infernap12/jdk
1
0
Fork 0
mirror of https://github.com/openjdk/jdk.git synced 2026-01-28 03:58:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge

Browse Source
This commit is contained in:
Phil Race 2018-05-25 12:12:24 -07:00
commit 5ff01c4163
5595 changed files with 549680 additions and 21763 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
.hgtags
View File

@ -485,3 +485,5 @@ f7363de371c9a1f668bd0a01b7df3d1ddb9cc58b jdk-11+7
e1e60f75cd39312a7f59d2a4f91d624e5aecc95e jdk-11+11
3ab6ba9f94a9045a526d645af26c933235371d6f jdk-11+12
758deedaae8406ae60147486107a54e9864aa7b0 jdk-11+13
3595bd343b65f8c37818ebe6a4c343ddeb1a5f88 jdk-11+14
a11c1cb542bbd1671d25b85efe7d09b983c48525 jdk-11+15

1
doc/building.html
View File

@ -555,6 +555,7 @@ CC: Sun C++ 5.13 SunOS_i386 151846-10 2015/10/30</code></pre>
</ul>
<h2 id="running-tests">Running Tests</h2>
<p>Most of the OpenJDK tests are using the <a href="http://openjdk.java.net/jtreg">JTReg</a> test framework. Make sure that your configuration knows where to find your installation of JTReg. If this is not picked up automatically, use the <code>--with-jtreg=&lt;path to jtreg home&gt;</code> option to point to the JTReg framework. Note that this option should point to the JTReg home, i.e. the top directory, containing <code>lib/jtreg.jar</code> etc.</p>
<p>The <a href="https://wiki.openjdk.java.net/display/Adoption">Adoption Group</a> provides recent builds of jtreg <a href="https://adopt-openjdk.ci.cloudbees.com/job/jtreg/lastSuccessfulBuild/artifact">here</a>. Download the latest <code>.tar.gz</code> file, unpack it, and point <code>--with-jtreg</code> to the <code>jtreg</code> directory that you just unpacked.</p>
<p>To execute the most basic tests (tier 1), use:</p>
<pre><code>make run-test-tier1</code></pre>
<p>For more details on how to run tests, please see the <a href="testing.html">Testing OpenJDK</a> document.</p>

6
doc/building.md
View File

@ -833,6 +833,12 @@ installation of JTReg. If this is not picked up automatically, use the
Note that this option should point to the JTReg home, i.e. the top directory,
containing `lib/jtreg.jar` etc.
The [Adoption Group](https://wiki.openjdk.java.net/display/Adoption) provides
recent builds of jtreg [here](
https://adopt-openjdk.ci.cloudbees.com/job/jtreg/lastSuccessfulBuild/artifact).
Download the latest `.tar.gz` file, unpack it, and point `--with-jtreg` to the
`jtreg` directory that you just unpacked.
To execute the most basic tests (tier 1), use:
```
make run-test-tier1

17
doc/testing.html
View File

@ -40,20 +40,21 @@
$ make run-test-jdk_lang JTREG=&quot;JOBS=8&quot;
$ make run-test TEST=jdk_lang
$ make run-test-only TEST=&quot;gtest:LogTagSet gtest:LogTagSetDescriptions&quot; GTEST=&quot;REPEAT=-1&quot;
$ make run-test TEST=&quot;hotspot/test:hotspot_gc&quot; JTREG=&quot;JOBS=1;TIMEOUT=8;VM_OPTIONS=-XshowSettings -Xlog:gc+ref=debug&quot;
$ make run-test TEST=&quot;jtreg:hotspot/test:hotspot_gc hotspot/test/native_sanity/JniVersion.java&quot;
$ make exploded-run-test TEST=hotspot_tier1</code></pre>
$ make run-test TEST=&quot;hotspot:hotspot_gc&quot; JTREG=&quot;JOBS=1;TIMEOUT=8;VM_OPTIONS=-XshowSettings -Xlog:gc+ref=debug&quot;
$ make run-test TEST=&quot;jtreg:test/hotspot:hotspot_gc test/hotspot/jtreg/native_sanity/JniVersion.java&quot;
$ make exploded-run-test TEST=tier2</code></pre>
<h3 id="configuration">Configuration</h3>
<p>To be able to run JTReg tests, <code>configure</code> needs to know where to find the JTReg test framework. If it is not picked up automatically by configure, use the <code>--with-jtreg=&lt;path to jtreg home&gt;</code> option to point to the JTReg framework. Note that this option should point to the JTReg home, i.e. the top directory, containing <code>lib/jtreg.jar</code> etc. (An alternative is to set the <code>JT_HOME</code> environment variable to point to the JTReg home before running <code>configure</code>.)</p>
<h2 id="test-selection">Test selection</h2>
<p>All functionality is available using the run-test make target. In this use case, the test or tests to be executed is controlled using the <code>TEST</code> variable. To speed up subsequent test runs with no source code changes, run-test-only can be used instead, which do not depend on the source and test image build.</p>
<p>For some common top-level tests, direct make targets have been generated. This includes all JTReg test groups, the hotspot gtest, and custom tests (if present). This means that <code>make run-test-tier1</code> is equivalent to <code>make run-test TEST=&quot;tier1&quot;</code>, but the latter is more tab-completion friendly. For more complex test runs, the <code>run-test TEST=&quot;x&quot;</code> solution needs to be used.</p>
<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:jdk/test:tier1 jtreg:langtools/test:tier1 jtreg:nashorn/test:tier1 jtreg:jaxp/test: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>
<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:$(TOPDIR)/test/jaxp:tier1</code>. You can always submit a list of fully qualified test descriptors in the <code>TEST</code> variable if you want to shortcut the parser.</p>
<h3 id="jtreg">JTReg</h3>
<p>JTReg test groups can be specified either without a test root, e.g. <code>:tier1</code> (or <code>tier1</code>, the initial colon is optional), or with, e.g. <code>hotspot/test:tier1</code>, <code>jdk/test:jdk_util</code>.</p>
<p>When specified without a test root, all matching groups from all tests roots will be added. Otherwise, only the group from the specified test root will be added.</p>
<p>Individual JTReg tests or directories containing JTReg tests can also be specified, like <code>hotspot/test/native_sanity/JniVersion.java</code> or <code>hotspot/test/native_sanity</code>. You can also specify an absolute path, to point to a JTReg test outside the source tree.</p>
<p>As long as the test groups or test paths can be uniquely resolved, you do not need to enter the <code>jtreg:</code> prefix. If this is not possible, or if you want to use a fully qualified test descriptor, add <code>jtreg:</code>, e.g. <code>jtreg:hotspot/test/native_sanity</code>.</p>
<p>JTReg tests can be selected either by picking a JTReg test group, or a selection of files or directories containing JTReg tests.</p>
<p>JTReg test groups can be specified either without a test root, e.g. <code>:tier1</code> (or <code>tier1</code>, the initial colon is optional), or with, e.g. <code>hotspot:tier1</code>, <code>test/jdk:jdk_util</code> or <code>$(TOPDIR)/test/hotspot/jtreg:hotspot_all</code>. The test root can be specified either as an absolute path, or a path relative to the OpenJDK top directory, or the <code>test</code> directory. For simplicity, the hotspot JTReg test root, which really is <code>hotspot/jtreg</code> can be abbreviated as just <code>hotspot</code>.</p>
<p>When specified without a test root, all matching groups from all test roots will be added. Otherwise, only the group from the specified test root will be added.</p>
<p>Individual JTReg tests or directories containing JTReg tests can also be specified, like <code>test/hotspot/jtreg/native_sanity/JniVersion.java</code> or <code>hotspot/jtreg/native_sanity</code>. Just like for test root selection, you can either specify an absolute path (which can even point to JTReg tests outside the source tree), or a path relative to either the OpenJDK top directory or the <code>test</code> directory. <code>hotspot</code> can be used as an alias for <code>hotspot/jtreg</code> here as well.</p>
<p>As long as the test groups or test paths can be uniquely resolved, you do not need to enter the <code>jtreg:</code> prefix. If this is not possible, or if you want to use a fully qualified test descriptor, add <code>jtreg:</code>, e.g. <code>jtreg:test/hotspot/jtreg/native_sanity</code>.</p>
<h3 id="gtest">Gtest</h3>
<p>Since the Hotspot Gtest suite is so quick, the default is to run all tests. This is specified by just <code>gtest</code>, or as a fully qualified test descriptor <code>gtest:all</code>.</p>
<p>If you want, you can single out an individual test or a group of tests, for instance <code>gtest:LogDecorations</code> or <code>gtest:LogDecorations.level_test_vm</code>. This can be particularly useful if you want to run a shaky test repeatedly.</p>

41
doc/testing.md
View File

@ -18,9 +18,9 @@ Some example command-lines:
$ make run-test-jdk_lang JTREG="JOBS=8"
$ make run-test TEST=jdk_lang
$ make run-test-only TEST="gtest:LogTagSet gtest:LogTagSetDescriptions" GTEST="REPEAT=-1"
$ make run-test TEST="hotspot/test:hotspot_gc" JTREG="JOBS=1;TIMEOUT=8;VM_OPTIONS=-XshowSettings -Xlog:gc+ref=debug"
$ make run-test TEST="jtreg:hotspot/test:hotspot_gc hotspot/test/native_sanity/JniVersion.java"
$ make exploded-run-test TEST=hotspot_tier1
$ make run-test TEST="hotspot:hotspot_gc" JTREG="JOBS=1;TIMEOUT=8;VM_OPTIONS=-XshowSettings -Xlog:gc+ref=debug"
$ make run-test TEST="jtreg:test/hotspot:hotspot_gc test/hotspot/jtreg/native_sanity/JniVersion.java"
$ make exploded-run-test TEST=tier2
### Configuration
@ -46,30 +46,41 @@ test runs, the `run-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:jdk/test:tier1
jtreg:langtools/test:tier1 jtreg:nashorn/test:tier1 jtreg:jaxp/test: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 `jtreg:$(TOPDIR)/test/hotspot/jtreg:tier1
jtreg:$(TOPDIR)/test/jdk:tier1 jtreg:$(TOPDIR)/test/langtools:tier1
jtreg:$(TOPDIR)/test/nashorn:tier1 jtreg:$(TOPDIR)/test/jaxp:tier1`. You can
always submit a list of fully qualified test descriptors in the `TEST` variable
if you want to shortcut the parser.
