Repository navigation
HOL-Light: Avoid repeated work in proof loops - #1962
Merged
Merged
Conversation
Instruction loops apply the same step and SIMD tactics after each instruction. Constructing them inside callbacks rebuilds rewrite conversions and helper predicates for every step, making long arithmetic proofs spend time preparing unchanged work. Prepare ARM/x86 step tactics and SIMD simplifiers once outside all matching ML-KEM loops. In SIMD_SIMPLIFY_ABBREV_TAC, also prepare the conversion and local-definition predicate once while still reading arithmetic equations from the current proof context. After NTT congruences are reduced to real polynomials, REAL_INTEGER_TAC needs no execution assumptions. Drop those assumptions before the four AArch64/x86_64 NTT and inverse NTT closure calls. The prepared conversions capture their rewrite rules when the tactic is built. Rebuild the tactic if those rules change. Signed-off-by: Hanno Becker <beckphan@amazon.co.uk>
Contributor
CBMC Results (ML-KEM-768)
Full Results (379 configurations)
|
Contributor
CBMC Results (ML-KEM-1024)
Full Results (379 configurations)
|
Contributor
CBMC Results (ML-KEM-512)
Full Results (379 configurations)
|
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Instruction loops apply the same step and SIMD tactics after each instruction. Constructing them inside callbacks rebuilds rewrite conversions and helper predicates for every step, making long arithmetic proofs spend time preparing unchanged work.
Prepare ARM/x86 step tactics and SIMD simplifiers once outside all matching ML-KEM loops. In SIMD_SIMPLIFY_ABBREV_TAC, also prepare the conversion and local-definition predicate once while still reading arithmetic equations from the current proof context.
After NTT congruences are reduced to real polynomials, REAL_INTEGER_TAC needs no execution assumptions. Drop those assumptions before the four AArch64/x86_64 NTT and inverse NTT closure calls.
The prepared conversions capture their rewrite rules when the tactic is built. Rebuild the tactic if those rules change.