CVE-2020-26270: CHECK-fail in LSTM with zero-length input in TensorFlow
(updated )
Running an LSTM/GRU model where the LSTM/GRU layer receives an input with zero-length results in a CHECK failure when using the CUDA backend.
This can result in a query-of-death vulnerability, via denial of service, if users can control the input to the layer.
References
- github.com/advisories/GHSA-m648-33qf-v3gp
- github.com/pypa/advisory-database/tree/main/vulns/tensorflow-cpu/PYSEC-2020-301.yaml
- github.com/pypa/advisory-database/tree/main/vulns/tensorflow-gpu/PYSEC-2020-336.yaml
- github.com/pypa/advisory-database/tree/main/vulns/tensorflow/PYSEC-2020-256.yaml
- github.com/tensorflow/tensorflow
- github.com/tensorflow/tensorflow/commit/14755416e364f17fb1870882fa778c7fec7f16e3
- github.com/tensorflow/tensorflow/security/advisories/GHSA-m648-33qf-v3gp
- nvd.nist.gov/vuln/detail/CVE-2020-26270
Code Behaviors & Features
Detect and mitigate CVE-2020-26270 with GitLab Dependency Scanning
Secure your software supply chain by verifying that all open source dependencies used in your projects contain no disclosed vulnerabilities. Learn more about Dependency Scanning →