Headline
CVE-2022-35996: Fix conv2d crash when input size is empty. · tensorflow/tensorflow@611d80d
TensorFlow is an open source platform for machine learning. If Conv2D is given empty input and the filter and padding sizes are valid, the output is all-zeros. This causes division-by-zero floating point exceptions that can be used to trigger a denial of service attack. We have patched the issue in GitHub commit 611d80db29dd7b0cfb755772c69d60ae5bca05f9. The fix will be included in TensorFlow 2.10.0. We will also cherrypick this commit on TensorFlow 2.9.1, TensorFlow 2.8.1, and TensorFlow 2.7.2, as these are also affected and still in supported range. There are no known workarounds for this issue.
@@ -759,6 +759,15 @@ def testConv2DExplicitPaddingWithDilations(self): padding=[[2, 1], [1, 2]], dilations=[2, 3])
@test_util.run_in_graph_and_eager_modes() def testConv2dOnlyPaddingReturnsZeros(self): self._VerifyValues( tensor_in_sizes=[1, 0, 2, 1], filter_in_sizes=[1, 1, 1, 1], strides=[1, 1], padding=[[1, 1], [1, 1]], expected=[0, 0, 0, 0, 0, 0, 0, 0])
def testConv2DExplicitPaddingWithLayoutOptimizer(self): # Test with Grappler’s layout optimizer, to ensure the layout optimizer # handles explicit padding correctly.
Related news
### Impact If `Conv2D` is given empty `input` and the `filter` and `padding` sizes are valid, the output is all-zeros. This causes division-by-zero floating point exceptions that can be used to trigger a denial of service attack. ```python import tensorflow as tf import numpy as np with tf.device("CPU"): # also can be triggerred on GPU input = np.ones([1, 0, 2, 1]) filter = np.ones([1, 1, 1, 1]) strides = ([1, 1, 1, 1]) padding = "EXPLICIT" explicit_paddings = [0 , 0, 1, 1, 1, 1, 0, 0] data_format = "NHWC" res = tf.raw_ops.Conv2D( input=input, filter=filter, strides=strides, padding=padding, explicit_paddings=explicit_paddings, data_format=data_format, ) ``` ### Patches We have patched the issue in GitHub commit [611d80db29dd7b0cfb755772c69d60ae5bca05f9](https://github.com/tensorflow/tensorflow/commit/611d80db29dd7b0cfb755772c69d60ae5bca05f9). The fix will be included in TensorFlow 2.10.0. We will also cherrypick this ...