[llvm] [GreedyRA]: Add flag to force local assignment heuristics (PR #102160)

[Matt Arsenault via llvm-commits]

================
@@ -321,10 +327,12 @@ unsigned DefaultPriorityAdvisor::getPriority(const LiveInterval &LI) const {
     // Giant live ranges fall back to the global assignment heuristic, which
     // prevents excessive spilling in pathological cases.
     const TargetRegisterClass &RC = *MRI->getRegClass(Reg);
-    bool ForceGlobal = RC.GlobalPriority ||
-                       (!ReverseLocalAssignment &&
-                        (Size / SlotIndex::InstrDist) >
-                            (2 * RegClassInfo.getNumAllocatableRegs(&RC)));
+
+    bool ForceGlobal =
+        RC.GlobalPriority ||
+        (!GreedyForceLocalAssignment && !ReverseLocalAssignment &&
+         (Size / SlotIndex::InstrDist) >
+             (2 * RegClassInfo.getNumAllocatableRegs(&RC)));
----------------
arsenm wrote:

This heuristic is broken and I've really wanted to fix this for a long time. getNumAllocatableRegs does not understand overlapping registers. What this really should be accounting for is number of independently allocatable registers. Can we fix that instead of just adding a flag to skip the heuristic? 

https://github.com/llvm/llvm-project/pull/102160