GHSA-h7ff-cfc9-wmmh TensorFlow vulnerable to `CHECK` fail in `FakeQuantWithMinMaxVarsPerChannelGradient`

• PyPI tensorflow: < 2.7.2 → fixed in 2.7.2 • PyPI tensorflow: >= 2.8.0, < 2.8.1 → fixed in 2.8.1 • PyPI tensorflow: >= 2.9.0, < 2.9.1 → fixed in 2.9.1 • PyPI tensorflow-cpu: < 2.7.2 → fixed in 2.7.2 • PyPI tensorflow-cpu: >= 2.8.0, < 2.8.1 → fixed in 2.8.1 • PyPI tensorflow-cpu: >= 2.9.0, < 2.9.1 → fixed in 2.9.1 • PyPI tensorflow-gpu: < 2.7.2 → fixed in 2.7.2 • PyPI tensorflow-gpu: >= 2.8.0, < 2.8.1 → fixed in 2.8.1 • PyPI tensorflow-gpu: >= 2.9.0, < 2.9.1 → fixed in 2.9.1

Impact When `tf.quantization.fake_quant_with_min_max_vars_per_channel_gradient` receives input `min` or `max` of rank other than 1, it gives a `CHECK` fail that can trigger a denial of service attack. ```python import tensorflow as tf arg_0=tf.random.uniform(shape=(1,1), dtype=tf.float32, maxval=None) arg_1=tf.random.uniform(shape=(1,1), dtype=tf.float32, maxval=None) arg_2=tf.random.uniform(shape=(1,1), dtype=tf.float32, maxval=None) arg_3=tf.random.uniform(shape=(1,1), dtype=tf.float32, maxval=None) arg_4=8 arg_5=False arg_6=None tf.quantization.fake_quant_with_min_max_vars_per_channel_gradient(gradients=arg_0, inputs=arg_1, min=arg_2, max=arg_3, num_bits=arg_4, narrow_range=arg_5, name=arg_6) ``` Patches We have patched the issue in GitHub commit f3cf67ac5705f4f04721d15e485e192bb319feed. 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. For more information Please consult our security guide for more information regarding the security model and how to contact us with issues and questions. Attribution This vulnerability has been reported by • 刘力源, Information System & Security and Countermeasures Experiments Center, Beijing Institute of Technology • Neophytos Christou, Secure Systems Labs, Brown University