Merge tag jb8u242-b1644.3HEADmastermain

Change-Id: I63388d7cef8a6b2d2fd1c64ee4e612e514e21409

71 files changed, 1536 insertions, 125 deletions

diff --git a/.hgtags b/.hgtags
index a933d7c01..31c3b58f0 100644
--- a/.hgtags
+++ b/.hgtags
@@ -1317,3 +1317,12 @@ c751303497d539aa85c6373aa0fa85580d3f3044 jdk8u232-b07
 4170228e11e6313e948e6ddcae9af3eed06b1fbe jdk8u232-b08
 12177d88b89c12c14daa5ad681030d7551e8a5a0 jdk8u232-b09
 12177d88b89c12c14daa5ad681030d7551e8a5a0 jdk8u232-ga
+ce42ae95d4d671f74246091f89bd101d5bcbfd91 jdk8u242-b01
+775e2bf92114e41365cc6baf1480c818454256a4 jdk8u242-b02
+ee19c358e3b8deeda2f64d660a0870df7b1abd49 jdk8u242-b03
+20258ba5a788da55485c0648bcc073f8ad2c26ef jdk8u242-b04
+2c1e9fab6964647f4eeffe55fe5592da6399a3ce jdk8u242-b05
+81ddc1072b923330f84c0ace3124226f63877582 jdk8u242-b06
+8b80409d5840142a27e274d33948f483a6406a50 jdk8u242-b07
+7c9f6b5f8d119dc1ba3c5536595ce3ae7414599d jdk8u242-b08
+7c9f6b5f8d119dc1ba3c5536595ce3ae7414599d jdk8u242-ga
diff --git a/THIRD_PARTY_README b/THIRD_PARTY_README
index 7dc54a057..643ea79ce 100644
--- a/THIRD_PARTY_README
+++ b/THIRD_PARTY_README
@@ -1334,11 +1334,13 @@ SUCH DAMAGE.
 
 --------------------------------------------------------------------------------
 
-%% This notice is provided with respect to Joni v1.1.9, which may be 
+%% This notice is provided with respect to Joni v2.1.16, which may be
 included with JRE 8, JDK 8, and OpenJDK 8.
 
 --- begin of LICENSE ---
 
+Copyright (c) 2017 JRuby Team
+
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal
 in the Software without restriction, including without limitation the rights
diff --git a/src/cpu/ppc/vm/vm_version_ppc.cpp b/src/cpu/ppc/vm/vm_version_ppc.cpp
index bbc52691b..3c59c96ca 100644
--- a/src/cpu/ppc/vm/vm_version_ppc.cpp
+++ b/src/cpu/ppc/vm/vm_version_ppc.cpp
@@ -194,6 +194,11 @@ void VM_Version::initialize() {
     FLAG_SET_DEFAULT(UseAESIntrinsics, false);
   }
 
+  if (UseGHASHIntrinsics) {
+    warning("GHASH intrinsics are not available on this CPU");
+    FLAG_SET_DEFAULT(UseGHASHIntrinsics, false);
+  }
+
   if (has_vshasig()) {
     if (FLAG_IS_DEFAULT(UseSHA)) {
       UseSHA = true;
diff --git a/src/cpu/sparc/vm/assembler_sparc.hpp b/src/cpu/sparc/vm/assembler_sparc.hpp
index dd83b092f..55f338754 100644
--- a/src/cpu/sparc/vm/assembler_sparc.hpp
+++ b/src/cpu/sparc/vm/assembler_sparc.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 1997, 2014, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 1997, 2015, Oracle and/or its affiliates. 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
@@ -129,6 +129,7 @@ class Assembler : public AbstractAssembler  {
     flog3_op3    = 0x36,
     edge_op3     = 0x36,
     fsrc_op3     = 0x36,
+    xmulx_op3    = 0x36,
     impdep2_op3  = 0x37,
     stpartialf_op3 = 0x37,
     jmpl_op3     = 0x38,
@@ -220,6 +221,8 @@ class Assembler : public AbstractAssembler  {
     mdtox_opf          = 0x110,
     mstouw_opf         = 0x111,
     mstosw_opf         = 0x113,
+    xmulx_opf          = 0x115,
+    xmulxhi_opf        = 0x116,
     mxtod_opf          = 0x118,
     mwtos_opf          = 0x119,
 
@@ -1212,6 +1215,9 @@ public:
   void movwtos( Register s, FloatRegister d ) { vis3_only();  emit_int32( op(arith_op) | fd(d, FloatRegisterImpl::S) | op3(mftoi_op3) | opf(mwtos_opf) | rs2(s)); }
   void movxtod( Register s, FloatRegister d ) { vis3_only();  emit_int32( op(arith_op) | fd(d, FloatRegisterImpl::D) | op3(mftoi_op3) | opf(mxtod_opf) | rs2(s)); }
 
+  void xmulx(Register s1, Register s2, Register d) { vis3_only(); emit_int32( op(arith_op) | rd(d) | op3(xmulx_op3) | rs1(s1) | opf(xmulx_opf) | rs2(s2)); }
+  void xmulxhi(Register s1, Register s2, Register d) { vis3_only(); emit_int32( op(arith_op) | rd(d) | op3(xmulx_op3) | rs1(s1) | opf(xmulxhi_opf) | rs2(s2)); }
+
   // Crypto SHA instructions
 
   void sha1()   { sha1_only();    emit_int32( op(arith_op) | op3(sha_op3) | opf(sha1_opf)); }
diff --git a/src/cpu/sparc/vm/stubGenerator_sparc.cpp b/src/cpu/sparc/vm/stubGenerator_sparc.cpp
index 6ad0b1a41..51064927e 100644
--- a/src/cpu/sparc/vm/stubGenerator_sparc.cpp
+++ b/src/cpu/sparc/vm/stubGenerator_sparc.cpp
@@ -4788,6 +4788,130 @@ class StubGenerator: public StubCodeGenerator {
     return start;
   }
 
+  /* Single and multi-block ghash operations */
+  address generate_ghash_processBlocks() {
+      __ align(CodeEntryAlignment);
+      Label L_ghash_loop, L_aligned, L_main;
+      StubCodeMark mark(this, "StubRoutines", "ghash_processBlocks");
+      address start = __ pc();
+
+      Register state = I0;
+      Register subkeyH = I1;
+      Register data = I2;
+      Register len = I3;
+
+      __ save_frame(0);
+
+      __ ldx(state, 0, O0);
+      __ ldx(state, 8, O1);
+
+      // Loop label for multiblock operations
+      __ BIND(L_ghash_loop);
+
+      // Check if 'data' is unaligned
+      __ andcc(data, 7, G1);
+      __ br(Assembler::zero, false, Assembler::pt, L_aligned);
+      __ delayed()->nop();
+
+      Register left_shift = L1;
+      Register right_shift = L2;
+      Register data_ptr = L3;
+
+      // Get left and right shift values in bits
+      __ sll(G1, LogBitsPerByte, left_shift);
+      __ mov(64, right_shift);
+      __ sub(right_shift, left_shift, right_shift);
+
+      // Align to read 'data'
+      __ sub(data, G1, data_ptr);
+
+      // Load first 8 bytes of 'data'
+      __ ldx(data_ptr, 0, O4);
+      __ sllx(O4, left_shift, O4);
+      __ ldx(data_ptr, 8, O5);
+      __ srlx(O5, right_shift, G4);
+      __ bset(G4, O4);
+
+      // Load second 8 bytes of 'data'
+      __ sllx(O5, left_shift, O5);
+      __ ldx(data_ptr, 16, G4);
+      __ srlx(G4, right_shift, G4);
+      __ ba(L_main);
+      __ delayed()->bset(G4, O5);
+
+      // If 'data' is aligned, load normally
+      __ BIND(L_aligned);
+      __ ldx(data, 0, O4);
+      __ ldx(data, 8, O5);
+
+      __ BIND(L_main);
+      __ ldx(subkeyH, 0, O2);
+      __ ldx(subkeyH, 8, O3);
+
+      __ xor3(O0, O4, O0);
+      __ xor3(O1, O5, O1);
+
+      __ xmulxhi(O0, O3, G3);
+      __ xmulx(O0, O2, O5);
+      __ xmulxhi(O1, O2, G4);
+      __ xmulxhi(O1, O3, G5);
+      __ xmulx(O0, O3, G1);
+      __ xmulx(O1, O3, G2);
+      __ xmulx(O1, O2, O3);
+      __ xmulxhi(O0, O2, O4);
+
+      __ mov(0xE1, O0);
+      __ sllx(O0, 56, O0);
+
+      __ xor3(O5, G3, O5);
+      __ xor3(O5, G4, O5);
+      __ xor3(G5, G1, G1);
+      __ xor3(G1, O3, G1);
+      __ srlx(G2, 63, O1);
+      __ srlx(G1, 63, G3);
+      __ sllx(G2, 63, O3);
+      __ sllx(G2, 58, O2);
+      __ xor3(O3, O2, O2);
+
+      __ sllx(G1, 1, G1);
+      __ or3(G1, O1, G1);
+
+      __ xor3(G1, O2, G1);
+
+      __ sllx(G2, 1, G2);
+
+      __ xmulxhi(G1, O0, O1);
+      __ xmulx(G1, O0, O2);
+      __ xmulxhi(G2, O0, O3);
+      __ xmulx(G2, O0, G1);
+
+      __ xor3(O4, O1, O4);
+      __ xor3(O5, O2, O5);
+      __ xor3(O5, O3, O5);
+
+      __ sllx(O4, 1, O2);
+      __ srlx(O5, 63, O3);
+
+      __ or3(O2, O3, O0);
+
+      __ sllx(O5, 1, O1);
+      __ srlx(G1, 63, O2);
+      __ or3(O1, O2, O1);
+      __ xor3(O1, G3, O1);
+
+      __ deccc(len);
+      __ br(Assembler::notZero, true, Assembler::pt, L_ghash_loop);
+      __ delayed()->add(data, 16, data);
+
+      __ stx(O0, I0, 0);
+      __ stx(O1, I0, 8);
+
+      __ ret();
+      __ delayed()->restore();
+
+      return start;
+  }
+
   void generate_initial() {
     // Generates all stubs and initializes the entry points
 
@@ -4860,6 +4984,10 @@ class StubGenerator: public StubCodeGenerator {
       StubRoutines::_cipherBlockChaining_encryptAESCrypt = generate_cipherBlockChaining_encryptAESCrypt();
       StubRoutines::_cipherBlockChaining_decryptAESCrypt = generate_cipherBlockChaining_decryptAESCrypt_Parallel();
     }
+    // generate GHASH intrinsics code
+    if (UseGHASHIntrinsics) {
+      StubRoutines::_ghash_processBlocks = generate_ghash_processBlocks();
+    }
 
     // generate SHA1/SHA256/SHA512 intrinsics code
     if (UseSHA1Intrinsics) {
diff --git a/src/cpu/sparc/vm/vm_version_sparc.cpp b/src/cpu/sparc/vm/vm_version_sparc.cpp
index c0cd16a18..d7240e876 100644
--- a/src/cpu/sparc/vm/vm_version_sparc.cpp
+++ b/src/cpu/sparc/vm/vm_version_sparc.cpp
@@ -286,39 +286,50 @@ void VM_Version::initialize() {
 
   // SPARC T4 and above should have support for AES instructions
   if (has_aes()) {
-    if (UseVIS > 2) { // AES intrinsics use MOVxTOd/MOVdTOx which are VIS3
-      if (FLAG_IS_DEFAULT(UseAES)) {
-        FLAG_SET_DEFAULT(UseAES, true);
-      }
-      if (FLAG_IS_DEFAULT(UseAESIntrinsics)) {
-        FLAG_SET_DEFAULT(UseAESIntrinsics, true);
-      }
-      // we disable both the AES flags if either of them is disabled on the command line
-      if (!UseAES || !UseAESIntrinsics) {
-        FLAG_SET_DEFAULT(UseAES, false);
-        FLAG_SET_DEFAULT(UseAESIntrinsics, false);
+    if (FLAG_IS_DEFAULT(UseAES)) {
+      FLAG_SET_DEFAULT(UseAES, true);
+    }
+    if (!UseAES) {
+      if (UseAESIntrinsics && !FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+        warning("AES intrinsics require UseAES flag to be enabled. Intrinsics will be disabled.");
       }
+      FLAG_SET_DEFAULT(UseAESIntrinsics, false);
     } else {
-        if (UseAES || UseAESIntrinsics) {
-          warning("SPARC AES intrinsics require VIS3 instruction support. Intrinsics will be disabled.");
-          if (UseAES) {
-            FLAG_SET_DEFAULT(UseAES, false);
-          }
-          if (UseAESIntrinsics) {
-            FLAG_SET_DEFAULT(UseAESIntrinsics, false);
-          }
+      // The AES intrinsic stubs require AES instruction support (of course)
+      // but also require VIS3 mode or higher for instructions it use.
+      if (UseVIS > 2) {
+        if (FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+          FLAG_SET_DEFAULT(UseAESIntrinsics, true);
         }
+      } else {
+        if (UseAESIntrinsics && !FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+          warning("SPARC AES intrinsics require VIS3 instructions. Intrinsics will be disabled.");
+        }
+        FLAG_SET_DEFAULT(UseAESIntrinsics, false);
+      }
     }
   } else if (UseAES || UseAESIntrinsics) {
-    warning("AES instructions are not available on this CPU");
-    if (UseAES) {
+    if (UseAES && !FLAG_IS_DEFAULT(UseAES)) {
+      warning("AES instructions are not available on this CPU");
       FLAG_SET_DEFAULT(UseAES, false);
     }
-    if (UseAESIntrinsics) {
+    if (UseAESIntrinsics && !FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+      warning("AES intrinsics are not available on this CPU");
       FLAG_SET_DEFAULT(UseAESIntrinsics, false);
     }
   }
 
+  // GHASH/GCM intrinsics
+  if (has_vis3() && (UseVIS > 2)) {
+    if (FLAG_IS_DEFAULT(UseGHASHIntrinsics)) {
+      UseGHASHIntrinsics = true;
+    }
+  } else if (UseGHASHIntrinsics) {
+    if (!FLAG_IS_DEFAULT(UseGHASHIntrinsics))
+      warning("GHASH intrinsics require VIS3 insructions support. Intriniscs will be disabled");
+    FLAG_SET_DEFAULT(UseGHASHIntrinsics, false);
+  }
+
   // SHA1, SHA256, and SHA512 instructions were added to SPARC T-series at different times
   if (has_sha1() || has_sha256() || has_sha512()) {
     if (UseVIS > 0) { // SHA intrinsics use VIS1 instructions
diff --git a/src/cpu/x86/vm/assembler_x86.cpp b/src/cpu/x86/vm/assembler_x86.cpp
index 7cbc47d60..1759ecdfd 100644
--- a/src/cpu/x86/vm/assembler_x86.cpp
+++ b/src/cpu/x86/vm/assembler_x86.cpp
@@ -2575,6 +2575,15 @@ void Assembler::psrldq(XMMRegister dst, int shift) {
   emit_int8(shift);
 }
 
+void Assembler::pslldq(XMMRegister dst, int shift) {
+  // Shift left 128 bit value in xmm register by number of bytes.
+  NOT_LP64(assert(VM_Version::supports_sse2(), ""));
+  int encode = simd_prefix_and_encode(xmm7, dst, dst, VEX_SIMD_66);
+  emit_int8(0x73);
+  emit_int8((unsigned char)(0xC0 | encode));
+  emit_int8(shift);
+}
+
 void Assembler::ptest(XMMRegister dst, Address src) {
   assert(VM_Version::supports_sse4_1(), "");
   assert((UseAVX > 0), "SSE mode requires address alignment 16 bytes");
diff --git a/src/cpu/x86/vm/assembler_x86.hpp b/src/cpu/x86/vm/assembler_x86.hpp
index 341d9e39b..5ea01311e 100644
--- a/src/cpu/x86/vm/assembler_x86.hpp
+++ b/src/cpu/x86/vm/assembler_x86.hpp
@@ -1527,6 +1527,8 @@ private:
 
   // Shift Right by bytes Logical DoubleQuadword Immediate
   void psrldq(XMMRegister dst, int shift);
+  // Shift Left by bytes Logical DoubleQuadword Immediate
+  void pslldq(XMMRegister dst, int shift);
 
