2026-09-01 - (/loop) scePthreadGetthreadid; PPSA21564 now runs at 0% stubs

With PPSA21564 booting (D451), its hot loop called scePthreadGetthreadid 22.5k times against the placeholder, so every thread reported one shared id. Implemented it in orbistoun-kernel as the calling thread's registry handle (thread::adopt) - the same unique-per-thread value scePthreadSelf answers, and a safe zeroed-block address if the guest dereferences it (the D151 reasoning). Recorded D452.

Result: PPSA21564 runs at 0% stubs - 499082 of 499087 calls implemented, the five remaining being singletons (_ZSt14_Random_devicev, _init_env, sceKernelGetGPI, pthread_key_create). Runs to the call budget with no fault, repeatably. Getting further now needs the budget raised or the loop's wait (video/input) wired, not another missing import.

fmt/clippy/tests/knowledge audit pass for the additions. The kernel crate carries pre-existing fmt drift from earlier-session work (benign line-wrapping in tests/comments, deliberately left un-reformatted); none of it is in this iteration's additions, which are clean. Additive change (one new symbol); other titles unaffected.