Hello, Related to TCG multithreaded, at Virtual Open Systems, we have also been working on it. More specifically we are looking on how to enable a basic multithreaded vCPU infrastructure with one cache per CPU. We intend to focus on the vCPU private tcg structures (instantiation/initialization), as initial activity. As soon as we have completed a first implementation, the plan is then to run bare-metal test cases to verify basic thread functionality and highlight future improvements (atomicity, tb invalidation, etc.).
Kind regards, alvise