[move] Refactor type caches in loader

[georgemitenkov]

Description

How Has This Been Tested?

Key Areas to Review

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
• Move Compiler
• Other (specify)

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]

⏱️ 12h 57m total CI duration on this PR

Slowest 15 Jobs	Cumulative Duration	Recent Runs
single-node-performance	8h 31m	🟥 🟥 🟥 🟥 🟥 (+15 more)
test-target-determinator	1h 37m	🟩 🟩 🟩 🟩 🟩 (+15 more)
execution-performance / single-node-performance	47m	🟩 🟩
check-dynamic-deps	33m	🟩 🟩 🟩 🟩 🟩 (+8 more)
rust-targeted-unit-tests	25m	🟩
rust-cargo-deny	15m	🟩 🟩 🟩 🟩 🟩 (+5 more)
execution-performance / test-target-determinator	9m	🟩 🟩
test-target-determinator	9m	🟩 🟩
rust-doc-tests	7m	🟩
rust-doc-tests	7m	🟩
semgrep/ci	5m	🟩 🟩 🟩 🟩 🟩 (+8 more)
general-lints	4m	🟩 🟩 🟩 🟩 🟩 (+5 more)
fetch-last-released-docker-image-tag	4m	🟩 🟩
file_change_determinator	2m	🟩 🟩 🟩 🟩 🟩 (+7 more)
permission-check	44s	🟩 🟩 🟩 🟩 🟩 (+8 more)

🚨 1 job on the last run was significantly faster/slower than expected

Job	Duration	vs 7d avg	Delta
test-target-determinator	4m	5m

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

[georgemitenkov]

• [move] Refactor type caches in loader #15818 👈 (View in Graphite)
• [move] Avoid double hashing in favour of equivalent keys for type tag cache #15740
• [aptos-vm][move] Avoid module loads when getting the struct name #15681
• main

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

[runtian-zhou]

Seems that this results in a performance regression? Do we have plans for mitigating that?

[runtian-zhou]

Changes looks good to me. Was mostly wondering on how we are planning to work on the perf issues.

[github-actions]

✅ Forge suite realistic_env_max_load success on 38246b8d21a43830a7cc9b4fc3920f262714fff5

two traffics test: inner traffic : committed: 12048.50 txn/s, submitted: 12050.67 txn/s, expired: 2.18 txn/s, latency: 3276.56 ms, (p50: 2700 ms, p70: 2700, p90: 3000 ms, p99: 27700 ms), latency samples: 4581132
two traffics test : committed: 99.96 txn/s, latency: 1561.12 ms, (p50: 1500 ms, p70: 1600, p90: 1700 ms, p99: 2000 ms), latency samples: 1780
Latency breakdown for phase 0: ["MempoolToBlockCreation: max: 1.988, avg: 1.752", "ConsensusProposalToOrdered: max: 0.345, avg: 0.330", "ConsensusOrderedToCommit: max: 0.545, avg: 0.515", "ConsensusProposalToCommit: max: 0.876, avg: 0.845"]
Max non-epoch-change gap was: 0 rounds at version 0 (avg 0.00) [limit 4], 1.31s no progress at version 19421 (avg 0.23s) [limit 15].
Max epoch-change gap was: 0 rounds at version 0 (avg 0.00) [limit 4], 0.64s no progress at version 1659989 (avg 0.60s) [limit 16].
Test Ok

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

[github-actions]

✅ Forge suite compat success on 60f7ca8827f5d64a148c3b163dc4126b0879279b ==> 38246b8d21a43830a7cc9b4fc3920f262714fff5

