Merge tag 'jdk-22.0.2+9' into ola/24.0_jdk-22.0.2+9

Added tag jdk-22.0.2+9 for changeset 8153097

doc/testing.html

@@ -179,8 +179,9 @@ <h2 id="test-selection">Test selection</h2>
 <p>The test specifications given in <code>TEST</code> is parsed into
 fully qualified test descriptors, which clearly and unambigously show
 which tests will be run. As an example, <code>:tier1</code> will expand
-to
-<code>jtreg:$(TOPDIR)/test/hotspot/jtreg:tier1 jtreg:$(TOPDIR)/test/jdk:tier1 jtreg:$(TOPDIR)/test/langtools:tier1 jtreg:$(TOPDIR)/test/nashorn:tier1 jtreg:$(TOPDIR)/test/jaxp:tier1</code>.
+to include all subcomponent test directories that define `tier1`,
+for example:
+<code>jtreg:$(TOPDIR)/test/hotspot/jtreg:tier1 jtreg:$(TOPDIR)/test/jdk:tier1 jtreg:$(TOPDIR)/test/langtools:tier1 ...</code>.
 You can always submit a list of fully qualified test descriptors in the
 <code>TEST</code> variable if you want to shortcut the parser.</p>
 <h3 id="common-test-groups">Common Test Groups</h3>

doc/testing.md

@@ -103,11 +103,11 @@ test runs, the `test TEST="x"` solution needs to be used.
 
 The test specifications given in `TEST` is parsed into fully qualified test
 descriptors, which clearly and unambigously show which tests will be run. As an
-example, `:tier1` will expand to `jtreg:$(TOPDIR)/test/hotspot/jtreg:tier1
-jtreg:$(TOPDIR)/test/jdk:tier1 jtreg:$(TOPDIR)/test/langtools:tier1
-jtreg:$(TOPDIR)/test/nashorn:tier1 jtreg:$(TOPDIR)/test/jaxp:tier1`. You can
-always submit a list of fully qualified test descriptors in the `TEST` variable
-if you want to shortcut the parser.
+example, `:tier1` will expand to include all subcomponent test directories
+that define `tier1`, for example: `jtreg:$(TOPDIR)/test/hotspot/jtreg:tier1
+jtreg:$(TOPDIR)/test/jdk:tier1 jtreg:$(TOPDIR)/test/langtools:tier1 ...`. You
+can always submit a list of fully qualified test descriptors in the `TEST`
+variable if you want to shortcut the parser.
 
 ### Common Test Groups
 

make/autoconf/jdk-options.m4

@@ -190,6 +190,17 @@ AC_DEFUN_ONCE([JDKOPT_SETUP_JDK_OPTIONS],
  fi
  AC_SUBST(INCLUDE_SA)
 
+ # Setup default CDS alignment. On platforms where one build may run on machines with different
+ # page sizes, the JVM choses a compatible alignment to fit all possible page sizes. This slightly
+ # increases archive size.
+ # The only platform having this problem at the moment is Linux on aarch64, which may encounter
+ # three different page sizes: 4K, 64K, and if run on Mac m1 hardware, 16K.
+ COMPATIBLE_CDS_ALIGNMENT_DEFAULT=false
+ if test "x$OPENJDK_TARGET_OS" = "xlinux" && test "x$OPENJDK_TARGET_CPU" = "xaarch64"; then
+ COMPATIBLE_CDS_ALIGNMENT_DEFAULT=true
+ fi
+ AC_SUBST(COMPATIBLE_CDS_ALIGNMENT_DEFAULT)
+
  # Compress jars
  COMPRESS_JARS=false
 
