Upgrade V8 to 5.1.281.57  DO NOT MERGE

FPIIM-449

Change-Id: Id981b686b4d587ac31697662eb98bb34be42ad90
(cherry picked from commit 3b9bc31999c9787eb726ecdbfd5796bfdec32a18)
diff --git a/src/s390/simulator-s390.h b/src/s390/simulator-s390.h
new file mode 100644
index 0000000..ae3dd58
--- /dev/null
+++ b/src/s390/simulator-s390.h
@@ -0,0 +1,552 @@
+// 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.
+
+// Declares a Simulator for S390 instructions if we are not generating a native
+// S390 binary. This Simulator allows us to run and debug S390 code generation
+// on regular desktop machines.
+// V8 calls into generated code by "calling" the CALL_GENERATED_CODE macro,
+// which will start execution in the Simulator or forwards to the real entry
+// on a S390 hardware platform.
+
+#ifndef V8_S390_SIMULATOR_S390_H_
+#define V8_S390_SIMULATOR_S390_H_
+
+#include "src/allocation.h"
+
+#if !defined(USE_SIMULATOR)
+// Running without a simulator on a native s390 platform.
+
+namespace v8 {
+namespace internal {
+
+// When running without a simulator we call the entry directly.
+#define CALL_GENERATED_CODE(isolate, entry, p0, p1, p2, p3, p4) \
+  (entry(p0, p1, p2, p3, p4))
+
+typedef int (*s390_regexp_matcher)(String*, int, const byte*, const byte*, int*,
+                                   int, Address, int, void*, Isolate*);
+
+// Call the generated regexp code directly. The code at the entry address
+// should act as a function matching the type ppc_regexp_matcher.
+// The ninth argument is a dummy that reserves the space used for
+// the return address added by the ExitFrame in native calls.
+#define CALL_GENERATED_REGEXP_CODE(isolate, entry, p0, p1, p2, p3, p4, p5, p6, \
+                                   p7, p8)                                     \
+  (FUNCTION_CAST<s390_regexp_matcher>(entry)(p0, p1, p2, p3, p4, p5, p6, p7,   \
+                                             NULL, p8))
+
+// The stack limit beyond which we will throw stack overflow errors in
+// generated code. Because generated code on s390 uses the C stack, we
+// just use the C stack limit.
+class SimulatorStack : public v8::internal::AllStatic {
+ public:
+  static inline uintptr_t JsLimitFromCLimit(v8::internal::Isolate* isolate,
+                                            uintptr_t c_limit) {
+    USE(isolate);
+    return c_limit;
+  }
+
+  static inline uintptr_t RegisterCTryCatch(v8::internal::Isolate* isolate,
+                                            uintptr_t try_catch_address) {
+    USE(isolate);
+    return try_catch_address;
+  }
+
+  static inline void UnregisterCTryCatch(v8::internal::Isolate* isolate) {
+    USE(isolate);
+  }
+};
+}  // namespace internal
+}  // namespace v8
+
+#else  // !defined(USE_SIMULATOR)
+// Running with a simulator.
+
+#include "src/assembler.h"
+#include "src/hashmap.h"
+#include "src/s390/constants-s390.h"
+
+namespace v8 {
+namespace internal {
+
+class CachePage {
+ public:
+  static const int LINE_VALID = 0;
+  static const int LINE_INVALID = 1;
+
+  static const int kPageShift = 12;
+  static const int kPageSize = 1 << kPageShift;
+  static const int kPageMask = kPageSize - 1;
+  static const int kLineShift = 2;  // The cache line is only 4 bytes right now.
+  static const int kLineLength = 1 << kLineShift;
+  static const int kLineMask = kLineLength - 1;
+
+  CachePage() { memset(&validity_map_, LINE_INVALID, sizeof(validity_map_)); }
+
+  char* ValidityByte(int offset) {
+    return &validity_map_[offset >> kLineShift];
+  }
+
+  char* CachedData(int offset) { return &data_[offset]; }
+
+ private:
+  char data_[kPageSize];  // The cached data.
+  static const int kValidityMapSize = kPageSize >> kLineShift;
+  char validity_map_[kValidityMapSize];  // One byte per line.
