Merge

make/jprt.properties

@@ -335,7 +335,19 @@ jprt.test.targets = \
 
 # The default test/Makefile targets that should be run
 
-# Example:
-#   jprt.make.rule.test.targets=*-*-*-packtest
 #jprt.make.rule.test.targets=*-product-*-packtest
 
+jprt.make.rule.test.targets = \
+  ${jprt.my.solaris.sparc}-*-c1-clienttest, \
+  ${jprt.my.solaris.i586}-*-c1-clienttest, \
+  ${jprt.my.linux.i586}-*-c1-clienttest, \
+  ${jprt.my.windows.i586}-*-c1-clienttest, \
+  ${jprt.my.solaris.sparc}-*-c2-servertest, \
+  ${jprt.my.solaris.sparcv9}-*-c2-servertest, \
+  ${jprt.my.solaris.i586}-*-c2-servertest, \
+  ${jprt.my.solaris.x64}-*-c2-servertest, \
+  ${jprt.my.linux.i586}-*-c2-servertest, \
+  ${jprt.my.linux.x64}-*-c2-servertest, \
+  ${jprt.my.windows.i586}-*-c2-servertest, \
+  ${jprt.my.windows.x64}-*-c2-servertest
+

make/linux/makefiles/gcc.make

@@ -171,3 +171,9 @@ DEBUG_CFLAGS += $(DEBUG_CFLAGS/$(BUILDARCH))
 ifeq ($(DEBUG_CFLAGS/$(BUILDARCH)),)
 DEBUG_CFLAGS += -gstabs
 endif
+
+# DEBUG_BINARIES overrides everything, use full -g debug information
+ifeq ($(DEBUG_BINARIES), true)
+  DEBUG_CFLAGS = -g
+  CFLAGS += $(DEBUG_CFLAGS)
+endif

make/linux/makefiles/jsig.make

@@ -41,10 +41,15 @@ LIBJSIG_MAPFILE = $(MAKEFILES_DIR)/mapfile-vers-jsig
 
 LFLAGS_JSIG += -D_GNU_SOURCE -D_REENTRANT $(LDFLAGS_HASH_STYLE)
 
+# DEBUG_BINARIES overrides everything, use full -g debug information
+ifeq ($(DEBUG_BINARIES), true)
+  JSIG_DEBUG_CFLAGS = -g
+endif
+
 $(LIBJSIG): $(JSIGSRCDIR)/jsig.c $(LIBJSIG_MAPFILE)
 	@echo Making signal interposition lib...
 	$(QUIETLY) $(CC) $(SYMFLAG) $(ARCHFLAG) $(SHARED_FLAG) $(PICFLAG) \
-                         $(LFLAGS_JSIG) -o $@ $< -ldl
+                         $(LFLAGS_JSIG) $(JSIG_DEBUG_CFLAGS) -o $@ $< -ldl
 
 install_jsig: $(LIBJSIG)
 	@echo "Copying $(LIBJSIG) to $(DEST_JSIG)"

make/linux/makefiles/saproc.make

@@ -43,6 +43,11 @@ SAMAPFILE = $(SASRCDIR)/mapfile
 
 DEST_SAPROC = $(JDK_LIBDIR)/$(LIBSAPROC)
 
+# DEBUG_BINARIES overrides everything, use full -g debug information
+ifeq ($(DEBUG_BINARIES), true)
+  SA_DEBUG_CFLAGS = -g
+endif
+
 # if $(AGENT_DIR) does not exist, we don't build SA
 # also, we don't build SA on Itanium.
 
@@ -67,6 +72,7 @@ $(LIBSAPROC): $(SASRCFILES) $(SAMAPFILE)
 	           -I$(BOOT_JAVA_HOME)/include/$(Platform_os_family)    \
 	           $(SASRCFILES)                                        \
 	           $(SA_LFLAGS)                                         \
+	           $(SA_DEBUG_CFLAGS)                                   \
 	           -o $@                                                \
 	           -lthread_db
 

src/cpu/sparc/vm/c1_LIRAssembler_sparc.cpp

@@ -378,7 +378,7 @@ void LIR_Assembler::emit_exception_handler() {
   compilation()->offsets()->set_value(CodeOffsets::Exceptions, code_offset());
 
 
-  if (compilation()->has_exception_handlers() || JvmtiExport::can_post_exceptions()) {
+  if (compilation()->has_exception_handlers() || compilation()->env()->jvmti_can_post_exceptions()) {
     __ call(Runtime1::entry_for(Runtime1::handle_exception_id), relocInfo::runtime_call_type);
     __ delayed()->nop();
   }

src/cpu/sparc/vm/c1_MacroAssembler_sparc.cpp

@@ -286,7 +286,7 @@ void C1_MacroAssembler::initialize_object(
     initialize_body(base, index);
   }
 
-  if (DTraceAllocProbes) {
+  if (CURRENT_ENV->dtrace_alloc_probes()) {
     assert(obj == O0, "must be");
     call(CAST_FROM_FN_PTR(address, Runtime1::entry_for(Runtime1::dtrace_object_alloc_id)),
          relocInfo::runtime_call_type);
@@ -355,7 +355,7 @@ void C1_MacroAssembler::allocate_array(
   sub(arr_size, hdr_size * wordSize, index);         // compute index = number of words to clear
   initialize_body(base, index);
 
-  if (DTraceAllocProbes) {
+  if (CURRENT_ENV->dtrace_alloc_probes()) {
     assert(obj == O0, "must be");
     call(CAST_FROM_FN_PTR(address, Runtime1::entry_for(Runtime1::dtrace_object_alloc_id)),
          relocInfo::runtime_call_type);

src/cpu/sparc/vm/sparc.ad

@@ -1712,6 +1712,23 @@ static FloatRegister reg_to_DoubleFloatRegister_object(int register_encoding) {
   return as_DoubleFloatRegister(register_encoding);
 }
 
+const bool Matcher::match_rule_supported(int opcode) {
+  if (!has_match_rule(opcode))
+    return false;
+
+  switch (opcode) {
+  case Op_CountLeadingZerosI:
+  case Op_CountLeadingZerosL:
+  case Op_CountTrailingZerosI:
+  case Op_CountTrailingZerosL:
+    if (!UsePopCountInstruction)
+      return false;
+    break;
+  }
+
+  return true;  // Per default match rules are supported.
+}
+
 int Matcher::regnum_to_fpu_offset(int regnum) {
   return regnum - 32; // The FP registers are in the second chunk
 }
@@ -1874,15 +1891,17 @@ RegMask Matcher::modL_proj_mask() {
 // The intptr_t operand types, defined by textual substitution.
 // (Cf. opto/type.hpp.  This lets us avoid many, many other ifdefs.)
 #ifdef _LP64
-#define immX    immL
-#define immX13  immL13
-#define iRegX   iRegL
-#define g1RegX  g1RegL
+#define immX     immL
+#define immX13   immL13
+#define immX13m7 immL13m7
+#define iRegX    iRegL
+#define g1RegX   g1RegL
 #else
-#define immX    immI
-#define immX13  immI13
-#define iRegX   iRegI
-#define g1RegX  g1RegI
+#define immX     immI
+#define immX13   immI13
+#define immX13m7 immI13m7
+#define iRegX    iRegI
+#define g1RegX   g1RegI
 #endif
 
 //----------ENCODING BLOCK-----------------------------------------------------
@@ -3437,6 +3456,16 @@ operand immI13() %{
   interface(CONST_INTER);
 %}
 
+// Integer Immediate: 13-bit minus 7
+operand immI13m7() %{
+  predicate((-4096 < n->get_int()) && ((n->get_int() + 7) <= 4095));
+  match(ConI);
+  op_cost(0);
+
+  format %{ %}
+  interface(CONST_INTER);
+%}
+
 // Unsigned (positive) Integer Immediate: 13-bit
 operand immU13() %{
   predicate((0 <= n->get_int()) && Assembler::is_simm13(n->get_int()));
@@ -3515,6 +3544,28 @@ operand immI_32_63() %{
   interface(CONST_INTER);
 %}
 
+// Immediates for special shifts (sign extend)
+
+// Integer Immediate: the value 16
+operand immI_16() %{
+  predicate(n->get_int() == 16);
+  match(ConI);
+  op_cost(0);
+
+  format %{ %}
+  interface(CONST_INTER);
+%}
+
+// Integer Immediate: the value 24
+operand immI_24() %{
+  predicate(n->get_int() == 24);
+  match(ConI);
+  op_cost(0);
+
+  format %{ %}
+  interface(CONST_INTER);
+%}
+
 // Integer Immediate: the value 255
 operand immI_255() %{
   predicate( n->get_int() == 255 );
@@ -3525,6 +3576,16 @@ operand immI_255() %{
   interface(CONST_INTER);
 %}
 
+// Integer Immediate: the value 65535
+operand immI_65535() %{
+  predicate(n->get_int() == 65535);
+  match(ConI);
+  op_cost(0);
+
+  format %{ %}
+  interface(CONST_INTER);
+%}
+
 // Long Immediate: the value FF
 operand immL_FF() %{
   predicate( n->get_long() == 0xFFL );
@@ -3630,6 +3691,16 @@ operand immL13() %{
   interface(CONST_INTER);
 %}
 
+// Long Immediate: 13-bit minus 7
+operand immL13m7() %{
+  predicate((-4096L < n->get_long()) && ((n->get_long() + 7L) <= 4095L));
+  match(ConL);
+  op_cost(0);
+
+  format %{ %}
+  interface(CONST_INTER);
+%}
+
 // Long Immediate: low 32-bit mask
 operand immL_32bits() %{
   predicate(n->get_long() == 0xFFFFFFFFL);
@@ -4067,7 +4138,7 @@ operand indirect(sp_ptr_RegP reg) %{
   %}
 %}
 
-// Indirect with Offset
+// Indirect with simm13 Offset
 operand indOffset13(sp_ptr_RegP reg, immX13 offset) %{
   constraint(ALLOC_IN_RC(sp_ptr_reg));
   match(AddP reg offset);
@@ -4082,6 +4153,21 @@ operand indOffset13(sp_ptr_RegP reg, immX13 offset) %{
   %}
 %}
 
+// Indirect with simm13 Offset minus 7
+operand indOffset13m7(sp_ptr_RegP reg, immX13m7 offset) %{
+  constraint(ALLOC_IN_RC(sp_ptr_reg));
+  match(AddP reg offset);
+
+  op_cost(100);
+  format %{ "[$reg + $offset]" %}
+  interface(MEMORY_INTER) %{
+    base($reg);
+    index(0x0);
+    scale(0x0);
+    disp($offset);
+  %}
+%}
+
 // Note:  Intel has a swapped version also, like this:
 //operand indOffsetX(iRegI reg, immP offset) %{
 //  constraint(ALLOC_IN_RC(int_reg));
@@ -5487,6 +5573,20 @@ instruct loadS(iRegI dst, memory mem) %{
   ins_pipe(iload_mask_mem);
 %}
 
+// Load Short (16 bit signed) to Byte (8 bit signed)
+instruct loadS2B(iRegI dst, indOffset13m7 mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadS mem) twentyfour) twentyfour));
+  ins_cost(MEMORY_REF_COST);
+
+  size(4);
+
+  format %{ "LDSB   $mem+1,$dst\t! short -> byte" %}
+  ins_encode %{
+    __ ldsb($mem$$Address, $dst$$Register, 1);
+  %}
+  ins_pipe(iload_mask_mem);
+%}
+
 // Load Short (16bit signed) into a Long Register
 instruct loadS2L(iRegL dst, memory mem) %{
   match(Set dst (ConvI2L (LoadS mem)));
@@ -5513,6 +5613,19 @@ instruct loadUS(iRegI dst, memory mem) %{
   ins_pipe(iload_mask_mem);
 %}
 
+// Load Unsigned Short/Char (16 bit UNsigned) to Byte (8 bit signed)
+instruct loadUS2B(iRegI dst, indOffset13m7 mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadUS mem) twentyfour) twentyfour));
+  ins_cost(MEMORY_REF_COST);
+
+  size(4);
+  format %{ "LDSB   $mem+1,$dst\t! ushort -> byte" %}
+  ins_encode %{
+    __ ldsb($mem$$Address, $dst$$Register, 1);
+  %}
+  ins_pipe(iload_mask_mem);
+%}
+
 // Load Unsigned Short/Char (16bit UNsigned) into a Long Register
 instruct loadUS2L(iRegL dst, memory mem) %{
   match(Set dst (ConvI2L (LoadUS mem)));
@@ -5539,6 +5652,62 @@ instruct loadI(iRegI dst, memory mem) %{
   ins_pipe(iload_mem);
 %}
 
+// Load Integer to Byte (8 bit signed)
+instruct loadI2B(iRegI dst, indOffset13m7 mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadI mem) twentyfour) twentyfour));
+  ins_cost(MEMORY_REF_COST);
+
+  size(4);
+
+  format %{ "LDSB   $mem+3,$dst\t! int -> byte" %}
+  ins_encode %{
+    __ ldsb($mem$$Address, $dst$$Register, 3);
+  %}
+  ins_pipe(iload_mask_mem);
+%}
+
+// Load Integer to Unsigned Byte (8 bit UNsigned)
+instruct loadI2UB(iRegI dst, indOffset13m7 mem, immI_255 mask) %{
+  match(Set dst (AndI (LoadI mem) mask));
+  ins_cost(MEMORY_REF_COST);
+
+  size(4);
+
+  format %{ "LDUB   $mem+3,$dst\t! int -> ubyte" %}
+  ins_encode %{
+    __ ldub($mem$$Address, $dst$$Register, 3);
+  %}
+  ins_pipe(iload_mask_mem);
+%}
+
+// Load Integer to Short (16 bit signed)
+instruct loadI2S(iRegI dst, indOffset13m7 mem, immI_16 sixteen) %{
+  match(Set dst (RShiftI (LShiftI (LoadI mem) sixteen) sixteen));
+  ins_cost(MEMORY_REF_COST);
+
+  size(4);
+
+  format %{ "LDSH   $mem+2,$dst\t! int -> short" %}
+  ins_encode %{
+    __ ldsh($mem$$Address, $dst$$Register, 2);
+  %}
+  ins_pipe(iload_mask_mem);
+%}
+
+// Load Integer to Unsigned Short (16 bit UNsigned)
+instruct loadI2US(iRegI dst, indOffset13m7 mem, immI_65535 mask) %{
+  match(Set dst (AndI (LoadI mem) mask));
+  ins_cost(MEMORY_REF_COST);
+
+  size(4);
+
+  format %{ "LDUH   $mem+2,$dst\t! int -> ushort/char" %}
+  ins_encode %{
+    __ lduh($mem$$Address, $dst$$Register, 2);
+  %}
+  ins_pipe(iload_mask_mem);
+%}
+
 // Load Integer into a Long Register
 instruct loadI2L(iRegL dst, memory mem) %{
   match(Set dst (ConvI2L (LoadI mem)));
@@ -9188,6 +9357,145 @@ instruct array_equals(o0RegP ary1, o1RegP ary2, g3RegP tmp1, g4RegP tmp2, notemp
   ins_pipe(long_memory_op);
 %}
 
+
+//---------- Zeros Count Instructions ------------------------------------------
+
+instruct countLeadingZerosI(iRegI dst, iRegI src, iRegI tmp, flagsReg cr) %{
+  predicate(UsePopCountInstruction);  // See Matcher::match_rule_supported
+  match(Set dst (CountLeadingZerosI src));
+  effect(TEMP dst, TEMP tmp, KILL cr);
+
+  // x |= (x >> 1);
+  // x |= (x >> 2);
+  // x |= (x >> 4);
+  // x |= (x >> 8);
+  // x |= (x >> 16);
+  // return (WORDBITS - popc(x));
+  format %{ "SRL     $src,1,$dst\t! count leading zeros (int)\n\t"
+            "OR      $src,$tmp,$dst\n\t"
+            "SRL     $dst,2,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "SRL     $dst,4,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "SRL     $dst,8,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "SRL     $dst,16,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "POPC    $dst,$dst\n\t"
+            "MOV     32,$tmp\n\t"
+            "SUB     $tmp,$dst,$dst" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Register Rtmp = $tmp$$Register;
+    __ srl(Rsrc, 1, Rtmp);
+    __ or3(Rsrc, Rtmp, Rdst);
+    __ srl(Rdst, 2, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ srl(Rdst, 4, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ srl(Rdst, 8, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ srl(Rdst, 16, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ popc(Rdst, Rdst);
+    __ mov(BitsPerInt, Rtmp);
+    __ sub(Rtmp, Rdst, Rdst);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countLeadingZerosL(iRegI dst, iRegL src, iRegL tmp, flagsReg cr) %{
+  predicate(UsePopCountInstruction);  // See Matcher::match_rule_supported
+  match(Set dst (CountLeadingZerosL src));
+  effect(TEMP dst, TEMP tmp, KILL cr);
+
+  // x |= (x >> 1);
+  // x |= (x >> 2);
+  // x |= (x >> 4);
+  // x |= (x >> 8);
+  // x |= (x >> 16);
+  // x |= (x >> 32);
+  // return (WORDBITS - popc(x));
+  format %{ "SRLX    $src,1,$dst\t! count leading zeros (long)\n\t"
+            "OR      $src,$tmp,$dst\n\t"
+            "SRLX    $dst,2,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "SRLX    $dst,4,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "SRLX    $dst,8,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "SRLX    $dst,16,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "SRLX    $dst,32,$tmp\n\t"
+            "OR      $dst,$tmp,$dst\n\t"
+            "POPC    $dst,$dst\n\t"
+            "MOV     64,$tmp\n\t"
+            "SUB     $tmp,$dst,$dst" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Register Rtmp = $tmp$$Register;
+    __ srlx(Rsrc, 1, Rtmp);
+    __ or3(Rsrc, Rtmp, Rdst);
+    __ srlx(Rdst, 2, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ srlx(Rdst, 4, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ srlx(Rdst, 8, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ srlx(Rdst, 16, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ srlx(Rdst, 32, Rtmp);
+    __ or3(Rdst, Rtmp, Rdst);
+    __ popc(Rdst, Rdst);
+    __ mov(BitsPerLong, Rtmp);
+    __ sub(Rtmp, Rdst, Rdst);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countTrailingZerosI(iRegI dst, iRegI src, flagsReg cr) %{
+  predicate(UsePopCountInstruction);  // See Matcher::match_rule_supported
+  match(Set dst (CountTrailingZerosI src));
+  effect(TEMP dst, KILL cr);
+
+  // return popc(~x & (x - 1));
+  format %{ "SUB     $src,1,$dst\t! count trailing zeros (int)\n\t"
+            "ANDN    $dst,$src,$dst\n\t"
+            "SRL     $dst,R_G0,$dst\n\t"
+            "POPC    $dst,$dst" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    __ sub(Rsrc, 1, Rdst);
+    __ andn(Rdst, Rsrc, Rdst);
+    __ srl(Rdst, G0, Rdst);
+    __ popc(Rdst, Rdst);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countTrailingZerosL(iRegI dst, iRegL src, flagsReg cr) %{
+  predicate(UsePopCountInstruction);  // See Matcher::match_rule_supported
+  match(Set dst (CountTrailingZerosL src));
+  effect(TEMP dst, KILL cr);
+
+  // return popc(~x & (x - 1));
+  format %{ "SUB     $src,1,$dst\t! count trailing zeros (long)\n\t"
+            "ANDN    $dst,$src,$dst\n\t"
+            "POPC    $dst,$dst" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    __ sub(Rsrc, 1, Rdst);
+    __ andn(Rdst, Rsrc, Rdst);
+    __ popc(Rdst, Rdst);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+
 //---------- Population Count Instructions -------------------------------------
 
 instruct popCountI(iRegI dst, iRegI src) %{

src/cpu/x86/vm/assembler_x86.cpp

@@ -952,6 +952,21 @@ void Assembler::andpd(XMMRegister dst, Address src) {
   emit_operand(dst, src);
 }
 
+void Assembler::bsfl(Register dst, Register src) {
+  int encode = prefix_and_encode(dst->encoding(), src->encoding());
+  emit_byte(0x0F);
+  emit_byte(0xBC);
+  emit_byte(0xC0 | encode);
+}
+
+void Assembler::bsrl(Register dst, Register src) {
+  assert(!VM_Version::supports_lzcnt(), "encoding is treated as LZCNT");
+  int encode = prefix_and_encode(dst->encoding(), src->encoding());
+  emit_byte(0x0F);
+  emit_byte(0xBD);
+  emit_byte(0xC0 | encode);
+}
+
 void Assembler::bswapl(Register reg) { // bswap
   int encode = prefix_and_encode(reg->encoding());
   emit_byte(0x0F);
@@ -1438,6 +1453,15 @@ void Assembler::lock() {
   }
 }
 
+void Assembler::lzcntl(Register dst, Register src) {
+  assert(VM_Version::supports_lzcnt(), "encoding is treated as BSR");
+  emit_byte(0xF3);
+  int encode = prefix_and_encode(dst->encoding(), src->encoding());
+  emit_byte(0x0F);
+  emit_byte(0xBD);
+  emit_byte(0xC0 | encode);
+}
+
 // Emit mfence instruction
 void Assembler::mfence() {
   NOT_LP64(assert(VM_Version::supports_sse2(), "unsupported");)
@@ -3688,6 +3712,21 @@ void Assembler::andq(Register dst, Register src) {
   emit_arith(0x23, 0xC0, dst, src);
 }
 
+void Assembler::bsfq(Register dst, Register src) {
+  int encode = prefixq_and_encode(dst->encoding(), src->encoding());
+  emit_byte(0x0F);
+  emit_byte(0xBC);
+  emit_byte(0xC0 | encode);
+}
+
+void Assembler::bsrq(Register dst, Register src) {
+  assert(!VM_Version::supports_lzcnt(), "encoding is treated as LZCNT");
+  int encode = prefixq_and_encode(dst->encoding(), src->encoding());
+  emit_byte(0x0F);
+  emit_byte(0xBD);
+  emit_byte(0xC0 | encode);
+}
+
 void Assembler::bswapq(Register reg) {
   int encode = prefixq_and_encode(reg->encoding());
   emit_byte(0x0F);
@@ -3941,6 +3980,15 @@ void Assembler::cmp_narrow_oop(Address src1, int32_t imm32, RelocationHolder con
   emit_data((int)imm32, rspec, narrow_oop_operand);
 }
 
+void Assembler::lzcntq(Register dst, Register src) {
+  assert(VM_Version::supports_lzcnt(), "encoding is treated as BSR");
+  emit_byte(0xF3);
+  int encode = prefixq_and_encode(dst->encoding(), src->encoding());
+  emit_byte(0x0F);
+  emit_byte(0xBD);
+  emit_byte(0xC0 | encode);
+}
+
 void Assembler::movdq(XMMRegister dst, Register src) {
   // table D-1 says MMX/SSE2
   NOT_LP64(assert(VM_Version::supports_sse2() || VM_Version::supports_mmx(), ""));

src/cpu/x86/vm/assembler_x86.hpp

@@ -757,6 +757,14 @@ private:
   void andpd(XMMRegister dst, Address src);
   void andpd(XMMRegister dst, XMMRegister src);
 
+  void bsfl(Register dst, Register src);
+  void bsrl(Register dst, Register src);
+
+#ifdef _LP64
+  void bsfq(Register dst, Register src);
+  void bsrq(Register dst, Register src);
+#endif
+
   void bswapl(Register reg);
 
   void bswapq(Register reg);
@@ -1061,6 +1069,12 @@ private:
 
   void lock();
 
+  void lzcntl(Register dst, Register src);
+
+#ifdef _LP64
+  void lzcntq(Register dst, Register src);
+#endif
+
   enum Membar_mask_bits {
     StoreStore = 1 << 3,
     LoadStore  = 1 << 2,

src/cpu/x86/vm/c1_LIRAssembler_x86.cpp

@@ -439,7 +439,7 @@ void LIR_Assembler::emit_exception_handler() {
 
   // if the method does not have an exception handler, then there is
   // no reason to search for one
-  if (compilation()->has_exception_handlers() || JvmtiExport::can_post_exceptions()) {
+  if (compilation()->has_exception_handlers() || compilation()->env()->jvmti_can_post_exceptions()) {
     // the exception oop and pc are in rax, and rdx
     // no other registers need to be preserved, so invalidate them
     __ invalidate_registers(false, true, true, false, true, true);

src/cpu/x86/vm/c1_MacroAssembler_x86.cpp

@@ -258,7 +258,7 @@ void C1_MacroAssembler::initialize_object(Register obj, Register klass, Register
     }
   }
 
-  if (DTraceAllocProbes) {
+  if (CURRENT_ENV->dtrace_alloc_probes()) {
     assert(obj == rax, "must be");
     call(RuntimeAddress(Runtime1::entry_for(Runtime1::dtrace_object_alloc_id)));
   }
@@ -291,7 +291,7 @@ void C1_MacroAssembler::allocate_array(Register obj, Register len, Register t1,
   const Register len_zero = len;
   initialize_body(obj, arr_size, header_size * BytesPerWord, len_zero);
 
-  if (DTraceAllocProbes) {
+  if (CURRENT_ENV->dtrace_alloc_probes()) {
     assert(obj == rax, "must be");
     call(RuntimeAddress(Runtime1::entry_for(Runtime1::dtrace_object_alloc_id)));
   }

src/cpu/x86/vm/vm_version_x86.cpp

@@ -284,7 +284,7 @@ void VM_Version::get_processor_features() {
   }
 
   char buf[256];
-  jio_snprintf(buf, sizeof(buf), "(%u cores per cpu, %u threads per core) family %d model %d stepping %d%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
+  jio_snprintf(buf, sizeof(buf), "(%u cores per cpu, %u threads per core) family %d model %d stepping %d%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
                cores_per_cpu(), threads_per_core(),
                cpu_family(), _model, _stepping,
                (supports_cmov() ? ", cmov" : ""),
@@ -301,6 +301,7 @@ void VM_Version::get_processor_features() {
                (supports_mmx_ext() ? ", mmxext" : ""),
                (supports_3dnow()   ? ", 3dnow"  : ""),
                (supports_3dnow2()  ? ", 3dnowext" : ""),
+               (supports_lzcnt()   ? ", lzcnt": ""),
                (supports_sse4a()   ? ", sse4a": ""),
                (supports_ht() ? ", ht": ""));
   _features_str = strdup(buf);
@@ -364,6 +365,13 @@ void VM_Version::get_processor_features() {
         UseXmmI2D = false;
       }
     }
+
+    // Use count leading zeros count instruction if available.
+    if (supports_lzcnt()) {
+      if (FLAG_IS_DEFAULT(UseCountLeadingZerosInstruction)) {
+        UseCountLeadingZerosInstruction = true;
+      }
+    }
   }
 
   if( is_intel() ) { // Intel cpus specific settings

src/cpu/x86/vm/vm_version_x86.hpp

@@ -120,7 +120,7 @@ public:
       uint32_t LahfSahf     : 1,
                CmpLegacy    : 1,
                             : 4,
-               abm          : 1,
+               lzcnt        : 1,
                sse4a        : 1,
                misalignsse  : 1,
                prefetchw    : 1,
@@ -182,7 +182,8 @@ protected:
      CPU_SSE4A  = (1 << 10),
      CPU_SSE4_1 = (1 << 11),
      CPU_SSE4_2 = (1 << 12),
-     CPU_POPCNT = (1 << 13)
+     CPU_POPCNT = (1 << 13),
+     CPU_LZCNT  = (1 << 14)
    } cpuFeatureFlags;
 
   // cpuid information block.  All info derived from executing cpuid with
@@ -277,8 +278,6 @@ protected:
     if (_cpuid_info.std_cpuid1_edx.bits.mmx != 0 || is_amd() &&
         _cpuid_info.ext_cpuid1_edx.bits.mmx != 0)
       result |= CPU_MMX;
-    if (is_amd() && _cpuid_info.ext_cpuid1_edx.bits.tdnow != 0)
-      result |= CPU_3DNOW;
     if (_cpuid_info.std_cpuid1_edx.bits.sse != 0)
       result |= CPU_SSE;
     if (_cpuid_info.std_cpuid1_edx.bits.sse2 != 0)
@@ -287,14 +286,23 @@ protected:
       result |= CPU_SSE3;
     if (_cpuid_info.std_cpuid1_ecx.bits.ssse3 != 0)
       result |= CPU_SSSE3;
-    if (is_amd() && _cpuid_info.ext_cpuid1_ecx.bits.sse4a != 0)
-      result |= CPU_SSE4A;
     if (_cpuid_info.std_cpuid1_ecx.bits.sse4_1 != 0)
       result |= CPU_SSE4_1;
     if (_cpuid_info.std_cpuid1_ecx.bits.sse4_2 != 0)
       result |= CPU_SSE4_2;
     if (_cpuid_info.std_cpuid1_ecx.bits.popcnt != 0)
       result |= CPU_POPCNT;
+
+    // AMD features.
+    if (is_amd()) {
+      if (_cpuid_info.ext_cpuid1_edx.bits.tdnow != 0)
+        result |= CPU_3DNOW;
+      if (_cpuid_info.ext_cpuid1_ecx.bits.lzcnt != 0)
+        result |= CPU_LZCNT;
+      if (_cpuid_info.ext_cpuid1_ecx.bits.sse4a != 0)
+        result |= CPU_SSE4A;
+    }
+
     return result;
   }
 
@@ -391,6 +399,7 @@ public:
   static bool supports_3dnow()    { return (_cpuFeatures & CPU_3DNOW) != 0; }
   static bool supports_mmx_ext()  { return is_amd() && _cpuid_info.ext_cpuid1_edx.bits.mmx_amd != 0; }
   static bool supports_3dnow2()   { return is_amd() && _cpuid_info.ext_cpuid1_edx.bits.tdnow2 != 0; }
+  static bool supports_lzcnt()    { return (_cpuFeatures & CPU_LZCNT) != 0; }
   static bool supports_sse4a()    { return (_cpuFeatures & CPU_SSE4A) != 0; }
 
   static bool supports_compare_and_exchange() { return true; }

src/cpu/x86/vm/x86_32.ad

@@ -1281,6 +1281,13 @@ static void emit_float_constant(CodeBuffer& cbuf, float x) {
 }
 
 
+const bool Matcher::match_rule_supported(int opcode) {
+  if (!has_match_rule(opcode))
+    return false;
+
+  return true;  // Per default match rules are supported.
+}
+
 int Matcher::regnum_to_fpu_offset(int regnum) {
   return regnum - 32; // The FP registers are in the second chunk
 }
@@ -5233,6 +5240,15 @@ operand immI_255() %{
   interface(CONST_INTER);
 %}
 
+// Constant for short-wide masking
+operand immI_65535() %{
+  predicate(n->get_int() == 65535);
+  match(ConI);
+
+  format %{ %}
+  interface(CONST_INTER);
+%}
+
 // Register Operands
 // Integer Register
 operand eRegI() %{
@@ -6644,6 +6660,153 @@ instruct bytes_reverse_long(eRegL dst) %{
 %}
 
 
+//---------- Zeros Count Instructions ------------------------------------------
+
+instruct countLeadingZerosI(eRegI dst, eRegI src, eFlagsReg cr) %{
+  predicate(UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosI src));
+  effect(KILL cr);
+
+  format %{ "LZCNT  $dst, $src\t# count leading zeros (int)" %}
+  ins_encode %{
+    __ lzcntl($dst$$Register, $src$$Register);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countLeadingZerosI_bsr(eRegI dst, eRegI src, eFlagsReg cr) %{
+  predicate(!UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosI src));
+  effect(KILL cr);
+
+  format %{ "BSR    $dst, $src\t# count leading zeros (int)\n\t"
+            "JNZ    skip\n\t"
+            "MOV    $dst, -1\n"
+      "skip:\n\t"
+            "NEG    $dst\n\t"
+            "ADD    $dst, 31" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Label skip;
+    __ bsrl(Rdst, Rsrc);
+    __ jccb(Assembler::notZero, skip);
+    __ movl(Rdst, -1);
+    __ bind(skip);
+    __ negl(Rdst);
+    __ addl(Rdst, BitsPerInt - 1);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countLeadingZerosL(eRegI dst, eRegL src, eFlagsReg cr) %{
+  predicate(UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosL src));
+  effect(TEMP dst, KILL cr);
+
+  format %{ "LZCNT  $dst, $src.hi\t# count leading zeros (long)\n\t"
+            "JNC    done\n\t"
+            "LZCNT  $dst, $src.lo\n\t"
+            "ADD    $dst, 32\n"
+      "done:" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Label done;
+    __ lzcntl(Rdst, HIGH_FROM_LOW(Rsrc));
+    __ jccb(Assembler::carryClear, done);
+    __ lzcntl(Rdst, Rsrc);
+    __ addl(Rdst, BitsPerInt);
+    __ bind(done);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countLeadingZerosL_bsr(eRegI dst, eRegL src, eFlagsReg cr) %{
+  predicate(!UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosL src));
+  effect(TEMP dst, KILL cr);
+
+  format %{ "BSR    $dst, $src.hi\t# count leading zeros (long)\n\t"
+            "JZ     msw_is_zero\n\t"
+            "ADD    $dst, 32\n\t"
+            "JMP    not_zero\n"
+      "msw_is_zero:\n\t"
+            "BSR    $dst, $src.lo\n\t"
+            "JNZ    not_zero\n\t"
+            "MOV    $dst, -1\n"
+      "not_zero:\n\t"
+            "NEG    $dst\n\t"
+            "ADD    $dst, 63\n" %}
+ ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Label msw_is_zero;
+    Label not_zero;
+    __ bsrl(Rdst, HIGH_FROM_LOW(Rsrc));
+    __ jccb(Assembler::zero, msw_is_zero);
+    __ addl(Rdst, BitsPerInt);
+    __ jmpb(not_zero);
+    __ bind(msw_is_zero);
+    __ bsrl(Rdst, Rsrc);
+    __ jccb(Assembler::notZero, not_zero);
+    __ movl(Rdst, -1);
+    __ bind(not_zero);
+    __ negl(Rdst);
+    __ addl(Rdst, BitsPerLong - 1);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countTrailingZerosI(eRegI dst, eRegI src, eFlagsReg cr) %{
+  match(Set dst (CountTrailingZerosI src));
+  effect(KILL cr);
+
+  format %{ "BSF    $dst, $src\t# count trailing zeros (int)\n\t"
+            "JNZ    done\n\t"
+            "MOV    $dst, 32\n"
+      "done:" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Label done;
+    __ bsfl(Rdst, $src$$Register);
+    __ jccb(Assembler::notZero, done);
+    __ movl(Rdst, BitsPerInt);
+    __ bind(done);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countTrailingZerosL(eRegI dst, eRegL src, eFlagsReg cr) %{
+  match(Set dst (CountTrailingZerosL src));
+  effect(TEMP dst, KILL cr);
+
+  format %{ "BSF    $dst, $src.lo\t# count trailing zeros (long)\n\t"
+            "JNZ    done\n\t"
+            "BSF    $dst, $src.hi\n\t"
+            "JNZ    msw_not_zero\n\t"
+            "MOV    $dst, 32\n"
+      "msw_not_zero:\n\t"
+            "ADD    $dst, 32\n"
+      "done:" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Label msw_not_zero;
+    Label done;
+    __ bsfl(Rdst, Rsrc);
+    __ jccb(Assembler::notZero, done);
+    __ bsfl(Rdst, HIGH_FROM_LOW(Rsrc));
+    __ jccb(Assembler::notZero, msw_not_zero);
+    __ movl(Rdst, BitsPerInt);
+    __ bind(msw_not_zero);
+    __ addl(Rdst, BitsPerInt);
+    __ bind(done);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+
 //---------- Population Count Instructions -------------------------------------
 
 instruct popCountI(eRegI dst, eRegI src) %{
@@ -6784,6 +6947,18 @@ instruct loadS(eRegI dst, memory mem) %{
   ins_pipe(ialu_reg_mem);
 %}
 
+// Load Short (16 bit signed) to Byte (8 bit signed)
+instruct loadS2B(eRegI dst, memory mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadS mem) twentyfour) twentyfour));
+
+  ins_cost(125);
+  format %{ "MOVSX  $dst, $mem\t# short -> byte" %}
+  ins_encode %{
+    __ movsbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
 // Load Short (16bit signed) into Long Register
 instruct loadS2L(eRegL dst, memory mem) %{
   match(Set dst (ConvI2L (LoadS mem)));
@@ -6816,9 +6991,20 @@ instruct loadUS(eRegI dst, memory mem) %{
   ins_pipe(ialu_reg_mem);
 %}
 
+// Load Unsigned Short/Char (16 bit UNsigned) to Byte (8 bit signed)
+instruct loadUS2B(eRegI dst, memory mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadUS mem) twentyfour) twentyfour));
+
+  ins_cost(125);
+  format %{ "MOVSX  $dst, $mem\t# ushort -> byte" %}
+  ins_encode %{
+    __ movsbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
 // Load Unsigned Short/Char (16 bit UNsigned) into Long Register
-instruct loadUS2L(eRegL dst, memory mem)
-%{
+instruct loadUS2L(eRegL dst, memory mem) %{
   match(Set dst (ConvI2L (LoadUS mem)));
 
   ins_cost(250);
@@ -6847,6 +7033,54 @@ instruct loadI(eRegI dst, memory mem) %{
   ins_pipe(ialu_reg_mem);
 %}
 
+// Load Integer (32 bit signed) to Byte (8 bit signed)
+instruct loadI2B(eRegI dst, memory mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadI mem) twentyfour) twentyfour));
+
+  ins_cost(125);
+  format %{ "MOVSX  $dst, $mem\t# int -> byte" %}
+  ins_encode %{
+    __ movsbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
+// Load Integer (32 bit signed) to Unsigned Byte (8 bit UNsigned)
+instruct loadI2UB(eRegI dst, memory mem, immI_255 mask) %{
+  match(Set dst (AndI (LoadI mem) mask));
+
+  ins_cost(125);
+  format %{ "MOVZX  $dst, $mem\t# int -> ubyte" %}
+  ins_encode %{
+    __ movzbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
+// Load Integer (32 bit signed) to Short (16 bit signed)
+instruct loadI2S(eRegI dst, memory mem, immI_16 sixteen) %{
+  match(Set dst (RShiftI (LShiftI (LoadI mem) sixteen) sixteen));
+
+  ins_cost(125);
+  format %{ "MOVSX  $dst, $mem\t# int -> short" %}
+  ins_encode %{
+    __ movswl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
+// Load Integer (32 bit signed) to Unsigned Short/Char (16 bit UNsigned)
+instruct loadI2US(eRegI dst, memory mem, immI_65535 mask) %{
+  match(Set dst (AndI (LoadI mem) mask));
+
+  ins_cost(125);
+  format %{ "MOVZX  $dst, $mem\t# int -> ushort/char" %}
+  ins_encode %{
+    __ movzwl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
 // Load Integer into Long Register
 instruct loadI2L(eRegL dst, memory mem) %{
   match(Set dst (ConvI2L (LoadI mem)));
@@ -8880,28 +9114,28 @@ instruct shrI_eReg_imm(eRegI dst, immI8 shift, eFlagsReg cr) %{
 
 // Logical Shift Right by 24, followed by Arithmetic Shift Left by 24.
 // This idiom is used by the compiler for the i2b bytecode.
-instruct i2b(eRegI dst, xRegI src, immI_24 twentyfour, eFlagsReg cr) %{
+instruct i2b(eRegI dst, xRegI src, immI_24 twentyfour) %{
   match(Set dst (RShiftI (LShiftI src twentyfour) twentyfour));
-  effect(KILL cr);
 
   size(3);
   format %{ "MOVSX  $dst,$src :8" %}
-  opcode(0xBE, 0x0F);
-  ins_encode( OpcS, OpcP, RegReg( dst, src));
-  ins_pipe( ialu_reg_reg );
+  ins_encode %{
+    __ movsbl($dst$$Register, $src$$Register);
+  %}
+  ins_pipe(ialu_reg_reg);
 %}
 
 // Logical Shift Right by 16, followed by Arithmetic Shift Left by 16.
 // This idiom is used by the compiler the i2s bytecode.
-instruct i2s(eRegI dst, xRegI src, immI_16 sixteen, eFlagsReg cr) %{
+instruct i2s(eRegI dst, xRegI src, immI_16 sixteen) %{
   match(Set dst (RShiftI (LShiftI src sixteen) sixteen));
-  effect(KILL cr);
 
   size(3);
   format %{ "MOVSX  $dst,$src :16" %}
-  opcode(0xBF, 0x0F);
-  ins_encode( OpcS, OpcP, RegReg( dst, src));
-  ins_pipe( ialu_reg_reg );
+  ins_encode %{
+    __ movswl($dst$$Register, $src$$Register);
+  %}
+  ins_pipe(ialu_reg_reg);
 %}
 
 

src/cpu/x86/vm/x86_64.ad

@@ -1980,6 +1980,13 @@ static void emit_float_constant(CodeBuffer& cbuf, float x) {
 }
 
 
+const bool Matcher::match_rule_supported(int opcode) {
+  if (!has_match_rule(opcode))
+    return false;
+
+  return true;  // Per default match rules are supported.
+}
+
 int Matcher::regnum_to_fpu_offset(int regnum)
 {
   return regnum - 32; // The FP registers are in the second chunk
@@ -6452,6 +6459,18 @@ instruct loadS(rRegI dst, memory mem)
   ins_pipe(ialu_reg_mem);
 %}
 
+// Load Short (16 bit signed) to Byte (8 bit signed)
+instruct loadS2B(rRegI dst, memory mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadS mem) twentyfour) twentyfour));
+
+  ins_cost(125);
+  format %{ "movsbl $dst, $mem\t# short -> byte" %}
+  ins_encode %{
+    __ movsbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
 // Load Short (16 bit signed) into Long Register
 instruct loadS2L(rRegL dst, memory mem)
 %{
@@ -6482,6 +6501,18 @@ instruct loadUS(rRegI dst, memory mem)
   ins_pipe(ialu_reg_mem);
 %}
 
+// Load Unsigned Short/Char (16 bit UNsigned) to Byte (8 bit signed)
+instruct loadUS2B(rRegI dst, memory mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadUS mem) twentyfour) twentyfour));
+
+  ins_cost(125);
+  format %{ "movsbl $dst, $mem\t# ushort -> byte" %}
+  ins_encode %{
+    __ movsbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
 // Load Unsigned Short/Char (16 bit UNsigned) into Long Register
 instruct loadUS2L(rRegL dst, memory mem)
 %{
@@ -6512,6 +6543,54 @@ instruct loadI(rRegI dst, memory mem)
   ins_pipe(ialu_reg_mem);
 %}
 
+// Load Integer (32 bit signed) to Byte (8 bit signed)
+instruct loadI2B(rRegI dst, memory mem, immI_24 twentyfour) %{
+  match(Set dst (RShiftI (LShiftI (LoadI mem) twentyfour) twentyfour));
+
+  ins_cost(125);
+  format %{ "movsbl  $dst, $mem\t# int -> byte" %}
+  ins_encode %{
+    __ movsbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
+// Load Integer (32 bit signed) to Unsigned Byte (8 bit UNsigned)
+instruct loadI2UB(rRegI dst, memory mem, immI_255 mask) %{
+  match(Set dst (AndI (LoadI mem) mask));
+
+  ins_cost(125);
+  format %{ "movzbl  $dst, $mem\t# int -> ubyte" %}
+  ins_encode %{
+    __ movzbl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
+// Load Integer (32 bit signed) to Short (16 bit signed)
+instruct loadI2S(rRegI dst, memory mem, immI_16 sixteen) %{
+  match(Set dst (RShiftI (LShiftI (LoadI mem) sixteen) sixteen));
+
+  ins_cost(125);
+  format %{ "movswl  $dst, $mem\t# int -> short" %}
+  ins_encode %{
+    __ movswl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
+// Load Integer (32 bit signed) to Unsigned Short/Char (16 bit UNsigned)
+instruct loadI2US(rRegI dst, memory mem, immI_65535 mask) %{
+  match(Set dst (AndI (LoadI mem) mask));
+
+  ins_cost(125);
+  format %{ "movzwl  $dst, $mem\t# int -> ushort/char" %}
+  ins_encode %{
+    __ movzwl($dst$$Register, $mem$$Address);
+  %}
+  ins_pipe(ialu_reg_mem);
+%}
+
 // Load Integer into Long Register
 instruct loadI2L(rRegL dst, memory mem)
 %{
@@ -7656,6 +7735,121 @@ instruct storeL_reversed(memory dst, rRegL src) %{
 %}
 
 
+//---------- Zeros Count Instructions ------------------------------------------
+
+instruct countLeadingZerosI(rRegI dst, rRegI src, rFlagsReg cr) %{
+  predicate(UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosI src));
+  effect(KILL cr);
+
+  format %{ "lzcntl  $dst, $src\t# count leading zeros (int)" %}
+  ins_encode %{
+    __ lzcntl($dst$$Register, $src$$Register);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countLeadingZerosI_bsr(rRegI dst, rRegI src, rFlagsReg cr) %{
+  predicate(!UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosI src));
+  effect(KILL cr);
+
+  format %{ "bsrl    $dst, $src\t# count leading zeros (int)\n\t"
+            "jnz     skip\n\t"
+            "movl    $dst, -1\n"
+      "skip:\n\t"
+            "negl    $dst\n\t"
+            "addl    $dst, 31" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Label skip;
+    __ bsrl(Rdst, Rsrc);
+    __ jccb(Assembler::notZero, skip);
+    __ movl(Rdst, -1);
+    __ bind(skip);
+    __ negl(Rdst);
+    __ addl(Rdst, BitsPerInt - 1);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countLeadingZerosL(rRegI dst, rRegL src, rFlagsReg cr) %{
+  predicate(UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosL src));
+  effect(KILL cr);
+
+  format %{ "lzcntq  $dst, $src\t# count leading zeros (long)" %}
+  ins_encode %{
+    __ lzcntq($dst$$Register, $src$$Register);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countLeadingZerosL_bsr(rRegI dst, rRegL src, rFlagsReg cr) %{
+  predicate(!UseCountLeadingZerosInstruction);
+  match(Set dst (CountLeadingZerosL src));
+  effect(KILL cr);
+
+  format %{ "bsrq    $dst, $src\t# count leading zeros (long)\n\t"
+            "jnz     skip\n\t"
+            "movl    $dst, -1\n"
+      "skip:\n\t"
+            "negl    $dst\n\t"
+            "addl    $dst, 63" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Register Rsrc = $src$$Register;
+    Label skip;
+    __ bsrq(Rdst, Rsrc);
+    __ jccb(Assembler::notZero, skip);
+    __ movl(Rdst, -1);
+    __ bind(skip);
+    __ negl(Rdst);
+    __ addl(Rdst, BitsPerLong - 1);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countTrailingZerosI(rRegI dst, rRegI src, rFlagsReg cr) %{
+  match(Set dst (CountTrailingZerosI src));
+  effect(KILL cr);
+
+  format %{ "bsfl    $dst, $src\t# count trailing zeros (int)\n\t"
+            "jnz     done\n\t"
+            "movl    $dst, 32\n"
+      "done:" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Label done;
+    __ bsfl(Rdst, $src$$Register);
+    __ jccb(Assembler::notZero, done);
+    __ movl(Rdst, BitsPerInt);
+    __ bind(done);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+instruct countTrailingZerosL(rRegI dst, rRegL src, rFlagsReg cr) %{
+  match(Set dst (CountTrailingZerosL src));
+  effect(KILL cr);
+
+  format %{ "bsfq    $dst, $src\t# count trailing zeros (long)\n\t"
+            "jnz     done\n\t"
+            "movl    $dst, 64\n"
+      "done:" %}
+  ins_encode %{
+    Register Rdst = $dst$$Register;
+    Label done;
+    __ bsfq(Rdst, $src$$Register);
+    __ jccb(Assembler::notZero, done);
+    __ movl(Rdst, BitsPerLong);
+    __ bind(done);
+  %}
+  ins_pipe(ialu_reg);
+%}
+
+
 //---------- Population Count Instructions -------------------------------------
 
 instruct popCountI(rRegI dst, rRegI src) %{

src/share/vm/adlc/filebuff.hpp

@@ -73,7 +73,7 @@ class FileBuff {
 
   // This converts a pointer into the buffer to a file offset.  It only works
   // when the pointer is valid (i.e. just obtained from getline()).
-  long getoff(const char* s) { return _bufoff + (s - _buf); }
+  long getoff(const char* s) { return _bufoff + (long)(s - _buf); }
 };
 
 //------------------------------FileBuffRegion---------------------------------

src/share/vm/adlc/output_c.cpp

@@ -1745,6 +1745,7 @@ void ArchDesc::defineExpand(FILE *fp, InstructForm *node) {
       fprintf(fp,"    del_req(i);\n");
       fprintf(fp,"  }\n");
       fprintf(fp,"  _num_opnds = %d;\n", new_num_opnds);
+      assert(new_num_opnds == node->num_unique_opnds(), "what?");
     }
   }
 
@@ -3761,6 +3762,12 @@ bool InstructForm::define_cisc_version(ArchDesc &AD, FILE *fp_cpp) {
     if ( this->captures_bottom_type() ) {
       fprintf(fp_cpp, "  node->_bottom_type = bottom_type();\n");
     }
+
+    uint cur_num_opnds = num_opnds();
+    if (cur_num_opnds > 1 && cur_num_opnds != num_unique_opnds()) {
+      fprintf(fp_cpp,"  node->_num_opnds = %d;\n", num_unique_opnds());
+    }
+
     fprintf(fp_cpp, "\n");
     fprintf(fp_cpp, "  // Copy _idx, inputs and operands to new node\n");
     fprintf(fp_cpp, "  fill_new_machnode(node, C);\n");

src/share/vm/c1/c1_Compilation.cpp

@@ -319,7 +319,7 @@ void Compilation::compile_method() {
     return;
   }
 
-  if (JvmtiExport::can_hotswap_or_post_breakpoint()) {
+  if (_env->jvmti_can_hotswap_or_post_breakpoint()) {
     // We can assert evol_method because method->can_be_compiled is true.
     dependency_recorder()->assert_evol_method(method());
   }
@@ -435,7 +435,7 @@ Compilation::Compilation(AbstractCompiler* compiler, ciEnv* env, ciMethod* metho
   assert(_arena == NULL, "shouldn't only one instance of Compilation in existence at a time");
   _arena = Thread::current()->resource_area();
   _compilation = this;
-  _needs_debug_information = JvmtiExport::can_examine_or_deopt_anywhere() ||
+  _needs_debug_information = _env->jvmti_can_examine_or_deopt_anywhere() ||
                                JavaMonitorsInStackTrace || AlwaysEmitDebugInfo || DeoptimizeALot;
   _exception_info_list = new ExceptionInfoList();
   _implicit_exception_table.set_size(0);

src/share/vm/c1/c1_GraphBuilder.cpp

@@ -1662,7 +1662,7 @@ void GraphBuilder::invoke(Bytecodes::Code code) {
         // Register dependence if JVMTI has either breakpoint
         // setting or hotswapping of methods capabilities since they may
         // cause deoptimization.
-        if (JvmtiExport::can_hotswap_or_post_breakpoint()) {
+        if (compilation()->env()->jvmti_can_hotswap_or_post_breakpoint()) {
           dependency_recorder()->assert_evol_method(inline_target);
         }
         return;
@@ -2863,7 +2863,7 @@ GraphBuilder::GraphBuilder(Compilation* compilation, IRScope* scope)
   start_block->merge(_initial_state);
 
   BlockBegin* sync_handler = NULL;
-  if (method()->is_synchronized() || DTraceMethodProbes) {
+  if (method()->is_synchronized() || _compilation->env()->dtrace_method_probes()) {
     // setup an exception handler to do the unlocking and/or notification
     sync_handler = new BlockBegin(-1);
     sync_handler->set(BlockBegin::exception_entry_flag);

src/share/vm/c1/c1_LIRGenerator.cpp

@@ -1064,7 +1064,7 @@ void LIRGenerator::do_IfInstanceOf(IfInstanceOf* x) {
 
 
 void LIRGenerator::do_Return(Return* x) {
-  if (DTraceMethodProbes) {
+  if (compilation()->env()->dtrace_method_probes()) {
     BasicTypeList signature;
     signature.append(T_INT);    // thread
     signature.append(T_OBJECT); // methodOop
@@ -1769,7 +1769,7 @@ void LIRGenerator::do_Throw(Throw* x) {
     __ null_check(exception_opr, new CodeEmitInfo(info, true));
   }
 
-  if (JvmtiExport::can_post_exceptions() &&
+  if (compilation()->env()->jvmti_can_post_exceptions() &&
       !block()->is_set(BlockBegin::default_exception_handler_flag)) {
     // we need to go through the exception lookup path to get JVMTI
     // notification done
@@ -1779,7 +1779,7 @@ void LIRGenerator::do_Throw(Throw* x) {
   assert(!block()->is_set(BlockBegin::default_exception_handler_flag) || unwind,
          "should be no more handlers to dispatch to");
 
-  if (DTraceMethodProbes &&
+  if (compilation()->env()->dtrace_method_probes() &&
       block()->is_set(BlockBegin::default_exception_handler_flag)) {
     // notify that this frame is unwinding
     BasicTypeList signature;
@@ -2204,7 +2204,7 @@ void LIRGenerator::do_Base(Base* x) {
     java_index += type2size[t];
   }
 
