat_icmp() unconditionally mangles the payload for embedded non-first IPv4 fragments. The problem is in the very old inverted pp->dont_defrag check which should not continue when embedded is a non-first TCP/UDP/SCTP fragment. Check for embedded non-first fragment is also missing from ip_vs_out_icmp_v6(), it is needed before any connection lookups that expect ports after the network headers. Drop the blocking code from ip_vs_in_icmp_v6() which prevents ICMPv6 from local clients to use non-MASQ forwarding. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Link: https://sashiko.dev/#/patchset/20260720201122.79882-1-ja%40ssi.bg Signed-off-by: Julian Anastasov Signed-off-by: Pablo Neira Ayuso --- include/net/ip_vs.h | 11 ++++---- net/netfilter/ipvs/ip_vs_core.c | 61 ++++++++++++++--------------------------- net/netfilter/ipvs/ip_vs_xmit.c | 28 +++++++++++++++---- 3 files changed, 48 insertions(+), 52 deletions(-) (limited to 'include') diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 4a10a01d6e2f..e6ca930a3507 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -1975,8 +1975,7 @@ int ip_vs_dr_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph); int ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, unsigned int toff, - unsigned int wlen, unsigned int hooknum, - struct ip_vs_iphdr *ciph); + unsigned int hooknum, struct ip_vs_iphdr *ciph); void ip_vs_dest_dst_rcu_free(struct rcu_head *head); #ifdef CONFIG_IP_VS_IPV6 @@ -1990,8 +1989,7 @@ int ip_vs_dr_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, struct ip_vs_iphdr *iph); int ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, unsigned int toff, - unsigned int wlen, unsigned int hooknum, - struct ip_vs_iphdr *ciph); + unsigned int hooknum, struct ip_vs_iphdr *ciph); #endif #ifdef CONFIG_SYSCTL @@ -2063,12 +2061,13 @@ static inline bool ip_vs_conn_use_hash2(struct ip_vs_conn *cp) } void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp, - struct ip_vs_conn *cp, int dir, unsigned int toff); + struct ip_vs_conn *cp, int dir, unsigned int toff, + bool has_ports); #ifdef CONFIG_IP_VS_IPV6 void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, struct ip_vs_conn *cp, int dir, unsigned int toff, - struct ip_vs_iphdr *ciph); + bool has_ports, struct ip_vs_iphdr *ciph); #endif static inline __wsum ip_vs_check_diff4(__be32 old, __be32 new, __wsum oldsum) diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c index cd5eb71543ec..7efa209a517b 100644 --- a/net/netfilter/ipvs/ip_vs_core.c +++ b/net/netfilter/ipvs/ip_vs_core.c @@ -924,7 +924,8 @@ static int ip_vs_route_me_harder(struct netns_ipvs *ipvs, int af, * - inout: 1=in->out, 0=out->in */ void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp, - struct ip_vs_conn *cp, int inout, unsigned int toff) + struct ip_vs_conn *cp, int inout, unsigned int toff, + bool has_ports) { struct iphdr *iph = ip_hdr(skb); struct icmphdr *icmph = (struct icmphdr *)(skb->data + toff); @@ -944,8 +945,7 @@ void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp, } /* the TCP/UDP/SCTP port */ - if (IPPROTO_TCP == ciph->protocol || IPPROTO_UDP == ciph->protocol || - IPPROTO_SCTP == ciph->protocol) { + if (has_ports) { __be16 *ports = (void *)ciph + ciph->ihl*4; if (inout) @@ -970,18 +970,15 @@ void ip_vs_nat_icmp(struct sk_buff *skb, struct ip_vs_protocol *pp, #ifdef CONFIG_IP_VS_IPV6 void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, struct ip_vs_conn *cp, int inout, unsigned int toff, - struct ip_vs_iphdr *ciph) + bool has_ports, struct ip_vs_iphdr *ciph) { struct ipv6hdr *iph = ipv6_hdr(skb); - int protocol; struct icmp6hdr *icmph; struct ipv6hdr *cih; icmph = (struct icmp6hdr *)(skb->data + toff); cih = (struct ipv6hdr *)(skb->data + ciph->off); - protocol = ciph->protocol; - if (inout) { iph->saddr = cp->vaddr.in6; cih->daddr = cp->vaddr.in6; @@ -991,9 +988,7 @@ void ip_vs_nat_icmp_v6(struct sk_buff *skb, struct ip_vs_protocol *pp, } /* the TCP/UDP/SCTP port */ - if (!ciph->fragoffs && - (protocol == IPPROTO_TCP || protocol == IPPROTO_UDP || - protocol == IPPROTO_SCTP)) { + if (has_ports) { __be16 *ports = (void *)(skb->data + ciph->len); IP_VS_DBG(11, "%s() changed port %d to %d\n", __func__, @@ -1035,6 +1030,7 @@ static int handle_response_icmp(int af, struct sk_buff *skb, int iproto = af == AF_INET6 ? IPPROTO_ICMPV6 : IPPROTO_ICMP; unsigned int verdict = NF_DROP; unsigned int ctoff = ciph->len; + bool has_ports = false; if (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ) goto after_nat; @@ -1048,17 +1044,19 @@ static int handle_response_icmp(int af, struct sk_buff *skb, } if (ciph->protocol == IPPROTO_TCP || ciph->protocol == IPPROTO_UDP || - ciph->protocol == IPPROTO_SCTP) + ciph->protocol == IPPROTO_SCTP) { ctoff += 2 * sizeof(__u16); + has_ports = true; + } if (skb_ensure_writable(skb, ctoff)) goto out; #ifdef CONFIG_IP_VS_IPV6 if (af == AF_INET6) - ip_vs_nat_icmp_v6(skb, pp, cp, 1, toff, ciph); + ip_vs_nat_icmp_v6(skb, pp, cp, 1, toff, has_ports, ciph); else #endif - ip_vs_nat_icmp(skb, pp, cp, 1, toff); + ip_vs_nat_icmp(skb, pp, cp, 1, toff, has_ports); if (ip_vs_route_me_harder(cp->ipvs, af, skb, hooknum)) goto out; @@ -1142,8 +1140,7 @@ static int ip_vs_out_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, return NF_ACCEPT; /* Is the embedded protocol header present? */ - if (unlikely(cih->frag_off & htons(IP_OFFSET) && - pp->dont_defrag)) + if (unlikely(cih->frag_off & htons(IP_OFFSET) && !pp->dont_defrag)) return NF_ACCEPT; IP_VS_DBG_PKT(11, AF_INET, pp, skb, offset, @@ -1207,6 +1204,10 @@ static int ip_vs_out_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, if (!pp) return NF_ACCEPT; + /* Is the embedded protocol header present? */ + if (unlikely(ciph.fragoffs && !pp->dont_defrag)) + return NF_ACCEPT; + /* The embedded headers contain source and dest in reverse order */ cp = INDIRECT_CALL_1(pp->conn_out_get, ip_vs_conn_out_get_proto, ipvs, AF_INET6, skb, &ciph); @@ -1865,8 +1866,7 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, pp = pd->pp; /* Is the embedded protocol header present? */ - if (unlikely(cih->frag_off & htons(IP_OFFSET) && - pp->dont_defrag)) + if (unlikely(cih->frag_off & htons(IP_OFFSET) && !pp->dont_defrag)) return NF_ACCEPT; IP_VS_DBG_PKT(11, AF_INET, pp, skb, offset, @@ -1874,7 +1874,6 @@ ip_vs_in_icmp(struct netns_ipvs *ipvs, struct sk_buff *skb, int *related, offset2 = offset; ip_vs_fill_iph_skb_icmp(AF_INET, skb, offset, !tunnel, &ciph); - offset = ciph.len; /* The embedded headers contain source and dest in reverse order. * For IPIP/UDP/GRE tunnel this is error for request, not for reply. @@ -1968,11 +1967,7 @@ ignore_tunnel: /* do the statistics and put it back */ ip_vs_in_stats(cp, skb); - if (IPPROTO_TCP == cih->protocol || IPPROTO_UDP == cih->protocol || - IPPROTO_SCTP == cih->protocol) - offset += 2 * sizeof(__u16); - verdict = ip_vs_icmp_xmit(skb, cp, pp, iph->len, offset, hooknum, - &ciph); + verdict = ip_vs_icmp_xmit(skb, cp, pp, iph->len, hooknum, &ciph); out: if (likely(!new_cp)) @@ -2032,8 +2027,8 @@ static int ip_vs_in_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, return NF_ACCEPT; pp = pd->pp; - /* Cannot handle fragmented embedded protocol */ - if (ciph.fragoffs) + /* Is the embedded protocol header present? */ + if (ciph.fragoffs && !pp->dont_defrag) return NF_ACCEPT; IP_VS_DBG_PKT(11, AF_INET6, pp, skb, offset, @@ -2057,13 +2052,6 @@ static int ip_vs_in_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, new_cp = true; } - /* VS/TUN, VS/DR and LOCALNODE just let it go */ - if ((hooknum == NF_INET_LOCAL_OUT) && - (IP_VS_FWD_METHOD(cp) != IP_VS_CONN_F_MASQ)) { - verdict = NF_ACCEPT; - goto out; - } - verdict = NF_DROP; /* Ensure the checksum is correct */ @@ -2079,14 +2067,7 @@ static int ip_vs_in_icmp_v6(struct netns_ipvs *ipvs, struct sk_buff *skb, /* do the statistics and put it back */ ip_vs_in_stats(cp, skb); - /* Need to mangle contained IPv6 header in ICMPv6 packet */ - offset = ciph.len; - if (IPPROTO_TCP == ciph.protocol || IPPROTO_UDP == ciph.protocol || - IPPROTO_SCTP == ciph.protocol) - offset += 2 * sizeof(__u16); /* Also mangle ports */ - - verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, iph->len, offset, hooknum, - &ciph); + verdict = ip_vs_icmp_xmit_v6(skb, cp, pp, iph->len, hooknum, &ciph); out: if (likely(!new_cp)) diff --git a/net/netfilter/ipvs/ip_vs_xmit.c b/net/netfilter/ipvs/ip_vs_xmit.c index c23401c789de..0b0c5304993a 100644 --- a/net/netfilter/ipvs/ip_vs_xmit.c +++ b/net/netfilter/ipvs/ip_vs_xmit.c @@ -1503,13 +1503,14 @@ tx_error: int ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, unsigned int toff, - unsigned int wlen, unsigned int hooknum, - struct ip_vs_iphdr *ciph) + unsigned int hooknum, struct ip_vs_iphdr *ciph) { struct rtable *rt; /* Route to the other host */ int rc; int local; int rt_mode, was_input; + bool has_ports = false; + unsigned int wlen; /* The ICMP packet for VS/TUN, VS/DR and LOCALNODE will be forwarded directly here, because there is no need to @@ -1565,6 +1566,13 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error; } + wlen = ciph->len; + if (ciph->protocol == IPPROTO_TCP || ciph->protocol == IPPROTO_UDP || + ciph->protocol == IPPROTO_SCTP) { + wlen += 2 * sizeof(__u16); /* Also mangle ports */ + has_ports = true; + } + /* copy-on-write the packet before mangling it */ if (skb_ensure_writable(skb, wlen)) goto tx_error; @@ -1572,7 +1580,7 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error; - ip_vs_nat_icmp(skb, pp, cp, 0, toff); + ip_vs_nat_icmp(skb, pp, cp, 0, toff, has_ports); /* Another hack: avoid icmp_send in ip_fragment */ skb->ignore_df = 1; @@ -1589,10 +1597,11 @@ ip_vs_icmp_xmit(struct sk_buff *skb, struct ip_vs_conn *cp, int ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, struct ip_vs_protocol *pp, unsigned int toff, - unsigned int wlen, unsigned int hooknum, - struct ip_vs_iphdr *ciph) + unsigned int hooknum, struct ip_vs_iphdr *ciph) { + bool has_ports = false; struct rt6_info *rt; /* Route to the other host */ + unsigned int wlen; int rc; int local; int rt_mode; @@ -1650,6 +1659,13 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, goto tx_error; } + wlen = ciph->len; + if (ciph->protocol == IPPROTO_TCP || ciph->protocol == IPPROTO_UDP || + ciph->protocol == IPPROTO_SCTP) { + wlen += 2 * sizeof(__u16); /* Also mangle ports */ + has_ports = true; + } + /* copy-on-write the packet before mangling it */ if (skb_ensure_writable(skb, wlen)) goto tx_error; @@ -1657,7 +1673,7 @@ ip_vs_icmp_xmit_v6(struct sk_buff *skb, struct ip_vs_conn *cp, if (skb_cow(skb, rt->dst.dev->hard_header_len)) goto tx_error; - ip_vs_nat_icmp_v6(skb, pp, cp, 0, toff, ciph); + ip_vs_nat_icmp_v6(skb, pp, cp, 0, toff, has_ports, ciph); /* Another hack: avoid icmp_send in ip_fragment */ skb->ignore_df = 1; -- cgit From bd44a6dcd4248883de90f5dad53ae80066e27096 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Sun, 19 Jul 2026 15:56:05 +0200 Subject: xsk: drain continuation descs after overflow in xsk_build_skb() Fix generic xmit path multi-buffer logic when packets are either too big (count of descriptors exceed MAX_SKB_FRAGS) or an invalid descriptor is included in fragmented packet. Introduce xdp_sock::drain_cont and act upon this flag - when it is set, keep on consuming descriptors from AF_XDP Tx ring and put them directly onto Cq. Previously these descriptors were silently lost and could never be reached again. Fixes: cf24f5a5feea ("xsk: add support for AF_XDP multi-buffer on Tx path") Closes: https://lore.kernel.org/all/20260425041726.85FB3C2BCB2@smtp.kernel.org/ Reviewed-by: Jason Xing Co-developed-by: Maciej Fijalkowski # wrapped cq addr submission onto routine Signed-off-by: Maciej Fijalkowski Signed-off-by: Jason Xing Acked-by: Stanislav Fomichev Link: https://patch.msgid.link/20260719135609.147823-3-maciej.fijalkowski@intel.com Signed-off-by: Jakub Kicinski --- include/net/xdp_sock.h | 1 + net/xdp/xsk.c | 45 ++++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 43 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/net/xdp_sock.h b/include/net/xdp_sock.h index ebac60a3d8a1..8b51876efbed 100644 --- a/include/net/xdp_sock.h +++ b/include/net/xdp_sock.h @@ -80,6 +80,7 @@ struct xdp_sock { * call of __xsk_generic_xmit(). */ struct sk_buff *skb; + bool drain_cont; struct list_head map_list; /* Protects map_list */ diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index a7a83dc4546a..12a845d012f6 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -737,6 +737,19 @@ static void xsk_cq_submit_addr_locked(struct xsk_buff_pool *pool, spin_unlock_irqrestore(&pool->cq_prod_lock, flags); } +static void xsk_cq_submit_addr_single_locked(struct xsk_buff_pool *pool, + struct xdp_desc *desc) +{ + unsigned long flags; + u32 idx; + + spin_lock_irqsave(&pool->cq_prod_lock, flags); + idx = xskq_get_prod(pool->cq); + xskq_prod_write_addr(pool->cq, idx, desc->addr); + xskq_prod_submit_n(pool->cq, 1); + spin_unlock_irqrestore(&pool->cq_prod_lock, flags); +} + static void xsk_cq_cancel_locked(struct xsk_buff_pool *pool, u32 n) { spin_lock(&pool->cq->cq_cached_prod_lock); @@ -1028,13 +1041,14 @@ free_err: static int __xsk_generic_xmit(struct sock *sk) { struct xdp_sock *xs = xdp_sk(sk); - bool sent_frame = false; struct xdp_desc desc; struct sk_buff *skb; + u32 cached_cons; u32 max_batch; int err = 0; mutex_lock(&xs->mutex); + cached_cons = xs->tx->cached_cons; /* Since we dropped the RCU read lock, the socket state might have changed. */ if (unlikely(!xsk_is_bound(xs))) { @@ -1063,11 +1077,21 @@ static int __xsk_generic_xmit(struct sock *sk) goto out; } + if (unlikely(xs->drain_cont)) { + xsk_cq_submit_addr_single_locked(xs->pool, &desc); + xs->tx->invalid_descs++; + xskq_cons_release(xs->tx); + xs->drain_cont = xp_mb_desc(&desc); + continue; + } + skb = xsk_build_skb(xs, &desc); if (IS_ERR(skb)) { err = PTR_ERR(skb); if (err != -EOVERFLOW) goto out; + if (xp_mb_desc(&desc)) + xs->drain_cont = true; err = 0; continue; } @@ -1096,18 +1120,33 @@ static int __xsk_generic_xmit(struct sock *sk) goto out; } - sent_frame = true; xs->skb = NULL; } if (xskq_has_descs(xs->tx)) { + bool drain = xs->skb || xs->drain_cont || xp_mb_desc(&desc); + + err = xsk_cq_reserve_locked(xs->pool); + if (err) { + xs->tx->invalid_descs--; + if (xs->skb) + xsk_drop_skb(xs->skb); + xs->drain_cont = drain; + err = -EAGAIN; + goto out; + } + if (xs->skb) xsk_drop_skb(xs->skb); + + xsk_cq_submit_addr_single_locked(xs->pool, &desc); + xskq_cons_release(xs->tx); + xs->drain_cont = xp_mb_desc(&desc); } out: - if (sent_frame) + if (xs->tx->cached_cons != cached_cons) __xsk_tx_release(xs); mutex_unlock(&xs->mutex); -- cgit From 08c9a8e794b4694c100dafcb80e069e29ad81b64 Mon Sep 17 00:00:00 2001 From: Maciej Fijalkowski Date: Sun, 19 Jul 2026 15:56:06 +0200 Subject: xsk: provide sufficient space in pool->tx_descs The temporary Tx descriptor array in an XSK buffer pool is currently sized from the Tx ring of the socket that creates the pool. This is insufficient for shared-UMEM Tx. A later socket may have a larger Tx ring and submit a valid multi-buffer packet containing more descriptors than the first socket's ring, while still remaining within the device's xdp_zc_max_segs limit. A packet-framed batch parser bounded by the temporary array cannot reach the end-of-packet descriptor in that case. It leaves the packet on the Tx ring and encounters the same packet on every subsequent attempt, stalling Tx processing for that socket. Size the temporary descriptor array to the larger of the first Tx ring and the device's xdp_zc_max_segs capability. This keeps the array large enough to inspect one maximum-sized valid packet. Larger shared Tx rings do not require further resizing, as they can be processed over multiple batches. Following commit will actually address the data path side. Fixes: d5581966040f ("xsk: support ZC Tx multi-buffer in batch API") Reviewed-by: Jason Xing Signed-off-by: Maciej Fijalkowski Acked-by: Stanislav Fomichev Link: https://patch.msgid.link/20260719135609.147823-4-maciej.fijalkowski@intel.com Signed-off-by: Jakub Kicinski --- include/net/xsk_buff_pool.h | 6 ++++-- net/xdp/xsk.c | 10 +++++++--- net/xdp/xsk_buff_pool.c | 12 ++++++++---- 3 files changed, 19 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/include/net/xsk_buff_pool.h b/include/net/xsk_buff_pool.h index ccb3b350001f..f5e737a83055 100644 --- a/include/net/xsk_buff_pool.h +++ b/include/net/xsk_buff_pool.h @@ -102,12 +102,14 @@ struct xsk_buff_pool { /* AF_XDP core. */ struct xsk_buff_pool *xp_create_and_assign_umem(struct xdp_sock *xs, - struct xdp_umem *umem); + struct xdp_umem *umem, + u32 max_segs); int xp_assign_dev(struct xsk_buff_pool *pool, struct net_device *dev, u16 queue_id, u16 flags); int xp_assign_dev_shared(struct xsk_buff_pool *pool, struct xdp_sock *umem_xs, struct net_device *dev, u16 queue_id); -int xp_alloc_tx_descs(struct xsk_buff_pool *pool, struct xdp_sock *xs); +int xp_alloc_tx_descs(struct xsk_buff_pool *pool, struct xdp_sock *xs, + u32 max_segs); void xp_destroy(struct xsk_buff_pool *pool); void xp_get_pool(struct xsk_buff_pool *pool); bool xp_put_pool(struct xsk_buff_pool *pool); diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 12a845d012f6..091792d1d82d 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -1525,7 +1525,8 @@ static int xsk_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr * and/or device. */ xs->pool = xp_create_and_assign_umem(xs, - umem_xs->umem); + umem_xs->umem, + dev->xdp_zc_max_segs); if (!xs->pool) { err = -ENOMEM; sockfd_put(sock); @@ -1557,7 +1558,8 @@ static int xsk_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr * utilizes */ if (xs->tx && !xs->pool->tx_descs) { - err = xp_alloc_tx_descs(xs->pool, xs); + err = xp_alloc_tx_descs(xs->pool, xs, + dev->xdp_zc_max_segs); if (err) { xp_put_pool(xs->pool); xs->pool = NULL; @@ -1575,7 +1577,9 @@ static int xsk_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr goto out_unlock; } else { /* This xsk has its own umem. */ - xs->pool = xp_create_and_assign_umem(xs, xs->umem); + xs->pool = xp_create_and_assign_umem(xs, xs->umem, + dev->xdp_zc_max_segs); + if (!xs->pool) { err = -ENOMEM; goto out_unlock; diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index 1f28a9641571..12c9fb29af05 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -42,9 +42,12 @@ void xp_destroy(struct xsk_buff_pool *pool) kvfree(pool); } -int xp_alloc_tx_descs(struct xsk_buff_pool *pool, struct xdp_sock *xs) +int xp_alloc_tx_descs(struct xsk_buff_pool *pool, struct xdp_sock *xs, + u32 max_segs) { - pool->tx_descs = kvzalloc_objs(*pool->tx_descs, xs->tx->nentries); + u32 nentries = max(xs->tx->nentries, max_segs); + + pool->tx_descs = kvzalloc_objs(*pool->tx_descs, nentries); if (!pool->tx_descs) return -ENOMEM; @@ -52,7 +55,8 @@ int xp_alloc_tx_descs(struct xsk_buff_pool *pool, struct xdp_sock *xs) } struct xsk_buff_pool *xp_create_and_assign_umem(struct xdp_sock *xs, - struct xdp_umem *umem) + struct xdp_umem *umem, + u32 max_segs) { bool unaligned = umem->flags & XDP_UMEM_UNALIGNED_CHUNK_FLAG; struct xsk_buff_pool *pool; @@ -69,7 +73,7 @@ struct xsk_buff_pool *xp_create_and_assign_umem(struct xdp_sock *xs, goto out; if (xs->tx) - if (xp_alloc_tx_descs(pool, xs)) + if (xp_alloc_tx_descs(pool, xs, max_segs)) goto out; pool->chunk_mask = ~((u64)umem->chunk_size - 1); -- cgit From 72f2b4516faf55d4dfac2414649d3cffa5fd2c5e Mon Sep 17 00:00:00 2001 From: Maciej Fijalkowski Date: Sun, 19 Jul 2026 15:56:07 +0200 Subject: xsk: reclaim invalid Tx descriptors in ZC batch path The zero-copy Tx batch parser stops when it encounters an invalid descriptor. If this happens after one or more continuation descriptors, the Tx consumer can be advanced past fragments that are neither submitted to the driver nor returned to userspace through the completion ring. A similar problem occurs when a packet exceeds xdp_zc_max_segs. The descriptors consumed up to the limit are released without completion, and the remaining continuation descriptors can subsequently be interpreted as the beginning of another packet. Parse Tx batches in packet units and distinguish descriptors belonging to complete valid packets from descriptors consumed while draining an invalid or oversized