Add XOP disassembler support. Fixes PR13933.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191874 91177308-0d34-0410-b5e6-96231b3b80d8

3 files changed, 174 insertions, 39 deletions

diff --git a/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c b/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c
index c26cfb5..480a869 100644
--- a/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c
+++ b/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c
@@ -81,6 +81,15 @@ static int modRMRequired(OpcodeType type,
   case THREEBYTE_A7:
     decision = &THREEBYTEA7_SYM;
     break;
+  case XOP8_MAP:
+    decision = &XOP8_MAP_SYM;
+    break;
+  case XOP9_MAP:
+    decision = &XOP9_MAP_SYM;
+    break;
+  case XOPA_MAP:
+    decision = &XOPA_MAP_SYM;
+    break;
   }
 
   return decision->opcodeDecisions[insnContext].modRMDecisions[opcode].
@@ -122,6 +131,15 @@ static InstrUID decode(OpcodeType type,
   case THREEBYTE_A7:
     dec = &THREEBYTEA7_SYM.opcodeDecisions[insnContext].modRMDecisions[opcode];
     break;
+  case XOP8_MAP:
+    dec = &XOP8_MAP_SYM.opcodeDecisions[insnContext].modRMDecisions[opcode];
+    break;
+  case XOP9_MAP:
+    dec = &XOP9_MAP_SYM.opcodeDecisions[insnContext].modRMDecisions[opcode];
+    break;
+  case XOPA_MAP:
+    dec = &XOPA_MAP_SYM.opcodeDecisions[insnContext].modRMDecisions[opcode];
+    break;
   }
 
   switch (dec->modrm_type) {
@@ -428,7 +446,7 @@ static int readPrefixes(struct InternalInstruction* insn) {
       dbgprintf(insn, "Found prefix 0x%hhx", byte);
   }
 
-  insn->vexSize = 0;
+  insn->vexXopType = TYPE_NO_VEX_XOP;
 
   if (byte == 0xc4) {
     uint8_t byte1;
@@ -439,7 +457,7 @@ static int readPrefixes(struct InternalInstruction* insn) {
     }
 
     if (insn->mode == MODE_64BIT || (byte1 & 0xc0) == 0xc0) {
-      insn->vexSize = 3;
+      insn->vexXopType = TYPE_VEX_3B;
       insn->necessaryPrefixLocation = insn->readerCursor - 1;
     }
     else {
@@ -447,22 +465,22 @@ static int readPrefixes(struct InternalInstruction* insn) {
       insn->necessaryPrefixLocation = insn->readerCursor - 1;
     }
 
-    if (insn->vexSize == 3) {
-      insn->vexPrefix[0] = byte;
-      consumeByte(insn, &insn->vexPrefix[1]);
-      consumeByte(insn, &insn->vexPrefix[2]);
+    if (insn->vexXopType == TYPE_VEX_3B) {
+      insn->vexXopPrefix[0] = byte;
+      consumeByte(insn, &insn->vexXopPrefix[1]);
+      consumeByte(insn, &insn->vexXopPrefix[2]);
 
       /* We simulate the REX prefix for simplicity's sake */
 
       if (insn->mode == MODE_64BIT) {
         insn->rexPrefix = 0x40
-                        | (wFromVEX3of3(insn->vexPrefix[2]) << 3)
-                        | (rFromVEX2of3(insn->vexPrefix[1]) << 2)
-                        | (xFromVEX2of3(insn->vexPrefix[1]) << 1)
-                        | (bFromVEX2of3(insn->vexPrefix[1]) << 0);
+                        | (wFromVEX3of3(insn->vexXopPrefix[2]) << 3)
+                        | (rFromVEX2of3(insn->vexXopPrefix[1]) << 2)
+                        | (xFromVEX2of3(insn->vexXopPrefix[1]) << 1)
+                        | (bFromVEX2of3(insn->vexXopPrefix[1]) << 0);
       }
 
-      switch (ppFromVEX3of3(insn->vexPrefix[2]))
+      switch (ppFromVEX3of3(insn->vexXopPrefix[2]))
       {
       default:
         break;
@@ -471,7 +489,9 @@ static int readPrefixes(struct InternalInstruction* insn) {
         break;
       }
 
