Correct way to thread with unicorn?

[coden-s]

I was wondering what is the correct way to use threading with unicorn. Currently, I’m working on a project emulating AArch64 on an x86 host. When threading, each thread gets its own unicorn instance and stack but with a shared code memory. Is this always safe? There haven’t been any issues with this approach so far, but I’m now testing on an RISC-V and there’s been emulation issues that appear exclusively when threading on RISC-V. (#2254)