- Kernel - mailweb.openeuler.org

[PATCH OLK-6.6 0/6] net/sched: fix UAF when destroying locked classifiers (CVE-2026-74700)
by superdcc97＠163.com 02 Sep '26

02 Sep '26

From: Dong Chenchen <dongchenchen2(a)huawei.com> This series fixes CVE-2026-74700, a slab-use-after-free (UAF) in the tc_new_tfilter() path of net/sched/cls_api.c. A short race window in tc_new_tfilter() allows a tcf_proto to be found and briefly referenced by a totally unrelated, unlocked classifier's request. When an unlocked classifier's request (e.g. flower on ingress) loses the tcf_chain_tp_insert_unique() race and ends up dropping the last reference on a locked classifier's proto, the destroy() callback runs without rtnl_lock held. That can corrupt shared classifier structures (e.g. the tc_u_common hash list) and causes a use-after-free. The final patch fixes this by having tcf_proto_destroy() take rtnl_lock around the destroy() callback for locked classifiers whenever rtnl is not already held. The earlier patches are the prerequisite backports that reworked the filter counting / software-bypass bookkeeping (skip_sw counter, filter counter, make skip_sw actually skip software, UAF fix in fl_dump_key, and refine software bypass handling). They are Stable-dep-of prerequisites for the main fix and are required for it to apply correctly on OLK-6.6. Asbjørn Sloth Tønnesen (3): net: sched: cls_api: add skip_sw counter net: sched: cls_api: add filter counter net: sched: make skip_sw actually skip software Jamal Hadi Salim (1): net/sched: cls_api: Always acquire rtnl_lock when destroying locked classifiers Jianbo Liu (1): net: sched: cls_api: fix slab-use-after-free in fl_dump_key Xin Long (1): net: sched: refine software bypass handling in tc_run include/net/pkt_cls.h | 18 +++++++++++++++ include/net/sch_generic.h | 3 +++ net/core/dev.c | 13 +++++++++++ net/sched/cls_api.c | 47 ++++++++++++++++++++++++++++++++++++++- net/sched/cls_bpf.c | 2 ++ net/sched/cls_flower.c | 2 ++ net/sched/cls_matchall.c | 2 ++ net/sched/cls_u32.c | 4 ++++ 8 files changed, 90 insertions(+), 1 deletion(-) -- 2.43.0

1 6

[PATCH OLK-5.10] netfilter: xt_connmark: reject invalid shift parameters
by superdcc97＠163.com 02 Sep '26

02 Sep '26

From: Dong Chenchen <dongchenchen2(a)huawei.com> mainline inclusion from mainline-v7.2-rc3 commit 1b47026fb4b35bac850ad6e8a4ad7fc018e09ebc category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/17361 CVE: CVE-2026-72347 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Revision 2 of the CONNMARK target accepts user-controlled shift parameters and applies them to 32-bit mark values in connmark_tg_shift(). A shift_bits value of 32 or more triggers an undefined-shift bug when the rule is evaluated. Invalid shift_dir values are also accepted and silently fall back to the left-shift path. Reject invalid revision-2 shift parameters in connmark_tg_check() so malformed rules fail at installation time, before they can reach the packet path. Fixes: 472a73e00757 ("netfilter: xt_conntrack: Support bit-shifting for CONNMARK & MARK targets.") Reported-by: Yuan Tan <yuantan098(a)gmail.com> Reported-by: Zhengchuan Liang <zcliangcn(a)gmail.com> Reported-by: Xin Liu <dstsmallbird(a)foxmail.com> Assisted-by: Codex:GPT-5.4 Signed-off-by: Wyatt Feng <bronzed_45_vested(a)icloud.com> Reviewed-by: Ren Wei <enjou1224z(a)gmail.com> Reviewed-by: Phil Sutter <phil(a)nwl.cc> Signed-off-by: Florian Westphal <fw(a)strlen.de> Conflicts: net/netfilter/xt_connmark.c [commit 0bfcb7b71e735 registers CONNMARK rev2 separately for NFPROTO_IPV4 and NFPROTO_IPV6, which is not merged] Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- net/netfilter/xt_connmark.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/net/netfilter/xt_connmark.c b/net/netfilter/xt_connmark.c index ad3c033db64e..3004c5229938 100644 --- a/net/netfilter/xt_connmark.c +++ b/net/netfilter/xt_connmark.c @@ -112,6 +112,16 @@ static int connmark_tg_check(const struct xt_tgchk_param *par) return ret; } +static int connmark_tg_check_v2(const struct xt_tgchk_param *par) +{ + const struct xt_connmark_tginfo2 *info = par->targinfo; + + if (info->shift_dir > D_SHIFT_RIGHT || info->shift_bits >= 32) + return -EINVAL; + + return connmark_tg_check(par); +} + static void connmark_tg_destroy(const struct xt_tgdtor_param *par) { nf_ct_netns_put(par->net, par->family); @@ -162,7 +172,7 @@ static struct xt_target connmark_tg_reg[] __read_mostly = { .name = "CONNMARK", .revision = 2, .family = NFPROTO_UNSPEC, - .checkentry = connmark_tg_check, + .checkentry = connmark_tg_check_v2, .target = connmark_tg_v2, .targetsize = sizeof(struct xt_connmark_tginfo2), .destroy = connmark_tg_destroy, -- 2.43.0

