Emit VMLA for multiply-add on ARM

src/arm/lithium-arm.h

 // Copyright 2012 the V8 project authors. All rights reserved.
 // Redistribution and use in source and binary forms, with or without
 // modification, are permitted provided that the following conditions are
 // met:
 //
 //     * Redistributions of source code must retain the above copyright
 //       notice, this list of conditions and the following disclaimer.
 //     * Redistributions in binary form must reproduce the above
 //       copyright notice, this list of conditions and the following
 //       disclaimer in the documentation and/or other materials provided
@@ skipping 117 matching lines @@
   V(LoadKeyed)                                  \
   V(LoadKeyedGeneric)                           \
   V(LoadNamedField)                             \
   V(LoadNamedFieldPolymorphic)                  \
   V(LoadNamedGeneric)                           \
   V(MapEnumLength)                              \
   V(MathFloorOfDiv)                             \
   V(MathMinMax)                                 \
   V(ModI)                                       \
   V(MulI)                                       \
+  V(MultiplyAddD)                               \
   V(NumberTagD)                                 \
   V(NumberTagI)                                 \
   V(NumberTagU)                                 \
   V(NumberUntagD)                               \
   V(ObjectLiteral)                              \
   V(OsrEntry)                                   \
   V(OuterContext)                               \
   V(Parameter)                                  \
   V(Power)                                      \
   V(PushArgument)                               \
@@ skipping 466 matching lines @@
 
   LOperand* left() { return inputs_[0]; }
   LOperand* right() { return inputs_[1]; }
   LOperand* temp() { return temps_[0]; }
 
   DECLARE_CONCRETE_INSTRUCTION(MulI, "mul-i")
   DECLARE_HYDROGEN_ACCESSOR(Mul)
 };
 
 
+class LMultiplyAddD: public LTemplateInstruction<1, 3, 0> {
+ public:
+  LMultiplyAddD(LOperand* c, LOperand *a, LOperand* b) {

[Sven Panne, 2012/11/13 13:03:52]

Can we use something less cryptic for the operands like factor1, factor2 and
summand? Or are a/b/c standard names in some ARM and/or GCC documentation?

[Jakob Kummerow, 2012/11/13 13:05:21]

I'd appreciate a comment here what the three operands are used for ("Computes c
+= a * b" or something like that).

Oh, and a nit: "LOperand* a"

[hans, 2012/11/13 14:59:58]

Done.

[hans, 2012/11/13 14:59:58]

Done.

+    inputs_[0] = c;
+    inputs_[1] = a;
+    inputs_[2] = b;
+  }
+
+  LOperand* c() { return inputs_[0]; }
+  LOperand* a() { return inputs_[1]; }
+  LOperand* b() { return inputs_[2]; }
+
+  DECLARE_CONCRETE_INSTRUCTION(MultiplyAddD, "multiply-add-d")
+};
+
+
 class LCmpIDAndBranch: public LControlInstruction<2, 0> {
  public:
   LCmpIDAndBranch(LOperand* left, LOperand* right) {
     inputs_[0] = left;
     inputs_[1] = right;
   }
 
   LOperand* left() { return inputs_[0]; }
   LOperand* right() { return inputs_[1]; }
 
@@ skipping 1890 matching lines @@
 
   DISALLOW_COPY_AND_ASSIGN(LChunkBuilder);
 };
 
 #undef DECLARE_HYDROGEN_ACCESSOR
 #undef DECLARE_CONCRETE_INSTRUCTION
 
 } }  // namespace v8::internal
 
 #endif  // V8_ARM_LITHIUM_ARM_H_