[P4] Corpus: recorded_threadStarvation_taskWithAThreadOfItsOwn can flip on a one-second stall #1183

The silent row in #1185 starts a task on a free pool thread, and ThreadStarvationDetector reads a clock, so a stall of 1000 ms between submitting and starting would fire it. No such stall has been seen: it passed every local run and the firing row cannot flake, since waiting longer only fires harder.

Why not now. There is no failure to fix, and the detector takes its time from nanoTime internally, so the row cannot remove the clock.

What it would take. If it ever flickers in CI, raise the threshold for the pair with setStarvationThresholdMs and lengthen the blocker to match. Close this if it stays quiet for a few months of runs.