8750186e55
Phase 26-H 完了内容: ✅ JoinIR 型定義実装（src/mir/join_ir.rs） ✅ MIR → JoinIR 自動変換実装（lower_min_loop_to_joinir） ✅ 自動変換テスト実装（mir_joinir_min_auto_lowering） ✅ PHI/Loop箱 → JoinIR 移行対応表追加（loopform_ssot.md） ドキュメント更新: - Phase 27 JoinIR タスク計画追加 - Phase 26-H タスク完了記録 - 各種 README 更新（進捗反映） - CURRENT_TASK.md 更新 コミット統計: $(git status --short | wc -l) files changed 次のステップ: Phase 27 一般化 MIR → JoinIR 変換
106 lines
6.7 KiB
Markdown
106 lines
6.7 KiB
Markdown
# Phase 21.8 — Numeric Core Integration & Builder Support
|
||
|
||
Status: partially completed（builder/imports導線まで・EXEベンチはPhase 25へ移管）
|
||
|
||
## Goal
|
||
|
||
Integrate the new numeric core boxes (IntArrayCore + MatI64) into the Hakorune selfhost chain so that:
|
||
|
||
- Stage‑B → MirBuilder → ny‑llvmc(crate) can emit MIR(JSON) for code that uses:
|
||
- `using nyash.core.numeric.intarray as IntArrayCore`
|
||
- `using nyash.core.numeric.matrix_i64 as MatI64`
|
||
- The `matmul_core` microbench（MatI64 + IntArrayCore）については、**構造的な導線（builder/imports）まで本フェーズで整備し、EXE/LLVM ベンチ統合は Phase 25 に移管する**。
|
||
|
||
21.6 provides the core boxes; 21.8 focuses on wiring them into the builder/runtime chain（Stage‑B/Bridge/MirBuilder）without changing default behaviour for other code. LLVM での数値コア実行経路は Phase 25（Ring0/Ring1 再編・numeric runtime AOT）で扱う。
|
||
|
||
## Scope (21.8, this host)
|
||
|
||
- Stage‑B / MirBuilder:
|
||
- Ensure `MatI64` and `IntArrayCore` are recognized as valid boxes when referenced via:
|
||
- `using nyash.core.numeric.matrix_i64 as MatI64`
|
||
- `using nyash.core.numeric.intarray as IntArrayCore`
|
||
- Fix the current provider‑emit failure:
|
||
- Error today: `[mirbuilder/parse/error] undefined variable: MatI64` during `env.mirbuilder.emit`.
|
||
- Diagnose and adjust Stage‑B / MirBuilder so that static box references (`MatI64.new`, `A.mul_naive`) compile in the same way as other boxes.
|
||
|
||
- AotPrep / emit pipeline:
|
||
- Keep AotPrep unchanged for now; the goal is to make `tools/hakorune_emit_mir.sh` succeed on `matmul_core` sources without special‑casing.
|
||
- Ensure `tools/hakorune_emit_mir.sh` with:
|
||
- `HAKO_APPLY_AOT_PREP=1 NYASH_AOT_COLLECTIONS_HOT=1 NYASH_LLVM_FAST=1 NYASH_MIR_LOOP_HOIST=1`
|
||
- can emit valid MIR(JSON) for MatI64/IntArrayCore code.
|
||
|
||
- Microbench integration:
|
||
- Finish wiring `matmul_core` in `tools/perf/microbench.sh`:
|
||
- Hako side: MatI64/IntArrayCore based O(n³) matmul (`MatI64.mul_naive`).
|
||
- C side: `MatI64Core { int64_t *ptr; rows; cols; stride; }` with identical algorithm.
|
||
- Accept that performance may still be far from the 80% target; 21.8 focuses on **structural integration and parity**, not tuning.
|
||
|
||
Out of scope:
|
||
|
||
- New optimizations inside AotPrep / CollectionsHot.
|
||
- SIMD/blocked matmul kernels (to be handled in a later optimization phase).
|
||
- f64/complex matrix variants.
|
||
|
||
## Tasks for implementation (Claude Code)
|
||
|
||
1) **Fix MatI64 visibility in Stage‑B / MirBuilder**
|
||
- Reproduce the current failure:
|
||
- Use a small `.hako` like:
|
||
- `using nyash.core.numeric.matrix_i64 as MatI64`
|
||
- `static box Main { method main(args) { local n = 4; local A = MatI64.new(n,n); return A.at(0,0); } }`
|
||
- Confirm `env.mirbuilder.emit` reports `undefined variable: MatI64`.
|
||
- Investigate how modules from `nyash.toml` (`"nyash.core.numeric.matrix_i64" = "lang/src/runtime/numeric/mat_i64_box.hako"`) are made visible to Stage‑B and MirBuilder.