2 1

[PATCH OLK-5.10] openvswitch: fix GSO userspace truncation underflow
by Yi Yang 02 Sep '26

02 Sep '26

From: Kyle Zeng <kylebot(a)openai.com> stable inclusion from stable-v5.10.265 commit 50a6a85f3d6b1d22d8436848606cdef5d2c490b4 category: bugfix bugzilla: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 4032f8ed10fcb84d41c508dfb04be96589f78dfe ] OVS_ACTION_ATTR_TRUNC currently stores a delta from the original skb length in OVS_CB(skb)->cutlen. When a later userspace action segments a GSO skb, queue_gso_packets() reuses that delta for each smaller segment. A segment can then reach queue_userspace_packet() with cutlen greater than skb->len, underflowing the length passed to skb_zerocopy(). Store the maximum preserved length instead and bound each consumer against the current skb length. Use U32_MAX as the no-truncation sentinel so the value remains valid if skb geometry changes before a consumer handles it. Fixes: f2a4d086ed4c ("openvswitch: Add packet truncation support.") Cc: stable(a)vger.kernel.org Assisted-by: Codex:gpt-5.5 Signed-off-by: Kyle Zeng <kylebot(a)openai.com> Reviewed-by: Ilya Maximets <i.maximets(a)ovn.org> Reviewed-by: Aaron Conole <aconole(a)redhat.com> Link: https://patch.msgid.link/20260707221635.27489-1-kylebot@openai.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> [5.10.y supports neither OVS_ACTION_ATTR_PSAMPLE nor OVS drop reasons] Signed-off-by: Ilya Maximets <i.maximets(a)ovn.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yi Yang <yiyang13(a)huawei.com> --- net/openvswitch/actions.c | 15 +++++---------- net/openvswitch/datapath.c | 25 ++++++++++++++----------- net/openvswitch/datapath.h | 2 +- net/openvswitch/vport.c | 2 +- 4 files changed, 21 insertions(+), 23 deletions(-) diff --git a/net/openvswitch/actions.c b/net/openvswitch/actions.c index a02ea493c269..1e087b7f6dc6 100644 --- a/net/openvswitch/actions.c +++ b/net/openvswitch/actions.c @@ -918,12 +918,8 @@ static void do_output(struct datapath *dp, struct sk_buff *skb, int out_port, u16 mru = OVS_CB(skb)->mru; u32 cutlen = OVS_CB(skb)->cutlen; - if (unlikely(cutlen > 0)) { - if (skb->len - cutlen > ovs_mac_header_len(key)) - pskb_trim(skb, skb->len - cutlen); - else - pskb_trim(skb, ovs_mac_header_len(key)); - } + if (unlikely(cutlen < skb->len)) + pskb_trim(skb, max(cutlen, ovs_mac_header_len(key))); if (likely(!mru || (skb->len <= mru + vport->dev->hard_header_len))) { @@ -1299,22 +1295,21 @@ static int do_execute_actions(struct datapath *dp, struct sk_buff *skb, clone = skb_clone(skb, GFP_ATOMIC); if (clone) do_output(dp, clone, port, key); - OVS_CB(skb)->cutlen = 0; + OVS_CB(skb)->cutlen = U32_MAX; break; } case OVS_ACTION_ATTR_TRUNC: { struct ovs_action_trunc *trunc = nla_data(a); - if (skb->len > trunc->max_len) - OVS_CB(skb)->cutlen = skb->len - trunc->max_len; + OVS_CB(skb)->cutlen = trunc->max_len; break; } case OVS_ACTION_ATTR_USERSPACE: output_userspace(dp, skb, key, a, attr, len, OVS_CB(skb)->cutlen); - OVS_CB(skb)->cutlen = 0; + OVS_CB(skb)->cutlen = U32_MAX; break; case OVS_ACTION_ATTR_HASH: diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c index 1c69aa986633..529eb9d6acab 100644 --- a/net/openvswitch/datapath.c +++ b/net/openvswitch/datapath.c @@ -240,7 +240,7 @@ void ovs_dp_process_packet(struct sk_buff *skb, struct sw_flow_key *key) upcall.cmd = OVS_PACKET_CMD_MISS; upcall.portid = ovs_vport_find_upcall_portid(p, skb); upcall.mru = OVS_CB(skb)->mru; - error = ovs_dp_upcall(dp, skb, key, &upcall, 0); + error = ovs_dp_upcall(dp, skb, key, &upcall, U32_MAX); switch (error) { case 0: case -EAGAIN: @@ -400,7 +400,8 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb, struct sk_buff *nskb = NULL; struct sk_buff *user_skb = NULL; /* to be queued to userspace */ struct nlattr *nla; - size_t len; + size_t msg_size; + size_t skb_len; unsigned int hlen; int err, dp_ifindex; u64 hash; @@ -421,7 +422,8 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb, skb = nskb; } - if (nla_attr_size(skb->len) > USHRT_MAX) { + skb_len = min(skb->len, cutlen); + if (nla_attr_size(skb_len) > USHRT_MAX) { err = -EFBIG; goto out; } @@ -436,13 +438,13 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb, * padding logic. Only perform zerocopy if padding is not required. */ if (dp->user_features & OVS_DP_F_UNALIGNED) - hlen = skb_zerocopy_headlen(skb); + hlen = min(skb_zerocopy_headlen(skb), cutlen); else - hlen = skb->len; + hlen = skb_len; - len = upcall_msg_size(upcall_info, hlen - cutlen, - OVS_CB(skb)->acts_origlen); - user_skb = genlmsg_new(len, GFP_ATOMIC); + msg_size = upcall_msg_size(upcall_info, hlen, + OVS_CB(skb)->acts_origlen); + user_skb = genlmsg_new(msg_size, GFP_ATOMIC); if (!user_skb) { err = -ENOMEM; goto out; @@ -503,7 +505,7 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb, } /* Add OVS_PACKET_ATTR_LEN when packet is truncated */ - if (cutlen > 0 && + if (skb_len < skb->len && nla_put_u32(user_skb, OVS_PACKET_ATTR_LEN, skb->len)) { err = -ENOBUFS; goto out; @@ -528,9 +530,9 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb, err = -ENOBUFS; goto out; } - nla->nla_len = nla_attr_size(skb->len - cutlen); + nla->nla_len = nla_attr_size(skb_len); - err = skb_zerocopy(user_skb, skb, skb->len - cutlen, hlen); + err = skb_zerocopy(user_skb, skb, skb_len, hlen); if (err) goto out; @@ -587,6 +589,7 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info) packet->ignore_df = 1; } OVS_CB(packet)->mru = mru; + OVS_CB(packet)->cutlen = U32_MAX; if (a[OVS_PACKET_ATTR_HASH]) { hash = nla_get_u64(a[OVS_PACKET_ATTR_HASH]); diff --git a/net/openvswitch/datapath.h b/net/openvswitch/datapath.h index 38f7d3e66ca6..62e7ad32f837 100644 --- a/net/openvswitch/datapath.h +++ b/net/openvswitch/datapath.h @@ -96,7 +96,7 @@ struct datapath { * @mru: The maximum received fragement size; 0 if the packet is not * fragmented. * @acts_origlen: The netlink size of the flow actions applied to this skb. - * @cutlen: The number of bytes from the packet end to be removed. + * @cutlen: The number of bytes in the packet to preserve on output. */ struct ovs_skb_cb { struct vport *input_vport; diff --git a/net/openvswitch/vport.c b/net/openvswitch/vport.c index da733b92ae8a..ced707ff6dd9 100644 --- a/net/openvswitch/vport.c +++ b/net/openvswitch/vport.c @@ -436,7 +436,7 @@ int ovs_vport_receive(struct vport *vport, struct sk_buff *skb, OVS_CB(skb)->input_vport = vport; OVS_CB(skb)->mru = 0; - OVS_CB(skb)->cutlen = 0; + OVS_CB(skb)->cutlen = U32_MAX; if (unlikely(dev_net(skb->dev) != ovs_dp_get_net(vport->dp))) { u32 mark; -- 2.25.1

