ModernUO/Projects/UOContent/Engines/Pathing/Cache/StepCacheFile.cs
Kamron Batman 7c9215d97c
feat(pathfinding): lazy .swb backing store for the step cache (#2448)
## Summary

Adds a binary disk format + lazy reader so the step cache can warm-start from a precomputed file without paying chunk-build cost on the first pathfind through a region. **Resident memory stays bounded by `MaxResidentChunks` regardless of file size** — opening a `.swb` reads only the header + chunk-offset index (~16 bytes per indexed chunk), and individual chunks are seeked + deserialized only when `ResolveMissingChunk` asks for them.

The lazy design (vs. an eager bulk load): a 250 MB bake on a RAM-constrained shard never materializes more than the LRU cap (~40 MB at the default 8192-chunk cap), and unwanted regions never enter memory at all.

Builds on PR #2447.

## What changed

- **`StepCacheFile`** — binary reader/writer module. Writer emits header → chunks (offsets recorded) → index trailer, then patches the header's `IndexOffset` field. Reader is `OpenForLazy(path)` returning a `LazyReader` that holds an open `FileStream` + offset dictionary.
- **`StepCacheFile.LazyReader`** — `TryReadChunk(chunkX, chunkY)` does a single seek + bulk read for one record. `Dispose` releases the underlying stream. Files are opened with `FileShare.Read | FileShare.Delete` so admin tooling can replace them.
- **TileData fingerprint via XxHash3.** The `.swb` header carries a hash of `LandTable + ItemTable` flags. Load rejects any file whose hash doesn't match the running server. Computed via `HashUtility.ComputeHash64` (engine-blessed hasher) — adds a `ReadOnlySpan<byte>` overload alongside the existing `ReadOnlySpan<char>` one for parity.
- **`StepCache.SaveToFile(path, mapId)`** — writes resident chunks for the given map.
- **`StepCache.TryOpenLazyReader(path, mapId)`** — opens the file, validates header, holds the reader for the map's lifetime.
- **`StepCache.ResolveMissingChunk`** — now consults the lazy reader before invoking the runtime baker. A loaded chunk whose `BuiltMultisVersion` doesn't match the live sector falls through to the baker (snapshot was made before a multi was added/removed in that sector).
- **`StepCache.Clear` closes lazy readers.** Test cleanup can delete `.swb` files cleanly.
- **Auto-load at startup.** `PathCacheCommands.Configure()` opens `Data/Pathfinding/<mapId>.swb` as a lazy reader for every map.
- **`[PathCacheSave`** / **`[PathCacheLoad`** — admin commands for the same workflow.
- **`pathfinding.maxResidentChunks` shard-tunable.** Read from `server.cfg` at boot via `ServerConfiguration.GetOrUpdateSetting` (default 8192 ≈ 40 MB). Small shards can tune down; large shards with substantial bakes can tune up to reduce eviction churn. Default is written back to `server.cfg` on first boot, matching the engine pattern used by other settings.

## File layout (v1)

```
Header (48 bytes):
  u32  Magic           = 0x42575300 ('SWB\0')
  u32  Version         = 1
  u32  MapId
  u64  TileDataHash    XxHash3 over LandTable + ItemTable flags (HashUtility)
  u64  BakeTimestamp   informational
  u32  ChunkCount
  u64  IndexOffset     file position where the chunk index begins

Chunk records (fixed size, ~5,393 bytes each, +32 if multi-Z):
  u16  ChunkX
  u16  ChunkY
  u32  BuiltMultisVersion
  u8   HasMultiZ
  byte WalkMask[256], WetMask[256]
  sbyte SourceZ[256], WalkZN..WalkZNW[256], SwimZN..SwimZNW[256]
  [byte MultiZCells[32] when HasMultiZ == 1]

Index trailer (16 × ChunkCount bytes):
  (u64 chunkKey, u64 fileOffset)
```