### JTReg
JTReg test groups can be specified either without a test root, e.g. `:tier1`
(or `tier1`, the initial colon is optional), or with, e.g.
`hotspot/test:tier1`, `jdk/test:jdk_util`.
JTReg tests can be selected either by picking a JTReg test group, or a selection
of files or directories containing JTReg tests.
When specified without a test root, all matching groups from all tests roots
JTReg test groups can be specified either without a test root, e.g. `:tier1`
(or `tier1`, the initial colon is optional), or with, e.g. `hotspot:tier1`,
`test/jdk:jdk_util` or `$(TOPDIR)/test/hotspot/jtreg:hotspot_all`. The test
root can be specified either as an absolute path, or a path relative to the
OpenJDK top directory, or the `test` directory. For simplicity, the hotspot
JTReg test root, which really is `hotspot/jtreg` can be abbreviated as
just `hotspot`.
When specified without a test root, all matching groups from all test roots
will be added. Otherwise, only the group from the specified test root will be
added.
Individual JTReg tests or directories containing JTReg tests can also be
specified, like `hotspot/test/native_sanity/JniVersion.java` or
`hotspot/test/native_sanity`. You can also specify an absolute path, to point
to a JTReg test outside the source tree.
specified, like `test/hotspot/jtreg/native_sanity/JniVersion.java` or
`hotspot/jtreg/native_sanity`. Just like for test root selection, you can
either specify an absolute path (which can even point to JTReg tests outside
the source tree), or a path relative to either the OpenJDK top directory or the
`test` directory. `hotspot` can be used as an alias for `hotspot/jtreg` here as
well.
As long as the test groups or test paths can be uniquely resolved, you do not
need to enter the `jtreg:` prefix. If this is not possible, or if you want to
use a fully qualified test descriptor, add `jtreg:`, e.g.
`jtreg:hotspot/test/native_sanity`.
`jtreg:test/hotspot/jtreg/native_sanity`.
### Gtest

1
make/Docs.gmk
View File

@ -85,7 +85,6 @@ JAVADOC_TAGS := \
-tag throws \
-taglet build.tools.taglet.ModuleGraph \
-tag since \
-tag version \
-tag serialData \
-tag factory \
-tag see \

27
make/autoconf/basics.m4
View File

@ -119,6 +119,33 @@ AC_DEFUN([BASIC_GET_NON_MATCHING_VALUES],
fi
])
###############################################################################
# Check if a list of space-separated words contains any word(s) from a list of
# space-separated illegal words. Typical use is to see if a user-specified
# set of words contains any from a set of illegal words.
#
# Sets the specified variable to list of matching illegal words, or to
# the empty string if no words are matching the illegal set.
#
# $1: result variable name
# $2: list of values to check
# $3: list of illegal values
AC_DEFUN([BASIC_GET_MATCHING_VALUES],
[
# grep filter function inspired by a comment to http://stackoverflow.com/a/1617326
# Notice that the original variant fails on SLES 10 and 11
# Some grep versions (at least bsd) behaves strangely on the base case with
# no legal_values, so make it explicit.
values_to_check=`$ECHO $2 | $TR ' ' '\n'`
illegal_values=`$ECHO $3 | $TR ' ' '\n'`
if test -z "$illegal_values"; then
$1=""
else
result=`$GREP -Fx "$illegal_values" <<< "$values_to_check" | $GREP -v '^$'`
$1=${result//$'\n'/ }
fi
])
###############################################################################
# Sort a space-separated list, and remove duplicates.
#

26
make/autoconf/hotspot.m4
View File

@ -28,6 +28,9 @@ VALID_JVM_FEATURES="compiler1 compiler2 zero minimal dtrace jvmti jvmci \
graal vm-structs jni-check services management cmsgc g1gc parallelgc serialgc nmt cds \
static-build link-time-opt aot jfr"
# Deprecated JVM features (these are ignored, but with a warning)
DEPRECATED_JVM_FEATURES="trace"
# All valid JVM variants
VALID_JVM_VARIANTS="server client minimal core zero custom"
@ -269,18 +272,27 @@ AC_DEFUN_ONCE([HOTSPOT_SETUP_JVM_FEATURES],
USER_JVM_FEATURE_LIST=`$ECHO $with_jvm_features | $SED -e 's/,/ /g'`
AC_MSG_RESULT([$user_jvm_feature_list])
# These features will be added to all variant defaults
JVM_FEATURES=`$ECHO $USER_JVM_FEATURE_LIST | $AWK '{ for (i=1; i<=NF; i++) if (!match($i, /^-.*/)) print $i }'`
JVM_FEATURES=`$ECHO $USER_JVM_FEATURE_LIST | $AWK '{ for (i=1; i<=NF; i++) if (!match($i, /^-.*/)) printf("%s ", $i) }'`
# These features will be removed from all variant defaults
DISABLED_JVM_FEATURES=`$ECHO $USER_JVM_FEATURE_LIST | $AWK '{ for (i=1; i<=NF; i++) if (match($i, /^-.*/)) print substr($i, 2) }'`
DISABLED_JVM_FEATURES=`$ECHO $USER_JVM_FEATURE_LIST | $AWK '{ for (i=1; i<=NF; i++) if (match($i, /^-.*/)) printf("%s ", substr($i, 2))}'`
# Verify that the user has provided valid features
BASIC_GET_NON_MATCHING_VALUES(INVALID_FEATURES, $JVM_FEATURES $DISABLED_JVM_FEATURES, $VALID_JVM_FEATURES)
BASIC_GET_NON_MATCHING_VALUES(INVALID_FEATURES, $JVM_FEATURES $DISABLED_JVM_FEATURES, $VALID_JVM_FEATURES $DEPRECATED_JVM_FEATURES)
if test "x$INVALID_FEATURES" != x; then
AC_MSG_NOTICE([Unknown JVM features specified: "$INVALID_FEATURES"])
AC_MSG_NOTICE([The available JVM features are: "$VALID_JVM_FEATURES"])
AC_MSG_ERROR([Cannot continue])
fi
# Check if the user has provided deprecated features
BASIC_GET_MATCHING_VALUES(DEPRECATED_FEATURES, $JVM_FEATURES $DISABLED_JVM_FEATURES, $DEPRECATED_JVM_FEATURES)
if test "x$DEPRECATED_FEATURES" != x; then
AC_MSG_WARN([Deprecated JVM features specified (will be ignored): "$DEPRECATED_FEATURES"])
# Filter out deprecated features
BASIC_GET_NON_MATCHING_VALUES(JVM_FEATURES, $JVM_FEATURES, $DEPRECATED_FEATURES)
BASIC_GET_NON_MATCHING_VALUES(DISABLED_JVM_FEATURES, $DISABLED_JVM_FEATURES, $DEPRECATED_FEATURES)
fi
fi
# Override hotspot cpu definitions for ARM platforms
@ -309,9 +321,11 @@ AC_DEFUN_ONCE([HOTSPOT_SETUP_JVM_FEATURES],
AC_MSG_ERROR([Specified JVM feature 'cmsgc' requires feature 'serialgc'])
