xfrm: move policy_bydst RCU sync from per-netns .exit to .pre_exit
In the Linux kernel, the following vulnerability has been resolved: xfrm: move policy_bydst RCU sync from per-netns .exit to .pre_exit The struct pernet_operations docstring in include/net/net_namespace.h explicitly warns against blocking RCU primitives in .exit handlers: Exit methods using blocking RCU primitives, such as synchronize_rcu(), should be implemented via exit_batch. [...] Please, avoid synchronize_rcu() at all, where it's possible. Note that a combination of pre_exit() and exit() can be used, since a synchronize_rcu() is guaranteed between the calls. xfrm_policy_fini() violates this: it calls synchronize_rcu() before freeing the policy_bydst hash tables (so no RCU reader is mid- traversal at free time), but runs from xfrm_net_ops.exit -- once per namespace -- so a cleanup_net() of N namespaces pays N full RCU grace periods serially. Use the documented pre_exit/exit split. Move the policy flush (and the workqueue drains it depends on) into a new .pre_exit handler; xfrm_policy_fini() then runs in .exit and frees the hash tables after the synchronize_rcu_expedited() that cleanup_net() guarantees between the two phases. Providing O(1) RCU grace periods per batch instead of O(N). Observed on Linux 6.18 with a workload doing unshare(CLONE_NEWNET) at ~13/sec sustained: cleanup_net() and the netns_wq rescuer kthread both stuck in xfrm_policy_fini()'s synchronize_rcu(), >300k struct net accumulated in the cleanup queue, Percpu in /proc/meminfo climbed to 130+ GB on 256-CP
- Products
- Kernel
- Affected versions
- Not specified
- Fixed versions
- 6.12.93, 6.18.35, 7.0.12
- CVEs
- CVE-2026-64525
- Confirmed fixed CVEs
- Not specified
- Updated in this feed
- 28 Oct 2026, 08:57 UTC · Revision 1
Source evidence
xfrm: move policy_bydst RCU sync from per-netns .exit to .pre_exit
core-osv · Original sourcePublished 25 Jul 2026, 09:20 UTC
In the Linux kernel, the following vulnerability has been resolved: xfrm: move policy_bydst RCU sync from per-netns .exit to .pre_exit The struct pernet_operations docstring in include/net/net_namespace.h explicitly warns against blocking RCU primitives in .exit handlers: Exit methods using blocking RCU primitives, such as synchronize_rcu(), should be implemented via exit_batch. [...] Please, avoid synchronize_rcu() at all, where it's possible. Note that a combination of pre_exit() and exit() can be used, since a synchronize_rcu() is guaranteed between the calls. xfrm_policy_fini() violates this: it calls synchronize_rcu() before freeing the policy_bydst hash tables (so no RCU reader is mid- traversal at free time), but runs from xfrm_net_ops.exit -- once per namespace -- so a cleanup_net() of N namespaces pays N full RCU grace periods serially. Use the documented pre_exit/exit split. Move the policy flush (and the workqueue drains it depends on) into a new .pre_exit handler; xfrm_policy_fini() then runs in .exit and frees the hash tables after the synchronize_rcu_expedited() that cleanup_net() guarantees between the two phases. Providing O(1) RCU grace periods per batch instead of O(N). Observed on Linux 6.18 with a workload doing unshare(CLONE_NEWNET) at ~13/sec sustained: cleanup_net() and the netns_wq rescuer kthread both stuck in xfrm_policy_fini()'s synchronize_rcu(), >300k struct net accumulated in the cleanup queue, Percpu in /proc/meminfo climbed to 130+ GB on 256-CP
osv:CVE-2026-64525
Kernel
Fixed versions: 6.12.93, 6.18.35, 7.0.12
Complete source record
{
"id": "7021b8c5a90a5f634ba3542c8c5a85cd6ea6003c315af93c71f17c0933d6b293",
"url": "https://osv.dev/vulnerability/CVE-2026-64525",
"method": "structured",
"payload": {
"kind": "advisory",
"title": "xfrm: move policy_bydst RCU sync from per-netns .exit to .pre_exit",
"summary": "In the Linux kernel, the following vulnerability has been resolved: xfrm: move policy_bydst RCU sync from per-netns .exit to .pre_exit The struct pernet_operations docstring in include/net/net_namespace.h explicitly warns against blocking RCU primitives in .exit handlers: Exit methods using blocking RCU primitives, such as synchronize_rcu(), should be implemented via exit_batch. [...] Please, avoid synchronize_rcu() at all, where it's possible. Note that a combination of pre_exit() and exit() can be used, since a synchronize_rcu() is guaranteed between the calls. xfrm_policy_fini() violates this: it calls synchronize_rcu() before freeing the policy_bydst hash tables (so no RCU reader is mid- traversal at free time), but runs from xfrm_net_ops.exit -- once per namespace -- so a cleanup_net() of N namespaces pays N full RCU grace periods serially. Use the documented pre_exit/exit split. Move the policy flush (and the workqueue drains it depends on) into a new .pre_exit handler; xfrm_policy_fini() then runs in .exit and frees the hash tables after the synchronize_rcu_expedited() that cleanup_net() guarantees between the two phases. Providing O(1) RCU grace periods per batch instead of O(N). Observed on Linux 6.18 with a workload doing unshare(CLONE_NEWNET) at ~13/sec sustained: cleanup_net() and the netns_wq rescuer kthread both stuck in xfrm_policy_fini()'s synchronize_rcu(), >300k struct net accumulated in the cleanup queue, Percpu in /proc/meminfo climbed to 130+ GB on 256-CP",
"evidence": "osv:CVE-2026-64525",
"products": [
{
"name": {
"value": "Kernel",
"evidence": "/affected/1/package/name"
},
"prerequisites": [],
"fixed_versions": [
{
"value": "6.12.93",
"evidence": "/affected/1/ranges"
},
{
"value": "6.18.35",
"evidence": "/affected/1/ranges"
},
{
"value": "7.0.12",
"evidence": "/affected/1/ranges"
}
],
"affected_versions": []
}
],
"severity": "unknown",
"structured": {
"id": "CVE-2026-64525",
"aliases": [
"CVE-2026-64525"
],
"affected": [
{
"ranges": [
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "438b1f668ad58f46ce699bb48e4698a7839e3f9e"
},
{
"fixed": "bca6386dc08750fc7cdcbc7683473748ba3114b9"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "3733fce2871c9bca9dd18a1a23b1432ea215a094"
},
{
"fixed": "91cc13978ab0bc6f669139f53e7e613a860d10e0"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "069daad4f2ae9c5c108131995529d5f02392c446"
},
{
"fixed": "d14ae8ef88c2c6590e107db61b6adce148cec7b3"
},
{
"fixed": "3e52417318473782012b236d0325bf7d2266a597"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "6.12.83"
},
{
"fixed": "6.12.93"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "6.18.24"
},
{
"fixed": "6.18.35"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "6.6.136"
},
{
"fixed": "6.7"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "6.19.14"
},
{
"fixed": "6.20"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "b66920a3348c0f63ba18365248fa21fbf0b3a937"
}
]
},
{
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"type": "GIT",
"events": [
{
"introduced": "33a3149dd81a1e2f52b80ee1e0fc380b39f3d028"
}
]
}
],
"versions": [
"v6.12.92",
"v6.12.91",
"v6.12.90",
"v6.12.89",
"v6.12.88",
"v6.12.87",
"v6.12.86",
"v6.12.85",
"v6.12.84",
"v6.12.83",
"v6.18.34",
"v6.18.33",
"v6.18.32",
"v6.18.31",
"v6.18.30",
"v6.18.29",
"v6.18.28",
"v6.18.27",
"v6.18.26",
"v6.18.25",
"v6.18.24",
"v6.6.158",
"v6.6.157",
"v6.6.156",
"v6.6.155",
"v6.6.154",
"v6.6.153",
"v6.6.152",
"v6.6.151",
"v6.6.150",
"v6.6.149",
"v6.6.148",
"v6.6.147",
"v6.6.146",
"v6.6.145",
"v6.6.144",
"v6.6.143",
"v6.6.142",
"v6.6.141",
"v6.6.140",
"v6.6.139",
"v6.6.138",
"v6.6.137",
"v6.6.136",
"v6.19.14"
],
"database_specific": {
"source": "https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-64525.json"
}
},
{
"ranges": [
{
"type": "ECOSYSTEM",
"events": [
{
"introduced": "0"
},
{
"fixed": "6.12.93"
}
]
},
{
"type": "ECOSYSTEM",
"events": [
{
"introduced": "6.13.0"
},
{
"fixed": "6.18.35"
}
]
},
{
"type": "ECOSYSTEM",
"events": [
{
"introduced": "6.19.0"
},
{
"fixed": "7.0.12"
}
]
}
],
"package": {
"name": "Kernel",
"ecosystem": "Linux"
},
"database_specific": {
"source": "https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2026-64525.json"
}
}
],
"modified": "2026-10-08T02:51:31.402477027Z",
"provider": "osv",
"references": [
"https://git.kernel.org/stable/c/3e52417318473782012b236d0325bf7d2266a597",
"https://git.kernel.org/stable/c/91cc13978ab0bc6f669139f53e7e613a860d10e0",
"https://git.kernel.org/stable/c/bca6386dc08750fc7cdcbc7683473748ba3114b9",
"https://git.kernel.org/stable/c/d14ae8ef88c2c6590e107db61b6adce148cec7b3",
"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/64xxx/CVE-2026-64525.json",
"https://nvd.nist.gov/vuln/detail/CVE-2026-64525",
"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git"
],
"provider_metadata": {
"cna_assigner": "Linux",
"osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/64xxx/CVE-2026-64525.json"
}
},
"related_cves": [
{
"value": "CVE-2026-64525",
"evidence": "/id"
}
],
"security_relevant": true,
"severity_evidence": null
},
"source_id": "core-osv",
"document_id": 98319,
"published_at": "2026-07-25T09:20:49.256+00:00"
}