com.cmpware.cmp.models
Class PowerPC64

java.lang.Object
  com.cmpware.cmp.Memory
      com.cmpware.cmp.Processor
          com.cmpware.tools.AutoModel
              com.cmpware.cmp.models.PowerPC64

All Implemented Interfaces:

ReadWriteInterface

public class PowerPC64
extends AutoModel

This is an AutoModel implementation of the PowerPC 64 bit processor istructions. It implements all of the standard PowerPC Instruction Set Architecture. In particular, no extensions such as floating point, embedded or the non-portable extensions are implemented in this version, although they can be added fairly easily to this model. The large number of special registers and the unusual addressing are perhaps the biggest complication in the PowerPC model. Only the registers actually used in the implemented instructions are implemented, all others are ignored. This information is taken from the IBM document "PPC405Fx Embedded Processor Core: User's Manual (Preliminary)", IBM Document Number SA14-2764-00, January 28, 2005. PowerPC 64-bit instructions added using the document: "Power ISA" version 2.04, April 3, 2007. Implementation notes: The document specifies Big Endian, but uses bit 0 as the MSB. Field numbers in this model will be (31-n) compared to the IBM document. BCA and BCL decodes swapped. Fixed 4/23/2008 Added missing MULHWU instruction 4/23/2008 AND instruction decoding incorrectly. Fixed 4/23/2008 AND, ANDC, ANDI and ANDIS implemented incorrectly. Fixed 4/24/2008. Added proper disassembly for ANDI, ANDIS, ORI, ORIS, XORI, XORIC. Fixed 4/24/2008. Added proper disassembly for AND, ANDC, NAND, NOR, OR, ORC, XOR. Fixed 4/24/2008.

Copyright (c) 2006 Cmpware, Inc. All Rights Reserved.

Author:

SAG

Nested Class Summary

static class	PowerPC64.CR0 This class defines the fields in the CR Special Register
static class	PowerPC64.MSR This class defines the fields in the MSR Special Register
static class	PowerPC64.SR This class defines the Special Register indicies
static class	PowerPC64.XER This class defines the fields in the XER Special Register

Nested classes/interfaces inherited from class com.cmpware.tools.AutoModel

AutoModel.Asm, AutoModel.Decode, AutoModel.Field, AutoModel.Format, AutoModel.Function, AutoModel.Instruction, AutoModel.Symbol

Field Summary

static java.lang.String	copyright Copyright string
static PowerPC_fp	fp The floating point instructions
long[]	r64 The 64-bit register file (shadow)
static java.lang.String[]	reg64Names The IDE names for the 32-bit half registers (two per register)
static AutoModel.Symbol[]	regSymbols The General Purpose Register Symbols
(package private) boolean	reserve The reserve semaphore for lwarx / stwcx / ldarx / stdcx.
(package private) long	reserveAddr The reserve address for ldarx / stdcx.
long[]	sr64 The 64-bit special register file (shadow)
static AutoModel.Symbol[]	sregSymbols The Special Purpose Register Symbols

Fields inherited from class com.cmpware.tools.AutoModel

currentInstr, currentInstrType, instructionSizes, TRACE

Fields inherited from class com.cmpware.cmp.Processor

bkpt, branchAddr, branchDelay, breakpointHit, currentBranchDelay, currentInstrCode, cycleCount, elf, idlePower, iMix, instructionSize, name, noop, opcodeName, pc, prevInstr, prevInstrCode, r, regName, runningPower, sr, sregName, stall, stallCount

Fields inherited from class com.cmpware.cmp.Memory

BIG, LITTLE

Constructor Summary

PowerPC64()
The constructor

Method Summary

static long	addCarry(long result, long src1, long src2) This method is used to determine if an add carry has occurred.
static long	addOverflow(long result, long src1, long src2) This method is used to determine if an add overflow has occurred.
int[]	getDebugRegisters() This method returns the current values of all of the general purpose registers as they are used by a debugger.
int	getRegister(int i) This method returns a general purpose register.
int[]	getRegisters() This method returns the current values of all of the general purpose registers.
static void	main(java.lang.String[] args) This enables the stand-alone assembler and disassembler.
void	postLoad() This method it used to perform any operations that may be required after a file is loaded.
void	preExecute(int instr) This method is called immediately before execution and may be overloaded to do things like save instruction field values to locals.
void	reset() This method resets the processor.
static long	rotateLeft(int a, int r) This method rotates the value a left by r.
static long	rotateLeft(long a, int r) This method rotates the value a left by r.
static long	sign(long a) This method returns the sign bit for 64 bit value.
static long	subCarry(long result, long src1, long src2) This method is used to determine if a sub carry has occurred.
static long	subOverflow(long result, long src1, long src2) This method is used to determine if a sub overflow has occurred.

Methods inherited from class com.cmpware.tools.AutoModel

advancePC, asmFixup, bitsToBytes, concatenate, dasm, dasm, dasmFixup, decode, defineInstructions, execute, fetch, fetch, getInstruction, getInstructions, getInstructionSizes, getPC, getSymbolNames, instrMatch, main, postExecute, setPC, signExtend, signExtend, toHexString, toHexString, toLong, traceOff, traceOn

Methods inherited from class com.cmpware.cmp.Processor

branch, branchNoDelay, defineBranchDelay, defineBreakpoint, defineInstructionSize, defineName, defineNoop, defineOpcodeNames, defineRegisterNames, defineRegisters, defineSpecialRegisterNames, defineSpecialRegisters, get, getBreakpoint, getBreakpointData, getBreakpoints, getCycleCount, getElf, getIdlePower, getInstructionMix, getInstructionSize, getName, getNoop, getOpcodeNames, getRegisterNames, getRunningPower, getSpecialRegister, getSpecialRegisterNames, getSpecialRegisters, getStallCount, isBreakpoint, isStalled, load, removeAllBreakpoints, removeBreakpoint, setBreakpoint, setIdlePower, setRegister, setRunningPower, setSpecialRegister, stall, statistics, step

