--- a/hotspot/src/cpu/aarch64/vm/c1_LIRAssembler_aarch64.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/cpu/aarch64/vm/c1_LIRAssembler_aarch64.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -3134,10 +3134,10 @@
       Register obj = as_reg(data);
       Register dst = as_reg(dest);
       if (is_oop && UseCompressedOops) {
-        __ encode_heap_oop(rscratch1, obj);
-        obj = rscratch1;
+        __ encode_heap_oop(rscratch2, obj);
+        obj = rscratch2;
       }
-      assert_different_registers(obj, addr.base(), tmp, rscratch2, dst);
+      assert_different_registers(obj, addr.base(), tmp, rscratch1, dst);
       __ lea(tmp, addr);
       (_masm->*xchg)(dst, obj, tmp);
       if (is_oop && UseCompressedOops) {

--- a/hotspot/src/cpu/ppc/vm/assembler_ppc.hpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/cpu/ppc/vm/assembler_ppc.hpp	Fri Apr 01 08:46:47 2016 +0000
@@ -624,6 +624,7 @@
     VNOR_OPCODE    = (4u  << OPCODE_SHIFT | 1284u     ),
     VOR_OPCODE     = (4u  << OPCODE_SHIFT | 1156u     ),
     VXOR_OPCODE    = (4u  << OPCODE_SHIFT | 1220u     ),
+    VRLD_OPCODE    = (4u  << OPCODE_SHIFT |  196u     ),
     VRLB_OPCODE    = (4u  << OPCODE_SHIFT |    4u     ),
     VRLW_OPCODE    = (4u  << OPCODE_SHIFT |  132u     ),
     VRLH_OPCODE    = (4u  << OPCODE_SHIFT |   68u     ),
@@ -2047,6 +2048,7 @@
   inline void vnor(     VectorRegister d, VectorRegister a, VectorRegister b);
   inline void vor(      VectorRegister d, VectorRegister a, VectorRegister b);
   inline void vxor(     VectorRegister d, VectorRegister a, VectorRegister b);
+  inline void vrld(     VectorRegister d, VectorRegister a, VectorRegister b);
   inline void vrlb(     VectorRegister d, VectorRegister a, VectorRegister b);
   inline void vrlw(     VectorRegister d, VectorRegister a, VectorRegister b);
   inline void vrlh(     VectorRegister d, VectorRegister a, VectorRegister b);

--- a/hotspot/src/cpu/ppc/vm/assembler_ppc.inline.hpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/cpu/ppc/vm/assembler_ppc.inline.hpp	Fri Apr 01 08:46:47 2016 +0000
@@ -839,6 +839,7 @@
 inline void Assembler::vnor(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VNOR_OPCODE     | vrt(d) | vra(a) | vrb(b)); }
 inline void Assembler::vor(     VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VOR_OPCODE      | vrt(d) | vra(a) | vrb(b)); }
 inline void Assembler::vxor(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VXOR_OPCODE     | vrt(d) | vra(a) | vrb(b)); }
+inline void Assembler::vrld(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VRLD_OPCODE     | vrt(d) | vra(a) | vrb(b)); }
 inline void Assembler::vrlb(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VRLB_OPCODE     | vrt(d) | vra(a) | vrb(b)); }
 inline void Assembler::vrlw(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VRLW_OPCODE     | vrt(d) | vra(a) | vrb(b)); }
 inline void Assembler::vrlh(    VectorRegister d, VectorRegister a, VectorRegister b) { emit_int32( VRLH_OPCODE     | vrt(d) | vra(a) | vrb(b)); }

--- a/hotspot/src/cpu/ppc/vm/stubGenerator_ppc.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/cpu/ppc/vm/stubGenerator_ppc.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -2417,6 +2417,433 @@
     return start;
   }
 
