Vulnerability · Published Sep 25, 2020 · Assigned by GitHub_M
CVE-2020-15200Segfault in Tensorflow
Medium 5.9
In plain English
A medium-severity improper input validation in tensorflow from tensorflow. There are no confirmed reports of exploitation so far. The chance it is exploited in the next 30 days is estimated at 0.9% (higher than 58% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.
Official description
In Tensorflow before version 2.3.1, 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 code sets up conditions to cause a heap buffer overflow. 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. The issue is patched in commit 3cbb917b4714766030b28eba9fb41bb97ce9ee02 and is released in TensorFlow version 2.3.1.
Affected packages & fixed versions
| Ecosystem | Package | Fixed in | Advisory |
|---|---|---|---|
| PyPI | tensorflow | 2.3.1 | GHSA-x7rp-74x2-mjf3 |
| PyPI | tensorflow | 2.3.1 | PYSEC-2020-123 |
| PyPI | tensorflow-cpu | 2.3.1 | GHSA-x7rp-74x2-mjf3 |
| PyPI | tensorflow-cpu | 2.3.1 | PYSEC-2020-280 |
| PyPI | tensorflow-gpu | 2.3.1 | GHSA-x7rp-74x2-mjf3 |
| PyPI | tensorflow-gpu | 2.3.1 | PYSEC-2020-315 |
Affected products (as reported)
| Vendor | Product | Versions |
|---|---|---|
| tensorflow | tensorflow | = 2.3.0 |
Weakness: CWE-20, CWE-122 · Vector: CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H