[all-commits] [llvm/llvm-project] 818256: [MLIR][GPUToLLVMSPV] Update ConvertGpuOpsToLLVMSPV...

[Jefferson Le Quellec via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 81825687b4b45e0a6839fd05cad7bedf18205315
      https://github.com/llvm/llvm-project/commit/81825687b4b45e0a6839fd05cad7bedf18205315
  Author: Jefferson Le Quellec <jefferson.lequellec at codeplay.com>
  Date:   2024-12-12 (Thu, 12 Dec 2024)

  Changed paths:
    M mlir/include/mlir/Conversion/Passes.td
    M mlir/lib/Conversion/GPUToLLVMSPV/GPUToLLVMSPV.cpp
    M mlir/test/Conversion/GPUToLLVMSPV/gpu-to-llvm-spv.mlir

  Log Message:
  -----------
  [MLIR][GPUToLLVMSPV] Update ConvertGpuOpsToLLVMSPVOps's option (#118818)

## Description

This PR updates the `ConvertGpuOpsToLLVMSPVOps`'s option by replacing
the `index-bitwidth` with a boolean option `use-64bit-index` (similar to
the `ConvertGPUToSPIRV` option).

The reason for this modification is because the
`ConvertGpuOpsToLLVMSPVOps`:
> Generate LLVM operations to be ingested by a SPIR-V backend for gpu
operations

In the context of SPIR-V specifications only two physical addressing
models are allowed: `Physical32` and `Physical64`.

This change guarantees output sanity by preventing invalid or
unsupported index bitwidths from being specified.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications