Source code for envi.archs.arm.thumb

import envi.bits as e_bits
from envi.bits import binary
import envi.bintree as e_btree

import envi.archs.arm.disasm as arm_dis
import envi.archs.arm.armdisasm as arm_armdis
import envi.archs.arm.regs as arm_reg

thumb_32 = [
        binary('11101'),
        binary('11110'),
        binary('11111'),
]

O_REG = 0
O_IMM = 1

def shmaskval(value, shval, mask):  #FIXME: unnecessary to make this another fn call.  will be called a bajillion times.
[docs] return (value >> shval) & mask class simpleops:
[docs] def __init__(self, *operdef): self.operdef = operdef def __call__(self, va, value): ret = [] for otype, shval, mask in self.operdef: oval = shmaskval(value, shval, mask) ret.append( (value >> shval) ) imm5_rm_rd = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7), (O_IMM, 6, 0x1f))
rm_rn_rd = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7), (O_REG, 6, 0x7)) imm3_rn_rd = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7), (O_IMM, 6, 0x7)) imm8_rd = simpleops((O_REG, 8, 0x7), (O_IMM, 0, 0xff)) rm_rd = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7)) rn_rdm = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7)) rm_rdn = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7)) rm_rd_imm0 = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7), (O_IMM, 0, 0)) rm4_shift3 = simpleops((O_REG, 3, 0xf)) rm_rn_rt = simpleops((O_REG, 0, 0x7), (O_REG, 3, 0x7), (O_REG, 6, 0x7)) imm8 = simpleops((O_IMM, 8, 0xff)) imm11 = simpleops((O_IMM, 11, 0x7ff)) sh4_imm1 = simpleops((O_IMM, 3, 0x1)) def d1_rm4_rd3(va, value):
[docs] # 0 1 0 0 0 1 0 0 DN(1) Rm(4) Rdn(3) rdbit = shmaskval(value, 4, 0x8) rd = shmaskval(value, 0, 0x7) + rdbit rm = shmaskval(value, 3, 0xf) return ArmRegOper(rd),ArmRegOper(rn) def rm_rn_rt(va, value):
[docs] rt = shmask(value, 0, 0x7) # target rn = shmask(value, 3, 0x7) # base rm = shmask(value, 6, 0x7) # offset oper0 = arm_dis.ArmRegOper(rt) oper1 = arm_dis.ArmRegOffsetOper(rn, rm, va) return oper0,oper1 def imm5_rn_rt(va, value):
[docs] imm = shmask(value, 6, 0x1f) rn = shmask(value, 3, 0x7) rt = shmask(value, 0, 0x7) oper0 = arm_dis.ArmRegOper(rt) oper1 = arm_dis.ArmImmOffsetOper(rn, imm, va) return oper0,oper1 def rd_sp_imm8(va, value):
[docs] rd = shmask(value, 8, 0x7) imm = shmask(value, 0, 0xff) oper0 = arm_dis.ArmRegOper(rd) # pre-compute PC relative addr oper1 = arm_dis.ArmImmOffsetOper(REG_SP, imm) return oper0,oper1 def rd_pc_imm8(va, value):
[docs] rd = shmask(value, 8, 0x7) imm = shmask(value, 0, 0xff) oper0 = arm_dis.ArmRegOper(rd) # pre-compute PC relative addr oper1 = arm_dis.ArmImmOper(va+imm) return oper0,oper1 def rt_pc_imm8(va, value):
[docs] rt = shmask(value, 8, 0x7) imm = shmask(value, 0, 0xff) oper0 = arm_dis.ArmRegOper(rt) oper1 = arm_dis.ArmImmOffsetOper() # FIXME offset from PC return oper0,oper1 def ldmia(va, value):
[docs] rd = shmask(value, 8, 0x7) reg_list = value & 0xff oper0 = arm_dis.ArmRegOper(rd) oper1 = arm_dis.ArmRegListOper(reg_list) flags = 1<<11 # W flag indicating that write back should occur (marked by "!") return oper0,oper1 def sp_sp_imm7(va, value):