+  // Arguments for generated stub (little endian only):
+  //   R3_ARG1   - source byte array address
+  //   R4_ARG2   - destination byte array address
+  //   R5_ARG3   - round key array
+  address generate_aescrypt_encryptBlock() {
+    assert(UseAES, "need AES instructions and misaligned SSE support");
+    StubCodeMark mark(this, "StubRoutines", "aescrypt_encryptBlock");
+
+    address start = __ function_entry();
+
+    Label L_doLast;
+
+    Register from           = R3_ARG1;  // source array address
+    Register to             = R4_ARG2;  // destination array address
+    Register key            = R5_ARG3;  // round key array
+
+    Register keylen         = R8;
+    Register temp           = R9;
+    Register keypos         = R10;
+    Register hex            = R11;
+    Register fifteen        = R12;
+
+    VectorRegister vRet     = VR0;
+
+    VectorRegister vKey1    = VR1;
+    VectorRegister vKey2    = VR2;
+    VectorRegister vKey3    = VR3;
+    VectorRegister vKey4    = VR4;
+
+    VectorRegister fromPerm = VR5;
+    VectorRegister keyPerm  = VR6;
+    VectorRegister toPerm   = VR7;
+    VectorRegister fSplt    = VR8;
+
+    VectorRegister vTmp1    = VR9;
+    VectorRegister vTmp2    = VR10;
+    VectorRegister vTmp3    = VR11;
+    VectorRegister vTmp4    = VR12;
+
+    VectorRegister vLow     = VR13;
+    VectorRegister vHigh    = VR14;
+
+    __ li              (hex, 16);
+    __ li              (fifteen, 15);
+    __ vspltisb        (fSplt, 0x0f);
+
+    // load unaligned from[0-15] to vsRet
+    __ lvx             (vRet, from);
+    __ lvx             (vTmp1, fifteen, from);
+    __ lvsl            (fromPerm, from);
+    __ vxor            (fromPerm, fromPerm, fSplt);
+    __ vperm           (vRet, vRet, vTmp1, fromPerm);
+
+    // load keylen (44 or 52 or 60)
+    __ lwz             (keylen, arrayOopDesc::length_offset_in_bytes() - arrayOopDesc::base_offset_in_bytes(T_INT), key);
+
+    // to load keys
+    __ lvsr            (keyPerm, key);
+    __ vxor            (vTmp2, vTmp2, vTmp2);
+    __ vspltisb        (vTmp2, -16);
+    __ vrld            (keyPerm, keyPerm, vTmp2);
+    __ vrld            (keyPerm, keyPerm, vTmp2);
+    __ vsldoi          (keyPerm, keyPerm, keyPerm, -8);
+
+    // load the 1st round key to vKey1
+    __ li              (keypos, 0);
+    __ lvx             (vKey1, keypos, key);
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey1, vTmp1, vKey1, keyPerm);
+
+    // 1st round
+    __ vxor (vRet, vRet, vKey1);
+
+    // load the 2nd round key to vKey1
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp2, vTmp1, keyPerm);
+
+    // load the 3rd round key to vKey2
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp1, vTmp2, keyPerm);
+
+    // load the 4th round key to vKey3
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey3, vTmp2, vTmp1, keyPerm);
+
+    // load the 5th round key to vKey4
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey4, vTmp1, vTmp2, keyPerm);
+
+    // 2nd - 5th rounds
+    __ vcipher (vRet, vRet, vKey1);
+    __ vcipher (vRet, vRet, vKey2);
+    __ vcipher (vRet, vRet, vKey3);
+    __ vcipher (vRet, vRet, vKey4);
+
+    // load the 6th round key to vKey1
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp2, vTmp1, keyPerm);
+
+    // load the 7th round key to vKey2
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp1, vTmp2, keyPerm);
+
+    // load the 8th round key to vKey3
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey3, vTmp2, vTmp1, keyPerm);
+
+    // load the 9th round key to vKey4
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey4, vTmp1, vTmp2, keyPerm);
+
+    // 6th - 9th rounds
+    __ vcipher (vRet, vRet, vKey1);
+    __ vcipher (vRet, vRet, vKey2);
+    __ vcipher (vRet, vRet, vKey3);
+    __ vcipher (vRet, vRet, vKey4);
+
+    // load the 10th round key to vKey1
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp2, vTmp1, keyPerm);
+
+    // load the 11th round key to vKey2
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp1, vTmp2, keyPerm);
+
+    // if all round keys are loaded, skip next 4 rounds
+    __ cmpwi           (CCR0, keylen, 44);
+    __ beq             (CCR0, L_doLast);
+