packet. Return the former to the driver and append the latter to the CQ address area so userspace can reclaim their UMEM frames. Treat a standalone invalid descriptor as a one-descriptor reclaim-only packet. Advancing the Tx-ring consumer releases the ring slot, but does not by itself return ownership of the referenced UMEM frame to userspace. Once draining starts, continue until the packet's end-of-packet descriptor is consumed. Preserve the drain state on the socket when EOP has not yet been supplied, so draining can continue during a later call. Leave incomplete but otherwise valid packets on the Tx ring. Shared-UMEM pools using multi-buffer Tx also need packet-framed parsing. Walk their Tx sockets one packet at a time, preserving the existing per-socket fairness scheme, instead of using the legacy one-descriptor fallback. Keep that fallback for shared pools that do not use multi-buffer Tx. Since the drain state is maintained per socket and both the singular and shared paths can resume an interrupted drain, changing the socket list from singular to shared requires no special bind-time transition. CQ entries are positional, and drivers may complete only part of the Tx work returned by xsk_tx_peek_release_desc_batch(). Therefore, reclaim-only entries cannot be published immediately when earlier driver-visible descriptors are still outstanding. Track the number of driver-visible CQ entries preceding the reclaim entries. Let xsk_tx_completed() publish partial hardware Tx completions, and publish the reclaim entries only after every earlier Tx descriptor has completed. Complete a reclaim-only batch immediately when there is no driver-visible work in front of it, and prevent another Tx batch from being appended while reclaim entries remain pending. Also cap batch processing by the size of the pool's temporary descriptor array, as Tx rings belonging to sockets sharing a UMEM may have different sizes. This ensures that every invalid Tx descriptor consumed by the ZC batch path is either submitted to the driver as part of a valid packet or returned to userspace without violating CQ completion ordering. Reviewed-by: Jason Xing Signed-off-by: Maciej Fijalkowski Acked-by: Stanislav Fomichev Fixes: cf24f5a5feea ("xsk: add support for AF_XDP multi-buffer on Tx path") Link: https://patch.msgid.link/20260719135609.147823-5-maciej.fijalkowski@intel.com Signed-off-by: Jakub Kicinski --- Documentation/networking/af_xdp.rst | 54 ++++++----- include/net/xsk_buff_pool.h | 3 + net/xdp/xsk.c | 187 ++++++++++++++++++++++++++++++++---- net/xdp/xsk_buff_pool.c | 1 + net/xdp/xsk_queue.h | 65 +++++++++---- 5 files changed, 248 insertions(+), 62 deletions(-) (limited to 'include') diff --git a/Documentation/networking/af_xdp.rst b/Documentation/networking/af_xdp.rst index 50d92084a49c..cc3f0d16b28f 100644 --- a/Documentation/networking/af_xdp.rst +++ b/Documentation/networking/af_xdp.rst @@ -43,12 +43,13 @@ UMEM also has two rings: the FILL ring and the COMPLETION ring. The FILL ring is used by the application to send down addr for the kernel to fill in with RX packet data. References to these frames will then appear in the RX ring once each packet has been received. The -COMPLETION ring, on the other hand, contains frame addr that the -kernel has transmitted completely and can now be used again by user -space, for either TX or RX. Thus, the frame addrs appearing in the -COMPLETION ring are addrs that were previously transmitted using the -TX ring. In summary, the RX and FILL rings are used for the RX path -and the TX and COMPLETION rings are used for the TX path. +COMPLETION ring, on the other hand, contains frame addresses from Tx +descriptors that the kernel has finished processing and that can now be +used again by user space, for either Tx or Rx. This includes frames whose +transmission has completed as well as frames referenced by invalid Tx +descriptors rejected by the kernel. A completion therefore returns +ownership of a frame to user space, but does not by itself guarantee that +the packet was successfully transmitted. The socket is then finally bound with a bind() call to a device and a specific queue id on that device, and it is not until bind is @@ -169,14 +170,15 @@ chunks mode, then the incoming addr will be left untouched. UMEM Completion Ring ~~~~~~~~~~~~~~~~~~~~ -The COMPLETION Ring is used transfer ownership of UMEM frames from +The COMPLETION Ring is used to transfer ownership of UMEM frames from kernel-space to user-space. Just like the FILL ring, UMEM indices are -used. - -Frames passed from the kernel to user-space are frames that has been -sent (TX ring) and can be used by user-space again. - -The user application consumes UMEM addrs from this ring. +used. Frames passed from the kernel to user-space are frames referenced +by Tx descriptors that the kernel has finished processing and can be +used by user-space again. This includes both frames whose transmission +has completed and frames referenced by invalid Tx descriptors that were +rejected and reclaimed by the kernel. A completion entry does not +guarantee successful packet transmission. The user application consumes +UMEM addrs from this ring. RX Ring @@ -504,21 +506,25 @@ will be treated as an invalid descriptor. These are the semantics for producing packets onto AF_XDP Tx ring consisting of multiple frames: -* When an invalid descriptor is found, all the other - descriptors/frames of this packet are marked as invalid and not - completed. The next descriptor is treated as the start of a new - packet, even if this was not the intent (because we cannot guess - the intent). As before, if your program is producing invalid - descriptors you have a bug that must be fixed. +* When an invalid descriptor is found, the complete packet is treated as + invalid. The kernel consumes descriptors through the descriptor marking + the end of the packet and returns all their frame addresses through the + COMPLETION ring. A standalone invalid descriptor is treated as a + one-descriptor invalid packet. The descriptor following the end of the + invalid packet is treated as the start of a new packet. As before, if + your program is producing invalid descriptors you have a bug that must + be fixed. Rejected descriptors are reported in the ``tx_invalid_descs`` + statistic. * Zero length descriptors are treated as invalid descriptors. * For copy mode, the maximum supported number of frames in a packet is - equal to CONFIG_MAX_SKB_FRAGS + 1. If it is exceeded, all - descriptors accumulated so far are dropped and treated as - invalid. To produce an application that will work on any system - regardless of this config setting, limit the number of frags to 18, - as the minimum value of the config is 