   // Logical Compare 128bit
   void ptest(XMMRegister dst, XMMRegister src);
diff --git a/src/cpu/x86/vm/stubGenerator_x86_32.cpp b/src/cpu/x86/vm/stubGenerator_x86_32.cpp
index 50a06d7a5..2e5599807 100644
--- a/src/cpu/x86/vm/stubGenerator_x86_32.cpp
+++ b/src/cpu/x86/vm/stubGenerator_x86_32.cpp
@@ -2719,6 +2719,169 @@ class StubGenerator: public StubCodeGenerator {
     return start;
   }
 
+  // byte swap x86 long
+  address generate_ghash_long_swap_mask() {
+    __ align(CodeEntryAlignment);
+    StubCodeMark mark(this, "StubRoutines", "ghash_long_swap_mask");
+    address start = __ pc();
+    __ emit_data(0x0b0a0908, relocInfo::none, 0);
+    __ emit_data(0x0f0e0d0c, relocInfo::none, 0);
+    __ emit_data(0x03020100, relocInfo::none, 0);
+    __ emit_data(0x07060504, relocInfo::none, 0);
+
+  return start;
+  }
+
+  // byte swap x86 byte array
+  address generate_ghash_byte_swap_mask() {
+    __ align(CodeEntryAlignment);
+    StubCodeMark mark(this, "StubRoutines", "ghash_byte_swap_mask");
+    address start = __ pc();
+    __ emit_data(0x0c0d0e0f, relocInfo::none, 0);
+    __ emit_data(0x08090a0b, relocInfo::none, 0);
+    __ emit_data(0x04050607, relocInfo::none, 0);
+    __ emit_data(0x00010203, relocInfo::none, 0);
+  return start;
+  }
+
+  /* Single and multi-block ghash operations */
+  address generate_ghash_processBlocks() {
+    assert(UseGHASHIntrinsics, "need GHASH intrinsics and CLMUL support");
+    __ align(CodeEntryAlignment);
+    Label L_ghash_loop, L_exit;
+    StubCodeMark mark(this, "StubRoutines", "ghash_processBlocks");
+    address start = __ pc();
+
+    const Register state        = rdi;
+    const Register subkeyH      = rsi;
+    const Register data         = rdx;
+    const Register blocks       = rcx;
+
+    const Address  state_param(rbp, 8+0);
+    const Address  subkeyH_param(rbp, 8+4);
+    const Address  data_param(rbp, 8+8);
+    const Address  blocks_param(rbp, 8+12);
+
+    const XMMRegister xmm_temp0 = xmm0;
+    const XMMRegister xmm_temp1 = xmm1;
+    const XMMRegister xmm_temp2 = xmm2;
+    const XMMRegister xmm_temp3 = xmm3;
+    const XMMRegister xmm_temp4 = xmm4;
+    const XMMRegister xmm_temp5 = xmm5;
+    const XMMRegister xmm_temp6 = xmm6;
+    const XMMRegister xmm_temp7 = xmm7;
+
+    __ enter();
+    handleSOERegisters(true);  // Save registers
+
+    __ movptr(state, state_param);
+    __ movptr(subkeyH, subkeyH_param);
+    __ movptr(data, data_param);
+    __ movptr(blocks, blocks_param);
+
+    __ movdqu(xmm_temp0, Address(state, 0));
+    __ pshufb(xmm_temp0, ExternalAddress(StubRoutines::x86::ghash_long_swap_mask_addr()));
+
+    __ movdqu(xmm_temp1, Address(subkeyH, 0));
+    __ pshufb(xmm_temp1, ExternalAddress(StubRoutines::x86::ghash_long_swap_mask_addr()));
+
+    __ BIND(L_ghash_loop);
+    __ movdqu(xmm_temp2, Address(data, 0));
+    __ pshufb(xmm_temp2, ExternalAddress(StubRoutines::x86::ghash_byte_swap_mask_addr()));
+
+    __ pxor(xmm_temp0, xmm_temp2);
+
+    //
+    // Multiply with the hash key
+    //
+    __ movdqu(xmm_temp3, xmm_temp0);
+    __ pclmulqdq(xmm_temp3, xmm_temp1, 0);      // xmm3 holds a0*b0
+    __ movdqu(xmm_temp4, xmm_temp0);
+    __ pclmulqdq(xmm_temp4, xmm_temp1, 16);     // xmm4 holds a0*b1
+
+    __ movdqu(xmm_temp5, xmm_temp0);
+    __ pclmulqdq(xmm_temp5, xmm_temp1, 1);      // xmm5 holds a1*b0
+    __ movdqu(xmm_temp6, xmm_temp0);
+    __ pclmulqdq(xmm_temp6, xmm_temp1, 17);     // xmm6 holds a1*b1
+
+    __ pxor(xmm_temp4, xmm_temp5);      // xmm4 holds a0*b1 + a1*b0
+
+    __ movdqu(xmm_temp5, xmm_temp4);    // move the contents of xmm4 to xmm5
+    __ psrldq(xmm_temp4, 8);    // shift by xmm4 64 bits to the right
+    __ pslldq(xmm_temp5, 8);    // shift by xmm5 64 bits to the left
+    __ pxor(xmm_temp3, xmm_temp5);
+    __ pxor(xmm_temp6, xmm_temp4);      // Register pair <xmm6:xmm3> holds the result
+                                        // of the carry-less multiplication of
+                                        // xmm0 by xmm1.
+
+    // We shift the result of the multiplication by one bit position
+    // to the left to cope for the fact that the bits are reversed.
+    __ movdqu(xmm_temp7, xmm_temp3);
+    __ movdqu(xmm_temp4, xmm_temp6);
+    __ pslld (xmm_temp3, 1);
+    __ pslld(xmm_temp6, 1);
+    __ psrld(xmm_temp7, 31);
+    __ psrld(xmm_temp4, 31);
+    __ movdqu(xmm_temp5, xmm_temp7);
+    __ pslldq(xmm_temp4, 4);
+    __ pslldq(xmm_temp7, 4);
+    __ psrldq(xmm_temp5, 12);
+    __ por(xmm_temp3, xmm_temp7);
+    __ por(xmm_temp6, xmm_temp4);
+    __ por(xmm_temp6, xmm_temp5);
+
+    //
+    // First phase of the reduction
+    //
+    // Move xmm3 into xmm4, xmm5, xmm7 in order to perform the shifts
+    // independently.
+    __ movdqu(xmm_temp7, xmm_temp3);
+    __ movdqu(xmm_temp4, xmm_temp3);
+    __ movdqu(xmm_temp5, xmm_temp3);
+    __ pslld(xmm_temp7, 31);    // packed right shift shifting << 31
+    __ pslld(xmm_temp4, 30);    // packed right shift shifting << 30
+    __ pslld(xmm_temp5, 25);    // packed right shift shifting << 25
+    __ pxor(xmm_temp7, xmm_temp4);      // xor the shifted versions
+    __ pxor(xmm_temp7, xmm_temp5);
+    __ movdqu(xmm_temp4, xmm_temp7);
+    __ pslldq(xmm_temp7, 12);
+    __ psrldq(xmm_temp4, 4);
+    __ pxor(xmm_temp3, xmm_temp7);      // first phase of the reduction complete
+
+    //
+    // Second phase of the reduction
+    //
+    // Make 3 copies of xmm3 in xmm2, xmm5, xmm7 for doing these
+    // shift operations.
+    __ movdqu(xmm_temp2, xmm_temp3);
+    __ movdqu(xmm_temp7, xmm_temp3);
+    __ movdqu(xmm_temp5, xmm_temp3);
+    __ psrld(xmm_temp2, 1);     // packed left shifting >> 1
+    __ psrld(xmm_temp7, 2);     // packed left shifting >> 2
+    __ psrld(xmm_temp5, 7);     // packed left shifting >> 7
+    __ pxor(xmm_temp2, xmm_temp7);      // xor the shifted versions
+    __ pxor(xmm_temp2, xmm_temp5);
+    __ pxor(xmm_temp2, xmm_temp4);
+    __ pxor(xmm_temp3, xmm_temp2);
+    __ pxor(xmm_temp6, xmm_temp3);      // the result is in xmm6
+
+    __ decrement(blocks);
+    __ jcc(Assembler::zero, L_exit);
+    __ movdqu(xmm_temp0, xmm_temp6);
+    __ addptr(data, 16);
+    __ jmp(L_ghash_loop);
+
+    __ BIND(L_exit);
+       // Byte swap 16-byte result
+    __ pshufb(xmm_temp6, ExternalAddress(StubRoutines::x86::ghash_long_swap_mask_addr()));
+    __ movdqu(Address(state, 0), xmm_temp6);   // store the result
+
+    handleSOERegisters(false);  // restore registers
+    __ leave();
+    __ ret(0);
+    return start;
+  }
+
   /**
    *  Arguments:
    *
@@ -3018,6 +3181,13 @@ class StubGenerator: public StubCodeGenerator {
       StubRoutines::_cipherBlockChaining_decryptAESCrypt = generate_cipherBlockChaining_decryptAESCrypt();
     }
 
+    // Generate GHASH intrinsics code
+    if (UseGHASHIntrinsics) {
+      StubRoutines::x86::_ghash_long_swap_mask_addr = generate_ghash_long_swap_mask();
+      StubRoutines::x86::_ghash_byte_swap_mask_addr = generate_ghash_byte_swap_mask();
+      StubRoutines::_ghash_processBlocks = generate_ghash_processBlocks();
+    }
+
     // Safefetch stubs.
     generate_safefetch("SafeFetch32", sizeof(int), &StubRoutines::_safefetch32_entry,
                                                    &StubRoutines::_safefetch32_fault_pc,
diff --git a/src/cpu/x86/vm/stubGenerator_x86_64.cpp b/src/cpu/x86/vm/stubGenerator_x86_64.cpp
index 1d38af799..c5811b28b 100644
--- a/src/cpu/x86/vm/stubGenerator_x86_64.cpp
+++ b/src/cpu/x86/vm/stubGenerator_x86_64.cpp
@@ -3639,6 +3639,175 @@ class StubGenerator: public StubCodeGenerator {
     return start;
   }
 
+
+  // byte swap x86 long
+  address generate_ghash_long_swap_mask() {
+    __ align(CodeEntryAlignment);
+    StubCodeMark mark(this, "StubRoutines", "ghash_long_swap_mask");
+    address start = __ pc();
+    __ emit_data64(0x0f0e0d0c0b0a0908, relocInfo::none );
+    __ emit_data64(0x0706050403020100, relocInfo::none );
+  return start;
+  }
+
+  // byte swap x86 byte array
+  address generate_ghash_byte_swap_mask() {
+    __ align(CodeEntryAlignment);
+    StubCodeMark mark(this, "StubRoutines", "ghash_byte_swap_mask");
+    address start = __ pc();
+    __ emit_data64(0x08090a0b0c0d0e0f, relocInfo::none );
+    __ emit_data64(0x0001020304050607, relocInfo::none );
+  return start;
+  }
+
+  /* Single and multi-block ghash operations */
+  address generate_ghash_processBlocks() {
+    __ align(CodeEntryAlignment);
+    Label L_ghash_loop, L_exit;
+    StubCodeMark mark(this, "StubRoutines", "ghash_processBlocks");
+    address start = __ pc();
+
+    const Register state        = c_rarg0;
+    const Register subkeyH      = c_rarg1;
+    const Register data         = c_rarg2;
+    const Register blocks       = c_rarg3;
+
+#ifdef _WIN64
+    const int XMM_REG_LAST  = 10;
+#endif
+
+    const XMMRegister xmm_temp0 = xmm0;
+    const XMMRegister xmm_temp1 = xmm1;
+    const XMMRegister xmm_temp2 = xmm2;
+    const XMMRegister xmm_temp3 = xmm3;
+    const XMMRegister xmm_temp4 = xmm4;
+    const XMMRegister xmm_temp5 = xmm5;
+    const XMMRegister xmm_temp6 = xmm6;
+    const XMMRegister xmm_temp7 = xmm7;
+    const XMMRegister xmm_temp8 = xmm8;
+    const XMMRegister xmm_temp9 = xmm9;
+    const XMMRegister xmm_temp10 = xmm10;
+
+    __ enter();
+
+#ifdef _WIN64
+    // save the xmm registers which must be preserved 6-10
+    __ subptr(rsp, -rsp_after_call_off * wordSize);
+    for (int i = 6; i <= XMM_REG_LAST; i++) {
+      __ movdqu(xmm_save(i), as_XMMRegister(i));
+    }
+#endif
+
+    __ movdqu(xmm_temp10, ExternalAddress(StubRoutines::x86::ghash_long_swap_mask_addr()));
+
+    __ movdqu(xmm_temp0, Address(state, 0));
+    __ pshufb(xmm_temp0, xmm_temp10);
+
+
+    __ BIND(L_ghash_loop);
+    __ movdqu(xmm_temp2, Address(data, 0));
+    __ pshufb(xmm_temp2, ExternalAddress(StubRoutines::x86::ghash_byte_swap_mask_addr()));
+
+    __ movdqu(xmm_temp1, Address(subkeyH, 0));
+    __ pshufb(xmm_temp1, xmm_temp10);
+
+    __ pxor(xmm_temp0, xmm_temp2);
+
+    //
+    // Multiply with the hash key
+    //
+    __ movdqu(xmm_temp3, xmm_temp0);
+    __ pclmulqdq(xmm_temp3, xmm_temp1, 0);      // xmm3 holds a0*b0
+    __ movdqu(xmm_temp4, xmm_temp0);
+    __ pclmulqdq(xmm_temp4, xmm_temp1, 16);     // xmm4 holds a0*b1
+
+    __ movdqu(xmm_temp5, xmm_temp0);
+    __ pclmulqdq(xmm_temp5, xmm_temp1, 1);      // xmm5 holds a1*b0
+    __ movdqu(xmm_temp6, xmm_temp0);
+    __ pclmulqdq(xmm_temp6, xmm_temp1, 17);     // xmm6 holds a1*b1
+
+    __ pxor(xmm_temp4, xmm_temp5);      // xmm4 holds a0*b1 + a1*b0
+
+    __ movdqu(xmm_temp5, xmm_temp4);    // move the contents of xmm4 to xmm5
+    __ psrldq(xmm_temp4, 8);    // shift by xmm4 64 bits to the right
+    __ pslldq(xmm_temp5, 8);    // shift by xmm5 64 bits to the left
+    __ pxor(xmm_temp3, xmm_temp5);
+    __ pxor(xmm_temp6, xmm_temp4);      // Register pair <xmm6:xmm3> holds the result
+                                        // of the carry-less multiplication of
+                                        // xmm0 by xmm1.
+
+    // We shift the result of the multiplication by one bit position
+    // to the left to cope for the fact that the bits are reversed.
+    __ movdqu(xmm_temp7, xmm_temp3);
+    __ movdqu(xmm_temp8, xmm_temp6);
+    __ pslld(xmm_temp3, 1);
+    __ pslld(xmm_temp6, 1);
+    __ psrld(xmm_temp7, 31);
+    __ psrld(xmm_temp8, 31);
+    __ movdqu(xmm_temp9, xmm_temp7);
+    __ pslldq(xmm_temp8, 4);
+    __ pslldq(xmm_temp7, 4);
+    __ psrldq(xmm_temp9, 12);
+    __ por(xmm_temp3, xmm_temp7);
+    __ por(xmm_temp6, xmm_temp8);
+    __ por(xmm_temp6, xmm_temp9);
+
+    //
+    // First phase of the reduction
+    //
+    // Move xmm3 into xmm7, xmm8, xmm9 in order to perform the shifts
+    // independently.
+    __ movdqu(xmm_temp7, xmm_temp3);
+    __ movdqu(xmm_temp8, xmm_temp3);
+    __ movdqu(xmm_temp9, xmm_temp3);
+    __ pslld(xmm_temp7, 31);    // packed right shift shifting << 31
+    __ pslld(xmm_temp8, 30);    // packed right shift shifting << 30
+    __ pslld(xmm_temp9, 25);    // packed right shift shifting << 25
+    __ pxor(xmm_temp7, xmm_temp8);      // xor the shifted versions
+    __ pxor(xmm_temp7, xmm_temp9);
+    __ movdqu(xmm_temp8, xmm_temp7);
+    __ pslldq(xmm_temp7, 12);
+    __ psrldq(xmm_temp8, 4);
+    __ pxor(xmm_temp3, xmm_temp7);      // first phase of the reduction complete
+
+    //
+    // Second phase of the reduction
+    //
+    // Make 3 copies of xmm3 in xmm2, xmm4, xmm5 for doing these
+    // shift operations.
+    __ movdqu(xmm_temp2, xmm_temp3);
+    __ movdqu(xmm_temp4, xmm_temp3);
+    __ movdqu(xmm_temp5, xmm_temp3);
+    __ psrld(xmm_temp2, 1);     // packed left shifting >> 1
+    __ psrld(xmm_temp4, 2);     // packed left shifting >> 2
+    __ psrld(xmm_temp5, 7);     // packed left shifting >> 7
+    __ pxor(xmm_temp2, xmm_temp4);      // xor the shifted versions
+    __ pxor(xmm_temp2, xmm_temp5);
+    __ pxor(xmm_temp2, xmm_temp8);
+    __ pxor(xmm_temp3, xmm_temp2);
+    __ pxor(xmm_temp6, xmm_temp3);      // the result is in xmm6
+
+    __ decrement(blocks);
+    __ jcc(Assembler::zero, L_exit);
+    __ movdqu(xmm_temp0, xmm_temp6);
+    __ addptr(data, 16);
+    __ jmp(L_ghash_loop);
+
+    __ BIND(L_exit);
+    __ pshufb(xmm_temp6, xmm_temp10);          // Byte swap 16-byte result
+    __ movdqu(Address(state, 0), xmm_temp6);   // store the result
+
+#ifdef _WIN64
+    // restore xmm regs belonging to calling function
+    for (int i = 6; i <= XMM_REG_LAST; i++) {
+      __ movdqu(as_XMMRegister(i), xmm_save(i));
+    }
+#endif
+    __ leave();
+    __ ret(0);
+    return start;
+  }
+
   /**
    *  Arguments:
    *
@@ -4077,6 +4246,13 @@ class StubGenerator: public StubCodeGenerator {
       StubRoutines::_cipherBlockChaining_decryptAESCrypt = generate_cipherBlockChaining_decryptAESCrypt_Parallel();
     }
 
