The nouveau GPU driver attempted an auxiliary transfer through the /dev/drm_dp_* interface while the GPU was in a runtime suspended state. Instead of indicating that the device was unavailable, the driver sent a request to the GPU’s GSP engine, which cannot process commands when the device is asleep. This invalid operation caused the kernel to panic and terminate the operating system. The crash demonstrates a fault that can be triggered by local code exercising the /dev/drm_dp_* interface during suspend, leading to a complete system halt.

All Linux kernel releases that include the nouveau driver and lack the 2026‑43381 patch are affected. The vulnerability was observed on kernel 6.18.10‑200.fc43.x86_64 and applies to other kernel versions that ship the unpatched driver, regardless of distribution or vendor. The affected kernels are listed in the provided CPEs, including kernel 7.0 rc1 and rc2 variants.

The CVSS score of 5.5 indicates medium severity, and the EPSS score of < 1% suggests a low probability of exploitation. The vulnerability is not listed in the CISA KEV catalog, implying no widespread exploitation has been reported. Based on the description, the attack vector is local; an attacker must run code that can invoke the /dev/drm_dp_* interface while the GPU is asleep, such as a privileged process or firmware‑upgrade client. The effect is a denial of service via a kernel crash, but the lack of a remote or privilege‑escalation path reduces the overall risk compared to a high‑impact remote exploit.