|
||
- Adjust the resolver / module prelude so that `MatI64` (and `IntArrayCore`) are treated like other core boxes:
|
||
- Either via explicit prelude inclusion,
|
||
- Or via module registry entries consumed by the builder.
|
||
|
||
2) **Ensure `tools/hakorune_emit_mir.sh` can emit MIR(JSON) for matmul_core**
|
||
- Once MatI64 is visible, run:
|
||
- `HAKO_APPLY_AOT_PREP=1 NYASH_AOT_COLLECTIONS_HOT=1 NYASH_LLVM_FAST=1 NYASH_MIR_LOOP_HOIST=1 NYASH_JSON_ONLY=1 tools/hakorune_emit_mir.sh <matmul_core.hako> tmp/matmul_core.json`
|
||
- Acceptance:
|
||
- No `undefined variable: MatI64` / `IntArrayCore` errors.
|
||
- `tmp/matmul_core.json` is valid MIR(JSON) (same schema as existing matmul case).
|
||
|
||
3) **Finish `matmul_core` microbench**
|
||
- Use the existing skeleton in `tools/perf/microbench.sh` (`case matmul_core`):
|
||
- Confirm Hako side compiles and runs under `--backend vm`.
|
||
- Confirm EXE path works:
|
||
- `NYASH_SKIP_TOML_ENV=1 NYASH_LLVM_SKIP_BUILD=1 tools/perf/microbench.sh --case matmul_core --backend llvm --exe --runs 1 --n 64`
|
||
- Update `benchmarks/README.md`:
|
||
- Add `matmul_core` row with a short description:
|
||
- “MatI64/IntArrayCore vs MatI64Core C struct (ptr+rows+cols+stride)”
|
||
- Record initial ratios (even if far from 80%).
|
||
|
||
4) **Keep existing behaviour stable**
|
||
- No changes to default user behaviour, env toggles, or existing benches beyond adding `matmul_core`.
|
||
- Ensure quick/profile smokes (where applicable) remain green with numeric core present.
|
||
|
||
## 結果（2025-11-14 時点）
|
||
|
||
本フェーズは「builder/imports 導線の整備」までを完了とし、EXE/LLVM ベンチ統合は Phase 25 に移管する。
|
||
|
||
**達成済み（21.8）**
|
||
|
||
- Stage‑B / Bridge / MirBuilder:
|
||
- `BridgeEnv` に `imports: HashMap<String,String>` フィールドを追加し、using 由来のエイリアス情報（alias → box_type）を保持できるようにした。
|
||
- `MapVars::resolve` を拡張し、`env.imports` を参照して `MatI64` / `IntArrayCore` などの using エイリアスを「静的 box 参照」として MIR 上で解決できるようにした。
|
||
- JSON v0 → MIR(JSON) 経路に `*_with_imports` 版を追加:
|
||
- `runner::json_v0_bridge::parse_json_v0_to_module_with_imports`
|
||
- `host_providers::mir_builder::program_json_to_mir_json_with_imports`
|
||
- `HAKO_MIRBUILDER_IMPORTS` 環境変数を経由して、Stage‑B 側で収集した imports を Rust 側の MirBuilder に渡す配線を追加（読み取り側）。
|
||
- Using / imports 収集:
|
||
- `collect_using_and_strip` の戻り値を `(cleaned, prelude_paths, imports)` に拡張し、using から alias 情報を収集できるようにした。
|
||
- 既存呼び出し側は `_imports` として無視するため、従来挙動は維持。
|
||
- MatI64/IntArrayCore:
|
||
- `using nyash.core.numeric.matrix_i64 as MatI64` / `using nyash.core.numeric.intarray as IntArrayCore` の構文が Stage‑B/Bridge で undefined variable エラーにならないところまで確認済み（builder 経路に乗る）。
|
||
|
||
**未完・移管（Phase 25 に送るもの）**
|
||
|
||
- `matmul_core` microbench の EXE/LLVM 統合:
|
||
- MIR 上では `MatI64` が BoxCall として現れるが、LLVM ラインには BoxCall 実行経路がなく、AotPrep も MatI64 向けの書き換えを持たない。
|
||
- 「MatI64/IntArrayCore を LLVM までどう運ぶか」（numeric ABI / AotPrep / Ring1 AOT 設計）は、Phase 25「脱Rustランタイム / Ring0-Ring1再編」で扱う。
|
||
- Numeric core AOT ライン:
|
||
- IntArrayCore/MatI64 の実装を `.hako`（Ring1）側に寄せ、LLVM からは汎用 `ExternCall`/numeric ABI のみを見る構造は Phase 25 の設計スコープとする。
|
||
|
||
詳細な将来計画・numeric runtime AOT 方針については `docs/private/roadmap2/phases/phase-25/README.md` を参照。
|