@@ -491,7 +502,7 @@ AC_DEFUN_ONCE([JDKOPT_SETUP_UNDEFINED_BEHAVIOR_SANITIZER],
 [
  # GCC reports lots of likely false positives for stringop-truncation and format-overflow.
  # Silence them for now.
- UBSAN_CHECKS="-fsanitize=undefined -fsanitize=float-divide-by-zero -fno-sanitize=shift-base"
+ UBSAN_CHECKS="-fsanitize=undefined -fsanitize=float-divide-by-zero -fno-sanitize=shift-base -fno-sanitize=alignment"
  UBSAN_CFLAGS="$UBSAN_CHECKS -Wno-stringop-truncation -Wno-format-overflow -fno-omit-frame-pointer -DUNDEFINED_BEHAVIOR_SANITIZER"
  UBSAN_LDFLAGS="$UBSAN_CHECKS"
  UTIL_ARG_ENABLE(NAME: ubsan, DEFAULT: false, RESULT: UBSAN_ENABLED,
@@ -673,7 +684,7 @@ AC_DEFUN([JDKOPT_ENABLE_DISABLE_CDS_ARCHIVE],
 #
 AC_DEFUN([JDKOPT_ENABLE_DISABLE_COMPATIBLE_CDS_ALIGNMENT],
 [
- UTIL_ARG_ENABLE(NAME: compatible-cds-alignment, DEFAULT: false,
+ UTIL_ARG_ENABLE(NAME: compatible-cds-alignment, DEFAULT: $COMPATIBLE_CDS_ALIGNMENT_DEFAULT,
  RESULT: ENABLE_COMPATIBLE_CDS_ALIGNMENT,
  DESC: [enable use alternative compatible cds core region alignment],
  DEFAULT_DESC: [disabled],

make/conf/version-numbers.conf

@@ -39,4 +39,4 @@ DEFAULT_VERSION_CLASSFILE_MINOR=0
 DEFAULT_VERSION_DOCS_API_SINCE=11
 DEFAULT_ACCEPTABLE_BOOT_VERSIONS="21 22"
 DEFAULT_JDK_SOURCE_TARGET_VERSION=22
-DEFAULT_PROMOTED_VERSION_PRE=ea
+DEFAULT_PROMOTED_VERSION_PRE=

make/devkit/createJMHBundle.sh

@@ -43,7 +43,7 @@ rm -f *
 fetchJar() {
  url="https://repo.maven.apache.org/maven2/$1/$2/$3/$2-$3.jar"
  if command -v curl > /dev/null; then
- curl -O --fail $url
+ curl -OL --fail $url
  elif command -v wget > /dev/null; then
  wget $url
  else

make/modules/java.base/Launcher.gmk

@@ -78,7 +78,8 @@ ifeq ($(call isTargetOs, macosx aix linux), true)
  NAME := jspawnhelper, \
  SRC := $(TOPDIR)/src/$(MODULE)/unix/native/jspawnhelper, \
  OPTIMIZATION := LOW, \
- CFLAGS := $(CFLAGS_JDKEXE) -I$(TOPDIR)/src/$(MODULE)/unix/native/libjava, \
+ CFLAGS := $(CFLAGS_JDKEXE) $(VERSION_CFLAGS) \
+ -I$(TOPDIR)/src/$(MODULE)/unix/native/libjava, \
  EXTRA_OBJECT_FILES := $(SUPPORT_OUTPUTDIR)/native/$(MODULE)/libjava/childproc$(OBJ_SUFFIX), \
  LDFLAGS := $(LDFLAGS_JDKEXE), \
  OUTPUT_DIR := $(SUPPORT_OUTPUTDIR)/modules_libs/$(MODULE), \

make/modules/java.base/lib/CoreLibraries.gmk

@@ -59,6 +59,7 @@ $(eval $(call SetupJdkLibrary, BUILD_LIBJAVA, \
  CFLAGS := $(CFLAGS_JDKLIB) \
  $(LIBJAVA_CFLAGS), \
  jdk_util.c_CFLAGS := $(VERSION_CFLAGS), \
+ ProcessImpl_md.c_CFLAGS := $(VERSION_CFLAGS), \
  WARNINGS_AS_ERRORS_xlc := false, \
  DISABLED_WARNINGS_gcc_ProcessImpl_md.c := unused-result, \
  LDFLAGS := $(LDFLAGS_JDKLIB) \

