First draft of the sh4 port

src/sh4/assembler-sh4.h

+// Copyright 2011-2012 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
+//       with the distribution.
+//     * Neither the name of Google Inc. nor the names of its
+//       contributors may be used to endorse or promote products derived
+//       from this software without specific prior written permission.
+//
+// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+// A light-weight SH4 Assembler.
+
+#ifndef V8_SH4_ASSEMBLER_SH4_H_
+#define V8_SH4_ASSEMBLER_SH4_H_
+
+#include "assembler.h"
+#include "serialize.h"
+
+namespace v8 {
+namespace internal {
+
+// CPU Registers.
+//
+// 1) We would prefer to use an enum, but enum values are assignment-
+// compatible with int, which has caused code-generation bugs.
+//
+// 2) We would prefer to use a class instead of a struct but we don't like
+// the register initialization to depend on the particular initialization
+// order (which appears to be different on OS X, Linux, and Windows for the
+// installed versions of C++ we tried). Using a struct permits C-style
+// "initialization". Also, the Register objects cannot be const as this
+// forces initialization stubs in MSVC, making us dependent on initialization
+// order.
+//
+// 3) By not using an enum, we are possibly preventing the compiler from
+// doing certain constant folds, which may significantly reduce the
+// code generated for some assembly instructions (because they boil down
+// to a few constants). If this is a problem, we could change the code
+// such that we use an enum in optimized mode, and the struct in debug
+// mode. This way we get the compile-time error checking in debug mode
+// and best performance in optimized code.
+
+// Core register
+struct Register {
+  static const int kNumRegisters = 16;
+  static const int kNumAllocatableRegisters = 8;
+  static const int kSizeInBytes = 4;
+
+  static int ToAllocationIndex(Register reg) {
+    ASSERT(reg.code() < kNumAllocatableRegisters);
+    return reg.code();
+  }
+
+  static Register FromAllocationIndex(int index) {
+    ASSERT(index >= 0 && index < kNumAllocatableRegisters);
+    return from_code(index);
+  }
+
+  static const char* AllocationIndexToString(int index) {
+    ASSERT(index >= 0 && index < kNumAllocatableRegisters);
+    const char* const names[] = {
+      "r0",
+      "r1",
+      "r2",
+      "r3",
+      "r4",
+      "r5",
+      "r6",
+      "r7",
+    };
+    return names[index];
+  }
+
+  static Register from_code(int code) {
+    Register r = { code };
+    return r;
+  }
+
+  bool is_valid() const { return 0 <= code_ && code_ < kNumRegisters; }
+  bool is(Register reg) const { return code_ == reg.code_; }
+  int code() const {
+    ASSERT(is_valid());
+    return code_;
+  }
+  int bit() const {
+    ASSERT(is_valid());
+    return 1 << code_;
+  }
+
+  void set_code(int code) {
+    code_ = code;
+    ASSERT(is_valid());
+  }
+
+  // Unfortunately we can't make this private in a struct.
+  int code_;
+};
+
+
+#define REGISTER(N, C) \
+  const int kRegister_ ## N ## _Code = C; \
+  const Register N = { C }
+
+REGISTER(no_reg, -1);
+REGISTER(r0, 0);      // ABI caller saved/return, idem for JIT
+REGISTER(r1, 1);      // ABI caller saved/return, idem for JIT
+REGISTER(r2, 2);      // ABI caller saved/return, idem for JIT
+REGISTER(r3, 3);      // ABI caller saved/return, idem for JIT
+REGISTER(r4, 4);      // ABI caller saved/param, idem for JIT
+REGISTER(r5, 5);      // ABI caller saved/param, idem for JIT
+REGISTER(r6, 6);      // ABI caller saved/param, idem for JIT
+REGISTER(r7, 7);      // ABI caller saved/param, idem for JIT
+REGISTER(r8, 8);      // ABI callee saved, idem for JIT
+REGISTER(r9, 9);      // ABI callee saved, idem for JIT
+REGISTER(r10, 10);    // ABI callee saved, idem for JIT
+REGISTER(r11, 11);    // ABI callee saved, idem for JIT
+REGISTER(roots, 12);  // ABI GP, root table pointer for JIT
+REGISTER(cp, 13);     // ABI callee saved, context pointer for JIT
+REGISTER(fp, 14);     // ABI FP, idem for JIT
+REGISTER(sp, 15);     // ABI SP, idem for JIT
+
+const Register pr = { -2 };     // Link register
+
+const Register sh4_r0 = r0;
+const Register sh4_r1 = r1;
+const Register sh4_r2 = r2;
+const Register sh4_r3 = r3;
+const Register sh4_r4 = r4;
+const Register sh4_r5 = r5;
+const Register sh4_r6 = r6;
+const Register sh4_r7 = r7;
+const Register sh4_r8 = r8;
+const Register sh4_r9 = r9;
+const Register sh4_r10 = r10;
+const Register sh4_r11 = r11;
+const Register sh4_r12 = roots;
+const Register sh4_r13 = cp;
+const Register sh4_rtmp = r11;  // Used for low level (assembler-sh4.cc)
+const Register sh4_ip = r10;    // Used as additional scratch in JS code
+
+
+// Single word VFP register.
+struct SwVfpRegister {
+  bool is_valid() const { return 0 <= code_ && code_ < 16; }
+  bool is(SwVfpRegister reg) const { return code_ == reg.code_; }
+  int code() const {
+    ASSERT(is_valid());
+    return code_;
+  }
+  int bit() const {
+    ASSERT(is_valid());
+    return 1 << code_;
+  }
+  static SwVfpRegister from_code(int code) {
+    SwVfpRegister r = { code };
+    return r;
+  }
+  void split_code(int* vm, int* m) const {
+    ASSERT(is_valid());
+    *m = code_ & 0x1;
+    *vm = code_ >> 1;
+  }
+
+  int code_;
+};
+
+
+// Double word VFP register.
