2026-09-01 - (/loop) time + fgets: PPSA21564 (Astro's Playroom) reaches main(), prints, parses args

Spine-crunching PPSA21564 toward code-correct headless emulation. Two libc walls in a row: printf("%s", asctime(localtime(&t))) (both time calls unimplemented -> placeholder followed as a %s), and a fgets read loop that never ended because unimplemented fgets never answered NULL.

Implemented, correctly:

• localtime/gmtime/asctime in the clock module - real epoch breakdown (Hinnant civil-from-days), FreeBSD struct tm layout (cited), ISO C asctime format, UTC (no timezone modelled). Tested against the epoch, a leap day, a recent weekday date, and a pre-epoch time.
• fgets on the read-only FILE model - one byte at a time, keeps the newline, NUL-terminates, answers NULL at EOF so the guest's read loop ends.

Result (the guest is the oracle): it runs into main() and prints its banner, a correct Current Time (our localtime/asctime), the build date, and ARGS: read from /app0/args.txt (our fgets) with the command line parsed. Verdict FURTHER; the spin is gone (0% stubs, ~500k real calls). Recorded D454.

New wall is threading: the engine asserts Cond.cpp:212 rc == 0 and aborts - a condition-variable call answers non-zero where success is expected (pthread_cond_wait -> InvalidHandle/BUSY, or the unserved pthread_cond_timedwait). That is a cond-handle-lifecycle / non-atomic-wait question for a focused turn, and it is reached only because startup now completes.

fmt/clippy/tests/knowledge pass for the additions (unsafe split one op per block per the workspace lint). Additive; other titles unaffected.