make/modules/java.desktop/lib/Awt2dLibraries.gmk

@@ -506,8 +506,10 @@ else
  # noexcept-type required for GCC 7 builds. Not required for GCC 8+.
  # expansion-to-defined required for GCC 9 builds. Not required for GCC 10+.
  # maybe-uninitialized required for GCC 8 builds. Not required for GCC 9+.
+ # calloc-transposed-args required for GCC 14 builds. (fixed upstream in Harfbuzz 032c931e1c0cfb20f18e5acb8ba005775242bd92)
  HARFBUZZ_DISABLED_WARNINGS_CXX_gcc := class-memaccess noexcept-type \
- expansion-to-defined dangling-reference maybe-uninitialized
+ expansion-to-defined dangling-reference maybe-uninitialized \
+ calloc-transposed-args
  HARFBUZZ_DISABLED_WARNINGS_clang := missing-field-initializers range-loop-analysis
  HARFBUZZ_DISABLED_WARNINGS_microsoft := 4267 4244
 

src/hotspot/cpu/aarch64/compressedKlass_aarch64.cpp

@@ -57,15 +57,15 @@ static char* reserve_at_eor_compatible_address(size_t size, bool aslr) {
  0x7800, 0x7c00, 0x7e00, 0x7f00, 0x7f80, 0x7fc0, 0x7fe0, 0x7ff0, 0x7ff8,
  0x7ffc, 0x7ffe, 0x7fff
  };
- static constexpr int num_immediates = sizeof(immediates) / sizeof(immediates[0]);
- const int start_index = aslr ? os::next_random((int)os::javaTimeNanos()) : 0;
+ static constexpr unsigned num_immediates = sizeof(immediates) / sizeof(immediates[0]);
+ const unsigned start_index = aslr ? os::next_random((int)os::javaTimeNanos()) : 0;
  constexpr int max_tries = 64;
  for (int ntry = 0; result == nullptr && ntry < max_tries; ntry ++) {
  // As in os::attempt_reserve_memory_between, we alternate between higher and lower
  // addresses; this maximizes the chance of early success if part of the address space
  // is not accessible (e.g. 39-bit address space).
- const int alt_index = (ntry & 1) ? 0 : num_immediates / 2;
- const int index = (start_index + ntry + alt_index) % num_immediates;
+ const unsigned alt_index = (ntry & 1) ? 0 : num_immediates / 2;
+ const unsigned index = (start_index + ntry + alt_index) % num_immediates;
  const uint64_t immediate = ((uint64_t)immediates[index]) << 32;
  assert(immediate > 0 && Assembler::operand_valid_for_logical_immediate(/*is32*/false, immediate),
  "Invalid immediate %d " UINT64_FORMAT, index, immediate);

src/hotspot/cpu/aarch64/gc/x/xGlobals_aarch64.cpp

@@ -142,9 +142,11 @@
 // * 63-48 Fixed (16-bits, always zero)
 //
 
-// Default value if probing is not implemented for a certain platform: 128TB
-static const size_t DEFAULT_MAX_ADDRESS_BIT = 47;
-// Minimum value returned, if probing fails: 64GB
+// Default value if probing is not implemented for a certain platform
+// Max address bit is restricted by implicit assumptions in the code, for instance
+// the bit layout of XForwardingEntry or Partial array entry (see XMarkStackEntry) in mark stack
+static const size_t DEFAULT_MAX_ADDRESS_BIT = 46;
+// Minimum value returned, if probing fails
 static const size_t MINIMUM_MAX_ADDRESS_BIT = 36;
 