17. + equal to CONFIG_MAX_SKB_FRAGS + 1. If it is exceeded, all descriptors + through the end of the oversized packet are consumed, treated as invalid, + and their frame addresses are returned through the COMPLETION ring. To + produce an application that will work on any system regardless of this + config setting, limit the number of frags to 18, as the minimum value of + the config is 17. * For zero-copy mode, the limit is up to what the NIC HW supports. Usually at least five on the NICs we have checked. We diff --git a/include/net/xsk_buff_pool.h b/include/net/xsk_buff_pool.h index f5e737a83055..2bb1d122b1bc 100644 --- a/include/net/xsk_buff_pool.h +++ b/include/net/xsk_buff_pool.h @@ -78,6 +78,9 @@ struct xsk_buff_pool { u32 chunk_size; u32 chunk_shift; u32 frame_len; + u32 tx_descs_nentries; + u32 reclaim_descs; + u32 tx_zc_pending_descs; u32 xdp_zc_max_segs; u8 tx_metadata_len; /* inherited from umem */ u8 cached_need_wakeup; diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 091792d1d82d..f906d51b6699 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -499,6 +499,23 @@ void __xsk_map_flush(struct list_head *flush_list) void xsk_tx_completed(struct xsk_buff_pool *pool, u32 nb_entries) { + u32 reclaim_descs = READ_ONCE(pool->reclaim_descs); + + if (unlikely(reclaim_descs)) { + u32 pending_descs = READ_ONCE(pool->tx_zc_pending_descs); + + if (nb_entries < pending_descs) { + WRITE_ONCE(pool->tx_zc_pending_descs, + pending_descs - nb_entries); + xskq_prod_submit_n(pool->cq, nb_entries); + return; + } + + WRITE_ONCE(pool->tx_zc_pending_descs, 0); + nb_entries += reclaim_descs; + WRITE_ONCE(pool->reclaim_descs, 0); + } + xskq_prod_submit_n(pool->cq, nb_entries); } EXPORT_SYMBOL(xsk_tx_completed); @@ -574,24 +591,157 @@ static u32 xsk_tx_peek_release_fallback(struct xsk_buff_pool *pool, u32 max_entr return nb_pkts; } +static void xsk_tx_commit_batch(struct xsk_buff_pool *pool, + struct xsk_tx_batch *batch) +{ + u32 nb_descs = xsk_tx_batch_cq_descs(batch); + u32 cq_cached_prod; + + if (!nb_descs) + return; + + cq_cached_prod = pool->cq->cached_prod; + xskq_prod_write_addr_batch(pool->cq, pool->tx_descs, nb_descs); + + if (unlikely(batch->reclaim_descs)) { + u32 cq_pending_descs; + + /* CQ is positional. Descriptors already written but not + * submitted must complete before any reclaim-only descriptors + * appended below. + */ + cq_pending_descs = cq_cached_prod - xskq_get_prod(pool->cq); + + WRITE_ONCE(pool->tx_zc_pending_descs, + batch->tx_descs + cq_pending_descs); + WRITE_ONCE(pool->reclaim_descs, batch->reclaim_descs); + if (unlikely(!pool->tx_zc_pending_descs)) + xsk_tx_completed(pool, 0); + } +} + +static struct xsk_tx_batch +__xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, struct xdp_sock *xs, + struct xdp_desc *descs, u32 max_descs) +{ + struct xsk_tx_batch batch = {}; + u32 entries; + + entries = xskq_cons_nb_entries(xs->tx, max_descs); + if (!entries) + return batch; + + batch = xskq_cons_read_desc_batch(xs, pool, descs, max_descs); + if (!xsk_tx_batch_cq_descs(&batch)) { + xs->tx->queue_empty_descs++; + } else { + __xskq_cons_release(xs->tx); + xs->sk.sk_write_space(&xs->sk); + } + return batch; +} + +static struct xsk_tx_batch +xsk_tx_peek_release_shared_desc_batch(struct xsk_buff_pool *pool, u32 max_descs) +{ + u32 cq_descs_before, cq_descs_after; + struct xsk_tx_batch sum_batch = {}; + bool budget_exhausted; + u32 per_socket_budget; + struct xdp_sock *xs; + + /* The fairness quota must allow one maximum-sized valid packet. */ + per_socket_budget = max_t(u32, MAX_PER_SOCKET_BUDGET, + pool->xdp_zc_max_segs); + +again: + budget_exhausted = false; + cq_descs_before = xsk_tx_batch_cq_descs(&sum_batch); + list_for_each_entry_rcu(xs, &pool->xsk_tx_list, tx_list) { + u32 budget, budget_left, offset, remaining, used; + struct xsk_tx_batch curr_batch; + + /* Once reclaim-only descriptors have been appended to the CQ + * address area, do not append driver-visible Tx descriptors + * from another socket after them. xsk_tx_completed() relies on + * all driver-visible descriptors preceding all reclaim-only + * descriptors in CQ order. + */ + if (sum_batch.reclaim_descs) + break; + + /* be gentle when playing with pool->tx_descs */ + offset = xsk_tx_batch_cq_descs(&sum_batch); + if (offset >= max_descs) + break; + + if (xs->tx_budget_spent >= per_socket_budget) { + if (xskq_cons_nb_entries(xs->tx, 1)) + budget_exhausted = true; + continue; + } + + budget_left = per_socket_budget - xs->tx_budget_spent; + remaining = max_descs - offset; + budget = min(remaining, budget_left); + + curr_batch = __xsk_tx_peek_release_desc_batch(pool, xs, + pool->tx_descs + offset, + budget); + used = xsk_tx_batch_cq_descs(&curr_batch); + if (!used) { + if (curr_batch.budget_limited && budget_left < remaining) + budget_exhausted = true; + continue; + } + + xs->tx_budget_spent += used; + sum_batch.tx_descs += curr_batch.tx_descs; + sum_batch.reclaim_descs = curr_batch.reclaim_descs; + } + + cq_descs_after = xsk_tx_batch_cq_descs(&sum_batch); + + if (sum_batch.reclaim_descs || cq_descs_after >= max_descs) + return sum_batch; + + /* Continue filling the batch while this pass made progress */ + if (cq_descs_before != cq_descs_after) + goto again; + + if (!budget_exhausted) + return sum_batch; + + list_for_each_entry_rcu(xs, &pool->xsk_tx_list, tx_list) + xs->tx_budget_spent = 0; + goto again; +} + u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, u32 nb_pkts) { + struct xsk_tx_batch batch = {}; struct xdp_sock *xs; + bool umem_shared; rcu_read_lock(); - if (!list_is_singular(&pool->xsk_tx_list)) { - /* Fallback to the non-batched version */ - rcu_read_unlock(); - return xsk_tx_peek_release_fallback(pool, nb_pkts); - } + if (unlikely(READ_ONCE(pool->reclaim_descs))) + goto out; - xs = list_first_or_null_rcu(&pool->xsk_tx_list, struct xdp_sock, tx_list); - if (!xs) { - nb_pkts = 0; + xs = list_first_or_null_rcu(&pool->xsk_tx_list, struct xdp_sock, + tx_list); + if (!xs) goto out; - } - nb_pkts = xskq_cons_nb_entries(xs->tx, nb_pkts); + nb_pkts = min(nb_pkts, pool->tx_descs_nentries); + if (!nb_pkts) + goto out; + + umem_shared = !list_is_singular(&pool->xsk_tx_list); + + if (umem_shared && !