+struct DwVfpRegister {
+  static const int kNumRegisters = 8;
+  static const int kNumAllocatableRegisters = 8;
+
+  static int ToAllocationIndex(DwVfpRegister reg) {
+    ASSERT(reg.code() != 0);
+    return reg.code();
+  }
+
+  static DwVfpRegister FromAllocationIndex(int index) {
+    ASSERT(index >= 0 && index < kNumAllocatableRegisters);
+    return from_code(index);
+  }
+
+  static const char* AllocationIndexToString(int index) {
+    ASSERT(index >= 0 && index < kNumAllocatableRegisters);
+    const char* const names[] = {
+      "dr0",
+      "dr2",
+      "dr4",
+      "dr6",
+      "dr8",
+      "dr10",
+      "dr12",
+      "dr14",
+    };
+    return names[index];
+  }
+
+  static DwVfpRegister from_code(int code) {
+    DwVfpRegister r = { code };
+    return r;
+  }
+
+  // Supporting dr0 to dr8
+  bool is_valid() const { return 0 <= code_ && code_ < kNumRegisters * 2 - 1; }
+  bool is(DwVfpRegister reg) const { return code_ == reg.code_; }
+  SwVfpRegister low() const {
+    SwVfpRegister reg;
+    reg.code_ = code_;
+
+    ASSERT(reg.is_valid());
+    return reg;
+  }
+  SwVfpRegister high() const {
+    SwVfpRegister reg;
+    reg.code_ = code_ + 1;
+
+    ASSERT(reg.is_valid());
+    return reg;
+  }
+  int code() const {
+    ASSERT(is_valid());
+    return code_;
+  }
+  int bit() const {
+    ASSERT(is_valid());
+    return 1 << code_;
+  }
+  void split_code(int* vm, int* m) const {
+    ASSERT(is_valid());
+    *m = (code_ & 0x10) >> 4;
+    *vm = code_ & 0x0F;
+  }
+
+  int code_;
+};
+
+typedef DwVfpRegister DoubleRegister;
+
+// Support for the VFP registers fr0 to fr15 (dr0 to dr7).
+// Note that "fr(N):fr(N+1)" is the same as "dr(N)".
+const SwVfpRegister no_freg = { -1 };
+const SwVfpRegister fr0  = {  0 };
+const SwVfpRegister fr1  = {  1 };
+const SwVfpRegister fr2  = {  2 };
+const SwVfpRegister fr3  = {  3 };
+const SwVfpRegister fr4  = {  4 };
+const SwVfpRegister fr5  = {  5 };
+const SwVfpRegister fr6  = {  6 };
+const SwVfpRegister fr7  = {  7 };
+const SwVfpRegister fr8  = {  8 };
+const SwVfpRegister fr9  = {  9 };
+const SwVfpRegister fr10 = { 10 };
+const SwVfpRegister fr11 = { 11 };
+// Callee saved registers
+// Using these registers is forbidden for the moment as we do not save/restaure
+// them on ABI frontiers.
+// const SwVfpRegister fr12 = { 12 };
+// const SwVfpRegister fr13 = { 13 };
+// const SwVfpRegister fr14 = { 14 };
+// const SwVfpRegister fr15 = { 15 };
+
+
+// Caller saved registers
+const DwVfpRegister no_dreg = { -1 };
+const DwVfpRegister dr0   = {  0  };
+const DwVfpRegister dr2   = {  2  };
+const DwVfpRegister dr4   = {  4  };
+const DwVfpRegister dr6   = {  6  };
+const DwVfpRegister dr8   = {  8  };
+const DwVfpRegister dr10  = {  10 };
+// Callee saved registers.
+// The uses of theses registers is forbidden for the moment as we do not
+// save/restaure them on ABI frontiers.
+// const DwVfpRegister dr12  = {  12 };
+// const DwVfpRegister dr14  = {  14 };
+
+enum Condition {
+  // any value < 0 is considered no_condition
+  kNoCondition = -1,
+
+  eq = 0 << 28,       // equal
+  ne = 1 << 28,       // not equal
+  gt = 2 << 28,       // greater
+  ge = 3 << 28,       // greater or equal
+  hi = 4 << 28,       // unsigned higher
+  hs = 5 << 28,       // unsigned higher or equal
+  lt = 6 << 28,       // lesser
+  le = 7 << 28,       // lesser or equal
+  ui = 8 << 28,       // unsigned lower
+  us = 9 << 28,       // unsigned lower or equal
+  pl = 10 << 28,      // positiv
+  pz = 11 << 28,      // positiv or null
+  ql = 12 << 28,      // negativ
+  qz = 13 << 28,      // negativ or null
+  al = 14 << 28,      // Always
+
+  // Aliases
+  t = eq,        // cmp eq; if SH4 cmpeq/cmp sets the T bit, t == eq
+  f = ne         // cmp ne: if SH4 cmpeq/cmp clears the T bit, f == ne
+};
+
+enum AddrMode {
+  PreIndex,
+  PostIndex,
+  Offset
+};
+
+
+// We use the cause fields bcause they are set to 1 or 0 depending on the
+// action. So they don't need to be reseted but they mist be use immediately
+static const uint32_t kFPUExceptionMask          = 0x1f << 12;
+static const uint32_t kFPUInexactExceptionBit    = 1 << 12;
+static const uint32_t kFPUUnderflowExceptionBit  = 1 << 13;
+static const uint32_t kFPUOverflowExceptionBit   = 1 << 14;
+static const uint32_t kFPUDividezeroExceptionBit = 1 << 15;
+static const uint32_t kFPUInvalidExceptionBit    = 1 << 16;
+
+// FPU rounding modes.
+enum FPURoundingMode {
+  RN = 0,   // Round to Nearest.
+  RZ = 1,   // Round towards zero.
+
+  // Aliases.