Methods inherited from class com.cmpware.cmp.Memory

addInput, addMemory, addOutput, getEndian, getLocalMemory, getMemoryManager, getMmio, getProfile, hexDump, hexDump, isReadable, isValidAddress, isWriteable, profile, read, read, read16, read24, read32, read64, removeMemory, resize, setEndian, size, toInt, toShort, write, write, write16, write24, write32, write64

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

copyright

public static final java.lang.String copyright

Copyright string

See Also:

Constant Field Values

fp

public static final PowerPC_fp fp

The floating point instructions

r64

public long[] r64

The 64-bit register file (shadow)

sr64

public long[] sr64

The 64-bit special register file (shadow)

reserve

boolean reserve

The reserve semaphore for lwarx / stwcx / ldarx / stdcx. instructions

reserveAddr

long reserveAddr

The reserve address for ldarx / stdcx.

reg64Names

public static final java.lang.String[] reg64Names

The IDE names for the 32-bit half registers (two per register)

regSymbols

public static final AutoModel.Symbol[] regSymbols

The General Purpose Register Symbols

sregSymbols

public static final AutoModel.Symbol[] sregSymbols

The Special Purpose Register Symbols

Constructor Detail

PowerPC64

public PowerPC64()

The constructor

Method Detail

main

public static void main(java.lang.String[] args)

This enables the stand-alone assembler and disassembler. The command line should be either: java PowerPC64 -dasm or java PowerPC64 -asm

Parameters:

args - the command line arguments.

reset

public void reset()

This method resets the processor. Note that super.reset() must me called to completely intialize the processor.

Overrides:

reset in class Processor

getRegisters

public int[] getRegisters()

Description copied from class: Processor

This method returns the current values of all of the general purpose registers.

Overrides:

getRegisters in class Processor

Returns:

This method returns the current values of all of the general purpose registers.

getRegister

public int getRegister(int i)

Description copied from class: Processor

This method returns a general purpose register.

Overrides:

getRegister in class Processor

Parameters:

i - The index of the general purpose register to get.

getDebugRegisters

public int[] getDebugRegisters()

Description copied from class: Processor

This method returns the current values of all of the general purpose registers as they are used by a debugger. This is added for machines that (for instance) have 128 bit registers of which the debugger only uses 32 bits. This is intended to be overloaded in these situations. This method is used exclusively by debuggers and debugger-like tools.

Overrides:

getDebugRegisters in class Processor

Returns:

This method returns the current values of all of the general purpose registers.

postLoad

public void postLoad()

Description copied from class: Processor

This method it used to perform any operations that may be required after a file is loaded. In the PowerPC, for instance, this method sets the TOC address for ELF files which use a TOC section. This method currently does nothing and must be overloaded by specific architectures.

Overrides:

postLoad in class Processor

preExecute

public void preExecute(int instr)

Description copied from class: AutoModel

This method is called immediately before execution and may be overloaded to do things like save instruction field values to locals.

Overrides:

preExecute in class AutoModel

Parameters:

instr - The instruction

addOverflow

public static long addOverflow(long result,
                               long src1,
                               long src2)

This method is used to determine if an add overflow has occurred. This is done by checking the signs of the source operands against the sign of the result.

Parameters:

result - The result of the operation on the src1 and src2 parameters.

src1 - The first source operand.

src2 - The second source operand.

Returns:

This method returns a true if an overflow was detected and false otherwise.

subOverflow

public static long subOverflow(long result,
                               long src1,
                               long src2)

This method is used to determine if a sub overflow has occurred. This is done by checking the signs of the source operands against the sign of the result.

Parameters:

result - The result of the operation on the src1 and src2 parameters.

src1 - The first source operand.

src2 - The second source operand.

Returns:

This method returns a true if an overflow was detected and a false otherwise.

addCarry

public static long addCarry(long result,
                            long src1,
                            long src2)

This method is used to determine if an add carry has occurred. This is done by checking the signs of the source operands against the sign of the result.

Parameters:

result - The result of the operation on the src1 and src2 parameters.

src1 - The first source operand.

src2 - The second source operand.

Returns:

This method returns a one if a carry was detected and a zero otherwise.

subCarry

public static long subCarry(long result,
                            long src1,
                            long src2)

This method is used to determine if a sub carry has occurred. This is done by checking the signs of the source operands against the sign of the result.

Parameters:

result - The result of the operation on the src1 and src2 parameters.

src1 - The first source operand.

src2 - The second source operand.

Returns:

This method returns a one if a carry was detected and a zero otherwise.

sign

public static long sign(long a)

This method returns the sign bit for 64 bit value.

Parameters:

a - The 64 bit value.

Returns:

This method returns the sign bit for 64 bit value.

rotateLeft

public static long rotateLeft(int a,
                              int r)

This method rotates the value a left by r. The value r should be between 0 and 31. The result is undefined for other values of r.

Parameters:

a - the value to be rotated left.

r - The number of bits to left rotate.

Returns:

this method returns the value a left rotated by r bits.

rotateLeft

public static long rotateLeft(long a,
                              int r)

This method rotates the value a left by r. The value r should be between 0 and 63. The result is undefined for other values of r.

Parameters:

a - the value to be rotated left.

r - The number of bits to left rotate.

Returns:

this method returns the value a left rotated by r bits.