2 Commits

Author	SHA1	Message	Date
Evan	5e4b2791aa	perf: reduce core-sim GC churn 42% and add GC-churn profiling to the perf harness (#4494) ... ## Summary Reduces core-simulation GC churn by **42%** on a 20-game-minute Giant World Map run, and extends the headless full-game perf harness so churn is measurable and regressions are visible. ### 1. GC-churn measurement (`tests/perf/fullgame/GcProfiler.ts`) `npm run perf:game` now reports: - **GC pauses** by kind (minor/major/incremental) via a `PerformanceObserver` on `'gc'` entries, bucketed into tick windows by timestamp (V8 only delivers these entries on a timer task, so they're flushed after the run) - **Allocation rate** per `--window N` ticks (default 1000) from used-heap deltas sampled every tick, so churn can be tracked across game phases - **Top allocating functions** from the V8 sampling heap profiler with `includeObjectsCollectedBy{Major,Minor}GC` — i.e. actual churn including short-lived garbage, not live memory — plus a `.heapprofile` loadable in Chrome DevTools (Memory → Allocation sampling) New flags: `--window N`, `--no-gc-profile`, `--no-alloc-profile`. ### 2. Allocation reductions in the hot paths it found | Site | Change | |---|---| | `GameMap.bfs` | inline neighbor enumeration instead of an array per visited tile | | `GameMap`/`Game` | new `forEachNeighborNSWE` — allocation-free iterator matching `neighbors()` N,S,W,E order for order-sensitive callers (`forEachNeighbor` visits W,E,N,S, so substituting it would change sim behavior) | | `PlayerImpl.nearby` / `sharesBorderWith` / `shoreReachableNeighbors` | no per-call neighbor arrays; no materialized shore-tile array | | `PlayerImpl.units(types)` | gather into a reusable scratch buffer, return one exact-size slice (still a fresh snapshot array per call) | | `AiAttackBehavior.maybeAttack` | single pass over border neighbors replacing the `flatMap`/`filter`/`map` chain over every border tile | | `AiAttackBehavior.isBorderingNukedTerritory` | reusable `neighbors4` buffer with early exit | | `SharedWaterCache.build` | allocation-free neighbor iteration | | `SpatialQuery.bfsNearest` | first-minimum scan instead of collect-then-stable-sort (identical result incl. tie-breaking) | ### Results (Giant World Map, 400 bots, 12,000 ticks ≈ 20 game-minutes, seed `perf-default`) | Metric | Before | After | |---|---|---| | Sampled allocations (incl. collected) | 97.7 GB | **56.9 GB (−42%)** | | GC count / total pause | 1,682 / 3,313 ms (1.8% of wall) | 1,058 / 2,087 ms (1.2%) | | Ticks/sec | 66 | 70 | | p99 / max tick | 49.9 ms / 988 ms | 43.5 ms / 689 ms | | Ticks over 100 ms budget | 31 | 19 | ## Determinism Every rewrite preserves exact iteration order (the new NSWE iterator exists precisely for the order-sensitive sites). Verified by identical final game-state hashes on three runs: Giant World Map 12,000 ticks (`67286276735690560`), Giant World Map 2,000 ticks, and World 1,800 ticks. ## Test plan - [x] Full suite green (1,896 tests) - [x] New tests: `forEachNeighborNSWE` order contract vs `neighbors()` over every tile; `units()` filtering semantics (insertion order, fresh-array guarantee, duplicate types, Set path) - [x] Final-hash equality on 3 seeded headless runs (2 maps) 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com>	2026-07-03 12:30:28 -07:00
Evan	2789db8b96	Optimize core simulation hot paths (no behavior change) (#4230) ... ## Summary Pure performance optimizations to the attack/conquer/cluster hot paths in `src/core`, driven by the full-game perf harness from #4228. **No behavior change**: the final game-state hash is identical before/after on every config tested — world quick run (2 different seeds), giantworldmap, and the default 1800-tick run. ### Changes - **Flat-arithmetic neighbor iteration**: `forEachNeighbor` / `forEachNeighborWithDiag` / `isBorder` / `isOceanShore` are now implemented inside `GameMapImpl` using raw `ref±1` / `ref±width` index math, skipping the per-neighbor `ref()` coordinate validation (`Number.isInteger` etc.). `GameImpl` and `GameView` delegate. - **New `neighbors4(ref, out)`**: zero-allocation, callback-free neighbor query for hot loops (W, E, N, S — same order as `forEachNeighbor`). - **`AttackExecution`**: the per-tile closures in `tick()` / `addNeighbors()` are replaced with reusable neighbor buffers, a cached `GameMap` reference, and integer `smallID()` owner comparisons instead of owner-object lookups. - **`GameImpl`**: the per-conquer `updateBorders` closure is hoisted to a method with a reusable buffer; `removeInactiveExecutions` compacts the executions array in place instead of allocating a new ~4200-element array every tick. - **`PlayerExecution`**: `surroundedBySamePlayer` / `isSurrounded` / `getCapturingPlayer` de-closured (`neighbors4` + integer compares; neighbor visit order preserved, so `getCapturingPlayer`'s Map-insertion-order tie-breaking is unchanged); flood-fill visit closure hoisted out of the while loop. - **`FlatBinaryHeap.dequeue`**: returns the tile directly instead of allocating a `[tile, priority]` tuple per dequeued tile (AttackExecution is the only caller). ### Performance (`npm run perf:game`, same machine, before → after) | run | mean tick | ticks/sec | max tick | |---|---|---|---| | default (world, 400 bots, 1800 ticks) | 9.04 → **7.98 ms** | 111 → **125** | 31.7 → 35.7 ms | | giantworldmap, 600 ticks | 22.5 → **17.4 ms** | 44 → **58** | 52.8 → **36.2 ms** | The giantworldmap tail improvement (max tick −31%) is the most relevant for the 100 ms tick budget. ### Determinism verification Identical `Final hash` before and after on all configs: | config | hash | |---|---| | `--map world --ticks 200 --bots 100` | `5455008589403520` | | same + `--seed second-seed-check` | `5580840142777488` | | `--map giantworldmap --ticks 600` | `37373734953428430` | | default run | `26773450321979388` | ### Tests - New `tests/NeighborIteration.test.ts` pins the exact neighbor iteration orders (W,E,N,S cardinal; dx-major diagonal — conquest order and RNG consumption depend on them) and conquer/border-tile invariants checked mid-battle. - New `tests/FlatBinaryHeap.test.ts` covers heap ordering, clear, and growth. - Full suite passes (122 files / 1386 tests + server tests); lint and prettier clean. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Claude Fable 5 <noreply@anthropic.com>	2026-06-11 19:58:42 -07:00