+  kRoundToNearest = RN,
+  kRoundToZero = RZ
+};
+
+static const uint32_t kFPURoundingModeMask = 1;
+
+enum CheckForInexactConversion {
+  kCheckForInexactConversion,
+  kDontCheckForInexactConversion
+};
+
+
+// Returns the equivalent of !cc.
+// Negation of the default no_condition (-1) results in a non-default
+// no_condition value (-2). As long as tests for no_condition check
+// for condition < 0, this will work as expected.
+inline Condition NegateCondition(Condition cc) {
+  switch (cc) {
+    case eq:
+      return ne;
+    case ne:
+      return eq;
+    case gt:
+      return le;
+    case ge:
+      return lt;
+    case hi:
+      return us;
+    case hs:
+      return ui;
+    case lt:
+      return ge;
+    case le:
+      return gt;
+    case ui:
+      return hs;
+    case us:
+      return hi;
+    case pl:
+      return qz;
+    case pz:
+      return ql;
+    case ql:
+      return pz;
+    case qz:
+      return pl;
+    default:
+      return cc;
+  }
+}
+
+
+// Corresponds to transposing the operands of a comparison.
+inline Condition ReverseCondition(Condition cc) {
+  UNIMPLEMENTED();
+  return ne;
+}
+
+
+// -----------------------------------------------------------------------------
+// Machine instruction Operands
+
+enum ScaleFactor {
+  times_1 = 0,
+  times_2 = 1,
+  times_4 = 2,
+  times_8 = 3,
+  times_int_size = times_4,
+  times_half_pointer_size = times_2,
+  times_pointer_size = times_4,
+  times_twice_pointer_size = times_8
+};
+
+
+class Operand BASE_EMBEDDED {
+ public:
+  inline explicit Operand(int32_t immediate,
+                          RelocInfo::Mode rmode = RelocInfo::NONE);
+  inline explicit Operand(const ExternalReference& f);
+  inline explicit Operand(Smi* value);
+  inline static Operand Zero() {
+    return Operand(static_cast<int32_t>(0));
+  }
+  explicit Operand(Handle<Object> handle);
+
+  bool is_int8() const {
+    return -128 <= imm32_ && imm32_ < 128 && rmode_ == RelocInfo::NONE;
+  }
+
+ private:
+  int32_t imm32_;
+  RelocInfo::Mode rmode_;
+
+  friend class Assembler;
+};
+
+
+class MemOperand BASE_EMBEDDED {
+ public:
+  INLINE(explicit MemOperand(Register Rd, int32_t offset = 0,
+                             AddrMode mode = Offset));
+  INLINE(explicit MemOperand(Register Rd, Register offset));
+
+  void set_offset(int32_t offset) {
+      ASSERT(rn_.is(no_reg));
+      offset_ = offset;
+  }
+
+  uint32_t offset() const {
+      ASSERT(rn_.is(no_reg));
+      return offset_;
+  }
+
+  Register rn() const { return rn_; }
+  Register rm() const { return rm_; }
+
+ private:
+  Register rm_;
+  Register rn_;
+  int32_t offset_;
+  AddrMode mode_;
+
+  friend class Assembler;
+};
+
+class CpuFeatures : public AllStatic {
+ public:
+  // Detect features of the target CPU. Set safe defaults if the serializer
+  // is enabled (snapshots must be portable).
+  static void Probe();
+
+  // Check whether a feature is supported by the target CPU.
+  static bool IsSupported(CpuFeature f) {
+    ASSERT(initialized_);
+    if (f == FPU && !FLAG_enable_fpu) return false;
+    return (supported_ & (1u << f)) != 0;
+  }
+
+#ifdef DEBUG
+  // Check whether a feature is currently enabled.
+  static bool IsEnabled(CpuFeature f) {
+    ASSERT(initialized_);
+    Isolate* isolate = Isolate::UncheckedCurrent();
+    if (isolate == NULL) {
+      // When no isolate is available, work as if we're running in
+      // release mode.
+      return IsSupported(f);
+    }
+    unsigned enabled = static_cast<unsigned>(isolate->enabled_cpu_features());
+    return (enabled & (1u << f)) != 0;
+  }
+#endif
+
+  // Enable a specified feature within a scope.
+  class Scope BASE_EMBEDDED {
+#ifdef DEBUG
+
+   public:
+    explicit Scope(CpuFeature f) {
+      unsigned mask = 1u << f;
+      ASSERT(CpuFeatures::IsSupported(f));
+      ASSERT(!Serializer::enabled() ||
+             (CpuFeatures::found_by_runtime_probing_ & mask) == 0);
+      isolate_ = Isolate::UncheckedCurrent();
+      old_enabled_ = 0;
+      if (isolate_ != NULL) {
+        old_enabled_ = static_cast<unsigned>(isolate_->enabled_cpu_features());
+        isolate_->set_enabled_cpu_features(old_enabled_ | mask);
+      }
+    }
+    ~Scope() {
+      ASSERT_EQ(Isolate::UncheckedCurrent(), isolate_);
+      if (isolate_ != NULL) {
+        isolate_->set_enabled_cpu_features(old_enabled_);
+      }
+    }
+
+   private:
+    Isolate* isolate_;
+    unsigned old_enabled_;
+#else
+
+   public:
+    explicit Scope(CpuFeature f) {}
+#endif
+  };
+
+  class TryForceFeatureScope BASE_EMBEDDED {
+   public:
+    explicit TryForceFeatureScope(CpuFeature f)
+        : old_supported_(CpuFeatures::supported_) {
+      if (CanForce()) {
+        CpuFeatures::supported_ |= (1u << f);
+      }
+    }
+
+    ~TryForceFeatureScope() {
+      if (CanForce()) {
+        CpuFeatures::supported_ = old_supported_;
+      }
+    }
+
+   private:
+    static bool CanForce() {
+      // It's only safe to temporarily force support of CPU features
+      // when there's only a single isolate, which is guaranteed when
+      // the serializer is enabled.
