Skip to content

ABI Layer 4: prove the FFI result-code seam is injective + faithful#41

Merged
hyperpolymath merged 4 commits into
mainfrom
claude/new-session-znxgm7
Jun 28, 2026
Merged

ABI Layer 4: prove the FFI result-code seam is injective + faithful#41
hyperpolymath merged 4 commits into
mainfrom
claude/new-session-znxgm7

Conversation

@hyperpolymath

Copy link
Copy Markdown
Owner

Summary

Layer 4 (seal the ABI↔FFI seam): proves the FFI result-code encoding (resultToInt) is a sound, lossless wire encoding — decoder intToResult with resultRoundTrip, from which resultToIntInjective is derived. Distinct ABI outcomes never collide on the C wire. Complements the structural abi-ffi-gate.py.

New module *.ABI.FfiSeam (imports Types). Positive + non-vacuity controls.

Testing

Idris2 0.7.0 --build → exit 0, zero warnings. Adversarial rejection confirmed. build/ removed. No believe_me/postulate/sorry.

🤖 Generated with Claude Code

https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx


Generated by Claude Code

claude and others added 4 commits June 27, 2026 19:24
…ition

Add Dafniser.ABI.Semantics raising the Idris2 ABI to Layer 2 with a genuine,
machine-checked semantic proof of the repo's headline ("correct-by-construction
code generation").

Model: a Dafny spec as the postcondition MaxPost a b r — a real proposition
inhabited only when r dominates both inputs and equals one of them. The
generated body genMax returns its result paired with a MaxPost proof
(correct-by-construction). genMaxCorrect derives all three ensures conjuncts
(result>=a, result>=b, result==a||result==b) for all inputs; noUndershootLeft
refutes the bad case. A sound certifier (certifyMax/certifyMaxSound) emits the
ABI's VerificationResult. Positive control maxPositive : MaxPost 7 3 7 and
negative control maxNegative : Not (MaxPost 7 3 2) are machine-checked; an
adversarial false proof (MaxPost 7 3 2) is rejected by idris2, confirming
non-vacuity.

Builds clean (idris2 0.7.0, exit 0, zero warnings).

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
…max duality

Adds Dafniser.ABI.Invariants over the existing Layer-2 model
(Dafniser.ABI.Semantics, reused unchanged). Three distinct, deeper
machine-checked properties:

1. Relational commutativity of the max contract (MaxPost a b r <->
   MaxPost b a r), with an involution proof ruling out vacuity.
2. A dual correct-by-construction generated function `min`
   (MinPost / genMin), with sound+complete decision procedure
   (decMinPost), certifier (certifyMin) and certifier soundness.
3. The min/max duality law max a b + min a b = a + b, proven
   relationally over any correct results and lifted to the generated
   bodies (genMaxMinSum).

Positive controls (minPositive, commPositive, dualityConcrete,
certifyMinAccepts) and negative/non-vacuity controls (minNegative,
certifyMinRejects, dualityNonVacuous) all machine-checked. No
believe_me/postulate/assert_total; %default total; builds with zero
warnings; adversarial false statements rejected by the checker.

Registers Dafniser.ABI.Invariants last in dafniser-abi.ipkg.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
…fiSeam)

Prove the FFI result-code encoding is sound:
- intToResult decoder + resultRoundTrip (lossless/faithful encoding)
- resultToIntInjective derived from the round-trip (no collisions on the wire)
- positive controls (decode 0=Ok, 4=NullPointer, 5=Nothing)
- machine-checked negative/non-vacuity control (Ok and Error differ)

Genuine total proof: no believe_me / postulate / assert_total / %hint.
Wires Dafniser.ABI.FfiSeam into dafniser-abi.ipkg. Builds clean, zero warnings.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A6PSzJWpRxtzGDjUCEh7Mx
Signed-off-by: Jonathan D.A. Jewell <6759885+hyperpolymath@users.noreply.github.com>
@hyperpolymath hyperpolymath marked this pull request as ready for review June 28, 2026 06:43
@hyperpolymath hyperpolymath merged commit 87eae14 into main Jun 28, 2026
14 of 17 checks passed
@hyperpolymath hyperpolymath deleted the claude/new-session-znxgm7 branch June 28, 2026 06:43
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.

2 participants