← board

origin/master has advanced 3 commit(s) since this sha. Re-verify at current HEAD before acting — the callback is tagged to the sha that was tested, which may no longer be the state of the tree.

regression: test-core#src:test/test_managed_block_meta.pas red at 86da0606d916 (auto-filed by twatch)

Repro

tools/testmgr.py --tier native --job 'test-core#src:test/test_managed_block_meta.pas' at 86da0606d9167dbdef14eed8da5e104e5f5bd9d4

Range

bad 86da0606d916, last good c0a49076d1d3, 3 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-392706/test_managed_block_meta26  [code=91020B  data=3328B  bss=9668B  procs=198]

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

TRIAGED 2026-08-14 by Track T — consolidated into a Track A bug

Reproduced at HEAD with a rebuilt compiler: FAIL grown ascii string stays ascii.

Cause is 9ffbba0bd perf(A): append in place for \s := s + x`, which is literally the construct the failing assertion exercises — and the test's own comment names the hazard ("growth through the inline resize path must not lose or invent the flag"). Filed as [[bug-a-in-place-append-loses-the-ascii-kind-flag-on-growth]] (Track A, p75, urgent — it is in pin v299, so every $(PXX_STABLE)` lane has it).