https://cve.radiocsirt.org Contains only the most 10 recent entries. http://www.rssboard.org/rss-specification python-feedgen en Tue, 20 Oct 2026 21:50:46 +0000 https://cve.radiocsirt.org/vuln/bit-tensorflow-2022-41894 <p><strong>Affected:</strong> Bitnami: tensorflow</p> <p>TensorFlow is an open source platform for machine learning. The reference kernel of the `CONV_3D_TRANSPOSE` TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result. Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels. An attacker can craft a model with a specific number of input channels. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter. We have patched the issue in GitHub commit 72c0bdcb25305b0b36842d746cc61d72658d2941. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range.</p> <p><strong>Affected:</strong> Bitnami: tensorflow</p> <p>TensorFlow is an open source platform for machine learning. The reference kernel of the `CONV_3D_TRANSPOSE` TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result. Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels. An attacker can craft a model with a specific number of input channels. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter. We have patched the issue in GitHub commit 72c0bdcb25305b0b36842d746cc61d72658d2941. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range.</p> https://cve.radiocsirt.org/vuln/bit-tensorflow-2022-41894 https://cve.radiocsirt.org/vuln/cnvd-2022-81227 cnvd-2022-81227 https://cve.radiocsirt.org/vuln/cnvd-2022-81227 https://cve.radiocsirt.org/vuln/euvd-2026-232531 EUVD-2026-232531 https://cve.radiocsirt.org/vuln/euvd-2026-232531 https://cve.radiocsirt.org/vuln/fkie_cve-2022-41894 <p>TensorFlow is an open source platform for machine learning. The reference kernel of the `CONV_3D_TRANSPOSE` TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result. Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels. An attacker can craft a model with a specific number of input channels. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter. We have patched the issue in GitHub commit 72c0bdcb25305b0b36842d746cc61d72658d2941. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range.</p> <p>TensorFlow is an open source platform for machine learning. The reference kernel of the `CONV_3D_TRANSPOSE` TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result. Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels. An attacker can craft a model with a specific number of input channels. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter. We have patched the issue in GitHub commit 72c0bdcb25305b0b36842d746cc61d72658d2941. The fix will be included in TensorFlow 2.11. We will also cherrypick this commit on TensorFlow 2.10.1, 2.9.3, and TensorFlow 2.8.4, as these are also affected and still in supported range.</p> https://cve.radiocsirt.org/vuln/fkie_cve-2022-41894 https://cve.radiocsirt.org/vuln/ghsa-h6q3-vv32-2cq5 <p><strong>Affected:</strong> PyPI: tensorflow</p> <p>### Impact The reference kernel of the [`CONV_3D_TRANSPOSE`](https://github.com/tensorflow/tensorflow/blob/091e63f0ea33def7ecad661a5ac01dcafbafa90b/tensorflow/lite/kernels/internal/reference/conv3d_transpose.h#L121) TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result.</p> <p>Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels.</p> <p>An attacker can craft a model with a specific number of input channels in a way similar to the attached example script. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter (i.e. `experimental_op_resolver_type=tf.lite.experimental.OpResolverType.BUILTIN_REF` is used). ```python import tensorflow as tf model = tf.keras.Sequential( [ tf.keras.layers.InputLayer(input_shape=(2, 2, 2, 1024), batch_size=1), tf.keras.layers.Conv3DTranspose( filters=8, kernel_size=(2, 2, 2), padding=&#34;same&#34;, data_format=&#34;channels_last&#34;, ), ] )</p> <p>converter = tf.lite.TFLiteConverter.from_keras_model(model) tflite_model = converter.convert()</p> <p>interpreter = tf.lite.Interpreter( model_content=tflite_model,…</p> <p><strong>Affected:</strong> PyPI: tensorflow</p> <p>### Impact The reference kernel of the [`CONV_3D_TRANSPOSE`](https://github.com/tensorflow/tensorflow/blob/091e63f0ea33def7ecad661a5ac01dcafbafa90b/tensorflow/lite/kernels/internal/reference/conv3d_transpose.h#L121) TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result.