8274527: Minimal VM build fails after JDK-8273459

Backport-of: a8edd1b360d4e5f35aff371a91fda42eeb00d395

Author: Sandhya Viswanathan

src/hotspot/cpu/x86/macroAssembler_x86.cpp

@@ -1178,6 +1178,10 @@ void MacroAssembler::align64() {
   align(64, (unsigned long long) pc());
 }
 
+void MacroAssembler::align32() {
+  align(32, (unsigned long long) pc());
+}
+
 void MacroAssembler::align(int modulus) {
   // 8273459: Ensure alignment is possible with current segment alignment
   assert(modulus <= CodeEntryAlignment, "Alignment must be <= CodeEntryAlignment");
@@ -7123,7 +7127,7 @@ void MacroAssembler::kernel_crc32(Register crc, Register buf, Register len, Regi
   // 128 bits per each of 4 parallel streams.
   movdqu(xmm0, ExternalAddress(StubRoutines::x86::crc_by128_masks_addr() + 32));
 
-  align(32);
+  align32();
   BIND(L_fold_512b_loop);
   fold_128bit_crc32(xmm1, xmm0, xmm5, buf,  0);
   fold_128bit_crc32(xmm2, xmm0, xmm5, buf, 16);

src/hotspot/cpu/x86/macroAssembler_x86.hpp

@@ -194,6 +194,7 @@ class MacroAssembler: public Assembler {
   void incrementq(AddressLiteral dst);
 
   // Alignment
+  void align32();
   void align64();
   void align(int modulus);
   void align(int modulus, int target);

src/hotspot/cpu/x86/macroAssembler_x86_adler.cpp

@@ -80,7 +80,7 @@ void MacroAssembler::updateBytesAdler32(Register init_d, Register data, Register
       cmpptr(data, end);
       jcc(Assembler::aboveEqual, SKIP_LOOP_1A);
 
-      align(32);
+      align32();
       bind(SLOOP1A);
       vbroadcastf128(ydata, Address(data, 0), Assembler::AVX_256bit);
       addptr(data, CHUNKSIZE);
@@ -178,7 +178,7 @@ void MacroAssembler::updateBytesAdler32(Register init_d, Register data, Register
       movdl(rax, xb);
       addl(b_d, rax);
 
-      align(32);
+      align32();
       bind(FINAL_LOOP);
       movzbl(rax, Address(data, 0)); //movzx   eax, byte[data]
       addl(a_d, rax);

src/hotspot/cpu/x86/stubGenerator_x86_64.cpp

@@ -1484,7 +1484,7 @@ class StubGenerator: public StubCodeGenerator {
         __ subq(temp1, loop_size[shift]);
 
         // Main loop with aligned copy block size of 192 bytes at 32 byte granularity.
-        __ align(32);
+        __ align32();
         __ BIND(L_main_loop);
            __ copy64_avx(to, from, temp4, xmm1, false, shift, 0);
            __ copy64_avx(to, from, temp4, xmm1, false, shift, 64);
@@ -1551,7 +1551,7 @@ class StubGenerator: public StubCodeGenerator {
 
         // Main loop with aligned copy block size of 192 bytes at
         // 64 byte copy granularity.
-        __ align(32);
+        __ align32();
         __ BIND(L_main_loop_64bytes);
            __ copy64_avx(to, from, temp4, xmm1, false, shift, 0 , true);
            __ copy64_avx(to, from, temp4, xmm1, false, shift, 64, true);
@@ -1691,7 +1691,7 @@ class StubGenerator: public StubCodeGenerator {
         __ BIND(L_main_pre_loop);
 
         // Main loop with aligned copy block size of 192 bytes at 32 byte granularity.
-        __ align(32);
+        __ align32();
         __ BIND(L_main_loop);
            __ copy64_avx(to, from, temp1, xmm1, true, shift, -64);
            __ copy64_avx(to, from, temp1, xmm1, true, shift, -128);
@@ -1724,7 +1724,7 @@ class StubGenerator: public StubCodeGenerator {
 