[docs] imm = shmask(value, 0, 0x7f) o0 = arm_dis.ArmRegOper(arm_reg.REG_SP) o1 = arm_dis.ArmRegOper(arm_reg.REG_SP) o2 = arm_dis.ArmImmOper(imm*4) return o0,o1,o2 def rm_reglist(va, value):
[docs] rm = shmask(value, 8, 0x7) reglist = value & 0xff oper0 = arm_dis.ArmRegOper(rm) oper1 = arm_dis.ArmReglistOper(reglist) return oper0,oper1 # opinfo is: # ( <mnem>, <operdef>, <flags> ) # operdef is: # ( (otype, oshift, omask), ...) thumb_table = [
('00000', ('lsl', imm5_rm_rd, 0)), # LSL<c> <Rd>,<Rm>,#<imm5> ('00001', ('lsr', imm5_rm_rd, 0)), # LSR<c> <Rd>,<Rm>,#<imm> ('00010', ('asr', imm5_rm_rd, 0)), # ASR<c> <Rd>,<Rm>,#<imm> ('0001100', ('add', rm_rn_rd, 0)), # ADD<c> <Rd>,<Rn>,<Rm> ('0001101', ('sub', rm_rn_rd, 0)), # SUB<c> <Rd>,<Rn>,<Rm> ('0001110', ('add', imm3_rn_rd, 0)), # ADD<c> <Rd>,<Rn>,#<imm3> ('0001111', ('sub', imm3_rn_rd, 0)), # SUB<c> <Rd>,<Rn>,#<imm3> ('00100', ('mov', imm8_rd, 0)), # MOV<c> <Rd>,#<imm8> ('00101', ('cmp', imm8_rd, 0)), # CMP<c> <Rn>,#<imm8> ('00110', ('add', imm8_rd, 0)), # ADD<c> <Rdn>,#<imm8> ('00111', ('sub', imm8_rd, 0)), # SUB<c> <Rdn>,#<imm8> # Data processing instructions ('0100000000', ('and', rm_rdn, 0)), # AND<c> <Rdn>,<Rm> ('0100000001', ('eor', rm_rdn, 0)), # EOR<c> <Rdn>,<Rm> ('0100000010', ('lsl', rm_rdn, 0)), # LSL<c> <Rdn>,<Rm> ('0100000011', ('lsr', rm_rdn, 0)), # LSR<c> <Rdn>,<Rm> ('0100000100', ('asr', rm_rdn, 0)), # ASR<c> <Rdn>,<Rm> ('0100000101', ('adc', rm_rdn, 0)), # ADC<c> <Rdn>,<Rm> ('0100000110', ('sbc', rm_rdn, 0)), # SBC<c> <Rdn>,<Rm> ('0100000111', ('ror', rm_rdn, 0)), # ROR<c> <Rdn>,<Rm> ('0100001000', ('tst', rm_rd, 0)), # TST<c> <Rn>,<Rm> ('0100001001', ('rsb', rm_rd_imm0, 0)), # RSB<c> <Rd>,<Rn>,#0 ('0100001010', ('cmp', rm_rd, 0)), # CMP<c> <Rn>,<Rm> ('0100001011', ('cmn', rm_rd, 0)), # CMN<c> <Rn>,<Rm> ('0100001100', ('orr', rm_rdn, 0)), # ORR<c> <Rdn>,<Rm> ('0100001101', ('mul', rn_rdm, 0)), # MUL<c> <Rdm>,<Rn>,<Rdm> ('0100001110', ('bic', rm_rdn, 0)), # BIC<c> <Rdn>,<Rm> ('0100001111', ('mvn', rm_rd, 0)), # MVN<c> <Rd>,<Rm> # Special data instructions and branch and exchange ('0100010000', ('add', d1_rm4_rd3, 0)), # ADD<c> <Rdn>,<Rm> ('0100010001', ('add', d1_rm4_rd3, 0)), # ADD<c> <Rdn>,<Rm> ('010001001', ('add', d1_rm4_rd3, 0)), # ADD<c> <Rdn>,<Rm> ('0100010101', ('cmp', d1_rm4_rd3, 0)), # CMP<c> <Rn>,<Rm> ('010001011', ('cmp', d1_rm4_rd3, 0)), # CMP<c> <Rn>,<Rm> ('0100011000', ('mov', d1_rm4_rd3, 0)), # MOV<c> <Rd>,<Rm> ('0100011001', ('mov', d1_rm4_rd3, 0)), # MOV<c> <Rd>,<Rm> ('0100011010', ('mov', d1_rm4_rd3, 0)), # MOV<c> <Rd>,<Rm> ('010001110', ('bx', rm4_shift3, 0)), # BX<c> <Rm> ('010001111', ('blx', rm4_shift3, 0)), # BLX<c> <Rm> # Load from Literal Pool ('01001', ('ldr', rt_pc_imm8, 0)), # LDR<c> <Rt>,<label> # Load/Stor single data item ('0101000', ('str', rm_rn_rt, 0)), # STR<c> <Rt>,[<Rn>,<Rm>] ('0101001', ('strh', rm_rn_rt, 0)), # STRH<c> <Rt>,[<Rn>,<Rm>] ('0101010', ('strb', rm_rn_rt, 0)), # STRB<c> <Rt>,[<Rn>,<Rm>] ('0101011', ('ldrsb', rm_rn_rt, 0)), # LDRSB<c> <Rt>,[<Rn>,<Rm>] ('0101100', ('ldr', rm_rn_rt, 0)), # LDR<c> <Rt>,[<Rn>,<Rm>] ('0101101', ('ldrh', rm_rn_rt, 0)), # LDRH<c> <Rt>,[<Rn>,<Rm>] ('0101110', ('ldrb', rm_rn_rt, 0)), # LDRB<c> <Rt>,[<Rn>,<Rm>] ('0101111', ('ldrsh', rm_rn_rt, 