Use ScopedArenaAllocator for building HGraph.

Memory needed to compile the two most expensive methods for
aosp_angler-userdebug boot image:
  BatteryStats.dumpCheckinLocked() : 21.1MiB -> 20.2MiB
  BatteryStats.dumpLocked(): 42.0MiB -> 40.3MiB
This is because all the memory previously used by the graph
builder is reused by later passes.

And finish the "arena"->"allocator" renaming; make renamed
allocator pointers that are members of classes const when
appropriate (and make a few more members around them const).

Test: m test-art-host-gtest
Test: testrunner.py --host
Bug: 64312607
Change-Id: Ia50aafc80c05941ae5b96984ba4f31ed4c78255e
diff --git a/compiler/optimizing/instruction_simplifier.cc b/compiler/optimizing/instruction_simplifier.cc
index f39acab..afe7484 100644
--- a/compiler/optimizing/instruction_simplifier.cc
+++ b/compiler/optimizing/instruction_simplifier.cc
@@ -1284,9 +1284,9 @@
         DCHECK(input_other->IsShr());  // For UShr, we would have taken the branch above.
         // Replace SHR+AND with USHR, for example "(x >> 24) & 0xff" -> "x >>> 24".
         HUShr* ushr = new (GetGraph()->GetAllocator()) HUShr(instruction->GetType(),
-                                                         input_other->InputAt(0),
-                                                         input_other->InputAt(1),
-                                                         input_other->GetDexPc());
+                                                             input_other->InputAt(0),
+                                                             input_other->InputAt(1),
+                                                             input_other->GetDexPc());
         instruction->GetBlock()->ReplaceAndRemoveInstructionWith(instruction, ushr);
         input_other->GetBlock()->RemoveInstruction(input_other);
         RecordSimplification();