Snyk has a proof-of-concept or detailed explanation of how to exploit this vulnerability.
The probability is the direct output of the EPSS model, and conveys an overall sense of the threat of exploitation in the wild. The percentile measures the EPSS probability relative to all known EPSS scores. Note: This data is updated daily, relying on the latest available EPSS model version. Check out the EPSS documentation for more details.
In a few clicks we can analyze your entire application and see what components are vulnerable in your application, and suggest you quick fixes.
Test your applicationsUpgrade github.com/NVIDIA/nvidia-container-toolkit/internal/discover to version 1.17.8 or higher.
Affected versions of this package are vulnerable to Untrusted Search Path though incorrect handling of the OCI hook createContainer during container initialization when enable-cuda-compat is used. An attacker with low privileges in a container can run arbitrary code with higher privileges, access sensitive data, alter information, or cause service disruptions. By setting LD_PRELOAD in the Dockerfile, the attacker could make the nvidia-ctk hook load a malicious library. Since the createContainer hook runs with its working directory at the container's root filesystem, the attacker can load the malicious library directly from the container image.
When using the NVIDIA Container Runtime in legacy mode, disable the hook by editing the /etc/nvidia-container-toolkit/config.toml file and setting the features.disable-cuda-compat-lib-hook flag to true.
[features]
disable-cuda-compat-lib-hook = true
When using the NVIDIA GPU Operator, disable the hook by adding disable-cuda-compat-lib-hook to the NVIDIA_CONTAINER_TOOLKIT_OPT_IN_FEATURES environment variable. This can be done by including the following arguments when installing or upgrading the GPU Operator with Helm:
--set
"toolkit.env[0].name=NVIDIA_CONTAINER_TOOLKIT_OPT_IN_FEATURES" \
--set
"toolkit.env[0].value=disable-cuda-compat-lib-hook"