LiveActive security incident?Get immediate response
CVE Record

CVE-2021-29520: Heap buffer overflow in `Conv3DBackprop*`

TensorFlow is an end-to-end open source platform for machine learning. Missing validation between arguments to `tf.raw_ops.Conv3DBackprop*` operations can result in heap buffer overflows. This is because the implementation(https://github.com/tensorflow/tensorflow/blob/4814fafb0ca6b5ab58a09411523b2193fed23fed/tensorflow/core/kernels/conv_grad_shape_utils.cc#L94-L153) assumes that the `input`, `filter_sizes` and `out_backprop` tensors have the same shape, as they are accessed in parallel. The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

LowCVSS 2.5Not KEV-listedUpdated
Discuss exposureBack to CVEs
Glexia's TakeAutomated analysislow

Security readout for executives and security teams

Plain-English summary

This is a low-severity TensorFlow memory safety flaw. Certain 3D convolution backpropagation operations can read/write beyond expected heap bounds when tensor shapes are inconsistent. Business impact is mainly limited service disruption in systems that let users influence TensorFlow operations or model execution.

Executive priority

Treat this as routine patching unless your environment accepts untrusted TensorFlow workloads. Prioritize internet-facing or multi-tenant ML systems first, then standard dependency maintenance cycles.

Technical view

CVE-2021-29520 is a CWE-120 heap buffer overflow in tf.raw_ops.Conv3DBackprop* shape handling. The implementation assumed input, filter_sizes, and out_backprop tensors had matching shapes while accessing them in parallel. CVSS 3.1 is 2.5: local access, high complexity, low availability impact.

Likely exposure

Exposure is most likely in TensorFlow deployments running affected versions and processing untrusted or user-controlled model graphs, tensors, or raw operation calls. Standard applications with fixed trusted models have lower practical exposure.

Exploitation context

The provided sources do not show active exploitation, and KEV status is false. The CVSS vector indicates local access, low privileges, high attack complexity, no confidentiality or integrity impact, and limited availability impact.

Researcher notes

The key validation point is whether an attacker can influence Conv3DBackprop* operation arguments in an affected TensorFlow runtime. The public advisory names the fix releases but does not provide evidence of exploitation or broader product impact.

Mitigation direction

  • Upgrade TensorFlow to 2.5.0 or a patched supported branch release.
  • For pinned 2.x branches, use 2.1.4, 2.2.3, 2.3.3, or 2.4.2.
  • Limit execution of untrusted TensorFlow graphs, tensors, and raw operations.
  • Check TensorFlow advisory guidance before applying nonstandard backports or vendor builds.

Validation and detection

  • Inventory TensorFlow versions across training, inference, CI, and notebook environments.
  • Flag versions below 2.1.4, 2.2.3, 2.3.3, or 2.4.2 in affected branches.
  • Identify services accepting externally supplied models, graphs, or tensor inputs.
  • Confirm patched versions are deployed in build manifests and runtime images.
Prepared
Confidence
high
Sources
4

Public sources used

Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.

Potential ATT&CK relevance

Conservative CVE-to-ATT&CK context

These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.

ATT&CK lookup starting points

Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.

cwe · low confidence lookup

CWE-120: Exact CWE lookup

Use the exact CWE identifier as the starting point before reviewing related ATT&CK behavior. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.

Open ATT&CK lookup
cve · low confidence lookup

CVE-2021-29520 mapping review

Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.

Open ATT&CK lookup
Vulnerability profileCVE Program record
Severity
Low
CVSS
2.5 (3.1)
Known Exploited
No
Published

Vector: CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L

Official CVE source material

CNA and ADP enrichment extracted from CVE v5

These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.

1CVSS vectors
0Timeline events
0ADP providers
3Source links

CVSS vector scores

1 official score

We collect every scored CVSS vector available in the official CNA and ADP containers. When more than one version is present, the table keeps the source vectors side by side instead of collapsing them into the highest score.

ScoreVersionSeverityVectorExploitImpactSource
2.5CVSS 3.1LowCVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L11.4Primary CVE score

Vulnerability scoring details

Base CVSS 3.1 score

2.5Low
CVSS 3.1 vector shape for CVE-2021-29520Attack VectorAttack ComplexityPrivileges RequiredUser InteractionScopeConfidentiality ImpactIntegrity ImpactAvailability Impact

Vector: CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L

Attack Vector
NetworkAdjacentLocalPhysical
Attack Complexity
LowHigh
Privileges Required
NoneLowHigh
User Interaction
NoneRequired
Scope
ChangedUnchanged
Confidentiality Impact
HighLowNone
Integrity Impact
HighLowNone
Availability Impact
HighLowNone

Source materials

Affected products

Products and packages named in the record

VendorProductVersion / packageStatus
tensorflowtensorflow< 2.1.4, >= 2.2.0, < 2.2.3, >= 2.3.0, < 2.3.3, >= 2.4.0, < 2.4.2Listed
Weakness

CWE details

CWE-120

CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.

CWE-120 · source CWE mapping

Buffer Copy without Checking Size of Input ('Classic Buffer Overflow')

Buffer Copy without Checking Size of Input ('Classic Buffer Overflow') represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.

Open CWE intelligenceSame-CWE CVEs