+};
+
+class Simulator {
+ public:
+  friend class S390Debugger;
+  enum Register {
+    no_reg = -1,
+    r0 = 0,
+    r1 = 1,
+    r2 = 2,
+    r3 = 3,
+    r4 = 4,
+    r5 = 5,
+    r6 = 6,
+    r7 = 7,
+    r8 = 8,
+    r9 = 9,
+    r10 = 10,
+    r11 = 11,
+    r12 = 12,
+    r13 = 13,
+    r14 = 14,
+    r15 = 15,
+    fp = r11,
+    ip = r12,
+    cp = r13,
+    ra = r14,
+    sp = r15,  // name aliases
+    kNumGPRs = 16,
+    d0 = 0,
+    d1,
+    d2,
+    d3,
+    d4,
+    d5,
+    d6,
+    d7,
+    d8,
+    d9,
+    d10,
+    d11,
+    d12,
+    d13,
+    d14,
+    d15,
+    kNumFPRs = 16
+  };
+
+  explicit Simulator(Isolate* isolate);
+  ~Simulator();
+
+  // The currently executing Simulator instance. Potentially there can be one
+  // for each native thread.
+  static Simulator* current(v8::internal::Isolate* isolate);
+
+  // Accessors for register state.
+  void set_register(int reg, uint64_t value);
+  uint64_t get_register(int reg) const;
+  template <typename T>
+  T get_low_register(int reg) const;
+  template <typename T>
+  T get_high_register(int reg) const;
+  void set_low_register(int reg, uint32_t value);
+  void set_high_register(int reg, uint32_t value);
+
+  double get_double_from_register_pair(int reg);
+  void set_d_register_from_double(int dreg, const double dbl) {
+    DCHECK(dreg >= 0 && dreg < kNumFPRs);
+    *bit_cast<double*>(&fp_registers_[dreg]) = dbl;
+  }
+
+  double get_double_from_d_register(int dreg) {
+    DCHECK(dreg >= 0 && dreg < kNumFPRs);
+    return *bit_cast<double*>(&fp_registers_[dreg]);
+  }
+  void set_d_register(int dreg, int64_t value) {
+    DCHECK(dreg >= 0 && dreg < kNumFPRs);
+    fp_registers_[dreg] = value;
+  }
+  int64_t get_d_register(int dreg) {
+    DCHECK(dreg >= 0 && dreg < kNumFPRs);
+    return fp_registers_[dreg];
+  }
+
+  void set_d_register_from_float32(int dreg, const float f) {
+    DCHECK(dreg >= 0 && dreg < kNumFPRs);
+
+    int32_t f_int = *bit_cast<int32_t*>(&f);
+    int64_t finalval = static_cast<int64_t>(f_int) << 32;
+    set_d_register(dreg, finalval);
+  }
+
+  float get_float32_from_d_register(int dreg) {
+    DCHECK(dreg >= 0 && dreg < kNumFPRs);
+
+    int64_t regval = get_d_register(dreg) >> 32;
+    int32_t regval32 = static_cast<int32_t>(regval);
+    return *bit_cast<float*>(&regval32);
+  }
+
+  // Special case of set_register and get_register to access the raw PC value.
+  void set_pc(intptr_t value);
+  intptr_t get_pc() const;
+
+  Address get_sp() const {
+    return reinterpret_cast<Address>(static_cast<intptr_t>(get_register(sp)));
+  }
+
+  // Accessor to the internal simulator stack area.
+  uintptr_t StackLimit(uintptr_t c_limit) const;
+
+  // Executes S390 instructions until the PC reaches end_sim_pc.
+  void Execute();
+
+  // Call on program start.
+  static void Initialize(Isolate* isolate);
+
+  static void TearDown(HashMap* i_cache, Redirection* first);
+
+  // V8 generally calls into generated JS code with 5 parameters and into
+  // generated RegExp code with 7 parameters. This is a convenience function,
+  // which sets up the simulator state and grabs the result on return.
+  intptr_t Call(byte* entry, int argument_count, ...);
+  // Alternative: call a 2-argument double function.
+  void CallFP(byte* entry, double d0, double d1);
+  int32_t CallFPReturnsInt(byte* entry, double d0, double d1);
+  double CallFPReturnsDouble(byte* entry, double d0, double d1);
+
+  // Push an address onto the JS stack.
+  uintptr_t PushAddress(uintptr_t address);
+
+  // Pop an address from the JS stack.
+  uintptr_t PopAddress();
+
+  // Debugger input.
+  void set_last_debugger_input(char* input);
+  char* last_debugger_input() { return last_debugger_input_; }
+
+  // ICache checking.
+  static void FlushICache(v8::internal::HashMap* i_cache, void* start,
+                          size_t size);
+
+  // Returns true if pc register contains one of the 'special_values' defined
+  // below (bad_lr, end_sim_pc).