2 1

[PATCH OLK-5.10 0/2] CVE-2026-72390: Fix race and UAF in sch_teql
by superdcc97＠163.com 02 Sep '26

02 Sep '26

From: Dong Chenchen <dongchenchen2(a)huawei.com> This series backports the CVE-2026-72390 fix for the OLK-5.10 kernel. Patch 1 introduces a per-master slaves_lock spinlock and RCU accessors around the master->slaves list to prevent concurrent mutation and use-after-free in the TEQL qdisc. Patch 2 is a follow-up that switches the rcu/spin primitives in teql_master_xmit() from the _bh variants to plain variants, since ndo_start_xmit is already called with BH disabled. Both patches preserve the original author sign-offs and add the adapter signature. Jamal Hadi Salim (2): net/sched: sch_teql: Introduce slaves_lock to avoid race condition and UAF net/sched: sch_teql: move rcu_read_lock()/spin_lock() from _bh variants net/sched/sch_teql.c | 128 +++++++++++++++++++++++++++++-------------- 1 file changed, 88 insertions(+), 40 deletions(-) -- 2.43.0

2 3

[PATCH OLK-5.10] net/sched: cls_u32: validate offshift to prevent shift-out-of-bounds
by superdcc97＠163.com 02 Sep '26

02 Sep '26

From: "Cen Zhang (Microsoft)" <blbllhy(a)gmail.com> mainline inclusion from mainline-v7.2-rc6 commit aef96eead2860cbfa371e4471d4f04412213b958 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/17793 CVE: CVE-2026-74544 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- u32_change() copies the user-provided tc_u32_sel.offshift (unsigned char, 0-255) into the kernel knode object without bounds validation. When a packet later hits u32_classify() with TC_U32_VAROFFSET set, it evaluates `ntohs(offmask & *data) >> offshift` where the left operand is a 16-bit value promoted to a 32-bit int. Any offshift >= 32 is undefined behavior per C11 6.5.7p3, triggerable by an unprivileged user via user/network namespaces. UBSAN: shift-out-of-bounds in net/sched/cls_u32.c:236:43 shift exponent 32 is too large for 32-bit type int Fix this by rejecting offshift >= 16 during filter creation in u32_change(). Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: AutonomousCodeSecurity(a)microsoft.com Link: https://lore.kernel.org/all/20260720034514.23053-1-blbllhy@gmail.com Signed-off-by: Cen Zhang (Microsoft) <blbllhy(a)gmail.com> Acked-by: Jamal Hadi Salim <jhs(a)mojatatu.com> Tested-by: Jamal Hadi Salim <jhs(a)mojatatu.com> Tested-by: Victor Nogueira <victor(a)mojatatu.com> Link: https://patch.msgid.link/20260723044955.89471-1-blbllhy@gmail.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Conflicts: net/sched/cls_u32.c [commit 69050f8d6d07 is not backport, which lead to context conflict] Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- net/sched/cls_u32.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/net/sched/cls_u32.c b/net/sched/cls_u32.c index e501390ccd75..0db59ea65e48 100644 --- a/net/sched/cls_u32.c +++ b/net/sched/cls_u32.c @@ -1067,6 +1067,13 @@ static int u32_change(struct net *net, struct sk_buff *in_skb, goto erridr; } + if (s->offshift >= 16) { + NL_SET_ERR_MSG_MOD(extack, + "offshift must be less than 16"); + err = -EINVAL; + goto erridr; + } + n = kzalloc(struct_size(n, sel.keys, s->nkeys), GFP_KERNEL); if (n == NULL) { err = -ENOBUFS; -- 2.43.0

2 1

[PATCH OLK-5.10] netfilter: nf_tables: make nft_object rhltable per table
by superdcc97＠163.com 02 Sep '26

02 Sep '26

From: Pablo Neira Ayuso <pablo(a)netfilter.org> mainline inclusion from mainline-v7.2-rc6 commit f4f699790590bd0896c48a71e9232a65198f92f0 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/17332 CVE: CVE-2026-74565 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- 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 <fw(a)strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: include/net/netfilter/nf_tables.h net/netfilter/nf_tables_api.c [commit f4f699790590bd0896c48a71e9232a65198f92f0 ("netfilter: nf_tables: make nft_object rhltable per table") is not backport, which lead to conflicts] [Structural conflict, no upstream prerequisite found in Fixes..CVE range: all flagged deviations are 5.10 branch code shape, not missing logic. classification: Class B forced-minimal adaptation; target control flow preserved; only CVE-related per-object-hashtable changes applied.] [Deviations: (1) 5.10 nf_tables_table_destroy() takes struct nft_ctx *ctx instead of struct nft_table *table, so objname_ht is destroyed via ctx->table->objname_ht, equivalent to upstream table->objname_ht; (2) 5.10 commit loop has no loop-local table variable (upstream has table = trans->table), so nft_obj_del() is called with