GHSA-x7rp-74x2-mjf3 Segfault in Tensorflow

• PyPI tensorflow: >= 2.3.0, < 2.3.1 → fixed in 2.3.1 • PyPI tensorflow-cpu: >= 2.3.0, < 2.3.1 → fixed in 2.3.1 • PyPI tensorflow-gpu: >= 2.3.0, < 2.3.1 → fixed in 2.3.1

Impact The `RaggedCountSparseOutput` implementation does not validate that the input arguments form a valid ragged tensor. In particular, there is no validation that the values in the `splits` tensor generate a valid partitioning of the `values` tensor. Thus, the [following code](https://github.com/tensorflow/tensorflow/blob/0e68f4d3295eb0281a517c3662f6698992b7b2cf/tensorflow/core/kernels/count_ops.cc#L248-L265 ) sets up conditions to cause a heap buffer overflow: ```cc auto per_batch_counts = BatchedMap(num_batches); int batch_idx = 0; for (int idx = 0; idx = splits_values(batch_idx)) { batch_idx++; } const auto& value = values_values(idx); if (value >= 0 && (maxlength_ <= 0 || value < maxlength_)) { per_batch_counts[batch_idx - 1][value] = 1; } } ``` A `BatchedMap` is equivalent to a vector where each element is a hashmap. However, if the first element of `splits_values` is not 0, `batch_idx` will never be 1, hence there will be no hashmap at index 0 in `per_batch_counts`. Trying to access that in the user code results in a segmentation fault. Patches We have patched the issue in 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and will release a patch release. We recommend users to upgrade to TensorFlow 2.3.1. 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 is a variant of GHSA-p5f8-gfw5-33w4