GHSA-HQ37-853P-G5CF
Vulnerability from github – Published: 2021-01-06 16:57 – Updated: 2024-09-13 17:42Doyensec Vulnerability Advisory
- Regular Expression Denial of Service (REDoS) in cairosvg
- Affected Product: CairoSVG v2.0.0+
- Vendor: https://github.com/Kozea
- Severity: Medium
- Vulnerability Class: Denial of Service
- Author(s): Ben Caller (Doyensec)
Summary
When processing SVG files, the python package CairoSVG uses two regular expressions which are vulnerable to Regular Expression Denial of Service (REDoS). If an attacker provides a malicious SVG, it can make cairosvg get stuck processing the file for a very long time.
Technical description
The vulnerable regular expressions are
https://github.com/Kozea/CairoSVG/blob/9c4a982b9a021280ad90e89707eacc1d114e4ac4/cairosvg/colors.py#L190-L191
The section between 'rgb(' and the final ')' contains multiple overlapping groups.
Since all three infinitely repeating groups accept spaces, a long string of spaces causes catastrophic backtracking when it is not followed by a closing parenthesis.
The complexity is cubic, so doubling the length of the malicious string of spaces makes processing take 8 times as long.
Reproduction steps
Create a malicious SVG of the form:
<svg width="1" height="1"><rect fill="rgb( ;"/></svg>
with the following code:
'<svg width="1" height="1"><rect fill="rgb(' + (' ' * 3456) + ';"/></svg>'
Note that there is no closing parenthesis before the semi-colon.
Run cairosvg e.g.:
cairosvg cairo-redos.svg -o x.png
and notice that it hangs at 100% CPU. Increasing the number of spaces increases the processing time with cubic complexity.
Remediation
Fix the regexes to avoid overlapping parts. Perhaps remove the [ \n\r\t]* groups from the regex, and use .strip() on the returned capture group.
Disclosure timeline
- 2020-12-30: Vulnerability disclosed via email to CourtBouillon
{
"affected": [
{
"package": {
"ecosystem": "PyPI",
"name": "CairoSVG"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "2.5.1"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2021-21236"
],
"database_specific": {
"cwe_ids": [
"CWE-400"
],
"github_reviewed": true,
"github_reviewed_at": "2021-01-06T16:57:38Z",
"nvd_published_at": "2021-01-06T17:15:00Z",
"severity": "HIGH"
},
"details": "# Doyensec Vulnerability Advisory \n\n* Regular Expression Denial of Service (REDoS) in cairosvg\n* Affected Product: CairoSVG v2.0.0+\n* Vendor: https://github.com/Kozea\n* Severity: Medium\n* Vulnerability Class: Denial of Service\n* Author(s): Ben Caller ([Doyensec](https://doyensec.com))\n\n## Summary\n\nWhen processing SVG files, the python package CairoSVG uses two regular expressions which are vulnerable to Regular Expression Denial of Service (REDoS).\nIf an attacker provides a malicious SVG, it can make cairosvg get stuck processing the file for a very long time.\n\n## Technical description\n\nThe vulnerable regular expressions are\n\nhttps://github.com/Kozea/CairoSVG/blob/9c4a982b9a021280ad90e89707eacc1d114e4ac4/cairosvg/colors.py#L190-L191\n\nThe section between \u0027rgb(\u0027 and the final \u0027)\u0027 contains multiple overlapping groups.\n\nSince all three infinitely repeating groups accept spaces, a long string of spaces causes catastrophic backtracking when it is not followed by a closing parenthesis.\n\nThe complexity is cubic, so doubling the length of the malicious string of spaces makes processing take 8 times as long.\n\n## Reproduction steps\n\nCreate a malicious SVG of the form:\n\n \u003csvg width=\"1\" height=\"1\"\u003e\u003crect fill=\"rgb( ;\"/\u003e\u003c/svg\u003e\n\nwith the following code:\n\n \u0027\u003csvg width=\"1\" height=\"1\"\u003e\u003crect fill=\"rgb(\u0027 + (\u0027 \u0027 * 3456) + \u0027;\"/\u003e\u003c/svg\u003e\u0027\n\nNote that there is no closing parenthesis before the semi-colon.\n\nRun cairosvg e.g.:\n\n cairosvg cairo-redos.svg -o x.png\n\nand notice that it hangs at 100% CPU. Increasing the number of spaces increases the processing time with cubic complexity.\n\n## Remediation\n\nFix the regexes to avoid overlapping parts. Perhaps remove the [ \\n\\r\\t]* groups from the regex, and use .strip() on the returned capture group.\n\n## Disclosure timeline\n\n- 2020-12-30: Vulnerability disclosed via email to CourtBouillon",
"id": "GHSA-hq37-853p-g5cf",
"modified": "2024-09-13T17:42:15Z",
"published": "2021-01-06T16:57:50Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/Kozea/CairoSVG/security/advisories/GHSA-hq37-853p-g5cf"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2021-21236"
},
{
"type": "WEB",
"url": "https://github.com/Kozea/CairoSVG/commit/cfc9175e590531d90384aa88845052de53d94bf3"
},
{
"type": "PACKAGE",
"url": "https://github.com/Kozea/CairoSVG"
},
{
"type": "WEB",
"url": "https://github.com/Kozea/CairoSVG/releases/tag/2.5.1"
},
{
"type": "WEB",
"url": "https://github.com/pypa/advisory-database/tree/main/vulns/cairosvg/PYSEC-2021-5.yaml"
},
{
"type": "WEB",
"url": "https://pypi.org/project/CairoSVG"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"type": "CVSS_V3"
},
{
"score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N/E:P",
"type": "CVSS_V4"
}
],
"summary": "Regular Expression Denial of Service in CairoSVG"
}
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.