fi
# Enable JFR by default, except on linux-sparcv9 and on minimal.
if test "x$OPENJDK_TARGET_OS" != xlinux || test "x$OPENJDK_TARGET_CPU" != xsparcv9; then
NON_MINIMAL_FEATURES="$NON_MINIMAL_FEATURES jfr"
# Enable JFR by default, except for Zero, linux-sparcv9 and on minimal.
if ! HOTSPOT_CHECK_JVM_VARIANT(zero); then
if test "x$OPENJDK_TARGET_OS" != xlinux || test "x$OPENJDK_TARGET_CPU" != xsparcv9; then
NON_MINIMAL_FEATURES="$NON_MINIMAL_FEATURES jfr"
fi
fi
# Turn on additional features based on other parts of configure

6
make/autoconf/toolchain_windows.m4
View File

@ -25,7 +25,7 @@
################################################################################
# The order of these defines the priority by which we try to find them.
VALID_VS_VERSIONS="2013 2012 2010 2015 2017"
VALID_VS_VERSIONS="2017 2013 2015 2012 2010"
VS_DESCRIPTION_2010="Microsoft Visual Studio 2010"
VS_VERSION_INTERNAL_2010=100
@ -58,7 +58,7 @@ VS_SDK_INSTALLDIR_2013=
VS_VS_PLATFORM_NAME_2013="v120"
VS_SDK_PLATFORM_NAME_2013=
VS_DESCRIPTION_2015="Microsoft Visual Studio 2015 - CURRENTLY NOT WORKING"
VS_DESCRIPTION_2015="Microsoft Visual Studio 2015"
VS_VERSION_INTERNAL_2015=140
VS_MSVCR_2015=vcruntime140.dll
VS_MSVCP_2015=msvcp140.dll
@ -71,7 +71,7 @@ VS_SDK_PLATFORM_NAME_2015=
# explicitly specifying Windows Kit 8.1 to be used.
VS_ENV_ARGS_2015="8.1"
VS_DESCRIPTION_2017="Microsoft Visual Studio 2017 - CURRENTLY NOT WORKING"
VS_DESCRIPTION_2017="Microsoft Visual Studio 2017"
VS_VERSION_INTERNAL_2017=141
VS_MSVCR_2017=vcruntime140.dll
VS_MSVCP_2017=msvcp140.dll

5
make/conf/jib-profiles.js
View File

@ -184,7 +184,8 @@ var getJibProfiles = function (input) {
// Identifies the version of this format to the tool reading it.
// 1.1 signifies that the publish, publish-src and get-src features are usable.
data.format_version = "1.1";
// 1.2 signifies that artifact uploads should fail on missing artifacts by default.
data.format_version = "1.2";
// Organization, product and version are used when uploading/publishing build results
data.organization = "";
@ -822,7 +823,7 @@ var getJibProfilesDependencies = function (input, common) {
macosx_x64: "Xcode6.3-MacOSX10.9+1.0",
solaris_x64: "SS12u4-Solaris11u1+1.0",
solaris_sparcv9: "SS12u4-Solaris11u1+1.1",
windows_x64: "VS2013SP4+1.0",
windows_x64: "VS2017-15.5.5+1.0",
linux_aarch64: (input.profile != null && input.profile.indexOf("arm64") >= 0
? "gcc-linaro-aarch64-linux-gnu-4.8-2013.11_linux+1.0"
: "gcc7.3.0-Fedora27+1.0"),

59
make/hotspot/gensrc/GensrcJfr.gmk
View File

@ -23,13 +23,11 @@
# questions.
#
$(eval $(call IncludeCustomExtension, hotspot/gensrc/GensrcJfr.gmk))
################################################################################
# Build tools needed for the Jfr source code generation
# Build tools needed for the JFR source code generation
JFR_TOOLS_SRCDIR := $(TOPDIR)/src/hotspot/share/jfr/metadata
JFR_TOOLS_OUTPUTDIR := $(JVM_VARIANT_OUTPUTDIR)/tools/jfr
JFR_TOOLS_SRCDIR := $(TOPDIR)/make/src/classes
JFR_TOOLS_OUTPUTDIR := $(OUTPUTDIR)/buildtools/tools_classes
$(eval $(call SetupJavaCompiler, GENERATE_JFRBYTECODE, \
JAVAC := $(JAVAC), \
@ -42,53 +40,28 @@ $(eval $(call SetupJavaCompiler, GENERATE_JFRBYTECODE, \
$(eval $(call SetupJavaCompilation, BUILD_JFR_TOOLS, \
SETUP := GENERATE_JFRBYTECODE, \
SRC := $(JFR_TOOLS_SRCDIR), \
INCLUDE_FILES := GenerateJfrFiles.java, \
BIN := $(JFR_TOOLS_OUTPUTDIR), \
))
TOOL_JFR_GEN := $(JAVA_SMALL) -cp $(JFR_TOOLS_OUTPUTDIR) GenerateJfrFiles
TARGETS += $(BUILD_JFR_TOOLS)
################################################################################
# Setup make rules for Jfr file file generation.
# Setup make rules for JFR gensrc file generation.
#
# Parameter 1 is the name of the rule. This name is used as variable prefix,
# and the targets generated are listed in a variable by that name. This name is
# also used as the name of the output file.
#
# Remaining parameters are named arguments. These include:
# XML_FILE -- The input source file to use
# XSD_FILE -- The input schema for validation
# OUTPUT_DIR -- The directory to put the generated file in
SetupJfrGeneration = $(NamedParamsMacroTemplate)
define SetupJfrGenerationBody
$$($1_OUTPUT_DIR)/$1: $$($1_XML_FILE) $$($1_XSD_FILE) $$(BUILD_JFR_TOOLS)
$$(call LogInfo, Generating $$(@F))
$$(call MakeDir, $$(@D))
$$(call ExecuteWithLog, $$@, $$(TOOL_JFR_GEN) $$($1_XML_FILE) $$($1_XSD_FILE) $$($1_OUTPUT_DIR))
test -f $$@
TARGETS += $$($1_OUTPUT_DIR)/$1
endef
################################################################################
# Create files in gensrc/jfrfiles
TOOL_JFR_GEN := $(JAVA_SMALL) -cp $(JFR_TOOLS_OUTPUTDIR) build.tools.jfr.GenerateJfrFiles