+  bool has_bad_pc() const;
+
+ private:
+  enum special_values {
+    // Known bad pc value to ensure that the simulator does not execute
+    // without being properly setup.
+    bad_lr = -1,
+    // A pc value used to signal the simulator to stop execution.  Generally
+    // the lr is set to this value on transition from native C code to
+    // simulated execution, so that the simulator can "return" to the native
+    // C code.
+    end_sim_pc = -2
+  };
+
+  // Unsupported instructions use Format to print an error and stop execution.
+  void Format(Instruction* instr, const char* format);
+
+  // Helper functions to set the conditional flags in the architecture state.
+  bool CarryFrom(int32_t left, int32_t right, int32_t carry = 0);
+  bool BorrowFrom(int32_t left, int32_t right);
+  template <typename T1>
+  inline bool OverflowFromSigned(T1 alu_out, T1 left, T1 right, bool addition);
+
+  // Helper functions to decode common "addressing" modes
+  int32_t GetShiftRm(Instruction* instr, bool* carry_out);
+  int32_t GetImm(Instruction* instr, bool* carry_out);
+  void ProcessPUW(Instruction* instr, int num_regs, int operand_size,
+                  intptr_t* start_address, intptr_t* end_address);
+  void HandleRList(Instruction* instr, bool load);
+  void HandleVList(Instruction* inst);
+  void SoftwareInterrupt(Instruction* instr);
+
+  // Stop helper functions.
+  inline bool isStopInstruction(Instruction* instr);
+  inline bool isWatchedStop(uint32_t bkpt_code);
+  inline bool isEnabledStop(uint32_t bkpt_code);
+  inline void EnableStop(uint32_t bkpt_code);
+  inline void DisableStop(uint32_t bkpt_code);
+  inline void IncreaseStopCounter(uint32_t bkpt_code);
+  void PrintStopInfo(uint32_t code);
+
+  // Byte Reverse
+  inline int16_t ByteReverse(int16_t hword);
+  inline int32_t ByteReverse(int32_t word);
+
+  // Read and write memory.
+  inline uint8_t ReadBU(intptr_t addr);
+  inline int8_t ReadB(intptr_t addr);
+  inline void WriteB(intptr_t addr, uint8_t value);
+  inline void WriteB(intptr_t addr, int8_t value);
+
+  inline uint16_t ReadHU(intptr_t addr, Instruction* instr);
+  inline int16_t ReadH(intptr_t addr, Instruction* instr);
+  // Note: Overloaded on the sign of the value.
+  inline void WriteH(intptr_t addr, uint16_t value, Instruction* instr);
+  inline void WriteH(intptr_t addr, int16_t value, Instruction* instr);
+
+  inline uint32_t ReadWU(intptr_t addr, Instruction* instr);
+  inline int32_t ReadW(intptr_t addr, Instruction* instr);
+  inline void WriteW(intptr_t addr, uint32_t value, Instruction* instr);
+  inline void WriteW(intptr_t addr, int32_t value, Instruction* instr);
+
+  inline int64_t ReadDW(intptr_t addr);
+  inline double ReadDouble(intptr_t addr);
+  inline void WriteDW(intptr_t addr, int64_t value);
+
+  // S390
+  void Trace(Instruction* instr);
+  bool DecodeTwoByte(Instruction* instr);
+  bool DecodeFourByte(Instruction* instr);
+  bool DecodeFourByteArithmetic(Instruction* instr);
+  bool DecodeFourByteArithmetic64Bit(Instruction* instr);
+  bool DecodeFourByteFloatingPoint(Instruction* instr);
+  void DecodeFourByteFloatingPointIntConversion(Instruction* instr);
+  void DecodeFourByteFloatingPointRound(Instruction* instr);
+
+  bool DecodeSixByte(Instruction* instr);
+  bool DecodeSixByteArithmetic(Instruction* instr);
+  bool S390InstructionDecode(Instruction* instr);
+  void DecodeSixByteBitShift(Instruction* instr);
+
+  // Used by the CL**BR instructions.
