pull request (net): ipsec 2022-11-23
Subsystems (4)
Steffen Klassert
maintainercover pull request (net): ipsec 2022-11-23
1) Fix "disable_policy" on ipv4 early demuxP Packets after the initial packet in a flow might be incorectly dropped on early demux if there are no matching policies. From Eyal Birger.
2) Fix a kernel warning in case XFRM encap type is not available. From Eyal Birger.
3) Fix ESN wrap around for GSO to avoid a double usage of a sequence number. From Christian Langrock.
4) Fix a send_acquire race with pfkey_register.
From Herbert Xu.
5) Fix a list corruption panic in __xfrm_state_delete(). Thomas Jarosch.
6) Fix an unchecked return value in xfrm6_init().
Chen Zhongjin.
Please pull or let me know if there are problems.
Thanks!
The following changes since commit 1d22f78d05737ce21bff7b88b6e58873f35e65ba:
Merge tag 'ieee802154-for-net-2022-10-05' of git://git.kernel.org/pub/scm/linux/kernel/git/sschmidt/wpan (2022-10-05 20:38:46 -0700)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec.git master
for you to fetch changes up to 40781bfb836eda57d19c0baa37c7e72590e05fdc:
xfrm: Fix ignored return value in xfrm6_init() (2022-11-22 07:16:34 +0100)
----------------------------------------------------------------
Chen Zhongjin (1):
xfrm: Fix ignored return value in xfrm6_init()
Christian Langrock (1):
xfrm: replay: Fix ESN wrap around for GSO
Eyal Birger (2):
xfrm: fix "disable_policy" on ipv4 early demux
xfrm: lwtunnel: squelch kernel warning in case XFRM encap type is not available
Herbert Xu (1):
af_key: Fix send_acquire race with pfkey_register
Thomas Jarosch (1):
xfrm: Fix oops in __xfrm_state_delete()
net/core/lwtunnel.c | 4 +++- net/ipv4/esp4_offload.c | 3 +++ net/ipv4/ip_input.c | 5 +++++ net/ipv6/esp6_offload.c | 3 +++ net/ipv6/xfrm6_policy.c | 6 +++++- net/key/af_key.c | 34 +++++++++++++++++++++++----------- net/xfrm/xfrm_device.c | 15 ++++++++++++++- net/xfrm/xfrm_replay.c | 2 +- 8 files changed, 57 insertions(+), 15 deletions(-)
Steffen Klassert
maintainerPATCH 1/6 xfrm: fix "disable_policy" on ipv4 early demux
From: Eyal Birger <[email protected]>
The commit in the "Fixes" tag tried to avoid a case where policy check is ignored due to dst caching in next hops.
However, when the traffic is locally consumed, the dst may be cached in a local TCP or UDP socket as part of early demux. In this case the "disable_policy" flag is not checked as ip_route_input_noref() was only called before caching, and thus, packets after the initial packet in a flow will be dropped if not matching policies.
Fix by checking the "disable_policy" flag also when a valid dst is already available.
----
v2: use dev instead of skb->dev
--- net/ipv4/ip_input.c | 5 +++++ 1 file changed, 5 insertions(+)
net/ipv4/ip_input.c +5 -0
@@ -366,6 +366,11 @@ static int ip_rcv_finish_core(struct net *net, struct sock *sk, iph->tos, dev);if (unlikely (err)) goto drop_error; + }else { +struct in_device *in_dev =__in_dev_get_rcu (dev); + +if (in_dev &&IN_DEV_ORCONF (in_dev,NOPOLICY )) +IPCB (skb)->flags |=IPSKB_NOPOLICY ; }#ifdef CONFIG_IP_ROUTE_CLASSID
Signature · 1 line
2.25.1
Re: [PATCH 1/6] xfrm: fix "disable_policy" on ipv4 early demux
This series was applied to netdev/net.git (master) by Steffen Klassert <[email protected]>:
Hello:
This series was applied to netdev/net.git (master) by Steffen Klassert <[email protected]>:
Quoted 5 lines · On Wed, 23 Nov 2022 10:31:11 +0100 you wrote:
From: Eyal Birger <[email protected]>
The commit in the "Fixes" tag tried to avoid a case where policy check is ignored due to dst caching in next hops.