+    // 10th - 11th rounds
+    __ vcipher (vRet, vRet, vKey1);
+    __ vcipher (vRet, vRet, vKey2);
+
+    // load the 12th round key to vKey1
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp2, vTmp1, keyPerm);
+
+    // load the 13th round key to vKey2
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp1, vTmp2, keyPerm);
+
+    // if all round keys are loaded, skip next 2 rounds
+    __ cmpwi           (CCR0, keylen, 52);
+    __ beq             (CCR0, L_doLast);
+
+    // 12th - 13th rounds
+    __ vcipher (vRet, vRet, vKey1);
+    __ vcipher (vRet, vRet, vKey2);
+
+    // load the 14th round key to vKey1
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp2, vTmp1, keyPerm);
+
+    // load the 15th round key to vKey2
+    __ addi            (keypos, keypos, 16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp1, vTmp2, keyPerm);
+
+    __ bind(L_doLast);
+
+    // last two rounds
+    __ vcipher (vRet, vRet, vKey1);
+    __ vcipherlast (vRet, vRet, vKey2);
+
+    __ neg             (temp, to);
+    __ lvsr            (toPerm, temp);
+    __ vspltisb        (vTmp2, -1);
+    __ vxor            (vTmp1, vTmp1, vTmp1);
+    __ vperm           (vTmp2, vTmp2, vTmp1, toPerm);
+    __ vxor            (toPerm, toPerm, fSplt);
+    __ lvx             (vTmp1, to);
+    __ vperm           (vRet, vRet, vRet, toPerm);
+    __ vsel            (vTmp1, vTmp1, vRet, vTmp2);
+    __ lvx             (vTmp4, fifteen, to);
+    __ stvx            (vTmp1, to);
+    __ vsel            (vRet, vRet, vTmp4, vTmp2);
+    __ stvx            (vRet, fifteen, to);
+
+    __ blr();
+     return start;
+  }
+
+  // Arguments for generated stub (little endian only):
+  //   R3_ARG1   - source byte array address
+  //   R4_ARG2   - destination byte array address
+  //   R5_ARG3   - K (key) in little endian int array
+  address generate_aescrypt_decryptBlock() {
+    assert(UseAES, "need AES instructions and misaligned SSE support");
+    StubCodeMark mark(this, "StubRoutines", "aescrypt_decryptBlock");
+
+    address start = __ function_entry();
+
+    Label L_doLast;
+    Label L_do44;
+    Label L_do52;
+    Label L_do60;
+
+    Register from           = R3_ARG1;  // source array address
+    Register to             = R4_ARG2;  // destination array address
+    Register key            = R5_ARG3;  // round key array
+
+    Register keylen         = R8;
+    Register temp           = R9;
+    Register keypos         = R10;
+    Register hex            = R11;
+    Register fifteen        = R12;
+
+    VectorRegister vRet     = VR0;
+
+    VectorRegister vKey1    = VR1;
+    VectorRegister vKey2    = VR2;
+    VectorRegister vKey3    = VR3;
+    VectorRegister vKey4    = VR4;
+    VectorRegister vKey5    = VR5;
+
+    VectorRegister fromPerm = VR6;
+    VectorRegister keyPerm  = VR7;
+    VectorRegister toPerm   = VR8;
+    VectorRegister fSplt    = VR9;
+
+    VectorRegister vTmp1    = VR10;
+    VectorRegister vTmp2    = VR11;
+    VectorRegister vTmp3    = VR12;
+    VectorRegister vTmp4    = VR13;
+
+    VectorRegister vLow     = VR14;
+    VectorRegister vHigh    = VR15;
+
+    __ li              (hex, 16);
+    __ li              (fifteen, 15);
+    __ vspltisb        (fSplt, 0x0f);
+
+    // load unaligned from[0-15] to vsRet
+    __ lvx             (vRet, from);
+    __ lvx             (vTmp1, fifteen, from);
+    __ lvsl            (fromPerm, from);
+    __ vxor            (fromPerm, fromPerm, fSplt);
+    __ vperm           (vRet, vRet, vTmp1, fromPerm); // align [and byte swap in LE]
+
+    // load keylen (44 or 52 or 60)
+    __ lwz             (keylen, arrayOopDesc::length_offset_in_bytes() - arrayOopDesc::base_offset_in_bytes(T_INT), key);
+
+    // to load keys
+    __ lvsr            (keyPerm, key);
+    __ vxor            (vTmp2, vTmp2, vTmp2);
+    __ vspltisb        (vTmp2, -16);
+    __ vrld            (keyPerm, keyPerm, vTmp2);
+    __ vrld            (keyPerm, keyPerm, vTmp2);
+    __ vsldoi          (keyPerm, keyPerm, keyPerm, -8);
+
+    __ cmpwi           (CCR0, keylen, 44);
+    __ beq             (CCR0, L_do44);
+
+    __ cmpwi           (CCR0, keylen, 52);
+    __ beq             (CCR0, L_do52);
+
+    // load the 15th round key to vKey11
