not ok 24 - flux-top quits on q keypress

[garlick]

Problem: this failure was observed on the bookworm gcc-12,distcheck builder in an urelated PR:

2026-02-03T18:28:05.4381184Z not ok 24 - flux-top quits on q keypress
2026-02-03T18:28:05.4381434Z FAIL: t2801-top-cmd.t 24 - flux-top quits on q keypress
2026-02-03T18:28:05.4381512Z #	
2026-02-03T18:28:05.4381671Z #		$runpty --quit-char=q --format=asciicast -o keys.log flux top &
2026-02-03T18:28:05.4381742Z #		pid=$! &&
2026-02-03T18:28:05.4381818Z #		sleep 1 &&
2026-02-03T18:28:05.4381896Z #		kill -USR1 $pid &&
2026-02-03T18:28:05.4381966Z #		wait $pid
2026-02-03T18:28:05.4382030Z #	

[chu11]

My guess is that sleep 1 is the problem, as we're racing for when flux-top is actually really ready to be sent a q. For all we know runpty is still starting up, hasn't setup signal handler, and SIGUSR1 just kills it.

Other tests like this have some type of output indicating "yup I'm ready" that we wait on. Perhaps that's all way have to do on the keys.log file.