Quoted 7 lines
However, when the traffic is locally consumed, the dst may be cached in a local TCP or UDP socket as part of early demux. In this case the "
disable_policy" flag is not checked asip_route_input_noref()was only called before caching, and thus, packets after the initial packet in a flow will be dropped if not matching policies.[...]
Here is the summary with links:
- [1/6] xfrm: fix "disable_policy" on ipv4 early demux
https://git.kernel.org/netdev/net/c/3a5913183aa1 - [2/6] xfrm: lwtunnel: squelch kernel warning in case XFRM encap type is not available https://git.kernel.org/netdev/net/c/d83f7040e184 - [3/6] xfrm: replay: Fix ESN wrap around for GSO https://git.kernel.org/netdev/net/c/4b549ccce941 - [4/6] af_key: Fix send_acquire race with pfkey_register https://git.kernel.org/netdev/net/c/7f57f8165cb6 - [5/6] xfrm: Fix oops in __xfrm_state_delete() https://git.kernel.org/netdev/net/c/b97df039a68b - [6/6] xfrm: Fix ignored return value in xfrm6_init() https://git.kernel.org/netdev/net/c/40781bfb836e
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html
Steffen Klassert
maintainerPATCH 2/6 xfrm: lwtunnel: squelch kernel warning in case XFRM encap type is not available
In reply to Steffen KlassertFrom: Eyal Birger <[email protected]>
Ido reported that a kernel warning [1] can be triggered from user space when the kernel is compiled with CONFIG_MODULES=y and CONFIG_XFRM=n when adding an xfrm encap type route, e.g:
$ ip route add 198.51.100.0/24 dev dummy1 encap xfrm if_id 1 Error: lwt encapsulation type not supported.
The reason for the warning is that the LWT infrastructure has an autoloading feature which is meant only for encap types that don't use a net device, which is not the case in xfrm encap.
Mute this warning for xfrm encap as there's no encap module to autoload in this case.
[1]
WARNING: CPU: 3 PID: 2746262 at net/core/lwtunnel.c:57 lwtunnel_valid_encap_type+0x4f/0x120 [...] Call Trace: <TASK> rtm_to_fib_config+0x211/0x350 inet_rtm_newroute+0x3a/0xa0 rtnetlink_rcv_msg+0x154/0x3c0 netlink_rcv_skb+0x49/0xf0 netlink_unicast+0x22f/0x350 netlink_sendmsg+0x208/0x440 ____sys_sendmsg+0x21f/0x250 ___sys_sendmsg+0x83/0xd0 __sys_sendmsg+0x54/0xa0 do_syscall_64+0x35/0x80 entry_SYSCALL_64_after_hwframe+0x63/0xcd
--- net/core/lwtunnel.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
net/core/lwtunnel.c +3 -1
@@ -48,9 +48,11 @@ static const char *lwtunnel_encap_str(enum lwtunnel_encap_types encap_type)return "RPL" ;case LWTUNNEL_ENCAP_IOAM6 :return "IOAM6" ; +case LWTUNNEL_ENCAP_XFRM : +/* module autoload not supported for encap type */ +return NULL ;case LWTUNNEL_ENCAP_IP6 :case LWTUNNEL_ENCAP_IP : -case LWTUNNEL_ENCAP_XFRM :case LWTUNNEL_ENCAP_NONE :case __LWTUNNEL_ENCAP_MAX:/* should not have got here */
Signature · 1 line
2.25.1
Steffen Klassert
maintainerPATCH 3/6 xfrm: replay: Fix ESN wrap around for GSO
In reply to Steffen KlassertFrom: Christian Langrock <[email protected]>
When using GSO it can happen that the wrong seq_hi is used for the last packets before the wrap around. This can lead to double usage of a sequence number. To avoid this, we should serialize this last GSO packet.