-      dbgprintf(insn, "Found VEX prefix 0x%hhx 0x%hhx 0x%hhx", insn->vexPrefix[0], insn->vexPrefix[1], insn->vexPrefix[2]);
+      dbgprintf(insn, "Found VEX prefix 0x%hhx 0x%hhx 0x%hhx",
+                insn->vexXopPrefix[0], insn->vexXopPrefix[1],
+                insn->vexXopPrefix[2]);
     }
   }
   else if (byte == 0xc5) {
@@ -483,22 +503,66 @@ static int readPrefixes(struct InternalInstruction* insn) {
     }
 
     if (insn->mode == MODE_64BIT || (byte1 & 0xc0) == 0xc0) {
-      insn->vexSize = 2;
+      insn->vexXopType = TYPE_VEX_2B;
+    }
+    else {
+      unconsumeByte(insn);
+    }
+
+    if (insn->vexXopType == TYPE_VEX_2B) {
+      insn->vexXopPrefix[0] = byte;
+      consumeByte(insn, &insn->vexXopPrefix[1]);
+
+      if (insn->mode == MODE_64BIT) {
+        insn->rexPrefix = 0x40
+                        | (rFromVEX2of2(insn->vexXopPrefix[1]) << 2);
+      }
+
+      switch (ppFromVEX2of2(insn->vexXopPrefix[1]))
+      {
+      default:
+        break;
+      case VEX_PREFIX_66:
+        hasOpSize = TRUE;
+        break;
+      }
+
+      dbgprintf(insn, "Found VEX prefix 0x%hhx 0x%hhx", insn->vexXopPrefix[0], insn->vexXopPrefix[1]);
+    }
+  }
+  else if (byte == 0x8f) {
+    uint8_t byte1;
+
+    if (lookAtByte(insn, &byte1)) {
+      dbgprintf(insn, "Couldn't read second byte of XOP");
+      return -1;
+    }
+
+    if ((byte1 & 0x38) != 0x0) { // 0 in these 3 bits is a POP instruction.
+      insn->vexXopType = TYPE_XOP;
+      insn->necessaryPrefixLocation = insn->readerCursor - 1;
     }
     else {
       unconsumeByte(insn);
+      insn->necessaryPrefixLocation = insn->readerCursor - 1;
     }
 
-    if (insn->vexSize == 2) {
-      insn->vexPrefix[0] = byte;
-      consumeByte(insn, &insn->vexPrefix[1]);
+    if (insn->vexXopType == TYPE_XOP) {
+      insn->vexXopPrefix[0] = byte;
+      consumeByte(insn, &insn->vexXopPrefix[1]);
+      consumeByte(insn, &insn->vexXopPrefix[2]);
+
+      /* We simulate the REX prefix for simplicity's sake */
 
       if (insn->mode == MODE_64BIT) {
         insn->rexPrefix = 0x40
-                        | (rFromVEX2of2(insn->vexPrefix[1]) << 2);
+                        | (wFromXOP3of3(insn->vexXopPrefix[2]) << 3)
+                        | (rFromXOP2of3(insn->vexXopPrefix[1]) << 2)
+                        | (xFromXOP2of3(insn->vexXopPrefix[1]) << 1)
+                        | (bFromXOP2of3(insn->vexXopPrefix[1]) << 0);
       }
 
-      switch (ppFromVEX2of2(insn->vexPrefix[1]))
+      switch (ppFromXOP3of3(insn->vexXopPrefix[2]))
       {
       default:
         break;
@@ -507,7 +571,9 @@ static int readPrefixes(struct InternalInstruction* insn) {
         break;
       }
 
-      dbgprintf(insn, "Found VEX prefix 0x%hhx 0x%hhx", insn->vexPrefix[0], insn->vexPrefix[1]);
+      dbgprintf(insn, "Found XOP prefix 0x%hhx 0x%hhx 0x%hhx",
+                insn->vexXopPrefix[0], insn->vexXopPrefix[1],
+                insn->vexXopPrefix[2]);
     }
   }
   else {
@@ -582,12 +648,13 @@ static int readOpcode(struct InternalInstruction* insn) {
 
   insn->opcodeType = ONEBYTE;
 
-  if (insn->vexSize == 3)
+  if (insn->vexXopType == TYPE_VEX_3B)
   {
-    switch (mmmmmFromVEX2of3(insn->vexPrefix[1]))
+    switch (mmmmmFromVEX2of3(insn->vexXopPrefix[1]))
     {
     default:
-      dbgprintf(insn, "Unhandled m-mmmm field for instruction (0x%hhx)", mmmmmFromVEX2of3(insn->vexPrefix[1]));
+      dbgprintf(insn, "Unhandled m-mmmm field for instruction (0x%hhx)",
+                mmmmmFromVEX2of3(insn->vexXopPrefix[1]));
       return -1;
     case VEX_LOB_0F:
       insn->opcodeType = TWOBYTE;
@@ -600,11 +667,30 @@ static int readOpcode(struct InternalInstruction* insn) {
       return consumeByte(insn, &insn->opcode);
     }
   }
-  else if (insn->vexSize == 2)
+  else if (insn->vexXopType == TYPE_VEX_2B)
   {
     insn->opcodeType = TWOBYTE;
     return consumeByte(insn, &insn->opcode);
   }
+  else if (insn->vexXopType == TYPE_XOP)
+  {
+    switch (mmmmmFromXOP2of3(insn->vexXopPrefix[1]))
+    {
+    default:
+      dbgprintf(insn, "Unhandled m-mmmm field for instruction (0x%hhx)",
+                mmmmmFromVEX2of3(insn->vexXopPrefix[1]));
+      return -1;
+    case XOP_MAP_SELECT_8:
+      insn->opcodeType = XOP8_MAP;
+      return consumeByte(insn, &insn->opcode);
+    case XOP_MAP_SELECT_9:
+      insn->opcodeType = XOP9_MAP;
+      return consumeByte(insn, &insn->opcode);
+    case XOP_MAP_SELECT_A:
+      insn->opcodeType = XOPA_MAP;
+      return consumeByte(insn, &insn->opcode);
+    }
+  }
 