JFR_OUTPUTDIR := $(JVM_VARIANT_OUTPUTDIR)/gensrc/jfrfiles
JFR_SRCDIR := $(TOPDIR)/src/hotspot/share/jfr/metadata
METADATA_XML ?= $(JFR_SRCDIR)/metadata.xml
METADATA_XSD ?= $(JFR_SRCDIR)/metadata.xsd
# Changing these will trigger a rebuild of generated jfr files.
JFR_DEPS += \
$(METADATA_XML) \
$(METADATA_XSD) \
#
METADATA_XML := $(JFR_SRCDIR)/metadata.xml
METADATA_XSD := $(JFR_SRCDIR)/metadata.xsd
# our generator will generate all files in one go, so only need to setup one target rule
$(eval $(call SetupJfrGeneration, jfrEventClasses.hpp, \
XML_FILE := $(METADATA_XML), \
XSD_FILE := $(METADATA_XSD), \
OUTPUT_DIR := $(JFR_OUTPUTDIR), \
))
$(JFR_OUTPUTDIR)/jfrEventClasses.hpp: $(METADATA_XML) $(METADATA_XSD) \
$(BUILD_JFR_TOOLS)
$(call LogInfo, Generating $(@F))
$(call MakeDir, $(@D))
$(call ExecuteWithLog, $@, $(TOOL_JFR_GEN) $(METADATA_XML) $(METADATA_XSD) $(JFR_OUTPUTDIR))
test -f $@
TARGETS += $(JFR_OUTPUTDIR)/jfrEventClasses.hpp

2
make/hotspot/lib/JvmFeatures.gmk
View File

@ -41,7 +41,7 @@ ifeq ($(call check-jvm-feature, compiler2), true)
else
JVM_EXCLUDES += opto libadt
JVM_EXCLUDE_FILES += bcEscapeAnalyzer.cpp ciTypeFlow.cpp
JVM_EXCLUDE_PATTERNS += c2_ runtime_
JVM_EXCLUDE_PATTERNS += c2_ runtime_ /c2/
endif
ifeq ($(call check-jvm-feature, zero), true)

1
make/lib/CoreLibraries.gmk
View File

@ -188,7 +188,6 @@ $(eval $(call SetupJdkLibrary, BUILD_LIBZIP, \
-I$(TOPDIR)/src/java.base/$(OPENJDK_TARGET_OS_TYPE)/native/libjava \
-I$(SUPPORT_OUTPUTDIR)/headers/java.base, \
CFLAGS_unix := $(BUILD_LIBZIP_MMAP) -UDEBUG, \
DISABLED_WARNINGS_gcc := implicit-fallthrough, \
LDFLAGS := $(LDFLAGS_JDKLIB) \
$(call SET_SHARED_LIBRARY_ORIGIN), \
LIBS_unix := -ljvm -ljava $(LIBZ_LIBS), \

2
src/hotspot/share/jfr/metadata/GenerateJfrFiles.java → make/src/classes/build/tools/jfr/GenerateJfrFiles.java
View File

@ -1,3 +1,5 @@
package build.tools.jfr;
import java.io.BufferedOutputStream;
import java.io.File;
import java.io.FileNotFoundException;

1366
make/test/JtregNativeHotspot.gmk
View File

File diff suppressed because it is too large Load Diff

2
src/hotspot/cpu/aarch64/aarch64.ad
View File

@ -3665,7 +3665,7 @@ const bool Matcher::strict_fp_requires_explicit_rounding = false;
// Are floats converted to double when stored to stack during
// deoptimization?
bool Matcher::float_in_double() { return true; }
bool Matcher::float_in_double() { return false; }
// Do ints take an entire long register or just half?
// The relevant question is how the int is callee-saved:

14
src/hotspot/cpu/aarch64/c1_LIRGenerator_aarch64.cpp
View File

@ -1284,17 +1284,19 @@ void LIRGenerator::do_If(If* x) {
yin->load_item();
}
// add safepoint before generating condition code so it can be recomputed
if (x->is_safepoint()) {
// increment backedge counter if needed
increment_backedge_counter(state_for(x, x->state_before()), x->profiled_bci());
__ safepoint(LIR_OprFact::illegalOpr, state_for(x, x->state_before()));
}
set_no_result(x);
LIR_Opr left = xin->result();
LIR_Opr right = yin->result();
// add safepoint before generating condition code so it can be recomputed
if (x->is_safepoint()) {
// increment backedge counter if needed
increment_backedge_counter_conditionally(lir_cond(cond), left, right, state_for(x, x->state_before()),
x->tsux()->bci(), x->fsux()->bci(), x->profiled_bci());
__ safepoint(LIR_OprFact::illegalOpr, state_for(x, x->state_before()));
}
__ cmp(lir_cond(cond), left, right);
// Generate branch profiling. Profiling code doesn't kill flags.
profile_branch(x, cond);

6
src/hotspot/cpu/aarch64/macroAssembler_aarch64.cpp
View File

@ -1224,7 +1224,6 @@ void MacroAssembler::check_klass_subtype_slow_path(Register sub_klass,
assert(sub_klass != r0, "killed reg"); // killed by mov(r0, super)
assert(sub_klass != r2, "killed reg"); // killed by lea(r2, &pst_counter)
// Get super_klass value into r0 (even if it was in r5 or r2).
RegSet pushed_registers;
if (!IS_A_TEMP(r2)) pushed_registers += r2;
if (!IS_A_TEMP(r5)) pushed_registers += r5;
@ -1235,6 +1234,11 @@ void MacroAssembler::check_klass_subtype_slow_path(Register sub_klass,
push(pushed_registers, sp);
// Get super_klass value into r0 (even if it was in r5 or r2).
if (super_klass != r0) {
mov(r0, super_klass);
}
#ifndef PRODUCT
mov(rscratch2, (address)&SharedRuntime::_partial_subtype_ctr);
Address pst_counter_addr(rscratch2);

2
src/hotspot/cpu/aarch64/sharedRuntime_aarch64.cpp
View File

@ -2937,7 +2937,7 @@ SafepointBlob* SharedRuntime::generate_handler_blob(address call_ptr, int poll_t
// Exception pending
RegisterSaver::restore_live_registers(masm);
RegisterSaver::restore_live_registers(masm, save_vectors);
__ far_jump(RuntimeAddress(StubRoutines::forward_exception_entry()));

16
src/hotspot/cpu/aarch64/stubGenerator_aarch64.cpp
View File

@ -1351,7 +1351,7 @@ class StubGenerator: public StubCodeGenerator {
BLOCK_COMMENT("Entry:");
}
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -1425,7 +1425,7 @@ class StubGenerator: public StubCodeGenerator {
__ cmp(rscratch1, count, Assembler::LSL, exact_log2(size));
__ br(Assembler::HS, nooverlap_target);
DecoratorSet decorators = 0;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -1789,7 +1789,7 @@ class StubGenerator: public StubCodeGenerator {
}
#endif //ASSERT
DecoratorSet decorators = ARRAYCOPY_CHECKCAST;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_CHECKCAST;
bool is_oop = true;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
@ -3851,7 +3851,9 @@ class StubGenerator: public StubCodeGenerator {
__ sub(cnt1, cnt1, 8 * wordSize);
__ eor(tmp7, tmp7, tmp8);
__ ldp(tmp2, tmp4, Address(__ post(a2, 2 * wordSize)));
__ cmp(cnt1, loopThreshold);
// tmp6 is not used. MacroAssembler::subs is used here (rather than
// cmp) because subs allows an unlimited range of immediate operand.
__ subs(tmp6, cnt1, loopThreshold);
__ orr(tmp5, tmp5, tmp7);
__ cbnz(tmp5, NOT_EQUAL);
__ br(__ GE, LOOP);
@ -3877,7 +3879,7 @@ class StubGenerator: public StubCodeGenerator {
__ ld1(v0, v1, v2, v3, __ T2D, Address(__ post(a1, 4 * 2 * wordSize)));
__ sub(cnt1, cnt1, 8 * wordSize);
__ ld1(v4, v5, v6, v7, __ T2D, Address(__ post(a2, 4 * 2 * wordSize)));
__ cmp(cnt1, loopThreshold);
__ subs(tmp1, cnt1, loopThreshold);
__ eor(v0, __ T16B, v0, v4);
__ eor(v1, __ T16B, v1, v5);
__ eor(v2, __ T16B, v2, v6);
@ -3936,7 +3938,7 @@ class StubGenerator: public StubCodeGenerator {
}
if (UseSIMDForArrayEquals) {
if (SoftwarePrefetchHintDistance >= 0) {
__ cmp(cnt1, prefetchLoopThreshold);
__ subs(tmp1, cnt1, prefetchLoopThreshold);
__ br(__ LE, NO_PREFETCH_LARGE_LOOP);
generate_large_array_equals_loop_simd(prefetchLoopThreshold,
/* prfm = */ true, NOT_EQUAL);
@ -3949,7 +3951,7 @@ class StubGenerator: public StubCodeGenerator {
} else {
__ push(spilled_regs, sp);
if (SoftwarePrefetchHintDistance >= 0) {
__ cmp(cnt1, prefetchLoopThreshold);
__ subs(tmp1, cnt1, prefetchLoopThreshold);
__ br(__ LE, NO_PREFETCH_LARGE_LOOP);
generate_large_array_equals_loop_nonsimd(prefetchLoopThreshold,
/* prfm = */ true, NOT_EQUAL);

3
src/hotspot/cpu/arm/c1_LIRGenerator_arm.cpp
View File

@ -1522,7 +1522,8 @@ void LIRGenerator::do_If(If* x) {
// add safepoint before generating condition code so it can be recomputed
if (x->is_safepoint()) {
increment_backedge_counter(state_for(x, x->state_before()), x->profiled_bci());
increment_backedge_counter_conditionally(lir_cond(cond), left, right, state_for(x, x->state_before()),
x->tsux()->bci(), x->fsux()->bci(), x->profiled_bci());
__ safepoint(LIR_OprFact::illegalOpr, state_for(x, x->state_before()));
}

4
src/hotspot/cpu/arm/stubGenerator_arm.cpp
View File

@ -2945,7 +2945,7 @@ class StubGenerator: public StubCodeGenerator {
__ push(LR);
#endif // AARCH64
DecoratorSet decorators = 0;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY;
if (disjoint) {
decorators |= ARRAYCOPY_DISJOINT;
}
@ -3217,7 +3217,7 @@ class StubGenerator: public StubCodeGenerator {
pushed+=1;
#endif // AARCH64
DecoratorSet decorators = ARRAYCOPY_CHECKCAST;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_CHECKCAST;
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->arraycopy_prologue(_masm, decorators, true, to, count, callee_saved_regs);

3
src/hotspot/cpu/ppc/c1_LIRGenerator_ppc.cpp
View File

@ -1157,7 +1157,8 @@ void LIRGenerator::do_If(If* x) {
// Add safepoint before generating condition code so it can be recomputed.