+    __ li              (keypos, 240);
+    __ lvx             (vTmp1, keypos, key);
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp1, vTmp2, keyPerm);
+
+    // load the 14th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp2, vTmp1, keyPerm);
+
+    // load the 13th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey3, vTmp1, vTmp2, keyPerm);
+
+    // load the 12th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey4, vTmp2, vTmp1, keyPerm);
+
+    // load the 11th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey5, vTmp1, vTmp2, keyPerm);
+
+    // 1st - 5th rounds
+    __ vxor            (vRet, vRet, vKey1);
+    __ vncipher        (vRet, vRet, vKey2);
+    __ vncipher        (vRet, vRet, vKey3);
+    __ vncipher        (vRet, vRet, vKey4);
+    __ vncipher        (vRet, vRet, vKey5);
+
+    __ b               (L_doLast);
+
+    __ bind            (L_do52);
+
+    // load the 13th round key to vKey11
+    __ li              (keypos, 208);
+    __ lvx             (vTmp1, keypos, key);
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp1, vTmp2, keyPerm);
+
+    // load the 12th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp2, vTmp1, keyPerm);
+
+    // load the 11th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey3, vTmp1, vTmp2, keyPerm);
+
+    // 1st - 3rd rounds
+    __ vxor            (vRet, vRet, vKey1);
+    __ vncipher        (vRet, vRet, vKey2);
+    __ vncipher        (vRet, vRet, vKey3);
+
+    __ b               (L_doLast);
+
+    __ bind            (L_do44);
+
+    // load the 11th round key to vKey11
+    __ li              (keypos, 176);
+    __ lvx             (vTmp1, keypos, key);
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp1, vTmp2, keyPerm);
+
+    // 1st round
+    __ vxor            (vRet, vRet, vKey1);
+
+    __ bind            (L_doLast);
+
+    // load the 10th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey1, vTmp2, vTmp1, keyPerm);
+
+    // load the 9th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey2, vTmp1, vTmp2, keyPerm);
+
+    // load the 8th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey3, vTmp2, vTmp1, keyPerm);
+
+    // load the 7th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey4, vTmp1, vTmp2, keyPerm);
+
+    // load the 6th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey5, vTmp2, vTmp1, keyPerm);
+
+    // last 10th - 6th rounds
+    __ vncipher        (vRet, vRet, vKey1);
+    __ vncipher        (vRet, vRet, vKey2);
+    __ vncipher        (vRet, vRet, vKey3);
+    __ vncipher        (vRet, vRet, vKey4);
+    __ vncipher        (vRet, vRet, vKey5);
+
+    // load the 5th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey1, vTmp1, vTmp2, keyPerm);
+
+    // load the 4th round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey2, vTmp2, vTmp1, keyPerm);
+
+    // load the 3rd round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey3, vTmp1, vTmp2, keyPerm);
+
+    // load the 2nd round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp1, keypos, key);
+    __ vperm           (vKey4, vTmp2, vTmp1, keyPerm);
+
+    // load the 1st round key to vKey10
+    __ addi            (keypos, keypos, -16);
+    __ lvx             (vTmp2, keypos, key);
+    __ vperm           (vKey5, vTmp1, vTmp2, keyPerm);
+
+    // last 5th - 1th rounds
+    __ vncipher        (vRet, vRet, vKey1);
+    __ vncipher        (vRet, vRet, vKey2);
+    __ vncipher        (vRet, vRet, vKey3);
+    __ vncipher        (vRet, vRet, vKey4);
+    __ vncipherlast    (vRet, vRet, vKey5);
+
+    __ neg             (temp, to);
+    __ lvsr            (toPerm, temp);
+    __ vspltisb        (vTmp2, -1);
+    __ vxor            (vTmp1, vTmp1, vTmp1);
+    __ vperm           (vTmp2, vTmp2, vTmp1, toPerm);
+    __ vxor            (toPerm, toPerm, fSplt);
+    __ lvx             (vTmp1, to);
+    __ vperm           (vRet, vRet, vRet, toPerm);
+    __ vsel            (vTmp1, vTmp1, vRet, vTmp2);
+    __ lvx             (vTmp4, fifteen, to);
+    __ stvx            (vTmp1, to);
+    __ vsel            (vRet, vRet, vTmp4, vTmp2);
+    __ stvx            (vRet, fifteen, to);
+
+    __ blr();
+     return start;
+  }
 