   if (consumeByte(insn, &current))
     return -1;
@@ -752,11 +838,27 @@ static int getID(struct InternalInstruction* insn, const void *miiArg) {
   if (insn->mode == MODE_64BIT)
     attrMask |= ATTR_64BIT;
 
-  if (insn->vexSize) {
+  if (insn->vexXopType != TYPE_NO_VEX_XOP) {
     attrMask |= ATTR_VEX;
 
-    if (insn->vexSize == 3) {
-      switch (ppFromVEX3of3(insn->vexPrefix[2])) {
+    if (insn->vexXopType == TYPE_VEX_3B) {
+      switch (ppFromVEX3of3(insn->vexXopPrefix[2])) {
+      case VEX_PREFIX_66:
+        attrMask |= ATTR_OPSIZE;
+        break;
+      case VEX_PREFIX_F3:
+        attrMask |= ATTR_XS;
+        break;
+      case VEX_PREFIX_F2:
+        attrMask |= ATTR_XD;
+        break;
+      }
+
+      if (lFromVEX3of3(insn->vexXopPrefix[2]))
+        attrMask |= ATTR_VEXL;
+    }
+    else if (insn->vexXopType == TYPE_VEX_2B) {
+      switch (ppFromVEX2of2(insn->vexXopPrefix[1])) {
       case VEX_PREFIX_66:
         attrMask |= ATTR_OPSIZE;
         break;
@@ -768,11 +870,11 @@ static int getID(struct InternalInstruction* insn, const void *miiArg) {
         break;
       }
 
-      if (lFromVEX3of3(insn->vexPrefix[2]))
+      if (lFromVEX2of2(insn->vexXopPrefix[1]))
         attrMask |= ATTR_VEXL;
     }
-    else if (insn->vexSize == 2) {
-      switch (ppFromVEX2of2(insn->vexPrefix[1])) {
+    else if (insn->vexXopType == TYPE_XOP) {
+      switch (ppFromXOP3of3(insn->vexXopPrefix[2])) {
       case VEX_PREFIX_66:
         attrMask |= ATTR_OPSIZE;
         break;
@@ -784,7 +886,7 @@ static int getID(struct InternalInstruction* insn, const void *miiArg) {
         break;
       }
 
-      if (lFromVEX2of2(insn->vexPrefix[1]))
+      if (lFromXOP3of3(insn->vexXopPrefix[2]))
         attrMask |= ATTR_VEXL;
     }
     else {
@@ -1450,10 +1552,12 @@ static int readImmediate(struct InternalInstruction* insn, uint8_t size) {
 static int readVVVV(struct InternalInstruction* insn) {
   dbgprintf(insn, "readVVVV()");
 
-  if (insn->vexSize == 3)
-    insn->vvvv = vvvvFromVEX3of3(insn->vexPrefix[2]);
-  else if (insn->vexSize == 2)
-    insn->vvvv = vvvvFromVEX2of2(insn->vexPrefix[1]);
+  if (insn->vexXopType == TYPE_VEX_3B)
+    insn->vvvv = vvvvFromVEX3of3(insn->vexXopPrefix[2]);
+  else if (insn->vexXopType == TYPE_VEX_2B)
+    insn->vvvv = vvvvFromVEX2of2(insn->vexXopPrefix[1]);
+  else if (insn->vexXopType == TYPE_XOP)
+    insn->vvvv = vvvvFromXOP3of3(insn->vexXopPrefix[2]);
   else
     return -1;
 
diff --git a/lib/Target/X86/Disassembler/X86DisassemblerDecoder.h b/lib/Target/X86/Disassembler/X86DisassemblerDecoder.h
index 4fd8fb7..bd9a651 100644
--- a/lib/Target/X86/Disassembler/X86DisassemblerDecoder.h
+++ b/lib/Target/X86/Disassembler/X86DisassemblerDecoder.h
@@ -59,6 +59,15 @@ extern "C" {
 #define lFromVEX2of2(vex)       (((vex) & 0x4) >> 2)
 #define ppFromVEX2of2(vex)      ((vex) & 0x3)
 
