The Linux kernel’s BPF Just‑In‑Time compiler for ARM64 allocates its executable buffer with only 4‑byte alignment, while the PLT target structure contains a 64‑bit field that requires 8‑byte alignment. When the JIT buffer is not properly aligned, a concurrent update of the target field and a 64‑bit load performed by the JITed code can result in a torn read. This misaligned access may cause the kernel to jump to a corrupted address. The flaw therefore poses a risk for kernel memory corruption and potential execution of unintended code. The description states that the issue leads to a corrupted jump target, which could be interpreted as a form of arbitrary kernel code execution; this conclusion is inferred from the text.

Any Linux kernel running on an ARM64 platform that incorporates the unpatched BPF JIT logic is affected. This includes all kernel versions that contain the original 4‑byte alignment for the JIT buffer and lack the recent change to enforce 8‑byte alignment. The scope covers all distributions shipping such kernels; no specific vendor or version is listed. The affected vendors are those providing the Linux kernel compiled for ARM64.

The vulnerability carries a CVSS score of 7.0, indicating high severity. EPSS analysis shows less than a 1 % chance of exploitation in the wild, and it is not recorded in the CISA Known Exploited Vulnerabilities catalog. Exploitation requires loading a BPF program, which is typically limited to privileged users or services; thus the attack vector is likely local. Based on the description, a local attacker with the ability to load BPF code could trigger the misaligned read and potentially gain kernel‑level execution. The impact is therefore significant but the current likelihood of exploitation remains low.