CVE-2025-65890: A device-ID validation flaw in OneFlow v0.9.0 allows attackers to cause a Denial of Service (DoS) by callin...
A device-ID validation flaw in OneFlow v0.9.0 allows attackers to cause a Denial of Service (DoS) by calling flow.cuda.synchronize() with an invalid or out-of-range GPU device index.
Security readout for executives and security teams
Plain-English summary
This CVE describes a denial-of-service risk in OneFlow v0.9.0. An invalid GPU device index passed to flow.cuda.synchronize() can crash or exhaust availability. The impact is service disruption, not data theft or data modification, based on the provided CVSS vector.
Executive priority
Treat this as high priority for GPU-backed OneFlow environments exposed to users or networks. The business risk is outage of AI, training, or inference services. Prioritize inventory, exposure reduction, and vendor guidance review.
Technical view
The flaw is improper device-ID validation in OneFlow v0.9.0 CUDA synchronization handling. The CVSS 3.1 vector is AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H, indicating unauthenticated network-reachable availability impact where the vulnerable call path is exposed.
Likely exposure
Exposure is most likely in GPU-backed OneFlow deployments where remote users, APIs, notebooks, or jobs can influence CUDA device indexes. The structured affected-product metadata is incomplete and lists n/a, while the description names OneFlow v0.9.0.
Exploitation context
The source bundle does not show CISA KEV listing or active exploitation evidence. The reported issue reference is public, but the bundle does not establish exploit availability, exploitation in the wild, or a named fixed version.
Researcher notes
Evidence is limited to the CVE description, CVSS vector, and public OneFlow issue reference. The bundle does not provide patch details, affected CPEs, proof of active exploitation, or broad version-range confirmation.
Mitigation direction
Check OneFlow project guidance for a fixed or recommended version.
Avoid exposing OneFlow GPU control paths directly to untrusted users.
Validate GPU device indexes before invoking OneFlow CUDA synchronization.
Restrict job submission and API access around GPU-backed workloads.
Monitor affected services for crashes, restarts, or abnormal GPU errors.
Validation and detection
Inventory services using OneFlow and identify deployed versions.
Confirm whether OneFlow v0.9.0 is present in production or research workloads.
Review API, notebook, and batch-job paths that accept GPU device IDs.
Check whether untrusted users can influence flow.cuda.synchronize() inputs.
Review logs for crashes tied to invalid CUDA device selection.
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Potential ATT&CK relevance
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
cwe · low confidence lookup
CWE-400: Exact CWE lookup
Use the exact CWE identifier as the starting point before reviewing related ATT&CK behavior. Open the exact CWE lookup page first, then review the ATT&CK searches from that MITRE weakness context. This is a Glexia lookup hint, not an official ATT&CK mapping.
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.
We collect every scored CVSS vector available in the official CNA and ADP containers. When more than one version is present, the table keeps the source vectors side by side instead of collapsing them into the highest score.
CWE links open Glexia weakness intelligence pages with official CWE context, developer remediation guidance, and related CVE mappings.
CWE-400 · source CWE mapping
Uncontrolled Resource Consumption
Uncontrolled Resource Consumption represents a recurring weakness pattern that can create exploitable paths when design, validation, or implementation controls are missing.