2026-09-01 - sceKernelVirtualQuery + the C-runtime threading family (D431)
(kernel) sceKernelVirtualQuery implemented against
mappings(): finds the region holding the queried address,
writes its start/end (offsets 0/8), or answers not-found. Cleared the
wall past sceKernelReserveVirtualRange for both real titles
- PPSA25872 20 imports, PPSA02664 27, both FURTHER.
(kernel/libc) The C-runtime threading family - _Mtx_*
(init/lock/unlock/trylock/destroy), _Cnd_*
(init/wait/timedwait/signal/broadcast/destroy),
_Xtime_get_ticks, _Thrd_sleep - the
std layer above _Execute_once. Each maps onto
the honest sync primitives (real mutual exclusion),
declared in libc, implemented in kernel, in the declared-elsewhere
exceptions. Removed _Throw_C_error and all 13 from the
unimplemented list at once: stubbed they answered the
0x7fff0001 placeholder, which the standard library threw
(D125 shape). SURPRISE worth recording: PPSA25872 went 20→19 imports
(BACK) as this landed - the throw was a side path that called imports
before dying, so removing it lowered the count while the fault PC stayed
put (image+0x1668a51). A correct fix that moves no wall and
lowers the metric; recorded as infrastructure, not progress (principle
3).
Next wall is instruction-level and the SELF/eboot wrapper hides the
bytes (ELF p_offset does not locate loaded code; only the
loader's wrapper decode does). Recorded a
dump/disasm verb as the honest next step
rather than guessing. Kernel tests green (74+33+33); orbistoun-kernel
clippy-clean (the orbistoun-fs escape/socket clippy debt is
pre-existing, not this session's).