+      return Serializer::enabled();
+    }
+
+    const unsigned old_supported_;
+  };
+
+ private:
+#ifdef DEBUG
+  static bool initialized_;
+#endif
+  static unsigned supported_;
+  static unsigned found_by_runtime_probing_;
+
+  DISALLOW_COPY_AND_ASSIGN(CpuFeatures);
+};
+
+
+typedef uint16_t Instr;
+
+// Use a target specfic value instead of kZapValue
+const int kSH4ZapValue = 0xbadbaffe;
+
+
+class Assembler : public AssemblerBase {
+ public:
+  // Create an assembler. Instructions and relocation information are emitted
+  // into a buffer, with the instructions starting from the beginning and the
+  // relocation information starting from the end of the buffer. See CodeDesc
+  // for a detailed comment on the layout (globals.h).
+  //
+  // If the provided buffer is NULL, the assembler allocates and grows its own
+  // buffer, and buffer_size determines the initial buffer size. The buffer is
+  // owned by the assembler and deallocated upon destruction of the assembler.
+  //
+  // If the provided buffer is not NULL, the assembler uses the provided buffer
+  // for code generation and assumes its size to be buffer_size. If the buffer
+  // is too small, a fatal error occurs. No deallocation of the buffer is done
+  // upon destruction of the assembler.
+  Assembler(Isolate* isolate, void* buffer, int buffer_size);
+  ~Assembler();
+
+  // Overrides the default provided by FLAG_debug_code.
+  void set_emit_debug_code(bool value) { emit_debug_code_ = value; }
+
+  // Avoids using instructions that vary in size in unpredictable ways between
+  // the snapshot and the running VM.  This is needed by the full compiler so
+  // that it can recompile code with debug support and fix the PC.
+  void set_predictable_code_size(bool value) { predictable_code_size_ = value; }
+
+  // GetCode emits any pending (non-emitted) code and fills the descriptor
+  // desc. GetCode() is idempotent; it returns the same result if no other
+  // Assembler functions are invoked in between GetCode() calls.
+  void GetCode(CodeDesc* desc);
+
+  // Label operations & relative jumps (PPUM Appendix D)
+  //
+  // Takes a branch opcode (cc) and a label (L) and generates
+  // either a backward branch or a forward branch and links it
+  // to the label fixup chain. Usage:
+  //
+  // Label L;    // unbound label
+  // j(cc, &L);  // forward branch to unbound label
+  // bind(&L);   // bind label to the current pc
+  // j(cc, &L);  // backward branch to bound label
+  // bind(&L);   // illegal: a label may be bound only once
+  //
+  // Note: The same Label can be used for forward and backward branches
+  // but it may be bound only once.
+
+  // binds an unbound label L to the current code position
+  void bind(Label* L);
+
+  // Puts a labels target address at the given position.
+  // This function is only used with not-bound labels
+  void load_label(Label* L);
+
+  // Return the address in the constant pool of the code target address used by
+  // the branch/call instruction at pc, or the object in a mov.
+  INLINE(static Address target_pointer_address_at(Address pc));
+
+  // Read/Modify the pointer in the branch/call/move instruction at pc.
+  INLINE(static Address target_pointer_at(Address pc));
+  INLINE(static void set_target_pointer_at(Address pc, Address target));
+
+  // Read/Modify the code target address in the branch/call instruction at pc.
+  INLINE(static Address target_address_at(Address pc));
+  INLINE(static void set_target_address_at(Address pc, Address target));
+
+  // Return the code target address at a call site from the return address
+  // of that call in the instruction stream.
+  INLINE(static Address target_address_from_return_address(Address pc));
+
+  // Given the address of the beginning of a call, return the address
+  // in the instruction stream that the call will return from.
+  INLINE(static Address return_address_from_call_start(Address pc));
+
+  // This sets the branch destination (which is in the constant pool on ARM).
+  // This is for calls and branches within generated code.
+  inline static void deserialization_set_special_target_at(
+      Address constant_pool_entry, Address target) {
+    // When serializing, the object visitor resolves the target_address_address
+    // and stops processing there to recursively serialize another object (or
+    // a reference to it).
+    // Thus when deserializing, the rewriting of targets directly uses the
+    // constant pool address. (same as on ARM)
+    Memory::Address_at(constant_pool_entry) = target;
+  }
+
+  // This sets the branch destination (which is in the constant pool on SH4).
+  // This is for calls and branches to runtime code.
+  inline static void set_external_target_at(Address constant_pool_entry,
+                                            Address target) {
+    // same as above, this function is currently not used anywhere.
+    UNREACHABLE();
+    deserialization_set_special_target_at(constant_pool_entry, target);
+  }
+
+  // Here we are patching the address in the constant pool, not the actual call
+  // instruction.  The address in the constant pool is the same size as a
+  // pointer.
+  static const int kSpecialTargetSize = kPointerSize;
+
+  // Size of an instruction.
+  static const int kInstrSize = sizeof(Instr);
+
+  // Distance between the instruction referring to the address of the call
+  // target and the return address.
+  // The call sequence is:
+  // mov.l const_pool, rx     @ call sequence start address
+  // nop
+  // bra skip
+  // nop
+  // const_pool:
+  // .long call_address
+  // skip:
+  // jsr rx
+  // nop
+  // ...                      @ return address (put in pr by the jsr)
+  static const int kCallTargetAddressOffset = 2 * kInstrSize + 4 +
+                                              4 * kInstrSize;
+
+  // Distance between start of patched return sequence and the emitted address
+  // to jump to.
+  static const int kPatchReturnSequenceAddressOffset = 0 * kInstrSize;
+
+  // Distance between start of patched debug break slot and the emitted address
+  // to jump to.
+  static const int kPatchDebugBreakSlotAddressOffset = 0 * kInstrSize;
+
+  // The debug break slot must be able to contain a call instruction.