--- net/ipv4/esp4_offload.c | 3 +++ net/ipv6/esp6_offload.c | 3 +++ net/xfrm/xfrm_device.c | 15 ++++++++++++++- net/xfrm/xfrm_replay.c | 2 +- 4 files changed, 21 insertions(+), 2 deletions(-)
net/ipv4/esp4_offload.c +3 -0
@@ -314,6 +314,9 @@ static int esp_xmit(struct xfrm_state *x, struct sk_buff *skb, netdev_features_ xo->seq.low +=skb_shinfo (skb)->gso_segs; } +if (xo->seq.low < seq) + xo->seq.hi++; + esp.seqno =cpu_to_be64 (seq + ((u64 )xo->seq.hi <<32 ));ip_hdr (skb)->tot_len =htons (skb->len);
net/ipv6/esp6_offload.c +3 -0
@@ -346,6 +346,9 @@ static int esp6_xmit(struct xfrm_state *x, struct sk_buff *skb, netdev_features xo->seq.low +=skb_shinfo (skb)->gso_segs; } +if (xo->seq.low < seq) + xo->seq.hi++; + esp.seqno =cpu_to_be64 (xo->seq.low + ((u64 )xo->seq.hi <<32 )); len = skb->len -sizeof (struct ipv6hdr );
net/xfrm/xfrm_device.c +14 -1
@@ -97,6 +97,18 @@ static void xfrm_outer_mode_prep(struct xfrm_state *x, struct sk_buff *skb) } } +static inline bool xmit_xfrm_check_overflow (struct sk_buff *skb) +{ +struct xfrm_offload *xo =xfrm_offload (skb); +__u32 seq = xo->seq.low; + + seq +=skb_shinfo (skb)->gso_segs; +if (unlikely (seq < xo->seq.low)) +return true; + +return false; +} +struct sk_buff *validate_xmit_xfrm (struct sk_buff *skb,netdev_features_t features,bool *again) {int err;
@@ -134,7 +146,8 @@ struct sk_buff *validate_xmit_xfrm(struct sk_buff *skb, netdev_features_t featurreturn skb; } -if (skb_is_gso (skb) &&unlikely (x->xso.dev != dev)) { +if (skb_is_gso (skb) && (unlikely (x->xso.dev != dev) || +unlikely (xmit_xfrm_check_overflow (skb)))) {struct sk_buff *segs;/* Packet got rerouted, fixup features and segment it. */
net/xfrm/xfrm_replay.c +1 -1
@@ -714,7 +714,7 @@ static int xfrm_replay_overflow_offload_esn(struct xfrm_state *x, struct sk_buff oseq +=skb_shinfo (skb)->gso_segs; } -if (unlikely (oseq < replay_esn->oseq)) { +if (unlikely (xo->seq.low < replay_esn->oseq)) {XFRM_SKB_CB (skb)->seq.output.hi = ++oseq_hi; xo->seq.hi = oseq_hi; replay_esn->oseq_hi = oseq_hi;
Signature · 1 line
2.25.1
Steffen Klassert
maintainerPATCH 4/6 af_key: Fix send_acquire race with pfkey_register
In reply to Steffen KlassertFrom: Herbert Xu <[email protected]>
The function pfkey_send_acquire may race with pfkey_register (which could even be in a different name space). This may result in a buffer overrun.
Allocating the maximum amount of memory that could be used prevents this.
