[all-commits] [llvm/llvm-project] f956e7: [AMDGPU] Prefer `s_memtime` for `readcyclecounter`...

[Joseph Huber via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: f956e7fbf161447b9236f7c4448a9d02d3564261
      https://github.com/llvm/llvm-project/commit/f956e7fbf161447b9236f7c4448a9d02d3564261
  Author: Joseph Huber <huberjn at outlook.com>
  Date:   2024-02-01 (Thu, 01 Feb 2024)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SMInstructions.td
    M llvm/test/CodeGen/AMDGPU/readcyclecounter.ll

  Log Message:
  -----------
  [AMDGPU] Prefer `s_memtime` for `readcyclecounter` on GFX10 (#80211)

Summary:
The old `s_memtime` instruction was supported until the GFX10
architecture. Although this instruction has a higher latency than the
new shader counter, it's much more usable as a processor clock as it is
a full 64-bit counter. The new shader counter is only a 20-bit counter,
which makes it difficult to use as a standard cycle counter as it will
overflow in a few milliseconds. This patch suggests preferring
`s_memtime` for this instrinsic if it is still available.