+  static const int kDebugBreakSlotLength = kInstrSize;
+
+
+  // branch type
+  enum branch_type {
+    branch_true          = 1 << 1,
+    branch_false         = 1 << 2,
+    branch_unconditional = 1 << 3,
+    branch_subroutine    = 1 << 4
+  };
+
+  static const RegList kAllRegisters = 0xffffffff;
+
+  // Negative pc_offset value (aligned value)
+  static const int kEndOfChain = -4;
+
+
+  // ---------------------------------------------------------------------------
+  // Wrappers around the code generators
+  void add(Register Rd, const Operand& imm, Register rtmp = sh4_rtmp);
+  void add(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp);
+  void add(Register Rd, Register Rs, Register Rt);
+
+  void bt(Label* L, Register rtmp = sh4_rtmp)
+        { ASSERT(!L->is_near_linked());
+          branch(L, rtmp, branch_true); }
+  void bt_near(Label* L, Register rtmp = sh4_rtmp)
+        { ASSERT(L->is_bound() || L->is_unused() || L->is_near_linked());
+          branch(L, rtmp, branch_true, Label::kNear); }
+
+  void bf(Label* L, Register rtmp = sh4_rtmp)
+        { ASSERT(!L->is_near_linked());
+          branch(L, rtmp, branch_false); }
+  void bf_near(Label* L, Register rtmp = sh4_rtmp)
+        { ASSERT(L->is_bound() || L->is_unused() || L->is_near_linked());
+          branch(L, rtmp, branch_false, Label::kNear); }
+
+  void jmp(Label* L, Register rtmp = sh4_rtmp)
+        { ASSERT(!L->is_near_linked());
+          branch(L, rtmp, branch_unconditional); }
+  void jmp_near(Label* L, Register rtmp = sh4_rtmp)
+        { ASSERT(L->is_bound() || L->is_unused() || L->is_near_linked());
+          branch(L, rtmp, branch_unconditional, Label::kNear); }
+
+  void b(Label* L, Register rtmp = sh4_rtmp)
+        { jmp(L, rtmp); }
+  void b_near(Label* L, Register rtmp = sh4_rtmp)
+        { jmp_near(L, rtmp); }
+
+  void b(Condition cond, Label* L, Label::Distance distance = Label::kFar,
+         Register rtmp = sh4_rtmp) {
+    ASSERT((distance == Label::kNear &&
+            (L->is_bound() || L->is_unused() || L->is_near_linked())) ||
+           (distance == Label::kFar  && (!L->is_near_linked())));
+    ASSERT(cond == ne || cond == eq);
+    branch(L, rtmp, cond == eq ? branch_true: branch_false, distance);
+  }
+
+  void jsr(Label* L, Register rtmp = sh4_rtmp)
+        { branch(L, rtmp, branch_subroutine); }
+
+
+  // Check the code size generated from label to here.
+  int SizeOfCodeGeneratedSince(Label* label) {
+    return pc_offset() - label->pos();
+  }
+
+  // Check the number of instructions generated from label to here.
+  int InstructionsGeneratedSince(Label* label) {
+    return SizeOfCodeGeneratedSince(label) / kInstrSize;
+  }
+
+  void jmp(Register Rd);
+  void jsr(Register Rd);
+  void jmp(Handle<Code> code, RelocInfo::Mode rmode, Register rtmp = sh4_rtmp);
+  void jsr(Handle<Code> code, RelocInfo::Mode rmode, Register rtmp = sh4_rtmp);
+
+  void cmpeq(Register Rd, Register Rs) { ASSERT(!Rs.is(Rd));
+                                         cmpeq_(Rs, Rd); }
+  void cmpgt(Register Rd, Register Rs) { ASSERT(!Rs.is(Rd));
+                                         cmpgt_(Rs, Rd); }     // is Rd > Rs ?
+  void cmpge(Register Rd, Register Rs) { ASSERT(!Rs.is(Rd));
+                                         cmpge_(Rs, Rd); }     // is Rd >= Rs ?
+  void cmphi(Register Rd, Register Rs) { ASSERT(!Rs.is(Rd));
+                                         cmphi_(Rs, Rd); }    // is Rd u> Rs ?
+  void cmphs(Register Rd, Register Rs) { ASSERT(!Rs.is(Rd));
+                                         cmphs_(Rs, Rd); }    // is Rd u>= Rs ?
