GHSA-2FM8-9QC5-4MQX
Vulnerability from github – Published: 2026-05-06 12:30 – Updated: 2026-05-06 12:30In the Linux kernel, the following vulnerability has been resolved:
xfrm: always flush state and policy upon NETDEV_UNREGISTER event
syzbot is reporting that "struct xfrm_state" refcount is leaking.
unregister_netdevice: waiting for netdevsim0 to become free. Usage count = 2 ref_tracker: netdev@ffff888052f24618 has 1/1 users at __netdev_tracker_alloc include/linux/netdevice.h:4400 [inline] netdev_tracker_alloc include/linux/netdevice.h:4412 [inline] xfrm_dev_state_add+0x3a5/0x1080 net/xfrm/xfrm_device.c:316 xfrm_state_construct net/xfrm/xfrm_user.c:986 [inline] xfrm_add_sa+0x34ff/0x5fa0 net/xfrm/xfrm_user.c:1022 xfrm_user_rcv_msg+0x58e/0xc00 net/xfrm/xfrm_user.c:3507 netlink_rcv_skb+0x158/0x420 net/netlink/af_netlink.c:2550 xfrm_netlink_rcv+0x71/0x90 net/xfrm/xfrm_user.c:3529 netlink_unicast_kernel net/netlink/af_netlink.c:1318 [inline] netlink_unicast+0x5aa/0x870 net/netlink/af_netlink.c:1344 netlink_sendmsg+0x8c8/0xdd0 net/netlink/af_netlink.c:1894 sock_sendmsg_nosec net/socket.c:727 [inline] __sock_sendmsg net/socket.c:742 [inline] _syssendmsg+0xa5d/0xc30 net/socket.c:2592 _sys_sendmsg+0x134/0x1d0 net/socket.c:2646 __sys_sendmsg+0x16d/0x220 net/socket.c:2678 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline] do_syscall_64+0xcd/0xf80 arch/x86/entry/syscall_64.c:94 entry_SYSCALL_64_after_hwframe+0x77/0x7f
This is because commit d77e38e612a0 ("xfrm: Add an IPsec hardware offloading API") implemented xfrm_dev_unregister() as no-op despite xfrm_dev_state_add() from xfrm_state_construct() acquires a reference to "struct net_device". I guess that that commit expected that NETDEV_DOWN event is fired before NETDEV_UNREGISTER event fires, and also assumed that xfrm_dev_state_add() is called only if (dev->features & NETIF_F_HW_ESP) != 0.
Sabrina Dubroca identified steps to reproduce the same symptoms as below.
echo 0 > /sys/bus/netdevsim/new_device dev=$(ls -1 /sys/bus/netdevsim/devices/netdevsim0/net/) ip xfrm state add src 192.168.13.1 dst 192.168.13.2 proto esp \ spi 0x1000 mode tunnel aead 'rfc4106(gcm(aes))' $key 128 \ offload crypto dev $dev dir out ethtool -K $dev esp-hw-offload off echo 0 > /sys/bus/netdevsim/del_device
Like these steps indicate, the NETIF_F_HW_ESP bit can be cleared after xfrm_dev_state_add() acquired a reference to "struct net_device". Also, xfrm_dev_state_add() does not check for the NETIF_F_HW_ESP bit when acquiring a reference to "struct net_device".
Commit 03891f820c21 ("xfrm: handle NETDEV_UNREGISTER for xfrm device") re-introduced the NETDEV_UNREGISTER event to xfrm_dev_event(), but that commit for unknown reason chose to share xfrm_dev_down() between the NETDEV_DOWN event and the NETDEV_UNREGISTER event. I guess that that commit missed the behavior in the previous paragraph.
Therefore, we need to re-introduce xfrm_dev_unregister() in order to release the reference to "struct net_device" by unconditionally flushing state and policy.