if (x->is_safepoint()) {
// Increment backedge counter if needed.
increment_backedge_counter(state_for(x, x->state_before()), x->profiled_bci());
increment_backedge_counter_conditionally(lir_cond(cond), left, right, state_for(x, x->state_before()),
x->tsux()->bci(), x->fsux()->bci(), x->profiled_bci());
__ safepoint(safepoint_poll_register(), state_for(x, x->state_before()));
}

6
src/hotspot/cpu/ppc/gc/g1/g1BarrierSetAssembler_ppc.cpp
View File

@ -304,15 +304,15 @@ void G1BarrierSetAssembler::oop_store_at(MacroAssembler* masm, DecoratorSet deco
void G1BarrierSetAssembler::load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
Register base, RegisterOrConstant ind_or_offs, Register dst,
Register tmp1, Register tmp2, bool needs_frame, Label *is_null) {
Register tmp1, Register tmp2, bool needs_frame, Label *L_handle_null) {
bool on_oop = type == T_OBJECT || type == T_ARRAY;
bool on_weak = (decorators & ON_WEAK_OOP_REF) != 0;
bool on_phantom = (decorators & ON_PHANTOM_OOP_REF) != 0;
bool on_reference = on_weak || on_phantom;
Label done;
if (on_oop && on_reference && is_null == NULL) { is_null = &done; }
if (on_oop && on_reference && L_handle_null == NULL) { L_handle_null = &done; }
// Load the value of the referent field.
ModRefBarrierSetAssembler::load_at(masm, decorators, type, base, ind_or_offs, dst, tmp1, tmp2, needs_frame, is_null);
ModRefBarrierSetAssembler::load_at(masm, decorators, type, base, ind_or_offs, dst, tmp1, tmp2, needs_frame, L_handle_null);
if (on_oop && on_reference) {
// Generate the G1 pre-barrier code to log the value of
// the referent field in an SATB buffer. Note with

2
src/hotspot/cpu/ppc/gc/g1/g1BarrierSetAssembler_ppc.hpp
View File

@ -61,7 +61,7 @@ public:
virtual void load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
Register base, RegisterOrConstant ind_or_offs, Register dst,
Register tmp1, Register tmp2, bool needs_frame, Label *is_null = NULL);
Register tmp1, Register tmp2, bool needs_frame, Label *L_handle_null = NULL);
virtual void resolve_jobject(MacroAssembler* masm, Register value, Register tmp1, Register tmp2, bool needs_frame);
};

24
src/hotspot/cpu/ppc/gc/shared/barrierSetAssembler_ppc.cpp
View File

@ -65,9 +65,10 @@ void BarrierSetAssembler::store_at(MacroAssembler* masm, DecoratorSet decorators
void BarrierSetAssembler::load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
Register base, RegisterOrConstant ind_or_offs, Register dst,
Register tmp1, Register tmp2, bool needs_frame, Label *is_null) {
bool on_heap = (decorators & IN_HEAP) != 0;
bool on_root = (decorators & IN_ROOT) != 0;
Register tmp1, Register tmp2, bool needs_frame, Label *L_handle_null) {
bool on_heap = (decorators & IN_HEAP) != 0;
bool on_root = (decorators & IN_ROOT) != 0;
bool not_null = (decorators & OOP_NOT_NULL) != 0;
assert(on_heap || on_root, "where?");
assert_different_registers(ind_or_offs.register_or_noreg(), dst, R0);
@ -75,19 +76,24 @@ void BarrierSetAssembler::load_at(MacroAssembler* masm, DecoratorSet decorators,
case T_ARRAY:
case T_OBJECT: {
if (UseCompressedOops && on_heap) {
__ lwz(dst, ind_or_offs, base);
if (is_null) {
if (L_handle_null != NULL) { // Label provided.
__ lwz(dst, ind_or_offs, base);
__ cmpwi(CCR0, dst, 0);
__ beq(CCR0, *is_null);
__ beq(CCR0, *L_handle_null);
__ decode_heap_oop_not_null(dst);
} else {
} else if (not_null) { // Guaranteed to be not null.
Register narrowOop = (tmp1 != noreg && Universe::narrow_oop_base_disjoint()) ? tmp1 : dst;
__ lwz(narrowOop, ind_or_offs, base);
__ decode_heap_oop_not_null(dst, narrowOop);
} else { // Any oop.
__ lwz(dst, ind_or_offs, base);
__ decode_heap_oop(dst);
}
} else {
__ ld(dst, ind_or_offs, base);
if (is_null) {
if (L_handle_null != NULL) {
__ cmpdi(CCR0, dst, 0);
__ beq(CCR0, *is_null);
__ beq(CCR0, *L_handle_null);
}
}
break;

2
src/hotspot/cpu/ppc/gc/shared/barrierSetAssembler_ppc.hpp
View File

@ -45,7 +45,7 @@ public:
virtual void load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
Register base, RegisterOrConstant ind_or_offs, Register dst,
Register tmp1, Register tmp2, bool needs_frame, Label *is_null = NULL);
Register tmp1, Register tmp2, bool needs_frame, Label *L_handle_null = NULL);
virtual void resolve_jobject(MacroAssembler* masm, Register value, Register tmp1, Register tmp2, bool needs_frame);

2
src/hotspot/cpu/ppc/interp_masm_ppc.hpp
View File

@ -77,7 +77,7 @@ class InterpreterMacroAssembler: public MacroAssembler {
Register tmp1, Register tmp2, Register tmp3, Label &ok_is_subtype);
// Load object from cpool->resolved_references(index).
void load_resolved_reference_at_index(Register result, Register index, Label *is_null = NULL);
void load_resolved_reference_at_index(Register result, Register index, Label *L_handle_null = NULL);
// load cpool->resolved_klass_at(index)
void load_resolved_klass_at_offset(Register Rcpool, Register Roffset, Register Rklass);

5
src/hotspot/cpu/ppc/interp_masm_ppc_64.cpp
View File

@ -471,7 +471,7 @@ void InterpreterMacroAssembler::get_u4(Register Rdst, Register Rsrc, int offset,
}
// Load object from cpool->resolved_references(index).
void InterpreterMacroAssembler::load_resolved_reference_at_index(Register result, Register index, Label *is_null) {
void InterpreterMacroAssembler::load_resolved_reference_at_index(Register result, Register index, Label *L_handle_null) {
assert_different_registers(result, index);
get_constant_pool(result);
@ -494,8 +494,7 @@ void InterpreterMacroAssembler::load_resolved_reference_at_index(Register result
#endif
// Add in the index.