-  if (DTraceMethodProbes) {
+  if (compilation()->env()->dtrace_method_probes()) {
     BasicTypeList signature;
     signature.append(T_INT);    // thread
     signature.append(T_OBJECT); // methodOop

src/share/vm/ci/ciEnv.cpp

@@ -170,6 +170,34 @@ ciEnv::~ciEnv() {
   current_thread->set_env(NULL);
 }
 
+// ------------------------------------------------------------------
+// Cache Jvmti state
+void ciEnv::cache_jvmti_state() {
+  VM_ENTRY_MARK;
+  // Get Jvmti capabilities under lock to get consistant values.
+  MutexLocker mu(JvmtiThreadState_lock);
+  _jvmti_can_hotswap_or_post_breakpoint = JvmtiExport::can_hotswap_or_post_breakpoint();
+  _jvmti_can_examine_or_deopt_anywhere  = JvmtiExport::can_examine_or_deopt_anywhere();
+  _jvmti_can_access_local_variables     = JvmtiExport::can_access_local_variables();
+  _jvmti_can_post_exceptions            = JvmtiExport::can_post_exceptions();
+}
+
+// ------------------------------------------------------------------
+// Cache DTrace flags
+void ciEnv::cache_dtrace_flags() {
+  // Need lock?
+  _dtrace_extended_probes = ExtendedDTraceProbes;
+  if (_dtrace_extended_probes) {
+    _dtrace_monitor_probes  = true;
+    _dtrace_method_probes   = true;
+    _dtrace_alloc_probes    = true;
+  } else {
+    _dtrace_monitor_probes  = DTraceMonitorProbes;
+    _dtrace_method_probes   = DTraceMethodProbes;
+    _dtrace_alloc_probes    = DTraceAllocProbes;
+  }
+}
+
 // ------------------------------------------------------------------
 // helper for lazy exception creation
 ciInstance* ciEnv::get_or_create_exception(jobject& handle, symbolHandle name) {
@@ -810,16 +838,39 @@ void ciEnv::register_method(ciMethod* target,
     // and invalidating our dependencies until we install this method.
     MutexLocker ml(Compile_lock);
 
-    if (log() != NULL) {
-      // Log the dependencies which this compilation declares.
-      dependencies()->log_all_dependencies();
+    // Change in Jvmti state may invalidate compilation.
+    if (!failing() &&
+        ( (!jvmti_can_hotswap_or_post_breakpoint() &&
+           JvmtiExport::can_hotswap_or_post_breakpoint()) ||
+          (!jvmti_can_examine_or_deopt_anywhere() &&
+           JvmtiExport::can_examine_or_deopt_anywhere()) ||
+          (!jvmti_can_access_local_variables() &&
+           JvmtiExport::can_access_local_variables()) ||
+          (!jvmti_can_post_exceptions() &&
+           JvmtiExport::can_post_exceptions()) )) {
+      record_failure("Jvmti state change invalidated dependencies");
     }
 
-    // Encode the dependencies now, so we can check them right away.
-    dependencies()->encode_content_bytes();
+    // Change in DTrace flags may invalidate compilation.
+    if (!failing() &&
+        ( (!dtrace_extended_probes() && ExtendedDTraceProbes) ||
+          (!dtrace_method_probes() && DTraceMethodProbes) ||
+          (!dtrace_alloc_probes() && DTraceAllocProbes) )) {
+      record_failure("DTrace flags change invalidated dependencies");
+    }
 
-    // Check for {class loads, evolution, breakpoints} during compilation
-    check_for_system_dictionary_modification(target);
+    if (!failing()) {
+      if (log() != NULL) {
+        // Log the dependencies which this compilation declares.
+        dependencies()->log_all_dependencies();
+      }
+
+      // Encode the dependencies now, so we can check them right away.
+      dependencies()->encode_content_bytes();
+
+      // Check for {class loads, evolution, breakpoints} during compilation
+      check_for_system_dictionary_modification(target);
+    }
 
     methodHandle method(THREAD, target->get_methodOop());
 

src/share/vm/ci/ciEnv.hpp

@@ -53,6 +53,18 @@ private:
   char* _name_buffer;
   int   _name_buffer_len;
 
+  // Cache Jvmti state
+  bool  _jvmti_can_hotswap_or_post_breakpoint;
+  bool  _jvmti_can_examine_or_deopt_anywhere;
+  bool  _jvmti_can_access_local_variables;
+  bool  _jvmti_can_post_exceptions;
+
+  // Cache DTrace flags
+  bool  _dtrace_extended_probes;
+  bool  _dtrace_monitor_probes;
+  bool  _dtrace_method_probes;
+  bool  _dtrace_alloc_probes;
+
   // Distinguished instances of certain ciObjects..
   static ciObject*              _null_object_instance;
   static ciMethodKlass*         _method_klass_instance;
@@ -236,6 +248,20 @@ public:
   bool break_at_compile() { return _break_at_compile; }
   void set_break_at_compile(bool z) { _break_at_compile = z; }
 
+  // Cache Jvmti state
+  void  cache_jvmti_state();
+  bool  jvmti_can_hotswap_or_post_breakpoint() const { return _jvmti_can_hotswap_or_post_breakpoint; }
+  bool  jvmti_can_examine_or_deopt_anywhere()  const { return _jvmti_can_examine_or_deopt_anywhere; }
+  bool  jvmti_can_access_local_variables()     const { return _jvmti_can_access_local_variables; }
+  bool  jvmti_can_post_exceptions()            const { return _jvmti_can_post_exceptions; }
+
+  // Cache DTrace flags
+  void  cache_dtrace_flags();
+  bool  dtrace_extended_probes() const { return _dtrace_extended_probes; }
+  bool  dtrace_monitor_probes()  const { return _dtrace_monitor_probes; }
+  bool  dtrace_method_probes()   const { return _dtrace_method_probes; }
+  bool  dtrace_alloc_probes()    const { return _dtrace_alloc_probes; }
+
   // The compiler task which has created this env.
   // May be useful to find out compile_id, comp_level, etc.
   CompileTask* task() { return _task; }

src/share/vm/ci/ciMethod.cpp

@@ -60,7 +60,8 @@ ciMethod::ciMethod(methodHandle h_m) : ciObject(h_m) {
   _flow               = NULL;
 #endif // COMPILER2
 
-  if (JvmtiExport::can_hotswap_or_post_breakpoint() && _is_compilable) {
+  ciEnv *env = CURRENT_ENV;
+  if (env->jvmti_can_hotswap_or_post_breakpoint() && _is_compilable) {
     // 6328518 check hotswap conditions under the right lock.
     MutexLocker locker(Compile_lock);
     if (Dependencies::check_evol_method(h_m()) != NULL) {
@@ -84,7 +85,6 @@ ciMethod::ciMethod(methodHandle h_m) : ciObject(h_m) {
   if (_can_be_statically_bound && h_m()->is_abstract())
     _can_be_statically_bound = false;
 
-  ciEnv *env = CURRENT_ENV;
   // generating _signature may allow GC and therefore move m.
   // These fields are always filled in.
   _name = env->get_object(h_m()->name())->as_symbol();
@@ -337,7 +337,7 @@ MethodLivenessResult ciMethod::liveness_at_bci(int bci) {
     _liveness->compute_liveness();
   }
   MethodLivenessResult result = _liveness->get_liveness_at(bci);
-  if (JvmtiExport::can_access_local_variables() || DeoptimizeALot || CompileTheWorld) {
+  if (CURRENT_ENV->jvmti_can_access_local_variables() || DeoptimizeALot || CompileTheWorld) {
     // Keep all locals live for the user's edification and amusement.
     result.at_put_range(0, result.size(), true);
   }

src/share/vm/classfile/vmSymbols.hpp

@@ -313,6 +313,8 @@
   template(value_name,                                "value")                                    \
   template(frontCacheEnabled_name,                    "frontCacheEnabled")                        \
   template(stringCacheEnabled_name,                   "stringCacheEnabled")                       \
+  template(numberOfLeadingZeros_name,                 "numberOfLeadingZeros")                     \
+  template(numberOfTrailingZeros_name,                "numberOfTrailingZeros")                    \
   template(bitCount_name,                             "bitCount")                                 \
   template(profile_name,                              "profile")                                  \
   template(equals_name,                               "equals")                                   \
@@ -559,6 +561,12 @@
   do_intrinsic(_longBitsToDouble,         java_lang_Double,       longBitsToDouble_name,    long_double_signature, F_S) \
    do_name(     longBitsToDouble_name,                           "longBitsToDouble")                                    \
                                                                                                                         \
+  do_intrinsic(_numberOfLeadingZeros_i,   java_lang_Integer,      numberOfLeadingZeros_name,int_int_signature,   F_S)   \
+  do_intrinsic(_numberOfLeadingZeros_l,   java_lang_Long,         numberOfLeadingZeros_name,long_int_signature,  F_S)   \
+                                                                                                                        \
+  do_intrinsic(_numberOfTrailingZeros_i,  java_lang_Integer,      numberOfTrailingZeros_name,int_int_signature,  F_S)   \
+  do_intrinsic(_numberOfTrailingZeros_l,  java_lang_Long,         numberOfTrailingZeros_name,long_int_signature, F_S)   \
+                                                                                                                        \
   do_intrinsic(_bitCount_i,               java_lang_Integer,      bitCount_name,            int_int_signature,   F_S)   \
   do_intrinsic(_bitCount_l,               java_lang_Long,         bitCount_name,            long_int_signature,  F_S)   \
                                                                                                                         \

src/share/vm/compiler/compileBroker.cpp

@@ -1530,6 +1530,12 @@ void CompileBroker::invoke_compiler_on_method(CompileTask* task) {
     assert(thread->env() == &ci_env, "set by ci_env");
     // The thread-env() field is cleared in ~CompileTaskWrapper.
 
+    // Cache Jvmti state
+    ci_env.cache_jvmti_state();
+
+    // Cache DTrace flags
+    ci_env.cache_dtrace_flags();
+
     ciMethod* target = ci_env.get_method_from_handle(target_handle);
 
     TraceTime t1("compilation", &time);

src/share/vm/opto/classes.hpp

@@ -104,6 +104,10 @@ macro(ConvL2I)
 macro(CosD)
 macro(CountedLoop)
 macro(CountedLoopEnd)
+macro(CountLeadingZerosI)
+macro(CountLeadingZerosL)
+macro(CountTrailingZerosI)
+macro(CountTrailingZerosL)
 macro(CreateEx)
 macro(DecodeN)
 macro(DivD)

src/share/vm/opto/connode.cpp

@@ -1255,3 +1255,93 @@ const Type *MoveD2LNode::Value( PhaseTransform *phase ) const {
   v.set_jdouble(td->getd());
   return TypeLong::make( v.get_jlong() );
 }
+
+//------------------------------Value------------------------------------------
+const Type* CountLeadingZerosINode::Value(PhaseTransform* phase) const {
+  const Type* t = phase->type(in(1));
+  if (t == Type::TOP) return Type::TOP;
+  const TypeInt* ti = t->isa_int();
+  if (ti && ti->is_con()) {
+    jint i = ti->get_con();
+    // HD, Figure 5-6
+    if (i == 0)
+      return TypeInt::make(BitsPerInt);
+    int n = 1;
+    unsigned int x = i;
+    if (x >> 16 == 0) { n += 16; x <<= 16; }
+    if (x >> 24 == 0) { n +=  8; x <<=  8; }
+    if (x >> 28 == 0) { n +=  4; x <<=  4; }
+    if (x >> 30 == 0) { n +=  2; x <<=  2; }
+    n -= x >> 31;
+    return TypeInt::make(n);
+  }
+  return TypeInt::INT;
+}
+
+//------------------------------Value------------------------------------------
+const Type* CountLeadingZerosLNode::Value(PhaseTransform* phase) const {
+  const Type* t = phase->type(in(1));
+  if (t == Type::TOP) return Type::TOP;
+  const TypeLong* tl = t->isa_long();
+  if (tl && tl->is_con()) {
+    jlong l = tl->get_con();
+    // HD, Figure 5-6
+    if (l == 0)
+      return TypeInt::make(BitsPerLong);
+    int n = 1;
+    unsigned int x = (((julong) l) >> 32);
+    if (x == 0) { n += 32; x = (int) l; }
+    if (x >> 16 == 0) { n += 16; x <<= 16; }
+    if (x >> 24 == 0) { n +=  8; x <<=  8; }
+    if (x >> 28 == 0) { n +=  4; x <<=  4; }
+    if (x >> 30 == 0) { n +=  2; x <<=  2; }
+    n -= x >> 31;
+    return TypeInt::make(n);
+  }
+  return TypeInt::INT;
+}
+
+//------------------------------Value------------------------------------------
+const Type* CountTrailingZerosINode::Value(PhaseTransform* phase) const {
+  const Type* t = phase->type(in(1));
+  if (t == Type::TOP) return Type::TOP;
+  const TypeInt* ti = t->isa_int();
+  if (ti && ti->is_con()) {
+    jint i = ti->get_con();
+    // HD, Figure 5-14
+    int y;
+    if (i == 0)
+      return TypeInt::make(BitsPerInt);
+    int n = 31;
+    y = i << 16; if (y != 0) { n = n - 16; i = y; }
+    y = i <<  8; if (y != 0) { n = n -  8; i = y; }
+    y = i <<  4; if (y != 0) { n = n -  4; i = y; }
+    y = i <<  2; if (y != 0) { n = n -  2; i = y; }
+    y = i <<  1; if (y != 0) { n = n -  1; }
+    return TypeInt::make(n);
+  }
+  return TypeInt::INT;
+}
+
+//------------------------------Value------------------------------------------
+const Type* CountTrailingZerosLNode::Value(PhaseTransform* phase) const {
+  const Type* t = phase->type(in(1));
+  if (t == Type::TOP) return Type::TOP;
+  const TypeLong* tl = t->isa_long();
+  if (tl && tl->is_con()) {
+    jlong l = tl->get_con();
+    // HD, Figure 5-14
+    int x, y;
+    if (l == 0)
+      return TypeInt::make(BitsPerLong);
+    int n = 63;
+    y = (int) l; if (y != 0) { n = n - 32; x = y; } else x = (((julong) l) >> 32);
+    y = x << 16; if (y != 0) { n = n - 16; x = y; }
+    y = x <<  8; if (y != 0) { n = n -  8; x = y; }
+    y = x <<  4; if (y != 0) { n = n -  4; x = y; }
+    y = x <<  2; if (y != 0) { n = n -  2; x = y; }
+    y = x <<  1; if (y != 0) { n = n -  1; }
+    return TypeInt::make(n);
+  }
+  return TypeInt::INT;
+}

src/share/vm/opto/connode.hpp

@@ -636,22 +636,62 @@ class MoveD2LNode : public Node {
   virtual const Type* Value( PhaseTransform *phase ) const;
 };
 
