In TensorFlow before version 2.6.0 an attacker can trigger a read from outside of bounds of heap allocated data by sending invalid arguments to tf.raw_ops.ResourceScatterUpdate. The implementation has an incomplete validation of the relationship between the shapes of indices and updates: instead of checking that the shape of indices is a prefix of the shape of updates (so that broadcasting can happen), code only checks that the number of elements in these two tensors are in a divisibility relationship.
In TensorFlow before version 2.6.0 an attacker can trigger a read from outside of bounds of heap allocated data by sending invalid arguments to tf.raw_ops.ResourceScatterUpdate. The implementation has an incomplete validation of the relationship between the shapes of indices and updates: instead of checking that the shape of indices is a prefix of the shape of updates (so that broadcasting can happen), code only checks that the number of elements in these two tensors are in a divisibility relationship.
https://github.com/tensorflow/tensorflow/security/advisories/GHSA-7fvx-3jfc-2cpc https://github.com/tensorflow/tensorflow/commit/01cff3f986259d661103412a20745928c727326f
Vulmon