trans->ctx.table, which is the same table the object was inserted into; (3) 5.10 merged nf_tables_getobj() (upstream split into nf_tables_getobj_single), the const removal on table applies there; (4) 5.10 module_init uses positional err1..err6 labels (upstream uses named labels); removing the global rhltable_init block requires renumbering so that nft_offload_init failure unwinds through err4 (unregister_netdevice_notifier) and nfnetlink_subsys_register failure through err5 (nft_offload_exit), preserving the exact upstream cleanup order and removing only the global rhltable_destroy for the object-name hashtable.] [Refcount/lifetime safety: object name-hashtable insertion and removal are paired at the table level (rhltable_insert in nf_tables_newobj vs rhltable_remove in nft_obj_del); the per-table hashtable is initialized in nf_tables_newtable and torn down in nf_tables_table_destroy / err_trans path; no dev_hold/refcount/lock operations were added, removed, or reordered relative to upstream; the CVE fix (isolating object lookup from concurrently-released objects across netns) is preserved verbatim.] Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- include/net/netfilter/nf_tables.h | 4 +++- net/netfilter/nf_tables_api.c | 40 +++++++++++++++---------------- 2 files changed, 22 insertions(+), 22 deletions(-) diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index 9168a6a8934c..fb9de90d00e7 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -1103,6 +1103,7 @@ 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 @@ -1118,6 +1119,7 @@ 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; @@ -1190,7 +1192,7 @@ 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 2beef989cced..fd498aecdd13 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -44,8 +44,6 @@ 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 *); @@ -1300,6 +1298,10 @@ static int nf_tables_newtable(struct net *net, struct sock *nlsk, 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); @@ -1316,6 +1318,8 @@ static int nf_tables_newtable(struct net *net, struct sock *nlsk, 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); @@ -1474,6 +1478,7 @@ static void nf_tables_table_destroy(struct nft_ctx *ctx) return; rhltable_destroy(&ctx->table->chains_ht); + rhltable_destroy(&ctx->table->objname_ht); kfree(ctx->table->name); kfree(ctx->table->udata); kfree(ctx->table); @@ -6150,7 +6155,7 @@ 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) { @@ -6166,7 +6171,7 @@ 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; @@ -6436,7 +6441,7 @@ static int nf_tables_newobj(struct net *net, struct sock *nlsk, 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; @@ -6622,7 +6627,7 @@ static int nf_tables_getobj(struct net *net, struct sock *nlsk, const struct nfgenmsg *nfmsg = nlmsg_data(nlh); u8 genmask = nft_genmask_cur(net); int family = nfmsg->nfgen_family; - const struct nft_table *table; + struct nft_table *table; struct nft_object *obj; struct sk_buff *skb2; bool reset = false; @@ -8159,9 +8164,9 @@ 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); } @@ -8741,7 +8746,7 @@ static int nf_tables_commit(struct net *net, struct sk_buff *skb) } break; case NFT_MSG_DELOBJ: - nft_obj_del(nft_trans_obj(trans)); + nft_obj_del(trans->ctx.table, nft_trans_obj(trans)); nf_tables_obj_notify(&trans->ctx, nft_trans_obj(trans), NFT_MSG_DELOBJ); break; @@ -8988,7 +8993,7 @@ static int __nf_tables_abort(struct net *net, enum nfnl_abort_action action) nft_trans_destroy(trans); } else { nft_use_dec_restore(&trans->ctx.table->use); - nft_obj_del(nft_trans_obj(trans)); + nft_obj_del(trans->ctx.table, nft_trans_obj(trans)); } break; case NFT_MSG_DELOBJ: @@ -9633,7 +9638,7 @@ 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); } @@ -9734,26 +9739,20 @@ static int __init nf_tables_module_init(void) if (err < 0) goto err3; - err = rhltable_init(&nft_objname_ht, &nft_objname_ht_params); - if (err < 0) - goto err4; - err = nft_offload_init(); if (err < 0) - goto err5; + goto err4; /* must be last */ err = nfnetlink_subsys_register(&nf_tables_subsys); if (err < 0) - goto err6; + goto err5; nft_chain_route_init(); return err; -err6: - nft_offload_exit(); err5: - rhltable_destroy(&nft_objname_ht); + nft_offload_exit(); err4: unregister_netdevice_notifier(&nf_tables_flowtable_notifier); err3: @@ -9777,7 +9776,6 @@ static void __exit nf_tables_module_exit(void) cancel_work_sync(&trans_gc_work); cancel_work_sync(&trans_destroy_work); rcu_barrier(); - rhltable_destroy(&nft_objname_ht); nf_tables_core_module_exit(); } -- 2.43.0

