Skip to content

feat(igla): Wave Loop 425 — OSCFSEL 0–7 sweep, PVT worst-case envelope theorems#1375

Open
gHashTag wants to merge 15 commits into
masterfrom
wave-loop-425
Open

feat(igla): Wave Loop 425 — OSCFSEL 0–7 sweep, PVT worst-case envelope theorems#1375
gHashTag wants to merge 15 commits into
masterfrom
wave-loop-425

Conversation

@gHashTag

@gHashTag gHashTag commented Jul 5, 2026

Copy link
Copy Markdown
Owner

Closes #1374

Wave Loop 425 executes Variant C: the bench remains blocked (P12 unwired, no relay gate), so work focused on formal/tooling hardening.

What changed

  • cli/tri/src/fpga.rs: default cclk-sweep and smoke-gate dry-run OSCFSEL range expanded from 0–5 to 0–7.
  • proofs/lean4/Trinity/TernaryFPGABoot.lean: added combined-worst-case envelope theorems pvt_half_ns_worst_case_is_upper_envelope and pvt_low_ns_worst_case_is_upper_envelope; moved OSCFSEL_WORST_CASE_PVT_CONTEXT earlier for proof visibility.
  • docs/reports/WAVE_LOOP_425_REPORT.md, FPGA_LOOP_EVIDENCE_W425_2026-07-05.md, FPGA_LOOP_COOPERATION_W426_2026-07-05.md.
  • .trinity/current-issue.md + .trinity/experience.md: W425 close-out and learnings.

Verification

🤖 Generated with Claude Code

Dmitrii Vasilev and others added 15 commits July 5, 2026 02:27
…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>
Refs: #1362
Closes #1361

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>
@github-actions

github-actions Bot commented Jul 5, 2026

Copy link
Copy Markdown
Contributor

📓 NotebookLM Notebook linked to this PR

This notebook contains session context, decisions, and artifacts for this work.

@github-actions

github-actions Bot commented Jul 5, 2026

Copy link
Copy Markdown
Contributor

PR Dashboard

Generated at: 2026-07-05 07:38:33 UTC

Summary

Status Count
Total Open PRs 19
PRs with Failing Checks 14
PRs with All Checks Green 5
READY 4
FAILING 14
PENDING 0

Seal Status

  • ⚠️ STALE -- sha256(compiler.rs)=2034a5f9760b != manifest seal=49e55df6d444.
    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).

gHashTag pushed a commit that referenced this pull request Jul 5, 2026
Inherit W425 changes (PR #1375, closes #1374) so W426 work builds on the
OSCFSEL 0–7 sweep and PVT worst-case envelope theorems.

[ref: #1374]
@gHashTag

gHashTag commented Jul 6, 2026

Copy link
Copy Markdown
Owner Author

Closing: stale wave-loop-4xx PR, superseded by later waves. Reopen if still relevant.

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.

Wave Loop 425 — FPGA board evidence / relay gate / PVT falsification

1 participant