Skip to content

verified: ternary_link sealed by the t27c 0.5.1 tag build; die A's 0.5.0 receipt moves to its own directory (Refs #7669) - #7787

Merged
dmitrii-f-t27 merged 1 commit into
masterfrom
verified/ternary-link-seal-051
Oct 8, 2026
Merged

dmitrii-f-t27 merged 1 commit into
masterfrom
verified/ternary-link-seal-051

Conversation

@dmitrii-f-t27

Copy link
Copy Markdown
Collaborator

Refs #7669, #7762, #7758

R3-3 receipts must all come from one t27c build, and run-record checks each receipt's toolchain against the seal's built_by. t27c 0.5.0 cannot name die B (#7758, fixed by #7761), so the run moves to the tag t27c-v0.5.1 (579bc55).

Changes

  1. fpga_ZeroDSP_TernaryLink.json is re-saved by the 0.5.1 tag build, the same step verified: ternary_link sealed by the t27c 0.5.0 tag build; R3-2 receipts move to their own run directory (Refs #7669) #7681 did for 0.5.0. Every gen hash and the spec hash are unchanged; only built_by, sealed_by and sealed_at move. Tests: 46/46.
  2. Die A's R3-3 receipt (verified: R3-3 receipt from die A 050d58218fd9854, seed 1, the #7669 nonce (Refs #7669) #7687, toolchain 0.5.0+cb9df57c5) moves unchanged into .trinity/receipts/r33-v050-2026-10-08/. Under this seal run-record would report it as toolchain is not the cited seal's built_by, and run-record reads .trinity/receipts/ without recursing. Die A needs a rerun with the 0.5.1 tag; its tail is 0x4F, so the DNA is named as before.
  3. Fix for my own mistake in verified: the FUSE_DNA tail is per-die data, so a second die is named (Closes #7758) #7761: the device_dna and die_binding seals recorded tests: blocked (runner.zig … has no member named 'Init') because zig 0.15 was on PATH when I sealed. They are re-saved by the tag build with zig 0.16.0: 15/15 and 22/22, with gen and spec hashes unchanged.

Check

With the 0.5.1 tag build, die B's rerun (#7759) prints DNA 0389c0c2d85e85c (FUSE_DNA and XSC_DNA agree, before and after the run). Its receipt: device_dna 0389c0c2d85e85c, toolchain t27c-bootstrap@0.5.1+579bc55aa, auth FRESH, die NAMED. t27c seal --verify reports all hashes MATCH for all three specs.

🤖 Generated with Claude Code

…5.0 receipt moves to its own directory (Refs #7669)

R3-3 receipts must come from one build, matched against the seal's
built_by. t27c 0.5.0 cannot name die B (#7758), so the run moves to
t27c-v0.5.1 (579bc55). As #7681 did for 0.5.0, the ternary_link seal
is re-saved by the tag build: every gen hash and the spec hash are
unchanged; built_by, sealed_by and sealed_at move.

Die A's R3-3 receipt (#7687) was built by 0.5.0, so under this seal it
is not complete. It moves unchanged into .trinity/receipts/r33-v050-2026-10-08/
until die A is rerun with the 0.5.1 tag.

The device_dna and die_binding seals from #7761 recorded tests as
blocked (sealed with zig 0.15 on PATH). They are re-saved by the tag
build with zig 0.16: 15/15 and 22/22, gen and spec hashes unchanged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions

github-actions Bot commented Oct 8, 2026

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-10-08 11:48:07 UTC

Summary

Status Count
Total Open PRs 50
PRs with Failing Checks 41
PRs with All Checks Green 9
READY 4
FAILING 41
PENDING 0
NO CHECKS YET 0

These columns do not partition: 4 + 41 + 0 + 0 = 45, and there are 50 open PRs. A PR is being counted twice or not at all.

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=39fa3908f9fd != manifest seal=87e5cbd3ad94.
    The committed NMSE numbers were certified against an older compiler.rs.
    Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

@dmitrii-f-t27
dmitrii-f-t27 merged commit ab544f4 into master Oct 8, 2026
25 checks passed
gHashTag added a commit that referenced this pull request Oct 8, 2026
)

Die 050d58218fd9854 (operator A), pnr seed 1, verifier nonce
e61798b78cbb6b287b8ff8523cea71a6, built by t27c-bootstrap@0.5.1+579bc55aa
(the tag t27c-v0.5.1), run in a tree at master a6c8ab1 whose ternary_link
seal names that build (#7787):
  A1 wrong part Done 0, B1 our bitstream Done 1,
  B2 read 0xa5a532bd design 10, clauses=1111 ok=1,
  DNA 050d58218fd9854 (FUSE_DNA and XSC_DNA agree, before and after),
  receipt signed by key 4d2ffce012238bd4, PASS.

t27c run-record on master with this receipt and die B's (#7759): 2 receipts,
both complete and FRESH, Die NAMED; first missing RUN_TOO_FEW_RECEIPTS until
die C (operator Phil, seed 42) reports. The 0.5.0 receipt of die A stays in
r33-v050-2026-10-08/.

Closes #7815

Co-authored-by: Claude <claude@anthropic.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
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