Description
In affected versions of TensorFlow 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. This is fixed in versions 1.15.5, 2.0.4, 2.1.3, 2.2.2, 2.3.2, and 2.4.0.
Affected products
| Vendor | Product | Affected | Fixed |
|---|---|---|---|
| tensorflow | tensorflow | < 1.15.5, >= 2.0.0, < 2.0.4, >= 2.1.0, < 2.1.3, >= 2.2.0, < 2.2.2, >= 2.3.0, < 2.3.2 | — |
Fixed in Linux distributions
Distributions backport fixes, so the version that fixes this vulnerability differs per release — upgrade to at least the version shown. From the distributions' security trackers and advisories.
SUSE
Security tracker ↗| Release | Package | Fixed in | Advisory |
|---|---|---|---|
| SUSE Package Hub 15 SP3 | bazel-skylib1.0.3 | 1.0.3-bp153.2.1 | openSUSE-SU-2022:10014-1 |
| SUSE Package Hub 15 SP3 | bazel3.7 | 3.7.2-bp153.4.1 | openSUSE-SU-2022:10014-1 |
| SUSE Package Hub 15 SP3 | tensorflow2 +3 | 2.6.0-bp153.2.3.1 | openSUSE-SU-2022:10014-1 |
openSUSE
Security tracker ↗| Release | Package | Fixed in | Advisory |
|---|---|---|---|
| openSUSE Leap 15.3 | bazel-skylib1.0.3 | 1.0.3-bp153.2.1 | openSUSE-SU-2022:10014-1 |
| openSUSE Leap 15.3 | bazel3.7 | 3.7.2-bp153.4.1 | openSUSE-SU-2022:10014-1 |
| openSUSE Leap 15.3 | tensorflow2 +3 | 2.6.0-bp153.2.3.1 | openSUSE-SU-2022:10014-1 |
Affected packages
| Ecosystem | Package | Affected | Fixed | Record |
|---|---|---|---|---|
| PyPI | tensorflow | < 1.15.5 | 1.15.5 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow | >= 2.0.0, < 2.0.4 | 2.0.4 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow | >= 2.1.0, < 2.1.3 | 2.1.3 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow | >= 2.2.0, < 2.2.2 | 2.2.2 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow | >= 2.3.0, < 2.3.2 | 2.3.2 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow | < 1.15.5 · >= 2.0.0, < 2.0.4 · >= 2.1.0, < 2.1.3 · >= 2.2.0, < 2.2.2 · >= 2.3.0, < 2.3.2 | 1.15.5, 2.0.4, 2.1.3, 2.2.2, 2.3.2 | PYSEC-2020-256 |
| PyPI | tensorflow-cpu | < 1.15.5 | 1.15.5 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-cpu | >= 2.0.0, < 2.0.4 | 2.0.4 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-cpu | >= 2.1.0, < 2.1.3 | 2.1.3 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-cpu | >= 2.2.0, < 2.2.2 | 2.2.2 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-cpu | >= 2.3.0, < 2.3.2 | 2.3.2 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-cpu | < 1.15.5 · >= 2.0.0, < 2.0.4 · >= 2.1.0, < 2.1.3 · >= 2.2.0, < 2.2.2 · >= 2.3.0, < 2.3.2 | 1.15.5, 2.0.4, 2.1.3, 2.2.2, 2.3.2 | PYSEC-2020-301 |
| PyPI | tensorflow-gpu | < 1.15.5 | 1.15.5 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-gpu | >= 2.0.0, < 2.0.4 | 2.0.4 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-gpu | >= 2.1.0, < 2.1.3 | 2.1.3 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-gpu | >= 2.2.0, < 2.2.2 | 2.2.2 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-gpu | >= 2.3.0, < 2.3.2 | 2.3.2 | GHSA-m648-33qf-v3gp |
| PyPI | tensorflow-gpu | < 1.15.5 · >= 2.0.0, < 2.0.4 · >= 2.1.0, < 2.1.3 · >= 2.2.0, < 2.2.2 · >= 2.3.0, < 2.3.2 | 1.15.5, 2.0.4, 2.1.3, 2.2.2, 2.3.2 | PYSEC-2020-336 |
Advisories 1
- openSUSE-SU-2022:10014-1 Security update for tensorflow2 SUSE / openSUSE Security Updates · Feb 4, 2026