Use ScopedArenaAllocator for register allocation.

Memory needed to compile the two most expensive methods for
aosp_angler-userdebug boot image:
  BatteryStats.dumpCheckinLocked() : 25.1MiB -> 21.1MiB
  BatteryStats.dumpLocked(): 49.6MiB -> 42.0MiB
This is because all the memory previously used by Scheduler
is reused by the register allocator; the register allocator
has a higher peak usage of the ArenaStack.

And continue the "arena"->"allocator" renaming.

Test: m test-art-host-gtest
Test: testrunner.py --host
Bug: 64312607
Change-Id: Idfd79a9901552b5147ec0bf591cb38120de86b01
diff --git a/compiler/utils/x86/assembler_x86.h b/compiler/utils/x86/assembler_x86.h
index dce3ad2..f3b516c 100644
--- a/compiler/utils/x86/assembler_x86.h
+++ b/compiler/utils/x86/assembler_x86.h
@@ -266,7 +266,8 @@
  */
 class ConstantArea {
  public:
-  explicit ConstantArea(ArenaAllocator* arena) : buffer_(arena->Adapter(kArenaAllocAssembler)) {}
+  explicit ConstantArea(ArenaAllocator* allocator)
+      : buffer_(allocator->Adapter(kArenaAllocAssembler)) {}
 
   // Add a double to the constant area, returning the offset into
   // the constant area where the literal resides.
@@ -307,7 +308,8 @@
 
 class X86Assembler FINAL : public Assembler {
  public:
-  explicit X86Assembler(ArenaAllocator* arena) : Assembler(arena), constant_area_(arena) {}
+  explicit X86Assembler(ArenaAllocator* allocator)
+      : Assembler(allocator), constant_area_(allocator) {}
   virtual ~X86Assembler() {}
 
   /*