Passes

Passes listed here are custom passes implemented within Pylir. Make sure to also checkout passes implemented in upstream MLIR: https://mlir.llvm.org/docs/Passes/

There are two kinds of passes:

• Transformation passes: Passes performing transformations on the IR, usually for optimization.

• Conversion passes: Passes converting from one MLIR Dialect to another.

Passes

• Transformation Passes
  • Generic Transformation Passes
    • -pylir-conditionals-implications
    • -pylir-dce
    • -pylir-fixpoint
    • -pylir-load-forwarding
    • -pylir-sroa
  • ‘py’ Transformation Passes
    • -pylir-expand-py-dialect
    • -pylir-fold-globals
    • -pylir-global-load-store-elimination
    • -pylir-global-sroa
    • -pylir-inliner
  • ‘pyMem’ Transformation Passes
    • -pylir-heap-to-stack
  • ‘pyHIR’ Transformation Passes
    • -pylir-class-body-outlining
    • -pylir-func-outlining
• Conversion passes
  • ‘py’ Conversion Passes
    • -convert-pylir-to-llvm
    • -convert-pylirHIR-to-pylirPy
    • -convert-pylirPy-to-pylirMem