add(result, tmp, result);
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->load_at(this, IN_HEAP, T_OBJECT, result, arrayOopDesc::base_offset_in_bytes(T_OBJECT), result, tmp, R0, false, is_null);
load_heap_oop(result, arrayOopDesc::base_offset_in_bytes(T_OBJECT), result, tmp, R0, false, 0, L_handle_null);
}
// load cpool->resolved_klass_at(index)

3
src/hotspot/cpu/ppc/macroAssembler_ppc.cpp
View File

@ -2045,7 +2045,8 @@ void MacroAssembler::check_method_handle_type(Register mtype_reg, Register mh_re
Label& wrong_method_type) {
assert_different_registers(mtype_reg, mh_reg, temp_reg);
// Compare method type against that of the receiver.
load_heap_oop_not_null(temp_reg, delayed_value(java_lang_invoke_MethodHandle::type_offset_in_bytes, temp_reg), mh_reg);
load_heap_oop(temp_reg, delayed_value(java_lang_invoke_MethodHandle::type_offset_in_bytes, temp_reg), mh_reg,
noreg, noreg, false, OOP_NOT_NULL);
cmpd(CCR0, temp_reg, mtype_reg);
bne(CCR0, wrong_method_type);
}

30
src/hotspot/cpu/ppc/macroAssembler_ppc.hpp
View File

@ -27,6 +27,7 @@
#define CPU_PPC_VM_MACROASSEMBLER_PPC_HPP
#include "asm/assembler.hpp"
#include "oops/accessDecorators.hpp"
#include "runtime/rtmLocking.hpp"
#include "utilities/macros.hpp"
@ -691,17 +692,26 @@ class MacroAssembler: public Assembler {
inline void null_check_throw(Register a, int offset, Register temp_reg, address exception_entry);
inline void null_check(Register a, int offset, Label *Lis_null); // implicit only if Lis_null not provided
// Load heap oop and decompress. Loaded oop may not be null.
// Specify tmp to save one cycle.
inline void load_heap_oop_not_null(Register d, RegisterOrConstant offs, Register s1 = noreg,
Register tmp = noreg);
// Store heap oop and decompress. Decompressed oop may not be null.
// Specify tmp register if d should not be changed.
inline void store_heap_oop_not_null(Register d, RegisterOrConstant offs, Register s1,
Register tmp = noreg);
// Access heap oop, handle encoding and GC barriers.
// Some GC barriers call C so use needs_frame = true if an extra frame is needed at the current call site.
private:
inline void access_store_at(BasicType type, DecoratorSet decorators,
Register base, RegisterOrConstant ind_or_offs, Register val,
Register tmp1, Register tmp2, Register tmp3, bool needs_frame);
inline void access_load_at(BasicType type, DecoratorSet decorators,
Register base, RegisterOrConstant ind_or_offs, Register dst,
Register tmp1, Register tmp2, bool needs_frame, Label *L_handle_null = NULL);
// Null allowed.
inline void load_heap_oop(Register d, RegisterOrConstant offs, Register s1 = noreg, Label *is_null = NULL);
public:
// Specify tmp1 for better code in certain compressed oops cases. Specify Label to bail out on null oop.
// tmp1, tmp2 and needs_frame are used with decorators ON_PHANTOM_OOP_REF or ON_WEAK_OOP_REF.
inline void load_heap_oop(Register d, RegisterOrConstant offs, Register s1,
Register tmp1, Register tmp2, bool needs_frame,
DecoratorSet decorators = 0, Label *L_handle_null = NULL);
inline void store_heap_oop(Register d, RegisterOrConstant offs, Register s1,
Register tmp1, Register tmp2, Register tmp3, bool needs_frame,
DecoratorSet decorators = 0);
// Encode/decode heap oop. Oop may not be null, else en/decoding goes wrong.
// src == d allowed.

71
src/hotspot/cpu/ppc/macroAssembler_ppc.inline.hpp
View File

@ -30,6 +30,8 @@
#include "asm/macroAssembler.hpp"
#include "asm/codeBuffer.hpp"
#include "code/codeCache.hpp"
#include "gc/shared/barrierSet.hpp"
#include "gc/shared/barrierSetAssembler.hpp"
#include "runtime/safepointMechanism.hpp"
inline bool MacroAssembler::is_ld_largeoffset(address a) {
@ -323,45 +325,52 @@ inline void MacroAssembler::null_check(Register a, int offset, Label *Lis_null)
}
}
inline void MacroAssembler::load_heap_oop_not_null(Register d, RegisterOrConstant offs, Register s1, Register tmp) {
if (UseCompressedOops) {
// In disjoint mode decoding can save a cycle if src != dst.
Register narrowOop = (tmp != noreg && Universe::narrow_oop_base_disjoint()) ? tmp : d;
lwz(narrowOop, offs, s1);
// Attention: no null check here!
Register res = decode_heap_oop_not_null(d, narrowOop);
assert(res == d, "caller will not consume loaded value");
inline void MacroAssembler::access_store_at(BasicType type, DecoratorSet decorators,
Register base, RegisterOrConstant ind_or_offs, Register val,
Register tmp1, Register tmp2, Register tmp3, bool needs_frame) {
assert((decorators & ~(AS_RAW | IN_HEAP | IN_HEAP_ARRAY | IN_ROOT | OOP_NOT_NULL |
ON_UNKNOWN_OOP_REF)) == 0, "unsupported decorator");
BarrierSetAssembler* bs = BarrierSet::barrier_set()->barrier_set_assembler();
bool as_raw = (decorators & AS_RAW) != 0;
if (as_raw) {
bs->BarrierSetAssembler::store_at(this, decorators, type,
base, ind_or_offs, val,
tmp1, tmp2, tmp3, needs_frame);
} else {
ld(d, offs, s1);
bs->store_at(this, decorators, type,
base, ind_or_offs, val,
tmp1, tmp2, tmp3, needs_frame);
}
}
inline void MacroAssembler::store_heap_oop_not_null(Register d, RegisterOrConstant offs, Register s1, Register tmp) {
if (UseCompressedOops) {
Register compressedOop = encode_heap_oop_not_null((tmp != noreg) ? tmp : d, d);
stw(compressedOop, offs, s1);
inline void MacroAssembler::access_load_at(BasicType type, DecoratorSet decorators,
Register base, RegisterOrConstant ind_or_offs, Register dst,
Register tmp1, Register tmp2, bool needs_frame, Label *L_handle_null) {
assert((decorators & ~(AS_RAW | IN_HEAP | IN_HEAP_ARRAY | IN_ROOT | OOP_NOT_NULL |
ON_PHANTOM_OOP_REF | ON_WEAK_OOP_REF)) == 0, "unsupported decorator");
BarrierSetAssembler* bs = BarrierSet::barrier_set()->barrier_set_assembler();
bool as_raw = (decorators & AS_RAW) != 0;
if (as_raw) {
bs->BarrierSetAssembler::load_at(this, decorators, type,
base, ind_or_offs, dst,
tmp1, tmp2, needs_frame, L_handle_null);
} else {
std(d, offs, s1);
bs->load_at(this, decorators, type,
base, ind_or_offs, dst,
tmp1, tmp2, needs_frame, L_handle_null);
}
}
inline void MacroAssembler::load_heap_oop(Register d, RegisterOrConstant offs, Register s1, Label *is_null) {
if (UseCompressedOops) {
lwz(d, offs, s1);
if (is_null != NULL) {
cmpwi(CCR0, d, 0);
beq(CCR0, *is_null);
decode_heap_oop_not_null(d);
} else {
decode_heap_oop(d);
}
} else {
ld(d, offs, s1);
if (is_null != NULL) {
cmpdi(CCR0, d, 0);
beq(CCR0, *is_null);
}
}
inline void MacroAssembler::load_heap_oop(Register d, RegisterOrConstant offs, Register s1,
Register tmp1, Register tmp2,
bool needs_frame, DecoratorSet decorators, Label *L_handle_null) {
access_load_at(T_OBJECT, IN_HEAP | decorators, s1, offs, d, tmp1, tmp2, needs_frame, L_handle_null);
}
inline void MacroAssembler::store_heap_oop(Register d, RegisterOrConstant offs, Register s1,
Register tmp1, Register tmp2, Register tmp3,
bool needs_frame, DecoratorSet decorators) {
access_store_at(T_OBJECT, IN_HEAP | decorators, s1, offs, d, tmp1, tmp2, tmp3, needs_frame);
}
inline Register MacroAssembler::encode_heap_oop_not_null(Register d, Register src) {

21
src/hotspot/cpu/ppc/methodHandles_ppc.cpp
View File

@ -173,11 +173,14 @@ void MethodHandles::jump_to_lambda_form(MacroAssembler* _masm,
// Load the invoker, as MH -> MH.form -> LF.vmentry
__ verify_oop(recv);
__ load_heap_oop_not_null(method_temp, NONZERO(java_lang_invoke_MethodHandle::form_offset_in_bytes()), recv, temp2);
__ load_heap_oop(method_temp, NONZERO(java_lang_invoke_MethodHandle::form_offset_in_bytes()), recv,
temp2, noreg, false, OOP_NOT_NULL);
