MFC 261991:

Upgrade our copy of llvm/clang to 3.4 release.  This version supports
all of the features in the current working draft of the upcoming C++
standard, provisionally named C++1y.

The code generator's performance is greatly increased, and the loop
auto-vectorizer is now enabled at -Os and -O2 in addition to -O3.  The
PowerPC backend has made several major improvements to code generation
quality and compile time, and the X86, SPARC, ARM32, Aarch64 and SystemZ
backends have all seen major feature work.

Release notes for llvm and clang can be found here:
<http://llvm.org/releases/3.4/docs/ReleaseNotes.html>
<http://llvm.org/releases/3.4/tools/clang/docs/ReleaseNotes.html>

MFC 262121 (by emaste):

Update lldb for clang/llvm 3.4 import

This commit largely restores the lldb source to the upstream r196259
snapshot with the addition of threaded inferior support and a few bug
fixes.

Specific upstream lldb revisions restored include:
   SVN      git
  181387  779e6ac
  181703  7bef4e2
  182099  b31044e
  182650  f2dcf35
  182683  0d91b80
  183862  15c1774
  183929  99447a6
  184177  0b2934b
  184948  4dc3761
  184954  007e7bc
  186990  eebd175

Sponsored by:	DARPA, AFRL

MFC 262186 (by emaste):

Fix mismerge in r262121

A break statement was lost in the merge.  The error had no functional
impact, but restore it to reduce the diff against upstream.

MFC 262303:

Pull in r197521 from upstream clang trunk (by rdivacky):

  Use the integrated assembler by default on FreeBSD/ppc and ppc64.

Requested by:	jhibbits

MFC 262611:

Pull in r196874 from upstream llvm trunk:

  Fix a crash that occurs when PWD is invalid.

  MCJIT needs to be able to run in hostile environments, even when PWD
  is invalid. There's no need to crash MCJIT in this case.

  The obvious fix is to simply leave MCContext's CompilationDir empty
  when PWD can't be determined. This way, MCJIT clients,
  and other clients that link with LLVM don't need a valid working directory.

  If we do want to guarantee valid CompilationDir, that should be done
  only for clients of getCompilationDir(). This is as simple as checking
  for an empty string.

  The only current use of getCompilationDir is EmitGenDwarfInfo, which
  won't conceivably run with an invalid working dir. However, in the
  purely hypothetically and untestable case that this happens, the
  AT_comp_dir will be omitted from the compilation_unit DIE.

This should help fix assertions occurring with ports-mgmt/tinderbox,
when it is using jails, and sometimes invalidates clang's current
working directory.

Reported by:	decke

MFC 262809:

Pull in r203007 from upstream clang trunk:

  Don't produce an alias between destructors with different calling conventions.

  Fixes pr19007.

(Please note that is an LLVM PR identifier, not a FreeBSD one.)

This should fix Firefox and/or libxul crashes (due to problems with
regparm/stdcall calling conventions) on i386.

Reported by:	multiple users on freebsd-current
PR:		bin/187103

MFC 263048:

Repair recognition of "CC" as an alias for the C++ compiler, since it
was silently broken by upstream for a Windows-specific use-case.

Apparently some versions of CMake still rely on this archaic feature...

Reported by:	rakuco

MFC 263049:

Garbage collect the old way of adding the libstdc++ include directories
in clang's InitHeaderSearch.cpp.  This has been superseded by David
Chisnall's commit in r255321.

Moreover, if libc++ is used, the libstdc++ include directories should
not be in the search path at all.  These directories are now only used
if you pass -stdlib=libstdc++.

1 files changed, 170 insertions, 97 deletions

diff --git a/contrib/llvm/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c b/contrib/llvm/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c
index e40edba..c81a857 100644
--- a/contrib/llvm/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c
+++ b/contrib/llvm/lib/Target/X86/Disassembler/X86DisassemblerDecoder.c
@@ -25,8 +25,6 @@
 #define TRUE  1
 #define FALSE 0
 
-typedef int8_t bool;
-
 #ifndef NDEBUG
 #define debug(s) do { x86DisassemblerDebug(__FILE__, __LINE__, s); } while (0)
 #else
@@ -81,6 +79,15 @@ 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 +129,15 @@ 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) {
@@ -305,6 +321,7 @@ static int readPrefixes(struct InternalInstruction* insn) {
   BOOL prefixGroups[4] = { FALSE };
   uint64_t prefixLocation;
   uint8_t byte = 0;
+  uint8_t nextByte;
 