 static size_t probe_valid_max_address_bit() {

src/hotspot/cpu/aarch64/gc/z/zAddress_aarch64.cpp

@@ -36,9 +36,11 @@
 #include <sys/mman.h>
 #endif // LINUX
 
-// Default value if probing is not implemented for a certain platform: 128TB
-static const size_t DEFAULT_MAX_ADDRESS_BIT = 47;
-// Minimum value returned, if probing fails: 64GB
+// Default value if probing is not implemented for a certain platform
+// Max address bit is restricted by implicit assumptions in the code, for instance
+// the bit layout of ZForwardingEntry or Partial array entry (see ZMarkStackEntry) in mark stack
+static const size_t DEFAULT_MAX_ADDRESS_BIT = 46;
+// Minimum value returned, if probing fail
 static const size_t MINIMUM_MAX_ADDRESS_BIT = 36;
 
 static size_t probe_valid_max_address_bit() {

src/hotspot/cpu/aarch64/macroAssembler_aarch64.cpp

@@ -390,13 +390,13 @@ static bool offset_for(uint32_t insn1, uint32_t insn2, ptrdiff_t &byte_offset) {
  return false;
 }
 
-class Decoder : public RelocActions {
- virtual reloc_insn adrpMem() { return &Decoder::adrpMem_impl; }
- virtual reloc_insn adrpAdd() { return &Decoder::adrpAdd_impl; }
- virtual reloc_insn adrpMovk() { return &Decoder::adrpMovk_impl; }
+class AArch64Decoder : public RelocActions {
+ virtual reloc_insn adrpMem() { return &AArch64Decoder::adrpMem_impl; }
+ virtual reloc_insn adrpAdd() { return &AArch64Decoder::adrpAdd_impl; }
+ virtual reloc_insn adrpMovk() { return &AArch64Decoder::adrpMovk_impl; }
 
 public:
- Decoder(address insn_addr, uint32_t insn) : RelocActions(insn_addr, insn) {}
+ AArch64Decoder(address insn_addr, uint32_t insn) : RelocActions(insn_addr, insn) {}
 
  virtual int loadStore(address insn_addr, address &target) {
  intptr_t offset = Instruction_aarch64::sextract(_insn, 23, 5);
@@ -492,7 +492,7 @@ class Decoder : public RelocActions {
 };
 
 address MacroAssembler::target_addr_for_insn(address insn_addr, uint32_t insn) {
- Decoder decoder(insn_addr, insn);
+ AArch64Decoder decoder(insn_addr, insn);
  address target;
  decoder.run(insn_addr, target);
  return target;
@@ -4258,108 +4258,117 @@ void MacroAssembler::kernel_crc32_common_fold_using_crypto_pmull(Register crc, R
  }
  add(table, table, table_offset);
 
+ // Registers v0..v7 are used as data registers.
+ // Registers v16..v31 are used as tmp registers.
  sub(buf, buf, 0x10);
- ldrq(v1, Address(buf, 0x10));
- ldrq(v2, Address(buf, 0x20));
- ldrq(v3, Address(buf, 0x30));
- ldrq(v4, Address(buf, 0x40));
- ldrq(v5, Address(buf, 0x50));
- ldrq(v6, Address(buf, 0x60));
- ldrq(v7, Address(buf, 0x70));
- ldrq(v8, Address(pre(buf, 0x80)));
-
- movi(v25, T4S, 0);
- mov(v25, S, 0, crc);
- eor(v1, T16B, v1, v25);
-
- ldrq(v0, Address(table));
+ ldrq(v0, Address(buf, 0x10));
+ ldrq(v1, Address(buf, 0x20));
+ ldrq(v2, Address(buf, 0x30));
+ ldrq(v3, Address(buf, 0x40));
+ ldrq(v4, Address(buf, 0x50));
+ ldrq(v5, Address(buf, 0x60));
+ ldrq(v6, Address(buf, 0x70));
+ ldrq(v7, Address(pre(buf, 0x80)));
+
+ movi(v31, T4S, 0);
+ mov(v31, S, 0, crc);
+ eor(v0, T16B, v0, v31);
+
+ // Register v16 contains constants from the crc table.
+ ldrq(v16, Address(table));
  b(CRC_by128_loop);
 