--- net/key/af_key.c | 32 ++++++++++++++++++++++---------- 1 file changed, 22 insertions(+), 10 deletions(-)
net/key/af_key.c +22 -10
@@ -2905,7 +2905,7 @@ static int count_ah_combs(const struct xfrm_tmpl *t)break ;if (!aalg->pfkey_supported)continue ; -if (aalg_tmpl_set (t, aalg) && aalg->available) +if (aalg_tmpl_set (t, aalg)) sz +=sizeof (struct sadb_comb ); }return sz +sizeof (struct sadb_prop );
@@ -2923,7 +2923,7 @@ static int count_esp_combs(const struct xfrm_tmpl *t)if (!ealg->pfkey_supported)continue ; -if (!(ealg_tmpl_set (t, ealg) && ealg->available)) +if (!(ealg_tmpl_set (t, ealg)))continue ;for (k =1 ; ; k++) {
@@ -2934,16 +2934,17 @@ static int count_esp_combs(const struct xfrm_tmpl *t)if (!aalg->pfkey_supported)continue ; -if (aalg_tmpl_set (t, aalg) && aalg->available) +if (aalg_tmpl_set (t, aalg)) sz +=sizeof (struct sadb_comb ); } }return sz +sizeof (struct sadb_prop ); } -static void dump_ah_combs (struct sk_buff *skb,const struct xfrm_tmpl *t) +static int dump_ah_combs (struct sk_buff *skb,const struct xfrm_tmpl *t) {struct sadb_prop *p; +int sz =0 ;int i; p =skb_put (skb,sizeof (struct sadb_prop ));
@@ -2971,13 +2972,17 @@ static void dump_ah_combs(struct sk_buff *skb, const struct xfrm_tmpl *t) c->sadb_comb_soft_addtime =20 *60 *60 ; c->sadb_comb_hard_usetime =8 *60 *60 ; c->sadb_comb_soft_usetime =7 *60 *60 ; + sz +=sizeof (*c); } } + +return sz +sizeof (*p); } -static void dump_esp_combs (struct sk_buff *skb,const struct xfrm_tmpl *t) +static int dump_esp_combs (struct sk_buff *skb,const struct xfrm_tmpl *t) {struct sadb_prop *p; +int sz =0 ;int i, k; p =skb_put (skb,sizeof (struct sadb_prop ));
@@ -3019,8 +3024,11 @@ static void dump_esp_combs(struct sk_buff *skb, const struct xfrm_tmpl *t) c->sadb_comb_soft_addtime =20 *60 *60 ; c->sadb_comb_hard_usetime =8 *60 *60 ; c->sadb_comb_soft_usetime =7 *60 *60 ; + sz +=sizeof (*c); } } + +return sz +sizeof (*p); }static int key_notify_policy_expire (struct xfrm_policy *xp,const struct km_event *c)
@@ -3150,6 +3158,7 @@ static int pfkey_send_acquire(struct xfrm_state *x, struct xfrm_tmpl *t, structstruct sadb_x_sec_ctx *sec_ctx;struct xfrm_sec_ctx *xfrm_ctx;int ctx_size =0 ; +int alg_size =0 ; sockaddr_size =pfkey_sockaddr_size (x->props.family);if (!sockaddr_size)
@@ -3161,16 +3170,16 @@ static int pfkey_send_acquire(struct xfrm_state *x, struct xfrm_tmpl *t, structsizeof (struct sadb_x_policy );if (x->id.proto ==IPPROTO_AH ) - size +=count_ah_combs (t); + alg_size =count_ah_combs (t);else if (x->id.proto ==IPPROTO_ESP ) - size +=count_esp_combs (t); + alg_size =count_esp_combs (t);if ((xfrm_ctx = x->security)) { ctx_size =PFKEY_ALIGN8 (xfrm_ctx->ctx_len); size +=sizeof (struct sadb_x_sec_ctx ) + ctx_size; } - skb =alloc_skb (size +16 ,GFP_ATOMIC ); + skb =alloc_skb (size + alg_size +16 ,GFP_ATOMIC );if (skb ==NULL )return -ENOMEM ;
@@ -3224,10 +3233,13 @@ static int pfkey_send_acquire(struct xfrm_state *x, struct xfrm_tmpl *t, struct pol->sadb_x_policy_priority = xp->priority;/* Set sadb_comb's. */ + alg_size =0 ;if (x->id.proto ==IPPROTO_AH ) -dump_ah_combs (skb, t); + alg_size =dump_ah_combs (skb, t);else if (x->id.proto ==IPPROTO_ESP ) -dump_esp_combs (skb, t); + alg_size =dump_esp_combs (skb, t); + + hdr->sadb_msg_len += alg_size /8 ;/* security context */ if (xfrm_ctx) {
Signature · 1 line
2.25.1
Steffen Klassert
maintainerPATCH 5/6 xfrm: Fix oops in __xfrm_state_delete()
In reply to Steffen KlassertFrom: Thomas Jarosch <[email protected]>
Kernel 5.14 added a new "byseq" index to speed up xfrm_state lookups by sequence number in commit fe9f1d8779cb ("xfrm: add state hashtable keyed by seq")
While the patch was thorough, the function pfkey_send_new_mapping() in net/af_key.c also modifies x->km.seq and never added the current xfrm_state to the "byseq" index.