2 1

[PATCH OLK-5.10] xfrm: fix stale skb->prev after async crypto steals a GSO segment
by superdcc97＠163.com 02 Sep '26

02 Sep '26

From: Petr Wozniak <petr.wozniak(a)gmail.com> mainline inclusion from mainline-v7.2-rc4 commit 3f4c3919baf0944ad96580467c302bc6c7758b00 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/17032 CVE: CVE-2026-68426 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- skb_gso_segment() leaves the segment list head with ->prev pointing at the last segment, an invariant validate_xmit_skb_list() relies on when it sets its tail pointer (tail = skb->prev). When validate_xmit_xfrm() walks a GSO list and some segments are stolen by async crypto (->xmit() returns -EINPROGRESS), those segments are unlinked from the list but the head ->prev is never updated. If the last segment is the one stolen, the returned head still has ->prev pointing at it, even though it is now owned by the crypto engine and may be freed. validate_xmit_skb_list() later does tail->next = skb, writing through that stale pointer -- a use-after-free. Repoint skb->prev at the last retained segment before returning. Fixes: f53c723902d1 ("net: Add asynchronous callbacks for xfrm on layer 2.") Signed-off-by: Petr Wozniak <petr.wozniak(a)gmail.com> Signed-off-by: Steffen Klassert <steffen.klassert(a)secunet.com> Conflicts: net/xfrm/xfrm_device.c [mainline commit 6860b467f569 ("xfrm: propagate -EINPROGRESS from validate_xmit_xfrm()") is not backport, which lead to conflicts: this is a Class B non-Stable-dep-of prerequisite (pure control-flow change turning NULL return into ERR_PTR(-EINPROGRESS), no dev_hold/refcount/ lock/UAF involvement). Applied forced minimal adaptation per P3: kept target's "return skb;" (5.10 callers already handle NULL via "if (!skb) continue;"), applied only the CVE change "if (skb) skb->prev = pskb;" to repoint the GSO head ->prev at the last retained segment so validate_xmit_skb_list() cannot chain onto a segment stolen by async crypto (use-after-free). Refcount/lifetime analysis: added code takes no reference and acquires no lock; ownership of stolen skbs remains with the crypto engine via the xfrm_dev_resume() path, unchanged] Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- net/xfrm/xfrm_device.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/net/xfrm/xfrm_device.c b/net/xfrm/xfrm_device.c index 0e7424b8a5fa..062caeb893a5 100644 --- a/net/xfrm/xfrm_device.c +++ b/net/xfrm/xfrm_device.c @@ -215,6 +215,14 @@ struct sk_buff *validate_xmit_xfrm(struct sk_buff *skb, netdev_features_t featur pskb = skb2; } + /* skb_gso_segment() set skb->prev to the last segment, but async + * crypto may have stolen it above without updating ->prev. Repoint + * it at the last retained segment so validate_xmit_skb_list() does + * not chain onto a segment now owned by the crypto engine. + */ + if (skb) + skb->prev = pskb; + return skb; } EXPORT_SYMBOL_GPL(validate_xmit_xfrm); -- 2.43.0