+  template <typename T1, typename T2>
+  void SetS390RoundConditionCode(T1 r2_val, T2 max, T2 min) {
+    condition_reg_ = 0;
+    double r2_dval = static_cast<double>(r2_val);
+    double dbl_min = static_cast<double>(min);
+    double dbl_max = static_cast<double>(max);
+
+    if (r2_dval == 0.0)
+      condition_reg_ = 8;
+    else if (r2_dval < 0.0 && r2_dval >= dbl_min && std::isfinite(r2_dval))
+      condition_reg_ = 4;
+    else if (r2_dval > 0.0 && r2_dval <= dbl_max && std::isfinite(r2_dval))
+      condition_reg_ = 2;
+    else
+      condition_reg_ = 1;
+  }
+
+  template <typename T1>
+  void SetS390RoundConditionCode(T1 r2_val, int64_t max, int64_t min) {
+    condition_reg_ = 0;
+    double r2_dval = static_cast<double>(r2_val);
+    double dbl_min = static_cast<double>(min);
+    double dbl_max = static_cast<double>(max);
+
+    // Note that the IEEE 754 floating-point representations (both 32 and
+    // 64 bit) cannot exactly represent INT64_MAX. The closest it can get
+    // is INT64_max + 1. IEEE 754 FP can, though, represent INT64_MIN
+    // exactly.
+
+    // This is not an issue for INT32, as IEEE754 64-bit can represent
+    // INT32_MAX and INT32_MIN with exact precision.
+
+    if (r2_dval == 0.0)
+      condition_reg_ = 8;
+    else if (r2_dval < 0.0 && r2_dval >= dbl_min && std::isfinite(r2_dval))
+      condition_reg_ = 4;
+    else if (r2_dval > 0.0 && r2_dval < dbl_max && std::isfinite(r2_dval))
+      condition_reg_ = 2;
+    else
+      condition_reg_ = 1;
+  }
+
+  // Used by the CL**BR instructions.
+  template <typename T1, typename T2, typename T3>
+  void SetS390ConvertConditionCode(T1 src, T2 dst, T3 max) {
+    condition_reg_ = 0;
+    if (src == static_cast<T1>(0.0)) {
+      condition_reg_ |= 8;
+    } else if (src < static_cast<T1>(0.0) && static_cast<T2>(src) == 0 &&
+               std::isfinite(src)) {
+      condition_reg_ |= 4;
+    } else if (src > static_cast<T1>(0.0) && std::isfinite(src) &&
+               src < static_cast<T1>(max)) {
+      condition_reg_ |= 2;
+    } else {
+      condition_reg_ |= 1;
+    }
+  }
+
+  template <typename T>
+  void SetS390ConditionCode(T lhs, T rhs) {
+    condition_reg_ = 0;
+    if (lhs == rhs) {
+      condition_reg_ |= CC_EQ;
+    } else if (lhs < rhs) {
+      condition_reg_ |= CC_LT;
+    } else if (lhs > rhs) {
+      condition_reg_ |= CC_GT;
+    }
+
+    // We get down here only for floating point
+    // comparisons and the values are unordered
+    // i.e. NaN
+    if (condition_reg_ == 0) condition_reg_ = unordered;
+  }
+
+  // Used by arithmetic operations that use carry.
+  template <typename T>
+  void SetS390ConditionCodeCarry(T result, bool overflow) {
+    condition_reg_ = 0;
+    bool zero_result = (result == static_cast<T>(0));
+    if (zero_result && !overflow) {
+      condition_reg_ |= 8;
+    } else if (!zero_result && !overflow) {
+      condition_reg_ |= 4;
+    } else if (zero_result && overflow) {
+      condition_reg_ |= 2;
+    } else if (!zero_result && overflow) {
+      condition_reg_ |= 1;
+    }
+    if (condition_reg_ == 0) UNREACHABLE();
+  }
+
+  bool isNaN(double value) { return (value != value); }
+
+  // Set the condition code for bitwise operations
+  // CC0 is set if value == 0.
+  // CC1 is set if value != 0.
+  // CC2/CC3 are not set.
+  template <typename T>
+  void SetS390BitWiseConditionCode(T value) {
+    condition_reg_ = 0;
+
+    if (value == 0)
+      condition_reg_ |= CC_EQ;
+    else
+      condition_reg_ |= CC_LT;
+  }
+
+  void SetS390OverflowCode(bool isOF) {
+    if (isOF) condition_reg_ = CC_OF;
+  }
+
+  bool TestConditionCode(Condition mask) {
+    // Check for unconditional branch
+    if (mask == 0xf) return true;
+
+    return (condition_reg_ & mask) != 0;
+  }
+
+  // Executes one instruction.
+  void ExecuteInstruction(Instruction* instr, bool auto_incr_pc = true);
+
+  // ICache.
