dasm_x86.lua

采用C语言写的Lua的解释器的代码！Lua不用介绍了吧

------------------------------------------------------------------------------
-- DynASM x86 module.
--
-- Copyright (C) 2005-2008 Mike Pall. All rights reserved.
-- See dynasm.lua for full copyright notice.
------------------------------------------------------------------------------

-- Module information:
local _info = {
  arch =	"x86",
  description =	"DynASM x86 (i386) module",
  version =	"1.1.4",
  vernum =	 10104,
  release =	"2008-01-29",
  author =	"Mike Pall",
  license =	"MIT",
}

-- Exported glue functions for the arch-specific module.
local _M = { _info = _info }

-- Cache library functions.
local type, tonumber, pairs, ipairs = type, tonumber, pairs, ipairs
local assert, unpack = assert, unpack
local _s = string
local sub, format, byte, char = _s.sub, _s.format, _s.byte, _s.char
local find, match, gmatch, gsub = _s.find, _s.match, _s.gmatch, _s.gsub
local concat, sort = table.concat, table.sort
local char, unpack = string.char, unpack

-- Inherited tables and callbacks.
local g_opt, g_arch
local wline, werror, wfatal, wwarn

-- Action name list.
-- CHECK: Keep this in sync with the C code!
local action_names = {
  -- int arg, 1 buffer pos:
  "DISP",  "IMM_S", "IMM_B", "IMM_W", "IMM_D",  "IMM_WB", "IMM_DB",
  -- action arg (1 byte), int arg, 1 buffer pos (num):
  "SPACE",
  -- ptrdiff_t arg, 1 buffer pos (address): !x64
  "SETLABEL", "REL_A",
  -- action arg (1 byte) or int arg, 2 buffer pos (link, offset):
  "REL_LG", "REL_PC",
  -- action arg (1 byte) or int arg, 1 buffer pos (link):
  "IMM_LG", "IMM_PC",
  -- action arg (1 byte) or int arg, 1 buffer pos (offset):
  "LABEL_LG", "LABEL_PC",
  -- action arg (1 byte), 1 buffer pos (offset):
  "ALIGN",
  -- action arg (1 byte), no buffer pos.
  "ESC",
  -- no action arg, no buffer pos.
  "MARK",
  -- action arg (1 byte), no buffer pos, terminal action:
  "SECTION",
  -- no args, no buffer pos, terminal action:
  "STOP"
}

-- Maximum number of section buffer positions for dasm_put().
-- CHECK: Keep this in sync with the C code!
local maxsecpos = 25 -- Keep this low, to avoid excessively long C lines.

-- Action name -> action number (dynamically generated below).
local map_action = {}
-- First action number. Everything below does not need to be escaped.
local actfirst = 256-#action_names

-- Action list buffer and string (only used to remove dupes).
local actlist = {}
local actstr = ""

-- Argument list for next dasm_put(). Start with offset 0 into action list.
local actargs = { 0 }

-- Current number of section buffer positions for dasm_put().
local secpos = 1

------------------------------------------------------------------------------

-- Compute action numbers for action names.
for n,name in ipairs(action_names) do
  local num = actfirst + n - 1
  map_action[name] = num
end

-- Dump action names and numbers.
local function dumpactions(out)
  out:write("DynASM encoding engine action codes:\n")
  for n,name in ipairs(action_names) do
    local num = map_action[name]
    out:write(format("  %-10s %02X  %d\n", name, num, num))
  end
  out:write("\n")
end

-- Write action list buffer as a huge static C array.
local function writeactions(out, name)
  local nn = #actlist
  local last = actlist[nn] or 255
  actlist[nn] = nil -- Remove last byte.
  if nn == 0 then nn = 1 end
  out:write("static const unsigned char ", name, "[", nn, "] = {\n")
  local s = "  "
  for n,b in ipairs(actlist) do
    s = s..b..","
    if #s >= 75 then
      assert(out:write(s, "\n"))
      s = "  "
    end
  end
  out:write(s, last, "\n};\n\n") -- Add last byte back.
end

------------------------------------------------------------------------------

-- Add byte to action list.
local function wputxb(n)
  assert(n >= 0 and n <= 255 and n % 1 == 0, "byte out of range")
  actlist[#actlist+1] = n
end

-- Add action to list with optional arg. Advance buffer pos, too.
local function waction(action, a, num)
  wputxb(assert(map_action[action], "bad action name `"..action.."'"))
  if a then actargs[#actargs+1] = a end
  if a or num then secpos = secpos + (num or 1) end
end

