thehackernews.com
Aug 7, 2026
New Interrupt Injection Attack Can Bypass Spectre v2 Defenses on Intel and AMD CPUs
AI-summarised brief · reviewed before publication
MIT CSAIL researchers discovered INTERRUPT INJECTION, a technique that lets an unprivileged Linux program time a hardware interrupt to re‑poison a processor’s branch predictor after Spectre v2 mitigations have run. On an AMD Zen 2 system with Linux 6.14 and all default defenses, the exploit leaked kernel memory at 5.47 bytes per second, successfully reading /etc/shadow in half of ten attempts. AMD and Intel were notified; AMD issued a bulletin and a kernel patch was merged on June 2. The fix prevents Safe‑RET from being disrupted by interrupts.
💡 Why It Matters
- · The attack shows that even hardened Spectre v2 defenses can be bypassed by precise interrupt timing, exposing kernel memory on shared systems.
- · It underscores the need for continuous vigilance and rapid patching of subtle microarchitectural vulnerabilities.
