build: transfer native CI workloads and correct RISC-V ECDSA selection - #110
Merged
Conversation
Centralize CI runner shapes and separate cross-built benchmark preparation from native measurement. Measure x86-64 on Intel and AMD, retain transfer provenance, and bound benchmark invocations to 90 minutes. Require physical Apple Silicon qualification through local commit hooks in place of hosted macOS CI, and configure Markdown linting.
Verify the original source and prepared binary, then retain three full-sample measurements on one pinned CPU with host snapshots. Keep every timing result and preserve failure status across repetitions; this diagnostic mode does not replace full CT qualification.
Keep the comb selection mask opaque to prevent LLVM from replacing the table scan with secret-dependent conditional loads. Add a single-case candidate diagnostic that preserves the original sampling budget and retains failures.
RunsOn resolves public-repository named runners from the default branch. Specify planning and cross-build shapes directly so branch diagnostics do not require those names to exist on main.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Cross-build the IBM and RISC-V CI workloads and source-bound runner tools on x86-64, then execute the transferred workloads on native hardware. Keep native CI in a fail-fast matrix and qualify Apple Silicon locally through mandatory Git hooks.
Fix LLVM's RISC-V ECDSA comb-table optimization by keeping the selection mask opaque. The original binary conditionally loaded coordinates based on the secret digit; the corrected measured binary scans the table with masked selection. Add retained diagnostic replay support and make CT preparation runners resolvable before named definitions reach the default branch.
Validation:
just check,just ct-test,just test-scripts, and workflow validation passed.9a10ef364c398241a28ea2b24360f43e056943f4.73f33d98827d21a309dac11d9e26f65a93d4acd303ff8a8aa3c2446f9aa59c05.This merges the validated baseline for code review. It does not bump versions, create a tag, publish, or claim full release qualification across every architecture. RISC-V BINSEC remains not applicable under the existing manifest policy.