+//---------- CountBitsNode -----------------------------------------------------
+class CountBitsNode : public Node {
+public:
+  CountBitsNode(Node* in1) : Node(0, in1) {}
+  const Type* bottom_type() const { return TypeInt::INT; }
+  virtual uint ideal_reg() const { return Op_RegI; }
+};
+
+//---------- CountLeadingZerosINode --------------------------------------------
+// Count leading zeros (0-bit count starting from MSB) of an integer.
+class CountLeadingZerosINode : public CountBitsNode {
+public:
+  CountLeadingZerosINode(Node* in1) : CountBitsNode(in1) {}
+  virtual int Opcode() const;
+  virtual const Type* Value(PhaseTransform* phase) const;
+};
+
+//---------- CountLeadingZerosLNode --------------------------------------------
+// Count leading zeros (0-bit count starting from MSB) of a long.
+class CountLeadingZerosLNode : public CountBitsNode {
+public:
+  CountLeadingZerosLNode(Node* in1) : CountBitsNode(in1) {}
+  virtual int Opcode() const;
+  virtual const Type* Value(PhaseTransform* phase) const;
+};
+
+//---------- CountTrailingZerosINode -------------------------------------------
+// Count trailing zeros (0-bit count starting from LSB) of an integer.
+class CountTrailingZerosINode : public CountBitsNode {
+public:
+  CountTrailingZerosINode(Node* in1) : CountBitsNode(in1) {}
+  virtual int Opcode() const;
+  virtual const Type* Value(PhaseTransform* phase) const;
+};
+
+//---------- CountTrailingZerosLNode -------------------------------------------
+// Count trailing zeros (0-bit count starting from LSB) of a long.
+class CountTrailingZerosLNode : public CountBitsNode {
+public:
+  CountTrailingZerosLNode(Node* in1) : CountBitsNode(in1) {}
+  virtual int Opcode() const;
+  virtual const Type* Value(PhaseTransform* phase) const;
+};
+
 //---------- PopCountINode -----------------------------------------------------
 // Population count (bit count) of an integer.
-class PopCountINode : public Node {
+class PopCountINode : public CountBitsNode {
 public:
-  PopCountINode(Node* in1) : Node(0, in1) {}
+  PopCountINode(Node* in1) : CountBitsNode(in1) {}
   virtual int Opcode() const;
-  const Type* bottom_type() const { return TypeInt::INT; }
-  virtual uint ideal_reg() const { return Op_RegI; }
 };
 
 //---------- PopCountLNode -----------------------------------------------------
 // Population count (bit count) of a long.
-class PopCountLNode : public Node {
+class PopCountLNode : public CountBitsNode {
 public:
-  PopCountLNode(Node* in1) : Node(0, in1) {}
+  PopCountLNode(Node* in1) : CountBitsNode(in1) {}
   virtual int Opcode() const;
-  const Type* bottom_type() const { return TypeInt::INT; }
-  virtual uint ideal_reg() const { return Op_RegI; }
 };

src/share/vm/opto/doCall.cpp

@@ -47,7 +47,7 @@ CallGenerator* Compile::call_generator(ciMethod* call_method, int vtable_index,
   CallGenerator* cg;
 
   // Dtrace currently doesn't work unless all calls are vanilla
-  if (DTraceMethodProbes) {
+  if (env()->dtrace_method_probes()) {
     allow_inline = false;
   }
 

src/share/vm/opto/escape.cpp

@@ -905,15 +905,22 @@ void ConnectionGraph::split_unique_types(GrowableArray<Node *>  &alloc_worklist)
       // see if it is unescaped.
       if (es != PointsToNode::NoEscape || !ptn->_scalar_replaceable)
         continue;
-      if (alloc->is_Allocate()) {
-        // Set the scalar_replaceable flag before the next check.
-        alloc->as_Allocate()->_is_scalar_replaceable = true;
-      }
-      // find CheckCastPP of call return value
+
+      // Find CheckCastPP for the allocate or for the return value of a call
       n = alloc->result_cast();
-      if (n == NULL ||          // No uses accept Initialize or
-          !n->is_CheckCastPP()) // not unique CheckCastPP.
+      if (n == NULL) {            // No uses except Initialize node
+        if (alloc->is_Allocate()) {
+          // Set the scalar_replaceable flag for allocation
+          // so it could be eliminated if it has no uses.
+          alloc->as_Allocate()->_is_scalar_replaceable = true;
+        }
+        continue;
+      }
+      if (!n->is_CheckCastPP()) { // not unique CheckCastPP.
+        assert(!alloc->is_Allocate(), "allocation should have unique type");
         continue;
+      }
+
       // The inline code for Object.clone() casts the allocation result to
       // java.lang.Object and then to the actual type of the allocated
       // object. Detect this case and use the second cast.
@@ -934,9 +941,17 @@ void ConnectionGraph::split_unique_types(GrowableArray<Node *>  &alloc_worklist)
         if (cast2 != NULL) {
           n = cast2;
         } else {
+          // Non-scalar replaceable if the allocation type is unknown statically
+          // (reflection allocation), the object can't be restored during
+          // deoptimization without precise type.
           continue;
         }
       }
+      if (alloc->is_Allocate()) {
+        // Set the scalar_replaceable flag for allocation
+        // so it could be eliminated.
+        alloc->as_Allocate()->_is_scalar_replaceable = true;
+      }
       set_escape_state(n->_idx, es);
       // in order for an object to be scalar-replaceable, it must be:
       //   - a direct allocation (not a call returning an object)

src/share/vm/opto/gcm.cpp

@@ -617,6 +617,9 @@ Block* PhaseCFG::insert_anti_dependences(Block* LCA, Node* load, bool verify) {
             assert(!LCA_orig->dominates(pred_block) ||
                    early->dominates(pred_block), "early is high enough");
             must_raise_LCA = true;
+          } else {
+            // anti-dependent upon PHI pinned below 'early', no edge needed
+            LCA = early;             // but can not schedule below 'early'
           }
         }
       }

src/share/vm/opto/graphKit.cpp

@@ -459,7 +459,7 @@ Bytecodes::Code GraphKit::java_bc() const {
 void GraphKit::builtin_throw(Deoptimization::DeoptReason reason, Node* arg) {
   bool must_throw = true;
 
-  if (JvmtiExport::can_post_exceptions()) {
+  if (env()->jvmti_can_post_exceptions()) {
     // Do not try anything fancy if we're notifying the VM on every throw.
     // Cf. case Bytecodes::_athrow in parse2.cpp.
     uncommon_trap(reason, Deoptimization::Action_none,
@@ -769,7 +769,7 @@ void GraphKit::add_safepoint_edges(SafePointNode* call, bool must_throw) {
     }
   }
 
-  if (JvmtiExport::can_examine_or_deopt_anywhere()) {
+  if (env()->jvmti_can_examine_or_deopt_anywhere()) {
     // At any safepoint, this method can get breakpointed, which would
     // then require an immediate deoptimization.
     full_info = true;

src/share/vm/opto/library_call.cpp

@@ -222,6 +222,8 @@ class LibraryCallKit : public GraphKit {
   bool inline_unsafe_CAS(BasicType type);
   bool inline_unsafe_ordered_store(BasicType type);
   bool inline_fp_conversions(vmIntrinsics::ID id);
+  bool inline_numberOfLeadingZeros(vmIntrinsics::ID id);
+  bool inline_numberOfTrailingZeros(vmIntrinsics::ID id);
   bool inline_bitCount(vmIntrinsics::ID id);
   bool inline_reverseBytes(vmIntrinsics::ID id);
 };
@@ -630,6 +632,14 @@ bool LibraryCallKit::try_to_inline() {
   case vmIntrinsics::_longBitsToDouble:
     return inline_fp_conversions(intrinsic_id());
 
+  case vmIntrinsics::_numberOfLeadingZeros_i:
+  case vmIntrinsics::_numberOfLeadingZeros_l:
+    return inline_numberOfLeadingZeros(intrinsic_id());
+
+  case vmIntrinsics::_numberOfTrailingZeros_i:
+  case vmIntrinsics::_numberOfTrailingZeros_l:
+    return inline_numberOfTrailingZeros(intrinsic_id());
+
   case vmIntrinsics::_bitCount_i:
   case vmIntrinsics::_bitCount_l:
     return inline_bitCount(intrinsic_id());
@@ -1844,6 +1854,48 @@ inline Node* LibraryCallKit::make_unsafe_address(Node* base, Node* offset) {
   }
 }
 
+//-------------------inline_numberOfLeadingZeros_int/long-----------------------
+// inline int Integer.numberOfLeadingZeros(int)
+// inline int Long.numberOfLeadingZeros(long)
+bool LibraryCallKit::inline_numberOfLeadingZeros(vmIntrinsics::ID id) {
+  assert(id == vmIntrinsics::_numberOfLeadingZeros_i || id == vmIntrinsics::_numberOfLeadingZeros_l, "not numberOfLeadingZeros");
+  if (id == vmIntrinsics::_numberOfLeadingZeros_i && !Matcher::match_rule_supported(Op_CountLeadingZerosI)) return false;
+  if (id == vmIntrinsics::_numberOfLeadingZeros_l && !Matcher::match_rule_supported(Op_CountLeadingZerosL)) return false;
+  _sp += arg_size();  // restore stack pointer
+  switch (id) {
+  case vmIntrinsics::_numberOfLeadingZeros_i:
+    push(_gvn.transform(new (C, 2) CountLeadingZerosINode(pop())));
+    break;
+  case vmIntrinsics::_numberOfLeadingZeros_l:
+    push(_gvn.transform(new (C, 2) CountLeadingZerosLNode(pop_pair())));
+    break;
+  default:
+    ShouldNotReachHere();
+  }
+  return true;
+}
+
+//-------------------inline_numberOfTrailingZeros_int/long----------------------
+// inline int Integer.numberOfTrailingZeros(int)
+// inline int Long.numberOfTrailingZeros(long)
+bool LibraryCallKit::inline_numberOfTrailingZeros(vmIntrinsics::ID id) {
+  assert(id == vmIntrinsics::_numberOfTrailingZeros_i || id == vmIntrinsics::_numberOfTrailingZeros_l, "not numberOfTrailingZeros");
+  if (id == vmIntrinsics::_numberOfTrailingZeros_i && !Matcher::match_rule_supported(Op_CountTrailingZerosI)) return false;
+  if (id == vmIntrinsics::_numberOfTrailingZeros_l && !Matcher::match_rule_supported(Op_CountTrailingZerosL)) return false;
+  _sp += arg_size();  // restore stack pointer
+  switch (id) {
+  case vmIntrinsics::_numberOfTrailingZeros_i:
+    push(_gvn.transform(new (C, 2) CountTrailingZerosINode(pop())));
+    break;
+  case vmIntrinsics::_numberOfTrailingZeros_l:
+    push(_gvn.transform(new (C, 2) CountTrailingZerosLNode(pop_pair())));
+    break;
+  default:
+    ShouldNotReachHere();
+  }
+  return true;
+}
+
 //----------------------------inline_bitCount_int/long-----------------------
 // inline int Integer.bitCount(int)
 // inline int Long.bitCount(long)
@@ -2541,7 +2593,8 @@ bool LibraryCallKit::inline_native_isInterrupted() {
   Node* p = basic_plus_adr(top()/*!oop*/, tls_ptr, in_bytes(JavaThread::osthread_offset()));
   Node* osthread = make_load(NULL, p, TypeRawPtr::NOTNULL, T_ADDRESS);
   p = basic_plus_adr(top()/*!oop*/, osthread, in_bytes(OSThread::interrupted_offset()));
-  Node* int_bit = make_load(NULL, p, TypeInt::BOOL, T_INT);
+  // Set the control input on the field _interrupted read to prevent it floating up.
+  Node* int_bit = make_load(control(), p, TypeInt::BOOL, T_INT);
   Node* cmp_bit = _gvn.transform( new (C, 3) CmpINode(int_bit, intcon(0)) );
   Node* bol_bit = _gvn.transform( new (C, 2) BoolNode(cmp_bit, BoolTest::ne) );
 

src/share/vm/opto/loopTransform.cpp

@@ -1630,6 +1630,10 @@ bool IdealLoopTree::iteration_split_impl( PhaseIdealLoop *phase, Node_List &old_
   // Before attempting fancy unrolling, RCE or alignment, see if we want
   // to completely unroll this loop or do loop unswitching.
   if( cl->is_normal_loop() ) {
+    if (should_unswitch) {
+      phase->do_unswitching(this, old_new);
+      return true;
+    }
     bool should_maximally_unroll =  policy_maximally_unroll(phase);
     if( should_maximally_unroll ) {
       // Here we did some unrolling and peeling.  Eventually we will
@@ -1637,10 +1641,6 @@ bool IdealLoopTree::iteration_split_impl( PhaseIdealLoop *phase, Node_List &old_
       phase->do_maximally_unroll(this,old_new);
       return true;
     }
-    if (should_unswitch) {
-      phase->do_unswitching(this, old_new);
-      return true;
-    }
   }
 
 

src/share/vm/opto/macro.cpp

@@ -988,7 +988,7 @@ void PhaseMacroExpand::expand_allocate_common(
     initial_slow_test = BoolNode::make_predicate(initial_slow_test, &_igvn);
   }
 