+    // Generate GHASH intrinsics code
+    if (UseGHASHIntrinsics) {
+      StubRoutines::x86::_ghash_long_swap_mask_addr = generate_ghash_long_swap_mask();
+      StubRoutines::x86::_ghash_byte_swap_mask_addr = generate_ghash_byte_swap_mask();
+      StubRoutines::_ghash_processBlocks = generate_ghash_processBlocks();
+    }
+
     // Safefetch stubs.
     generate_safefetch("SafeFetch32", sizeof(int),     &StubRoutines::_safefetch32_entry,
                                                        &StubRoutines::_safefetch32_fault_pc,
diff --git a/src/cpu/x86/vm/stubRoutines_x86.cpp b/src/cpu/x86/vm/stubRoutines_x86.cpp
index 200f2aff8..9b0d8fc75 100644
--- a/src/cpu/x86/vm/stubRoutines_x86.cpp
+++ b/src/cpu/x86/vm/stubRoutines_x86.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2013, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2013, 2015, Oracle and/or its affiliates. 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
@@ -33,6 +33,8 @@
 
 address StubRoutines::x86::_verify_mxcsr_entry = NULL;
 address StubRoutines::x86::_key_shuffle_mask_addr = NULL;
+address StubRoutines::x86::_ghash_long_swap_mask_addr = NULL;
+address StubRoutines::x86::_ghash_byte_swap_mask_addr = NULL;
 
 uint64_t StubRoutines::x86::_crc_by128_masks[] =
 {
diff --git a/src/cpu/x86/vm/stubRoutines_x86.hpp b/src/cpu/x86/vm/stubRoutines_x86.hpp
index d8e52ab3b..bb160486c 100644
--- a/src/cpu/x86/vm/stubRoutines_x86.hpp
+++ b/src/cpu/x86/vm/stubRoutines_x86.hpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2013, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2013, 2015, Oracle and/or its affiliates. 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
@@ -36,10 +36,15 @@
   // masks and table for CRC32
   static uint64_t _crc_by128_masks[];
   static juint    _crc_table[];
+  // swap mask for ghash
+  static address _ghash_long_swap_mask_addr;
+  static address _ghash_byte_swap_mask_addr;
 
  public:
   static address verify_mxcsr_entry()    { return _verify_mxcsr_entry; }
   static address key_shuffle_mask_addr() { return _key_shuffle_mask_addr; }
   static address crc_by128_masks_addr()  { return (address)_crc_by128_masks; }
+  static address ghash_long_swap_mask_addr() { return _ghash_long_swap_mask_addr; }
+  static address ghash_byte_swap_mask_addr() { return _ghash_byte_swap_mask_addr; }
 
 #endif // CPU_X86_VM_STUBROUTINES_X86_32_HPP
diff --git a/src/cpu/x86/vm/vm_version_x86.cpp b/src/cpu/x86/vm/vm_version_x86.cpp
index fd0a68d10..3a4246c68 100644
--- a/src/cpu/x86/vm/vm_version_x86.cpp
+++ b/src/cpu/x86/vm/vm_version_x86.cpp
@@ -553,12 +553,36 @@ void VM_Version::get_processor_features() {
   // Use AES instructions if available.
   if (supports_aes()) {
     if (FLAG_IS_DEFAULT(UseAES)) {
-      UseAES = true;
+      FLAG_SET_DEFAULT(UseAES, true);
     }
-  } else if (UseAES) {
-    if (!FLAG_IS_DEFAULT(UseAES))
+    if (!UseAES) {
+      if (UseAESIntrinsics && !FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+        warning("AES intrinsics require UseAES flag to be enabled. Intrinsics will be disabled.");
+      }
+      FLAG_SET_DEFAULT(UseAESIntrinsics, false);
+    } else {
+      if (UseSSE > 2) {
+        if (FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+          FLAG_SET_DEFAULT(UseAESIntrinsics, true);
+        }
+      } else {
+        // The AES intrinsic stubs require AES instruction support (of course)
+        // but also require sse3 mode or higher for instructions it use.
+        if (UseAESIntrinsics && !FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+          warning("X86 AES intrinsics require SSE3 instructions or higher. Intrinsics will be disabled.");
+        }
+        FLAG_SET_DEFAULT(UseAESIntrinsics, false);
+      }
+    }
+  } else if (UseAES || UseAESIntrinsics) {
+    if (UseAES && !FLAG_IS_DEFAULT(UseAES)) {
       warning("AES instructions are not available on this CPU");
-    FLAG_SET_DEFAULT(UseAES, false);
+      FLAG_SET_DEFAULT(UseAES, false);
+    }
+    if (UseAESIntrinsics && !FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+      warning("AES intrinsics are not available on this CPU");
+      FLAG_SET_DEFAULT(UseAESIntrinsics, false);
+    }
   }
 
   // Use CLMUL instructions if available.
@@ -582,16 +606,15 @@ void VM_Version::get_processor_features() {
     FLAG_SET_DEFAULT(UseCRC32Intrinsics, false);
   }
 
-  // The AES intrinsic stubs require AES instruction support (of course)
-  // but also require sse3 mode for instructions it use.
-  if (UseAES && (UseSSE > 2)) {
-    if (FLAG_IS_DEFAULT(UseAESIntrinsics)) {
-      UseAESIntrinsics = true;
+  // GHASH/GCM intrinsics
+  if (UseCLMUL && (UseSSE > 2)) {
+    if (FLAG_IS_DEFAULT(UseGHASHIntrinsics)) {
+      UseGHASHIntrinsics = true;
     }
-  } else if (UseAESIntrinsics) {
-    if (!FLAG_IS_DEFAULT(UseAESIntrinsics))
-      warning("AES intrinsics are not available on this CPU");
-    FLAG_SET_DEFAULT(UseAESIntrinsics, false);
+  } else if (UseGHASHIntrinsics) {
+    if (!FLAG_IS_DEFAULT(UseGHASHIntrinsics))
+      warning("GHASH intrinsic requires CLMUL and SSE2 instructions on this CPU");
+    FLAG_SET_DEFAULT(UseGHASHIntrinsics, false);
   }
 
   if (UseSHA) {
diff --git a/src/share/vm/c1/c1_LIR.cpp b/src/share/vm/c1/c1_LIR.cpp
index 76086794f..900bd827a 100644
--- a/src/share/vm/c1/c1_LIR.cpp
+++ b/src/share/vm/c1/c1_LIR.cpp
@@ -2112,8 +2112,14 @@ void LIR_OpProfileCall::print_instr(outputStream* out) const {
 
 // LIR_OpProfileType
 void LIR_OpProfileType::print_instr(outputStream* out) const {
-  out->print("exact = "); exact_klass()->print_name_on(out);
-  out->print("current = "); ciTypeEntries::print_ciklass(out, current_klass());
+  out->print("exact = ");
+  if  (exact_klass() == NULL) {
+    out->print("unknown");
+  } else {
+    exact_klass()->print_name_on(out);
+  }
+  out->print(" current = "); ciTypeEntries::print_ciklass(out, current_klass());
+  out->print(" ");
   mdp()->print(out);          out->print(" ");
   obj()->print(out);          out->print(" ");
   tmp()->print(out);          out->print(" ");
diff --git a/src/share/vm/classfile/vmSymbols.hpp b/src/share/vm/classfile/vmSymbols.hpp
index 7f0e82071..46f2e2d8b 100644
--- a/src/share/vm/classfile/vmSymbols.hpp
+++ b/src/share/vm/classfile/vmSymbols.hpp
@@ -863,6 +863,12 @@
    do_name(     implCompressMB_name,                               "implCompressMultiBlock0")                           \
    do_signature(implCompressMB_signature,                          "([BII)I")                                           \
                                                                                                                         \
+  /* support for com.sun.crypto.provider.GHASH */                                                                       \
+  do_class(com_sun_crypto_provider_ghash, "com/sun/crypto/provider/GHASH")                                              \
+  do_intrinsic(_ghash_processBlocks, com_sun_crypto_provider_ghash, processBlocks_name, ghash_processBlocks_signature, F_S) \
+   do_name(processBlocks_name, "processBlocks")                                                                         \
+   do_signature(ghash_processBlocks_signature, "([BII[J[J)V")                                                           \
+                                                                                                                        \
   /* support for java.util.zip */                                                                                       \
   do_class(java_util_zip_CRC32,           "java/util/zip/CRC32")                                                        \
   do_intrinsic(_updateCRC32,               java_util_zip_CRC32,   update_name, int2_int_signature,               F_SN)  \
diff --git a/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.cpp b/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.cpp
index e5a3a1106..2a2faa69a 100644
--- a/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.cpp
+++ b/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.cpp
@@ -161,6 +161,8 @@ CompactibleFreeListSpace::CompactibleFreeListSpace(BlockOffsetSharedArray* bs,
     }
     _dictionary->set_par_lock(&_parDictionaryAllocLock);
   }
+
+  _used_stable = 0;
 }
 
 // Like CompactibleSpace forward() but always calls cross_threshold() to
@@ -377,6 +379,14 @@ size_t CompactibleFreeListSpace::used() const {
   return capacity() - free();
 }
 
+size_t CompactibleFreeListSpace::used_stable() const {
+  return _used_stable;
+}
+
+void CompactibleFreeListSpace::recalculate_used_stable() {
+  _used_stable = used();
+}
+
 size_t CompactibleFreeListSpace::free() const {
   // "MT-safe, but not MT-precise"(TM), if you will: i.e.
   // if you do this while the structures are in flux you
@@ -1218,6 +1228,13 @@ HeapWord* CompactibleFreeListSpace::allocate(size_t size) {
     debug_only(fc->mangleAllocated(size));
   }
 
+  // During GC we do not need to recalculate the stable used value for
+  // every allocation in old gen. It is done once at the end of GC instead
+  // for performance reasons.
+  if (!Universe::heap()->is_gc_active()) {
+    recalculate_used_stable();
+  }
+
   return res;
 }
 
diff --git a/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.hpp b/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.hpp
index 7269068f6..121299d4b 100644
--- a/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.hpp
+++ b/src/share/vm/gc_implementation/concurrentMarkSweep/compactibleFreeListSpace.hpp
@@ -148,6 +148,9 @@ class CompactibleFreeListSpace: public CompactibleSpace {
   // Used to keep track of limit of sweep for the space
   HeapWord* _sweep_limit;
 
+  // Stable value of used().
+  size_t _used_stable;
+
   // Support for compacting cms
   HeapWord* cross_threshold(HeapWord* start, HeapWord* end);
   HeapWord* forward(oop q, size_t size, CompactPoint* cp, HeapWord* compact_top);
@@ -343,6 +346,17 @@ class CompactibleFreeListSpace: public CompactibleSpace {
   // which overestimates the region by returning the entire
   // committed region (this is safe, but inefficient).
 
+  // Returns monotonically increasing stable used space bytes for CMS.
+  // This is required for jstat and other memory monitoring tools
+  // that might otherwise see inconsistent used space values during a garbage
+  // collection, promotion or allocation into compactibleFreeListSpace.
+  // The value returned by this function might be smaller than the
+  // actual value.
+  size_t used_stable() const;
+  // Recalculate and cache the current stable used() value. Only to be called
+  // in places where we can be sure that the result is stable.
+  void recalculate_used_stable();
+
   // Returns a subregion of the space containing all the objects in
   // the space.
   MemRegion used_region() const {
diff --git a/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.cpp b/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.cpp
index 1872e72e3..e7228d313 100644
--- a/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.cpp
+++ b/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.cpp
@@ -869,6 +869,10 @@ ConcurrentMarkSweepGeneration::unsafe_max_alloc_nogc() const {
   return _cmsSpace->max_alloc_in_words() * HeapWordSize;
 }
 
+size_t ConcurrentMarkSweepGeneration::used_stable() const {
+  return cmsSpace()->used_stable();
+}
+
 size_t ConcurrentMarkSweepGeneration::max_available() const {
   return free() + _virtual_space.uncommitted_size();
 }
@@ -1955,6 +1959,8 @@ void CMSCollector::compute_new_size() {
   FreelistLocker z(this);
   MetaspaceGC::compute_new_size();
   _cmsGen->compute_new_size_free_list();
+  // recalculate CMS used space after CMS collection
+  _cmsGen->cmsSpace()->recalculate_used_stable();
 }
 
 // A work method used by foreground collection to determine
@@ -2768,6 +2774,7 @@ void ConcurrentMarkSweepGeneration::gc_prologue(bool full) {
 
   _capacity_at_prologue = capacity();
   _used_at_prologue = used();
+  _cmsSpace->recalculate_used_stable();
 
   // Delegate to CMScollector which knows how to coordinate between
   // this and any other CMS generations that it is responsible for
@@ -2837,6 +2844,7 @@ void CMSCollector::gc_epilogue(bool full) {
   _eden_chunk_index = 0;
 
   size_t cms_used   = _cmsGen->cmsSpace()->used();
+  _cmsGen->cmsSpace()->recalculate_used_stable();
 
   // update performance counters - this uses a special version of
   // update_counters() that allows the utilization to be passed as a
@@ -3672,6 +3680,7 @@ void CMSCollector::checkpointRootsInitial(bool asynch) {
     _collectorState = Marking;
   }
   SpecializationStats::print();
+  _cmsGen->cmsSpace()->recalculate_used_stable();
 }
 
 void CMSCollector::checkpointRootsInitialWork(bool asynch) {
@@ -5066,10 +5075,12 @@ void CMSCollector::checkpointRootsFinal(bool asynch,
                     Mutex::_no_safepoint_check_flag);
     assert(!init_mark_was_synchronous, "but that's impossible!");
     checkpointRootsFinalWork(asynch, clear_all_soft_refs, false);
+    _cmsGen->cmsSpace()->recalculate_used_stable();
   } else {
     // already have all the locks
     checkpointRootsFinalWork(asynch, clear_all_soft_refs,
                              init_mark_was_synchronous);
+    _cmsGen->cmsSpace()->recalculate_used_stable();
   }
   verify_work_stacks_empty();
   verify_overflow_empty();
@@ -6368,6 +6379,10 @@ void CMSCollector::sweep(bool asynch) {
       // Update heap occupancy information which is used as
       // input to soft ref clearing policy at the next gc.
       Universe::update_heap_info_at_gc();
+
+      // recalculate CMS used space after CMS collection
+      _cmsGen->cmsSpace()->recalculate_used_stable();
+
       _collectorState = Resizing;
     }
   } else {
@@ -6467,6 +6482,7 @@ void ConcurrentMarkSweepGeneration::update_gc_stats(int current_level,
     // Gather statistics on the young generation collection.
     collector()->stats().record_gc0_end(used());
   }
+  _cmsSpace->recalculate_used_stable();
 }
 