-- Add call to embedded DynASM C code.
local function wcall(func, args)
  wline(format("dasm_%s(Dst, %s);", func, concat(args, ", ")), true)
end

-- Delete duplicate action list chunks. A tad slow, but so what.
local function dedupechunk(offset)
  local al, as = actlist, actstr
  local chunk = char(unpack(al, offset+1, #al))
  local orig = find(as, chunk, 1, true)
  if orig then
    actargs[1] = orig-1 -- Replace with original offset.
    for i=offset+1,#al do al[i] = nil end -- Kill dupe.
  else
    actstr = as..chunk
  end
end

-- Flush action list (intervening C code or buffer pos overflow).
local function wflush(term)
  local offset = actargs[1]
  if #actlist == offset then return end -- Nothing to flush.
  if not term then waction("STOP") end -- Terminate action list.
  dedupechunk(offset)
  wcall("put", actargs) -- Add call to dasm_put().
  actargs = { #actlist } -- Actionlist offset is 1st arg to next dasm_put().
  secpos = 1 -- The actionlist offset occupies a buffer position, too.
end

-- Put escaped byte.
local function wputb(n)
  if n >= actfirst then waction("ESC") end -- Need to escape byte.
  wputxb(n)
end

------------------------------------------------------------------------------

-- Global label name -> global label number. With auto assignment on 1st use.
local next_global = 10
local map_global = setmetatable({}, { __index = function(t, name)
  if not match(name, "^[%a_][%w_]*$") then werror("bad global label") end
  local n = next_global
  if n > 246 then werror("too many global labels") end
  next_global = n + 1
  t[name] = n
  return n
end})

-- Dump global labels.
local function dumpglobals(out, lvl)
  local t = {}
  for name, n in pairs(map_global) do t[n] = name end
  out:write("Global labels:\n")
  for i=10,next_global-1 do
    out:write(format("  %s\n", t[i]))
  end
  out:write("\n")
end

-- Write global label enum.
local function writeglobals(out, prefix)
  local t = {}
  for name, n in pairs(map_global) do t[n] = name end
  out:write("enum {\n")
  for i=10,next_global-1 do
    out:write("  ", prefix, t[i], ",\n")
  end
  out:write("  ", prefix, "_MAX\n};\n")
end

------------------------------------------------------------------------------

-- Arch-specific maps.
local map_archdef = {}		-- Ext. register name -> int. name.
local map_reg_rev = {}		-- Int. register name -> ext. name.
local map_reg_num = {}		-- Int. register name -> register number.
local map_reg_opsize = {}	-- Int. register name -> operand size.
local map_reg_valid_base = {}	-- Int. register name -> valid base register?
local map_reg_valid_index = {}	-- Int. register name -> valid index register?
local reg_list = {}		-- Canonical list of int. register names.

local map_type = {}		-- Type name -> { ctype, reg }
local ctypenum = 0		-- Type number (for _PTx macros).

local addrsize = "d"		-- Size for address operands. !x64

-- Helper function to fill register maps.
local function mkrmap(sz, names)
  for n,name in ipairs(names) do
    local iname = format("@%s%x", sz, n-1)
    reg_list[#reg_list+1] = iname
    map_archdef[name] = iname
    map_reg_rev[iname] = name
    map_reg_num[iname] = n-1
    map_reg_opsize[iname] = sz
    if sz == addrsize then
      map_reg_valid_base[iname] = true
      map_reg_valid_index[iname] = true
    end
  end
  reg_list[#reg_list+1] = ""
end

-- Integer registers (dword, word and byte sized).
mkrmap("d", {"eax", "ecx", "edx", "ebx", "esp", "ebp", "esi", "edi"})
map_reg_valid_index[map_archdef.esp] = nil
mkrmap("w", {"ax", "cx", "dx", "bx", "sp", "bp", "si", "di"})
mkrmap("b", {"al", "cl", "dl", "bl", "ah", "ch", "dh", "bh"})

-- FP registers (internally tword sized, but use "f" as operand size).
mkrmap("f", {"st0", "st1", "st2", "st3", "st4", "st5", "st6", "st7"})

-- SSE registers (oword sized, but qword and dword accessible).
mkrmap("o", {"xmm0", "xmm1", "xmm2", "xmm3", "xmm4", "xmm5", "xmm6", "xmm7"})

-- Operand size prefixes to codes.
local map_opsize = {
  byte = "b", word = "w", dword = "d", qword = "q", oword = "o", tword = "t",
  aword = addrsize,
}

-- Operand size code to number.
local map_opsizenum = {
  b = 1, w = 2, d = 4, q = 8, o = 16, t = 10,
}