-  if (DTraceAllocProbes ||
+  if (C->env()->dtrace_alloc_probes() ||
       !UseTLAB && (!Universe::heap()->supports_inline_contig_alloc() ||
                    (UseConcMarkSweepGC && CMSIncrementalMode))) {
     // Force slow-path allocation
@@ -1150,7 +1150,7 @@ void PhaseMacroExpand::expand_allocate_common(
                                         fast_oop_ctrl, fast_oop_rawmem, fast_oop,
                                         klass_node, length, size_in_bytes);
 
-    if (ExtendedDTraceProbes) {
+    if (C->env()->dtrace_extended_probes()) {
       // Slow-path call
       int size = TypeFunc::Parms + 2;
       CallLeafNode *call = new (C, size) CallLeafNode(OptoRuntime::dtrace_object_alloc_Type(),

src/share/vm/opto/matcher.hpp

@@ -225,10 +225,16 @@ public:
   OptoRegPair *_parm_regs;        // Array of machine registers per argument
   RegMask *_calling_convention_mask; // Array of RegMasks per argument
 
-  // Does matcher support this ideal node?
+  // Does matcher have a match rule for this ideal node?
   static const bool has_match_rule(int opcode);
   static const bool _hasMatchRule[_last_opcode];
 
+  // Does matcher have a match rule for this ideal node and is the
+  // predicate (if there is one) true?
+  // NOTE: If this function is used more commonly in the future, ADLC
+  // should generate this one.
+  static const bool match_rule_supported(int opcode);
+
   // Used to determine if we have fast l2f conversion
   // USII has it, USIII doesn't
   static const bool convL2FSupported(void);

src/share/vm/opto/parse1.cpp

@@ -439,7 +439,7 @@ Parse::Parse(JVMState* caller, ciMethod* parse_method, float expected_uses)
   // Always register dependence if JVMTI is enabled, because
   // either breakpoint setting or hotswapping of methods may
   // cause deoptimization.
-  if (JvmtiExport::can_hotswap_or_post_breakpoint()) {
+  if (C->env()->jvmti_can_hotswap_or_post_breakpoint()) {
     C->dependencies()->assert_evol_method(method());
   }
 
@@ -953,7 +953,7 @@ void Parse::do_exits() {
   bool do_synch = method()->is_synchronized() && GenerateSynchronizationCode;
 
   // record exit from a method if compiled while Dtrace is turned on.
-  if (do_synch || DTraceMethodProbes) {
+  if (do_synch || C->env()->dtrace_method_probes()) {
     // First move the exception list out of _exits:
     GraphKit kit(_exits.transfer_exceptions_into_jvms());
     SafePointNode* normal_map = kit.map();  // keep this guy safe
@@ -975,7 +975,7 @@ void Parse::do_exits() {
         // Unlock!
         kit.shared_unlock(_synch_lock->box_node(), _synch_lock->obj_node());
       }
-      if (DTraceMethodProbes) {
+      if (C->env()->dtrace_method_probes()) {
         kit.make_dtrace_method_exit(method());
       }
       // Done with exception-path processing.
@@ -1074,7 +1074,7 @@ void Parse::do_method_entry() {
 
   NOT_PRODUCT( count_compiled_calls(true/*at_method_entry*/, false/*is_inline*/); )
 
-  if (DTraceMethodProbes) {
+  if (C->env()->dtrace_method_probes()) {
     make_dtrace_method_entry(method());
   }
 
@@ -1960,7 +1960,7 @@ void Parse::return_current(Node* value) {
   if (method()->is_synchronized() && GenerateSynchronizationCode) {
     shared_unlock(_synch_lock->box_node(), _synch_lock->obj_node());
   }
-  if (DTraceMethodProbes) {
+  if (C->env()->dtrace_method_probes()) {
     make_dtrace_method_exit(method());
   }
   SafePointNode* exit_return = _exits.map();

src/share/vm/opto/parse2.cpp

@@ -2052,7 +2052,7 @@ void Parse::do_one_bytecode() {
     // null exception oop throws NULL pointer exception
     do_null_check(peek(), T_OBJECT);
     if (stopped())  return;
-    if (JvmtiExport::can_post_exceptions()) {
+    if (env()->jvmti_can_post_exceptions()) {
       // "Full-speed throwing" is not necessary here,
       // since we're notifying the VM on every throw.
       uncommon_trap(Deoptimization::Reason_unhandled,

src/share/vm/opto/subnode.cpp

@@ -640,7 +640,11 @@ const Type *CmpPNode::sub( const Type *t1, const Type *t2 ) const {
     if (klass0 && klass1 &&
         kps != 1 &&             // both or neither are klass pointers
         klass0->is_loaded() && !klass0->is_interface() && // do not trust interfaces
-        klass1->is_loaded() && !klass1->is_interface()) {
+        klass1->is_loaded() && !klass1->is_interface() &&
+        (!klass0->is_obj_array_klass() ||
+         !klass0->as_obj_array_klass()->base_element_klass()->is_interface()) &&
+        (!klass1->is_obj_array_klass() ||
+         !klass1->as_obj_array_klass()->base_element_klass()->is_interface())) {
       bool unrelated_classes = false;
       // See if neither subclasses the other, or if the class on top
       // is precise.  In either of these cases, the compare is known

src/share/vm/runtime/globals.hpp

@@ -2189,6 +2189,9 @@ class CommandLineFlags {
   diagnostic(bool, PrintIntrinsics, false,                                  \
           "prints attempted and successful inlining of intrinsics")         \
                                                                             \
+  product(bool, UseCountLeadingZerosInstruction, false,                     \
+          "Use count leading zeros instruction")                            \
+                                                                            \
   product(bool, UsePopCountInstruction, false,                              \
           "Use population count instruction")                               \
                                                                             \

test/Makefile

@@ -172,6 +172,33 @@ PHONY_LIST += jtreg_tests
 
 ################################################################
 
+# clienttest (make sure various basic java client options work)
+
+clienttest: prep $(PRODUCT_HOME)
+	$(PRODUCT_HOME)/bin/java $(JAVA_OPTIONS) -version
+	$(PRODUCT_HOME)/bin/java $(JAVA_OPTIONS) -help
+	$(PRODUCT_HOME)/bin/java $(JAVA_OPTIONS) -X
+	$(RM) $(PRODUCT_HOME)/jre/lib/*/client/classes.jsa
+	$(RM) $(PRODUCT_HOME)/jre/lib/*/client/classes_g.jsa
+	$(RM) $(PRODUCT_HOME)/jre/bin/client/classes.jsa
+	$(RM) $(PRODUCT_HOME)/jre/bin/client/classes_g.jsa
+	$(PRODUCT_HOME)/bin/java $(JAVA_OPTIONS) -Xshare:dump
+
+PHONY_LIST += clienttest
+
+################################################################
+
+# servertest (make sure various basic java server options work)
+
+servertest: prep $(PRODUCT_HOME)
+	$(PRODUCT_HOME)/bin/java $(JAVA_OPTIONS) -version
+	$(PRODUCT_HOME)/bin/java $(JAVA_OPTIONS) -help
+	$(PRODUCT_HOME)/bin/java $(JAVA_OPTIONS) -X
+
+PHONY_LIST += servertest
+
+################################################################
+
 # packtest
 
 # Expect JPRT to set JPRT_PACKTEST_HOME.

test/compiler/6636138/Test1.java

@@ -29,7 +29,7 @@
  * @run main/othervm -server -Xbatch -XX:CompileOnly=Test1.init Test1
  */
 
-class Test1 {
+public class Test1 {
 
     public static void init(int src[], int [] dst, int[] ref) {
         // initialize the arrays

test/compiler/6636138/Test2.java

@@ -29,7 +29,7 @@
  * @run main/othervm -server -Xbatch -XX:CompileOnly=Test2.shift Test2
  */
 
-class Test2 {
+public class Test2 {
 
     public static void init(int src[]) {
         // Initialize the array

test/compiler/6772683/InterruptedTest.java

+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All Rights Reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This code is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ *
+ */
+
+/*
+ * @test
+ * @bug 6772683
+ * @summary Thread.isInterrupted() fails to return true on multiprocessor PC
+ * @run main/othervm InterruptedTest
+ */
+
+public class InterruptedTest {
+
+    public static void main(String[] args) throws Exception {
+        Thread workerThread = new Thread("worker") {
+            public void run() {
+                System.out.println("Worker thread: running...");
+                while (!Thread.currentThread().isInterrupted()) {
+                }
+                System.out.println("Worker thread: bye");
+            }
+        };
+        System.out.println("Main thread: starts a worker thread...");
+        workerThread.start();
+        System.out.println("Main thread: waits at most 5s for the worker thread to die...");
+        workerThread.join(5000); // Wait 5 sec to let run() method to be compiled
+        int ntries = 0;
+        while (workerThread.isAlive() && ntries < 5) {
+            System.out.println("Main thread: interrupts the worker thread...");
+            workerThread.interrupt();
+            if (workerThread.isInterrupted()) {
+                System.out.println("Main thread: worker thread is interrupted");
+            }
+            ntries++;
+            System.out.println("Main thread: waits for the worker thread to die...");
+            workerThread.join(1000); // Wait 1 sec and try again
+        }
+        if (ntries == 5) {
+          System.out.println("Main thread: the worker thread dod not die");
+          System.exit(97);
+        }
+        System.out.println("Main thread: bye");
+    }
+
+}

test/compiler/6814842/Test6814842.java

+/*
+ * Copyright 2009 Sun Microsystems, Inc.  All Rights Reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This code is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ */
+
+/**
+ * @test
+ * @bug 6814842
+ * @summary Load shortening optimizations
+ *
+ * @run main/othervm -Xcomp -XX:CompileOnly=Test6814842.loadS2B,Test6814842.loadS2Bmask255,Test6814842.loadUS2B,Test6814842.loadUS2Bmask255,Test6814842.loadI2B,Test6814842.loadI2Bmask255,Test6814842.loadI2S,Test6814842.loadI2Smask255,Test6814842.loadI2Smask65535,Test6814842.loadI2US,Test6814842.loadI2USmask255,Test6814842.loadI2USmask65535 Test6814842
+ */
+
+public class Test6814842 {
+    static final short[] sa = new short[] { (short) 0xF1F2 };
+    static final char[]  ca = new char[]  { (char) 0xF3F4  };
+    static final int[]   ia = new int[]   { 0xF1F2F3F4     };
+
+    public static void main(String[] args)
+    {
+        byte s2b = loadS2B(sa);
+        if (s2b != (byte) 0xF2)
+            throw new InternalError("loadS2B failed: " + s2b + " != " + (byte) 0xF2);
+
+        byte s2bmask255 = loadS2Bmask255(sa);
+        if (s2bmask255 != (byte) 0xF2)
+            throw new InternalError("loadS2Bmask255 failed: " + s2bmask255 + " != " + (byte) 0xF2);
+
+        byte us2b = loadUS2B(ca);
+        if (us2b != (byte) 0xF4)
+            throw new InternalError("loadUS2B failed: " + us2b + " != " + (byte) 0xF4);
+
+        byte us2bmask255 = loadUS2Bmask255(ca);
+        if (us2bmask255 != (byte) 0xF4)
+            throw new InternalError("loadUS2Bmask255 failed: " + us2bmask255 + " != " + (byte) 0xF4);
+
+        byte i2b = loadI2B(ia);
+        if (i2b != (byte) 0xF4)
+            throw new InternalError("loadI2B failed: " + i2b + " != " + (byte) 0xF4);
+
+        byte i2bmask255 = loadI2Bmask255(ia);
+        if (i2bmask255 != (byte) 0xF4)
+            throw new InternalError("loadI2Bmask255 failed: " + i2bmask255 + " != " + (byte) 0xF4);
+
+        short i2s = loadI2S(ia);
+        if (i2s != (short) 0xF3F4)
+            throw new InternalError("loadI2S failed: " + i2s + " != " + (short) 0xF3F4);
+
+        short i2smask255 = loadI2Smask255(ia);
+        if (i2smask255 != (short) 0xF4)
+            throw new InternalError("loadI2Smask255 failed: " + i2smask255 + " != " + (short) 0xF4);
+
+        short i2smask65535 = loadI2Smask65535(ia);
+        if (i2smask65535 != (short) 0xF3F4)
+            throw new InternalError("loadI2Smask65535 failed: " + i2smask65535 + " != " + (short) 0xF3F4);
+
+        char i2us = loadI2US(ia);
+        if (i2us != (char) 0xF3F4)
+            throw new InternalError("loadI2US failed: " + (int) i2us + " != " + (char) 0xF3F4);
+
+        char i2usmask255 = loadI2USmask255(ia);
+        if (i2usmask255 != (char) 0xF4)
+            throw new InternalError("loadI2USmask255 failed: " + (int) i2usmask255 + " != " + (char) 0xF4);
+
+        char i2usmask65535 = loadI2USmask65535(ia);
+        if (i2usmask65535 != (char) 0xF3F4)
+            throw new InternalError("loadI2USmask65535 failed: " + (int) i2usmask65535 + " != " + (char) 0xF3F4);
+    }
+
+    static byte  loadS2B          (short[] sa) { return (byte)  (sa[0]         ); }
+    static byte  loadS2Bmask255   (short[] sa) { return (byte)  (sa[0] & 0xFF  ); }
+
+    static byte  loadUS2B         (char[]  ca) { return (byte)  (ca[0]         ); }
+    static byte  loadUS2Bmask255  (char[]  ca) { return (byte)  (ca[0] & 0xFF  ); }
+
+    static byte  loadI2B          (int[]   ia) { return (byte)  (ia[0]         ); }
+    static byte  loadI2Bmask255   (int[]   ia) { return (byte)  (ia[0] & 0xFF  ); }
+
+    static short loadI2S          (int[]   ia) { return (short) (ia[0]         ); }
+    static short loadI2Smask255   (int[]   ia) { return (short) (ia[0] & 0xFF  ); }
+    static short loadI2Smask65535 (int[]   ia) { return (short) (ia[0] & 0xFFFF); }
+
+    static char  loadI2US         (int[]   ia) { return (char)  (ia[0]         ); }
+    static char  loadI2USmask255  (int[]   ia) { return (char)  (ia[0] & 0xFF  ); }
+    static char  loadI2USmask65535(int[]   ia) { return (char)  (ia[0] & 0xFFFF); }
+}

