[llvm] [AMDGPU] Track async events for vm_vsrc simplifications (PR #211333)

[Krzysztof Drewniak via llvm-commits]

================
@@ -1322,8 +1314,8 @@ void WaitcntBrackets::simplifyVmVsrc(const AMDGPU::Waitcnt &CheckWait,
   // have read its VGPR sources, but only if there are no other outstanding VMEM
   // operations that use a different counter (like SAMPLE_CNT).
   static constexpr AMDGPU::InstCounterType VmemCounters[] = {
-      AMDGPU::LOAD_CNT, AMDGPU::STORE_CNT, AMDGPU::SAMPLE_CNT, AMDGPU::BVH_CNT,
-      AMDGPU::DS_CNT};
+      AMDGPU::LOAD_CNT, AMDGPU::STORE_CNT, AMDGPU::SAMPLE_CNT,
+      AMDGPU::BVH_CNT,  AMDGPU::DS_CNT,    AMDGPU::ASYNC_CNT};
----------------
krzysz00 wrote:

Because my memory is having a moment, this runs meaningfully after the asyncmark/asyncwait intrinsics get lowered to real counter usages, yes?

Also, while we're at it, is there a TENSORCNT hazard? That instruction's SGPRs-only, but I thought I might as well ask.

https://github.com/llvm/llvm-project/pull/211333