   BOOL hasAdSize = FALSE;
   BOOL hasOpSize = FALSE;
@@ -314,20 +331,42 @@ static int readPrefixes(struct InternalInstruction* insn) {
   while (isPrefix) {
     prefixLocation = insn->readerCursor;
 
+    /* If we fail reading prefixes, just stop here and let the opcode reader deal with it */
     if (consumeByte(insn, &byte))
-      return -1;
+      break;
 
     /*
      * If the byte is a LOCK/REP/REPNE prefix and not a part of the opcode, then
      * break and let it be disassembled as a normal "instruction".
      */
+    if (insn->readerCursor - 1 == insn->startLocation && byte == 0xf0)
+      break;
+
     if (insn->readerCursor - 1 == insn->startLocation
-        && (byte == 0xf0 || byte == 0xf2 || byte == 0xf3)) {
-      uint8_t nextByte;
-      if (byte == 0xf0)
-        break;
-      if (lookAtByte(insn, &nextByte))
-        return -1;
+        && (byte == 0xf2 || byte == 0xf3)
+        && !lookAtByte(insn, &nextByte))
+    {
+      /*
+       * If the byte is 0xf2 or 0xf3, and any of the following conditions are
+       * met:
+       * - it is followed by a LOCK (0xf0) prefix
+       * - it is followed by an xchg instruction
+       * then it should be disassembled as a xacquire/xrelease not repne/rep.
+       */
+      if ((byte == 0xf2 || byte == 0xf3) &&
+          ((nextByte == 0xf0) |
+          ((nextByte & 0xfe) == 0x86 || (nextByte & 0xf8) == 0x90)))
+        insn->xAcquireRelease = TRUE;
+      /*
+       * Also if the byte is 0xf3, and the following condition is met:
+       * - it is followed by a "mov mem, reg" (opcode 0x88/0x89) or
+       *                       "mov mem, imm" (opcode 0xc6/0xc7) instructions.
+       * then it should be disassembled as an xrelease not rep.
+       */
+      if (byte == 0xf3 &&
+          (nextByte == 0x88 || nextByte == 0x89 ||
+           nextByte == 0xc6 || nextByte == 0xc7))
+        insn->xAcquireRelease = TRUE;
       if (insn->mode == MODE_64BIT && (nextByte & 0xf0) == 0x40) {
         if (consumeByte(insn, &nextByte))
           return -1;
@@ -405,7 +444,7 @@ 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;
@@ -416,7 +455,7 @@ 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 {
@@ -424,22 +463,22 @@ 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;
@@ -448,7 +487,9 @@ 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) {
@@ -460,22 +501,22 @@ 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->vexSize == 2) {
-      insn->vexPrefix[0] = byte;
-      consumeByte(insn, &insn->vexPrefix[1]);
+    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->vexPrefix[1]) << 2);
+                        | (rFromVEX2of2(insn->vexXopPrefix[1]) << 2);
       }
 
-      switch (ppFromVEX2of2(insn->vexPrefix[1]))
+      switch (ppFromVEX2of2(insn->vexXopPrefix[1]))
       {
       default:
         break;
@@ -484,7 +525,53 @@ 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 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->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
+                        | (wFromXOP3of3(insn->vexXopPrefix[2]) << 3)
+                        | (rFromXOP2of3(insn->vexXopPrefix[1]) << 2)
+                        | (xFromXOP2of3(insn->vexXopPrefix[1]) << 1)
+                        | (bFromXOP2of3(insn->vexXopPrefix[1]) << 0);
+      }
+
+      switch (ppFromXOP3of3(insn->vexXopPrefix[2]))
+      {
+      default:
+        break;
+      case VEX_PREFIX_66:
+        hasOpSize = TRUE;
+        break;
+      }
+
+      dbgprintf(insn, "Found XOP prefix 0x%hhx 0x%hhx 0x%hhx",
+                insn->vexXopPrefix[0], insn->vexXopPrefix[1],
+                insn->vexXopPrefix[2]);
     }
   }
   else {
@@ -559,37 +646,49 @@ 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 0:
-      break;
     case VEX_LOB_0F:
-      insn->twoByteEscape = 0x0f;
       insn->opcodeType = TWOBYTE;
       return consumeByte(insn, &insn->opcode);
     case VEX_LOB_0F38:
-      insn->twoByteEscape = 0x0f;
-      insn->threeByteEscape = 0x38;
       insn->opcodeType = THREEBYTE_38;
       return consumeByte(insn, &insn->opcode);
     case VEX_LOB_0F3A:
-      insn->twoByteEscape = 0x0f;
-      insn->threeByteEscape = 0x3a;
       insn->opcodeType = THREEBYTE_3A;
       return consumeByte(insn, &insn->opcode);
     }
   }
-  else if (insn->vexSize == 2)
+  else if (insn->vexXopType == TYPE_VEX_2B)
   {
-    insn->twoByteEscape = 0x0f;
     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;
@@ -597,16 +696,12 @@ static int readOpcode(struct InternalInstruction* insn) {
   if (current == 0x0f) {
     dbgprintf(insn, "Found a two-byte escape prefix (0x%hhx)", current);
 
