GHSA-3RQ9-MJHW-Q3G6
Vulnerability from github – Published: 2026-10-06 09:31 – Updated: 2026-10-06 09:31In the Linux kernel, the following vulnerability has been resolved:
af_unix: Unify scc_index when finalising SCC in __unix_walk_scc().
Commit bfdb01283ee8 ("af_unix: Assign a unique index to SCC.") changed Tarjan's algorithm to update lowlink with lowlink, which is called lowpoint (unix_vertex.scc_index).
unix_vertex_dead() assumes all vertices in an SCC share the same lowpoint, but this is not always true if an SCC has two or more back edges, depending on the order of DFS.
For example, the graph below has two back edges from B to A and from C to B.
A --> B --> C
^ | ^ |
----'----'
If DFS walks through A -> B -> C -> B (-> C -> B) -> A (-> B -> A), each index and scc_index will be updated as follows.
A --> B --> C C = (3, 3) (index, scc_index) B = (2, 2) A = (1, 1)
A ... B ... C C = (3, 2)<-. ^ | B = (2, 2) -' `----' A = (1, 1)
A ... B ... C C = (3, 2) ^ | . . B = (2, 1)<-. `----' .... A = (1, 1) -'
Then, unix_vertex_dead() thinks that B is passed to another SCC with scc_index 2, and the SCC is not garbage-collected.
This does not happen if DFS walks in a different order below or starts from B.
1 3
A --> B --> C
^ | ^ |
----'----'
2 4
Let's unify scc_index across the SCC when finalising it.
Note that updating v->index was previously done in unix_scc_dead(), when called from __unix_walk_scc(), just to save one loop. Since __unix_walk_scc() now iterates over the SCC anyway, the update is moved back to __unix_walk_scc() and 'fast' argument is dropped.
{
"affected": [],
"aliases": [
"CVE-2026-98289"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-10-06T09:18:19Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\naf_unix: Unify scc_index when finalising SCC in __unix_walk_scc().\n\nCommit bfdb01283ee8 (\"af_unix: Assign a unique index to SCC.\")\nchanged Tarjan\u0027s algorithm to update lowlink with lowlink,\nwhich is called lowpoint (unix_vertex.scc_index).\n\nunix_vertex_dead() assumes all vertices in an SCC share the same\nlowpoint, but this is not always true if an SCC has two or more\nback edges, depending on the order of DFS.\n\nFor example, the graph below has two back edges from B to A\nand from C to B.\n\n A --\u003e B --\u003e C\n ^ | ^ |\n `----\u0027 `----\u0027\n\nIf DFS walks through A -\u003e B -\u003e C -\u003e B (-\u003e C -\u003e B) -\u003e A (-\u003e B -\u003e A),\neach index and scc_index will be updated as follows.\n\n A --\u003e B --\u003e C C = (3, 3) (index, scc_index)\n B = (2, 2)\n A = (1, 1)\n\n A ... B ... C C = (3, 2)\u003c-.\n ^ | B = (2, 2) -\u0027\n `----\u0027 A = (1, 1)\n\n A ... B ... C C = (3, 2)\n ^ | . . B = (2, 1)\u003c-.\n `----\u0027 .... A = (1, 1) -\u0027\n\nThen, unix_vertex_dead() thinks that B is passed to another\nSCC with scc_index 2, and the SCC is not garbage-collected.\n\nThis does not happen if DFS walks in a different order below\nor starts from B.\n\n 1 3\n A --\u003e B --\u003e C\n ^ | ^ |\n `----\u0027 `----\u0027\n 2 4\n\nLet\u0027s unify scc_index across the SCC when finalising it.\n\nNote that updating v-\u003eindex was previously done in unix_scc_dead(),\nwhen called from __unix_walk_scc(), just to save one loop. Since\n__unix_walk_scc() now iterates over the SCC anyway, the update is\nmoved back to __unix_walk_scc() and \u0027fast\u0027 argument is dropped.",
"id": "GHSA-3rq9-mjhw-q3g6",
"modified": "2026-10-06T09:31:33Z",
"published": "2026-10-06T09:31:33Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-98289"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/4a4263dfeabad72f95e8ab6e15146861fa4144dd"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/b8f71d69380a5fbc734c868acbeb560f1de1766f"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/be550ee108ad80828aa61280400bedfc7e7c39d3"
}
],
"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.
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.