starknet_os: proper base measurement in OS resources test

[dorimedini-starkware]

No description provided.

[reviewable-StarkWare]

This change is 

[dorimedini-starkware]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• starknet_os: proper base measurement in OS resources test #14294 👈 (View in Graphite)
• starknet_os: os resources test - add sha512 #14256
• starknet_os: generalize sha resource test to include sha512 #14255
• starknet_os: os resources test - add secp syscalls #14144 : 1 other dependent PR (#14145 )
• starknet_os: os resources test - add send_message_to_l1 #14143 : 1 other dependent PR (#14153 )
• starknet_os: os resources test - add replace_class #14142
• starknet_os: os resources test - support virtual builtins, add sha256 #14141
• starknet_os: define sha256 batch resources constants #14140
• starknet_os: os resources test - add keccak #14139
• starknet_os: os resources test - add 3 more syscalls #14138
• starknet_os: os resources test - remove all fee transfer syscalls #14137
• starknet_os: os resources test - add emit event #14136
• starknet_os: os resources test - add meta tx #14135
• starknet_os: os resources test - measure linear factor, add deploy #14134
• starknet_os: os resources test - library call #14129
• starknet_os: initial test, measure call contract resources #14128
• blockifier_test_utils: add option to deploy from zero in test accounts #14258
• starknet_os_flow_tests: helper function to declare explicit classes #14257
• starknet_os: os resources test - create fixed deployable account #14173
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[github-actions]

Artifacts upload workflows:

• Started at 2026-06-02T10:22:33Z

[cursor]

PR Summary

Low Risk
Test methodology and fee/resource constant JSON only; no runtime execution path changes beyond aligned syscall cost tables.

Overview
Fixes how the OS resources regression test derives constant (base) costs for syscalls with a calldata linear factor (Deploy, MetaTxV0, and the Keccak path).

The test now tracks how many linear elements are baked into the first paired measurement (e.g. deploy calldata length) and subtracts linear_factor × count from that trace before recording the base cost. Previously it treated the first syscall trace as the base even when it still included per-element overhead.

Versioned constants for 0.14.4 are refreshed to match the corrected measurements: lower Deploy / MetaTxV0 constant steps and Pedersen counts, and a higher Deploy per-calldata step factor.

^{Reviewed by Cursor Bugbot for commit c8fdcf3. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit c8fdcf3. Configure here.}

[cursor]

const with LazyLock defeats lazy caching purpose

Low Severity

SYSCALLS_WITH_LINEAR_FACTOR is declared as const but wraps a LazyLock<HashMap<...>>. In Rust, const items are inlined at each use site, so every reference to this identifier creates a brand-new LazyLock and re-allocates the HashMap — completely defeating the lazy-initialization caching that LazyLock provides. Since it's referenced inside a while loop (at line 436), a fresh HashMap is built on every iteration. Every other LazyLock in this crate correctly uses static. This is the well-known declare_interior_mutable_const anti-pattern; changing const to static fixes it.

 

^{Reviewed by Cursor Bugbot for commit c8fdcf3. Configure here.}