[all-commits] [llvm/llvm-project] a2d3bb: Revert "[NVPTX] Lower 16xi8 and 8xi8 stores effici...

[Artem Belevich via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: a2d3bb1fa96aeba610273908d09fefad6d6032f8
      https://github.com/llvm/llvm-project/commit/a2d3bb1fa96aeba610273908d09fefad6d6032f8
  Author: Artem Belevich <tra at google.com>
  Date:   2023-12-06 (Wed, 06 Dec 2023)

  Changed paths:
    M llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp
    M llvm/test/CodeGen/NVPTX/i8x4-instructions.ll
    M llvm/test/CodeGen/NVPTX/vector-stores.ll

  Log Message:
  -----------
  Revert "[NVPTX] Lower 16xi8 and 8xi8 stores efficiently (#73646)" (#74518)

This reverts commit 173fcf7da592acd284dc50749558fd36928861f0.

We need to constrain the optimization to properly aligned loads/stores
only.
https://github.com/llvm/llvm-project/pull/73646#issuecomment-1841454559