2 2

[PATCH OLK-6.6] tcp: clear sock_ops cb flags before force-closing a child socket
by JiangJieHua 02 Sep '26

02 Sep '26

From: Sechang Lim <rhkrqnwk98(a)gmail.com> mainline inclusion from mainline-v7.2-rc1 commit 990348e5bb457697c2f1f7f7b65154a3334d9d2b category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/17954 CVE: CVE-2026-74268 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- A child socket inherits the listener's bpf_sock_ops_cb_flags via sk_clone_lock(). If its setup fails in tcp_v4_syn_recv_sock() / tcp_v6_syn_recv_sock(), the child is freed through put_and_exit, where inet_csk_prepare_forced_close() drops the socket lock and tcp_done() runs without it. If BPF_SOCK_OPS_STATE_CB_FLAG was inherited, tcp_done() -> tcp_set_state() calls tcp_call_bpf(), which expects the lock and trips sock_owned_by_me(): WARNING: include/net/sock.h:1799 at tcp_set_state+0x433/0x550 RIP: 0010:tcp_set_state+0x433/0x550 include/net/sock.h:1799 Call Trace: <IRQ> tcp_done+0xba/0x250 net/ipv4/tcp.c:5095 tcp_v4_syn_recv_sock+0x850/0xa50 net/ipv4/tcp_ipv4.c:1787 tcp_check_req+0xf30/0x1360 net/ipv4/tcp_minisocks.c:926 tcp_v4_rcv+0x1047/0x1b50 net/ipv4/tcp_ipv4.c:2164 </IRQ> The child is freed before it is ever established, so it should run no sock_ops callback. Clear its cb flags in inet_csk_prepare_for_destroy_sock(), the common point for the IPv4, IPv6 and chtls forced-close paths and for the MPTCP ->syn_recv_sock() failure path (dispose_child), which reaches tcp_done() on a child that was never established too. Suggested-by: Jiayuan Chen <jiayuan.chen(a)linux.dev> Fixes: d44874910a26 ("bpf: Add BPF_SOCK_OPS_STATE_CB") Signed-off-by: Sechang Lim <rhkrqnwk98(a)gmail.com> Reviewed-by: Jiayuan Chen <jiayuan.chen(a)linux.dev> Reviewed-by: Kuniyuki Iwashima <kuniyu(a)google.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Link: https://patch.msgid.link/20260611092923.1895982-1-rhkrqnwk98@gmail.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Conflicts: include/net/inet_connection_sock.h net/ipv4/inet_connection_sock.c [In OLK-6.6, inet_csk_prepare_for_destroy_sock() remains a static inline function in include/net/inet_connection_sock.h, while upstream already moved it into inet_connection_sock.c via commit 7051b54fb5aa (due to header dependency). This backport cherry-picks the inline-to-non-inline conversion (changing the definition in inet_connection_sock.h to a declaration), retains OLK-6.6's this_cpu_inc(*sk->sk_prot->orphan_count) implementation without pulling in the orphan_count refactoring, and adds a sk->sk_protocol == IPPROTO_TCP guard to isolate the DCCP path.] Signed-off-by: JiangJieHua <jiangjiehua1(a)huawei.com> --- include/net/inet_connection_sock.h | 8 +------- include/net/tcp.h | 9 +++++++++ net/ipv4/inet_connection_sock.c | 9 +++++++++ 3 files changed, 19 insertions(+), 7 deletions(-) diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 327ae14ca82b2..c16f600e5c5c3 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -303,14 +303,8 @@ reqsk_timeout(struct request_sock *req, unsigned long max_timeout) return (unsigned long)min_t(u64, timeout, max_timeout); } -static inline void inet_csk_prepare_for_destroy_sock(struct sock *sk) -{ - /* The below has to be done to allow calling inet_csk_destroy_sock */ - sock_set_flag(sk, SOCK_DEAD); - this_cpu_inc(*sk->sk_prot->orphan_count); -} - void inet_csk_destroy_sock(struct sock *sk); +void inet_csk_prepare_for_destroy_sock(struct sock *sk); void inet_csk_prepare_forced_close(struct sock *sk); /* diff --git a/include/net/tcp.h b/include/net/tcp.h index 8674996bd39c6..e1a345d5a40b1 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -2515,6 +2515,11 @@ static inline int tcp_call_bpf_3arg(struct sock *sk, int op, u32 arg1, u32 arg2, return tcp_call_bpf(sk, op, 3, args); } +static inline void tcp_clear_sock_ops_cb_flags(struct sock *sk) +{ + tcp_sk(sk)->bpf_sock_ops_cb_flags = 0; +} + #else static inline int tcp_call_bpf(struct sock *sk, int op, u32 nargs, u32 *args) { @@ -2532,6 +2537,10 @@ static inline int tcp_call_bpf_3arg(struct sock *sk, int op, u32 arg1, u32 arg2, return -EPERM; } +static inline void tcp_clear_sock_ops_cb_flags(struct sock *sk) +{ +} + #endif static inline u32 tcp_timeout_init(struct sock *sk) diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 9f1d17cd5eebd..bca256306c9ee 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1240,6 +1240,15 @@ void inet_csk_destroy_sock(struct sock *sk) } EXPORT_SYMBOL(inet_csk_destroy_sock); +void inet_csk_prepare_for_destroy_sock(struct sock *sk) +{ + /* The below has to be done to allow calling inet_csk_destroy_sock */ + if (sk->sk_protocol == IPPROTO_TCP) + tcp_clear_sock_ops_cb_flags(sk); + sock_set_flag(sk, SOCK_DEAD); + this_cpu_inc(*sk->sk_prot->orphan_count); +} + /* This function allows to force a closure of a socket after the call to * tcp/dccp_create_openreq_child(). */ -- 2.33.8

2 1

[PATCH OLK-6.6] tcp: clear sock_ops cb flags before force-closing a child socket
by JiangJieHua 02 Sep '26

02 Sep '26

From: Sechang Lim <rhkrqnwk98(a)gmail.com> mainline inclusion from mainline-v7.2-rc1 commit 990348e5bb457697c2f1f7f7b65154a3334d9d2b category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/17954 CVE: CVE-2026-74268 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- A child socket inherits the listener's bpf_sock_ops_cb_flags via sk_clone_lock(). If its setup fails in tcp_v4_syn_recv_sock() / tcp_v6_syn_recv_sock(), the child is freed through put_and_exit, where inet_csk_prepare_forced_close() drops the socket lock and tcp_done() runs without it. If BPF_SOCK_OPS_STATE_CB_FLAG was inherited, tcp_done() -> tcp_set_state() calls tcp_call_bpf(), which expects the lock and trips sock_owned_by_me(): WARNING: include/net/sock.h:1799 at tcp_set_state+0x433/0x550 RIP: 0010:tcp_set_state+0x433/0x550 include/net/sock.h:1799 Call Trace: <IRQ> tcp_done+0xba/0x250 net/ipv4/tcp.c:5095 tcp_v4_syn_recv_sock+0x850/0xa50 net/ipv4/tcp_ipv4.c:1787 tcp_check_req+0xf30/0x1360 net/ipv4/tcp_minisocks.c:926 tcp_v4_rcv+0x1047/0x1b50 net/ipv4/tcp_ipv4.c:2164 </IRQ> The child is freed before it is ever established, so it should run no sock_ops callback. Clear its cb flags in inet_csk_prepare_for_destroy_sock(), the common point for the IPv4, IPv6 and chtls forced-close paths and for the MPTCP ->syn_recv_sock() failure path (dispose_child), which reaches tcp_done() on a child that was never established too. Suggested-by: Jiayuan Chen <jiayuan.chen(a)linux.dev> Fixes: d44874910a26 ("bpf: Add BPF_SOCK_OPS_STATE_CB") Signed-off-by: Sechang Lim <rhkrqnwk98(a)gmail.com> Reviewed-by: Jiayuan Chen <jiayuan.chen(a)linux.dev> Reviewed-by: Kuniyuki Iwashima <kuniyu(a)google.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Link: https://patch.msgid.link/20260611092923.1895982-1-rhkrqnwk98@gmail.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Conflicts: net/ipv4/inet_connection_sock.h net/ipv4/inet_connection_sock.c [In OLK-6.6, inet_csk_prepare_for_destroy_sock() remains a static inline function in include/net/inet_connection_sock.h, while upstream already moved it into inet_connection_sock.c via commit 7051b54fb5aa (due to header dependency). This backport cherry-picks the inline-to-non-inline conversion (changing the definition in inet_connection_sock.h to a declaration), retains OLK-6.6's this_cpu_inc(*sk->sk_prot->orphan_count) implementation without pulling in the orphan_count refactoring, and adds a sk->sk_protocol == IPPROTO_TCP guard to isolate the DCCP path.] Signed-off-by: JiangJieHua <jiangjiehua1(a)huawei.com> --- include/net/inet_connection_sock.h | 8 +------- include/net/tcp.h | 9 +++++++++ net/ipv4/inet_connection_sock.c | 9 +++++++++ 3 files changed, 19 insertions(+), 7 deletions(-) diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 327ae14ca82b2..c16f600e5c5c3 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -303,14 +303,8 @@ reqsk_timeout(struct request_sock *req, unsigned long max_timeout) return (unsigned long)min_t(u64, timeout, max_timeout); } -static inline void inet_csk_prepare_for_destroy_sock(struct sock *sk) -{ - /* The below has to be done to allow calling inet_csk_destroy_sock */ - sock_set_flag(sk, SOCK_DEAD); - this_cpu_inc(*sk->sk_prot->orphan_count); -} - void inet_csk_destroy_sock(struct sock *sk); +void inet_csk_prepare_for_destroy_sock(struct sock *sk); void inet_csk_prepare_forced_close(struct sock *sk); /* diff --git a/include/net/tcp.h b/include/net/tcp.h index 8674996bd39c6..e1a345d5a40b1 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -2515,6 +2515,11 @@ static inline int tcp_call_bpf_3arg(struct sock *sk, int op, u32 arg1, u32 arg2, return tcp_call_bpf(sk, op, 3, args); } +static inline void tcp_clear_sock_ops_cb_flags(struct sock *sk) +{ + tcp_sk(sk)->bpf_sock_ops_cb_flags = 0; +} + #else static inline int tcp_call_bpf(struct sock *sk, int op, u32 nargs, u32 *args) { @@ -2532,6 +2537,10 @@ static inline int tcp_call_bpf_3arg(struct sock *sk, int op, u32 arg1, u32 arg2, return -EPERM; } +static inline void tcp_clear_sock_ops_cb_flags(struct sock *sk) +{ +} + #endif static inline u32 tcp_timeout_init(struct sock *sk) diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 9f1d17cd5eebd..bca256306c9ee 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1240,6 +1240,15 @@ void inet_csk_destroy_sock(struct sock *sk) } EXPORT_SYMBOL(inet_csk_destroy_sock); +void inet_csk_prepare_for_destroy_sock(struct sock *sk) +{ + /* The below has to be done to allow calling inet_csk_destroy_sock */ + if (sk->sk_protocol == IPPROTO_TCP) + tcp_clear_sock_ops_cb_flags(sk); + sock_set_flag(sk, SOCK_DEAD); + this_cpu_inc(*sk->sk_prot->orphan_count); +} + /* This function allows to force a closure of a socket after the call to * tcp/dccp_create_openreq_child(). */ -- 2.33.8

