Skip to content

Resync the five drifted stage0 mirrors through the regen round - #12157

Merged
gunbai-bot[bot] merged 6 commits into
mainfrom
session/neat-carp-265-stage0-resync
Sep 24, 2026
Merged

gunbai-bot[bot] merged 6 commits into
mainfrom
session/neat-carp-265-stage0-resync

Conversation

@gunbai-bot

@gunbai-bot gunbai-bot Bot commented Sep 23, 2026

Copy link
Copy Markdown
Contributor

Regenerates the stage0 mirrors that had drifted on main. The route is claim_executor --regen-round-cost; nothing is hand-edited. Main can land this now that #12146 made the round able to install a crate-less manifest drift. It is a prerequisite for #12104: that PR adds a merge_group-only step running claim_executor --required-regen, and the step would go red on these mirrors.

Fixed point, measured on this branch's head (merged with main at the #12146 landing)

claim_executor --regen-round-cost --source-root dag --source-root src/v2: required-regen: first_generation_equal=true planned=158 executed=158 adjudicated=158, changed_paths=0, exit 0. The installing round before it (on #12146's head): changed_paths=5, convergence_stages=2, exit 0.

The five mirrors, and what each had drifted from

Each attribution below is read from the mirror's own diff against the last commit that touched it. None of them was bisected.

Why the gate did not refuse any of these

No CI step has adjudicated the stage0 mirrors since #11742. #12104 restores that check on merge_group.

🤖 Generated with Claude Code

gunbc-ci-auto-heal and others added 3 commits September 23, 2026 13:01
…ter: no crate declares it

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…12011) and regenerate to the stage0 fixed point

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@gunbai-bot

gunbai-bot Bot commented Sep 23, 2026

Copy link
Copy Markdown
Contributor Author

Merged origin/main and re-ran the regen to the fixed point. New head: ca6ab98a496.

  • The merge conflicted on v1_compiler_emit_rust.rs (GeneratedArtifactConcurrentDivergence). I followed the driver's printed recipe; neither side's bytes are hand-picked.
  • After merging, main's own --required-regen refused CommittedMirrorNoLongerEmitted [release_locus_seed_constants_generated.rs]. microVM slot controller on srv1: the unit execs run_controller as MainPID, the 28 GiB shakedown slot, the reservation-authorized planner arm, the controller's release locus #12011 committed that main_wet-generated seed file and hand-added its pub mod line to the emitted lib.rs, but never registered it. This PR adds the one missing row, WetActuatorGeneratedRegistration { basename: "release_locus_seed_constants_generated", declaration_route: PlainPubMod }, in v2.compiler.self_host.stage0_crate_layout wet_actuator_generated_registrations. Every other main_wet seed file already carries such a row. main_wet then regenerates bootstrap_stage0_crate_layout_generated.rs, gunbc_stage0_crate_layout_generated.rs and stage0_crate_layout_generated.dag.
  • Result on this head: required-regen: first_generation_equal=true planned=159 executed=159 adjudicated=159 (seed rebuilt from the installed mirrors), and required-regen-fixed-point: fixed_point_equal=true referenced_first_generation_equal=true. The PR diff is 9 files: the 5 original mirrors, plus the conflicted v1_compiler_emit_rust.rs now holding generated bytes, plus the registration row and its 3 projections.

— sent from neat-carp-265

… are not modeled (§4c)

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@briansrls
briansrls added this pull request to the merge queue Sep 23, 2026
…fixed point

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@gunbai-bot
gunbai-bot Bot removed this pull request from the merge queue due to a manual request Sep 23, 2026
@gunbai-bot
gunbai-bot Bot added this pull request to the merge queue Sep 23, 2026
Merged via the queue into main with commit 16356f4 Sep 24, 2026
5 checks passed
@gunbai-bot
gunbai-bot Bot deleted the session/neat-carp-265-stage0-resync branch September 24, 2026 03:01
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

0 participants