← board

regression: test-smoke#11 red at 163ffea562fa (auto-filed by twatch)

Repro

tools/testmgr.py --tier native --job 'test-smoke#11' at 163ffea562fa3ae292765bc5ecf21fd2a0f45ac0

Range

bad 163ffea562fa, last good 7b95a7470ebc, 8 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-3482618/pascal26-self  [code=4101239B  data=132128B  bss=320541336B  procs=1789]
ok: /tmp/testmgr-scratch-3482618/pascal26-next  [code=4101239B  data=132128B  bss=320541336B  procs=1789]
ok: /tmp/testmgr-scratch-3482618/smoke_boot26  [code=32300B  data=232B  bss=8932B  procs=65]
ok: /tmp/testmgr-scratch-3482618/pascal26-fixedpoint  [code=4101239B  data=132128B  bss=320541336B  procs=1789]
cmp: EOF on /tmp/testmgr-scratch-3482618/pascal26-fixedpoint after byte 2371584, in line 1465

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

Triage (2026-07-12, opus-night)

Transient, not a regression: the watcher itself re-tested the SAME SHA (163ffea562fa) and went GREEN (full) with FIXED:test-smoke#11 (tstate commit d8bfac14); TSTATE.md shows no open regressions and everything through 7bf9e80393ed GREEN. The failing log's cmp: EOF ... after byte 2371584 = a TRUNCATED fixedpoint binary while self/next both built at full size — an I/O-pressure artifact (the range contained the accidental 1450-file IDF build-tree commit + its removal, i.e. heavy checkout churn on the same box), same family as [[project_borg_red_harness_race_not_regression]]. No commit in the range touches compiler/ at all (lib/rtl DNS, examples/esp32, tickets, Makefile lib-test lines), so a genuine self-host regression was not possible from source. Closing as transient-fixed.