From 77320d5e4110918aff0954e53d4085d995b6ada7 Mon Sep 17 00:00:00 2001 From: Sylpheed port agent Date: Wed, 2 Sep 2026 16:40:55 +0000 Subject: [PATCH] port: put verify-menu-audio's control in check-all Every other assertion in check-all has a control; audio was the one that did not, which is how it went weeks unable to fail. Control measured just now: with move absent from the baseline the diagonal reads +0.000, the check reports 'at least one cue is not sounding' and exits 1. --- tools/port/check-all | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/tools/port/check-all b/tools/port/check-all index 1c883ca2..8050e866 100755 --- a/tools/port/check-all +++ b/tools/port/check-all @@ -117,6 +117,12 @@ step transcode-bands must-pass tools/port/verify-transcode-fidelity step bands-harness must-pass tools/port/verify-transcode-fidelity --selftest step capture-controls must-pass tools/port/check-capture-controls step menu-audio must-pass env OUT="$OUT/audio" tools/port/verify-menu-audio +# 🔴 ADDED 2026-09-02, because `menu-audio` above SPENT WEEKS UNABLE TO FAIL. It +# computed its verdict, printed a red line when a cue was silent, and its python +# had no exit path -- so it returned 0 while registered `must-pass` here. Every +# other assertion in this file has a control for exactly this reason and audio +# was the one that did not. It costs a second set of runs and that is the price. +step menu-audio-ctl must-pass env OUT="$OUT/audioctl" tools/port/verify-menu-audio --control # 🔴 ADDED 2026-09-01 after a human found Ⓐ dead on a real controller while the # unattended P5 walk passed. `--script` sends `InputEventAction`, which BYPASSES # the input map, so every check here asserted the code BELOW the map and nothing