[all-commits] [llvm/llvm-project] da7462: [NVPTX] Add tex.grad.cube{array} intrinsics (#77693)

[Alex MacLean via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: da7462a6ae3730551efe0be5d7d525195dc3678f
      https://github.com/llvm/llvm-project/commit/da7462a6ae3730551efe0be5d7d525195dc3678f
  Author: Alex MacLean <amaclean at nvidia.com>
  Date:   2024-01-17 (Wed, 17 Jan 2024)

  Changed paths:
    M llvm/include/llvm/IR/IntrinsicsNVVM.td
    M llvm/lib/Target/NVPTX/NVPTXISelDAGToDAG.cpp
    M llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp
    M llvm/lib/Target/NVPTX/NVPTXISelLowering.h
    M llvm/lib/Target/NVPTX/NVPTXIntrinsics.td
    M llvm/lib/Target/NVPTX/NVPTXReplaceImageHandles.cpp
    M llvm/test/CodeGen/NVPTX/surf-tex.py

  Log Message:
  -----------
  [NVPTX] Add tex.grad.cube{array} intrinsics (#77693)

Extend IR support for PTX `tex` instruction described in [PTX ISA.
9.7.9.3. Texture Instructions:
tex](https://docs.nvidia.com/cuda/parallel-thread-execution/index.html#texture-instructions-tex).
Add support for unified-move versions of `tex.grad.cube{array}` variants
added in PTX ISA 4.3.