This leads to the following kernel Ooops:
BUG: kernel NULL pointer dereference, address: 0000000000000000
..
RIP: 0010:__xfrm_state_delete+0xc9/0x1c0
..
Call Trace:
<TASK>
xfrm_state_delete+0x1e/0x40
xfrm_del_sa+0xb0/0x110 [xfrm_user]
xfrm_user_rcv_msg+0x12d/0x270 [xfrm_user]
? remove_entity_load_avg+0x8a/0xa0
? copy_to_user_state_extra+0x580/0x580 [xfrm_user]
netlink_rcv_skb+0x51/0x100
xfrm_netlink_rcv+0x30/0x50 [xfrm_user]
netlink_unicast+0x1a6/0x270
netlink_sendmsg+0x22a/0x480
__sys_sendto+0x1a6/0x1c0
? __audit_syscall_entry+0xd8/0x130
? __audit_syscall_exit+0x249/0x2b0
__x64_sys_sendto+0x23/0x30
do_syscall_64+0x3a/0x90
entry_SYSCALL_64_after_hwframe+0x61/0xcbExact location of the crash in __xfrm_state_delete():
if (x->km.seq)
hlist_del_rcu(&x->byseq);The hlist_node "byseq" was never populated.
The bug only triggers if a new NAT traversal mapping (changed IP or port) is detected in esp_input_done2() / esp6_input_done2(), which in turn indirectly calls pfkey_send_new_mapping() *if* the kernel is compiled with CONFIG_NET_KEY and "af_key" is active.
The PF_KEYv2 message SADB_X_NAT_T_NEW_MAPPING is not part of RFC 2367. Various implementations have been examined how they handle the "sadb_msg_seq" header field:
- racoon (Android): does not process SADB_X_NAT_T_NEW_MAPPING
- strongswan: does not care about sadb_msg_seq
- openswan: does not care about sadb_msg_seq
There is no standard how PF_KEYv2 sadb_msg_seq should be populated for SADB_X_NAT_T_NEW_MAPPING and it's not used in popular implementations either. Herbert Xu suggested we should just use the current km.seq value as is. This fixes the root cause of the oops since we no longer modify km.seq itself.
The update of "km.seq" looks like a copy'n'paste error
from pfkey_send_acquire(). SADB_ACQUIRE must indeed assign a unique km.seq number according to RFC 2367. It has been verified that code paths involving pfkey_send_acquire() don't cause the same Oops.
PF_KEYv2 SADB_X_NAT_T_NEW_MAPPING support was originally added here:
https://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git
commit cbc3488685b20e7b2a98ad387a1a816aada569d8
Author: Derek Atkins <[email protected]>
AuthorDate: Wed Apr 2 13:21:02 2003 -0800
[IPSEC]: Implement UDP Encapsulation framework.
In particular, implement ESPinUDP encapsulation for IPsec Nat Traversal.
A note on triggering the bug: I was not able to trigger it using VMs. There is one VPN using a high latency link on our production VPN server that triggered it like once a day though.
--- net/key/af_key.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/key/af_key.c +1 -1
@@ -3394,7 +3394,7 @@ static int pfkey_send_new_mapping(struct xfrm_state *x, xfrm_address_t *ipaddr, hdr->sadb_msg_len = size /sizeof (uint64_t ); hdr->sadb_msg_errno =0 ; hdr->sadb_msg_reserved =0 ; - hdr->sadb_msg_seq = x->km.seq =get_acqseq (); + hdr->sadb_msg_seq = x->km.seq; hdr->sadb_msg_pid =0 ;/* SA */
Signature · 1 line
2.25.1
Steffen Klassert
maintainerPATCH 6/6 xfrm: Fix ignored return value in xfrm6_init()
In reply to Steffen KlassertFrom: Chen Zhongjin <[email protected]>
When IPv6 module initializing in xfrm6_init(), register_pernet_subsys() is possible to fail but its return value is ignored.
If IPv6 initialization fails later and xfrm6_fini() is called, removing uninitialized list in xfrm6_net_ops will cause null-ptr-deref:
KASAN: null-ptr-deref in range [0x0000000000000008-0x000000000000000f] CPU: 1 PID: 330 Comm: insmod RIP: 0010:unregister_pernet_operations+0xc9/0x450 Call Trace: <TASK> unregister_pernet_subsys+0x31/0x3e xfrm6_fini+0x16/0x30 [ipv6] ip6_route_init+0xcd/0x128 [ipv6] inet6_init+0x29c/0x602 [ipv6] ...
Fix it by catching the error return value of register_pernet_subsys().
--- net/ipv6/xfrm6_policy.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
net/ipv6/xfrm6_policy.c +5 -1
@@ -287,9 +287,13 @@ int __init xfrm6_init(void)if (ret) goto out_state; -register_pernet_subsys (&xfrm6_net_ops); + ret =register_pernet_subsys (&xfrm6_net_ops); +if (ret) + goto out_protocol; out:return ret; +out_protocol: +xfrm6_protocol_fini (); out_state:xfrm6_state_fini (); out_policy:
Signature · 1 line
2.25.1
From: Steffen Klassert <[email protected]>
Date: Wed, 23 Nov 2022 10:31:10 +0100
Quoted 53 lines
1) Fix "
disable_policy" on ipv4 early demuxP Packets after the initial packet in a flow might be incorectly dropped on early demux if there are no matching policies. From Eyal Birger.2) Fix a kernel warning in case XFRM encap type is not available. From Eyal Birger.
3) Fix ESN wrap around for GSO to avoid a double usage of a sequence number. From Christian Langrock.
4) Fix a
send_acquirerace withpfkey_register.
From Herbert Xu.5) Fix a list corruption panic in
__xfrm_state_delete(). Thomas Jarosch.6) Fix an unchecked return value in
xfrm6_init().
Chen Zhongjin.Please pull or let me know if there are problems.
Thanks!
The following changes since commit
1d22f78d05737ce21bff7b88b6e58873f35e65ba:Merge tag 'ieee802154-for-net-2022-10-05' of
git://git.kernel.org/pub/scm/linux/kernel/git/sschmidt/wpan(2022-10-05 20:38:46 -0700)are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec.gitmasterfor you to fetch changes up to
40781bfb836eda57d19c0baa37c7e72590e05fdc:xfrm: Fix ignored return value in
xfrm6_init()(2022-11-22 07:16:34 +0100)----------------------------------------------------------------
Chen Zhongjin (1):
xfrm: Fix ignored return value inxfrm6_init()Christian Langrock (1):
xfrm: replay: Fix ESN wrap around for GSOEyal Birger (2):
xfrm: fix "disable_policy" on ipv4 early demux
xfrm: lwtunnel: squelch kernel warning in case XFRM encap type is not availableHerbert Xu (1):
af_key: Fixsend_acquirerace withpfkey_registerThomas Jarosch (1):
xfrm: Fix oops in__xfrm_state_delete()
Quoted 9 lines
net/core/lwtunnel.c | 4 +++- net/ipv4/esp4_offload.c | 3 +++ net/ipv4/ip_input.c | 5 +++++ net/ipv6/esp6_offload.c | 3 +++ net/ipv6/xfrm6_policy.c | 6 +++++- net/key/af_key.c | 34 +++++++++++++++++++++++----------- net/xfrm/xfrm_device.c | 15 ++++++++++++++- net/xfrm/xfrm_replay.c | 2 +- 8 files changed, 57 insertions(+), 15 deletions(-)
(for the whole PR)
Thanks,
Olek