   void generate_arraycopy_stubs() {
     // Note: the disjoint stubs must be generated first, some of
@@ -2693,10 +3120,6 @@
     // arraycopy stubs used by compilers
     generate_arraycopy_stubs();
 
-    if (UseAESIntrinsics) {
-      guarantee(!UseAESIntrinsics, "not yet implemented.");
-    }
-
     // Safefetch stubs.
     generate_safefetch("SafeFetch32", sizeof(int),     &StubRoutines::_safefetch32_entry,
                                                        &StubRoutines::_safefetch32_fault_pc,
@@ -2719,6 +3142,12 @@
       StubRoutines::_montgomerySquare
         = CAST_FROM_FN_PTR(address, SharedRuntime::montgomery_square);
     }
+
+    if (UseAESIntrinsics) {
+      StubRoutines::_aescrypt_encryptBlock = generate_aescrypt_encryptBlock();
+      StubRoutines::_aescrypt_decryptBlock = generate_aescrypt_decryptBlock();
+    }
+
   }
 
  public:

--- a/hotspot/src/cpu/ppc/vm/vm_version_ppc.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/cpu/ppc/vm/vm_version_ppc.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -122,7 +122,7 @@
                (has_fcfids()  ? " fcfids"  : ""),
                (has_vand()    ? " vand"    : ""),
                (has_lqarx()   ? " lqarx"   : ""),
-               (has_vcipher() ? " vcipher" : ""),
+               (has_vcipher() ? " aes"     : ""),
                (has_vpmsumb() ? " vpmsumb" : ""),
                (has_tcheck()  ? " tcheck"  : ""),
                (has_mfdscr()  ? " mfdscr"  : "")
@@ -186,6 +186,28 @@
   }
 
   // The AES intrinsic stubs require AES instruction support.
+#if defined(VM_LITTLE_ENDIAN)
+  if (has_vcipher()) {
+    if (FLAG_IS_DEFAULT(UseAES)) {
+      UseAES = true;
+    }
+  } else if (UseAES) {
+    if (!FLAG_IS_DEFAULT(UseAES))
+      warning("AES instructions are not available on this CPU");
+    FLAG_SET_DEFAULT(UseAES, false);
+  }
+
+  if (UseAES && has_vcipher()) {
+    if (FLAG_IS_DEFAULT(UseAESIntrinsics)) {
+      UseAESIntrinsics = 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 (UseAES) {
     warning("AES instructions are not available on this CPU");
     FLAG_SET_DEFAULT(UseAES, false);
@@ -195,6 +217,7 @@
       warning("AES intrinsics are not available on this CPU");
     FLAG_SET_DEFAULT(UseAESIntrinsics, false);
   }
+#endif
 
   if (UseAESCTRIntrinsics) {
     warning("AES/CTR intrinsics are not available on this CPU");

--- a/hotspot/src/jdk.vm.ci/share/classes/jdk.vm.ci.hotspot/src/jdk/vm/ci/hotspot/HotSpotMemoryAccessProviderImpl.java	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/jdk.vm.ci/share/classes/jdk.vm.ci.hotspot/src/jdk/vm/ci/hotspot/HotSpotMemoryAccessProviderImpl.java	Fri Apr 01 08:46:47 2016 +0000
@@ -23,7 +23,6 @@
 package jdk.vm.ci.hotspot;
 
 import static jdk.vm.ci.hotspot.UnsafeAccess.UNSAFE;
-import jdk.vm.ci.common.JVMCIError;
 import jdk.vm.ci.hotspot.HotSpotVMConfig.CompressEncoding;
 import jdk.vm.ci.meta.Constant;
 import jdk.vm.ci.meta.JavaConstant;
@@ -59,7 +58,7 @@
                     return true;
                 }
             } else {
-                throw new JVMCIError("%s", metaspaceObject);
+                throw new IllegalArgumentException(String.valueOf(metaspaceObject));
             }
         }
         return false;
@@ -75,7 +74,7 @@
                 return prim.asLong();
             }
         }
-        throw new JVMCIError("%s", base);
+        throw new IllegalArgumentException(String.valueOf(base));
     }
 
