toir.c

The Valgrind distribution has multiple tools. The most popular is the memory checking tool (called M

      /* SIB, no displacement.  */
      case 0x04: {
         UChar sib    = getIByte(delta);
         UChar base_r = toUChar(sib & 7);
         if (base_r == R_EBP) return 6; else return 2;
      }
      /* SIB, with 8-bit displacement.  */
      case 0x0C: return 3;

      /* SIB, with 32-bit displacement.  */
      case 0x14: return 6;

      default:
         vpanic("lengthAMode");
         return 0; /*notreached*/
   }
}

/*------------------------------------------------------------*/
/*--- Disassembling common idioms                          ---*/
/*------------------------------------------------------------*/

/* Handle binary integer instructions of the form
      op E, G  meaning
      op reg-or-mem, reg
   Is passed the a ptr to the modRM byte, the actual operation, and the
   data size.  Returns the address advanced completely over this
   instruction.

   E(src) is reg-or-mem
   G(dst) is reg.

   If E is reg, -->    GET %G,  tmp
                       OP %E,   tmp
                       PUT tmp, %G
 
   If E is mem and OP is not reversible, 
                -->    (getAddr E) -> tmpa
                       LD (tmpa), tmpa
                       GET %G, tmp2
                       OP tmpa, tmp2
                       PUT tmp2, %G

   If E is mem and OP is reversible
                -->    (getAddr E) -> tmpa
                       LD (tmpa), tmpa
                       OP %G, tmpa
                       PUT tmpa, %G
*/
static
UInt dis_op2_E_G ( UChar       sorb,
                   Bool        addSubCarry,
                   IROp        op8, 
                   Bool        keep,
                   Int         size, 
                   Int         delta0,
                   HChar*      t_x86opc )
{
   HChar   dis_buf[50];
   Int     len;
   IRType  ty   = szToITy(size);
   IRTemp  dst1 = newTemp(ty);
   IRTemp  src  = newTemp(ty);
   IRTemp  dst0 = newTemp(ty);
   UChar   rm   = getUChar(delta0);
   IRTemp  addr = IRTemp_INVALID;

   /* addSubCarry == True indicates the intended operation is
      add-with-carry or subtract-with-borrow. */
   if (addSubCarry) {
      vassert(op8 == Iop_Add8 || op8 == Iop_Sub8);
      vassert(keep);
   }

   if (epartIsReg(rm)) {
      /* Specially handle XOR reg,reg, because that doesn't really
         depend on reg, and doing the obvious thing potentially
         generates a spurious value check failure due to the bogus
         dependency.  Ditto SBB reg,reg. */
      if ((op8 == Iop_Xor8 || (op8 == Iop_Sub8 && addSubCarry))
          && gregOfRM(rm) == eregOfRM(rm)) {
         putIReg(size, gregOfRM(rm), mkU(ty,0));
      }
      assign( dst0, getIReg(size,gregOfRM(rm)) );
      assign( src,  getIReg(size,eregOfRM(rm)) );

      if (addSubCarry && op8 == Iop_Add8) {
         helper_ADC( size, dst1, dst0, src );
         putIReg(size, gregOfRM(rm), mkexpr(dst1));
      } else
      if (addSubCarry && op8 == Iop_Sub8) {
         helper_SBB( size, dst1, dst0, src );
         putIReg(size, gregOfRM(rm), mkexpr(dst1));
      } else {
         assign( dst1, binop(mkSizedOp(ty,op8), mkexpr(dst0), mkexpr(src)) );
         if (isAddSub(op8))
            setFlags_DEP1_DEP2(op8, dst0, src, ty);
         else
            setFlags_DEP1(op8, dst1, ty);
         if (keep)
            putIReg(size, gregOfRM(rm), mkexpr(dst1));
      }

      DIP("%s%c %s,%s\n", t_x86opc, nameISize(size), 
                          nameIReg(size,eregOfRM(rm)),
                          nameIReg(size,gregOfRM(rm)));
      return 1+delta0;
   } else {
      /* E refers to memory */
      addr = disAMode ( &len, sorb, delta0, dis_buf);
      assign( dst0, getIReg(size,gregOfRM(rm)) );
      assign( src,  loadLE(szToITy(size), mkexpr(addr)) );

      if (addSubCarry && op8 == Iop_Add8) {
         helper_ADC( size, dst1, dst0, src );
         putIReg(size, gregOfRM(rm), mkexpr(dst1));
      } else
      if (addSubCarry && op8 == Iop_Sub8) {
         helper_SBB( size, dst1, dst0, src );
         putIReg(size, gregOfRM(rm), mkexpr(dst1));
      } else {
         assign( dst1, binop(mkSizedOp(ty,op8), mkexpr(dst0), mkexpr(src)) );
         if (isAddSub(op8))
            setFlags_DEP1_DEP2(op8, dst0, src, ty);
         else
            setFlags_DEP1(op8, dst1, ty);
         if (keep)
            putIReg(size, gregOfRM(rm), mkexpr(dst1));
      }

      DIP("%s%c %s,%s\n", t_x86opc, nameISize(size), 
                          dis_buf,nameIReg(size,gregOfRM(rm)));
      return len+delta0;
   }
}



/* Handle binary integer instructions of the form
      op G, E  meaning
      op reg, reg-or-mem
   Is passed the a ptr to the modRM byte, the actual operation, and the
   data size.  Returns the address advanced completely over this
   instruction.

   G(src) is reg.
   E(dst) is reg-or-mem

   If E is reg, -->    GET %E,  tmp
                       OP %G,   tmp
                       PUT tmp, %E
 
   If E is mem, -->    (getAddr E) -> tmpa
                       LD (tmpa), tmpv
                       OP %G, tmpv
                       ST tmpv, (tmpa)
*/
static
UInt dis_op2_G_E ( UChar       sorb,
                   Bool        addSubCarry,
                   IROp        op8, 
                   Bool        keep,
                   Int         size, 
                   Int         delta0,
                   HChar*      t_x86opc )
{
   HChar   dis_buf[50];
   Int     len;
   IRType  ty   = szToITy(size);
   IRTemp  dst1 = newTemp(ty);
   IRTemp  src  = newTemp(ty);
   IRTemp  dst0 = newTemp(ty);
   UChar   rm   = getIByte(delta0);
   IRTemp  addr = IRTemp_INVALID;

   /* addSubCarry == True indicates the intended operation is
      add-with-carry or subtract-with-borrow. */
   if (addSubCarry) {
      vassert(op8 == Iop_Add8 || op8 == Iop_Sub8);
      vassert(keep);
   }

   if (epartIsReg(rm)) {
      /* Specially handle XOR reg,reg, because that doesn't really
         depend on reg, and doing the obvious thing potentially
         generates a spurious value check failure due to the bogus
         dependency.  Ditto SBB reg,reg.*/
      if ((op8 == Iop_Xor8 || (op8 == Iop_Sub8 && addSubCarry))
          && gregOfRM(rm) == eregOfRM(rm)) {
         putIReg(size, eregOfRM(rm), mkU(ty,0));
      }
      assign(dst0, getIReg(size,eregOfRM(rm)));
      assign(src,  getIReg(size,gregOfRM(rm)));

      if (addSubCarry && op8 == Iop_Add8) {
         helper_ADC( size, dst1, dst0, src );
         putIReg(size, eregOfRM(rm), mkexpr(dst1));
      } else
      if (addSubCarry && op8 == Iop_Sub8) {
         helper_SBB( size, dst1, dst0, src );
         putIReg(size, eregOfRM(rm), mkexpr(dst1));
      } else {
         assign(dst1, binop(mkSizedOp(ty,op8), mkexpr(dst0), mkexpr(src)));
         if (isAddSub(op8))
            setFlags_DEP1_DEP2(op8, dst0, src, ty);
         else
            setFlags_DEP1(op8, dst1, ty);
         if (keep)
            putIReg(size, eregOfRM(rm), mkexpr(dst1));
      }

      DIP("%s%c %s,%s\n", t_x86opc, nameISize(size), 
                          nameIReg(size,gregOfRM(rm)),
                          nameIReg(size,eregOfRM(rm)));
      return 1+delta0;
   }

   /* E refers to memory */    
   {
      addr = disAMode ( &len, sorb, delta0, dis_buf);
      assign(dst0, loadLE(ty,mkexpr(addr)));
      assign(src,  getIReg(size,gregOfRM(rm)));

      if (addSubCarry && op8 == Iop_Add8) {
         helper_ADC( size, dst1, dst0, src );
         storeLE(mkexpr(addr), mkexpr(dst1));
      } else
      if (addSubCarry && op8 == Iop_Sub8) {
         helper_SBB( size, dst1, dst0, src );
         storeLE(mkexpr(addr), mkexpr(dst1));
      } else {
         assign(dst1, binop(mkSizedOp(ty,op8), mkexpr(dst0), mkexpr(src)));
         if (isAddSub(op8))
            setFlags_DEP1_DEP2(op8, dst0, src, ty);
         else
            setFlags_DEP1(op8, dst1, ty);
         if (keep)
            storeLE(mkexpr(addr), mkexpr(dst1));
      }

      DIP("%s%c %s,%s\n", t_x86opc, nameISize(size), 
                          nameIReg(size,gregOfRM(rm)), dis_buf);
      return len+delta0;
   }
}


/* Handle move instructions of the form
      mov E, G  meaning
      mov reg-or-mem, reg
   Is passed the a ptr to the modRM byte, and the data size.  Returns
   the address advanced completely over this instruction.

   E(src) is reg-or-mem
   G(dst) is reg.