+
+  inline void cmpeq(Register Rd, const Operand& imm,
+                    Register rtmp = sh4_rtmp);
+  inline void cmpgt(Register Rd, const Operand& imm,
+                    Register rtmp = sh4_rtmp);
+  inline void cmpge(Register Rd, const Operand& imm,
+                    Register rtmp = sh4_rtmp);
+  inline void cmphi(Register Rd, const Operand& imm,
+                    Register rtmp = sh4_rtmp);
+  inline void cmphs(Register Rd, const Operand& imm,
+                    Register rtmp = sh4_rtmp);
+
+  // ALiases for cmpeq
+  void cmp(Register Rd, Register Rs) { cmpeq_(Rs, Rd); }
+  void cmp(Register Rd, const Operand& imm, Register rtmp = sh4_rtmp)
+        { cmpeq(Rd, imm, rtmp); }
+
+  inline void cmp(Condition *cond, Register Rd, Register Rs);
+  void cmp(Condition *cond, Register Rd, const Operand& imm,
+           Register rtmp = sh4_rtmp) {
+    mov(rtmp, imm);
+    cmp(cond, Rd, rtmp);
+  }
+
+  void cmpeq_r0_unsigned_imm(int imm) {
+    ASSERT(is_uint8(imm));
+    cmpeq_imm_R0_((int8_t)imm); }
+  bool fits_cmp_unsigned_imm(int imm) { return is_uint8(imm); }
+
+  void dt(Register Rd)  { dt_(Rd); }
+
+  // FPU support
+  // Load float
+  void fldr(SwVfpRegister dst, const MemOperand& src, Register rtmp = sh4_rtmp);
+  // Load double
+  void dldr(DwVfpRegister dst, const MemOperand& src, Register rtmp = sh4_rtmp);
+  // Store float
+  void fstr(SwVfpRegister src, const MemOperand& dst, Register rtmp = sh4_rtmp);
+  // Store double
+  void dstr(DwVfpRegister src, const MemOperand& dst, Register rtmp = sh4_rtmp);
+
+  // Double conversion from register: Dd = (double)Rs
+  void dfloat(DwVfpRegister Dd, Register Rs);
+  // Double conversion from int operand: Dd = (double)imm
+  void dfloat(DwVfpRegister Dd, const Operand &imm, Register rtmp = sh4_rtmp);
+
+  // Double conversion from unsigned int register: Dd = (double)Rs(unsigned)
+  void dufloat(DwVfpRegister Dd, Register Rs, DwVfpRegister drtmp,
+               Register rtmp);
+
+  // Interger conversion from double: Rs = (int)Dd
+  void idouble(Register Rd, DwVfpRegister Ds, Register fpscr = no_reg);
+  // Interger conversion from dingle: Rs = (int)Frs
+  void isingle(Register Rd, SwVfpRegister Frs);
+
+  // Conversion from simple to double
+  void fcnvsd(DwVfpRegister Dd, SwVfpRegister Fs) { flds_FPUL_(Fs);
+                                                    fcnvsd_FPUL_double_(Dd); }
+  // Conversion from double to simple
+  void fcnvds(SwVfpRegister Fd, DwVfpRegister Ds) { fcnvds_double_FPUL_(Ds);
+                                                    fsts_FPUL_(Fd); }
+
+  // Double comparisons
+  void dcmpeq(DwVfpRegister Dd, DwVfpRegister Ds) { fcmpeq_double_(Ds, Dd); }
+  void dcmpgt(DwVfpRegister Dd, DwVfpRegister Ds) { fcmpgt_double_(Ds, Dd); }
+
+  // FPU operations
+  void fadd(DwVfpRegister Dd, DwVfpRegister Ds)   { fadd_double_(Ds, Dd); }
+  void fsub(DwVfpRegister Dd, DwVfpRegister Ds)   { fsub_double_(Ds, Dd); }
+  void fmul(DwVfpRegister Dd, DwVfpRegister Ds)   { fmul_double_(Ds, Dd); }
+  void fdiv(DwVfpRegister Dd, DwVfpRegister Ds)   { fdiv_double_(Ds, Dd); }
+
+  // Read/patch instructions
+  static Instr instr_at(byte* pc)
+        { return *reinterpret_cast<Instr*>(pc); }
+  static void instr_at_put(byte* pc, Instr instr)
+        { *reinterpret_cast<Instr*>(pc) = instr; }
+  static Condition GetCondition(Instr instr);
+  static bool IsBranch(Instr instr);
+  static Register GetRn(Instr instr);
+  static Register GetRm(Instr instr);
+  static bool IsCmpRegister(Instr instr);
+  static bool IsCmpImmediate(Instr instr);
+  static Register GetCmpImmediateRegister(Instr instr);
+  static int GetCmpImmediateAsUnsigned(Instr instr);
+  static bool IsMovImmediate(Instr instr);
+  static bool IsMovlPcRelative(Instr instr)
+      { return (instr & (0xf << 12)) == 0xd000; }
+
+  void sub(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp);
+  void sub(Register Rd, Register Rs, Register Rt);
+
+  // Reverse sub: imm - Rs
+  inline void rsb(Register Rd, Register Rs, const Operand& imm,
+                  Register rtmp = sh4_rtmp);
+  // Reverse sub: Rt - Rs
+  inline void rsb(Register Rd, Register Rs, Register Rt);
+  inline void rsb(Register Rd, Register Rs, const Operand& imm,
+                  Condition cond, Register rtmp = sh4_rtmp);
+
+  void addv(Register Rd, Register Rs, Register Rt);
+  void addv(Register Rd, Register Rs, const Operand& imm,
+            Register rtmp = sh4_rtmp);
+  void subv(Register Rd, Register Rs, Register Rt, Register rtmp = sh4_rtmp);
+
+  void addc(Register Rd, Register Rs, Register Rt);
+  void subc(Register Rd, Register Rs, Register Rt, Register rtmp = sh4_rtmp);
+
+  // Note for shifts with shift amount in register:
+  // the default behavior is mapped on ARM behavior which flushes when shift
+  // amount is >= 32. The implementation is quite slow in this case as SH4
+  // shifts do not flush.
+  // In case where the register shift amount is known to be in range [0,31] one
+  // can call the shift methods with in_range parameter set to true. This case
+  // generates faster code.
+  // In addition, in the case of lsl (but not asr nor lsr) the semantic is to
+  // wrap on SH4 (i.e. the least 5 bits are extracted before doing the shift),
+  // thus in this case the boolean parameter is called wrap and this semantic
+  // can be used on purpose whatever the shift amount.
+
+  // arithmetic shift right
+  void asr(Register Rd, Register Rs, Register Rt, bool in_range = false,
+           Register rtmp = sh4_rtmp);
+  void asr(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp);
+  // arithmetic shift left
+  void asl(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp);
+
+  void lsl(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp);
+  void lsl(Register Rd, Register Rs, Register Rt, bool wrap = false,
+           Register rtmp = sh4_rtmp);
+  void lsr(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp);
+  void lsr(Register Rd, Register Rs, Register Rt, bool in_range = false,
+           Register rtmp = sh4_rtmp);
+
+  void land(Register Rd, Register Rs, const Operand& imm,
+            Register rtmp = sh4_rtmp);
+  void land(Register Rd, Register Rs, Register Rt);
+
+  // bit clear
+  void bic(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp)
+        { land(Rd, Rs, Operand(~imm.imm32_), rtmp); }
+  void bic(Register Rd, Register Rs, Register Rt, Register rtmp = sh4_rtmp) {
+    lnot(rtmp, Rt);
+    land(Rd, Rs, rtmp);
+  }
+
+  void lnot(Register Rd, Register Rs) { not_(Rs, Rd); }
+  void mvn(Register Rd, Register Rs)  { lnot(Rd, Rs); }  // Alias for lnot()
+
+  void lor(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp);
+  void lor(Register Rd, Register Rs, Register Rt);
+  void lor(Register Rd, Register Rs, const Operand& imm, Condition cond,
+           Register rtmp = sh4_rtmp);
+  void lor(Register Rd, Register Rs, Register Rt, Condition cond);
+
+  void lxor(Register Rd, Register Rs, const Operand& imm,
+            Register rtmp = sh4_rtmp);
+  void lxor(Register Rd, Register Rs, Register Rt);
+
+  // Aliases for lxor
+  void eor(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp) { lxor(Rd, Rs, imm, rtmp); }
+  void eor(Register Rd, Register Rs, Register Rt)  { lxor(Rd, Rs, Rt); }
+
+  // Aliases for lor
+  void orr(Register Rd, Register Rs, const Operand& imm,
+           Register rtmp = sh4_rtmp) { lor(Rd, Rs, imm, rtmp); }
+  void orr(Register Rd, Register Rs, Register Rt)  { lor(Rd, Rs, Rt); }
+  void orr(Register Rd, Register Rs, const Operand& imm,
+           Condition cond, Register rtmp = sh4_rtmp)
+        { lor(Rd, Rs, imm, cond, rtmp); }
+  void orr(Register Rd, Register Rs, Register Rt, Condition cond)
+        { lor(Rd, Rs, Rt, cond); }
+
+  void tst(Register Rd, Register Rs) { tst_(Rs, Rd); }
+  void tst(Register Rd, const Operand& imm, Register rtmp = sh4_rtmp);
+
+  void teq(Register Rd, const Operand& imm, Register rtmp = sh4_rtmp) {
+    lxor(rtmp, Rd, imm);
+    tst(rtmp, rtmp);
+  }
+  void teq(Register Rd, Register Rs, Register rtmp = sh4_rtmp) {
+    lxor(rtmp, Rd, Rs);
+    tst(rtmp, rtmp);
+  }
+
+  // Moves and conditional moves.
+  // This one allows pr as src or dst.
+  void mov(Register Rd, Register Rs, Condition cond = al);
+  void mov(Register Rd, const Operand& src, bool force = false);
+  void mov(Register Rd, const Operand& imm, Condition cond);
+
+  // load op.
+  void mov(Register Rd, const MemOperand& src, Register rtmp = sh4_rtmp);
+  // unsigned 8 bit load op.
+  void movb(Register Rd, const MemOperand& src, Register rtmp = sh4_rtmp);
+  // unsigned 16 bit load op.
+  void movw(Register Rd, const MemOperand& src, Register rtmp = sh4_rtmp);
+  // store op.
+  void mov(const MemOperand& dst, Register Rd, Register rtmp = sh4_rtmp);
+  // store 8 bits op.
+  void movb(const MemOperand& dst, Register Rd, Register rtmp = sh4_rtmp);
+  // store 16 bits op.
+  void movw(const MemOperand& dst, Register Rd, Register rtmp = sh4_rtmp);
+
+  void movd(DwVfpRegister Dd, Register Rs1, Register Rs2);
+  void movd(Register Rd1, Register Rd2, DwVfpRegister Ds);
+
+  inline void ldr(Register Rd, const MemOperand& src,
+                  Register rtmp = sh4_rtmp);
+  void ldrb(Register Rd, const MemOperand& src, Register rtmp = sh4_rtmp)
+        { movb(Rd, src, rtmp); }
+  void ldrh(Register Rd, const MemOperand& src, Register rtmp = sh4_rtmp)
+        { movw(Rd, src, rtmp); }
+  // signed 8 bit load op.
+  void ldrsb(Register Rd, const MemOperand& src, Register rtmp = sh4_rtmp);
+  // signed 16 bit load op.
+  void ldrsh(Register Rd, const MemOperand& src, Register rtmp = sh4_rtmp);
+
+  inline void str(Register Rs, const MemOperand& dst,
+                  Register rtmp = sh4_rtmp);
+  void strh(Register Rs, const MemOperand& dst, Register rtmp = sh4_rtmp)
+        { movw(dst, Rs, rtmp); }
+  void strb(Register Rs, const MemOperand& dst, Register rtmp = sh4_rtmp)
+        { movb(dst, Rs, rtmp); }
+
+  void ldrpr(Register Rd) { lds_PR_(Rd); }
+  void strpr(Register Rs) { sts_PR_(Rs); }
+
+  void ldr_fpscr(Register Rs) { lds_FPSCR_(Rs); }
+  void str_fpscr(Register Rd) { sts_FPSCR_(Rd); }
+
+  void mul(Register Rd, Register Rs, Register Rt);
+  void dmuls(Register dstL, Register dstH, Register src1, Register src2);
+
+  void nop() { nop_(); }
+
+  void push(Register src);
+  void push(DwVfpRegister src);
+  // push an immediate on the stack: use rtmp register for that
+  void push(const Operand& op, Register rtmp = sh4_rtmp);
+  void pushm(RegList src, bool doubles = false);
+
+  void pop(Register dst);
+  void pop() { add(sp, sp, Operand(kPointerSize)); }
+  void pop(DwVfpRegister dst);
+  void popm(RegList dst, bool doubles = false);
+
+  inline void rts();
+
+  // Exception-generating instructions and debugging support
+  void stop(const char* msg);
+  void bkpt();
+
+  // Align the code
+  inline int align();
+  inline int misalign();
+
+  // Copy some bytes
+  // The count argument is scratched
+  void memcpy(Register dst, Register src, Register count,
+              Register scratch1, Register scratch2,
+              Register scratch3, Register scratch4);
+
+  // Compare some bytes using a loop
+  void memcmp(Register left, Register right, Register length,
+              Register scratch1, Register scratch2, Label *not_equal);
+
+  bool predictable_code_size() const { return predictable_code_size_; }
+
+  // Insert the smallest number of nop instructions
+  // possible to align the pc offset to a multiple
+  // of m. m must be a power of 2 (>= 4).