 CMSAdaptiveSizePolicy* ConcurrentMarkSweepGeneration::size_policy() {
diff --git a/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.hpp b/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.hpp
index a6d06a5dc..a023b9fb9 100644
--- a/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.hpp
+++ b/src/share/vm/gc_implementation/concurrentMarkSweep/concurrentMarkSweepGeneration.hpp
@@ -1190,6 +1190,7 @@ class ConcurrentMarkSweepGeneration: public CardGeneration {
   double occupancy() const { return ((double)used())/((double)capacity()); }
   size_t contiguous_available() const;
   size_t unsafe_max_alloc_nogc() const;
+  size_t used_stable() const;
 
   // over-rides
   MemRegion used_region() const;
diff --git a/src/share/vm/gc_implementation/g1/g1CollectedHeap.cpp b/src/share/vm/gc_implementation/g1/g1CollectedHeap.cpp
index aa6bbc81a..4100b83ef 100644
--- a/src/share/vm/gc_implementation/g1/g1CollectedHeap.cpp
+++ b/src/share/vm/gc_implementation/g1/g1CollectedHeap.cpp
@@ -2520,6 +2520,12 @@ void G1CollectedHeap::collect(GCCause::Cause cause) {
           }
         }
       }
+    } else if (GC_locker::should_discard(cause, gc_count_before)) {
+      // Return to be consistent with VMOp failure due to another
+      // collection slipping in after our gc_count but before our
+      // request is processed.  _gc_locker collections upgraded by
+      // GCLockerInvokesConcurrent are handled above and never discarded.
+      return;
     } else {
       if (cause == GCCause::_gc_locker || cause == GCCause::_wb_young_gc
           DEBUG_ONLY(|| cause == GCCause::_scavenge_alot)) {
diff --git a/src/share/vm/gc_implementation/g1/g1CollectorPolicy.cpp b/src/share/vm/gc_implementation/g1/g1CollectorPolicy.cpp
index c4463f50f..b25ab59f2 100644
--- a/src/share/vm/gc_implementation/g1/g1CollectorPolicy.cpp
+++ b/src/share/vm/gc_implementation/g1/g1CollectorPolicy.cpp
@@ -376,7 +376,7 @@ G1YoungGenSizer::G1YoungGenSizer() : _sizer_kind(SizerDefaults), _adaptive_size(
                              MAX2((uint) (MaxNewSize / HeapRegion::GrainBytes),
                                   1U);
       _sizer_kind = SizerMaxAndNewSize;
-      _adaptive_size = _min_desired_young_length == _max_desired_young_length;
+      _adaptive_size = _min_desired_young_length != _max_desired_young_length;
     } else {
       _sizer_kind = SizerNewSizeOnly;
     }
diff --git a/src/share/vm/gc_implementation/g1/g1CollectorPolicy.hpp b/src/share/vm/gc_implementation/g1/g1CollectorPolicy.hpp
index 96e4dc63d..d15dee26c 100644
--- a/src/share/vm/gc_implementation/g1/g1CollectorPolicy.hpp
+++ b/src/share/vm/gc_implementation/g1/g1CollectorPolicy.hpp
@@ -133,7 +133,11 @@ private:
   SizerKind _sizer_kind;
   uint _min_desired_young_length;
   uint _max_desired_young_length;
+
+  // False when using a fixed young generation size due to command-line options,
+  // true otherwise.
   bool _adaptive_size;
+
   uint calculate_default_min_length(uint new_number_of_heap_regions);
   uint calculate_default_max_length(uint new_number_of_heap_regions);
 
diff --git a/src/share/vm/gc_implementation/parallelScavenge/parallelScavengeHeap.cpp b/src/share/vm/gc_implementation/parallelScavenge/parallelScavengeHeap.cpp
index f407f20b7..e13fefa2c 100644
--- a/src/share/vm/gc_implementation/parallelScavenge/parallelScavengeHeap.cpp
+++ b/src/share/vm/gc_implementation/parallelScavenge/parallelScavengeHeap.cpp
@@ -530,6 +530,10 @@ void ParallelScavengeHeap::collect(GCCause::Cause cause) {
     full_gc_count = Universe::heap()->total_full_collections();
   }
 
+  if (GC_locker::should_discard(cause, gc_count)) {
+    return;
+  }
+
   VM_ParallelGCSystemGC op(gc_count, full_gc_count, cause);
   VMThread::execute(&op);
 }
diff --git a/src/share/vm/gc_implementation/parallelScavenge/vmPSOperations.cpp b/src/share/vm/gc_implementation/parallelScavenge/vmPSOperations.cpp
index 9148bb4ff..2d554b397 100644
--- a/src/share/vm/gc_implementation/parallelScavenge/vmPSOperations.cpp
+++ b/src/share/vm/gc_implementation/parallelScavenge/vmPSOperations.cpp
@@ -52,11 +52,16 @@ void VM_ParallelGCFailedAllocation::doit() {
   }
 }
 
+static bool is_cause_full(GCCause::Cause cause) {
+  return (cause != GCCause::_gc_locker) && (cause != GCCause::_wb_young_gc)
+         DEBUG_ONLY(&& (cause != GCCause::_scavenge_alot));
+}
+
 // Only used for System.gc() calls
 VM_ParallelGCSystemGC::VM_ParallelGCSystemGC(uint gc_count,
                                              uint full_gc_count,
                                              GCCause::Cause gc_cause) :
-  VM_GC_Operation(gc_count, gc_cause, full_gc_count, true /* full */)
+  VM_GC_Operation(gc_count, gc_cause, full_gc_count, is_cause_full(gc_cause))
 {
 }
 
@@ -68,8 +73,7 @@ void VM_ParallelGCSystemGC::doit() {
     "must be a ParallelScavengeHeap");
 
   GCCauseSetter gccs(heap, _gc_cause);
-  if (_gc_cause == GCCause::_gc_locker || _gc_cause == GCCause::_wb_young_gc
-      DEBUG_ONLY(|| _gc_cause == GCCause::_scavenge_alot)) {
+  if (!_full) {
     // If (and only if) the scavenge fails, this will invoke a full gc.
     heap->invoke_scavenge();
   } else {
diff --git a/src/share/vm/gc_implementation/shared/gSpaceCounters.hpp b/src/share/vm/gc_implementation/shared/gSpaceCounters.hpp
index 7a5fdf9ee..b03430e27 100644
--- a/src/share/vm/gc_implementation/shared/gSpaceCounters.hpp
+++ b/src/share/vm/gc_implementation/shared/gSpaceCounters.hpp
@@ -63,7 +63,7 @@ class GSpaceCounters: public CHeapObj<mtGC> {
   }
 
   inline void update_used() {
-    _used->set_value(_gen->used());
+    _used->set_value(_gen->used_stable());
   }
 
   // special version of update_used() to allow the used value to be
@@ -107,7 +107,7 @@ class GenerationUsedHelper : public PerfLongSampleHelper {
     GenerationUsedHelper(Generation* g) : _gen(g) { }
 
     inline jlong take_sample() {
-      return _gen->used();
+      return _gen->used_stable();
     }
 };
 
diff --git a/src/share/vm/gc_implementation/shared/vmGCOperations.cpp b/src/share/vm/gc_implementation/shared/vmGCOperations.cpp
index 972099b9c..d60f751af 100644
--- a/src/share/vm/gc_implementation/shared/vmGCOperations.cpp
+++ b/src/share/vm/gc_implementation/shared/vmGCOperations.cpp
@@ -201,6 +201,19 @@ void VM_GenCollectForAllocation::doit() {
   }
 }
 
+static bool is_full_gc(int max_level) {
+  // Return true if max_level is all generations
+  return (max_level == (GenCollectedHeap::heap()->n_gens() - 1));
+}
+
+VM_GenCollectFull::VM_GenCollectFull(uint gc_count_before,
+                                     uint full_gc_count_before,
+                                     GCCause::Cause gc_cause,
+                                     int max_level) :
+  VM_GC_Operation(gc_count_before, gc_cause, full_gc_count_before,
+                  is_full_gc(max_level) /* full */),
+  _max_level(max_level) { }
+
 void VM_GenCollectFull::doit() {
   SvcGCMarker sgcm(SvcGCMarker::FULL);
 
diff --git a/src/share/vm/gc_implementation/shared/vmGCOperations.hpp b/src/share/vm/gc_implementation/shared/vmGCOperations.hpp
index b8027a972..cb070bd73 100644
--- a/src/share/vm/gc_implementation/shared/vmGCOperations.hpp
+++ b/src/share/vm/gc_implementation/shared/vmGCOperations.hpp
@@ -201,9 +201,7 @@ class VM_GenCollectFull: public VM_GC_Operation {
   VM_GenCollectFull(uint gc_count_before,
                     uint full_gc_count_before,
                     GCCause::Cause gc_cause,
-                    int max_level)
-    : VM_GC_Operation(gc_count_before, gc_cause, full_gc_count_before, true /* full */),
-      _max_level(max_level) { }
+                    int max_level);
   ~VM_GenCollectFull() {}
   virtual VMOp_Type type() const { return VMOp_GenCollectFull; }
   virtual void doit();
diff --git a/src/share/vm/memory/gcLocker.cpp b/src/share/vm/memory/gcLocker.cpp
index 967426379..df8914e50 100644
--- a/src/share/vm/memory/gcLocker.cpp
+++ b/src/share/vm/memory/gcLocker.cpp
@@ -31,6 +31,7 @@
 volatile jint GC_locker::_jni_lock_count = 0;
 volatile bool GC_locker::_needs_gc       = false;
 volatile bool GC_locker::_doing_gc       = false;
+unsigned int  GC_locker::_total_collections = 0;
 
 #ifdef ASSERT
 volatile jint GC_locker::_debug_jni_lock_count = 0;
@@ -94,6 +95,11 @@ void GC_locker::stall_until_clear() {
   }
 }
 
+bool GC_locker::should_discard(GCCause::Cause cause, uint total_collections) {
+  return (cause == GCCause::_gc_locker) &&
+         (_total_collections != total_collections);
+}
+
 void GC_locker::jni_lock(JavaThread* thread) {
   assert(!thread->in_critical(), "shouldn't currently be in a critical region");
   MutexLocker mu(JNICritical_lock);
@@ -117,7 +123,13 @@ void GC_locker::jni_unlock(JavaThread* thread) {
   decrement_debug_jni_lock_count();
   thread->exit_critical();
   if (needs_gc() && !is_active_internal()) {
-    // We're the last thread out. Cause a GC to occur.
+    // We're the last thread out. Request a GC.
+    // Capture the current total collections, to allow detection of
+    // other collections that make this one unnecessary.  The value of
+    // total_collections() is only changed at a safepoint, so there
+    // must not be a safepoint between the lock becoming inactive and
+    // getting the count, else there may be unnecessary GCLocker GCs.
+    _total_collections = Universe::heap()->total_collections();
     _doing_gc = true;
     {
       // Must give up the lock while at a safepoint
diff --git a/src/share/vm/memory/gcLocker.hpp b/src/share/vm/memory/gcLocker.hpp
index f12aa6755..068688e93 100644
--- a/src/share/vm/memory/gcLocker.hpp
+++ b/src/share/vm/memory/gcLocker.hpp
@@ -26,6 +26,7 @@
 #define SHARE_VM_MEMORY_GCLOCKER_HPP
 
 #include "gc_interface/collectedHeap.hpp"
+#include "gc_interface/gcCause.hpp"
 #include "memory/genCollectedHeap.hpp"
 #include "memory/universe.hpp"
 #include "oops/oop.hpp"
@@ -57,6 +58,7 @@ class GC_locker: public AllStatic {
   static volatile bool _needs_gc;        // heap is filling, we need a GC
                                          // note: bool is typedef'd as jint
   static volatile bool _doing_gc;        // unlock_critical() is doing a GC
+  static uint _total_collections;        // value for _gc_locker collection
 
 #ifdef ASSERT
   // This lock count is updated for all operations and is used to
@@ -116,6 +118,12 @@ class GC_locker: public AllStatic {
   // Sets _needs_gc if is_active() is true. Returns is_active().
   static bool check_active_before_gc();
 
+  // Return true if the designated collection is a GCLocker request
+  // that should be discarded.  Returns true if cause == GCCause::_gc_locker
+  // and the given total collection value indicates a collection has been
+  // done since the GCLocker request was made.
+  static bool should_discard(GCCause::Cause cause, uint total_collections);
+
   // Stalls the caller (who should not be in a jni critical section)
   // until needs_gc() clears. Note however that needs_gc() may be
   // set at a subsequent safepoint and/or cleared under the
diff --git a/src/share/vm/memory/genCollectedHeap.cpp b/src/share/vm/memory/genCollectedHeap.cpp
index ddf0a860c..daa0a06bf 100644
--- a/src/share/vm/memory/genCollectedHeap.cpp
+++ b/src/share/vm/memory/genCollectedHeap.cpp
@@ -796,8 +796,11 @@ void GenCollectedHeap::collect(GCCause::Cause cause) {
 #else  // INCLUDE_ALL_GCS
     ShouldNotReachHere();
 #endif // INCLUDE_ALL_GCS
-  } else if (cause == GCCause::_wb_young_gc) {
-    // minor collection for WhiteBox API
+  } else if ((cause == GCCause::_wb_young_gc) ||
+             (cause == GCCause::_gc_locker)) {
+    // minor collection for WhiteBox or GCLocker.
+    // _gc_locker collections upgraded by GCLockerInvokesConcurrent
+    // are handled above and never discarded.
     collect(cause, 0);
   } else {
 #ifdef ASSERT
@@ -835,6 +838,11 @@ void GenCollectedHeap::collect_locked(GCCause::Cause cause, int max_level) {
   // Read the GC count while holding the Heap_lock
   unsigned int gc_count_before      = total_collections();
   unsigned int full_gc_count_before = total_full_collections();
+
+  if (GC_locker::should_discard(cause, gc_count_before)) {
+    return;
+  }
+
   {
     MutexUnlocker mu(Heap_lock);  // give up heap lock, execute gets it back
     VM_GenCollectFull op(gc_count_before, full_gc_count_before,
@@ -887,24 +895,16 @@ void GenCollectedHeap::do_full_collection(bool clear_all_soft_refs) {
 
 void GenCollectedHeap::do_full_collection(bool clear_all_soft_refs,
                                           int max_level) {
-  int local_max_level;
-  if (!incremental_collection_will_fail(false /* don't consult_young */) &&
-      gc_cause() == GCCause::_gc_locker) {
-    local_max_level = 0;
-  } else {
-    local_max_level = max_level;
-  }
 
   do_collection(true                 /* full */,
                 clear_all_soft_refs  /* clear_all_soft_refs */,
                 0                    /* size */,
                 false                /* is_tlab */,
-                local_max_level      /* max_level */);
+                max_level            /* max_level */);
   // Hack XXX FIX ME !!!
   // A scavenge may not have been attempted, or may have
   // been attempted and failed, because the old gen was too full
-  if (local_max_level == 0 && gc_cause() == GCCause::_gc_locker &&
-      incremental_collection_will_fail(false /* don't consult_young */)) {
+  if (gc_cause() == GCCause::_gc_locker && incremental_collection_failed()) {
     if (PrintGCDetails) {
       gclog_or_tty->print_cr("GC locker: Trying a full collection "
                              "because scavenge failed");
diff --git a/src/share/vm/memory/generation.cpp b/src/share/vm/memory/generation.cpp
index 3c45a9192..dc4ac0869 100644
--- a/src/share/vm/memory/generation.cpp
+++ b/src/share/vm/memory/generation.cpp
@@ -68,6 +68,12 @@ GenerationSpec* Generation::spec() {
   return gch->_gen_specs[level()];
 }
 
+// This is for CMS. It returns stable monotonic used space size.
+// Remove this when CMS is removed.
+size_t Generation::used_stable() const {
+  return used();
+}
+
 size_t Generation::max_capacity() const {
   return reserved().byte_size();
 }
diff --git a/src/share/vm/memory/generation.hpp b/src/share/vm/memory/generation.hpp
index 63dccb70f..ca3fe949d 100644
--- a/src/share/vm/memory/generation.hpp
+++ b/src/share/vm/memory/generation.hpp
@@ -168,6 +168,7 @@ class Generation: public CHeapObj<mtGC> {
   virtual size_t capacity() const = 0;  // The maximum number of object bytes the
                                         // generation can currently hold.
   virtual size_t used() const = 0;      // The number of used bytes in the gen.
+  virtual size_t used_stable() const;   // The number of used bytes for memory monitoring tools.
   virtual size_t free() const = 0;      // The number of free bytes in the gen.
 
