Skip to content

fix(code): bind aligned native CPU source provenance - #16

Merged
erikleblansch merged 1 commit into
feature/native-cpu-capabilityfrom
feature/native-cpu-alignment-join
Oct 15, 2026
Merged

erikleblansch merged 1 commit into
feature/native-cpu-capabilityfrom
feature/native-cpu-alignment-join

Conversation

@erikleblansch

Copy link
Copy Markdown
Contributor

Scope

Bind the native trial to core 6d65bb12ed7b8720d43868c9306313133a5b96ae and its reviewed Q1 aligned-load correction. Validate the atomic CPU source/compile pair, exact eight-file wrapper inventory, compiler-input verification and copied-build projections. Historical 4/6/7-file formats remain readable but cannot authorize the new trial.

Verification

  • 51 targeted inert Python tests pass without skips when the retained exact core and loader sources are supplied; no native/model build or execution on the development host.
  • Independent read-only review passed.
  • Core source run 38802248241 passes both normal and ASan/UBSan: all six native targets and 98 unique Python cases per mode, including the complete upstream tensor test.
  • Hosted Code source checks remain required.

No change to model/toolset, resource bounds, permissions, handshake or cleanup. This is not a fix or success claim for the separate failed actual trial 38797816570; its rejected edit is being investigated independently. No new actual trial is dispatched by this change.

@erikleblansch
erikleblansch merged commit 10deb66 into feature/native-cpu-capability Oct 15, 2026
1 check passed
@erikleblansch
erikleblansch deleted the feature/native-cpu-alignment-join branch October 15, 2026 20:46
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.

1 participant