batch: merge Wave Loops 420–459 + W774-W776 ladder backlog - #1783
Merged
Merged
Conversation
…shold, PVT process-corner monotonicity Closes #1361 Variant C fallback (bench still blocked: P12 unwired, DLC10 cable missing, no relay). Instrument-import depth and formal guarding only. - cli/tri/src/fpga.rs: - exact-token VCD $date/$version/$comment terminator with regression test for embedded $end-like tokens (closes W419 report/actual gap) - real-valued VCD net auto-threshold from observed voltage swing - PVT half-period process-corner monotonicity regression test - proofs/lean4/Trinity/TernaryFPGABoot.lean: - pvt_half_ns_monotone_in_process_corner (ff ≤ tt ≤ ss) - fpga/HARDWARE_SSOT.md: §3.6.17 documenting W420 VCD/PVT improvements - .trinity/experience.md: W420 learnings - docs/reports: W420 report, evidence, and W421 cooperation variants Verification: - cargo test -p tri vcd: 13/13 PASS - cargo test -p tri pvt: 10/10 PASS - cargo test -p tri fpga::tests: 48/48 PASS - lake build Trinity.TernaryFPGABoot: PASS (2967 jobs) - ./scripts/tri test: 16 pre-existing yosys failures (#1245), no new ones Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes #1363 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
… PVT monotonicity, competitor snapshot Closes #1363 Variant C fallback (bench still blocked: openFPGALoader --detect reports 0 devices). - cli/tri/src/fpga.rs: - apply vcd_line_ends_with_token to $timescale terminator - add test_parse_vcd_timescale_with_embedded_end_token - add test_parse_vcd_real_auto_threshold_us_timescale - add test_pvt_half_ns_monotone_combined - proofs/lean4/Trinity/TernaryFPGABoot.lean: - add pvt_half_ns_monotone_combined - fpga/HARDWARE_SSOT.md: §3.6.18 documenting W421 improvements - docs/reports/T27_VS_FORMAL_HDL_2026.md: competitor comparison - .trinity/experience.md: W421 learnings - docs/reports: W421 report, evidence, and W422 cooperation variants Verification: - cargo test -p tri vcd: 15/15 PASS - cargo test -p tri pvt: 11/11 PASS - cargo test -p tri fpga::tests: 51/51 PASS - lake build Trinity.TernaryFPGABoot: PASS (2967 jobs) - ./scripts/tri test: 16 pre-existing yosys failures (#1245), no new ones Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…escape, PVT worst-case bound Closes #1365 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes #1368 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…st-case, competitor refresh (Closes #1368) - cli/tri/src/fpga.rs: CSV ms/us/ns/sample-number normalization, VCD real-net slope filter, unknown timescale fallback, dumpoff/dumpon without timestamp, --pvt-worstcase mode; +10 regression tests (60/60 PASS). - fpga/HARDWARE_SSOT.md: add §3.6.20 documenting W423 instrument-import depth. - docs/reports/T27_VS_FORMAL_HDL_2026.md: refresh Sparkle/Verilean, Clash, CIRCT. - docs/reports/WAVE_LOOP_423_REPORT.md, FPGA_LOOP_EVIDENCE_W423_2026-07-05.md, FPGA_LOOP_COOPERATION_W424_2026-07-05.md: W423 close-out + W424 variants. - docs/NOW.md, .trinity/experience.md, .trinity/current-issue.md: W423 close-out and W424 setup. Full sweep: 576 passed, 0 seal mismatches, 7 pre-existing gen-verilog yosys smoke failures, 0 FPGA smoke failures. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Update .trinity/current-issue.md and cooperation file with the new W424 issue number. Reference #1371 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…voltage units, non-blocking continue, W425 setup - Add non-blocking wait_for_continue helper for boot-log/cold-por/cclk-sweep - Parse CSV voltage units (v/mv) and scale to volts before threshold detection - Embed optional PVT context + XADC placeholder into boot-log JSON - Expand default CCLK sweep to OSCFSEL 0..7 - Add Lean 4 ProcessCorner decidability/severity helpers - W424 report, evidence, and W425 cooperation variants Closes #1371 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Update .trinity/current-issue.md with W425 goal, plan, acceptance criteria - Prepend W424 close-out and W425 setup to docs/NOW.md Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes #1371 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes #1374 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…e theorems (closes #1374) - cli/tri/src/fpga.rs: expand cclk-sweep/smoke-gate default OSCFSEL range to 0–7 - proofs/lean4/Trinity/TernaryFPGABoot.lean: add pvt_half_ns_worst_case_is_upper_envelope and pvt_low_ns_worst_case_is_upper_envelope; move OSCFSEL_WORST_CASE_PVT_CONTEXT earlier - docs/reports: W425 report, evidence, and W426 cooperation variants - .trinity/current-issue.md + experience.md: W425 close-out learnings - Variant C executed because P12 is still unwired and no relay gate is available Verification: cargo test -p tri PASS, lake build PASS, tri test PASS (7 pre-existing gen-verilog yosys smoke failures unchanged) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…e tri fpga JSON, competitor refresh (Variant C) Wave 426 executed Variant C because the physical bench is still blocked: P12 CCLK probe unwired, no relay gate, DLC10 missing. Deliverables: - proofs/lean4/Trinity/TernaryFPGABoot.lean: - pvt_half_ns_operating_rectangle_grid_bounded - pvt_low_ns_operating_rectangle_grid_bounded - exhaustive 75-point worst-case corner dominance proof - cli/tri/src/fpga.rs: - cclk_nominal_hz mirror of the Lean CCLK table - pvt_envelope_margin_ns and recommendation_from_conclusion helpers - SweepLog, boot-log, and cold-por JSON now emit recommendation and margin - 8 new unit tests (101/101 pass) - docs/reports/T27_VS_FORMAL_HDL_2026.md: Sparkle July 2026 Functional Matsuri talk, Clash 1.8.5 verification fixes, CIRCT/firtool notes - docs/reports/W426_WEAK_POINTS_AND_COMPETITORS.md: updated weak-point scan - docs/reports/WAVE_LOOP_426_REPORT.md and FPGA_LOOP_EVIDENCE_W426_2026-07-05.md - docs/reports/FPGA_LOOP_COOPERATION_W427_2026-07-05.md: next-wave variants - .trinity/experience.md: W426 learnings - docs/NOW.md: W426 close-out / W427 setup - .trinity/current-issue.md: mark Variant C acceptance criteria complete Verification: - cargo test -p tri: 101/101 pass - lake build Trinity.TernaryFPGABoot: 2967 jobs, 0 errors - ./scripts/tri test: parse/typecheck/gen/seal-verify PASS; 7 pre-existing gen-verilog-yosys-smoke failures (#1245) unchanged Closes #1376 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…ga sweep-report --json, competitor refresh (Variant C) Closes #1379 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…t-envelope --json, competitor refresh Closes #1383 - Add four quantified OSCFSEL 0..7 PVT/transaction theorems in proofs/lean4/Trinity/TernaryFPGABoot.lean. - Add --json output to tri fpga pvt-envelope with shared report builder and unit tests. - Refresh T27_VS_FORMAL_HDL_2026.md with 2026 releases and emerging signals. - Document W428 gen-verilog #1245 deferral. - Add close-out reports and W429 cooperation variants. - Update docs/NOW.md and .trinity/experience.md. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Sets up W429 tracking issue #1385 and default Variant C rationale. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
… --json, W429 report and W430 cooperation (Closes #1385) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes #1385 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…431 cooperation\n\n- Add tri fpga read-xadc and --xadc flags on boot-log/cold-por/cclk-sweep.\n- Parse openFPGALoader --read-xadc output and embed live temp/vccint/vccaux in JSON logs.\n- Add XadcOperatingPoint formal bridge and worst-case-bound theorem in TernaryFPGABoot.lean.\n- Update fpga/HARDWARE_SSOT.md, T27_VS_FORMAL_HDL_2026.md, GEN_VERILOG_DEFECTS_REPRO.md.\n- Create WAVE_LOOP_430_REPORT.md and FPGA_LOOP_COOPERATION_W431_2026-07-01.md.\n- Verification: cargo test fpga:: 79/79, lake build pass, tri test pass except 7 pre-existing #1245 failures.\n\nCloses #1388\n\nCo-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
… 430 close-out)\n\n- Point .trinity/current-issue.md and docs/NOW.md to W431 (#1389).\n- Add W430 learnings to .trinity/experience.md.\n\nCo-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
…ope, measured-to-lean --json summary hardening Closes #1389 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Contributor
|
📓 NotebookLM Notebook linked to this PR
This notebook contains session context, decisions, and artifacts for this work. |
Contributor
gHashTag
added a commit
that referenced
this pull request
Aug 6, 2026
[GOLD-RING] fix(compiler): restore declarations dropped by batch merge #1783 — master does not compile
gHashTag
added a commit
that referenced
this pull request
Aug 6, 2026
… + master diagnosis specs/ternary/gft_dot8.t27: a spec-first GF-T16 8-term dot product (realistic inference / attention-head tile length), balanced reduction tree. Verified: typecheck 0 err; in-spec test (8.0 = 22016); iverilog cross-check vs the same tree from silicon-proven gft_dot2+gft_add = ALL_PASS on 2000 random inputs (bit-exact). GF-T ladder: dot2->dot4->dot8. NOW.md documents a master codegen diagnosis: current master is red (7 internal tests + yosys broken) from #1783 batch-merge drops that #1788 only partially repaired. The tests-section emits `// synthesis translate_off` but closes with `endif (unbalanced -> breaks yosys); one-line fix = emit `ifndef SIMULATION (verified). 6 other red tests are distinct dropped features. Not pushed (multi-bug repair + reseal sweep in an active-repair area = owner decision). The spec-first hardware delta applies + builds clean on master and lands once those go green. Refs #1764 Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Batch-merge the 22 remaining open wave-loop PRs that target
master. Each wave branch was merged into this integration branch withdocs/NOW.mdconflicts resolved in favor of the currentmasterternary/gen-rust work, preserving all wave-specific specs, generators, seals, reports, plans, tests, and tracker updates.Supersedes: Closes #1362 Closes #1364 Closes #1369 Closes #1372 Closes #1373 Closes #1375 Closes #1378 Closes #1382 Closes #1384 Closes #1387 Closes #1390 Closes #1392 Closes #1394 Closes #1396 Closes #1400 Closes #1403 Closes #1406 Closes #1426 Closes #1430 Closes #1432 Closes #1434 Closes #1489