+  static void CheckICache(v8::internal::HashMap* i_cache, Instruction* instr);
+  static void FlushOnePage(v8::internal::HashMap* i_cache, intptr_t start,
+                           int size);
+  static CachePage* GetCachePage(v8::internal::HashMap* i_cache, void* page);
+
+  // Runtime call support.
+  static void* RedirectExternalReference(
+      Isolate* isolate, void* external_function,
+      v8::internal::ExternalReference::Type type);
+
+  // Handle arguments and return value for runtime FP functions.
+  void GetFpArgs(double* x, double* y, intptr_t* z);
+  void SetFpResult(const double& result);
+  void TrashCallerSaveRegisters();
+
+  void CallInternal(byte* entry, int reg_arg_count = 3);
+
+  // Architecture state.
+  // On z9 and higher and supported Linux on z Systems platforms, all registers
+  // are 64-bit, even in 31-bit mode.
+  uint64_t registers_[kNumGPRs];
+  int64_t fp_registers_[kNumFPRs];
+
+  // Condition Code register. In S390, the last 4 bits are used.
+  int32_t condition_reg_;
+  // Special register to track PC.
+  intptr_t special_reg_pc_;
+
+  // Simulator support.
+  char* stack_;
+  static const size_t stack_protection_size_ = 256 * kPointerSize;
+  bool pc_modified_;
+  int64_t icount_;
+
+  // Debugger input.
+  char* last_debugger_input_;
+
+  // Icache simulation
+  v8::internal::HashMap* i_cache_;
+
+  // Registered breakpoints.
+  Instruction* break_pc_;
+  Instr break_instr_;
+
+  v8::internal::Isolate* isolate_;
+
+  // A stop is watched if its code is less than kNumOfWatchedStops.
+  // Only watched stops support enabling/disabling and the counter feature.
+  static const uint32_t kNumOfWatchedStops = 256;
+
+  // Breakpoint is disabled if bit 31 is set.
+  static const uint32_t kStopDisabledBit = 1 << 31;
+
+  // A stop is enabled, meaning the simulator will stop when meeting the
+  // instruction, if bit 31 of watched_stops_[code].count is unset.
+  // The value watched_stops_[code].count & ~(1 << 31) indicates how many times
+  // the breakpoint was hit or gone through.
+  struct StopCountAndDesc {
+    uint32_t count;
+    char* desc;
+  };
+  StopCountAndDesc watched_stops_[kNumOfWatchedStops];
+  void DebugStart();
+};
+
+// When running with the simulator transition into simulated execution at this
+// point.
+#define CALL_GENERATED_CODE(isolate, entry, p0, p1, p2, p3, p4)          \
+  reinterpret_cast<Object*>(Simulator::current(isolate)->Call(           \
+      FUNCTION_ADDR(entry), 5, (intptr_t)p0, (intptr_t)p1, (intptr_t)p2, \
+      (intptr_t)p3, (intptr_t)p4))
+
+#define CALL_GENERATED_REGEXP_CODE(isolate, entry, p0, p1, p2, p3, p4, p5, p6, \
+                                   p7, p8)                                     \
+  Simulator::current(isolate)->Call(entry, 10, (intptr_t)p0, (intptr_t)p1,     \
+                                    (intptr_t)p2, (intptr_t)p3, (intptr_t)p4,  \
+                                    (intptr_t)p5, (intptr_t)p6, (intptr_t)p7,  \
+                                    (intptr_t)NULL, (intptr_t)p8)
+
+// The simulator has its own stack. Thus it has a different stack limit from
+// the C-based native code.  The JS-based limit normally points near the end of
+// the simulator stack.  When the C-based limit is exhausted we reflect that by
+// lowering the JS-based limit as well, to make stack checks trigger.
+class SimulatorStack : public v8::internal::AllStatic {
+ public:
+  static inline uintptr_t JsLimitFromCLimit(v8::internal::Isolate* isolate,
+                                            uintptr_t c_limit) {
+    return Simulator::current(isolate)->StackLimit(c_limit);
+  }
+
+  static inline uintptr_t RegisterCTryCatch(v8::internal::Isolate* isolate,
+                                            uintptr_t try_catch_address) {
+    Simulator* sim = Simulator::current(isolate);
+    return sim->PushAddress(try_catch_address);
+  }
+
+  static inline void UnregisterCTryCatch(v8::internal::Isolate* isolate) {
+    Simulator::current(isolate)->PopAddress();
+  }
+};
+
+}  // namespace internal
+}  // namespace v8
+
+#endif  // !defined(USE_SIMULATOR)
+#endif  // V8_S390_SIMULATOR_S390_H_