2 1

[PATCH OLK-5.10] bpf: Fix UAF in sock clone early bailouts
by superdcc97＠163.com 02 Sep '26

02 Sep '26

From: Matt Bobrowski <mattbobrowski(a)google.com> mainline inclusion from mainline-v7.2-rc4 commit 7cbd0c4cebe4c9f678d15e6b9ba975e1155a107f category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/17028 CVE: CVE-2026-68399 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Similar to recent commit 9b51a6155d14 ("bpf,fork: wipe ->bpf_storage before bailouts that access it"), sk_clone() performs an initial shallow copy of the socket field ->sk_bpf_storage via sock_copy() for the cloned socket newsk. If sk_clone() bails out early (e.g. if sk_filter_charge() fails) prior to calling bpf_sk_storage_clone(), newsk->sk_bpf_storage still points to the parent socket's BPF local storage. When newsk is subsequently freed via sk_free(), the deallocation path (__sk_destruct() -> bpf_sk_storage_free()) destroys the parent socket's BPF local storage, leading to a use-after-free (UAF) on the parent socket. Fix this by resetting newsk->sk_bpf_storage to NULL immediately after sock_copy() in sk_clone(), and remove the now redundant initialization from bpf_sk_storage_clone(). Fixes: 6ac99e8f23d4 ("bpf: Introduce bpf sk local storage") Fixes: f12dd75959b0 ("bpf: net: Set sk_bpf_storage back to NULL for cloned sk") Signed-off-by: Matt Bobrowski <mattbobrowski(a)google.com> Signed-off-by: Daniel Borkmann <daniel(a)iogearbox.net> Reviewed-by: Kuniyuki Iwashima <kuniyu(a)google.com> Acked-by: Daniel Borkmann <daniel(a)iogearbox.net> Link: https://lore.kernel.org/bpf/20260709025316.999913-1-mattbobrowski@google.com Conflicts: net/core/bpf_sk_storage.c net/core/sock.c [commit 0db4941d9dae1 uses rcu_read_lock_dont_migrate() which does not exist on 5.10 (uses rcu_read_lock()), so the hunk was adapted manually] Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- net/core/bpf_sk_storage.c | 2 -- net/core/sock.c | 3 +++ 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/net/core/bpf_sk_storage.c b/net/core/bpf_sk_storage.c index 3fad2f5b920e..7923d1a52fe4 100644 --- a/net/core/bpf_sk_storage.c +++ b/net/core/bpf_sk_storage.c @@ -190,8 +190,6 @@ int bpf_sk_storage_clone(const struct sock *sk, struct sock *newsk) struct bpf_local_storage_elem *selem; int ret = 0; - RCU_INIT_POINTER(newsk->sk_bpf_storage, NULL); - rcu_read_lock(); sk_storage = rcu_dereference(sk->sk_bpf_storage); diff --git a/net/core/sock.c b/net/core/sock.c index 8904b0d87751..e1380c3ebf17 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1939,6 +1939,9 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) sock_copy(newsk, sk); newsk->sk_prot_creator = prot; +#ifdef CONFIG_BPF_SYSCALL + RCU_INIT_POINTER(newsk->sk_bpf_storage, NULL); +#endif /* SANITY */ if (likely(newsk->sk_net_refcnt)) { -- 2.43.0

2 1