[arm][turbofan] Use NEON for unaligned float64 memory accesses

src/compiler/arm/code-generator-arm.cc

 // Copyright 2014 the V8 project authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "src/compiler/code-generator.h"
 
 #include "src/arm/macro-assembler-arm.h"
 #include "src/assembler-inl.h"
 #include "src/compilation-info.h"
 #include "src/compiler/code-generator-impl.h"
@@ skipping 1425 matching lines @@
       break;
     case kArmVldrF32: {
       __ vldr(i.OutputFloatRegister(), i.InputOffset());
       DCHECK_EQ(LeaveCC, i.OutputSBit());
       break;
     }
     case kArmVstrF32:
       __ vstr(i.InputFloatRegister(0), i.InputOffset(1));
       DCHECK_EQ(LeaveCC, i.OutputSBit());
       break;
+    case kArmVld1F64: {
+      MemOperand src = i.InputOffset();
+      __ vld1(NeonSize::Neon8, NeonListOperand(i.OutputDoubleRegister()),
+              NeonMemOperand(src.rn(), src.rm()));

[martyn.capewell, 2017/03/23 11:11:53]

I don't think this works, as the addressing mode that accepts two registers uses
the second register as a post index (ie. *rn; rn += rm) not an offset (ie. *(rn
+ rm)).

It will need either an address calculation in a scratch here, or emitting an
extra addition in the selector. Store will be similar.

+      break;
+    }
+    case kArmVst1F64: {
+      MemOperand src = i.InputOffset(1);
+      __ vst1(Neon8, NeonListOperand(i.InputDoubleRegister(0)),
+              NeonMemOperand(src.rn(), src.rm()));
+      break;
+    }
     case kArmVldrF64:
       __ vldr(i.OutputDoubleRegister(), i.InputOffset());
       DCHECK_EQ(LeaveCC, i.OutputSBit());
       break;
     case kArmVstrF64:
       __ vstr(i.InputDoubleRegister(0), i.InputOffset(1));
       DCHECK_EQ(LeaveCC, i.OutputSBit());
       break;
     case kArmFloat32Max: {
       SwVfpRegister result = i.OutputFloatRegister();
@@ skipping 1326 matching lines @@
       padding_size -= v8::internal::Assembler::kInstrSize;
     }
   }
 }
 
 #undef __
 
 }  // namespace compiler
 }  // namespace internal
 }  // namespace v8