  align(OptoLoopAlignment);
  BIND(CRC_by128_loop);
- pmull (v9, T1Q, v1, v0, T1D);
- pmull2(v10, T1Q, v1, v0, T2D);
- ldrq(v1, Address(buf, 0x10));
- eor3(v1, T16B, v9, v10, v1);
-
- pmull (v11, T1Q, v2, v0, T1D);
- pmull2(v12, T1Q, v2, v0, T2D);
- ldrq(v2, Address(buf, 0x20));
- eor3(v2, T16B, v11, v12, v2);
-
- pmull (v13, T1Q, v3, v0, T1D);
- pmull2(v14, T1Q, v3, v0, T2D);
- ldrq(v3, Address(buf, 0x30));
- eor3(v3, T16B, v13, v14, v3);
-
- pmull (v15, T1Q, v4, v0, T1D);
- pmull2(v16, T1Q, v4, v0, T2D);
- ldrq(v4, Address(buf, 0x40));
- eor3(v4, T16B, v15, v16, v4);
-
- pmull (v17, T1Q, v5, v0, T1D);
- pmull2(v18, T1Q, v5, v0, T2D);
- ldrq(v5, Address(buf, 0x50));
- eor3(v5, T16B, v17, v18, v5);
-
- pmull (v19, T1Q, v6, v0, T1D);
- pmull2(v20, T1Q, v6, v0, T2D);
- ldrq(v6, Address(buf, 0x60));
- eor3(v6, T16B, v19, v20, v6);
-
- pmull (v21, T1Q, v7, v0, T1D);
- pmull2(v22, T1Q, v7, v0, T2D);
- ldrq(v7, Address(buf, 0x70));
- eor3(v7, T16B, v21, v22, v7);
-
- pmull (v23, T1Q, v8, v0, T1D);
- pmull2(v24, T1Q, v8, v0, T2D);
- ldrq(v8, Address(pre(buf, 0x80)));
- eor3(v8, T16B, v23, v24, v8);
+ pmull (v17, T1Q, v0, v16, T1D);
+ pmull2(v18, T1Q, v0, v16, T2D);
+ ldrq(v0, Address(buf, 0x10));
+ eor3(v0, T16B, v17, v18, v0);
+
+ pmull (v19, T1Q, v1, v16, T1D);
+ pmull2(v20, T1Q, v1, v16, T2D);
+ ldrq(v1, Address(buf, 0x20));
+ eor3(v1, T16B, v19, v20, v1);
+
+ pmull (v21, T1Q, v2, v16, T1D);
+ pmull2(v22, T1Q, v2, v16, T2D);
+ ldrq(v2, Address(buf, 0x30));
+ eor3(v2, T16B, v21, v22, v2);
+
+ pmull (v23, T1Q, v3, v16, T1D);
+ pmull2(v24, T1Q, v3, v16, T2D);
+ ldrq(v3, Address(buf, 0x40));
+ eor3(v3, T16B, v23, v24, v3);
+
+ pmull (v25, T1Q, v4, v16, T1D);
+ pmull2(v26, T1Q, v4, v16, T2D);
+ ldrq(v4, Address(buf, 0x50));
+ eor3(v4, T16B, v25, v26, v4);
+
+ pmull (v27, T1Q, v5, v16, T1D);
+ pmull2(v28, T1Q, v5, v16, T2D);
+ ldrq(v5, Address(buf, 0x60));
+ eor3(v5, T16B, v27, v28, v5);
+
+ pmull (v29, T1Q, v6, v16, T1D);
+ pmull2(v30, T1Q, v6, v16, T2D);
+ ldrq(v6, Address(buf, 0x70));
+ eor3(v6, T16B, v29, v30, v6);
+
+ // Reuse registers v23, v24.
+ // Using them won't block the first instruction of the next iteration.
+ pmull (v23, T1Q, v7, v16, T1D);
+ pmull2(v24, T1Q, v7, v16, T2D);
+ ldrq(v7, Address(pre(buf, 0x80)));
+ eor3(v7, T16B, v23, v24, v7);
 
  subs(len, len, 0x80);
  br(Assembler::GE, CRC_by128_loop);
 