     private static long readRawValue(Constant baseConstant, long displacement, int bits) {
@@ -91,7 +90,7 @@
                 case Long.SIZE:
                     return UNSAFE.getLong(base, displacement);
                 default:
-                    throw new JVMCIError("%d", bits);
+                    throw new IllegalArgumentException(String.valueOf(bits));
             }
         } else {
             long pointer = asRawPointer(baseConstant);
@@ -105,7 +104,7 @@
                 case Long.SIZE:
                     return UNSAFE.getLong(pointer + displacement);
                 default:
-                    throw new JVMCIError("%d", bits);
+                    throw new IllegalArgumentException(String.valueOf(bits));
             }
         }
     }
@@ -178,7 +177,7 @@
                 case Double:
                     return JavaConstant.forDouble(Double.longBitsToDouble(rawValue));
                 default:
-                    throw new JVMCIError("Unsupported kind: %s", kind);
+                    throw new IllegalArgumentException("Unsupported kind: " + kind);
             }
         } catch (NullPointerException e) {
             return null;

--- a/hotspot/src/jdk.vm.ci/share/classes/jdk.vm.ci.meta/src/jdk/vm/ci/meta/MemoryAccessProvider.java	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/jdk.vm.ci/share/classes/jdk.vm.ci.meta/src/jdk/vm/ci/meta/MemoryAccessProvider.java	Fri Apr 01 08:46:47 2016 +0000
@@ -35,8 +35,10 @@
      * @param displacement the displacement within the object in bytes
      * @return the read value encapsulated in a {@link JavaConstant} object, or {@code null} if the
      *         value cannot be read.
+     * @throws IllegalArgumentException if {@code kind} is {@link JavaKind#Void} or not
+     *             {@linkplain JavaKind#isPrimitive() primitive} kind
      */
-    JavaConstant readUnsafeConstant(JavaKind kind, JavaConstant base, long displacement);
+    JavaConstant readUnsafeConstant(JavaKind kind, JavaConstant base, long displacement) throws IllegalArgumentException;
 
     /**
      * Reads a primitive value using a base address and a displacement.
@@ -46,8 +48,11 @@
      * @param displacement the displacement within the object in bytes
      * @param bits the number of bits to read from memory
      * @return the read value encapsulated in a {@link JavaConstant} object of {@link JavaKind} kind
+     * @throws IllegalArgumentException if {@code kind} is {@link JavaKind#Void} or not
+     *             {@linkplain JavaKind#isPrimitive() primitive} kind or {@code bits} is not 8, 16,
+     *             32 or 64
      */
-    JavaConstant readPrimitiveConstant(JavaKind kind, Constant base, long displacement, int bits);
+    JavaConstant readPrimitiveConstant(JavaKind kind, Constant base, long displacement, int bits) throws IllegalArgumentException;
 
     /**
      * Reads a Java {@link Object} value using a base address and a displacement.

--- a/hotspot/src/jdk.vm.ci/share/classes/jdk.vm.ci.meta/src/jdk/vm/ci/meta/MethodHandleAccessProvider.java	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/jdk.vm.ci/share/classes/jdk.vm.ci.meta/src/jdk/vm/ci/meta/MethodHandleAccessProvider.java	Fri Apr 01 08:46:47 2016 +0000
@@ -51,6 +51,8 @@
     /**
      * Returns the method handle method intrinsic identifier for the provided method, or
      * {@code null} if the method is not an intrinsic processed by this interface.
+     *
+     * @throws NullPointerException if {@code method} is null
      */
     IntrinsicMethod lookupMethodHandleIntrinsic(ResolvedJavaMethod method);
 
@@ -58,19 +60,27 @@
      * Resolves the invocation target for an invocation of {@link IntrinsicMethod#INVOKE_BASIC
      * MethodHandle.invokeBasic} with the given constant receiver {@link MethodHandle}. Returns
      * {@code null} if the invocation target is not available at this time.
-     * <p>
+     *
      * The first invocations of a method handle can use an interpreter to lookup the actual invoked
      * method; frequently executed method handles can use Java bytecode generation to avoid the
      * interpreter overhead. If the parameter forceBytecodeGeneration is set to true, the VM should
      * try to generate bytecodes before this method returns.
+     *
+     * @returns {@code null} if {@code methodHandle} is not a {@link MethodHandle} or the invocation
+     *          target is not available at this time
+     * @throws NullPointerException if {@code methodHandle} is null
      */
     ResolvedJavaMethod resolveInvokeBasicTarget(JavaConstant methodHandle, boolean forceBytecodeGeneration);
 