   // Support for java.lang.Runtime.maxMemory(); see CollectedHeap.
diff --git a/src/share/vm/oops/instanceKlass.cpp b/src/share/vm/oops/instanceKlass.cpp
index 0204188ad..4c973d38b 100644
--- a/src/share/vm/oops/instanceKlass.cpp
+++ b/src/share/vm/oops/instanceKlass.cpp
@@ -294,6 +294,7 @@ InstanceKlass::InstanceKlass(int vtable_len,
   set_has_unloaded_dependent(false);
   set_init_state(InstanceKlass::allocated);
   set_init_thread(NULL);
+  set_init_state(allocated);
   set_reference_type(rt);
   set_oop_map_cache(NULL);
   set_jni_ids(NULL);
@@ -978,11 +979,13 @@ void InstanceKlass::set_initialization_state_and_notify_impl(instanceKlassHandle
   oop init_lock = this_oop->init_lock();
   if (init_lock != NULL) {
     ObjectLocker ol(init_lock, THREAD);
+    this_oop->set_init_thread(NULL); // reset _init_thread before changing _init_state
     this_oop->set_init_state(state);
     this_oop->fence_and_clear_init_lock();
     ol.notify_all(CHECK);
   } else {
     assert(init_lock != NULL, "The initialization state should never be set twice");
+    this_oop->set_init_thread(NULL); // reset _init_thread before changing _init_state
     this_oop->set_init_state(state);
   }
 }
@@ -3602,6 +3605,7 @@ void InstanceKlass::set_init_state(ClassState state) {
   bool good_state = is_shared() ? (_init_state <= state)
                                                : (_init_state < state);
   assert(good_state || state == allocated, "illegal state transition");
+  assert(_init_thread == NULL, "should be cleared before state change");
   _init_state = (u1)state;
 }
 #endif
diff --git a/src/share/vm/oops/instanceKlass.hpp b/src/share/vm/oops/instanceKlass.hpp
index 444eadd38..49d552822 100644
--- a/src/share/vm/oops/instanceKlass.hpp
+++ b/src/share/vm/oops/instanceKlass.hpp
@@ -241,7 +241,7 @@ class InstanceKlass: public Klass {
   u2              _misc_flags;
   u2              _minor_version;        // minor version number of class file
   u2              _major_version;        // major version number of class file
-  Thread*         _init_thread;          // Pointer to current thread doing initialization (to handle recusive initialization)
+  Thread*         _init_thread;          // Pointer to current thread doing initialization (to handle recursive initialization)
   int             _vtable_len;           // length of Java vtable (in words)
   int             _itable_len;           // length of Java itable (in words)
   OopMapCache*    volatile _oop_map_cache;   // OopMapCache for all methods in the klass (allocated lazily)
diff --git a/src/share/vm/oops/klassVtable.cpp b/src/share/vm/oops/klassVtable.cpp
index 61dd4c21d..b3f477cd0 100644
--- a/src/share/vm/oops/klassVtable.cpp
+++ b/src/share/vm/oops/klassVtable.cpp
@@ -289,22 +289,25 @@ InstanceKlass* klassVtable::find_transitive_override(InstanceKlass* initialsuper
                             int vtable_index, Handle target_loader, Symbol* target_classname, Thread * THREAD) {
   InstanceKlass* superk = initialsuper;
   while (superk != NULL && superk->super() != NULL) {
-    InstanceKlass* supersuperklass = InstanceKlass::cast(superk->super());
-    klassVtable* ssVtable = supersuperklass->vtable();
+    klassVtable* ssVtable = (superk->super())->vtable();
     if (vtable_index < ssVtable->length()) {
       Method* super_method = ssVtable->method_at(vtable_index);
+      // get the class holding the matching method
+      // make sure you use that class for is_override
+      InstanceKlass* supermethodholder = super_method->method_holder();
 #ifndef PRODUCT
       Symbol* name= target_method()->name();
       Symbol* signature = target_method()->signature();
       assert(super_method->name() == name && super_method->signature() == signature, "vtable entry name/sig mismatch");
 #endif
-      if (supersuperklass->is_override(super_method, target_loader, target_classname, THREAD)) {
+
+      if (supermethodholder->is_override(super_method, target_loader, target_classname, THREAD)) {
 #ifndef PRODUCT
         if (PrintVtables && Verbose) {
           ResourceMark rm(THREAD);
           char* sig = target_method()->name_and_sig_as_C_string();
           tty->print("transitive overriding superclass %s with %s::%s index %d, original flags: ",
-           supersuperklass->internal_name(),
+           supermethodholder->internal_name(),
            _klass->internal_name(), sig, vtable_index);
            super_method->access_flags().print_on(tty);
            if (super_method->is_default_method()) {
@@ -656,7 +659,7 @@ bool klassVtable::needs_new_vtable_entry(methodHandle target_method,
 
   // search through the super class hierarchy to see if we need
   // a new entry
-  ResourceMark rm;
+  ResourceMark rm(THREAD);
   Symbol* name = target_method()->name();
   Symbol* signature = target_method()->signature();
   Klass* k = super;
diff --git a/src/share/vm/opto/escape.cpp b/src/share/vm/opto/escape.cpp
index 77c49ba72..66e5c3365 100644
--- a/src/share/vm/opto/escape.cpp
+++ b/src/share/vm/opto/escape.cpp
@@ -952,6 +952,7 @@ void ConnectionGraph::process_call_arguments(CallNode *call) {
                   strcmp(call->as_CallLeaf()->_name, "aescrypt_decryptBlock") == 0 ||
                   strcmp(call->as_CallLeaf()->_name, "cipherBlockChaining_encryptAESCrypt") == 0 ||
                   strcmp(call->as_CallLeaf()->_name, "cipherBlockChaining_decryptAESCrypt") == 0 ||
+                  strcmp(call->as_CallLeaf()->_name, "ghash_processBlocks") == 0 ||
                   strcmp(call->as_CallLeaf()->_name, "sha1_implCompress") == 0 ||
                   strcmp(call->as_CallLeaf()->_name, "sha1_implCompressMB") == 0 ||
                   strcmp(call->as_CallLeaf()->_name, "sha256_implCompress") == 0 ||
@@ -2114,6 +2115,9 @@ bool ConnectionGraph::not_global_escape(Node *n) {
     return false;
   }
   PointsToNode* ptn = ptnode_adr(idx);
+  if (ptn == NULL) {
+    return false; // not in congraph (e.g. ConI)
+  }
   PointsToNode::EscapeState es = ptn->escape_state();
   // If we have already computed a value, return it.
   if (es >= PointsToNode::GlobalEscape)
diff --git a/src/share/vm/opto/library_call.cpp b/src/share/vm/opto/library_call.cpp
index 4bb5ca886..5c5fec586 100644
--- a/src/share/vm/opto/library_call.cpp
+++ b/src/share/vm/opto/library_call.cpp
@@ -311,6 +311,7 @@ class LibraryCallKit : public GraphKit {
   Node* inline_cipherBlockChaining_AESCrypt_predicate(bool decrypting);
   Node* get_key_start_from_aescrypt_object(Node* aescrypt_object);
   Node* get_original_key_start_from_aescrypt_object(Node* aescrypt_object);
+  bool inline_ghash_processBlocks();
   bool inline_sha_implCompress(vmIntrinsics::ID id);
   bool inline_digestBase_implCompressMB(int predicate);
   bool inline_sha_implCompressMB(Node* digestBaseObj, ciInstanceKlass* instklass_SHA,
@@ -570,6 +571,10 @@ CallGenerator* Compile::make_vm_intrinsic(ciMethod* m, bool is_virtual) {
     predicates = 3;
     break;
 
+  case vmIntrinsics::_ghash_processBlocks:
+    if (!UseGHASHIntrinsics) return NULL;
+    break;
+
   case vmIntrinsics::_updateCRC32:
   case vmIntrinsics::_updateBytesCRC32:
   case vmIntrinsics::_updateByteBufferCRC32:
@@ -957,6 +962,9 @@ bool LibraryCallKit::try_to_inline(int predicate) {
   case vmIntrinsics::_montgomerySquare:
     return inline_montgomerySquare();
 
+  case vmIntrinsics::_ghash_processBlocks:
+    return inline_ghash_processBlocks();
+
   case vmIntrinsics::_encodeISOArray:
     return inline_encodeISOArray();
 
@@ -6599,6 +6607,35 @@ Node* LibraryCallKit::inline_cipherBlockChaining_AESCrypt_predicate(bool decrypt
   return _gvn.transform(region);
 }
 
+//------------------------------inline_ghash_processBlocks
+bool LibraryCallKit::inline_ghash_processBlocks() {
+  address stubAddr;
+  const char *stubName;
+  assert(UseGHASHIntrinsics, "need GHASH intrinsics support");
+
+  stubAddr = StubRoutines::ghash_processBlocks();
+  stubName = "ghash_processBlocks";
+
+  Node* data           = argument(0);
+  Node* offset         = argument(1);
+  Node* len            = argument(2);
+  Node* state          = argument(3);
+  Node* subkeyH        = argument(4);
+
+  Node* state_start  = array_element_address(state, intcon(0), T_LONG);
+  assert(state_start, "state is NULL");
+  Node* subkeyH_start  = array_element_address(subkeyH, intcon(0), T_LONG);
+  assert(subkeyH_start, "subkeyH is NULL");
+  Node* data_start  = array_element_address(data, offset, T_BYTE);
+  assert(data_start, "data is NULL");
+
+  Node* ghash = make_runtime_call(RC_LEAF|RC_NO_FP,
+                                  OptoRuntime::ghash_processBlocks_Type(),
+                                  stubAddr, stubName, TypePtr::BOTTOM,
+                                  state_start, subkeyH_start, data_start, len);
+  return true;
+}
+
 //------------------------------inline_sha_implCompress-----------------------
 //
 // Calculate SHA (i.e., SHA-1) for single-block byte[] array.
diff --git a/src/share/vm/opto/loopTransform.cpp b/src/share/vm/opto/loopTransform.cpp
index 7ef67c155..e3637b652 100644
--- a/src/share/vm/opto/loopTransform.cpp
+++ b/src/share/vm/opto/loopTransform.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2000, 2013, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2000, 2019, Oracle and/or its affiliates. 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
@@ -2231,6 +2231,13 @@ bool IdealLoopTree::policy_do_remove_empty_loop( PhaseIdealLoop *phase ) {
     // We also need to replace the original limit to collapse loop exit.
     Node* cmp = cl->loopexit()->cmp_node();
     assert(cl->limit() == cmp->in(2), "sanity");
+    // Duplicate cmp node if it has other users
+    if (cmp->outcnt() > 1) {
+      cmp = cmp->clone();
+      cmp = phase->_igvn.register_new_node_with_optimizer(cmp);
+      BoolNode *bol = cl->loopexit()->in(CountedLoopEndNode::TestValue)->as_Bool();
+      phase->_igvn.replace_input_of(bol, 1, cmp); // put bol on worklist
+    }
     phase->_igvn._worklist.push(cmp->in(2)); // put limit on worklist
     phase->_igvn.replace_input_of(cmp, 2, exact_limit); // put cmp on worklist
   }
diff --git a/src/share/vm/opto/loopnode.hpp b/src/share/vm/opto/loopnode.hpp
index 0ce98a063..032f94e67 100644
--- a/src/share/vm/opto/loopnode.hpp
+++ b/src/share/vm/opto/loopnode.hpp
@@ -279,6 +279,7 @@ public:
     if (iv_phi == NULL) {
       return NULL;
     }
+    assert(iv_phi->is_Phi(), "should be PhiNode");
     Node *ln = iv_phi->in(0);
     if (ln->is_CountedLoop() && ln->as_CountedLoop()->loopexit() == this) {
       return (CountedLoopNode*)ln;
diff --git a/src/share/vm/opto/loopopts.cpp b/src/share/vm/opto/loopopts.cpp
index 0452a236e..788e55ef9 100644
--- a/src/share/vm/opto/loopopts.cpp
+++ b/src/share/vm/opto/loopopts.cpp
@@ -309,7 +309,7 @@ Node *PhaseIdealLoop::has_local_phi_input( Node *n ) {
       }
       return NULL;
     }
-    assert(m->is_Phi() || is_dominator(get_ctrl(m), n_ctrl), "m has strange control");
+    assert(n->is_Phi() || m->is_Phi() || is_dominator(get_ctrl(m), n_ctrl), "m has strange control");
   }
 
   return n_ctrl;
diff --git a/src/share/vm/opto/runtime.cpp b/src/share/vm/opto/runtime.cpp
index 57d2f5764..4562dbcd5 100644
--- a/src/share/vm/opto/runtime.cpp
+++ b/src/share/vm/opto/runtime.cpp
@@ -92,7 +92,25 @@
 // At command line specify the parameters: -XX:+FullGCALot -XX:FullGCALotStart=100000000
 
 
-
+// GHASH block processing
+const TypeFunc* OptoRuntime::ghash_processBlocks_Type() {
+    int argcnt = 4;
+
+    const Type** fields = TypeTuple::fields(argcnt);
+    int argp = TypeFunc::Parms;
+    fields[argp++] = TypePtr::NOTNULL;    // state
+    fields[argp++] = TypePtr::NOTNULL;    // subkeyH
+    fields[argp++] = TypePtr::NOTNULL;    // data
+    fields[argp++] = TypeInt::INT;        // blocks
+    assert(argp == TypeFunc::Parms+argcnt, "correct decoding");
+    const TypeTuple* domain = TypeTuple::make(TypeFunc::Parms+argcnt, fields);
+
+    // result type needed
+    fields = TypeTuple::fields(1);
+    fields[TypeFunc::Parms+0] = NULL; // void
+    const TypeTuple* range = TypeTuple::make(TypeFunc::Parms, fields);
+    return TypeFunc::make(domain, range);
+}
 
 // Compiled code entry points
 address OptoRuntime::_new_instance_Java                           = NULL;
diff --git a/src/share/vm/opto/runtime.hpp b/src/share/vm/opto/runtime.hpp
index 99f03f408..58c6bd501 100644
--- a/src/share/vm/opto/runtime.hpp
+++ b/src/share/vm/opto/runtime.hpp
@@ -311,6 +311,8 @@ private:
   static const TypeFunc* montgomeryMultiply_Type();
   static const TypeFunc* montgomerySquare_Type();
 
+  static const TypeFunc* ghash_processBlocks_Type();
+
   static const TypeFunc* updateBytesCRC32_Type();
 
   // leaf on stack replacement interpreter accessor types
diff --git a/src/share/vm/opto/superword.cpp b/src/share/vm/opto/superword.cpp
index 718f077d0..78d9743ec 100644
--- a/src/share/vm/opto/superword.cpp
+++ b/src/share/vm/opto/superword.cpp
@@ -448,6 +448,7 @@ bool SuperWord::ref_is_alignable(SWPointer& p) {
     return true;   // no induction variable
   }
   CountedLoopEndNode* pre_end = get_pre_loop_end(lp()->as_CountedLoop());
+  assert(pre_end != NULL, "we must have a correct pre-loop");
   assert(pre_end->stride_is_con(), "pre loop stride is constant");
   int preloop_stride = pre_end->stride_con();
 
@@ -2052,7 +2053,7 @@ void SuperWord::align_initial_loop_index(MemNode* align_to_ref) {
   CountedLoopNode *main_head = lp()->as_CountedLoop();
   assert(main_head->is_main_loop(), "");
   CountedLoopEndNode* pre_end = get_pre_loop_end(main_head);
-  assert(pre_end != NULL, "");
+  assert(pre_end != NULL, "we must have a correct pre-loop");
   Node *pre_opaq1 = pre_end->limit();
   assert(pre_opaq1->Opcode() == Op_Opaque1, "");
   Opaque1Node *pre_opaq = (Opaque1Node*)pre_opaq1;
@@ -2207,16 +2208,27 @@ void SuperWord::align_initial_loop_index(MemNode* align_to_ref) {
 