+#define rFromXOP2of3(xop)       (((~(xop)) & 0x80) >> 7)
+#define xFromXOP2of3(xop)       (((~(xop)) & 0x40) >> 6)
+#define bFromXOP2of3(xop)       (((~(xop)) & 0x20) >> 5)
+#define mmmmmFromXOP2of3(xop)   ((xop) & 0x1f)
+#define wFromXOP3of3(xop)       (((xop) & 0x80) >> 7)
+#define vvvvFromXOP3of3(vex)    (((~(vex)) & 0x78) >> 3)
+#define lFromXOP3of3(xop)       (((xop) & 0x4) >> 2)
+#define ppFromXOP3of3(xop)      ((xop) & 0x3)
+
 /*
  * These enums represent Intel registers for use by the decoder.
  */
@@ -444,9 +453,15 @@ typedef enum {
 typedef enum {
   VEX_LOB_0F = 0x1,
   VEX_LOB_0F38 = 0x2,
-  VEX_LOB_0F3A = 0x3
+  VEX_LOB_0F3A = 0x3,
 } VEXLeadingOpcodeByte;
 
+typedef enum {
+  XOP_MAP_SELECT_8 = 0x8,
+  XOP_MAP_SELECT_9 = 0x9,
+  XOP_MAP_SELECT_A = 0xA
+} XOPMapSelect;
+
 /*
  * VEXPrefixCode - Possible values for the VEX.pp field
  */
@@ -458,6 +473,13 @@ typedef enum {
   VEX_PREFIX_F2 = 0x3
 } VEXPrefixCode;
 
+typedef enum {
+  TYPE_NO_VEX_XOP = 0x0,
+  TYPE_VEX_2B = 0x1,
+  TYPE_VEX_3B = 0x2,
+  TYPE_XOP = 0x3
+} VEXXOPType;
+
 typedef uint8_t BOOL;
 
 /*
@@ -514,10 +536,10 @@ struct InternalInstruction {
   uint8_t prefixPresent[0x100];
   /* contains the location (for use with the reader) of the prefix byte */
   uint64_t prefixLocations[0x100];
-  /* The value of the VEX prefix, if present */
-  uint8_t vexPrefix[3];
+  /* The value of the VEX/XOP prefix, if present */
+  uint8_t vexXopPrefix[3];
   /* The length of the VEX prefix (0 if not present) */
-  uint8_t vexSize;
+  VEXXOPType vexXopType;
   /* The value of the REX prefix, if present */
   uint8_t rexPrefix;
   /* The location where a mandatory prefix would have to be (i.e., right before
diff --git a/lib/Target/X86/Disassembler/X86DisassemblerDecoderCommon.h b/lib/Target/X86/Disassembler/X86DisassemblerDecoderCommon.h
index b2f053b..86a90ee 100644
--- a/lib/Target/X86/Disassembler/X86DisassemblerDecoderCommon.h
+++ b/lib/Target/X86/Disassembler/X86DisassemblerDecoderCommon.h
@@ -32,6 +32,9 @@
 #define THREEBYTE3A_SYM   x86DisassemblerThreeByte3AOpcodes
 #define THREEBYTEA6_SYM   x86DisassemblerThreeByteA6Opcodes
 #define THREEBYTEA7_SYM   x86DisassemblerThreeByteA7Opcodes
+#define XOP8_MAP_SYM      x86DisassemblerXOP8Opcodes
+#define XOP9_MAP_SYM      x86DisassemblerXOP9Opcodes
+#define XOPA_MAP_SYM      x86DisassemblerXOPAOpcodes
 
 #define INSTRUCTIONS_STR  "x86DisassemblerInstrSpecifiers"
 #define CONTEXTS_STR      "x86DisassemblerContexts"
@@ -41,6 +44,9 @@
 #define THREEBYTE3A_STR   "x86DisassemblerThreeByte3AOpcodes"
 #define THREEBYTEA6_STR   "x86DisassemblerThreeByteA6Opcodes"
 #define THREEBYTEA7_STR   "x86DisassemblerThreeByteA7Opcodes"
+#define XOP8_MAP_STR      "x86DisassemblerXOP8Opcodes"
+#define XOP9_MAP_STR      "x86DisassemblerXOP9Opcodes"
+#define XOPA_MAP_STR      "x86DisassemblerXOPAOpcodes"
 
 /*
  * Attributes of an instruction that must be known before the opcode can be
@@ -234,7 +240,10 @@ typedef enum {
   THREEBYTE_38  = 2,
   THREEBYTE_3A  = 3,
   THREEBYTE_A6  = 4,
-  THREEBYTE_A7  = 5
+  THREEBYTE_A7  = 5,
+  XOP8_MAP      = 6,
+  XOP9_MAP      = 7,
+  XOPA_MAP      = 8
 } OpcodeType;
 
 /*