x86emul: improve LOCK handling

Certain opcodes would so far not have got #UD when a LOCK prefix was
present. Adjust this by
- moving the too early generic check into destination operand decoding,
  where DstNone and DstReg already have respective handling
- switching source and destination of TEST r,r/m, for it to be taken
  care of by aforementioned generic checks
- explicitly dealing with all forms of CMP, SHLD, SHRD, as well as
  TEST $imm,r/m

To make the handling of opcodes F6 and F7 more obvious, reduce the
amount of state set in the table, and adjust the respective switch()
statement accordingly.

Also eliminate the latent bug of the check in DstNone handling not
considering the opcode extension set.

Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
---
This will only apply cleanly on top of
https://lists.xenproject.org/archives/html/xen-devel/2016-08/msg01975.html.

--- a/xen/arch/x86/x86_emulate/x86_emulate.c
+++ b/xen/arch/x86/x86_emulate/x86_emulate.c
@@ -110,7 +110,7 @@ static uint8_t opcode_table[256] = {
     /* 0x80 - 0x87 */
     ByteOp|DstMem|SrcImm|ModRM, DstMem|SrcImm|ModRM,
     ByteOp|DstMem|SrcImm|ModRM, DstMem|SrcImmByte|ModRM,
-    ByteOp|DstMem|SrcReg|ModRM, DstMem|SrcReg|ModRM,
+    ByteOp|DstReg|SrcMem|ModRM, DstReg|SrcMem|ModRM,
     ByteOp|DstMem|SrcReg|ModRM, DstMem|SrcReg|ModRM,
     /* 0x88 - 0x8F */
     ByteOp|DstMem|SrcReg|ModRM|Mov, DstMem|SrcReg|ModRM|Mov,
@@ -169,8 +169,7 @@ static uint8_t opcode_table[256] = {
     DstEax|SrcImplicit, DstEax|SrcImplicit, ImplicitOps, ImplicitOps,
     /* 0xF0 - 0xF7 */
     0, ImplicitOps, 0, 0,
-    ImplicitOps, ImplicitOps,
-    ByteOp|DstMem|SrcNone|ModRM, DstMem|SrcNone|ModRM,
+    ImplicitOps, ImplicitOps, ByteOp|ModRM, ModRM,
     /* 0xF8 - 0xFF */
     ImplicitOps, ImplicitOps, ImplicitOps, ImplicitOps,
     ImplicitOps, ImplicitOps, ByteOp|DstMem|SrcNone|ModRM, DstMem|SrcNone|ModRM
@@ -1651,9 +1650,6 @@ x86_emulate(
         }
     }
 
-    /* Lock prefix is allowed only on RMW instructions. */
-    generate_exception_if((d & Mov) && lock_prefix, EXC_UD, -1);
-
     /* ModRM and SIB bytes. */
     if ( d & ModRM )
     {
@@ -1729,13 +1725,17 @@ x86_emulate(
                 switch ( modrm_reg & 7 )
                 {
                 case 0 ... 1: /* test */
-                    d = (d & ~SrcMask) | SrcImm;
+                    d |= DstMem | SrcImm;
+                    break;
+                case 2: /* not */
+                case 3: /* neg */
+                    d |= DstMem;
                     break;
                 case 4: /* mul */
                 case 5: /* imul */
                 case 6: /* div */
                 case 7: /* idiv */
-                    d = (d & (ByteOp | ModRM)) | DstEax | SrcMem;
+                    d |= DstEax | SrcMem;
                     break;
                 }
                 break;
@@ -1983,8 +1983,9 @@ x86_emulate(
          */
         generate_exception_if(
             lock_prefix &&
-            ((b < 0x20) || (b > 0x23)) && /* MOV CRn/DRn */
-            (b != 0xc7),                  /* CMPXCHG{8,16}B */
+            (ext != ext_0f ||
+             (((b < 0x20) || (b > 0x23)) && /* MOV CRn/DRn */
+              (b != 0xc7))),                /* CMPXCHG{8,16}B */
             EXC_UD, -1);
         dst.type = OP_NONE;
         break;
@@ -2062,6 +2063,8 @@ x86_emulate(
                 goto done;
             dst.orig_val = dst.val;
         }
+        else /* Lock prefix is allowed only on RMW instructions. */
+            generate_exception_if(lock_prefix, EXC_UD, -1);
         break;
     }
 
@@ -2111,6 +2114,7 @@ x86_emulate(
         break;
 
     case 0x38 ... 0x3d: cmp: /* cmp */
+        generate_exception_if(lock_prefix, EXC_UD, -1);
         emulate_2op_SrcV("cmp", src, dst, _regs.eflags);
         dst.type = OP_NONE;
         break;
@@ -3545,6 +3549,7 @@ x86_emulate(
             unsigned long u[2], v;
 
         case 0 ... 1: /* test */
+            generate_exception_if(lock_prefix, EXC_UD, -1);
             goto test;
         case 2: /* not */
             dst.val = ~dst.val;
@@ -4507,6 +4512,7 @@ x86_emulate(
     case 0xad: /* shrd %%cl,r,r/m */ {
         uint8_t shift, width = dst.bytes << 3;
 
+        generate_exception_if(lock_prefix, EXC_UD, -1);
         if ( b & 1 )
             shift = _regs.ecx;
         else