+  void Align(int m);
+
+  void call(Label* L);
+
+  inline void emit(Instr x);
+
+  // Mark address of the ExitJSFrame code.
+  void RecordJSReturn();
+
+  // Record the AST id of the CallIC being compiled, so that it can be placed
+  // in the relocation information.
+  void SetRecordedAstId(TypeFeedbackId ast_id) {
+    ASSERT(recorded_ast_id_.IsNone());
+    recorded_ast_id_ = ast_id;
+  }
+
+  TypeFeedbackId RecordedAstId() {
+    ASSERT(!recorded_ast_id_.IsNone());
+    return recorded_ast_id_;
+  }
+
+  void ClearRecordedAstId() { recorded_ast_id_ = TypeFeedbackId::None(); }
+
+  // Use -code_comments to enable, or provide "force = true" flag to always
+  // write a comment.
+  void RecordComment(const char* msg, bool force = false);
+
+  // Writes a single byte or word of data in the code stream.  Used for
+  // inline tables, e.g., jump-tables.
+  void db(uint8_t data);
+  void dw(uint16_t data);
+  void dd(uint32_t data);
+
+  int pc_offset() const { return pc_ - buffer_; }
+
+  // Return in Rd the value of pc_after + offset.
+  // Where pc_after is the pc after this operation.
+  // It clobbers pr which must be always passed in the Pr parameter
+  void addpc(Register Rd, int offset, Register Pr);
+
+  // Check if there is less than kGap bytes available in the buffer.
+  // If this is the case, we need to grow the buffer before emitting
+  // an instruction or relocation information.
+  inline bool overflow() const { return pc_ >= reloc_info_writer.pos() - kGap; }
+
+  // Get the number of bytes available in the buffer.
+  inline int available_space() const { return reloc_info_writer.pos() - pc_; }
+
+  PositionsRecorder* positions_recorder() { return &positions_recorder_; }
+
+
+ protected:
+  // Relocation for a type-recording IC has the AST id added to it.  This
+  // member variable is a way to pass the information from the call site to
+  // the relocation info.
+  TypeFeedbackId recorded_ast_id_;
+
+  bool emit_debug_code() const { return emit_debug_code_; }
+
+  int buffer_space() const { return reloc_info_writer.pos() - pc_; }
+
+
+ private:
+  // code generation wrappers
+  void branch(Label* L, Register rtmp, branch_type type,
+              Label::Distance distance = Label::kFar);
+  void branch(int offset, Register rtmp, branch_type type,
+              Label::Distance distance, bool patched_later);
+  void conditional_branch(int offset, Register rtmp, Label::Distance distance,
+                          bool patched_later, bool type);
+  void jmp(int offset, Register rtmp, Label::Distance distance,
+           bool patched_later);
+  void jsr(int offset, Register rtmp, bool patched_later);
+
+  void writeBranchTag(int nop_count, branch_type type);
+  void patchBranchOffset(int fixup_pos, uint16_t *p_pos, int is_near_linked);
+
+  // The bound position, before this we cannot do instruction elimination.
+  int last_bound_pos_;
+
+  // Code emission
+  inline void CheckBuffer();
+  void GrowBuffer();
+
+  void next(Label *L, Label::Distance distance = Label::kFar);
+
+  // record reloc info for current pc_
+  void RecordRelocInfo(RelocInfo::Mode rmode, intptr_t data = 0);
+
+  friend class CodePatcher;
+  friend class EnsureSpace;
+
+  // Code buffer:
+  // The buffer into which code and relocation info are generated.
+  byte* buffer_;
+  int buffer_size_;
+  // True if the assembler owns the buffer, false if buffer is external.
+  bool own_buffer_;
+
+  // Code generation
+  // The relocation writer's position is at least kGap bytes below the end of
+  // the generated instructions. This is so that multi-instruction sequences do
+  // not have to check for overflow. The same is true for writes of large
+  // relocation info entries.
+  static const int kGap = 32;
+  byte* pc_;  // the program counter; moves forward
+  RelocInfoWriter reloc_info_writer;
+
+  PositionsRecorder positions_recorder_;
+
+  bool emit_debug_code_;
+  bool predictable_code_size_;
+
+  friend class PositionsRecorder;
+
+
+  // ---------------------------------------------------------------------------
+  // low level code generation (opcodes)
+  #include "opcodes-sh4.h"
+};
+
+
+// Helper class that ensures that there is enough space for generating
+// instructions and relocation information.  The constructor makes
+// sure that there is enough space and (in debug mode) the destructor
+// checks that we did not generate too much.
+class EnsureSpace BASE_EMBEDDED {
+ public:
+  explicit EnsureSpace(Assembler* assembler) : assembler_(assembler) {
+    if (assembler_->overflow()) assembler_->GrowBuffer();
+#ifdef DEBUG
+    space_before_ = assembler_->available_space();
+#endif
+  }
+
+#ifdef DEBUG
+  ~EnsureSpace() {
+    int bytes_generated = space_before_ - assembler_->available_space();
+    ASSERT(bytes_generated < assembler_->kGap);
+  }
+#endif
+
+ private:
+  Assembler* assembler_;
+#ifdef DEBUG
+  int space_before_;
+#endif
+};
+
+} }  // namespace v8::internal
+
+#endif  // V8_SH4_ASSEMBLER_SH4_H_