Common architecture

This section introduces the common architecture between FlagTree and Triton. The common architecture comprises AST Processing, Backend Compilation, and Runtime System modules that handle the compilation process.

The following list introduces directories in FlagTree and the functions of each module:

• AST Processing:

  • Directory: python/triton/compiler/ and python/triton/language/

  • Function: Understands your code and translates Python kernel code into TTIR（Triton IR）in MLIR（Multi-Level Intermediate Representation）format. This module comprises the following submodules:

  • Language Definition:

    • Directory: python/triton/language/

    • Function: Defines Triton’s language constructs, including core types (core.py), standard operations (standard.py), math functions (math.py), and semantic rules (semantic.py). These provide the building blocks that kernels use.

  • Code Generation:

    • Directory: python/triton/compiler/code_generator.py

    • Function: Transforms Python AST（Abstract Syntax Tree）into TTIR （that is Triton IR） operations by recognizing Triton language constructs (defined in python/triton/language/) and generating the initial Intermediate Representation (IR).

  • Compiler Coordination:

    • Directory: python/triton/compiler/compiler.py

    • Function: Orchestrates the compilation process, managing AST sources, IR sources, and coordinating with backends.

• Backend Compilation:

  • Directory: third_party/[backend]/backend/compiler.py

  • Function: Each backend defines its compilation pipeline through the add_stages() method, which specifies how to transform TTIR into executable code. The typical flow includes: TTIR → TTGPU IR/ Linalg IR → LLVM IR → target assembly → binary. Each backend implements these stages with hardware-specific optimizations and code generation.

• Runtime System:

  • Directory: python/triton/runtime/

  • Function: Handles JIT compilation, kernel caching, and kernel launch. The runtime compiles kernels on-demand when first called, caches compiled results to avoid recompilation, and manages kernel execution on the GPU through backend drivers.