</p> <p>Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels.</p> <p>An attacker can craft a model with a specific number of input channels in a way similar to the attached example script. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter (i.e. `experimental_op_resolver_type=tf.lite.experimental.OpResolverType.BUILTIN_REF` is used). ```python import tensorflow as tf model = tf.keras.Sequential( [ tf.keras.layers.InputLayer(input_shape=(2, 2, 2, 1024), batch_size=1), tf.keras.layers.Conv3DTranspose( filters=8, kernel_size=(2, 2, 2), padding=&#34;same&#34;, data_format=&#34;channels_last&#34;, ), ] )</p> <p>converter = tf.lite.TFLiteConverter.from_keras_model(model) tflite_model = converter.convert()</p> <p>interpreter = tf.lite.Interpreter( model_content=tflite_model,…</p> https://cve.radiocsirt.org/vuln/ghsa-h6q3-vv32-2cq5 https://cve.radiocsirt.org/vuln/gsd-2022-41894 gsd-2022-41894 https://cve.radiocsirt.org/vuln/gsd-2022-41894 https://cve.radiocsirt.org/vuln/msrc_cve-2022-41894 msrc_CVE-2022-41894 https://cve.radiocsirt.org/vuln/msrc_cve-2022-41894 https://cve.radiocsirt.org/vuln/pysec-2026-936 <p><strong>Affected:</strong> PyPI: tensorflow</p> <p>### Impact The reference kernel of the [`CONV_3D_TRANSPOSE`](https://github.com/tensorflow/tensorflow/blob/091e63f0ea33def7ecad661a5ac01dcafbafa90b/tensorflow/lite/kernels/internal/reference/conv3d_transpose.h#L121) TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result.</p> <p>Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels.</p> <p>An attacker can craft a model with a specific number of input channels in a way similar to the attached example script. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter (i.e. `experimental_op_resolver_type=tf.lite.experimental.OpResolverType.BUILTIN_REF` is used). ```python import tensorflow as tf model = tf.keras.Sequential( [ tf.keras.layers.InputLayer(input_shape=(2, 2, 2, 1024), batch_size=1), tf.keras.layers.Conv3DTranspose( filters=8, kernel_size=(2, 2, 2), padding=&#34;same&#34;, data_format=&#34;channels_last&#34;, ), ] )</p> <p>converter = tf.lite.TFLiteConverter.from_keras_model(model) tflite_model = converter.convert()</p> <p>interpreter = tf.lite.Interpreter( model_content=tflite_model,…</p> <p><strong>Affected:</strong> PyPI: tensorflow</p> <p>### Impact The reference kernel of the [`CONV_3D_TRANSPOSE`](https://github.com/tensorflow/tensorflow/blob/091e63f0ea33def7ecad661a5ac01dcafbafa90b/tensorflow/lite/kernels/internal/reference/conv3d_transpose.h#L121) TensorFlow Lite operator wrongly increments the data_ptr when adding the bias to the result.</p> <p>Instead of `data_ptr += num_channels;` it should be `data_ptr += output_num_channels;` as if the number of input channels is different than the number of output channels, the wrong result will be returned and a buffer overflow will occur if num_channels &gt; output_num_channels.</p> <p>An attacker can craft a model with a specific number of input channels in a way similar to the attached example script. It is then possible to write specific values through the bias of the layer outside the bounds of the buffer. This attack only works if the reference kernel resolver is used in the interpreter (i.e. `experimental_op_resolver_type=tf.lite.experimental.OpResolverType.BUILTIN_REF` is used). ```python import tensorflow as tf model = tf.keras.Sequential( [ tf.keras.layers.InputLayer(input_shape=(2, 2, 2, 1024), batch_size=1), tf.keras.layers.Conv3DTranspose( filters=8, kernel_size=(2, 2, 2), padding=&#34;same&#34;, data_format=&#34;channels_last&#34;, ), ] )</p> <p>converter = tf.lite.TFLiteConverter.from_keras_model(model) tflite_model = converter.convert()</p> <p>interpreter = tf.lite.Interpreter( model_content=tflite_model,…</p> https://cve.radiocsirt.org/vuln/pysec-2026-936