__ verify_oop(method_temp);
__ load_heap_oop_not_null(method_temp, NONZERO(java_lang_invoke_LambdaForm::vmentry_offset_in_bytes()), method_temp, temp2);
__ load_heap_oop(method_temp, NONZERO(java_lang_invoke_LambdaForm::vmentry_offset_in_bytes()), method_temp,
temp2, noreg, false, OOP_NOT_NULL);
__ verify_oop(method_temp);
__ load_heap_oop_not_null(method_temp, NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes()), method_temp);
__ load_heap_oop(method_temp, NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes()), method_temp,
temp2, noreg, false, OOP_NOT_NULL);
__ verify_oop(method_temp);
__ ld(method_temp, NONZERO(java_lang_invoke_ResolvedMethodName::vmtarget_offset_in_bytes()), method_temp);
@ -338,7 +341,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
if (VerifyMethodHandles && iid != vmIntrinsics::_linkToInterface) {
Label L_ok;
Register temp2_defc = temp2;
__ load_heap_oop_not_null(temp2_defc, NONZERO(java_lang_invoke_MemberName::clazz_offset_in_bytes()), member_reg, temp3);
__ load_heap_oop(temp2_defc, NONZERO(java_lang_invoke_MemberName::clazz_offset_in_bytes()), member_reg,
temp3, noreg, false, OOP_NOT_NULL);
load_klass_from_Class(_masm, temp2_defc, temp3, temp4);
__ verify_klass_ptr(temp2_defc);
__ check_klass_subtype(temp1_recv_klass, temp2_defc, temp3, temp4, L_ok);
@ -365,7 +369,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
if (VerifyMethodHandles) {
verify_ref_kind(_masm, JVM_REF_invokeSpecial, member_reg, temp2);
}
__ load_heap_oop(R19_method, NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes()), member_reg);
__ load_heap_oop(R19_method, NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes()), member_reg,
temp3, noreg, false, OOP_NOT_NULL);
__ ld(R19_method, NONZERO(java_lang_invoke_ResolvedMethodName::vmtarget_offset_in_bytes()), R19_method);
break;
@ -373,7 +378,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
if (VerifyMethodHandles) {
verify_ref_kind(_masm, JVM_REF_invokeStatic, member_reg, temp2);
}
__ load_heap_oop(R19_method, NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes()), member_reg);
__ load_heap_oop(R19_method, NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes()), member_reg,
temp3, noreg, false, OOP_NOT_NULL);
__ ld(R19_method, NONZERO(java_lang_invoke_ResolvedMethodName::vmtarget_offset_in_bytes()), R19_method);
break;
@ -415,7 +421,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
}
Register temp2_intf = temp2;
__ load_heap_oop_not_null(temp2_intf, NONZERO(java_lang_invoke_MemberName::clazz_offset_in_bytes()), member_reg, temp3);
__ load_heap_oop(temp2_intf, NONZERO(java_lang_invoke_MemberName::clazz_offset_in_bytes()), member_reg,
temp3, noreg, false, OOP_NOT_NULL);
load_klass_from_Class(_masm, temp2_intf, temp3, temp4);
__ verify_klass_ptr(temp2_intf);

8
src/hotspot/cpu/ppc/stubGenerator_ppc.cpp
View File

@ -2024,7 +2024,7 @@ class StubGenerator: public StubCodeGenerator {
STUB_ENTRY(arrayof_oop_disjoint_arraycopy) :
STUB_ENTRY(oop_disjoint_arraycopy);
DecoratorSet decorators = 0;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2063,7 +2063,7 @@ class StubGenerator: public StubCodeGenerator {
address start = __ function_entry();
assert_positive_int(R5_ARG3);
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2159,7 +2159,7 @@ class StubGenerator: public StubCodeGenerator {
}
#endif
DecoratorSet decorators = ARRAYCOPY_CHECKCAST;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_CHECKCAST;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2198,7 +2198,7 @@ class StubGenerator: public StubCodeGenerator {
// ======== loop entry is here ========
__ bind(load_element);
__ load_heap_oop(R10_oop, R8_offset, R3_from, &store_null); // Load the oop.
__ load_heap_oop(R10_oop, R8_offset, R3_from, R12_tmp, noreg, false, AS_RAW, &store_null);
__ load_klass(R11_klass, R10_oop); // Query the object klass.

7
src/hotspot/cpu/ppc/templateInterpreterGenerator_ppc.cpp
View File

@ -524,11 +524,8 @@ address TemplateInterpreterGenerator::generate_Reference_get_entry(void) {
__ cmpdi(CCR0, R3_RET, 0);
__ beq(CCR0, slow_path);
// Load the value of the referent field.
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->load_at(_masm, IN_HEAP | ON_WEAK_OOP_REF, T_OBJECT,
R3_RET, referent_offset, R3_RET,
/* non-volatile temp */ R31, R11_scratch1, true);
__ load_heap_oop(R3_RET, referent_offset, R3_RET,
/* non-volatile temp */ R31, R11_scratch1, true, ON_WEAK_OOP_REF);
// Generate the G1 pre-barrier code to log the value of
// the referent field in an SATB buffer. Note with

6
src/hotspot/cpu/ppc/templateTable_ppc_64.cpp
View File

@ -62,8 +62,7 @@ static void do_oop_store(InterpreterMacroAssembler* _masm,
Register tmp3,
DecoratorSet decorators) {
assert_different_registers(tmp1, tmp2, tmp3, val, base);
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->store_at(_masm, decorators, T_OBJECT, base, offset, val, tmp1, tmp2, tmp3, false);
__ store_heap_oop(val, offset, base, tmp1, tmp2, tmp3, false, decorators);
}
static void do_oop_load(InterpreterMacroAssembler* _masm,
@ -75,8 +74,7 @@ static void do_oop_load(InterpreterMacroAssembler* _masm,
DecoratorSet decorators) {
assert_different_registers(base, tmp1, tmp2);
assert_different_registers(dst, tmp1, tmp2);
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->load_at(_masm, decorators, T_OBJECT, base, offset, dst, tmp1, tmp2, false);
__ load_heap_oop(dst, offset, base, tmp1, tmp2, false, decorators);
}
// ============================================================================

80
src/hotspot/cpu/ppc/vm_version_ext_ppc.cpp Normal file
View File

@ -0,0 +1,80 @@
/*
* Copyright (c) 2013, 2018, 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.
*
*/
#include "jvm.h"
#include "memory/allocation.hpp"
#include "memory/allocation.inline.hpp"
#include "vm_version_ext_ppc.hpp"
// VM_Version_Ext statics
int VM_Version_Ext::_no_of_threads = 0;
int VM_Version_Ext::_no_of_cores = 0;
int VM_Version_Ext::_no_of_sockets = 0;
char VM_Version_Ext::_cpu_name[CPU_TYPE_DESC_BUF_SIZE] = {0};
char VM_Version_Ext::_cpu_desc[CPU_DETAILED_DESC_BUF_SIZE] = {0};
// get cpu information.
bool VM_Version_Ext::initialize_cpu_information(void) {
// Not yet implemented.
return false;
}
int VM_Version_Ext::number_of_threads(void) {
initialize_cpu_information();
return _no_of_threads;
}
int VM_Version_Ext::number_of_cores(void) {
initialize_cpu_information();
return _no_of_cores;
}
int VM_Version_Ext::number_of_sockets(void) {
initialize_cpu_information();
return _no_of_sockets;
}
const char* VM_Version_Ext::cpu_name(void) {
if (!initialize_cpu_information()) {
return NULL;
}
char* tmp = NEW_C_HEAP_ARRAY_RETURN_NULL(char, CPU_TYPE_DESC_BUF_SIZE, mtTracing);
if (NULL == tmp) {
return NULL;
}
strncpy(tmp, _cpu_name, CPU_TYPE_DESC_BUF_SIZE);
return tmp;
}
const char* VM_Version_Ext::cpu_description(void) {
if (!initialize_cpu_information()) {
return NULL;
}
char* tmp = NEW_C_HEAP_ARRAY_RETURN_NULL(char, CPU_DETAILED_DESC_BUF_SIZE, mtTracing);
if (NULL == tmp) {
return NULL;
}
strncpy(tmp, _cpu_desc, CPU_DETAILED_DESC_BUF_SIZE);
return tmp;
}

61
src/hotspot/cpu/ppc/vm_version_ext_ppc.hpp Normal file
View File

@ -0,0 +1,61 @@
/*
* Copyright (c) 2013, 2018, 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.
