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/runtime/base/arena_allocator_test.cc b/runtime/base/arena_allocator_test.cc
index e2c2e2f..6bf56c8 100644
--- a/runtime/base/arena_allocator_test.cc
+++ b/runtime/base/arena_allocator_test.cc
@@ -23,9 +23,9 @@
 
 class ArenaAllocatorTest : public testing::Test {
  protected:
-  size_t NumberOfArenas(ArenaAllocator* arena) {
+  size_t NumberOfArenas(ArenaAllocator* allocator) {
     size_t result = 0u;
-    for (Arena* a = arena->arena_head_; a != nullptr; a = a->next_) {
+    for (Arena* a = allocator->arena_head_; a != nullptr; a = a->next_) {
       ++result;
     }
     return result;