The vulnerability originates in the Linux kernel's BPF JIT blind instruction handling, where rewriting an "if rX op imm goto off" constructs can overflow the signed 16‑bit offset field. A wrap‑around occurs when values below –32768 are used, causing the jump to leap past the end of the program. In hardened mode (bpf_jit_harden=2), this fault triggers a kernel crash rather than providing code execution or privilege escalation. The impact is a loss of kernel availability for the affected system.

The flaw affects the Linux kernel wherever BPF JIT compilation with hardening is enabled. Specific version ranges are not listed, so any kernel builds that include this BPF code path are potentially vulnerable. The issue is tied to the BPF subsystem, not to particular distributions or kernel releases.

The CVSS score is 4.1 and the EPSS score is <1%, but the vulnerability is not listed in the CISA KEV catalog. Exploitation requires the ability to load a crafted BPF program, which typically requires root or capabilities such as CAP_SYS_ADMIN. The attack vector is therefore local or potentially remote if privilege escalation is achievable. The risk is moderate due to the requirement of BPF program injection and the fact that the flaw only causes a crash, not arbitrary code execution.