## Memory math

| Scenario | Disk file | RAM at boot | Notes |
|---|---|---|---|
| Empty / no `.swb` files | — | 0 | Silent; cache builds on demand. |
| Admin-curated towns (5K chunks) | 25 MB | 0 + per-query | Index ≈ 80 KB. Resident grows to the configured cap under steady-state queries. |
| Full-map bake (50K chunks) | 250 MB | 0 + per-query | Index ≈ 800 KB. Same configured cap. Cold areas never load. |
| All 5 maps fully baked | 1.25 GB | 0 + per-query | Index ≈ 4 MB total. Same configured cap. |

## Hash choice (FNV-1a → XxHash3)

The original draft used inlined FNV-1a-64. Switched to XxHash3 via `HashUtility`:

- ~30× faster on this workload (~30 GB/s SIMD vs ~2 GB/s byte-by-byte). Boot-time only, so absolute saving is microseconds — the real wins are elsewhere.
- Stronger collision resistance and distribution.
- Drops ~25 lines of inlined hash code; matches the rest of the codebase's hashing pattern.
- Hash is stable as long as `HashUtility`'s `xxHash3Seed` constant doesn't change (already marked `// DO NOT CHANGE THIS NUMBER`).
2026-05-06 01:32:44 -07:00

408 lines
15 KiB
C#
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

using System;
using System.Buffers.Binary;
using System.Collections.Generic;
using System.IO;
using System.Runtime.InteropServices;
namespace Server.Engines.Pathing.Cache;
/// <summary>
/// Binary serializer + lazy reader for the step cache. Persists chunk records to disk
/// so a server warm-starts without paying chunk-build cost on the first pathfind through
/// a region. Lazy: opening a file reads only the header + chunk-offset index (~few KB
/// for tens of thousands of chunks), then individual chunks are seeked + deserialized
/// only when the cache asks for them. RAM stays bounded by MaxResidentChunks regardless
/// of file size.
///
/// File layout (little-endian, BufferWriter / BufferReader convention):
///
/// Header (48 bytes):
/// u32 Magic = 0x42575300 ('SWB\0')
/// u32 Version = current FormatVersion
/// u32 MapId
/// u64 TileDataHash XxHash3 over LandTable + ItemTable flags (via
/// HashUtility); rejects a load when client tile data has
/// shifted under us.
/// u64 BakeTimestamp DateTime.UtcNow.Ticks at write time (informational).
/// u32 ChunkCount
/// u64 IndexOffset File position where the chunk index begins.
///
/// Per chunk (ChunkCount times, variable size):
/// u16 ChunkX
/// u16 ChunkY
/// u32 BuiltMultisVersion
/// u8 HasMultiZ 0 = no MultiZCells follow; 1 = 32 bytes of MultiZCells follow
/// byte WalkMask[256]
/// byte WetMask[256]
/// sbyte SourceZ[256]
/// sbyte WalkZN[256]..WalkZNW[256] (8 arrays in N,NE,E,SE,S,SW,W,NW order)
/// sbyte SwimZN[256]..SwimZNW[256] (8 arrays in same order)
/// [byte MultiZCells[32] — only when HasMultiZ == 1]
///
/// Index trailer (16 × ChunkCount bytes):
/// For each chunk: { u64 chunkKey, u64 fileOffset }
///
/// Per-chunk size: ~5,393 bytes (no multi-Z) or ~5,425 bytes (with multi-Z).
/// LRU bookkeeping (LastTouchedTicks) is intentionally not persisted.
/// </summary>
internal static class StepCacheFile
{
public const uint Magic = 0x42575300; // 'SWB\0'
public const uint FormatVersion = 1;
private const int HeaderSize =
sizeof(uint) // Magic
+ sizeof(uint) // Version
+ sizeof(uint) // MapId
+ sizeof(ulong) // TileDataHash
+ sizeof(ulong) // BakeTimestamp
+ sizeof(uint) // ChunkCount