     /**
      * Resolves the invocation target for an invocation of a {@code MethodHandle.linkTo*} method
      * with the given constant member name. The member name is the last parameter of the
-     * {@code linkTo*} method. Returns {@code null} if the invocation target is not available at
-     * this time.
+     * {@code linkTo*} method.
+     *
+     * @returns {@code null} if the invocation target is not available at this time
+     * @throws NullPointerException if {@code memberName} is null
+     * @throws IllegalArgumentException if {@code memberName} is not a
+     *             {@code java.lang.invoke.MemberName}
      */
     ResolvedJavaMethod resolveLinkToTarget(JavaConstant memberName);
 }

--- a/hotspot/src/share/vm/classfile/systemDictionary.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/classfile/systemDictionary.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -2063,7 +2063,18 @@
   int  sid  = (info >> CEIL_LG_OPTION_LIMIT);
   Symbol* symbol = vmSymbols::symbol_at((vmSymbols::SID)sid);
   InstanceKlass** klassp = &_well_known_klasses[id];
-  bool must_load = (init_opt < SystemDictionary::Opt);
+
+  bool must_load;
+#if INCLUDE_JVMCI
+  if (EnableJVMCI) {
+    // If JVMCI is enabled we require its classes to be found.
+    must_load = (init_opt < SystemDictionary::Opt) || (init_opt == SystemDictionary::Jvmci);
+  } else
+#endif
+  {
+    must_load = (init_opt < SystemDictionary::Opt);
+  }
+
   if ((*klassp) == NULL) {
     Klass* k;
     if (must_load) {

--- a/hotspot/src/share/vm/classfile/systemDictionary.hpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/classfile/systemDictionary.hpp	Fri Apr 01 08:46:47 2016 +0000
@@ -241,7 +241,7 @@
 
     Opt,                        // preload tried; NULL if not present
 #if INCLUDE_JVMCI
-    Jvmci,                      // preload tried; error if not present, use only with JVMCI
+    Jvmci,                      // preload tried; error if not present if JVMCI enabled
 #endif
     OPTION_LIMIT,
     CEIL_LG_OPTION_LIMIT = 2    // OPTION_LIMIT <= (1<<CEIL_LG_OPTION_LIMIT)

--- a/hotspot/src/share/vm/compiler/compileBroker.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/compiler/compileBroker.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -386,13 +386,16 @@
     task = CompilationPolicy::policy()->select_task(this);
   }
 
-  // Save method pointers across unlock safepoint.  The task is removed from
-  // the compilation queue, which is walked during RedefineClasses.
-  save_method = methodHandle(task->method());
-  save_hot_method = methodHandle(task->hot_method());
+  if (task != NULL) {
+    // Save method pointers across unlock safepoint.  The task is removed from
+    // the compilation queue, which is walked during RedefineClasses.
+    save_method = methodHandle(task->method());
+    save_hot_method = methodHandle(task->hot_method());
 
-  remove(task);
-  purge_stale_tasks(); // may temporarily release MCQ lock
+    remove(task);
+    purge_stale_tasks(); // may temporarily release MCQ lock
+  }
+
   return task;
 }
 
@@ -1781,7 +1784,8 @@
   bool is_osr = (osr_bci != standard_entry_bci);
   bool should_log = (thread->log() != NULL);
   bool should_break = false;
-  int task_level = task->comp_level();
+  const int task_level = task->comp_level();
+  AbstractCompiler* comp = task->compiler();
 
   DirectiveSet* directive;
   {
@@ -1793,7 +1797,7 @@
     assert(!method->is_native(), "no longer compile natives");
 
     // Look up matching directives
-    directive = DirectivesStack::getMatchingDirective(method, compiler(task_level));
+    directive = DirectivesStack::getMatchingDirective(method, comp);
 
     // Save information about this method in case of failure.
     set_last_compile(thread, method, is_osr, task_level);
@@ -1812,13 +1816,13 @@
   int compilable = ciEnv::MethodCompilable;
   const char* failure_reason = NULL;
   const char* retry_message = NULL;
-  AbstractCompiler *comp = compiler(task_level);
 
   int system_dictionary_modification_counter;
   {
     MutexLocker locker(Compile_lock, thread);
     system_dictionary_modification_counter = SystemDictionary::number_of_modifications();
   }
+
 #if INCLUDE_JVMCI
   if (UseJVMCICompiler && comp != NULL && comp->is_jvmci()) {
     JVMCICompiler* jvmci = (JVMCICompiler*) comp;

