mirror of
https://github.com/openglow-org/forgefirm.git
synced 2026-10-15 17:19:55 -07:00
- grblHAL-glowforge b34636f -> badc832 (0.1.20 -> 0.1.21): the host tests trimmed to one proof per rule, with the cases the cut dropped back (the tray, session-budget and envelope clamps, the switch bits as numbers, an M5 then an M3 inside the armed window); a host build may shorten the M-code answer wait (-DGFMCODE_WAIT_S, CI's 5 s, for mcode_test.py's no-answer case) while the image keeps 30 s; the homing comment covers a setup step's hand-back. The driver the image builds is unchanged in behavior. It follows205f9f2on its own because the driver's CI runs this repository's harnesses from master, and its mcode_test expects the 5 s host build. Proof: image 20261015101725 carried the driver built from badc832 (its manifest's srcrev), and campaign c-20261015103714-0c91 passed all 53 catalog tests on it with none inherited. bitbake -c fetch grblhal-glowforge from this pin succeeds (2 tasks). The tree manifest from the pins holds grblhal-glowforge at badc832 (176 files); the coverage lint over it finds no uncovered path, no empty entry and no host-tested row problem (53 tests); test_tree_manifest passes. The driver's CI on badc832, with this repository's harnesses at205f9f2, passes.