   If E is reg, -->    GET %E,  tmpv
                       PUT tmpv, %G
 
   If E is mem  -->    (getAddr E) -> tmpa
                       LD (tmpa), tmpb
                       PUT tmpb, %G
*/
static
UInt dis_mov_E_G ( UChar       sorb,
                   Int         size, 
                   Int         delta0 )
{
   Int len;
   UChar rm = getIByte(delta0);
   HChar dis_buf[50];

   if (epartIsReg(rm)) {
      putIReg(size, gregOfRM(rm), getIReg(size, eregOfRM(rm)));
      DIP("mov%c %s,%s\n", nameISize(size), 
                           nameIReg(size,eregOfRM(rm)),
                           nameIReg(size,gregOfRM(rm)));
      return 1+delta0;
   }

   /* E refers to memory */    
   {
      IRTemp addr = disAMode ( &len, sorb, delta0, dis_buf );
      putIReg(size, gregOfRM(rm), loadLE(szToITy(size), mkexpr(addr)));
      DIP("mov%c %s,%s\n", nameISize(size), 
                           dis_buf,nameIReg(size,gregOfRM(rm)));
      return delta0+len;
   }
}


/* Handle move instructions of the form
      mov G, E  meaning
      mov reg, reg-or-mem
   Is passed the a ptr to the modRM byte, and the data size.  Returns
   the address advanced completely over this instruction.

   G(src) is reg.
   E(dst) is reg-or-mem

   If E is reg, -->    GET %G,  tmp
                       PUT tmp, %E
 
   If E is mem, -->    (getAddr E) -> tmpa
                       GET %G, tmpv
                       ST tmpv, (tmpa) 
*/
static
UInt dis_mov_G_E ( UChar       sorb,
                   Int         size, 
                   Int         delta0 )
{
   Int len;
   UChar rm = getIByte(delta0);
   HChar dis_buf[50];

   if (epartIsReg(rm)) {
      putIReg(size, eregOfRM(rm), getIReg(size, gregOfRM(rm)));
      DIP("mov%c %s,%s\n", nameISize(size), 
                           nameIReg(size,gregOfRM(rm)),
                           nameIReg(size,eregOfRM(rm)));
      return 1+delta0;
   }

   /* E refers to memory */    
   {
      IRTemp addr = disAMode ( &len, sorb, delta0, dis_buf);
      storeLE( mkexpr(addr), getIReg(size, gregOfRM(rm)) );
      DIP("mov%c %s,%s\n", nameISize(size), 
                           nameIReg(size,gregOfRM(rm)), dis_buf);
      return len+delta0;
   }
}


/* op $immediate, AL/AX/EAX. */
static
UInt dis_op_imm_A ( Int    size,
                    Bool   carrying,
                    IROp   op8,
                    Bool   keep,
                    Int    delta,
                    HChar* t_x86opc )
{
   IRType ty   = szToITy(size);
   IRTemp dst0 = newTemp(ty);
   IRTemp src  = newTemp(ty);
   IRTemp dst1 = newTemp(ty);
   UInt lit    = getUDisp(size,delta);
   assign(dst0, getIReg(size,R_EAX));
   assign(src,  mkU(ty,lit));

   if (isAddSub(op8) && !carrying) {
      assign(dst1, binop(mkSizedOp(ty,op8), mkexpr(dst0), mkexpr(src)) );
      setFlags_DEP1_DEP2(op8, dst0, src, ty);
   } 
   else
   if (isLogic(op8)) {
      vassert(!carrying);
      assign(dst1, binop(mkSizedOp(ty,op8), mkexpr(dst0), mkexpr(src)) );
      setFlags_DEP1(op8, dst1, ty);
   } 
   else
   if (op8 == Iop_Add8 && carrying) {
      helper_ADC( size, dst1, dst0, src );
   }
   else
   if (op8 == Iop_Sub8 && carrying) {
      helper_SBB( size, dst1, dst0, src );
   }
   else
      vpanic("dis_op_imm_A(x86,guest)");

   if (keep)
      putIReg(size, R_EAX, mkexpr(dst1));

   DIP("%s%c $0x%x, %s\n", t_x86opc, nameISize(size), 
                           lit, nameIReg(size,R_EAX));
   return delta+size;
}


/* Sign- and Zero-extending moves. */
static
UInt dis_movx_E_G ( UChar      sorb,
                    Int delta, Int szs, Int szd, Bool sign_extend )
{
   UChar rm = getIByte(delta);
   if (epartIsReg(rm)) {
      putIReg(szd, gregOfRM(rm),
                   unop(mkWidenOp(szs,szd,sign_extend), 
                        getIReg(szs,eregOfRM(rm))));
      DIP("mov%c%c%c %s,%s\n", sign_extend ? 's' : 'z',
                               nameISize(szs), nameISize(szd),
                               nameIReg(szs,eregOfRM(rm)),
                               nameIReg(szd,gregOfRM(rm)));
      return 1+delta;
   }

   /* E refers to memory */    
   {
      Int    len;
      HChar  dis_buf[50];
      IRTemp addr = disAMode ( &len, sorb, delta, dis_buf );

      putIReg(szd, gregOfRM(rm),
                   unop(m