Skip to main content

Linux Kernel EUVDEUVD-2026-55605

| CVE-2026-68419 HIGH
2026-08-10 Linux GHSA-f527-f42m-rc5q
7.8
CVSS 3.1 · Vendor: Linux
Share

Severity by source

Vendor (Linux) PRIMARY
7.8 HIGH
AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
vuln.today AI
7.8 HIGH

Local RDMA device access required (AV:L, PR:L); no race or special timing needed beyond standard RDMA verbs sequence (AC:L); full memory impact plausible via zero mkey CQP dispatch.

3.1 AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
4.0 AV:L/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
SUSE
5.5 HIGH
AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Red Hat
5.5 MEDIUM
qualitative

Primary rating from Vendor (Linux).

CVSS VectorVendor: Linux

Attack Vector
Local
Attack Complexity
Low
Privileges Required
Low
User Interaction
None
Scope
Unchanged
Confidentiality
High
Integrity
High
Availability
High

Lifecycle Timeline

5
Analysis Generated
Aug 14, 2026 - 01:51 vuln.today
CVSS changed
Aug 13, 2026 - 23:37 NVD
7.8 (HIGH)
Patch available
Aug 10, 2026 - 14:18 EUVD
CVE Published
Aug 10, 2026 - 12:04 cve.org
HIGH 7.8
CVE Published
Aug 10, 2026 - 12:04 cve.org
UNKNOWN (no severity yet)

DescriptionCVE.org

In the Linux kernel, the following vulnerability has been resolved:

RDMA/irdma: Prevent rereg_mr for non-mem regions

When a QP/CQ/SRQ is created, a two step process is used where the buffer is allocated in userspace and explicitly registered with the normal reg_mr mechanism prior to creating the actual QP/CQ/SRQ object.

These special registrations are indicated via an ABI field so the driver knows that they do not have a valid mkey and to skip the actual CQP command submission.

Since these are real MR objects from the core's perspective, it is possible for a user application to invoke rereg_mr on them and cause a real CQP op to be emitted with the zero-initialized mkey value of 0.

Fix this by preventing rereg_mr on these special regions.

AnalysisAI

Remote memory access corruption in the Linux kernel's RDMA/irdma subsystem allows a local user with RDMA device access to trigger an invalid CQP (Control Queue Pair) command carrying a zero-initialized mkey by invoking rereg_mr on special QP/CQ/SRQ buffer registrations that were never intended for re-registration. Affected systems include Intel RDMA-capable hardware environments running Linux 6.6.x through 6.6.147, 6.7, and related stable branches prior to the fix commits. No public exploit code has been identified at time of analysis, and EPSS is very low at 0.16% (6th percentile), indicating limited exploitation interest despite the 7.8 CVSS score.

Technical ContextAI

The irdma driver is Intel's kernel-space implementation of iWARP and RoCE (RDMA over Converged Ethernet) protocols for Intel Ethernet adapters. RDMA object creation (QP - Queue Pair, CQ - Completion Queue, SRQ - Shared Receive Queue) follows a two-phase model: userspace allocates the backing buffer first and registers it as a Memory Region (MR) via reg_mr; the driver uses an ABI field to mark these as 'non-mem' special regions with no valid mkey (Memory Key - RDMA's memory access token), intentionally suppressing CQP command submission during this phase. The CQP (Control Queue Pair) is the management control path used to issue commands to the Intel RDMA hardware. The flaw is that the rereg_mr code path does not check the ABI flag indicating a special non-mem region, so when rereg_mr is invoked on such an MR, the driver emits a live CQP command with the zero-initialized mkey value 0. In RDMA semantics, mkey 0 can be an invalid or privileged key whose hardware interpretation may permit unauthorized access to arbitrary registered memory. No CWE is officially assigned; structurally this maps to improper input validation or missing check on a driver-internal state flag.

RemediationAI

The primary fix is upgrading to a patched Linux kernel version: 6.6.148 or later in the 6.6 stable series, 6.12.101 or later in the 6.12 series, 7.1.6 or later in the 7.1 series, 7.2-rc4 or later, or 6.18.42 or later in the 6.18 series. Upstream commits are confirmed available at the git.kernel.org stable tree links in the references. Linux distribution vendors (RHEL, SUSE, Ubuntu, Debian, etc.) should be consulted for distribution-specific kernel updates incorporating these patches. Where immediate kernel upgrade is not feasible, the most specific compensating control is to restrict access to RDMA device nodes: remove non-root users from the rdma group and apply restrictive permissions to /dev/infiniband/* devices (e.g., chmod 600 and chown root:root), effectively preventing unprivileged RDMA verbs invocation. This mitigation disables all userspace RDMA application functionality for non-root users and is unsuitable where RDMA workloads are required. Alternatively, if the irdma module is not required, blacklisting it via /etc/modprobe.d eliminates the attack surface entirely at the cost of Intel RDMA NIC functionality.

Vendor StatusVendor

SUSE

Severity: Important
Product Status
SUSE Linux Enterprise Desktop 15 SP7 Affected
SUSE Linux Enterprise Desktop 15 SP7 Affected
SUSE Linux Enterprise High Availability Extension 15 SP7 Affected
SUSE Linux Enterprise High Availability Extension 15 SP7 Affected
SUSE Linux Enterprise High Availability Extension 16.0 Affected

Share

EUVD-2026-55605 vulnerability details – vuln.today

This site uses cookies essential for authentication and security. No tracking or analytics cookies are used. Privacy Policy