 //----------------------------get_pre_loop_end---------------------------
 // Find pre loop end from main loop.  Returns null if none.
-CountedLoopEndNode* SuperWord::get_pre_loop_end(CountedLoopNode *cl) {
-  Node *ctrl = cl->in(LoopNode::EntryControl);
+CountedLoopEndNode* SuperWord::get_pre_loop_end(CountedLoopNode* cl) {
+  Node* ctrl = cl->in(LoopNode::EntryControl);
   if (!ctrl->is_IfTrue() && !ctrl->is_IfFalse()) return NULL;
-  Node *iffm = ctrl->in(0);
+  Node* iffm = ctrl->in(0);
   if (!iffm->is_If()) return NULL;
-  Node *p_f = iffm->in(0);
+  Node* bolzm = iffm->in(1);
+  if (!bolzm->is_Bool()) return NULL;
+  Node* cmpzm = bolzm->in(1);
+  if (!cmpzm->is_Cmp()) return NULL;
+  Node* opqzm = cmpzm->in(2);
+  // Can not optimize a loop if zero-trip Opaque1 node is optimized
+  // away and then another round of loop opts attempted.
+  if (opqzm->Opcode() != Op_Opaque1) {
+    return NULL;
+  }
+  Node* p_f = iffm->in(0);
   if (!p_f->is_IfFalse()) return NULL;
   if (!p_f->in(0)->is_CountedLoopEnd()) return NULL;
-  CountedLoopEndNode *pre_end = p_f->in(0)->as_CountedLoopEnd();
-  if (!pre_end->loopnode()->is_pre_loop()) return NULL;
+  CountedLoopEndNode* pre_end = p_f->in(0)->as_CountedLoopEnd();
+  CountedLoopNode* loop_node = pre_end->loopnode();
+  if (loop_node == NULL || !loop_node->is_pre_loop()) return NULL;
   return pre_end;
 }
 
diff --git a/src/share/vm/runtime/globals.hpp b/src/share/vm/runtime/globals.hpp
index 421a3ba13..0fdf47ca9 100644
--- a/src/share/vm/runtime/globals.hpp
+++ b/src/share/vm/runtime/globals.hpp
@@ -602,6 +602,9 @@ class CommandLineFlags {
   product(bool, UseSHA, false,                                              \
           "Control whether SHA instructions can be used on SPARC")          \
                                                                             \
+  product(bool, UseGHASHIntrinsics, false,                                  \
+          "Use intrinsics for GHASH versions of crypto")                    \
+                                                                            \
   product(uintx, LargePageSizeInBytes, 0,                                   \
           "Large page size (0 to let VM choose the page size)")             \
                                                                             \
diff --git a/src/share/vm/runtime/safepoint.cpp b/src/share/vm/runtime/safepoint.cpp
index be4bb457a..0e82c8b73 100644
--- a/src/share/vm/runtime/safepoint.cpp
+++ b/src/share/vm/runtime/safepoint.cpp
@@ -537,6 +537,7 @@ void SafepointSynchronize::do_cleanup_tasks() {
 
   // rotate log files?
   if (UseGCLogFileRotation) {
+    TraceTime t8("rotating gc logs", TraceSafepointCleanupTime);
     gclog_or_tty->rotate_log(false);
   }
 
diff --git a/src/share/vm/runtime/stubRoutines.cpp b/src/share/vm/runtime/stubRoutines.cpp
index 6b8f7e754..d943248da 100644
--- a/src/share/vm/runtime/stubRoutines.cpp
+++ b/src/share/vm/runtime/stubRoutines.cpp
@@ -124,6 +124,7 @@ address StubRoutines::_aescrypt_encryptBlock               = NULL;
 address StubRoutines::_aescrypt_decryptBlock               = NULL;
 address StubRoutines::_cipherBlockChaining_encryptAESCrypt = NULL;
 address StubRoutines::_cipherBlockChaining_decryptAESCrypt = NULL;
+address StubRoutines::_ghash_processBlocks                 = NULL;
 
 address StubRoutines::_sha1_implCompress     = NULL;
 address StubRoutines::_sha1_implCompressMB   = NULL;
@@ -176,7 +177,7 @@ void StubRoutines::initialize1() {
     StubGenerator_generate(&buffer, false);
     // When new stubs added we need to make sure there is some space left
     // to catch situation when we should increase size again.
-    assert(buffer.insts_remaining() > 200, "increase code_size1");
+    assert(code_size1 == 0 || buffer.insts_remaining() > 200, "increase code_size1");
   }
 }
 
@@ -231,7 +232,7 @@ void StubRoutines::initialize2() {
     StubGenerator_generate(&buffer, true);
     // When new stubs added we need to make sure there is some space left
     // to catch situation when we should increase size again.
-    assert(buffer.insts_remaining() > 200, "increase code_size2");
+    assert(code_size2 == 0 || buffer.insts_remaining() > 200, "increase code_size2");
   }
 
 #ifdef ASSERT
diff --git a/src/share/vm/runtime/stubRoutines.hpp b/src/share/vm/runtime/stubRoutines.hpp
index 42808a4c6..0f6641c17 100644
--- a/src/share/vm/runtime/stubRoutines.hpp
+++ b/src/share/vm/runtime/stubRoutines.hpp
@@ -197,6 +197,7 @@ class StubRoutines: AllStatic {
   static address _aescrypt_decryptBlock;
   static address _cipherBlockChaining_encryptAESCrypt;
   static address _cipherBlockChaining_decryptAESCrypt;
+  static address _ghash_processBlocks;
 
   static address _sha1_implCompress;
   static address _sha1_implCompressMB;
@@ -359,6 +360,7 @@ class StubRoutines: AllStatic {
   static address aescrypt_decryptBlock()                { return _aescrypt_decryptBlock; }
   static address cipherBlockChaining_encryptAESCrypt()  { return _cipherBlockChaining_encryptAESCrypt; }
   static address cipherBlockChaining_decryptAESCrypt()  { return _cipherBlockChaining_decryptAESCrypt; }
+  static address ghash_processBlocks() { return _ghash_processBlocks; }
 
   static address sha1_implCompress()     { return _sha1_implCompress; }
   static address sha1_implCompressMB()   { return _sha1_implCompressMB; }
diff --git a/src/share/vm/runtime/vmStructs.cpp b/src/share/vm/runtime/vmStructs.cpp
index 161a4c401..7f6f84887 100644
--- a/src/share/vm/runtime/vmStructs.cpp
+++ b/src/share/vm/runtime/vmStructs.cpp
@@ -810,6 +810,7 @@ typedef TwoOopHashtable<Symbol*, mtClass>     SymbolTwoOopHashtable;
      static_field(StubRoutines,                _aescrypt_decryptBlock,                        address)                               \
      static_field(StubRoutines,                _cipherBlockChaining_encryptAESCrypt,          address)                               \
      static_field(StubRoutines,                _cipherBlockChaining_decryptAESCrypt,          address)                               \
+     static_field(StubRoutines,                _ghash_processBlocks,                          address)                               \
      static_field(StubRoutines,                _updateBytesCRC32,                             address)                               \
      static_field(StubRoutines,                _crc_table_adr,                                address)                               \
      static_field(StubRoutines,                _multiplyToLen,                                address)                               \
diff --git a/src/share/vm/services/allocationSite.hpp b/src/share/vm/services/allocationSite.hpp
index 07f8595f8..d747c1b72 100644
--- a/src/share/vm/services/allocationSite.hpp
+++ b/src/share/vm/services/allocationSite.hpp
@@ -34,8 +34,9 @@ template <class E> class AllocationSite VALUE_OBJ_CLASS_SPEC {
  private:
   NativeCallStack  _call_stack;
   E                e;
+  MEMFLAGS         _flag;
  public:
-  AllocationSite(const NativeCallStack& stack) : _call_stack(stack) { }
+  AllocationSite(const NativeCallStack& stack, MEMFLAGS flag) : _call_stack(stack), _flag(flag) { }
   int hash() const { return _call_stack.hash(); }
   bool equals(const NativeCallStack& stack) const {
     return _call_stack.equals(stack);
@@ -52,6 +53,8 @@ template <class E> class AllocationSite VALUE_OBJ_CLASS_SPEC {
   // Information regarding this allocation
   E* data()             { return &e; }
   const E* peek() const { return &e; }
+
+  MEMFLAGS flag() const { return _flag; }
 };
 
 #endif  // SHARE_VM_SERVICES_ALLOCATION_SITE_HPP
diff --git a/src/share/vm/services/mallocSiteTable.cpp b/src/share/vm/services/mallocSiteTable.cpp
index 94fe1b96f..cb80b3e18 100644
--- a/src/share/vm/services/mallocSiteTable.cpp
+++ b/src/share/vm/services/mallocSiteTable.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2014, 2017, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2014, 2019, Oracle and/or its affiliates. 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
@@ -84,12 +84,18 @@ bool MallocSiteTable::initialize() {
   // Create pseudo call stack for hashtable entry allocation
   address pc[3];
   if (NMT_TrackingStackDepth >= 3) {
-    pc[2] = (address)MallocSiteTable::allocation_at;
+    uintx *fp = (uintx*)MallocSiteTable::allocation_at;
+    // On ppc64, 'fp' is a pointer to a function descriptor which is a struct  of
+    // three native pointers where the first pointer is the real function address.
+    // See: http://refspecs.linuxfoundation.org/ELF/ppc64/PPC-elf64abi-1.9.html#FUNC-DES
+    pc[2] = (address)(fp PPC64_ONLY(BIG_ENDIAN_ONLY([0])));
   }
   if (NMT_TrackingStackDepth >= 2) {
-    pc[1] = (address)MallocSiteTable::lookup_or_add;
+    uintx *fp = (uintx*)MallocSiteTable::lookup_or_add;
+    pc[1] = (address)(fp PPC64_ONLY(BIG_ENDIAN_ONLY([0])));
   }
-  pc[0] = (address)MallocSiteTable::new_entry;
+  uintx *fp = (uintx*)MallocSiteTable::new_entry;
+  pc[0] = (address)(fp PPC64_ONLY(BIG_ENDIAN_ONLY([0])));
 
   // Instantiate NativeCallStack object, have to use placement new operator. (see comments above)
   NativeCallStack* stack = ::new ((void*)_hash_entry_allocation_stack)
@@ -158,7 +164,7 @@ MallocSite* MallocSiteTable::lookup_or_add(const NativeCallStack& key, size_t* b
   MallocSiteHashtableEntry* head = _table[index];
   while (head != NULL && (*pos_idx) <= MAX_BUCKET_LENGTH) {
     MallocSite* site = head->data();
-    if (site->flags() == flags && site->equals(key)) {
+    if (site->flag() == flags && site->equals(key)) {
       return head->data();
     }
 
diff --git a/src/share/vm/services/mallocSiteTable.hpp b/src/share/vm/services/mallocSiteTable.hpp
index b2e34affc..819f4e737 100644
--- a/src/share/vm/services/mallocSiteTable.hpp
+++ b/src/share/vm/services/mallocSiteTable.hpp
@@ -37,15 +37,12 @@
 // MallocSite represents a code path that eventually calls
 // os::malloc() to allocate memory
 class MallocSite : public AllocationSite<MemoryCounter> {
- private:
-  MEMFLAGS _flags;
-
  public:
   MallocSite() :
-    AllocationSite<MemoryCounter>(NativeCallStack::empty_stack()), _flags(mtNone) {}
+    AllocationSite<MemoryCounter>(NativeCallStack::empty_stack(), mtNone) {}
 
   MallocSite(const NativeCallStack& stack, MEMFLAGS flags) :
-    AllocationSite<MemoryCounter>(stack), _flags(flags) {}
+    AllocationSite<MemoryCounter>(stack, flags) {}
 
 
   void allocate(size_t size)      { data()->allocate(size);   }
@@ -55,7 +52,6 @@ class MallocSite : public AllocationSite<MemoryCounter> {
   size_t size()  const { return peek()->size(); }
   // The number of calls were made
   size_t count() const { return peek()->count(); }
-  MEMFLAGS flags() const { return (MEMFLAGS)_flags; }
 };
 
 // Malloc site hashtable entry
diff --git a/src/share/vm/services/memBaseline.cpp b/src/share/vm/services/memBaseline.cpp
index f580aada2..1856749a7 100644
--- a/src/share/vm/services/memBaseline.cpp
+++ b/src/share/vm/services/memBaseline.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2012, 2017, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2012, 2019, Oracle and/or its affiliates. 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
@@ -63,7 +63,7 @@ int compare_malloc_site(const MallocSite& s1, const MallocSite& s2) {
 int compare_malloc_site_and_type(const MallocSite& s1, const MallocSite& s2) {
   int res = compare_malloc_site(s1, s2);
   if (res == 0) {
-    res = (int)(s1.flags() - s2.flags());
+    res = (int)(s1.flag() - s2.flag());
   }
 
   return res;
@@ -209,7 +209,7 @@ bool MemBaseline::aggregate_virtual_memory_allocation_sites() {
   const ReservedMemoryRegion* rgn;
   VirtualMemoryAllocationSite* site;
   while ((rgn = itr.next()) != NULL) {
-    VirtualMemoryAllocationSite tmp(*rgn->call_stack());
+    VirtualMemoryAllocationSite tmp(*rgn->call_stack(), rgn->flag());
     site = allocation_sites.find(tmp);
     if (site == NULL) {
       LinkedListNode<VirtualMemoryAllocationSite>* node =
diff --git a/src/share/vm/services/memReporter.cpp b/src/share/vm/services/memReporter.cpp
index 67e3eea09..a324890d3 100644
--- a/src/share/vm/services/memReporter.cpp
+++ b/src/share/vm/services/memReporter.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2012, 2017, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2012, 2019, Oracle and/or its affiliates. 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
@@ -205,7 +205,7 @@ void MemDetailReporter::report_malloc_sites() {
     const NativeCallStack* stack = malloc_site->call_stack();
     stack->print_on(out);
     out->print("%29s", " ");
-    MEMFLAGS flag = malloc_site->flags();
+    MEMFLAGS flag = malloc_site->flag();
     assert((flag >= 0 && flag < (int)mt_number_of_types) && flag != mtNone,
       "Must have a valid memory type");
     print_malloc(malloc_site->size(), malloc_site->count(),flag);
@@ -231,6 +231,10 @@ void MemDetailReporter::report_virtual_memory_allocation_sites()  {
     stack->print_on(out);
     out->print("%28s (", " ");
     print_total(virtual_memory_site->reserved(), virtual_memory_site->committed());
+    MEMFLAGS flag = virtual_memory_site->flag();
+    if (flag != mtNone) {
+      out->print(" Type=%s", NMTUtil::flag_to_name(flag));
+    }
     out->print_cr(")\n");
   }
 }
@@ -562,24 +566,24 @@ void MemDetailDiffReporter::diff_virtual_memory_sites() const {
 
 void MemDetailDiffReporter::new_malloc_site(const MallocSite* malloc_site) const {
   diff_malloc_site(malloc_site->call_stack(), malloc_site->size(), malloc_site->count(),
-    0, 0, malloc_site->flags());
+    0, 0, malloc_site->flag());
 }
 
 void MemDetailDiffReporter::old_malloc_site(const MallocSite* malloc_site) const {
   diff_malloc_site(malloc_site->call_stack(), 0, 0, malloc_site->size(),
-    malloc_site->count(), malloc_site->flags());
+    malloc_site->count(), malloc_site->flag());
 }
 
 void MemDetailDiffReporter::diff_malloc_site(const MallocSite* early,
   const MallocSite* current)  const {
-  if (early->flags() != current->flags()) {
+  if (early->flag() != current->flag()) {
     // If malloc site type changed, treat it as deallocation of old type and
     // allocation of new type.
     old_malloc_site(early);
     new_malloc_site(current);
   } else {
     diff_malloc_site(current->call_stack(), current->size(), current->count(),
-      early->size(), early->count(), early->flags());
+      early->size(), early->count(), early->flag());
   }
 }
 
@@ -603,21 +607,22 @@ void MemDetailDiffReporter::diff_malloc_site(const NativeCallStack* stack, size_
 
 
 void MemDetailDiffReporter::new_virtual_memory_site(const VirtualMemoryAllocationSite* site) const {
-  diff_virtual_memory_site(site->call_stack(), site->reserved(), site->committed(), 0, 0);
+  diff_virtual_memory_site(site->call_stack(), site->reserved(), site->committed(), 0, 0, site->flag());
 }
 
 void MemDetailDiffReporter::old_virtual_memory_site(const VirtualMemoryAllocationSite* site) const {
-  diff_virtual_memory_site(site->call_stack(), 0, 0, site->reserved(), site->committed());
+  diff_virtual_memory_site(site->call_stack(), 0, 0, site->reserved(), site->committed(), site->flag());
 }
 
 void MemDetailDiffReporter::diff_virtual_memory_site(const VirtualMemoryAllocationSite* early,
   const VirtualMemoryAllocationSite* current) const {
+  assert(early->flag() == current->flag(), "Should be the same");
   diff_virtual_memory_site(current->call_stack(), current->reserved(), current->committed(),
-    early->reserved(), early->committed());
+    early->reserved(), early->committed(), current->flag());
 }
 
 void MemDetailDiffReporter::diff_virtual_memory_site(const NativeCallStack* stack, size_t current_reserved,
-  size_t current_committed, size_t early_reserved, size_t early_committed) const  {
+  size_t current_committed, size_t early_reserved, size_t early_committed, MEMFLAGS flag) const  {
   outputStream* out = output();
 