+ sizeof(ulong); // IndexOffset
private const int IndexEntryBytes = sizeof(ulong) + sizeof(ulong); // chunkKey + offset
private const int BytesPerChunkBase =
sizeof(ushort) + sizeof(ushort) + sizeof(uint) + sizeof(byte)
+ StepChunk.CellsPerChunk // WalkMask
+ StepChunk.CellsPerChunk // WetMask
+ StepChunk.CellsPerChunk // SourceZ
+ 8 * StepChunk.CellsPerChunk // WalkZ[8]
+ 8 * StepChunk.CellsPerChunk; // SwimZ[8]
private const int BytesPerMultiZ = 32;
/// <summary>
/// Byte offset of the IndexOffset u64 within the header
/// (Magic+Version+MapId+TileDataHash+BakeTimestamp+ChunkCount = 32). Patched after chunks land.
/// </summary>
private const int IndexOffsetFieldPosition = 32;
public delegate bool ChunkEnumerator(out int chunkX, out int chunkY, out StepChunk chunk);
/// <summary>
/// Computes a stable hash of the loaded TileData flags via XxHash3 (HashUtility).
/// Bake files carry this hash so a load can refuse to populate the cache when tile
/// data has shifted (client patch, mismatched version) — mismatched data would
/// silently skew walkability answers. Hash is stable as long as HashUtility's seed
/// constant doesn't change.
/// </summary>
public static ulong ComputeTileDataHash()
{
var landTable = TileData.LandTable;
var itemTable = TileData.ItemTable;
// Project just the Flags ulong from each entry into a contiguous byte buffer.
// The struct itself contains a string Name (reference) whose object identity isn't
// stable across runs, so we can't MemoryMarshal.Cast the whole struct.
var bytes = new byte[(landTable.Length + itemTable.Length) * sizeof(ulong)];
var span = bytes.AsSpan();
for (var i = 0; i < landTable.Length; i++)
{
BinaryPrimitives.WriteUInt64LittleEndian(span[(i * 8)..], (ulong)landTable[i].Flags);
}
var itemOffset = landTable.Length * 8;
for (var i = 0; i < itemTable.Length; i++)
{
BinaryPrimitives.WriteUInt64LittleEndian(span[(itemOffset + i * 8)..], (ulong)itemTable[i].Flags);
}
return HashUtility.ComputeHash64(bytes);
}
/// <summary>
/// Writes the file: header (with placeholder IndexOffset) → chunks (offsets recorded)
/// → index trailer → patches the header IndexOffset. <paramref name="chunkCount"/> must
/// equal the actual number of chunks <paramref name="next"/> will yield.
/// </summary>
public static void Write(string path, uint mapId, uint chunkCount, ChunkEnumerator next)
{
Directory.CreateDirectory(Path.GetDirectoryName(path) ?? ".");
var capacity = HeaderSize
+ (BytesPerChunkBase + BytesPerMultiZ) * (int)chunkCount
+ IndexEntryBytes * (int)chunkCount;
var buffer = new byte[capacity];
var w = new BufferWriter(buffer, prefixStr: false);
w.Write(Magic);
w.Write(FormatVersion);
w.Write(mapId);
w.Write(ComputeTileDataHash());
w.Write((ulong)DateTime.UtcNow.Ticks);
w.Write(chunkCount);
w.Write(0UL); // IndexOffset placeholder, patched after chunks
var indexEntries = new (ulong key, ulong offset)[chunkCount];
var written = 0u;
while (next(out var chunkX, out var chunkY, out var chunk))
{
if (written >= chunkCount)
{
throw new InvalidOperationException(
$"StepCacheFile.Write: enumerator yielded more than the declared {chunkCount} chunks"
);
}
var chunkOffset = (ulong)w.Position;
WriteChunk(w, chunkX, chunkY, chunk);
indexEntries[written] = (PackChunkKey(chunkX, chunkY), chunkOffset);
written++;
}
if (written != chunkCount)
{
