# CVE-2020-15211: Out of bounds access in tensorflow-lite > An out-of-bounds access vulnerability in TensorFlow Lite allows attackers to read or write heap memory via crafted models, potentially leading to code execution. - **Severity:** Medium 4.8 (`CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:N`) - **Exploited in the wild:** Not known - **Login needed:** Not needed - **Attackable:** Over the internet - **Victim action needed:** Not needed - **Fix available:** Yes - **Published:** 25 Sep 2020 - **Updated:** 4 Aug 2024 - **Also known as:** BIT-tensorflow-2020-15211, GHSA-cvpc-8phh-8f45, PYSEC-2020-134, PYSEC-2020-291, PYSEC-2020-326 ## What is CVE-2020-15211? CVE-2020-15211 is a medium-severity memory corruption vulnerability in Google Tensorflow, published on 25 Sep 2020. An attacker can exploit it over the internet without logging in, with no action from the victim. ### Official description In TensorFlow Lite before versions 1.15.4, 2.0.3, 2.1.2, 2.2.1 and 2.3.1, saved models in the flatbuffer format use a double indexing scheme: a model has a set of subgraphs, each subgraph has a set of operators and each operator has a set of input/output tensors. The flatbuffer format uses indices for the tensors, indexing into an array of tensors that is owned by the subgraph. This results in a pattern of double array indexing when trying to get the data of each tensor. However, some operators can have some tensors be optional. To handle this scenario, the flatbuffer model uses a negative `-1` value as index for these tensors. This results in special casing during validation at model loading time. Unfortunately, this means that the `-1` index is a valid tensor index for any operator, including those that don't expect optional inputs and including for output tensors. Thus, this allows writing and reading from outside the bounds of heap allocated arrays, although only at a specific offset from the start of these arrays. This results in both read and write gadgets, albeit very limited in scope. The issue is patched in several commits (46d5b0852, 00302787b7, e11f5558, cd31fd0ce, 1970c21, and fff2c83), and is released in TensorFlow versions 1.15.4, 2.0.3, 2.1.2, 2.2.1, or 2.3.1. A potential workaround would be to add a custom `Verifier` to the model loading code to ensure that only operators which accept optional inputs use the `-1` special value and only for the tensors that they expect to be optional. Since this allow-list type approach is erro-prone, we advise upgrading to the patched code. Weakness: CWE-787 (Out-of-bounds Write), CWE-125 (Out-of-bounds Read) ## Which versions are affected by CVE-2020-15211? 34 open source packages ship an affected version, including tensorflow, tensorflow-cpu and tensorflow-gpu. It affects 2 products. | Package | Ecosystem | Affected | Fixed in | | --- | --- | --- | --- | | `tensorflow` | Bitnami | All versions | 1.15.4, 2.0.3, 2.1.2, 2.2.1, 2.3.1 | | `tensorflow` | PyPI | All versions | 1.15.4, 2.0.3, 2.1.2, 2.2.1, 2.3.1 | | `tensorflow` | PyPI | All versions | 1.15.4 | | `tensorflow` | PyPI | All versions | 1.15.4 | | `tensorflow` | PyPI | From 2.0.0 | 2.0.3 | | `tensorflow` | PyPI | From 2.0.0 | 2.0.3 | | `tensorflow` | PyPI | From 2.1.0 | 2.1.2 | | `tensorflow` | PyPI | From 2.1.0 | 2.1.2 | | `tensorflow` | PyPI | From 2.2.0 | 2.2.1 | | `tensorflow` | PyPI | From 2.2.0 | 2.2.1 | | `tensorflow` | PyPI | From 2.3.0 | 2.3.1 | | `tensorflow` | PyPI | From 2.3.0 | 2.3.1 | | `tensorflow-cpu` | PyPI | All versions | 1.15.4, 2.0.3, 2.1.2, 2.2.1, 2.3.1 | | `tensorflow-cpu` | PyPI | All versions | 1.15.4 | | `tensorflow-cpu` | PyPI | All versions | 1.15.4 | | `tensorflow-cpu` | PyPI | From 2.0.0 | 2.0.3 | | `tensorflow-cpu` | PyPI | From 2.0.0 | 2.0.3 | | `tensorflow-cpu` | PyPI | From 2.1.0 | 2.1.2 | | `tensorflow-cpu` | PyPI | From 2.1.0 | 2.1.2 | | `tensorflow-cpu` | PyPI | From 2.2.0 | 2.2.1 | | `tensorflow-cpu` | PyPI | From 2.2.0 | 2.2.1 | | `tensorflow-cpu` | PyPI | From 2.3.0 | 2.3.1 | | `tensorflow-cpu` | PyPI | From 