CVE-2021-29540Public exploit

Heap buffer overflow via unchecked filter dimensions

Published May 14, 2021 · Updated Aug 3, 2024

Heap buffer overflow in TensorFlow allows local users to access heap memory outside an allocated filter buffer via crafted tensors. Conv2DCustomBackpropFilterOp computes a filter size without verifying that it equals the filter_backprop tensor's element count, then uses that size for buffer access. Reachability requires low-privileged local execution of a TensorFlow graph containing the raw operation; malformed inputs can corrupt memory or crash the process.

CVSS severity7.8
High
EPSS probability0.21%
Next 30 days · Sep 16, 2026
Known exploitationUnconfirmed
Based on sourced intelligence
Hinoki checkNot available
Coverage for this vulnerability

See if you're affected

Explore vulnerability checks for your environment with Hinoki.

Book a demo

Summary

What happened

Heap buffer overflow in TensorFlow allows local users to access heap memory outside an allocated filter buffer via crafted tensors. Conv2DCustomBackpropFilterOp computes a filter size without verifying that it equals the filter_backprop tensor's element count, then uses that size for buffer access. Reachability requires low-privileged local execution of a TensorFlow graph containing the raw operation; malformed inputs can corrupt memory or crash the process.

The record

CVE
CVE-2021-29540
Published
May 14, 2021
Updated
Aug 3, 2024
Vendor
Google
Product
TensorFlow
Classifications
CWE-787, CWE-120
Attack vector
local
Privileges
authenticated

Timeline

How it unfolded

  1. May 14, 2021CVE publishedPublication date reported by the CVE source.
  2. Aug 3, 2024Record updatedLatest update available in the CVE record.

Exploitability

Present is not the same as exploitable

Compare your product and version with the public record. A matching version still requires validation against your environment.

Is a vulnerable build present?

Compare these published version ranges with your installed build and any vendor patches.

  1. Affected versionversion=< 2.1.4
  2. Affected versionversion=>= 2.2.0, < 2.2.3
  3. Affected versionversion=>= 2.3.0, < 2.3.3
  4. Affected versionversion=>= 2.4.0, < 2.4.2

What conditions does exploitation require?

Attack vectorlocal
Required privilegesauthenticated

What is affected?

Google · TensorFlowversion=< 2.1.4; version=>= 2.2.0, < 2.2.3; version=>= 2.3.0, < 2.3.3; version=>= 2.4.0, < 2.4.2

Published CVSS scores

7.8NIST NVDCVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
4.6NIST NVDAV:L/AC:L/Au:N/C:P/I:P/A:P
2.5GitHub CNACVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L

CVSS describes severity. EPSS estimates exploitation probability.

Attacks

What attackers are doing with it

Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.

Daily unique IPsNo honeypot observations are available for this CVE in the selected window.

No observations available

Sep 9, 2026Sep 15, 2026
Latest reporting daySep 15, 2026
Latest daily unique IPsUnavailable
Prior 30-day averageUnavailable
SourceShadowserver honeypots (KEV)
Vectorlocal
Privilegesauthenticated
Known exploitationUnconfirmed
Public exploitPublished

Weakness, pattern, technique

CWE-787Out-of-bounds Write
CWE-120Buffer Copy without Checking Size of Input ('Classic Buffer Overflow')

Public exploit references

Labels summarize the accepted research assessment. They do not indicate a test against your environment.

Technologies

Your stack

See the directory against your own environment.

Your stack

Check the software in your environment

Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.

Book a demo