         // Main loop with aligned copy block size of 192 bytes at
         // 64 byte copy granularity.
-        __ align(32);
+        __ align32();
         __ BIND(L_main_loop_64bytes);
            __ copy64_avx(to, from, temp1, xmm1, true, shift, -64 , true);
            __ copy64_avx(to, from, temp1, xmm1, true, shift, -128, true);
@@ -4274,7 +4274,7 @@ class StubGenerator: public StubCodeGenerator {
 
   //Mask for byte-swapping a couple of qwords in an XMM register using (v)pshufb.
   address generate_pshuffle_byte_flip_mask_sha512() {
-    __ align(32);
+    __ align32();
     StubCodeMark mark(this, "StubRoutines", "pshuffle_byte_flip_mask_sha512");
     address start = __ pc();
     if (VM_Version::supports_avx2()) {
@@ -5307,7 +5307,7 @@ address generate_avx_ghash_processBlocks() {
 
   address base64_avx2_shuffle_addr()
   {
-    __ align(32);
+    __ align32();
     StubCodeMark mark(this, "StubRoutines", "avx2_shuffle_base64");
     address start = __ pc();
     __ emit_data64(0x0809070805060405, relocInfo::none);
@@ -5319,7 +5319,7 @@ address generate_avx_ghash_processBlocks() {
 
   address base64_avx2_input_mask_addr()
   {
-    __ align(32);
+    __ align32();
     StubCodeMark mark(this, "StubRoutines", "avx2_input_mask_base64");
     address start = __ pc();
     __ emit_data64(0x8000000000000000, relocInfo::none);
@@ -5331,7 +5331,7 @@ address generate_avx_ghash_processBlocks() {
 
   address base64_avx2_lut_addr()
   {
-    __ align(32);
+    __ align32();
     StubCodeMark mark(this, "StubRoutines", "avx2_lut_base64");
     address start = __ pc();
     __ emit_data64(0xfcfcfcfcfcfc4741, relocInfo::none);
@@ -5436,7 +5436,7 @@ address generate_avx_ghash_processBlocks() {
       __ evmovdquq(xmm2, Address(encode_table, 0), Assembler::AVX_512bit);
       __ evpbroadcastq(xmm1, rax, Assembler::AVX_512bit);
 
-      __ align(32);
+      __ align32();
       __ BIND(L_vbmiLoop);
 
       __ vpermb(xmm0, xmm3, Address(source, start_offset), Assembler::AVX_512bit);
@@ -5636,7 +5636,7 @@ address generate_avx_ghash_processBlocks() {
       __ cmpl(length, 31);
       __ jcc(Assembler::belowEqual, L_process3);
 
-      __ align(32);
+      __ align32();
       __ BIND(L_32byteLoop);
 
       // Get next 32 bytes
@@ -6083,7 +6083,7 @@ address generate_avx_ghash_processBlocks() {
       __ evmovdquq(join12, ExternalAddress(StubRoutines::x86::base64_vbmi_join_1_2_addr()), Assembler::AVX_512bit, r13);
       __ evmovdquq(join23, ExternalAddress(StubRoutines::x86::base64_vbmi_join_2_3_addr()), Assembler::AVX_512bit, r13);
 
-      __ align(32);
+      __ align32();
       __ BIND(L_process256);
       // Grab input data
       __ evmovdquq(input0, Address(source, start_offset, Address::times_1, 0x00), Assembler::AVX_512bit);
@@ -6165,7 +6165,7 @@ address generate_avx_ghash_processBlocks() {
       __ cmpl(length, 63);
       __ jcc(Assembler::lessEqual, L_finalBit);
 
-      __ align(32);
+      __ align32();
       __ BIND(L_process64Loop);
 
       // Handle first 64-byte block
@@ -6301,7 +6301,7 @@ address generate_avx_ghash_processBlocks() {
       __ shrq(rax, 1);
       __ jmp(L_donePadding);
 
-      __ align(32);
+      __ align32();
       __ BIND(L_bruteForce);
     }   // End of if(avx512_vbmi)
 
@@ -6345,7 +6345,7 @@ address generate_avx_ghash_processBlocks() {
 
     __ jmp(L_bottomLoop);
 
-    __ align(32);
+    __ align32();
     __ BIND(L_forceLoop);
     __ shll(byte1, 18);
     __ shll(byte2, 12);