0)), # LDRSH<c> <Rt>,[<Rn>,<Rm>] ('01100', ('str', imm5_rn_rt, 0)), # STR<c> <Rt>, [<Rn>{,#<imm5>}] ('01101', ('ldr', imm5_rn_rt, 0)), # LDR<c> <Rt>, [<Rn>{,#<imm5>}] ('01110', ('strb', imm5_rn_rt, 0)), # STRB<c> <Rt>,[<Rn>,#<imm5>] ('01111', ('ldrb', imm5_rn_rt, 0)), # LDRB<c> <Rt>,[<Rn>{,#<imm5>}] ('10000', ('strh', imm5_rn_rt, 0)), # STRH<c> <Rt>,[<Rn>{,#<imm>}] ('10001', ('ldrh', imm5_rn_rt, 0)), # LDRH<c> <Rt>,[<Rn>{,#<imm>}] ('10010', ('str', imm5_rn_rt, 0)), # STR<c> <Rt>, [<Rn>{,#<imm>}] ('10011', ('ldr', imm5_rn_rt, 0)), # LDR<c> <Rt>, [<Rn>{,#<imm>}] # Generate PC relative address ('10100', ('add', rd_pc_imm8, 0)), # ADD<c> <Rd>,<label> # Generate SP relative address ('10101', ('add', rd_sp_imm8, 0)), # ADD<c> <Rd>,SP,#<imm> # Miscellaneous instructions ('10110110010', ('setend', sh4_imm1, 0)), # SETEND <endian_specifier> ('10110110011', ('cps', simpleops(),0)), # CPS<effect> <iflags> FIXME ('1011101000', ('rev', rn_rdm, 0)), # REV Rd, Rn ('1011101001', ('rev16', rn_rdm, 0)), # REV16 Rd, Rn ('1011101011', ('revsh', rn_rdm, 0)), # REVSH Rd, Rn ('101100000', ('add', sp_sp_imm7, 0)), # ADD<c> SP,SP,#<imm> ('101100001', ('sub', sp_sp_imm7, 0)), # SUB<c> SP,SP,#<imm> ('10111110', ('bkpt', imm8, 0)), # BKPT <blahblah> # Load / Store Multiple ('11000', ('stmia', rm_reglist, 0x800)), # LDMIA Rd!, reg_list ('11001', ('ldmia', rm_reglist, 0x800)), # STMIA Rd!, reg_list # Conditional Branches ('11010000', ('b', imm8, 0)), ('11010001', ('bn', imm8, 0)), ('11010010', ('bz', imm8, 0)), ('11010011', ('bnz', imm8, 0)), ('11010100', ('bc', imm8, 0)), ('11010101', ('bnc', imm8, 0)), ('11010100', ('bzc', imm8, 0)), ('11010111', ('bnzc', imm8, 0)), ('11011000', ('bv', imm8, 0)), ('11011001', ('bnv', imm8, 0)), ('11011010', ('bzv', imm8, 0)), ('11011011', ('bnzv', imm8, 0)), ('11011100', ('bcv', imm8, 0)), ('11011101', ('bncv', imm8, 0)), ('11011110', ('bzcv', imm8, 0)), ('11011111', ('bnzcv', imm8, 0)), # Software Interrupt ('11011111', ('swi', imm8, 0)), # SWI <blahblah> ('11100', ('b', imm11, 0)), # B <addr11> ('11101', ('blx', imm11, 0)), # BLX suffix <addr11> -- SEE p542 of 14218.pdf manual for how this if gonna fuck with emulation. ('11110', ('bl', imm11, 0)), # BL/BLX prefix <addr11> -- SEE p542 of 14218.pdf manual for how this if gonna fuck with emulation. ('11111', ('blx', imm11, 0)), # BL suffix <addr11> -- SEE p542 of 14218.pdf manual for how this if gonna fuck with emulation. ] ttree = e_btree.BinaryTree() for binstr, opinfo in thumb_table: ttree.addBinstr(binstr, opinfo) thumb32mask = binary('11111') thumb32min = binary('11100') def is_thumb32(val):
[docs] ''' Take a 16 bit integer (opcode) value and determine if it is really the first 16 bits of a 32 bit instruction. ''' bval = val >> 11 return (bval & thumb32mask) > thumb32min class ThumbOpcode(arm_armdis.ArmOpcode):
[docs] pass #class ArmThumbDisasm(arm_armdis.ArmDisasmChild): #def disasm(self, bytes, offset, va, trackMode=True): #val = struct.unpack("<L", bytes[offset:offset+2]) #mnem, opermkr, flags = ttree.getInstr(val) #olist = opermkr(va, val) #op = ThumbOpcode(va, opcode, mnem, 0xe, 2, olist, flags) #return op #raise Exception("ummm. you could try Implementing disasm first... duh.")