Compatibility test results for 60f7ca8827f5d64a148c3b163dc4126b0879279b ==> 38246b8d21a43830a7cc9b4fc3920f262714fff5 (PR)
1. Check liveness of validators at old version: 60f7ca8827f5d64a148c3b163dc4126b0879279b
compatibility::simple-validator-upgrade::liveness-check : committed: 9815.78 txn/s, latency: 3290.37 ms, (p50: 2700 ms, p70: 3500, p90: 5400 ms, p99: 6300 ms), latency samples: 334380
2. Upgrading first Validator to new version: 38246b8d21a43830a7cc9b4fc3920f262714fff5
compatibility::simple-validator-upgrade::single-validator-upgrading : committed: 3214.45 txn/s, latency: 9051.43 ms, (p50: 9700 ms, p70: 10900, p90: 11900 ms, p99: 12000 ms), latency samples: 70000
compatibility::simple-validator-upgrade::single-validator-upgrade : committed: 3201.06 txn/s, latency: 10759.63 ms, (p50: 11600 ms, p70: 12400, p90: 12400 ms, p99: 12500 ms), latency samples: 125880
3. Upgrading rest of first batch to new version: 38246b8d21a43830a7cc9b4fc3920f262714fff5
compatibility::simple-validator-upgrade::half-validator-upgrading : committed: 1053.78 txn/s, submitted: 1053.96 txn/s, expired: 0.18 txn/s, latency: 9947.06 ms, (p50: 10700 ms, p70: 12200, p90: 12800 ms, p99: 13000 ms), latency samples: 64629
compatibility::simple-validator-upgrade::half-validator-upgrade : committed: 2980.17 txn/s, latency: 11247.77 ms, (p50: 12400 ms, p70: 13300, p90: 13400 ms, p99: 13400 ms), latency samples: 112640
4. upgrading second batch to new version: 38246b8d21a43830a7cc9b4fc3920f262714fff5
compatibility::simple-validator-upgrade::rest-validator-upgrading : committed: 5263.73 txn/s, latency: 5697.49 ms, (p50: 6400 ms, p70: 6600, p90: 7800 ms, p99: 8000 ms), latency samples: 104500
compatibility::simple-validator-upgrade::rest-validator-upgrade : committed: 5148.89 txn/s, latency: 6540.37 ms, (p50: 6500 ms, p70: 7600, p90: 8000 ms, p99: 8300 ms), latency samples: 186160
5. check swarm health
Compatibility test for 60f7ca8827f5d64a148c3b163dc4126b0879279b ==> 38246b8d21a43830a7cc9b4fc3920f262714fff5 passed
Test Ok

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

[github-actions]

✅ Forge suite framework_upgrade success on 60f7ca8827f5d64a148c3b163dc4126b0879279b ==> 38246b8d21a43830a7cc9b4fc3920f262714fff5

Compatibility test results for 60f7ca8827f5d64a148c3b163dc4126b0879279b ==> 38246b8d21a43830a7cc9b4fc3920f262714fff5 (PR)
Upgrade the nodes to version: 38246b8d21a43830a7cc9b4fc3920f262714fff5
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 1500.17 txn/s, submitted: 1506.16 txn/s, failed submission: 5.99 txn/s, expired: 5.99 txn/s, latency: 2031.35 ms, (p50: 1800 ms, p70: 2200, p90: 3300 ms, p99: 4400 ms), latency samples: 135302
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 1442.32 txn/s, submitted: 1448.07 txn/s, failed submission: 5.75 txn/s, expired: 5.75 txn/s, latency: 2516.68 ms, (p50: 1800 ms, p70: 2100, p90: 3600 ms, p99: 12600 ms), latency samples: 125400
5. check swarm health
Compatibility test for 60f7ca8827f5d64a148c3b163dc4126b0879279b ==> 38246b8d21a43830a7cc9b4fc3920f262714fff5 passed
Upgrade the remaining nodes to version: 38246b8d21a43830a7cc9b4fc3920f262714fff5
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 1502.64 txn/s, submitted: 1505.81 txn/s, failed submission: 3.17 txn/s, expired: 3.17 txn/s, latency: 1933.48 ms, (p50: 1700 ms, p70: 2100, p90: 3100 ms, p99: 4300 ms), latency samples: 132540
Test Ok

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