Mike-Solar
48e99e56b7
docs/zh/plans/render-pipeline-threads.md M2: the graph's textures stay on the GPU from evaluation through presentation, and presentation runs on the UI's own wgpu device. - wgpu 25 -> 29 (naga 29) across the engine, unifying it with gpui_wgpu so engine textures are directly sampleable by the presenter (a single wgpu remains in the lockfile). - GpuContext::adopt/install_shared: the app registers the window's device at startup and the render thread renders on it; texture_handle hands the raw Arc<wgpu::Texture> to SurfaceSource::Texture - zero-copy present on Linux/FreeBSD. The shared slot replaces an engine context that has not touched the GPU yet (startup-order guard) and refuses once it has. - Texture::Gpu shares a GpuLease so clones release the registry token exactly once; the compositor, transitions and adjustment sweeps keep GPU textures end to end (no per-clip readbacks; GPU clears for black/generated frames). - Color management stays on the GPU: the output node + display ICC chain is baked into a 65^3 3D LUT with the exact CPU reference and applied by the present WGSL pass (manual trilinear); ColorTransformJob bakes its OCIO processor the same way. Neither path skips color management. - The explicit readback boundaries accept GPU textures: export encoder, CLI, worker shm, disk cache; CPU OpenFX already read back. - M5 dependency: the YUV->RGB GPU pass (BT.601/709/2020 x limited/full) matches colormath::yuv444p16_to_rgb_f32. - Acceptance: gpu_transfer_counters; single-clip and layered (multi-track + transition + adjustment) playback tests assert zero GPU->CPU readbacks, and the app test asserts adopted-device present is zero-copy. GPU tests hard-fail when OAK_REQUIRE_GPU is set (CI lavapipe) instead of skipping silently.
长期计划（plans）
本目录收纳 Oak 的中长期规划文档与并行执行计划。已完成的战役
文档归档在 completed/；进行中的计划在下方目录。
目录
| 文档 | 内容 | 启动前提 |
|---|---|---|
ai-agent-design.md |
AI Agent 插件设计（已按 OPP/1 重写）：多模态 LLM 作为外部插件经策展工具面自动剪辑，render.* 取帧回喂形成"编辑→看图→再编辑"视觉闭环；事务化编辑、双层确认、声明式 AI 面板、Mock LLM/回放夹具测试、A1–A5 里程碑 |
external-plugin-system P1–P3 完成（面板需 P4） |
external-plugin-system.md |
外部功能插件系统：插件=独立进程（非库加载），JSON-RPC over stdio 控制面 + shm 数据面（泛化 M15 render-worker 传输）；策展宿主 API（事务化可撤销编辑、取帧回喂 AI）、声明式/像素面双 UI 路径、能力位与确认模式；含与 ai-agent-design.md 的关系与 P1–P6 里程碑 | 已解锁（RIIR + M15 完成），随时启动 |
external-plugin-protocol.md |
插件协议规范 OPP/1：NDJSON 分帧 + JSON-RPC 2.0 双向信封、握手/心跳/关闭、全量方法/事件/错误码、编辑事务协议、shm 数据面（无头部无锁）、声明式与像素面 UI 协议、限流配额、版本演进规则、AI 粗剪报文示例 | 随 external-plugin-system 启动 |
render-pipeline-threads.md |
渲染管线改造：单解码线程 + 单渲染线程 + 主进程上屏（GPU 单队列，多进程无意义）；解码硬解优先且零拷贝（FFmpeg hwaccel 经 DMA-BUF/DXGI/IOSurface/CUDA 导入，CPU 解码上传仅 fallback，手写 GPU 解码为次选）；OpenFX 收编进唯一隔离进程；decode→render→present 三队列流水线；内置效果全 GPU 零拷贝；resolve 重写为 match+Job 单循环并升级为 Job 图（GraphInput/GraphOutput 虚拟端点可见、禁删，从输入节点 Kahn 形态 BFS 至全部分支汇聚输出节点）；M0a–M5 里程碑 | 已解锁，随时启动（M0a 独立先行） |
其他参考
- 构建：
docs/zh/build.md、docs/zh/build_macos-zh.md - 工程文件：
docs/zh/project-file-reference.md - 代码风格与 Google Test 要求：
CONTRIBUTING.md（仓库根）