[clang] [llvm] [AMDGPU] Emit a waitcnt instruction after each memory instruction (PR #79236)

[Pierre van Houtryve via cfe-commits]

================
@@ -167,6 +167,10 @@ def FeatureCuMode : SubtargetFeature<"cumode",
   "Enable CU wavefront execution mode"
 >;
 
+def FeaturePreciseMemory
----------------
Pierre-vh wrote:

It's only called once per run by the driver, yes

We already do this for wavefrontsize64, and pretty much all other options that aren't automatically handled. We even call `hasFlag` twice on the same option in some cases instead of saving the result somewhere, so I doubt `hasFlag` can be considered expensive enough to justify not dropping `amdgpu-` from the feature name

https://github.com/llvm/llvm-project/pull/79236