test/compiler/6823354/Test6823354.java

+/*
+ * Copyright 2009 Sun Microsystems, Inc.  All Rights Reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This code is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ */
+
+/**
+ * @test
+ * @bug 6823354
+ * @summary These methods can be instrinsified by using bit scan, bit test, and population count instructions.
+ *
+ * @run main/othervm -Xcomp -XX:CompileOnly=Test6823354.lzcomp,Test6823354.tzcomp,.dolzcomp,.dotzcomp Test6823354
+ */
+
+import java.net.URLClassLoader;
+
+public class Test6823354 {
+    // Arrays of corner case values.
+    static final int[]  ia = new int[]  { 0,  1,  -1,  Integer.MIN_VALUE, Integer.MAX_VALUE };
+    static final long[] la = new long[] { 0L, 1L, -1L, Long.MIN_VALUE,    Long.MAX_VALUE    };
+
+    public static void main(String[] args) throws Exception {
+        // Load the classes and the methods.
+        Integer.numberOfLeadingZeros(0);
+        Integer.numberOfTrailingZeros(0);
+        Long.numberOfLeadingZeros(0);
+        Long.numberOfTrailingZeros(0);
+
+        lz();
+        tz();
+    }
+
+    static void lz() throws Exception {
+        // int
+
+        // Test corner cases.
+        for (int i = 0; i < ia.length; i++) {
+            int x = ia[i];
+            check(x, lzcomp(x), lzint(x));
+        }
+
+        // Test all possible return values.
+        for (int i = 0; i < Integer.SIZE; i++) {
+            int x = 1 << i;
+            check(x, lzcomp(x), lzint(x));
+        }
+
+        String classname = "Test6823354$lzconI";
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < ia.length; i++) {
+            testclass(classname, ia[i]);
+        }
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < Integer.SIZE; i++) {
+            int x = 1 << i;
+            testclass(classname, x);
+        }
+
+
+        // long
+
+        // Test corner cases.
+        for (int i = 0; i < ia.length; i++) {
+            long x = la[i];
+            check(x, lzcomp(x), lzint(x));
+        }
+
+        // Test all possible return values.
+        for (int i = 0; i < Long.SIZE; i++) {
+            long x = 1L << i;
+            check(x, lzcomp(x), lzint(x));
+        }
+
+        classname = "Test6823354$lzconL";
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < la.length; i++) {
+            testclass(classname, la[i]);
+        }
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < Long.SIZE; i++) {
+            long x = 1L << i;
+            testclass(classname, x);
+        }
+    }
+
+    static void tz() throws Exception {
+        // int
+
+        // Test corner cases.
+        for (int i = 0; i < ia.length; i++) {
+            int x = ia[i];
+            check(x, tzcomp(x), tzint(x));
+        }
+
+        // Test all possible return values.
+        for (int i = 0; i < Integer.SIZE; i++) {
+            int x = 1 << i;
+            check(x, tzcomp(x), tzint(x));
+        }
+
+        String classname = "Test6823354$tzconI";
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < ia.length; i++) {
+            testclass(classname, ia[i]);
+        }
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < Integer.SIZE; i++) {
+            int x = 1 << i;
+            testclass(classname, x);
+        }
+
+
+        // long
+
+        // Test corner cases.
+        for (int i = 0; i < la.length; i++) {
+            long x = la[i];
+            check(x, tzcomp(x), tzint(x));
+        }
+
+        // Test all possible return values.
+        for (int i = 0; i < Long.SIZE; i++) {
+            long x = 1L << i;
+            check(x, tzcomp(x), tzint(x));
+        }
+
+        classname = "Test6823354$tzconL";
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < la.length; i++) {
+            testclass(classname, la[i]);
+        }
+
+        // Test Ideal optimizations (constant values).
+        for (int i = 0; i < Long.SIZE; i++) {
+            long x = 1L << i;
+            testclass(classname, x);
+        }
+    }
+
+    static void check(int value, int result, int expected) {
+        //System.out.println(value + ": " + result + ", " + expected);
+        if (result != expected)
+            throw new InternalError(value + " failed: " + result + " != " + expected);
+    }
+
+    static void check(long value, long result, long expected) {
+        //System.out.println(value + ": " + result + ", " + expected);
+        if (result != expected)
+            throw new InternalError(value + " failed: " + result + " != " + expected);
+    }
+
+    static int lzint( int i)  { return Integer.numberOfLeadingZeros(i); }
+    static int lzcomp(int i)  { return Integer.numberOfLeadingZeros(i); }
+
+    static int lzint( long l) { return Long.numberOfLeadingZeros(l); }
+    static int lzcomp(long l) { return Long.numberOfLeadingZeros(l); }
+
+    static int tzint( int i)  { return Integer.numberOfTrailingZeros(i); }
+    static int tzcomp(int i)  { return Integer.numberOfTrailingZeros(i); }
+
+    static int tzint( long l) { return Long.numberOfTrailingZeros(l); }
+    static int tzcomp(long l) { return Long.numberOfTrailingZeros(l); }
+
+    static void testclass(String classname, int x) throws Exception {
+        System.setProperty("value", "" + x);
+        loadandrunclass(classname);
+    }
+
+    static void testclass(String classname, long x) throws Exception {
+        System.setProperty("value", "" + x);
+        loadandrunclass(classname);
+    }
+
+    static void loadandrunclass(String classname) throws Exception {
+        Class cl = Class.forName(classname);
+        URLClassLoader apploader = (URLClassLoader) cl.getClassLoader();
+        ClassLoader loader = new URLClassLoader(apploader.getURLs(), apploader.getParent());
+        Class c = loader.loadClass(classname);
+        Runnable r = (Runnable) c.newInstance();
+        r.run();
+    }
+
+    public static class lzconI implements Runnable {
+        static final int VALUE;
+
+        static {
+            int value = 0;
+            try {
+                value = Integer.decode(System.getProperty("value"));
+            } catch (Throwable e) {}
+            VALUE = value;
+        }
+
+        public void run() { check(VALUE, lzint(VALUE), dolzcomp()); }
+        static int dolzcomp() { return lzcomp(VALUE); }
+    }
+
+    public static class lzconL implements Runnable {
+        static final long VALUE;
+
+        static {
+            long value = 0;
+            try {
+                value = Long.decode(System.getProperty("value"));
+            } catch (Throwable e) {}
+            VALUE = value;
+        }
+
+        public void run() { check(VALUE, lzint(VALUE), dolzcomp()); }
+        static int dolzcomp() { return lzcomp(VALUE); }
+    }
+
+    public static class tzconI implements Runnable {
+        static final int VALUE;
+
+        static {
+            int value = 0;
+            try {
+                value = Integer.decode(System.getProperty("value"));
+            } catch (Throwable e) {}
+            VALUE = value;
+        }
+
+        public void run() { check(VALUE, tzint(VALUE), dotzcomp()); }
+        static int dotzcomp() { return tzcomp(VALUE); }
+    }
+
+    public static class tzconL implements Runnable {
+        static final long VALUE;
+
+        static {
+            long value = 0;
+            try {
+                value = Long.decode(System.getProperty("value"));
+            } catch (Throwable e) {}
+            VALUE = value;
+        }
+
+        public void run() { check(VALUE, tzint(VALUE), dotzcomp()); }
+        static int dotzcomp() { return tzcomp(VALUE); }
+    }
+}

test/compiler/6832293/Test.java

+/*
+ * Copyright 2008 Sun Microsystems, Inc.  All Rights Reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This code is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ *
+ */
+
+/*
+ * @test
+ * @bug 6832293
+ * @summary JIT compiler got wrong result in type checking with -server
+ * @run main/othervm  -Xcomp -XX:CompileOnly=Test.run Test
+ */
+
+import java.io.PrintStream;
+
+interface SomeInterface {
+    int SEVENS = 777;
+}
+
+interface AnotherInterface {
+    int THIRDS = 33;
+}
+
+class SomeClass implements SomeInterface {
+    int i;
+
+    SomeClass(int i) {
+        this.i = i;
+    }
+}
+
+class ImmediateSubclass extends SomeClass implements SomeInterface {
+    float f;
+
+    ImmediateSubclass(int i, float f) {
+        super(i);
+        this.f = f;
+    }
+}
+
+final class FinalSubclass extends ImmediateSubclass implements AnotherInterface {
+    double d;
+
+     FinalSubclass(int i, float f, double d) {
+        super(i, f);
+        this.d = d;
+    }
+}
+
+public class Test {
+
+    public static void main(String args[]) throws Exception{
+        /* try to pre initialize */
+        SomeClass[] a=new SomeClass[10];
+        Class.forName("ImmediateSubclass");
+        Class.forName("FinalSubclass");
+        System.exit(run(args, System.out) + 95/*STATUS_TEMP*/);
+    }
+
+    static int errorStatus = 0/*STATUS_PASSED*/;
+
+    static void errorAlert(PrintStream out, int errorLevel) {
+        out.println("Test: failure #" + errorLevel);
+        errorStatus = 2/*STATUS_FAILED*/;
+    }
+
+    static SomeClass[] v2 = new FinalSubclass[4];
+
+    public static int run(String args[],PrintStream out) {
+        int i [], j [];
+        SomeInterface u [], v[] [];
+        AnotherInterface w [];
+        SomeClass x [] [];
+
+        i = new int [10];
+        i[0] = 777;
+        j = (int []) i;
+        if (j != i)
+            errorAlert(out, 2);
+        else if (j.length != 10)
+            errorAlert(out, 3);
+        else if (j[0] != 777)
+            errorAlert(out, 4);
+
+        v = new SomeClass [3] [];
+        x = (SomeClass [] []) v;
+        if (! (x instanceof SomeInterface [] []))
+            errorAlert(out, 5);
+        else if (! (x instanceof SomeClass [] []))
+            errorAlert(out, 6);
+        else if (x != v)
+            errorAlert(out, 7);
+
+        x[0] = (SomeClass []) new ImmediateSubclass [4];
+        if (! (x[0] instanceof ImmediateSubclass []))
+            errorAlert(out, 8);
+        else if (x[0].length != 4)
+            errorAlert(out, 9);
+
+        x[1] = (SomeClass []) v2;
+        if (! (x[1] instanceof FinalSubclass []))
+            errorAlert(out, 10);
+        else if (x[1].length != 4)
+            errorAlert(out, 11);
+
+        w = (AnotherInterface []) x[1];
+        if (! (w instanceof FinalSubclass []))
+            errorAlert(out, 12);
+        else if (w != x[1])
+            errorAlert(out, 13);
+        else if (w.length != 4)
+            errorAlert(out, 14);
+
+        return errorStatus;
+    }
+}
+

test/compiler/6843752/Test.java

+/*
+ * Copyright 2009 Sun Microsystems, Inc.  All Rights Reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This code is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
+ * CA 95054 USA or visit www.sun.com if you need additional information or
+ * have any questions.
+ */
+
+/*
+ * @test
+ * @bug 6843752
+ * @summary missing code for an anti-dependent Phi in GCM
+ * @run main/othervm -Xbatch Test
+ */
+
+public class Test {
+
+    Item list;
+
+    static class Item {
+        public Item    next;
+        public Item    prev;
+        public boolean remove;
+
+        Item(boolean r) { remove = r; }
+    }
+
+    private void linkIn(Item item) {
+        Item head = list;
+        if (head == null) {
+            item.next = item;
+            item.prev = item;
+            list = item;
+        } else {
+            item.next = head;
+            item.prev = head.prev;
+            head.prev.next = item;
+            head.prev = item;
+        }
+    }
+
+    private void linkOut(Item item) {
+        Item head = list;
+        if (item.next == item) {
+            list = null;
+        } else {
+            item.prev.next = item.next;
+            item.next.prev = item.prev;
+            if (head == item) {
+                list = item.next;
+            }
+        }
+        item.next = null;
+        item.prev = null; // this is the null pointer we are seeing
+    }
+
+    private void removeItems(int numItems) {
+        Item item = list;
+        if (item == null) {
+            return;
+        }
+        Item last = item.prev;
+        boolean done = false;
+        while (!done && numItems > 1) {
+            // the original code "done = (item == last);" triggered an infinite loop
+            // and was changed slightly in order to produce an exception instead.
+            done = (item.next == last.next);
+            item = item.next;
+            if (item.prev.remove) {
+                linkOut(item.prev);
+            }
+        }
+    }
+
+    public void perform(int numItems) {
+        for (int i = 0; i < numItems; i++) {
+            linkIn(new Item(i == 0));
+        }
+        removeItems(numItems);
+        list = null;
+    }
+
+    static public void main(String[] args) {
+        int caseCnt = 0;
+        Test bj = new Test();
+        try {
+            for (; caseCnt < 500000;) {
+                int numItems = (++caseCnt % 2);
+                if ((caseCnt % 64) == 0) {
+                    numItems = 5;
+                }
+                bj.perform(numItems);
+                if ((caseCnt % 100000) == 0) {
+                    System.out.println("successfully performed " + caseCnt + " cases");
+                }
+            }
+        } catch (Exception e) {
+            System.out.println("ERROR: crashed during case " + caseCnt);
+            e.printStackTrace(System.out);
+            System.exit(97);
+        }
+    }
+}
+