   // no change
@@ -631,6 +636,10 @@ void MemDetailDiffReporter::diff_virtual_memory_site(const NativeCallStack* stac
   print_virtual_memory_diff(current_reserved, current_committed,
     early_reserved, early_committed);
 
+  if (flag != mtNone) {
+    out->print(" Type=%s", NMTUtil::flag_to_name(flag));
+  }
+
   out->print_cr(")\n");
  }
 
diff --git a/src/share/vm/services/memReporter.hpp b/src/share/vm/services/memReporter.hpp
index 9b83f8ce4..794804dc2 100644
--- a/src/share/vm/services/memReporter.hpp
+++ b/src/share/vm/services/memReporter.hpp
@@ -218,7 +218,7 @@ class MemDetailDiffReporter : public MemSummaryDiffReporter {
   void diff_malloc_site(const NativeCallStack* stack, size_t current_size,
     size_t currrent_count, size_t early_size, size_t early_count, MEMFLAGS flags) const;
   void diff_virtual_memory_site(const NativeCallStack* stack, size_t current_reserved,
-    size_t current_committed, size_t early_reserved, size_t early_committed) const;
+    size_t current_committed, size_t early_reserved, size_t early_committed, MEMFLAGS flag) const;
 };
 
 #endif // INCLUDE_NMT
diff --git a/src/share/vm/services/memoryPool.hpp b/src/share/vm/services/memoryPool.hpp
index 007366e0b..157b54e73 100644
--- a/src/share/vm/services/memoryPool.hpp
+++ b/src/share/vm/services/memoryPool.hpp
@@ -198,7 +198,7 @@ public:
                                bool support_usage_threshold);
 
   MemoryUsage get_memory_usage();
-  size_t used_in_bytes()            { return _space->used(); }
+  size_t used_in_bytes()            { return _space->used_stable(); }
 };
 #endif // INCLUDE_ALL_GCS
 
diff --git a/src/share/vm/services/virtualMemoryTracker.hpp b/src/share/vm/services/virtualMemoryTracker.hpp
index 1be43b554..16c9be7f9 100644
--- a/src/share/vm/services/virtualMemoryTracker.hpp
+++ b/src/share/vm/services/virtualMemoryTracker.hpp
@@ -69,8 +69,8 @@ class VirtualMemory VALUE_OBJ_CLASS_SPEC {
 // Virtual memory allocation site, keeps track where the virtual memory is reserved.
 class VirtualMemoryAllocationSite : public AllocationSite<VirtualMemory> {
  public:
-  VirtualMemoryAllocationSite(const NativeCallStack& stack) :
-    AllocationSite<VirtualMemory>(stack) { }
+  VirtualMemoryAllocationSite(const NativeCallStack& stack, MEMFLAGS flag) :
+    AllocationSite<VirtualMemory>(stack, flag) { }
 
   inline void reserve_memory(size_t sz)  { data()->reserve_memory(sz);  }
   inline void commit_memory (size_t sz)  { data()->commit_memory(sz);   }
diff --git a/src/share/vm/utilities/macros.hpp b/src/share/vm/utilities/macros.hpp
index da206d398..1f1c4df2a 100644
--- a/src/share/vm/utilities/macros.hpp
+++ b/src/share/vm/utilities/macros.hpp
@@ -416,6 +416,14 @@
 #define NOT_EMBEDDED(code) code
 #endif
 
+#ifdef VM_LITTLE_ENDIAN
+#define LITTLE_ENDIAN_ONLY(code) code
+#define BIG_ENDIAN_ONLY(code)
+#else
+#define LITTLE_ENDIAN_ONLY(code)
+#define BIG_ENDIAN_ONLY(code) code
+#endif
+
 #define define_pd_global(type, name, value) const type pd_##name = value;
 
 #endif // SHARE_VM_UTILITIES_MACROS_HPP
diff --git a/src/share/vm/utilities/taskqueue.hpp b/src/share/vm/utilities/taskqueue.hpp
index d2b39c943..0a2af4cdb 100644
--- a/src/share/vm/utilities/taskqueue.hpp
+++ b/src/share/vm/utilities/taskqueue.hpp
@@ -714,6 +714,11 @@ GenericTaskQueue<E, F, N>::pop_local(volatile E& t) {
   } else {
     // Otherwise, the queue contained exactly one element; we take the slow
     // path.
+
+    // The barrier is required to prevent reordering the two reads of _age:
+    // one is the _age.get() below, and the other is _age.top() above the if-stmt.
+    // The algorithm may fail if _age.get() reads an older value than _age.top().
+    OrderAccess::loadload();
     return pop_local_slow(localBot, _age.get());
   }
 }
diff --git a/src/share/vm/utilities/workgroup.cpp b/src/share/vm/utilities/workgroup.cpp
index 3d1f1eef7..08b97aa77 100644
--- a/src/share/vm/utilities/workgroup.cpp
+++ b/src/share/vm/utilities/workgroup.cpp
@@ -464,7 +464,6 @@ bool SubTasksDone::is_task_claimed(uint t) {
   if (old == 0) {
     old = Atomic::cmpxchg(1, &_tasks[t], 0);
   }
-  assert(_tasks[t] == 1, "What else?");
   bool res = old != 0;
 #ifdef ASSERT
   if (!res) {
diff --git a/test/compiler/7184394/TestAESBase.java b/test/compiler/7184394/TestAESBase.java
index 4d3204880..5c3e6881e 100644
--- a/test/compiler/7184394/TestAESBase.java
+++ b/test/compiler/7184394/TestAESBase.java
@@ -29,6 +29,7 @@
 import javax.crypto.Cipher;
 import javax.crypto.KeyGenerator;
 import javax.crypto.SecretKey;
+import javax.crypto.spec.GCMParameterSpec;
 import javax.crypto.spec.IvParameterSpec;
 import javax.crypto.spec.SecretKeySpec;
 import java.security.AlgorithmParameters;
@@ -62,8 +63,12 @@ abstract public class TestAESBase {
   Random random = new Random(0);
   Cipher cipher;
   Cipher dCipher;
-  AlgorithmParameters algParams;
+  AlgorithmParameters algParams = null;
   SecretKey key;
+  GCMParameterSpec gcm_spec;
+  byte[] aad = { 0x11, 0x22, 0x33, 0x44, 0x55 };
+  int tlen = 12;
+  byte[] iv = new byte[16];
 
   static int numThreads = 0;
   int  threadId;
@@ -77,7 +82,10 @@ abstract public class TestAESBase {
 
   public void prepare() {
     try {
-    System.out.println("\nalgorithm=" + algorithm + ", mode=" + mode + ", paddingStr=" + paddingStr + ", msgSize=" + msgSize + ", keySize=" + keySize + ", noReinit=" + noReinit + ", checkOutput=" + checkOutput + ", encInputOffset=" + encInputOffset + ", encOutputOffset=" + encOutputOffset + ", decOutputOffset=" + decOutputOffset + ", lastChunkSize=" +lastChunkSize );
+      System.out.println("\nalgorithm=" + algorithm + ", mode=" + mode + ", paddingStr=" + paddingStr +
+              ", msgSize=" + msgSize + ", keySize=" + keySize + ", noReinit=" + noReinit +
+              ", checkOutput=" + checkOutput + ", encInputOffset=" + encInputOffset + ", encOutputOffset=" +
+              encOutputOffset + ", decOutputOffset=" + decOutputOffset + ", lastChunkSize=" +lastChunkSize );
 
       if (encInputOffset % ALIGN != 0 || encOutputOffset % ALIGN != 0 || decOutputOffset % ALIGN !=0 )
         testingMisalignment = true;
@@ -98,16 +106,24 @@ abstract public class TestAESBase {
       cipher = Cipher.getInstance(algorithm + "/" + mode + "/" + paddingStr, "SunJCE");
       dCipher = Cipher.getInstance(algorithm + "/" + mode + "/" + paddingStr, "SunJCE");
 
+      // CBC init
       if (mode.equals("CBC")) {
-        int ivLen = (algorithm.equals("AES") ? 16 : algorithm.equals("DES") ? 8 : 0);
-        IvParameterSpec initVector = new IvParameterSpec(new byte[ivLen]);
+        IvParameterSpec initVector = new IvParameterSpec(iv);
         cipher.init(Cipher.ENCRYPT_MODE, key, initVector);
-      } else {
         algParams = cipher.getParameters();
+        dCipher.init(Cipher.DECRYPT_MODE, key, initVector);
+
+      // GCM init
+      } else if (mode.equals("GCM")) {
+        gcm_init(true);
+        gcm_init(false);
+
+      // ECB init
+      } else {
         cipher.init(Cipher.ENCRYPT_MODE, key, algParams);
+        dCipher.init(Cipher.DECRYPT_MODE, key, algParams);
       }
-      algParams = cipher.getParameters();
-      dCipher.init(Cipher.DECRYPT_MODE, key, algParams);
+
       if (threadId == 0) {
         childShowCipher();
       }
@@ -188,4 +204,19 @@ abstract public class TestAESBase {
   }
 
   abstract void childShowCipher();
+
+  void gcm_init(boolean encrypt) throws Exception {
+    gcm_spec = new GCMParameterSpec(tlen * 8, iv);
+    if (encrypt) {
+      // Get a new instance everytime because of reuse IV restrictions
+      cipher = Cipher.getInstance(algorithm + "/" + mode + "/" + paddingStr, "SunJCE");
+      cipher.init(Cipher.ENCRYPT_MODE, key, gcm_spec);
+      cipher.updateAAD(aad);
+    } else {
+      dCipher.init(Cipher.DECRYPT_MODE, key, gcm_spec);
+      dCipher.updateAAD(aad);
+
+
+    }
+  }
 }
diff --git a/test/compiler/7184394/TestAESDecode.java b/test/compiler/7184394/TestAESDecode.java
index 21f1f5559..e90ef767e 100644
--- a/test/compiler/7184394/TestAESDecode.java
+++ b/test/compiler/7184394/TestAESDecode.java
@@ -32,7 +32,11 @@ public class TestAESDecode extends TestAESBase {
   @Override
   public void run() {
     try {
-      if (!noReinit) dCipher.init(Cipher.DECRYPT_MODE, key, algParams);
+      if (mode.equals("GCM")) {
+        gcm_init(false);
+      } else if (!noReinit) {
+        dCipher.init(Cipher.DECRYPT_MODE, key, algParams);
+      }
       decode = new byte[decodeLength];
       if (testingMisalignment) {
         int tempSize = dCipher.update(encode, encOutputOffset, (decodeMsgSize - lastChunkSize), decode, decOutputOffset);
diff --git a/test/compiler/7184394/TestAESEncode.java b/test/compiler/7184394/TestAESEncode.java
index f1a35bde0..cbfb81795 100644
--- a/test/compiler/7184394/TestAESEncode.java
+++ b/test/compiler/7184394/TestAESEncode.java
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2012, 2014, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2012, 2015, Oracle and/or its affiliates. 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
@@ -32,7 +32,11 @@ public class TestAESEncode extends TestAESBase {
   @Override
   public void run() {
     try {
-      if (!noReinit) cipher.init(Cipher.ENCRYPT_MODE, key, algParams);
+      if (mode.equals("GCM")) {
+        gcm_init(true);
+      } else if (!noReinit) {
+        cipher.init(Cipher.ENCRYPT_MODE, key, algParams);
+      }
       encode = new byte[encodeLength];
       if (testingMisalignment) {
         int tempSize = cipher.update(input, encInputOffset, (msgSize - lastChunkSize), encode, encOutputOffset);
diff --git a/test/compiler/7184394/TestAESMain.java b/test/compiler/7184394/TestAESMain.java
index 20929e8ba..ddd8eeaef 100644
--- a/test/compiler/7184394/TestAESMain.java
+++ b/test/compiler/7184394/TestAESMain.java
@@ -41,6 +41,13 @@
  * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=ECB -DencInputOffset=1 -DencOutputOffset=1 TestAESMain
  * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=ECB -DencInputOffset=1 -DencOutputOffset=1 -DdecOutputOffset=1 TestAESMain
  * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=ECB -DencInputOffset=1 -DencOutputOffset=1 -DdecOutputOffset=1 -DpaddingStr=NoPadding -DmsgSize=640 TestAESMain
+ * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=GCM TestAESMain
+ * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=GCM -DencInputOffset=1 TestAESMain
+ * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=GCM -DencOutputOffset=1 TestAESMain
+ * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=GCM -DdecOutputOffset=1 TestAESMain
+ * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=GCM -DencInputOffset=1 -DencOutputOffset=1 TestAESMain
+ * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=GCM -DencInputOffset=1 -DencOutputOffset=1 -DdecOutputOffset=1 TestAESMain
+ * @run main/othervm/timeout=600 -Xbatch -DcheckOutput=true -Dmode=GCM -DencInputOffset=1 -DencOutputOffset=1 -DdecOutputOffset=1 -DpaddingStr=NoPadding -DmsgSize=640 TestAESMain
  *
  * @author Tom Deneau
  */
diff --git a/test/compiler/classUnloading/anonymousClass/TestAnonymousClassUnloading.java b/test/compiler/classUnloading/anonymousClass/TestAnonymousClassUnloading.java
index 6fbaf9cef..9ff45492f 100644
--- a/test/compiler/classUnloading/anonymousClass/TestAnonymousClassUnloading.java
+++ b/test/compiler/classUnloading/anonymousClass/TestAnonymousClassUnloading.java
@@ -22,9 +22,10 @@
  */
 
 import sun.hotspot.WhiteBox;
-import sun.misc.Unsafe;
 import sun.misc.IOUtils;
+import sun.misc.Unsafe;
 
+import java.io.IOException;
 import java.lang.reflect.Method;
 import java.net.URL;
 import java.net.URLConnection;
@@ -108,7 +109,13 @@ public class TestAnonymousClassUnloading {
         // (1) Load an anonymous version of this class using the corresponding Unsafe method
         URL classUrl = TestAnonymousClassUnloading.class.getResource("TestAnonymousClassUnloading.class");
         URLConnection connection = classUrl.openConnection();
-        byte[] classBytes = IOUtils.readFully(connection.getInputStream(), connection.getContentLength(), true);
+
+        int length = connection.getContentLength();
+        byte[] classBytes = IOUtils.readAllBytes(connection.getInputStream());
+        if (length != -1 && classBytes.length != length) {
+            throw new IOException("Expected:" + length + ", actual: " + classBytes.length);
+        }
+
         Class<?> anonymousClass = UNSAFE.defineAnonymousClass(TestAnonymousClassUnloading.class, classBytes, null);
 
