regression: test-core#665 red at 8d1e694a9d8d (auto-filed by twatch)

• Type: regression (auto-filed by Track T watcher, host borg). Untriaged.
• Found: 2026-07-12T07:43:51Z
• Test source: test/test_c_gtk_window.pas

Repro

tools/testmgr.py --tier native --job 'test-core#665' at 8d1e694a9d8d4d8c639603644f260dd467fe9e9f

Range

bad 8d1e694a9d8d, last good 2b2a0c29fdc0, 7 commit(s) in range — the watcher narrows this by idle bisect; check tstate/TSTATE.md for the current range.

Log tail

ok: /tmp/testmgr-scratch-438560/test_c_gtk_window26  [code=43578B  data=1268B  bss=8924B  procs=13619]
xvfb-run: error: Xvfb failed to start

Stub ticket: signal only. Track T agent (face 2) enriches or a dev track takes it from the repro line.

Triage (Track T, 2026-07-12)

Not a compiler regression — test-infra flake. Log tail is xvfb-run: error: Xvfb failed to start, after the test binary already built ok. The gtk recipes (Makefile test-core) called bare xvfb-run, which pins servernum :99. borg runs two testmgr checkouts (dev tree + watcher clone), so whoever grabbed :99 first won and the other's gtk jobs died. testmgr's {"xvfb"} exclusive resource only serializes jobs within one run, not across runs.

Reproduced by occupying :99 and running the old recipe (same error verbatim); xvfb-run -a passes. Fixed in fe462098. All four gtk jobs green at HEAD.

• 2026-07-12 — resolved, commit fe462098.