--- a/hotspot/src/share/vm/compiler/compileTask.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/compiler/compileTask.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 1998, 2015, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 1998, 2016, 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
@@ -122,6 +122,13 @@
   _next = NULL;
 }
 
+/**
+ * Returns the compiler for this task.
+ */
+AbstractCompiler* CompileTask::compiler() {
+  return CompileBroker::compiler(_comp_level);
+}
+
 // ------------------------------------------------------------------
 // CompileTask::code/set_code
 //

--- a/hotspot/src/share/vm/compiler/compileTask.hpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/compiler/compileTask.hpp	Fri Apr 01 08:46:47 2016 +0000
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 1998, 2015, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 1998, 2016, 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
@@ -115,6 +115,8 @@
   int          comp_level()                      { return _comp_level;}
   void         set_comp_level(int comp_level)    { _comp_level = comp_level;}
 
+  AbstractCompiler* compiler();
+
   int          num_inlined_bytecodes() const     { return _num_inlined_bytecodes; }
   void         set_num_inlined_bytecodes(int n)  { _num_inlined_bytecodes = n; }
 

--- a/hotspot/src/share/vm/opto/graphKit.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/opto/graphKit.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -1686,6 +1686,9 @@
   const Type* elemtype = arytype->elem();
   BasicType elembt = elemtype->array_element_basic_type();
   Node* adr = array_element_address(ary, idx, elembt, arytype->size());
+  if (elembt == T_NARROWOOP) {
+    elembt = T_OBJECT; // To satisfy switch in LoadNode::make()
+  }
   Node* ld = make_load(ctl, adr, elemtype, elembt, arytype, MemNode::unordered);
   return ld;
 }

--- a/hotspot/src/share/vm/opto/library_call.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/opto/library_call.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -6272,7 +6272,20 @@
 
 //------------------------------get_key_start_from_aescrypt_object-----------------------
 Node * LibraryCallKit::get_key_start_from_aescrypt_object(Node *aescrypt_object) {
+#ifdef PPC64
+  // MixColumns for decryption can be reduced by preprocessing MixColumns with round keys.
+  // Intel's extention is based on this optimization and AESCrypt generates round keys by preprocessing MixColumns.
+  // However, ppc64 vncipher processes MixColumns and requires the same round keys with encryption.
+  // The ppc64 stubs of encryption and decryption use the same round keys (sessionK[0]).
+  Node* objSessionK = load_field_from_object(aescrypt_object, "sessionK", "[[I", /*is_exact*/ false);
+  assert (objSessionK != NULL, "wrong version of com.sun.crypto.provider.AESCrypt");
+  if (objSessionK == NULL) {
+    return (Node *) NULL;
+  }
+  Node* objAESCryptKey = load_array_element(control(), objSessionK, intcon(0), TypeAryPtr::OOPS);
+#else
   Node* objAESCryptKey = load_field_from_object(aescrypt_object, "K", "[I", /*is_exact*/ false);
+#endif // PPC64
   assert (objAESCryptKey != NULL, "wrong version of com.sun.crypto.provider.AESCrypt");
   if (objAESCryptKey == NULL) return (Node *) NULL;
 

--- a/hotspot/src/share/vm/runtime/simpleThresholdPolicy.cpp	Fri Apr 01 01:34:00 2016 -0700
+++ b/hotspot/src/share/vm/runtime/simpleThresholdPolicy.cpp	Fri Apr 01 08:46:47 2016 +0000
@@ -233,6 +233,14 @@
   if (level == CompLevel_none) {
     return;
   }
+
+#if INCLUDE_JVMCI
+  // We can't compile with a JVMCI compiler until the module system is initialized.
+  if (level == CompLevel_full_optimization && UseJVMCICompiler && !Universe::is_module_initialized()) {
+    return;
+  }
+#endif
+
   // Check if the method can be compiled. If it cannot be compiled with C1, continue profiling
   // in the interpreter and then compile with C2 (the transition function will request that,
   // see common() ). If the method cannot be compiled with C2 but still can with C1, compile it with