  // fold into 512 bits
- ldrq(v0, Address(table, 0x10));
+ // Use v31 for constants because v16 can be still in use.
+ ldrq(v31, Address(table, 0x10));
 
- pmull (v10, T1Q, v1, v0, T1D);
- pmull2(v11, T1Q, v1, v0, T2D);
- eor3(v1, T16B, v10, v11, v5);
+ pmull (v17, T1Q, v0, v31, T1D);
+ pmull2(v18, T1Q, v0, v31, T2D);
+ eor3(v0, T16B, v17, v18, v4);
 
- pmull (v12, T1Q, v2, v0, T1D);
- pmull2(v13, T1Q, v2, v0, T2D);
- eor3(v2, T16B, v12, v13, v6);
+ pmull (v19, T1Q, v1, v31, T1D);
+ pmull2(v20, T1Q, v1, v31, T2D);
+ eor3(v1, T16B, v19, v20, v5);
 
- pmull (v14, T1Q, v3, v0, T1D);
- pmull2(v15, T1Q, v3, v0, T2D);
- eor3(v3, T16B, v14, v15, v7);
+ pmull (v21, T1Q, v2, v31, T1D);
+ pmull2(v22, T1Q, v2, v31, T2D);
+ eor3(v2, T16B, v21, v22, v6);
 
- pmull (v16, T1Q, v4, v0, T1D);
- pmull2(v17, T1Q, v4, v0, T2D);
- eor3(v4, T16B, v16, v17, v8);
+ pmull (v23, T1Q, v3, v31, T1D);
+ pmull2(v24, T1Q, v3, v31, T2D);
+ eor3(v3, T16B, v23, v24, v7);
 
  // fold into 128 bits
- ldrq(v5, Address(table, 0x20));
- pmull (v10, T1Q, v1, v5, T1D);
- pmull2(v11, T1Q, v1, v5, T2D);
- eor3(v4, T16B, v4, v10, v11);
-
- ldrq(v6, Address(table, 0x30));
- pmull (v12, T1Q, v2, v6, T1D);
- pmull2(v13, T1Q, v2, v6, T2D);
- eor3(v4, T16B, v4, v12, v13);
-
- ldrq(v7, Address(table, 0x40));
- pmull (v14, T1Q, v3, v7, T1D);
- pmull2(v15, T1Q, v3, v7, T2D);
- eor3(v1, T16B, v4, v14, v15);
+ // Use v17 for constants because v31 can be still in use.
+ ldrq(v17, Address(table, 0x20));
+ pmull (v25, T1Q, v0, v17, T1D);
+ pmull2(v26, T1Q, v0, v17, T2D);
+ eor3(v3, T16B, v3, v25, v26);
+
+ // Use v18 for constants because v17 can be still in use.
+ ldrq(v18, Address(table, 0x30));
+ pmull (v27, T1Q, v1, v18, T1D);
+ pmull2(v28, T1Q, v1, v18, T2D);
+ eor3(v3, T16B, v3, v27, v28);
+
+ // Use v19 for constants because v18 can be still in use.
+ ldrq(v19, Address(table, 0x40));
+ pmull (v29, T1Q, v2, v19, T1D);
+ pmull2(v30, T1Q, v2, v19, T2D);
+ eor3(v0, T16B, v3, v29, v30);
 
  add(len, len, 0x80);
  add(buf, buf, 0x10);
 
- mov(tmp0, v1, D, 0);
- mov(tmp1, v1, D, 1);
+ mov(tmp0, v0, D, 0);
+ mov(tmp1, v0, D, 1);
 }
 
 SkipIfEqual::SkipIfEqual(