OSEC-2026-21
Vulnerability from osv_ocaml – Published: 2026-10-23 21:30 – Updated: 2026-10-26 14:15 – Source websiteANT-2026-KKC4QS4S Str partial-match stale capture registers underflow
Str.string_partial_match records the beginning and end of a captured group.
There was an implicit assumption that beginning <= end.
However if it reached the end of the matched string but not the end of the regexp, it was possible for the beginning of the group to be matched at iteration N of the star operator, and the end of the group to be matched at iteration N-1, resulting in beginning > end.
Str.matched_group is safe, it fails with an Invalid_argument on groups with beginning > end, but Str.replace_matched can crash, trying to memcpy end - beginning bytes.
The example in the ANT is ab partially matched against \(\(a\)b\)+. Group 2 has beginning = 2 and end = 1.
Fix
The fix is to record groups with beginning > end as not matched.
Am I affected?
All of the following conditions are necessary:
* you need to use Str.string_partial_match
* if you don't use this function you are not affected
* you need to use one of Str.{replace_matched,global_replace,replace_first} on the same regexp after a call to Str.string_partial_match.
* If you use Str.group_beginning and Str.group_end and do your own string manipulation that is safe, because the String module would raise on out of bounds accesses. Similarly Str.matched_group is safe (uses String functions in its OCaml implementation)
* the match replacement must reference a group > 0
* if the replacement string argument is fixed, and doesn't reference any groups, or references only group 0, then you are not affected
* the referenced replacement group must be under a repetition operator in the original regexp
* if there is no repetition operator in the regexp, or there are no groups under the repetition operator then you are not affected
For example {|\(\(a\)b\)+|} as the regular expression in a call to Str.string_partial_match, followed by a reference to \2 in the argument of Str.replace_matched could be affected.
Timeline
- 2026-08-24: vulnerability was found
- 2026-08-27: forwarded report to security@ocaml.org
- 2026-09-29: fix merged to trunk
{
"affected": [
{
"ecosystem_specific": {
"affected_bindings": [
"Str.string_partial_match",
"Str.replace_matched",
"Str.global_replace",
"Str.replace_first"
],
"opam_constraint": "ocaml {\u003c \"4.14.5\" | \u003e= \"5\" \u0026 \u003c \"5.5.2\"}"
},
"package": {
"ecosystem": "opam",
"name": "ocaml",
"purl": "pkg:opam/ocaml"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "4.14.5"
}
],
"type": "ECOSYSTEM"
},
{
"events": [
{
"introduced": "5"
},
{
"fixed": "5.5.2"
}
],
"type": "ECOSYSTEM"
}
],
"versions": [
"3.07",
"3.07+1",
"3.07+2",
"3.08.0",
"3.08.1",
"3.08.2",
"3.08.3",
"3.08.4",
"3.09.0",
"3.09.1",
"3.09.2",
"3.09.3",
"3.10.0",
"3.10.1",
"3.10.2",
"3.11.0",
"3.11.1",
"3.11.2",
"3.12.0",
"3.12.1",
"4.00.0",
"4.00.1",
"4.01.0",
"4.02.0",
"4.02.1",
"4.02.2",
"4.02.3",
"4.02.4",
"4.03.0",
"4.03.1",
"4.04.0",
"4.04.1",
"4.04.2",
"4.04.3",
"4.05.0",
"4.05.1",
"4.06.0",
"4.06.1",
"4.06.2",
"4.07.0",
"4.07.1",
"4.07.2",
"4.08.0",
"4.08.1",
"4.08.2",
"4.09.0",
"4.09.1",
"4.09.2",
"4.10.0",
"4.10.1",
"4.10.2",
"4.10.3",
"4.11.0",
"4.11.1",
"4.11.2",
"4.11.3",
"4.12.0",
"4.12.1",
"4.12.2",
"4.13.0",
"4.13.1",
"4.13.2",
"4.14.0",
"4.14.1",
"4.14.2",
"4.14.3",
"4.14.4",
"5.0.0",
"5.0.1",
"5.1.0",
"5.1.1",
"5.1.2",
"5.2.0",
"5.2.1",
"5.2.2",
"5.3.0",
"5.3.1",
"5.4.0",
"5.4.1",
"5.4.2",
"5.5.0",
"5.5.1"
]
},
{
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "824dadcbde77aea27d84b8d70a72442b8fcb344a"
}
],
"repo": "https://github.com/ocaml/ocaml",
"type": "GIT"
},
{
"events": [
{
"introduced": "0"
},
{
"fixed": "feccd4744ebe25b90208e569d03efd514b964dce"
}
],
"repo": "https://github.com/ocaml/ocaml",