throw new InvalidOperationException(
$"StepCacheFile.Write: declared {chunkCount} chunks but enumerator yielded {written}"
);
}
var indexOffset = (ulong)w.Position;
for (var i = 0u; i < chunkCount; i++)
{
w.Write(indexEntries[i].key);
w.Write(indexEntries[i].offset);
}
// Patch IndexOffset directly into the buffer (BufferWriter has no Seek).
BinaryPrimitives.WriteUInt64LittleEndian(buffer.AsSpan(IndexOffsetFieldPosition, 8), indexOffset);
var totalBytes = (int)w.Position;
File.WriteAllBytes(path, buffer.AsSpan(0, totalBytes).ToArray());
}
/// <summary>
/// Opens a .swb file and reads only its header + chunk-offset index. Returns null on
/// missing file, magic / version mismatch, or TileDataHash mismatch (a stale bake
/// against a freshly patched client). Callers own disposal of the returned reader.
/// </summary>
public static LazyReader OpenForLazy(string path)
{
if (!File.Exists(path))
{
return null;
}
FileStream stream = null;
try
{
stream = new FileStream(
path,
FileMode.Open,
FileAccess.Read,
FileShare.Read | FileShare.Delete
);
Span<byte> headerBuf = stackalloc byte[HeaderSize];
if (stream.Read(headerBuf) != HeaderSize)
{
stream.Dispose();
return null;
}
var magic = BinaryPrimitives.ReadUInt32LittleEndian(headerBuf);
if (magic != Magic)
{
stream.Dispose();
return null;
}
var version = BinaryPrimitives.ReadUInt32LittleEndian(headerBuf[4..]);
if (version != FormatVersion)
{
stream.Dispose();
return null;
}
var mapId = BinaryPrimitives.ReadUInt32LittleEndian(headerBuf[8..]);
var tileDataHash = BinaryPrimitives.ReadUInt64LittleEndian(headerBuf[12..]);
var bakeTimestamp = BinaryPrimitives.ReadUInt64LittleEndian(headerBuf[20..]);
var chunkCount = BinaryPrimitives.ReadUInt32LittleEndian(headerBuf[28..]);
var indexOffset = BinaryPrimitives.ReadUInt64LittleEndian(headerBuf[32..]);
if (tileDataHash != ComputeTileDataHash())
{
stream.Dispose();
return null;
}
// Read the chunk-offset index in one shot.
var indexBytes = (int)chunkCount * IndexEntryBytes;
var indexBuf = new byte[indexBytes];
stream.Position = (long)indexOffset;
if (stream.Read(indexBuf, 0, indexBytes) != indexBytes)
{
stream.Dispose();
return null;
}
var offsets = new Dictionary<ulong, ulong>((int)chunkCount);
for (var i = 0; i < chunkCount; i++)
{
var entry = indexBuf.AsSpan(i * IndexEntryBytes);
var key = BinaryPrimitives.ReadUInt64LittleEndian(entry);
var off = BinaryPrimitives.ReadUInt64LittleEndian(entry[8..]);
offsets[key] = off;
}
return new LazyReader(stream, mapId, tileDataHash, bakeTimestamp, chunkCount, offsets);
}
catch
{
stream?.Dispose();
return null;
}
}
private static ulong PackChunkKey(int chunkX, int chunkY) => ((ulong)(uint)chunkX << 32) | (uint)chunkY;
private static void WriteChunk(BufferWriter w, int chunkX, int chunkY, StepChunk chunk)
{
w.Write((ushort)chunkX);
w.Write((ushort)chunkY);
w.Write((uint)chunk.BuiltMultisVersion);
var multiZ = chunk.GetMultiZCellsForSerialization();
w.Write((byte)(multiZ != null ? 1 : 0));
w.Write(chunk.WalkMask);
w.Write(chunk.WetMask);
WriteSBytes(w, chunk.SourceZ);
WriteSBytes(w, chunk.WalkZN);
WriteSBytes(w, chunk.WalkZNE);
WriteSBytes(w, chunk.WalkZE);
WriteSBytes(w, chunk.WalkZSE);
WriteSBytes(w, chunk.WalkZS);
WriteSBytes(w, chunk.WalkZSW);
WriteSBytes(w, chunk.WalkZW);
WriteSBytes(w, chunk.WalkZNW);
WriteSBytes(w, chunk.SwimZN);
WriteSBytes(w, chunk.SwimZNE);