-- Operand size code to name.
local map_opsizename = {
  b = "byte", w = "word", d = "dword", q = "qword", o = "oword", t = "tword",
  f = "fpword",
}

-- Valid index register scale factors.
local map_xsc = {
  ["1"] = 0, ["2"] = 1, ["4"] = 2, ["8"] = 3,
}

-- Condition codes.
local map_cc = {
  o = 0, no = 1, b = 2, nb = 3, e = 4, ne = 5, be = 6, nbe = 7,
  s = 8, ns = 9, p = 10, np = 11, l = 12, nl = 13, le = 14, nle = 15,
  c = 2, nae = 2, nc = 3, ae = 3, z = 4, nz = 5, na = 6, a = 7,
  nge = 12, ge = 13, ng = 14, g = 15,
}


-- Reverse defines for registers.
function _M.revdef(s)
  return gsub(s, "@%w+", map_reg_rev)
end

-- Dump register names and numbers
local function dumpregs(out)
  out:write("Register names, sizes and internal numbers:\n")
  for _,reg in ipairs(reg_list) do
    if reg == "" then
      out:write("\n")
    else
      local name = map_reg_rev[reg]
      local num = map_reg_num[reg]
      local opsize = map_opsizename[map_reg_opsize[reg]]
      out:write(format("  %-5s %-8s %d\n", name, opsize, num))
    end
  end
end

------------------------------------------------------------------------------

-- Put action for label arg (IMM_LG, IMM_PC, REL_LG, REL_PC).
local function wputlabel(aprefix, imm, num)
  if type(imm) == "number" then
    waction(aprefix.."LG", nil, num);
    wputxb(imm)
  else
    waction(aprefix.."PC", imm, num)
  end
end

-- Put signed byte or arg.
local function wputsbarg(n)
  if type(n) == "number" then
    if n < -128 or n > 127 then
      werror("signed immediate byte out of range")
    end
    if n < 0 then n = n + 256 end
    wputb(n)
  else waction("IMM_S", n) end
end

-- Put unsigned byte or arg.
local function wputbarg(n)
  if type(n) == "number" then
    if n < 0 or n > 255 then
      werror("unsigned immediate byte out of range")
    end
    wputb(n)
  else waction("IMM_B", n) end
end

-- Put unsigned word or arg.
local function wputwarg(n)
  if type(n) == "number" then
    if n < 0 or n > 65535 then
      werror("unsigned immediate word out of range")
    end
    local r = n%256; n = (n-r)/256; wputb(r); wputb(n);
  else waction("IMM_W", n) end
end

-- Put signed or unsigned dword or arg.
local function wputdarg(n)
  local tn = type(n)
  if tn == "number" then
    if n < 0 then n = n + 4294967296 end
    local r = n%256; n = (n-r)/256; wputb(r);
    r = n%256; n = (n-r)/256; wputb(r);
    r = n%256; n = (n-r)/256; wputb(r); wputb(n);
  elseif tn == "table" then
    wputlabel("IMM_", n[1], 1)
  else
    waction("IMM_D", n)
  end
end

-- Put operand-size dependent number or arg (defaults to dword).
local function wputszarg(sz, n)
  if not sz or sz == "d" then wputdarg(n)
  elseif sz == "w" then wputwarg(n)
  elseif sz == "b" then wputbarg(n)
  elseif sz == "s" then wputsbarg(n)
  else werror("bad operand size") end
end

-- Put multi-byte opcode with operand-size dependent modifications.
local function wputop(sz, op)
  local r
  if sz == "w" then wputb(102) end
  if op >= 16777216 then r = op % 16777216 wputb((op-r) / 16777216) op = r end
  if op >= 65536 then r = op % 65536 wputb((op-r) / 65536) op = r end
  if op >= 256 then r = op % 256 wputb((op-r) / 256) op = r end
  if sz == "b" then op = op - 1 end
  wputb(op)
end

-- Put ModRM or SIB formatted byte.
local function wputmodrm(m, s, rm)
  assert(m < 4 and s < 8 and rm < 8, "bad modrm operands")
  wputb(64*m + 8*s + rm)
end

-- Put ModRM/SIB plus optional displacement.
local function wputmrmsib(t, s, imark)
  -- Register mode.
  if sub(t.mode, 1, 1) == "r" then
    wputmodrm(3, s, t.reg)
    return
  end

  local disp = t.disp
  local tdisp = type(disp)
  -- No base register?
  if not t.reg then
    if t.xreg then
      -- Indexed mode with index register only.
      wputmodrm(0, s, 4) -- [xreg*xsc+disp] -> (0, s, esp) (xsc, xreg, ebp)