[llvm] [AMDGPU] Skip s_delay_alu for gfx12+ WMMA C-reuse chains (PR #214101)

[Matt Arsenault via llvm-commits]

================
@@ -444,6 +455,17 @@ class AMDGPUInsertDelayAlu {
       // twice?
       State.advance(Type, Cycles);
 
+      // Track the preceding WMMA's dst for C-reuse; reset on anything else.
+      if (ST->hasGFX12Insts()) {
----------------
arsenm wrote:

Don't need this?

https://github.com/llvm/llvm-project/pull/214101