WriteSBytes(w, chunk.SwimZE);
WriteSBytes(w, chunk.SwimZSE);
WriteSBytes(w, chunk.SwimZS);
WriteSBytes(w, chunk.SwimZSW);
WriteSBytes(w, chunk.SwimZW);
WriteSBytes(w, chunk.SwimZNW);
if (multiZ != null)
{
w.Write(multiZ);
}
}
private static StepChunk ReadChunk(byte[] buffer)
{
var r = new BufferReader(buffer);
// Skip ChunkX + ChunkY (already known via the index lookup).
r.ReadUShort();
r.ReadUShort();
var multisVersion = (int)r.ReadUInt();
var hasMultiZ = r.ReadByte() != 0;
var chunk = new StepChunk { BuiltMultisVersion = multisVersion };
r.Read(chunk.WalkMask);
r.Read(chunk.WetMask);
ReadSBytes(r, chunk.SourceZ);
ReadSBytes(r, chunk.WalkZN);
ReadSBytes(r, chunk.WalkZNE);
ReadSBytes(r, chunk.WalkZE);
ReadSBytes(r, chunk.WalkZSE);
ReadSBytes(r, chunk.WalkZS);
ReadSBytes(r, chunk.WalkZSW);
ReadSBytes(r, chunk.WalkZW);
ReadSBytes(r, chunk.WalkZNW);
ReadSBytes(r, chunk.SwimZN);
ReadSBytes(r, chunk.SwimZNE);
ReadSBytes(r, chunk.SwimZE);
ReadSBytes(r, chunk.SwimZSE);
ReadSBytes(r, chunk.SwimZS);
ReadSBytes(r, chunk.SwimZSW);
ReadSBytes(r, chunk.SwimZW);
ReadSBytes(r, chunk.SwimZNW);
if (hasMultiZ)
{
var multiZ = new byte[BytesPerMultiZ];
r.Read(multiZ);
chunk.RestoreMultiZCellsFromSerialization(multiZ);
}
return chunk;
}
private static void WriteSBytes(BufferWriter w, sbyte[] arr) =>
w.Write(MemoryMarshal.Cast<sbyte, byte>(arr.AsSpan()));
private static void ReadSBytes(BufferReader r, sbyte[] arr) =>
r.Read(MemoryMarshal.Cast<sbyte, byte>(arr.AsSpan()));
/// <summary>
/// Open handle on a .swb file. Holds the FileStream + chunk-offset index. Chunks are
/// fetched on demand via <see cref="TryReadChunk"/>; only the records actually queried
/// are ever materialized. Dispose releases the underlying stream.
/// </summary>
internal sealed class LazyReader : IDisposable
{
private FileStream _stream;
private readonly Dictionary<ulong, ulong> _offsets;
private byte[] _buffer;
public uint MapId { get; }
public ulong TileDataHash { get; }
public ulong BakeTimestamp { get; }
public uint ChunkCount { get; }
public int IndexedChunkCount => _offsets.Count;
public bool Has(int chunkX, int chunkY) => _offsets.ContainsKey(PackChunkKey(chunkX, chunkY));
internal LazyReader(
FileStream stream, uint mapId, ulong tileDataHash, ulong bakeTimestamp,
uint chunkCount, Dictionary<ulong, ulong> offsets
)
{
_stream = stream;
MapId = mapId;
TileDataHash = tileDataHash;
BakeTimestamp = bakeTimestamp;
ChunkCount = chunkCount;
_offsets = offsets;
_buffer = new byte[BytesPerChunkBase + BytesPerMultiZ];
}
/// <summary>
/// Returns the chunk record at (<paramref name="chunkX"/>, <paramref name="chunkY"/>)
/// from the file, or null if the file doesn't contain it. Single seek + bulk read;
/// no allocations beyond the returned StepChunk and its arrays.
/// </summary>
public StepChunk TryReadChunk(int chunkX, int chunkY)
{
if (_stream == null)
{
return null;
}
var key = PackChunkKey(chunkX, chunkY);
if (!_offsets.TryGetValue(key, out var offset))
{
return null;
}
_stream.Position = (long)offset;
// Try to read the maximum size; the file may have less remaining, which is OK
// since BufferReader stops at the bytes it actually needs.
var read = _stream.Read(_buffer, 0, _buffer.Length);
return read < BytesPerChunkBase ? null : ReadChunk(_buffer);
}
public void Dispose()
{
_stream?.Dispose();
_stream = null;
_buffer = null;
}
}
}