         // (2) Make sure all paths of doWork are profiled and compiled
diff --git a/test/compiler/loopopts/StrangeControl.jasm b/test/compiler/loopopts/StrangeControl.jasm
new file mode 100644
index 000000000..c36217ddb
--- /dev/null
+++ b/test/compiler/loopopts/StrangeControl.jasm
@@ -0,0 +1,48 @@
+/*
+ * Copyright (c) 2019, Oracle and/or its affiliates. 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 Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ *
+ */
+
+super public class compiler/loopopts/StrangeControl
+  version 51:0
+{
+
+static Field field:"I";
+
+public static Method test:"(I)V"
+  stack 2 locals 2
+{
+    iconst_0;
+    istore 1;
+  L1: stack_frame_type append;
+    locals_map int;
+    iinc  1, 1;
+    iload 1;
+    iconst_2;
+    if_icmple L1;
+  L2: stack_frame_type same;
+    iload_0;
+    putstatic Field field:"I";
+    goto L1;
+}
+
+} // end Class StrangeControl
diff --git a/test/compiler/loopopts/TestRemoveEmptyLoop.java b/test/compiler/loopopts/TestRemoveEmptyLoop.java
new file mode 100644
index 000000000..d9828e56f
--- /dev/null
+++ b/test/compiler/loopopts/TestRemoveEmptyLoop.java
@@ -0,0 +1,53 @@
+/*
+ * Copyright (c) 2019, Huawei Technologies Co. Ltd. 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 Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ */
+
+/**
+ * @test
+ * @bug 8231988
+ * @summary Unexpected test result caused by C2 IdealLoopTree::do_remove_empty_loop
+ *
+ * @run main/othervm -XX:-TieredCompilation -XX:-BackgroundCompilation
+ *      TestRemoveEmptyLoop
+ */
+
+public class TestRemoveEmptyLoop {
+
+    public void test() {
+        int i = 34;
+        for (; i > 0; i -= 11);
+        if (i < 0) {
+            // do nothing
+        } else {
+            throw new RuntimeException("Test failed.");
+        }
+    }
+
+    public static void main(String[] args) {
+        TestRemoveEmptyLoop _instance = new TestRemoveEmptyLoop();
+        for (int i = 0; i < 50000; i++) {
+            _instance.test();
+        }
+        System.out.println("Test passed.");
+    }
+
+}
diff --git a/test/compiler/loopopts/TestStrangeControl.java b/test/compiler/loopopts/TestStrangeControl.java
new file mode 100644
index 000000000..d93ed9910
--- /dev/null
+++ b/test/compiler/loopopts/TestStrangeControl.java
@@ -0,0 +1,49 @@
+/*
+ * Copyright (c) 2019, Oracle and/or its affiliates. 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 Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ *
+ */
+
+/*
+ * @test
+ * @bug 8228888
+ * @summary Test PhaseIdealLoop::has_local_phi_input() with phi input with non-dominating control.
+ * @compile StrangeControl.jasm
+ * @run main/othervm -Xbatch -XX:CompileCommand=inline,compiler.loopopts.StrangeControl::test
+ *                   compiler.loopopts.TestStrangeControl
+ */
+
+package compiler.loopopts;
+
+public class TestStrangeControl {
+
+    public static void main(String[] args) throws Exception {
+        Thread thread = new Thread() {
+            public void run() {
+                // Run this in an own thread because it's basically an endless loop
+                StrangeControl.test(42);
+            }
+        };
+        thread.start();
+        // Give thread executing strange control loop enough time to trigger OSR compilation
+        Thread.sleep(4000);
+    }
+}
diff --git a/test/compiler/loopopts/superword/TestFuzzPreLoop.java b/test/compiler/loopopts/superword/TestFuzzPreLoop.java
new file mode 100644
index 000000000..61fd78d3a
--- /dev/null
+++ b/test/compiler/loopopts/superword/TestFuzzPreLoop.java
@@ -0,0 +1,65 @@
+/*
+ * Copyright (c) 2019, Red Hat, 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 Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ */
+
+/*
+ * @test
+ * @bug 8134739 8010500
+ * @summary SEGV in SuperWord::get_pre_loop_end
+ * @run main/othervm compiler.loopopts.superword.TestFuzzPreLoop
+ */
+
+package compiler.loopopts.superword;
+
+public class TestFuzzPreLoop {
+    static Object sink;
+    short sFld = -19206;
+
+    void doTest() {
+        int[] arr = new int[400];
+
+        for (int i1 = 0; i1 < 200; i1++) {
+            for (int i2 = 0; i2 < 100; i2++) {
+                sink = new int[400];
+            }
+            arr[i1] = 0;
+        }
+
+        float f1 = 0;
+        for (int i3 = 0; i3 < 200; i3++) {
+            f1 += i3 * i3;
+        }
+        for (int i4 = 0; i4 < 200; i4++) {
+            f1 += i4 - sFld;
+        }
+
+        System.out.println(arr);
+        System.out.println(f1);
+    }
+
+    public static void main(String... args) throws Exception {
+        TestFuzzPreLoop test = new TestFuzzPreLoop();
+        for (int i = 0; i < 100; i++) {
+            test.doTest();
+        }
+    }
+}
diff --git a/test/compiler/print/TestProfileReturnTypePrinting.java b/test/compiler/print/TestProfileReturnTypePrinting.java
new file mode 100644
index 000000000..ca57139c3
--- /dev/null
+++ b/test/compiler/print/TestProfileReturnTypePrinting.java
@@ -0,0 +1,68 @@
+/*
+ * Copyright (c) 2015, Oracle and/or its affiliates. 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 Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ */
+
+/**
+ * @test
+ * @bug 8073154
+ * @build TestProfileReturnTypePrinting
+ * @run main/othervm -XX:TypeProfileLevel=020
+ *                   -XX:CompileOnly=TestProfileReturnTypePrinting.testMethod
+ *                   -XX:+IgnoreUnrecognizedVMOptions -XX:+PrintLIR
+ *                   TestProfileReturnTypePrinting
+ * @summary Verify that c1's LIR that contains ProfileType node could be dumped
+ *          without a crash disregard to an exact class knowledge.
+ */
+public class TestProfileReturnTypePrinting {
+    private static final int ITERATIONS = 1_000_000;
+
+    public static void main(String args[]) {
+        for (int i = 0; i < ITERATIONS; i++) {
+            TestProfileReturnTypePrinting.testMethod(i);
+        }
+    }
+
+    private static int testMethod(int i) {
+        return TestProfileReturnTypePrinting.foo().hashCode()
+                + TestProfileReturnTypePrinting.bar(i).hashCode();
+    }
+
+    /* Exact class of returned value is known statically. */
+    private static B foo() {
+        return new B();
+    }
+
+    /* Exact class of returned value is not known statically. */
+    private static Object bar(int i) {
+        if (i % 2 == 0) {
+            return new A();
+        } else {
+            return new B();
+        }
+    }
+
+    private static class A {
+    }
+
+    private static class B extends A {
+    }
+}
diff --git a/test/gc/stress/gclocker/TestExcessGCLockerCollections.java b/test/gc/stress/gclocker/TestExcessGCLockerCollections.java
new file mode 100644
index 000000000..7495d5b9e
--- /dev/null
+++ b/test/gc/stress/gclocker/TestExcessGCLockerCollections.java
@@ -0,0 +1,285 @@
+/*
+ * Copyright (c) 2019, Oracle and/or its affiliates. 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 Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ */
+
+package gc.stress.gclocker;
+
+// Based on Kim Barrett;s test for JDK-8048556
+
+/*
+ * @test TestExcessGCLockerCollections
+ * @key gc
+ * @bug 8048556
+ * @summary Check for GC Locker initiated GCs that immediately follow another
+ * GC and so have very little needing to be collected.
+ * @library /testlibrary
+ * @run driver/timeout=1000 gc.stress.gclocker.TestExcessGCLockerCollections 300 4 2
+ */
+
+import java.util.HashMap;
+import java.util.Map;
+
+import java.util.zip.Deflater;
+
+import java.util.ArrayList;
+import java.util.Arrays;
+
+import javax.management.MBeanServer;
+import javax.management.Notification;
+import javax.management.NotificationListener;
+import javax.management.openmbean.CompositeData;
+import java.lang.management.ManagementFactory;
+import java.lang.management.GarbageCollectorMXBean;
+import java.lang.management.MemoryUsage;
+import java.util.List;
+import com.sun.management.GarbageCollectionNotificationInfo;
+import com.sun.management.GcInfo;
+
+import com.oracle.java.testlibrary.Asserts;
+import com.oracle.java.testlibrary.ProcessTools;
+import com.oracle.java.testlibrary.OutputAnalyzer;
+
+class TestExcessGCLockerCollectionsStringConstants {
+    // Some constant strings used in both GC logging and error detection
+    static public final String GCLOCKER_CAUSE = "GCLocker Initiated GC";
+    static public final String USED_TOO_LOW = "TOO LOW";
+    static public final String USED_OK = "OK";
+}
+
+class TestExcessGCLockerCollectionsAux {
+    static private final int LARGE_MAP_SIZE = 64 * 1024;
+
+    static private final int MAP_ARRAY_LENGTH = 4;
+    static private final int MAP_SIZE = 1024;
+
+    static private final int BYTE_ARRAY_LENGTH = 128 * 1024;
+
+    static private void println(String str) { System.out.println(str); }
+    static private void println()           { System.out.println();    }
+
+    static private volatile boolean keepRunning = true;
+
+    static Map<Integer,String> populateMap(int size) {
+        Map<Integer,String> map = new HashMap<Integer,String>();
+        for (int i = 0; i < size; i += 1) {
+            Integer keyInt = Integer.valueOf(i);
+            String valStr = "value is [" + i + "]";
+            map.put(keyInt,valStr);
+        }
+        return map;
+    }
+
+    static private class AllocatingWorker implements Runnable {
+        private final Object[] array = new Object[MAP_ARRAY_LENGTH];
+        private int arrayIndex = 0;
+
+        private void doStep() {
+            Map<Integer,String> map = populateMap(MAP_SIZE);
+            array[arrayIndex] = map;
+            arrayIndex = (arrayIndex + 1) % MAP_ARRAY_LENGTH;
+        }
+
+        public void run() {
+            while (keepRunning) {
+                doStep();
+            }
+        }
+    }
+
+    static private class JNICriticalWorker implements Runnable {
+        private int count;
+
+        private void doStep() {
+            byte[] inputArray = new byte[BYTE_ARRAY_LENGTH];
+            for (int i = 0; i < inputArray.length; i += 1) {
+                inputArray[i] = (byte) (count + i);
+            }
+
+            Deflater deflater = new Deflater();
+            deflater.setInput(inputArray);
+            deflater.finish();
+
+            byte[] outputArray = new byte[2 * inputArray.length];
+            deflater.deflate(outputArray);
+
+            count += 1;
+        }
+
+        public void run() {
+            while (keepRunning) {
+                doStep();
+            }
+        }
+    }
+
+    static class GCNotificationListener implements NotificationListener {
+        static private final double MIN_USED_PERCENT = 40.0;
+
+        static private final List<String> newGenPoolNames = Arrays.asList(
+                "G1 Eden Space",           // OpenJDK G1GC: -XX:+UseG1GC
+                "PS Eden Space",           // OpenJDK ParallelGC: -XX:+ParallelGC
+                "Par Eden Space",          // OpenJDK ConcMarkSweepGC: -XX:+ConcMarkSweepGC
+                "Eden Space"               // OpenJDK SerialGC: -XX:+UseSerialGC
+                                           // OpenJDK ConcMarkSweepGC: -XX:+ConcMarkSweepGC -XX:-UseParNewGC
+        );
+
+        @Override
+        public void handleNotification(Notification notification, Object handback) {
+            try {
+                if (notification.getType().equals(GarbageCollectionNotificationInfo.GARBAGE_COLLECTION_NOTIFICATION)) {
+                    GarbageCollectionNotificationInfo info =
+                            GarbageCollectionNotificationInfo.from((CompositeData) notification.getUserData());
+
+                    String gc_cause = info.getGcCause();
+
+                    if (gc_cause.equals(TestExcessGCLockerCollectionsStringConstants.GCLOCKER_CAUSE)) {
+                        Map<String, MemoryUsage> memory_before_gc = info.getGcInfo().getMemoryUsageBeforeGc();
+
+                        for (String newGenPoolName : newGenPoolNames) {
+                            MemoryUsage usage = memory_before_gc.get(newGenPoolName);
+                            if (usage == null) continue;
+
+                            double startTime = ((double) info.getGcInfo().getStartTime()) / 1000.0;
+                            long used = usage.getUsed();
+                            long committed = usage.getCommitted();
+                            long max = usage.getMax();
+                            double used_percent = (((double) used) / Math.max(committed, max)) * 100.0;
+
+                            System.out.printf("%6.3f: (%s) %d/%d/%d, %8.4f%% (%s)\n",
+                                              startTime, gc_cause, used, committed, max, used_percent,
+                                              ((used_percent < MIN_USED_PERCENT) ? TestExcessGCLockerCollectionsStringConstants.USED_TOO_LOW
+                                                                                 : TestExcessGCLockerCollectionsStringConstants.USED_OK));
+                        }
+                    }
+                }
+            } catch (RuntimeException ex) {
+                System.err.println("Exception during notification processing:" + ex);
+                ex.printStackTrace();
+            }
+        }
+
+        public static boolean register() {
+            try {
+                MBeanServer mbeanServer = ManagementFactory.getPlatformMBeanServer();
+
+                // Get the list of MX
+                List<GarbageCollectorMXBean> gc_mxbeans = ManagementFactory.getGarbageCollectorMXBeans();
+
+                // Create the notification listener
+                GCNotificationListener gcNotificationListener = new GCNotificationListener();
+
+                for (GarbageCollectorMXBean gcbean : gc_mxbeans) {
+                  // Add notification listener for the MXBean
+                  mbeanServer.addNotificationListener(gcbean.getObjectName(), gcNotificationListener, null, null);
+                }
+            } catch (Exception ex) {
+                System.err.println("Exception during mbean registration:" + ex);
+                ex.printStackTrace();
+                // We've failed to set up, terminate
+                return false;
+            }
+
+            return true;
+        }
+    }
+
+    static public Map<Integer,String> largeMap;
+
+    static public void main(String args[]) {
+        long durationSec = Long.parseLong(args[0]);
+        int allocThreadNum = Integer.parseInt(args[1]);
+        int jniCriticalThreadNum = Integer.parseInt(args[2]);
+
+        println("Running for " + durationSec + " secs");
+
+        if (!GCNotificationListener.register()) {
+          println("failed to register GC notification listener");
+          System.exit(-1);
+        }
+
+        largeMap = populateMap(LARGE_MAP_SIZE);
+
+        println("Starting " + allocThreadNum + " allocating threads");
+        for (int i = 0; i < allocThreadNum; i += 1) {
+            new Thread(new AllocatingWorker()).start();
+        }
+
+        println("Starting " + jniCriticalThreadNum + " jni critical threads");
+        for (int i = 0; i < jniCriticalThreadNum; i += 1) {
+            new Thread(new JNICriticalWorker()).start();
+        }
+
+        long durationMS = (long) (1000 * durationSec);
+        long start = System.currentTimeMillis();
+        long now = start;
+        long soFar = now - start;
+        while (soFar < durationMS) {
+            try {
+                Thread.sleep(durationMS - soFar);
+            } catch (Exception e) {
+            }
+            now = System.currentTimeMillis();
+            soFar = now - start;
+        }
+        println("Done.");
+        keepRunning = false;
+    }
+}
+
+public class TestExcessGCLockerCollections {
+    private static final String USED_OK_LINE =
+        "\\(" + TestExcessGCLockerCollectionsStringConstants.GCLOCKER_CAUSE + "\\)"
+              + " .* " +
+        "\\(" + TestExcessGCLockerCollectionsStringConstants.USED_OK + "\\)";
+    private static final String USED_TOO_LOW_LINE =
+        "\\(" + TestExcessGCLockerCollectionsStringConstants.GCLOCKER_CAUSE + "\\)"
+              + " .* " +
+        "\\(" + TestExcessGCLockerCollectionsStringConstants.USED_TOO_LOW + "\\)";
+
+    private static final String[] COMMON_OPTIONS = new String[] {
+        "-Xmx1G", "-Xms1G", "-Xmn256M" };
+
+    public static void main(String args[]) throws Exception {
+        if (args.length < 3) {
+            System.out.println("usage: TestExcessGCLockerCollections" +
+                               " <duration sec> <alloc threads>" +
+                               " <jni critical threads>");
+            throw new RuntimeException("Invalid arguments");
+        }
+
+        ArrayList<String> finalArgs = new ArrayList<String>();
+        finalArgs.addAll(Arrays.asList(COMMON_OPTIONS));
+        finalArgs.add(TestExcessGCLockerCollectionsAux.class.getName());
+        finalArgs.addAll(Arrays.asList(args));
+
+        // GC and other options obtained from test framework.
+        ProcessBuilder pb = ProcessTools.createJavaProcessBuilder(
+            true, finalArgs.toArray(new String[0]));
+        OutputAnalyzer output = new OutputAnalyzer(pb.start());
+        output.shouldHaveExitValue(0);
+        //System.out.println("------------- begin stdout ----------------");
+        //System.out.println(output.getStdout());
+        //System.out.println("------------- end stdout ----------------");
+        output.stdoutShouldMatch(USED_OK_LINE);
+        output.stdoutShouldNotMatch(USED_TOO_LOW_LINE);
+    }
+}
diff --git a/test/runtime/8003720/VictimClassLoader.java b/test/runtime/8003720/VictimClassLoader.java
index 3505f7832..e219fd60f 100644
--- a/test/runtime/8003720/VictimClassLoader.java
+++ b/test/runtime/8003720/VictimClassLoader.java
@@ -22,6 +22,8 @@
  *
  */
 
+import sun.misc.IOUtils;
+
 public class VictimClassLoader extends ClassLoader {
     public static long counter = 0;
 
@@ -72,8 +74,10 @@ public class VictimClassLoader extends ClassLoader {
     }
 
     static byte[] readFully(java.io.InputStream in, int len) throws java.io.IOException {
-        // Warning here:
-        return sun.misc.IOUtils.readFully(in, len, true);
+        byte[] b = IOUtils.readAllBytes(in);
+        if (len != -1 && b.length != len)
+            throw new java.io.IOException("Expected:" + len + ", actual:" + b.length);
+        return b;
     }
 
     public void finalize() {