[PATCH] D89754: [AMDGPU] Avoid inserting noops during scheduling

[Austin Kerbow via Phabricator via llvm-commits]

kerbowa added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIShrinkInstructions.cpp:605-633
-      // Combine adjacent s_nops to use the immediate operand encoding how long
-      // to wait.
-      //
-      // s_nop N
-      // s_nop M
-      //  =>
-      // s_nop (N + M)
----------------
foad wrote:
> Shouldn't this be part of the previous patch?
The scheduler still emits noops one at a time. So we cannot remove this in the earlier patch.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D89754/new/

https://reviews.llvm.org/D89754