[execution window] use blocking txn provider

[bchocho]

Description

Optimizes the prepare and execute stages for better performance for larger block windows. Mostly the optimizations are reducing the computation per transaction or number of transactions that have to be touched.

Prepare:

• Return an Arc of cached batches, and lazily flatten the block to vector of transactions to avoid unnecessary clones.
• Re-sort the batches by gas bucket (as a stable sort, to maintain relative order).
• Filter out already committed transactions. Uses the cached per-pipelined block of committed hashes. Except for the most recent block in the window.
• Move block shuffle to execute stage.

Execute:

• Filter out already committed transactions from the latest recent block in the window.
• Do block shuffle using a BlockingTxnProvider; the shuffler incrementally writes shuffled transactions and execution reads the shuffle transactions (blocking if not yet available).

Type of Change

• New feature
• Bug fix
• Breaking change
• Performance improvement
• Refactoring
• Dependency update
• Documentation update
• Tests

Which Components or Systems Does This Change Impact?

• Validator Node
• Full Node (API, Indexer, etc.)
• Move/Aptos Virtual Machine
• Aptos Framework
• Aptos CLI/SDK
• Developer Infrastructure
• Other (specify)

How Has This Been Tested?

Key Areas to Review

Checklist

• I have read and followed the CONTRIBUTING doc
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I identified and added all stakeholders and component owners affected by this change as reviewers
• I tested both happy and unhappy path of the functionality
• I have made corresponding changes to the documentation

[trunk-io]

⏱️ 5h 28m total CI duration on this PR

Slowest 15 Jobs	Cumulative Duration	Recent Runs
forge-e2e-test / forge	3h 50m	🟥 ⬜ 🟥 🟥 🟥 (+2 more)
test-target-determinator	26m	🟩 🟩 🟩 🟩
rust-images / rust-all	14m	🟩
rust-move-tests	8m	🟩
rust-move-tests	8m	🟥
rust-cargo-deny	8m	🟩 🟩 🟩 🟩
rust-move-tests	8m	🟩
general-lints	7m	🟩 🟩 🟩 🟩
check-dynamic-deps	7m	🟩 🟩 🟩 🟩 🟩 (+2 more)
rust-move-tests	5m	🟥
semgrep/ci	2m	🟩 🟩 🟩 🟩 🟩
file_change_determinator	53s	⬜ 🟩 🟩 🟩 🟩
file_change_determinator	48s	🟩 🟩 🟩 🟩
file_change_determinator	46s	🟩 🟩 🟩 🟩
permission-check	19s	🟩 🟩 🟩 🟩 🟩

_{settings ⋅ feedback ⋅ docs ⋅ learn more about trunk.io}

[github-actions]

✅ Forge suite realistic_env_max_load success on 20a72fee519e5fd2e8ba452fba7081abaae49407

two traffics test: inner traffic : committed: 14146.62 txn/s, latency: 2811.52 ms, (p50: 2700 ms, p70: 2700, p90: 3000 ms, p99: 3500 ms), latency samples: 5378880
two traffics test : committed: 100.02 txn/s, latency: 2477.10 ms, (p50: 2300 ms, p70: 2400, p90: 2700 ms, p99: 9600 ms), latency samples: 1800
Latency breakdown for phase 0: ["MempoolToBlockCreation: max: 1.158, avg: 0.755", "ConsensusProposalToOrdered: max: 0.318, avg: 0.298", "ConsensusOrderedToCommit: max: 0.414, avg: 0.402", "ConsensusProposalToCommit: max: 0.713, avg: 0.699"]
Max non-epoch-change gap was: 0 rounds at version 0 (avg 0.00) [limit 4], 0.89s no progress at version 2763041 (avg 0.21s) [limit 15].
Max epoch-change gap was: 0 rounds at version 0 (avg 0.00) [limit 4], 8.04s no progress at version 2763039 (avg 8.04s) [limit 15].
Test Ok

• Grafana dashboard
• Humio Logs
• Axiom Logs
• Validator CPU Profile
• Fullnode CPU Profile
• Test runner output
• Test run is land-blocking