Great they're getting on it. I'll note that the early security kernels attempted to mitigate timing channels. GEMSOS used event counts and sequences IIRC. It wa

Great they're getting on it. I'll note that the early security kernels attempted to mitigate timing channels. GEMSOS used event counts and sequences IIRC. It was also mandatory for high-assurance security under TCSEC regulations to look for them and try to mitigate them. There was no reason to think it would actually work over time at the rate bypasses were discovered. They were trying, though. More recently, under the MILS model for partitioning kernels, they've applied stuff like ARINC time and space partitioning to separation kernels for both safety- and security-critical applications. Big in aerospace for DO-178C regulations. One example [1] describing some timing protections. That can mitigate some timing channels. The one good thing about Meltdown/Spectre's publicity is CompSci folks are coming up with all kinds of stuff on this problem. [1] https://www.ghs.com/products/safety_critical/integrity-do-178b.html https://www.ghs.com/products/safety_critical/integrity-do-17...