{
"affected": [],
"aliases": [
"CVE-2026-43167"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-05-06T12:16:34Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nxfrm: always flush state and policy upon NETDEV_UNREGISTER event\n\nsyzbot is reporting that \"struct xfrm_state\" refcount is leaking.\n\n unregister_netdevice: waiting for netdevsim0 to become free. Usage count = 2\n ref_tracker: netdev@ffff888052f24618 has 1/1 users at\n __netdev_tracker_alloc include/linux/netdevice.h:4400 [inline]\n netdev_tracker_alloc include/linux/netdevice.h:4412 [inline]\n xfrm_dev_state_add+0x3a5/0x1080 net/xfrm/xfrm_device.c:316\n xfrm_state_construct net/xfrm/xfrm_user.c:986 [inline]\n xfrm_add_sa+0x34ff/0x5fa0 net/xfrm/xfrm_user.c:1022\n xfrm_user_rcv_msg+0x58e/0xc00 net/xfrm/xfrm_user.c:3507\n netlink_rcv_skb+0x158/0x420 net/netlink/af_netlink.c:2550\n xfrm_netlink_rcv+0x71/0x90 net/xfrm/xfrm_user.c:3529\n netlink_unicast_kernel net/netlink/af_netlink.c:1318 [inline]\n netlink_unicast+0x5aa/0x870 net/netlink/af_netlink.c:1344\n netlink_sendmsg+0x8c8/0xdd0 net/netlink/af_netlink.c:1894\n sock_sendmsg_nosec net/socket.c:727 [inline]\n __sock_sendmsg net/socket.c:742 [inline]\n ____sys_sendmsg+0xa5d/0xc30 net/socket.c:2592\n ___sys_sendmsg+0x134/0x1d0 net/socket.c:2646\n __sys_sendmsg+0x16d/0x220 net/socket.c:2678\n do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]\n do_syscall_64+0xcd/0xf80 arch/x86/entry/syscall_64.c:94\n entry_SYSCALL_64_after_hwframe+0x77/0x7f\n\nThis is because commit d77e38e612a0 (\"xfrm: Add an IPsec hardware\noffloading API\") implemented xfrm_dev_unregister() as no-op despite\nxfrm_dev_state_add() from xfrm_state_construct() acquires a reference\nto \"struct net_device\".\nI guess that that commit expected that NETDEV_DOWN event is fired before\nNETDEV_UNREGISTER event fires, and also assumed that xfrm_dev_state_add()\nis called only if (dev-\u003efeatures \u0026 NETIF_F_HW_ESP) != 0.\n\nSabrina Dubroca identified steps to reproduce the same symptoms as below.\n\n echo 0 \u003e /sys/bus/netdevsim/new_device\n dev=$(ls -1 /sys/bus/netdevsim/devices/netdevsim0/net/)\n ip xfrm state add src 192.168.13.1 dst 192.168.13.2 proto esp \\\n spi 0x1000 mode tunnel aead \u0027rfc4106(gcm(aes))\u0027 $key 128 \\\n offload crypto dev $dev dir out\n ethtool -K $dev esp-hw-offload off\n echo 0 \u003e /sys/bus/netdevsim/del_device\n\nLike these steps indicate, the NETIF_F_HW_ESP bit can be cleared after\nxfrm_dev_state_add() acquired a reference to \"struct net_device\".\nAlso, xfrm_dev_state_add() does not check for the NETIF_F_HW_ESP bit\nwhen acquiring a reference to \"struct net_device\".\n\nCommit 03891f820c21 (\"xfrm: handle NETDEV_UNREGISTER for xfrm device\")\nre-introduced the NETDEV_UNREGISTER event to xfrm_dev_event(), but that\ncommit for unknown reason chose to share xfrm_dev_down() between the\nNETDEV_DOWN event and the NETDEV_UNREGISTER event.\nI guess that that commit missed the behavior in the previous paragraph.\n\nTherefore, we need to re-introduce xfrm_dev_unregister() in order to\nrelease the reference to \"struct net_device\" by unconditionally flushing\nstate and policy.",
"id": "GHSA-2fm8-9qc5-4mqx",
"modified": "2026-05-06T12:30:31Z",
"published": "2026-05-06T12:30:31Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-43167"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/166801e49a5b5fc127b8c9e2f110f303cfddfbc3"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/4efa91a28576054aae0e6dad9cba8fed8293aef8"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/59581778792cbaf8ad788f4a21dc663ce986050e"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/8c75c455ecd3bfd2f36abf66edb7021c4fa19ec4"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/a3c8fede034fa27892f87c863cbd5493167d17ed"
}
],
"schema_version": "1.4.0",
"severity": []
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.