[llvm] [AMDGPU] Per-chain MFMA->AGPR conversion (PR #217328)

[Matt Arsenault via llvm-commits]

================
@@ -104,6 +106,12 @@ static cl::opt<bool> DisableRewriteMFMAFormSchedStage(
     "amdgpu-disable-rewrite-mfma-form-sched-stage", cl::Hidden,
     cl::desc("Disable rewrite mfma rewrite scheduling stage"), cl::init(true));
 
+static cl::opt<bool> EnableMFMAChainSplitting(
----------------
arsenm wrote:

Options should only be added for concrete needs, not just in case. There is a major cost to the large set of unused cl::opts 

I'd also much rather maintain only one path 

https://github.com/llvm/llvm-project/pull/217328