"type": "GIT"
},
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2f8dc3dfa3bd99c1b8af68bbd5cb9f5593711b16"
}
],
"repo": "https://github.com/ocaml/ocaml",
"type": "GIT"
}
]
}
],
"credits": [
{
"name": "Anthropic",
"type": "REPORTER"
},
{
"name": "Hannes Mehnert",
"type": "COORDINATOR"
},
{
"name": "Edwin T\u00f6r\u00f6k",
"type": "COORDINATOR"
},
{
"name": "Nicol\u00e1s Ojeda B\u00e4r",
"type": "REMEDIATION_REVIEWER"
},
{
"name": "Edwin T\u00f6r\u00f6k",
"type": "REMEDIATION_REVIEWER"
},
{
"name": "Xavier Leroy",
"type": "REMEDIATION_DEVELOPER"
}
],
"database_specific": {
"cwe": [
"CWE-122"
],
"human_link": "https://github.com/ocaml/security-advisories/tree/main/advisories/2026/OSEC-2026-21.md",
"osv": "https://github.com/ocaml/security-advisories/tree/generated-osv/2026/OSEC-2026-21.json"
},
"details": "ANT-2026-KKC4QS4S Str partial-match stale capture registers underflow \n\n`Str.string_partial_match` records the beginning and end of a captured group.\nThere was an implicit assumption that `beginning \u003c= end`.\nHowever if it reached the end of the matched string but not the end of the regexp, it was possible for the beginning of the group to be matched at iteration N of the star operator, and the end of the group to be matched at iteration N-1, resulting in `beginning \u003e end`.\n`Str.matched_group` is safe, it fails with an `Invalid_argument` on groups with `beginning \u003e end`, but `Str.replace_matched` can crash, trying to memcpy `end - beginning` bytes.\n\nThe example in the ANT is `ab` partially matched against `\\(\\(a\\)b\\)+`. Group 2 has `beginning = 2` and `end = 1`.\n\n## Fix\n\nThe fix is to record groups with `beginning \u003e end` as not matched.\n\n## Am I affected?\n\nAll of the following conditions are necessary:\n * you need to use `Str.string_partial_match`\n * if you don\u0027t use this function you are not affected\n * you need to use one of `Str.{replace_matched,global_replace,replace_first}` on the same regexp after a call to `Str.string_partial_match`.\n * If you use `Str.group_beginning` and `Str.group_end` and do your own string manipulation that is safe, because the String module would raise on out of bounds accesses. Similarly `Str.matched_group` is safe (uses `String` functions in its OCaml implementation)\n * the match replacement must reference a group \u003e 0\n * if the replacement string argument is fixed, and doesn\u0027t reference any groups, or references only group 0, then you are not affected\n * the referenced replacement group must be under a repetition operator in the original regexp\n * if there is no repetition operator in the regexp, or there are no groups under the repetition operator then you are not affected\n\nFor example `{|\\(\\(a\\)b\\)+|}` as the regular expression in a call to `Str.string_partial_match`, followed by a reference to `\\2` in the argument of `Str.replace_matched` could be affected.\n\n## Timeline\n\n- 2026-08-24: vulnerability was found\n- 2026-08-27: forwarded report to security@ocaml.org\n- 2026-09-29: fix merged to trunk",
"id": "OSEC-2026-21",
"modified": "2026-10-26T14:15:00Z",
"published": "2026-10-23T21:30:00Z",
"references": [],
"schema_version": "1.7.4",
"severity": [
{
"score": "CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:N/VC:N/VI:L/VA:H/SC:N/SI:N/SA:N/E:P/AU:Y/R:U",
"type": "CVSS_V4"
}
],
"summary": "Str partial-match stale capture registers underflow"
}
Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.
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.
Vulnerability timeline
Sightings and recorded dates, grouped by day · UTC
How durations are calculated
Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.
Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.
Loading timeline…
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.