-    insn->twoByteEscape = current;
-
     if (consumeByte(insn, &current))
       return -1;
 
     if (current == 0x38) {
       dbgprintf(insn, "Found a three-byte escape prefix (0x%hhx)", current);
 
-      insn->threeByteEscape = current;
-
       if (consumeByte(insn, &current))
         return -1;
 
@@ -614,8 +709,6 @@ static int readOpcode(struct InternalInstruction* insn) {
     } else if (current == 0x3a) {
       dbgprintf(insn, "Found a three-byte escape prefix (0x%hhx)", current);
 
-      insn->threeByteEscape = current;
-
       if (consumeByte(insn, &current))
         return -1;
 
@@ -623,8 +716,6 @@ static int readOpcode(struct InternalInstruction* insn) {
     } else if (current == 0xa6) {
       dbgprintf(insn, "Found a three-byte escape prefix (0x%hhx)", current);
 
-      insn->threeByteEscape = current;
-
       if (consumeByte(insn, &current))
         return -1;
 
@@ -632,8 +723,6 @@ static int readOpcode(struct InternalInstruction* insn) {
     } else if (current == 0xa7) {
       dbgprintf(insn, "Found a three-byte escape prefix (0x%hhx)", current);
 
-      insn->threeByteEscape = current;
-
       if (consumeByte(insn, &current))
         return -1;
 
@@ -747,11 +836,27 @@ 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;
@@ -763,11 +868,11 @@ 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;
@@ -779,7 +884,7 @@ static int getID(struct InternalInstruction* insn, const void *miiArg) {
         break;
       }
 
-      if (lFromVEX2of2(insn->vexPrefix[1]))
+      if (lFromXOP3of3(insn->vexXopPrefix[2]))
         attrMask |= ATTR_VEXL;
     }
     else {
@@ -805,42 +910,6 @@ static int getID(struct InternalInstruction* insn, const void *miiArg) {
 
   /* The following clauses compensate for limitations of the tables. */
 
-  if ((attrMask & ATTR_VEXL) && (attrMask & ATTR_REXW) &&
-      !(attrMask & ATTR_OPSIZE)) {
-    /*
-     * Some VEX instructions ignore the L-bit, but use the W-bit. Normally L-bit
-     * has precedence since there are no L-bit with W-bit entries in the tables.
-     * So if the L-bit isn't significant we should use the W-bit instead.
-     * We only need to do this if the instruction doesn't specify OpSize since
-     * there is a VEX_L_W_OPSIZE table.
-     */
-
-    const struct InstructionSpecifier *spec;
-    uint16_t instructionIDWithWBit;
-    const struct InstructionSpecifier *specWithWBit;
-
-    spec = specifierForUID(instructionID);
-
-    if (getIDWithAttrMask(&instructionIDWithWBit,
-                          insn,
-                          (attrMask & (~ATTR_VEXL)) | ATTR_REXW)) {
-      insn->instructionID = instructionID;
-      insn->spec = spec;
-      return 0;
-    }
-
-    specWithWBit = specifierForUID(instructionIDWithWBit);
-
-    if (instructionID != instructionIDWithWBit) {
-      insn->instructionID = instructionIDWithWBit;
-      insn->spec = specWithWBit;
-    } else {
-      insn->instructionID = instructionID;
-      insn->spec = spec;
-    }
-    return 0;
-  }
-
   if (insn->prefixPresent[0x66] && !(attrMask & ATTR_OPSIZE)) {
     /*
      * The instruction tables make no distinction between instructions that
@@ -1234,6 +1303,8 @@ static int readModRM(struct InternalInstruction* insn) {
       return prefix##_EAX + index;                        \
     case TYPE_R64:                                        \
       return prefix##_RAX + index;                        \
+    case TYPE_XMM512:                                     \
+      return prefix##_ZMM0 + index;                       \
     case TYPE_XMM256:                                     \
       return prefix##_YMM0 + index;                       \
     case TYPE_XMM128:                                     \
@@ -1479,10 +1550,12 @@ 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;