2.3.0 | 2.3.1 | | `tensorflow-gpu` | PyPI | All versions | 1.15.4, 2.0.3, 2.1.2, 2.2.1, 2.3.1 | | `tensorflow-gpu` | PyPI | All versions | 1.15.4 | | `tensorflow-gpu` | PyPI | All versions | 1.15.4 | | `tensorflow-gpu` | PyPI | From 2.0.0 | 2.0.3 | | `tensorflow-gpu` | PyPI | From 2.0.0 | 2.0.3 | | `tensorflow-gpu` | PyPI | From 2.1.0 | 2.1.2 | | `tensorflow-gpu` | PyPI | From 2.1.0 | 2.1.2 | | `tensorflow-gpu` | PyPI | From 2.2.0 | 2.2.1 | | `tensorflow-gpu` | PyPI | From 2.2.0 | 2.2.1 | | `tensorflow-gpu` | PyPI | From 2.3.0 | 2.3.1 | | `tensorflow-gpu` | PyPI | From 2.3.0 | 2.3.1 | Products: Google Tensorflow, Opensuse Leap. ## How do I fix CVE-2020-15211? Upgrade tensorflow to a fixed version for your release line: 1.15.4, 2.0.3, 2.1.2 or 2.2.1. ## Is CVE-2020-15211 being exploited? There is no known exploitation of CVE-2020-15211 in the wild and no public exploit code. ## Other vulnerabilities in Google Tensorflow - [CVE-2023-25668](https://insights.sbomify.com/CVE-2023-25668/): TensorFlow vulnerable to heap out-of-buffer read in the QuantizeAndDequantize operation (critical 9.8) - [CVE-2018-7575](https://insights.sbomify.com/CVE-2018-7575/): Integer Overflow or Wraparound in Google TensorFlow (critical 9.8) - [CVE-2021-37678](https://insights.sbomify.com/CVE-2021-37678/): Arbitrary code execution due to YAML deserialization (critical 9.3) - [CVE-2021-35958](https://insights.sbomify.com/CVE-2021-35958/): Path traversal in n/a (critical 9.1) - [CVE-2020-15202](https://insights.sbomify.com/CVE-2020-15202/): Integer truncation in Shard API usage (critical 9.0) ## Fix prompt for AI coding agents ```text Check this codebase for CVE-2020-15211 and fix it if it is affected. What it is: CVE-2020-15211 is a medium-severity memory corruption vulnerability in Google Tensorflow, published on 25 Sep 2020. An attacker can exploit it over the internet without logging in, with no action from the victim. Exploitation: There is no known exploitation of CVE-2020-15211 in the wild and no public exploit code. Affected packages and fixed versions: - tensorflow (Bitnami): affected all versions; fixed in 1.15.4, 2.0.3, 2.1.2, 2.2.1, 2.3.1 - tensorflow (PyPI): affected all versions; fixed in 1.15.4, 2.0.3, 2.1.2, 2.2.1, 2.3.1 - tensorflow (PyPI): affected all versions; fixed in 1.15.4 - tensorflow (PyPI): affected all versions; fixed in 1.15.4 - tensorflow (PyPI): affected from 2.0.0; fixed in 2.0.3 - tensorflow (PyPI): affected from 2.0.0; fixed in 2.0.3 - tensorflow (PyPI): affected from 2.1.0; fixed in 2.1.2 - tensorflow (PyPI): affected from 2.1.0; fixed in 2.1.2 - tensorflow (PyPI): affected from 2.2.0; fixed in 2.2.1 - tensorflow (PyPI): affected from 2.2.0; fixed in 2.2.1 - tensorflow (PyPI): affected from 2.3.0; fixed in 2.3.1 - tensorflow (PyPI): affected from 2.3.0; fixed in 2.3.1 Task: 1. Search the dependency manifests and lockfiles (for example package.json, package-lock.json, requirements.txt, poetry.lock, uv.lock, pom.xml, build.gradle, go.mod, Cargo.lock, Gemfile.lock) and any Dockerfiles for the affected packages, including transitive dependencies. 2. If an affected version is in use, upgrade it to the fixed version for its release line with the project's package manager, so the lockfile is regenerated. Prefer the smallest upgrade that fixes it. 3. If there is no fix, or the upgrade would break the build, do not guess: explain the risk and the safest workaround instead. 4. Run the tests and summarise what changed and why. 5. If nothing in the project is affected, say so and change nothing. Reference: https://insights.sbomify.com/CVE-2020-15211/ (Markdown: https://insights.sbomify.com/CVE-2020-15211/index.md) ``` ## Sources - CVE record: https://www.cve.org/CVERecord?id=CVE-2020-15211 - NVD: https://nvd.nist.gov/vuln/detail/CVE-2020-15211 - OSV: https://osv.dev/vulnerability/CVE-2020-15211 --- From sbomify Insights: https://insights.sbomify.com/CVE-2020-15211/ Summaries are generated and may contain mistakes; the official description above is the record.