GHSA-C5Q6-H25M-P95J
Vulnerability from github – Published: 2026-06-24 18:32 – Updated: 2026-06-28 09:31In the Linux kernel, the following vulnerability has been resolved:
bpf: Fix stale offload->prog pointer after constant blinding
When a dev-bound-only BPF program (BPF_F_XDP_DEV_BOUND_ONLY) undergoes JIT compilation with constant blinding enabled (bpf_jit_harden >= 2), bpf_jit_blind_constants() clones the program. The original prog is then freed in bpf_jit_prog_release_other(), which updates aux->prog to point to the surviving clone, but fails to update offload->prog.
This leaves offload->prog pointing to the freed original program. When the network namespace is subsequently destroyed, cleanup_net() triggers bpf_dev_bound_netdev_unregister(), which iterates ondev->progs and calls __bpf_prog_offload_destroy(offload->prog). Accessing the freed prog causes a page fault:
BUG: unable to handle page fault for address: ffffc900085f1038 Workqueue: netns cleanup_net RIP: 0010:__bpf_prog_offload_destroy+0xc/0x80 Call Trace: __bpf_offload_dev_netdev_unregister+0x257/0x350 bpf_dev_bound_netdev_unregister+0x4a/0x90 unregister_netdevice_many_notify+0x2a2/0x660 ... cleanup_net+0x21a/0x320
The test sequence that triggers this reliably is:
- Set net.core.bpf_jit_harden=2 (echo 2 > /proc/sys/net/core/bpf_jit_harden)
- Run xdp_metadata selftest, which creates a dev-bound-only XDP program on a veth inside a netns (./test_progs -t xdp_metadata)
- cleanup_net -> page fault in __bpf_prog_offload_destroy
Dev-bound-only programs are unique in that they have an offload structure but go through the normal JIT path instead of bpf_prog_offload_compile(). This means they are subject to constant blinding's prog clone-and-replace, while also having offload->prog that must stay in sync.
Fix this by updating offload->prog in bpf_jit_prog_release_other(), alongside the existing aux->prog update. Both are back-pointers to the prog that must be kept in sync when the prog is replaced.
{
"affected": [],
"aliases": [
"CVE-2026-53094"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-06-24T17:17:23Z",
"severity": "HIGH"
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nbpf: Fix stale offload-\u003eprog pointer after constant blinding\n\nWhen a dev-bound-only BPF program (BPF_F_XDP_DEV_BOUND_ONLY) undergoes\nJIT compilation with constant blinding enabled (bpf_jit_harden \u003e= 2),\nbpf_jit_blind_constants() clones the program. The original prog is then\nfreed in bpf_jit_prog_release_other(), which updates aux-\u003eprog to point\nto the surviving clone, but fails to update offload-\u003eprog.\n\nThis leaves offload-\u003eprog pointing to the freed original program. When\nthe network namespace is subsequently destroyed, cleanup_net() triggers\nbpf_dev_bound_netdev_unregister(), which iterates ondev-\u003eprogs and calls\n__bpf_prog_offload_destroy(offload-\u003eprog). Accessing the freed prog\ncauses a page fault:\n\nBUG: unable to handle page fault for address: ffffc900085f1038\nWorkqueue: netns cleanup_net\nRIP: 0010:__bpf_prog_offload_destroy+0xc/0x80\nCall Trace:\n__bpf_offload_dev_netdev_unregister+0x257/0x350\nbpf_dev_bound_netdev_unregister+0x4a/0x90\nunregister_netdevice_many_notify+0x2a2/0x660\n...\ncleanup_net+0x21a/0x320\n\nThe test sequence that triggers this reliably is:\n\n1. Set net.core.bpf_jit_harden=2 (echo 2 \u003e /proc/sys/net/core/bpf_jit_harden)\n2. Run xdp_metadata selftest, which creates a dev-bound-only XDP\n program on a veth inside a netns (./test_progs -t xdp_metadata)\n3. cleanup_net -\u003e page fault in __bpf_prog_offload_destroy\n\nDev-bound-only programs are unique in that they have an offload structure\nbut go through the normal JIT path instead of bpf_prog_offload_compile().\nThis means they are subject to constant blinding\u0027s prog clone-and-replace,\nwhile also having offload-\u003eprog that must stay in sync.\n\nFix this by updating offload-\u003eprog in bpf_jit_prog_release_other(),\nalongside the existing aux-\u003eprog update. Both are back-pointers to\nthe prog that must be kept in sync when the prog is replaced.",
"id": "GHSA-c5q6-h25m-p95j",
"modified": "2026-06-28T09:31:41Z",
"published": "2026-06-24T18:32:46Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-53094"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/059525cf18e69a9313baf947d8898c6ee7ca6b65"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/25484c39d1ec82a0368798d956da3de5039b3fe8"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/a1aa9ef47c299c5bbc30594d3c2f0589edf908e6"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/a713b72ff88cdab4d5d692908ab1259ada511f4d"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/c79f8503d83d4665be461fb9e45e215d0380c67b"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
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.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.