(pool->umem->flags & XDP_UMEM_SG_FLAG)) { + rcu_read_unlock(); + return xsk_tx_peek_release_fallback(pool, nb_pkts); + } /* This is the backpressure mechanism for the Tx path. Try to * reserve space in the completion queue for all packets, but @@ -603,19 +753,16 @@ u32 xsk_tx_peek_release_desc_batch(struct xsk_buff_pool *pool, u32 nb_pkts) if (!nb_pkts) goto out; - nb_pkts = xskq_cons_read_desc_batch(xs->tx, pool, nb_pkts); - if (!nb_pkts) { - xs->tx->queue_empty_descs++; - goto out; - } - - __xskq_cons_release(xs->tx); - xskq_prod_write_addr_batch(pool->cq, pool->tx_descs, nb_pkts); - xs->sk.sk_write_space(&xs->sk); + batch = umem_shared ? + xsk_tx_peek_release_shared_desc_batch(pool, nb_pkts) : + __xsk_tx_peek_release_desc_batch(pool, xs, + pool->tx_descs, + nb_pkts); + xsk_tx_commit_batch(pool, &batch); out: rcu_read_unlock(); - return nb_pkts; + return batch.tx_descs; } EXPORT_SYMBOL(xsk_tx_peek_release_desc_batch); diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index 12c9fb29af05..a4089480b22b 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -51,6 +51,7 @@ int xp_alloc_tx_descs(struct xsk_buff_pool *pool, struct xdp_sock *xs, if (!pool->tx_descs) return -ENOMEM; + pool->tx_descs_nentries = nentries; return 0; } diff --git a/net/xdp/xsk_queue.h b/net/xdp/xsk_queue.h index 3e3fbb73d23e..1bc42c8902f4 100644 --- a/net/xdp/xsk_queue.h +++ b/net/xdp/xsk_queue.h @@ -58,6 +58,17 @@ struct parsed_desc { u32 valid; }; +struct xsk_tx_batch { + u32 tx_descs; + u32 reclaim_descs; + bool budget_limited; +}; + +static inline u32 xsk_tx_batch_cq_descs(const struct xsk_tx_batch *batch) +{ + return batch->tx_descs + batch->reclaim_descs; +} + /* The structure of the shared state of the rings are a simple * circular buffer, as outlined in * Documentation/core-api/circular-buffers.rst. For the Rx and @@ -263,17 +274,18 @@ static inline void parse_desc(struct xsk_queue *q, struct xsk_buff_pool *pool, parsed->mb = xp_mb_desc(desc); } -static inline -u32 xskq_cons_read_desc_batch(struct xsk_queue *q, struct xsk_buff_pool *pool, - u32 max) +static inline struct xsk_tx_batch +xskq_cons_read_desc_batch(struct xdp_sock *xs, struct xsk_buff_pool *pool, + struct xdp_desc *descs, u32 max) { - u32 cached_cons = q->cached_cons, nb_entries = 0; - struct xdp_desc *descs = pool->tx_descs; - u32 total_descs = 0, nr_frags = 0; + bool drain = READ_ONCE(xs->drain_cont); + u32 cached_cons, nb_entries = 0; + struct xsk_tx_batch batch = {}; + struct xsk_queue *q = xs->tx; + u32 nr_frags = 0; + + cached_cons = q->cached_cons; - /* track first entry, if stumble upon *any* invalid descriptor, rewind - * current packet that consists of frags and stop the processing - */ while (cached_cons != q->cached_prod && nb_entries < max) { struct xdp_rxtx_ring *ring = (struct xdp_rxtx_ring *)q->ring; u32 idx = cached_cons & q->ring_mask; @@ -283,25 +295,42 @@ u32 xskq_cons_read_desc_batch(struct xsk_queue *q, struct xsk_buff_pool *pool, cached_cons++; parse_desc(q, pool, &descs[nb_entries], &parsed); if (unlikely(!parsed.valid)) - break; + drain = true; + + nr_frags++; + nb_entries++; if (likely(!parsed.mb)) { - total_descs += (nr_frags + 1); - nr_frags = 0; - } else { - nr_frags++; - if (nr_frags == pool->xdp_zc_max_segs) { + if (unlikely(drain)) { + batch.reclaim_descs = nr_frags; + WRITE_ONCE(xs->drain_cont, false); nr_frags = 0; break; } + + batch.tx_descs += nr_frags; + nr_frags = 0; + continue; + } + + if (nr_frags == pool->xdp_zc_max_segs) + drain = true; + } + + if (nr_frags) { + if (drain) { + batch.reclaim_descs = nr_frags; + WRITE_ONCE(xs->drain_cont, true); + } else { + if (nb_entries == max) + batch.budget_limited = true; + cached_cons -= nr_frags; } - nb_entries++; } - cached_cons -= nr_frags; /* Release valid plus any invalid entries */ xskq_cons_release_n(q, cached_cons - q->cached_cons); - return total_descs; + return batch; } /* Functions for consumers */ -- cgit From 97ac08560d236ca17f6606d9e671118e5eae5721 Mon Sep 17 00:00:00 2001 From: Eric Joyner Date: Wed, 22 Jul 2026 21:13:42 -0700 Subject: ethtool: Embed FEC hist ranges as buffer in struct When a driver's .get_fec_stats() handler is called and the driver supports FEC histogram stats, the driver supplies the histogram bin ranges via a pointer. This pointer is assigned while under the netdev ops lock in fec_prepare_data(), but the actual data is only read after the lock is released; so this allows the driver to change the ranges (e.g. from another .get_fec_stats() call) while the current call chain is reading them in fec_fill_reply(). Fix this by adding an ethtool core-owned buffer, ranges_buf, to struct ethtool_fec_hist. Drivers whose ranges are built dynamically (currently just mlx5) fill ranges_buf and then point the existing ranges pointer at it, giving ethtool a consistent copy that stays valid after the netdev ops lock is dropped and later in fec_fill_reply(). Drivers whose ranges are compile-time constants (bnxt, netdevsim) are unaffected by the potential race and keep setting the existing ranges pointer to their constant array, without making copies. Fixes: cc2f08129925 ("ethtool: add FEC bins histogram report") Signed-off-by: Eric Joyner Reviewed-by: Vadim Fedorenko Link: https://patch.msgid.link/20260723041342.39238-1-eric.joyner@amd.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en.h | 1 - drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 7 ------- drivers/net/ethernet/mellanox/mlx5/core/en_stats.c | 19 +++++++++---------- include/linux/ethtool.h | 1 + 4 files changed, 10 insertions(+), 18 deletions(-) (limited to 'include') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index d507289096c2..6867a5aed42c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -984,7 +984,6 @@ struct mlx5e_priv { struct mlx5e_mqprio_rl *mqprio_rl; struct dentry *dfs_root; struct mlx5_devcom_comp_dev *devcom; - struct ethtool_fec_hist_range *fec_ranges; }; static inline u16 mlx5e_stats_nch_read(const struct mlx5e_priv *priv) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index c1acb9012d3f..7d47a1da8b6b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -6415,14 +6415,8 @@ int mlx5e_priv_init(struct mlx5e_priv *priv, if (!priv->channel_stats) goto err_free_tx_rates; - priv->fec_ranges = kzalloc_objs(*priv->fec_ranges, ETHTOOL_FEC_HIST_MAX); - if (!priv->fec_ranges) - goto err_free_channel_stats; - return 0; -err_free_channel_stats: - kfree(priv->channel_stats); err_free_tx_rates: kfree(priv->tx_rates); err_free_txq2sq_stats: @@ -6447,7 +6441,6 @@ void mlx5e_priv_cleanup(struct mlx5e_priv *priv) if (!priv->mdev) return; - kfree(priv->fec_ranges); for (i = 0; i < priv->stats_nch; i++) kvfree(priv->channel_stats[i]); kfree(priv->channel_stats); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index de38b60806c2..e7e6db7f6bf1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -1550,7 +1550,7 @@ static bool fec_rs_validate_hist_type(int mode, int hist_type) static u8 fec_rs_histogram_fill_ranges(struct mlx5e_priv *priv, int mode, - const struct ethtool_fec_hist_range **ranges) + struct ethtool_fec_hist_range *ranges) { struct mlx5_core_dev *mdev = priv->mdev; u32 out[MLX5_ST_SZ_DW(pphcr_reg)] = {0}; @@ -1558,8 +1558,6 @@ fec_rs_histogram_fill_ranges(struct mlx5e_priv *priv, int mode, int sz = MLX5_ST_SZ_BYTES(pphcr_reg); u8 