Remove the old ARM code generator from ART's Optimizing compiler.

The AArch32 VIXL-based code generator has been the default
ARM code generator in ART for some time now. The old ARM
code generator does not compile anymore; retiring it.

Test: test.py
Bug: 63316036
Change-Id: Iab8fbc4ac73eac2c1a809cd7b22fec6b619755db
diff --git a/compiler/optimizing/intrinsics_arm_vixl.cc b/compiler/optimizing/intrinsics_arm_vixl.cc
index ced931b..8b4044d 100644
--- a/compiler/optimizing/intrinsics_arm_vixl.cc
+++ b/compiler/optimizing/intrinsics_arm_vixl.cc
@@ -1058,7 +1058,7 @@
                     (can_call ? Location::kOutputOverlap : Location::kNoOutputOverlap));
   if (type == Primitive::kPrimNot && kEmitCompilerReadBarrier && kUseBakerReadBarrier) {
     // We need a temporary register for the read barrier marking slow
-    // path in InstructionCodeGeneratorARM::GenerateReferenceLoadWithBakerReadBarrier.
+    // path in CodeGeneratorARMVIXL::GenerateReferenceLoadWithBakerReadBarrier.
     locations->AddTemp(Location::RequiresRegister());
   }
 }
@@ -2377,7 +2377,7 @@
       // TODO: Also convert this intrinsic to the IsGcMarking strategy?
 
       // SystemArrayCopy implementation for Baker read barriers (see
-      // also CodeGeneratorARM::GenerateReferenceLoadWithBakerReadBarrier):
+      // also CodeGeneratorARMVIXL::GenerateReferenceLoadWithBakerReadBarrier):
       //
       //   uint32_t rb_state = Lockword(src->monitor_).ReadBarrierState();
       //   lfence;  // Load fence or artificial data dependency to prevent load-load reordering