Expertise in MLIR (Multi-Level Intermediate Representation) and CIR (Clang IR) development for domain-specific compilation and high-level optimizations. Use this skill when building ML compilers, domain-specific languages, or working with multi-level compilation pipelines.
This skill covers MLIR (Multi-Level Intermediate Representation) development for building domain-specific compilers and high-level optimization pipelines.
MLIR is a compiler infrastructure that enables building reusable and extensible compiler components. It provides:
High-Level DSL
↓
Domain-Specific Dialects (e.g., TensorFlow, PyTorch)
↓
Mid-Level Dialects (e.g., Linalg, Affine)
↓
Low-Level Dialects (e.g., LLVM, GPU)
↓
Target Code
Dialects are groupings of operations, types, and attributes:
// Define a custom dialect
class MyDialect : public mlir::Dialect {