hist_type, num_of_bins; - memset(priv->fec_ranges, 0, - ETHTOOL_FEC_HIST_MAX * sizeof(*priv->fec_ranges)); MLX5_SET(pphcr_reg, in, local_port, 1); if (mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPHCR, 0, 0)) return 0; @@ -1575,12 +1573,11 @@ fec_rs_histogram_fill_ranges(struct mlx5e_priv *priv, int mode, for (int i = 0; i < num_of_bins; i++) { void *bin_range = MLX5_ADDR_OF(pphcr_reg, out, bin_range[i]); - priv->fec_ranges[i].high = MLX5_GET(bin_range_layout, bin_range, - high_val); - priv->fec_ranges[i].low = MLX5_GET(bin_range_layout, bin_range, - low_val); + ranges[i].high = MLX5_GET(bin_range_layout, bin_range, + high_val); + ranges[i].low = MLX5_GET(bin_range_layout, bin_range, + low_val); } - *ranges = priv->fec_ranges; return num_of_bins; } @@ -1622,10 +1619,12 @@ static void fec_set_histograms_stats(struct mlx5e_priv *priv, int mode, case MLX5E_FEC_LLRS_272_257_1: case MLX5E_FEC_RS_544_514_INTERLEAVED_QUAD: num_of_bins = - fec_rs_histogram_fill_ranges(priv, mode, &hist->ranges); - if (num_of_bins) + fec_rs_histogram_fill_ranges(priv, mode, hist->ranges_buf); + if (num_of_bins) { + hist->ranges = hist->ranges_buf; return fec_rs_histogram_fill_stats(priv, num_of_bins, hist); + } break; default: return; diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index 5d491a98265e..12683b5d125e 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -562,6 +562,7 @@ struct ethtool_fec_hist { u64 per_lane[ETHTOOL_MAX_LANES]; } values[ETHTOOL_FEC_HIST_MAX]; const struct ethtool_fec_hist_range *ranges; + struct ethtool_fec_hist_range ranges_buf[ETHTOOL_FEC_HIST_MAX]; }; /** * struct ethtool_fec_stats - statistics for IEEE 802.3 FEC -- cgit From af24e338bf5dafb80f42baa9a0b9e9b57b1c5d9c Mon Sep 17 00:00:00 2001 From: Pauli Virtanen Date: Fri, 24 Jul 2026 23:20:34 +0300 Subject: Bluetooth: ISO: fix race of kfree vs kref_get_unless_zero hci_conn::iso_data is accessed and modified without lock or RCU. This leads to a race [Task hdev->workqueue] [Task 2] iso_recv iso_conn_put(conn) conn = LOAD hcon->iso_data iso_conn_free(conn) iso_conn_hold_unless_zero(conn) hcon->iso_data = NULL kfree(conn) kref_get_unless_zero(&conn->ref) /* UAF */ and also to races in iso_conn_add() vs. iso_conn_free(). Fix by adding spinlock hci_conn::proto_lock and using it to guard hci_conn::iso_data. Fixes: dc26097bdb86 ("Bluetooth: ISO: Use kref to track lifetime of iso_conn") Signed-off-by: Pauli Virtanen Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 4 ++- net/bluetooth/hci_conn.c | 2 ++ net/bluetooth/iso.c | 64 +++++++++++++++++++++++++++++++++------- 3 files changed, 58 insertions(+), 12 deletions(-) (limited to 'include') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index e7133ff87fbf..3df59849dcbe 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -767,9 +767,11 @@ struct hci_conn { struct dentry *debugfs; struct hci_dev *hdev; + + spinlock_t proto_lock; /* lock guarding protocol data */ void *l2cap_data; void *sco_data; - void *iso_data; + void *iso_data __guarded_by(&proto_lock); struct list_head link_list; struct hci_conn *parent; diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index 1966cd153d97..ebb04badf10c 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -1123,6 +1123,8 @@ static struct hci_conn *__hci_conn_add(struct hci_dev *hdev, int type, INIT_DELAYED_WORK(&conn->idle_work, hci_conn_idle); INIT_DELAYED_WORK(&conn->le_conn_timeout, le_conn_timeout); + spin_lock_init(&conn->proto_lock); + atomic_set(&conn->refcnt, 0); hci_dev_hold(hdev); diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index 30de99ba4b30..a461c8a4efed 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -109,9 +109,16 @@ static void iso_conn_free(struct kref *ref) BT_DBG("conn %p", conn); if (conn->hcon) { - conn->hcon->iso_data = NULL; - if (!test_and_set_bit(ISO_CONN_DROPPED, conn->flags)) - hci_conn_drop(conn->hcon); + spin_lock(&conn->hcon->proto_lock); + + /* Check we are not racing with iso_conn_add */ + if (conn->hcon->iso_data == conn) { + conn->hcon->iso_data = NULL; + if (!test_and_set_bit(ISO_CONN_DROPPED, conn->flags)) + hci_conn_drop(conn->hcon); + } + + spin_unlock(&conn->hcon->proto_lock); } kfree_skb(conn->rx_skb); @@ -126,7 +133,21 @@ static void iso_conn_put(struct iso_conn *conn) BT_DBG("conn %p refcnt %d", conn, kref_read(&conn->ref)); + /* The following race vs. iso_conn_del() is possible: + * + * 1. conn->hcon != NULL here + * 2. kref_put puts the last reference + * 3. concurrent iso_conn_del() gets iso_conn_hold_unless_zero() -> NULL + * and returns immediately, so conn->hcon is not cleared + * 4. iso_conn_free() dereferences conn->hcon + * + * To avoid UAF in step 4, take RCU before decrementing the refcount. + */ + rcu_read_lock(); + kref_put(&conn->ref, iso_conn_free); + + rcu_read_unlock(); } static struct iso_conn *iso_conn_hold_unless_zero(struct iso_conn *conn) @@ -205,22 +226,28 @@ static void iso_sock_disable_timer(struct sock *sk) /* ---- ISO connections ---- */ static struct iso_conn *iso_conn_add(struct hci_conn *hcon) + __must_hold(&hcon->hdev->lock) { - struct iso_conn *conn = hcon->iso_data; + struct iso_conn *conn; + + spin_lock(&hcon->proto_lock); - conn = iso_conn_hold_unless_zero(conn); + conn = iso_conn_hold_unless_zero(hcon->iso_data); if (conn) { if (!conn->hcon) { iso_conn_lock(conn); conn->hcon = hcon; iso_conn_unlock(conn); } + spin_unlock(&hcon->proto_lock); return conn; } - conn = kzalloc_obj(*conn); - if (!conn) + conn = kzalloc_obj(*conn, GFP_ATOMIC); + if (!conn) { + spin_unlock(&hcon->proto_lock); return NULL; + } kref_init(&conn->ref); spin_lock_init(&conn->lock); @@ -229,6 +256,8 @@ static struct iso_conn *iso_conn_add(struct hci_conn *hcon) conn->hcon = hcon; conn->tx_sn = 0; + spin_unlock(&hcon->proto_lock); + BT_DBG("hcon %p conn %p", hcon, conn); return conn; @@ -269,10 +298,12 @@ static void iso_chan_del(struct sock *sk, int err) static void iso_conn_del(struct hci_conn *hcon, int err) __must_hold(&hcon->hdev->lock) { - struct iso_conn *conn = hcon->iso_data; + struct iso_conn *conn; struct sock *sk; - conn = iso_conn_hold_unless_zero(conn); + spin_lock(&hcon->proto_lock); + conn = iso_conn_hold_unless_zero(hcon->iso_data); + spin_unlock(&hcon->proto_lock); if (!conn) return; @@ -296,10 +327,12 @@ static void iso_conn_del(struct hci_conn *hcon, int err) done: /* No sk access to conn->hcon any more (lock_sock + hdev->lock) */ + spin_lock(&hcon->proto_lock); iso_conn_lock(conn); conn->hcon = NULL; hcon->iso_data = NULL; iso_conn_unlock(conn); + spin_unlock(&hcon->proto_lock); iso_conn_put(conn); } @@ -421,6 +454,8 @@ static int iso_connect_bis(struct sock *sk) iso_pi(sk)->bc_sid = hcon->sid; } + lockdep_assert_held(&hcon->hdev->lock); + conn = iso_conn_add(hcon); if (!conn) { hci_conn_drop(hcon); @@ -524,6 +559,8 @@ static int iso_connect_cis(struct sock *sk) } } + lockdep_assert_held(&hcon->hdev->lock); + conn = iso_conn_add(hcon); if (!conn) { hci_conn_drop(hcon); @@ -855,8 +892,8 @@ static void iso_sock_disconn(struct sock *sk) */ if (bis_sk) { hcon->state = BT_OPEN; - hcon->iso_data = NULL; - iso_pi(sk)->conn->hcon = NULL; + set_bit(ISO_CONN_DROPPED, iso_pi(sk)->conn->flags); + iso_sock_clear_timer(sk); iso_chan_del(sk, bt_to_errno(hcon->abort_reason)); sock_put(bis_sk); @@ -1306,6 +1343,8 @@ static int iso_listen_bis(struct sock *sk) goto unlock; } + lockdep_assert_held(&hcon->hdev->lock); + conn = iso_conn_add(hcon); if (!conn) { hci_conn_drop(hcon); @@ -2591,7 +2630,10 @@ int iso_recv(struct hci_dev *hdev, u16 handle, struct sk_buff *skb, u16 flags) return -ENOENT; } + spin_lock(&hcon->proto_lock); conn = iso_conn_hold_unless_zero(hcon->iso_data); + spin_unlock(&hcon->proto_lock); + hcon = NULL; hci_dev_unlock(hdev); -- cgit From 6bc85579c3bbb2f088cbac849c5dc2a134dda736 Mon Sep 17 00:00:00 2001 From: Simon Schippers Date: Tue, 28 Jul 2026 11:22:38 +0200 Subject: Revert "ptr_ring: move free-space check into separate helper" This reverts commit fba362c17d9d9211fc51f272156bb84fc23bdf98. __ptr_ring_check_produce() has no users left after reverting commit 1d6e569b7d0c ("tun/tap & vhost-net: avoid ptr_ring tail-drop when a qdisc is present"). Signed-off-by: Simon Schippers Acked-by: Michael S. Tsirkin Link: https://patch.msgid.link/20260728092240.250257-3-simon.schippers@tu-dortmund.de Signed-off-by: Jakub Kicinski --- include/linux/ptr_ring.h | 20 ++------------------ 1 file changed, 2 insertions(+), 18 deletions(-) (limited to 'include') diff --git a/include/linux/ptr_ring.h b/include/linux/ptr_ring.h index c95e891903f0..d2c3629bbe45 100644 --- a/include/linux/ptr_ring.h +++ b/include/linux/ptr_ring.h @@ -96,20 +96,6 @@ static inline bool ptr_ring_full_bh(struct ptr_ring *r) return ret; } -/* Note: callers invoking this in a loop must use a compiler barrier, - * for example cpu_relax(). Callers must hold producer_lock. - */ -static inline int __ptr_ring_check_produce(struct ptr_ring *r) -{ - if (unlikely(!r->size)) - return -EINVAL; - - if (data_race(r->queue[r->producer])) - return -ENOSPC; - - return 0; -} - /* Note: callers invoking this in a loop must use a compiler barrier, * for example cpu_relax(). Callers must hold producer_lock. * Callers are responsible for making sure pointer that is being queued @@ -117,10 +103,8 @@ static inline int __ptr_ring_check_produce(struct ptr_ring *r) */ static inline int __ptr_ring_produce(struct ptr_ring *r, void *ptr) { - int p = __ptr_ring_check_produce(r); - - if (p) - return p; + if (unlikely(!r->size) || data_race(r->queue[r->producer])) + return -ENOSPC; /* Make sure the pointer we are storing points to a valid data. */ /* Pairs with the dependency ordering in __ptr_ring_consume. */ -- cgit From 8f83be72d9f5ef16c4a908450d0d993e8ec99d34 Mon Sep 17 00:00:00 2001 From: Simon Schippers Date: Tue, 28 Jul 2026 11:22:39 +0200 Subject: Revert "vhost-net: wake queue of tun/tap after ptr_ring consume" This reverts commit baf808fe4fcd35767ab732b4ab2ea80dabfd97a6. There is no netdev queue left to wake after reverting commit 1d6e569b7d0c ("tun/tap & vhost-net: avoid ptr_ring tail-drop when a qdisc is present"). Signed-off-by: Simon Schippers Acked-by: Michael S. Tsirkin Link: https://patch.msgid.link/20260728092240.250257-4-simon.schippers@tu-dortmund.de Signed-off-by: Jakub Kicinski --- drivers/net/tun.c | 23 ----------------------- drivers/vhost/net.c | 21 ++++++--------------- include/linux/if_tun.h | 3 --- 3 files changed, 6 insertions(+), 41 deletions(-) (limited to 'include') diff --git a/drivers/net/tun.c b/drivers/net/tun.c index ec5573f545af..39abc3078097 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -3787,29 +3787,6 @@ struct ptr_ring *tun_get_tx_ring(struct file *file) } EXPORT_SYMBOL_GPL(tun_get_tx_ring); -/* Callers must hold ring.consumer_lock */ -void tun_wake_queue(struct file *file, int consumed) -{ - struct tun_file *tfile; - struct tun_struct *tun; - - if (file->f_op != &tun_fops) - return; - - tfile = file->private_data; - if (!tfile) - return; - - rcu_read_lock(); - - tun = rcu_dereference(tfile->tun); - if (tun) - __tun_wake_queue(tun, tfile, consumed); - - rcu_read_unlock(); -} -EXPORT_SYMBOL_GPL(tun_wake_queue); - module_init(tun_init); module_exit(tun_cleanup); MODULE_DESCRIPTION(DRV_DESCRIPTION); diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c index 3e72b9c6af0c..6949b704166d 100644 --- a/drivers/vhost/net.c +++ b/drivers/vhost/net.c @@ -176,21 +176,13 @@ static void *vhost_net_buf_consume(struct vhost_net_buf *rxq) return ret; } -static int vhost_net_buf_produce(struct sock *sk, - struct vhost_net_virtqueue *nvq) +static int vhost_net_buf_produce(struct vhost_net_virtqueue *nvq) { - struct file *file = sk->sk_socket->file; struct vhost_net_buf *rxq = &nvq->rxq; rxq->head = 0; - spin_lock(&nvq->rx_ring->consumer_lock); - rxq->tail = __ptr_ring_consume_batched(nvq->rx_ring, rxq->queue, - VHOST_NET_BATCH); - - if (rxq->tail) - tun_wake_queue(file, rxq->tail); - - spin_unlock(&nvq->rx_ring->consumer_lock); + rxq->tail = ptr_ring_consume_batched(nvq->rx_ring, rxq->queue, + VHOST_NET_BATCH); return rxq->tail; } @@ -217,15 +209,14 @@ static int vhost_net_buf_peek_len(void *ptr) return __skb_array_len_with_tag(ptr); } -static int vhost_net_buf_peek(struct sock *sk, - struct vhost_net_virtqueue *nvq) +static int vhost_net_buf_peek(struct vhost_net_virtqueue *nvq) { struct vhost_net_buf *rxq = &nvq->rxq; if (!vhost_net_buf_is_empty(rxq)) goto out; - if (!vhost_net_buf_produce(sk, nvq)) + if (!vhost_net_buf_produce(nvq)) return 0; out: @@ -1013,7 +1004,7 @@ static int peek_head_len(struct vhost_net_virtqueue *rvq, struct sock *sk) unsigned long flags; if (rvq->rx_ring) - return vhost_net_buf_peek(sk, rvq); + return vhost_net_buf_peek(rvq); spin_lock_irqsave(&sk->sk_receive_queue.lock, flags); head = skb_peek(&sk->sk_receive_queue); diff --git a/include/linux/if_tun.h b/include/linux/if_tun.h index 5f3e206c7a73..80166eb62f41 100644 --- a/include/linux/if_tun.h +++ b/include/linux/if_tun.h @@ -22,7 +22,6 @@ struct tun_msg_ctl { #if defined(CONFIG_TUN) || defined(CONFIG_TUN_MODULE) struct socket *tun_get_socket(struct file *); struct ptr_ring *tun_get_tx_ring(struct file *file); -void tun_wake_queue(struct file *file, int consumed); static inline bool tun_is_xdp_frame(void *ptr) { @@ -56,8 +55,6 @@ static inline struct ptr_ring *tun_get_tx_ring(struct file *f) return ERR_PTR(-EINVAL); } -static inline void tun_wake_queue(struct file *f, int consumed) {} - static inline bool tun_is_xdp_frame(void *ptr) { return false; -- cgit