docs: plan the decode/render thread pipeline with GPU zero-copy

Task book for the render-pipeline rearchitecture: one decode thread
and one render thread feeding queue-linked stages with the main
process presenting (the GPU's single queue makes the multi-process
backend dead weight), one dedicated OpenFX host process with
bounded respawn, GPU-resident frames end-to-end except at the CPU
OFX/export/cache boundaries, a per-backend interop table, and the
resolve rewrite to a single-loop match over a completed Job enum
(CacheJob included) following upstream Olive's
NodeTraverser::ResolveJobs. Milestones M0-M5 with the thread backend
kept behind an OAK_PIPELINE fallback switch.
This commit is contained in:
2026-09-11 09:48:02 +08:00
parent d216567bcd
commit 87ed45ffee
2 changed files with 276 additions and 0 deletions
+1
View File
@@ -10,6 +10,7 @@
| [`ai-agent-design.md`](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`](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`](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`](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)、OpenFX 收编进唯一隔离进程(崩溃重生)、decode→render→present 三队列流水线、内置效果全 GPU 零拷贝(CPU OFX 才回读,平台分支 CUDA/Vulkan/OpenGL/DirectX)、resolve 重写为 match+Job 单循环(含 CacheJob);M0–M5 里程碑 | 已解锁,随时启动(M0 独立先行) |
## 其他参考