*
*/
#ifndef CPU_PPC_VM_VM_VERSION_EXT_PPC_HPP
#define CPU_PPC_VM_VM_VERSION_EXT_PPC_HPP
#include "utilities/macros.hpp"
#include "vm_version_ppc.hpp"
#define CPU_INFO "cpu_info"
#define CPU_TYPE "fpu_type"
#define CPU_DESCRIPTION "implementation"
#define CHIP_ID "chip_id"
#define CORE_ID "core_id"
class VM_Version_Ext : public VM_Version {
private:
static const size_t CPU_TYPE_DESC_BUF_SIZE = 256;
static const size_t CPU_DETAILED_DESC_BUF_SIZE = 4096;
static int _no_of_threads;
static int _no_of_cores;
static int _no_of_sockets;
static char _cpu_name[CPU_TYPE_DESC_BUF_SIZE];
static char _cpu_desc[CPU_DETAILED_DESC_BUF_SIZE];
static bool initialize_cpu_information(void);
public:
static int number_of_threads(void);
static int number_of_cores(void);
static int number_of_sockets(void);
static const char* cpu_name(void);
static const char* cpu_description(void);
};
#endif // CPU_PPC_VM_VM_VERSION_EXT_PPC_HPP

60
src/hotspot/cpu/ppc/vm_version_ppc.cpp
View File

@ -37,7 +37,15 @@
#include "utilities/globalDefinitions.hpp"
#include "vm_version_ppc.hpp"
# include <sys/sysinfo.h>
#include <sys/sysinfo.h>
#if defined(LINUX) && defined(VM_LITTLE_ENDIAN)
#include <sys/auxv.h>
#ifndef PPC_FEATURE2_HTM_NOSC
#define PPC_FEATURE2_HTM_NOSC (1 << 24)
#endif
#endif
bool VM_Version::_is_determine_features_test_running = false;
uint64_t VM_Version::_dscr_val = 0;
@ -123,7 +131,7 @@ void VM_Version::initialize() {
// Create and print feature-string.
char buf[(num_features+1) * 16]; // Max 16 chars per feature.
jio_snprintf(buf, sizeof(buf),
"ppc64%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
"ppc64%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
(has_fsqrt() ? " fsqrt" : ""),
(has_isel() ? " isel" : ""),
(has_lxarxeh() ? " lxarxeh" : ""),
@ -136,7 +144,6 @@ void VM_Version::initialize() {
(has_lqarx() ? " lqarx" : ""),
(has_vcipher() ? " aes" : ""),
(has_vpmsumb() ? " vpmsumb" : ""),
(has_tcheck() ? " tcheck" : ""),
(has_mfdscr() ? " mfdscr" : ""),
(has_vsx() ? " vsx" : ""),
(has_ldbrx() ? " ldbrx" : ""),
@ -304,15 +311,15 @@ void VM_Version::initialize() {
// Adjust RTM (Restricted Transactional Memory) flags.
if (UseRTMLocking) {
// If CPU or OS are too old:
// If CPU or OS do not support TM:
// Can't continue because UseRTMLocking affects UseBiasedLocking flag
// setting during arguments processing. See use_biased_locking().
// VM_Version_init() is executed after UseBiasedLocking is used
// in Thread::allocate().
if (!has_tcheck()) {
vm_exit_during_initialization("RTM instructions are not available on this CPU");
if (PowerArchitecturePPC64 < 8) {
vm_exit_during_initialization("RTM instructions are not available on this CPU.");
}
bool os_too_old = true;
bool os_support_tm = false;
#ifdef AIX
// Actually, this is supported since AIX 7.1.. Unfortunately, this first
// contained bugs, so that it can only be enabled after AIX 7.1.3.30.
@ -322,22 +329,25 @@ void VM_Version::initialize() {
// So the tests can not check on subversion 3.30, and we only enable RTM
// with AIX 7.2.
if (os::Aix::os_version() >= 0x07020000) { // At least AIX 7.2.
os_too_old = false;
os_support_tm = true;
}
#endif
#ifdef LINUX
// At least Linux kernel 4.2, as the problematic behavior of syscalls
// being called in the middle of a transaction has been addressed.
// Please, refer to commit b4b56f9ecab40f3b4ef53e130c9f6663be491894
// in Linux kernel source tree: https://goo.gl/Kc5i7A
if (os::Linux::os_version_is_known()) {
if (os::Linux::os_version() >= 0x040200)
os_too_old = false;
} else {
vm_exit_during_initialization("RTM can not be enabled: kernel version is unknown.");
#if defined(LINUX) && defined(VM_LITTLE_ENDIAN)
unsigned long auxv = getauxval(AT_HWCAP2);
if (auxv & PPC_FEATURE2_HTM_NOSC) {
if (auxv & PPC_FEATURE2_HAS_HTM) {
// TM on POWER8 and POWER9 in compat mode (VM) is supported by the JVM.
// TM on POWER9 DD2.1 NV (baremetal) is not supported by the JVM (TM on
// POWER9 DD2.1 NV has a few issues that need a couple of firmware
// and kernel workarounds, so there is a new mode only supported
// on non-virtualized P9 machines called HTM with no Suspend Mode).
// TM on POWER9 D2.2+ NV is not supported at all by Linux.
os_support_tm = true;
}
}
#endif
if (os_too_old) {
if (!os_support_tm) {
vm_exit_during_initialization("RTM is not supported on this OS version.");
}
}
@ -680,12 +690,11 @@ void VM_Version::determine_features() {
a->lqarx_unchecked(R6, R3_ARG1, R4_ARG2, 1); // code[9] -> lqarx_m
a->vcipher(VR0, VR1, VR2); // code[10] -> vcipher
a->vpmsumb(VR0, VR1, VR2); // code[11] -> vpmsumb
a->tcheck(0); // code[12] -> tcheck
a->mfdscr(R0); // code[13] -> mfdscr
a->lxvd2x(VSR0, R3_ARG1); // code[14] -> vsx
a->ldbrx(R7, R3_ARG1, R4_ARG2); // code[15] -> ldbrx
a->stdbrx(R7, R3_ARG1, R4_ARG2); // code[16] -> stdbrx
a->vshasigmaw(VR0, VR1, 1, 0xF); // code[17] -> vshasig
a->mfdscr(R0); // code[12] -> mfdscr
a->lxvd2x(VSR0, R3_ARG1); // code[13] -> vsx
a->ldbrx(R7, R3_ARG1, R4_ARG2); // code[14] -> ldbrx
a->stdbrx(R7, R3_ARG1, R4_ARG2); // code[15] -> stdbrx
a->vshasigmaw(VR0, VR1, 1, 0xF); // code[16] -> vshasig
a->blr();
// Emit function to set one cache line to zero. Emit function descriptor and get pointer to it.
@ -732,7 +741,6 @@ void VM_Version::determine_features() {
if (code[feature_cntr++]) features |= lqarx_m;
if (code[feature_cntr++]) features |= vcipher_m;
if (code[feature_cntr++]) features |= vpmsumb_m;
if (code[feature_cntr++]) features |= tcheck_m;
if (code[feature_cntr++]) features |= mfdscr_m;
if (code[feature_cntr++]) features |= vsx_m;
if (code[feature_cntr++]) features |= ldbrx_m;

3
src/hotspot/cpu/ppc/vm_version_ppc.hpp
View File

@ -44,7 +44,6 @@ protected:
lqarx,
vcipher,
vpmsumb,
tcheck,
mfdscr,
vsx,
ldbrx,
@ -67,7 +66,6 @@ protected:
vcipher_m = (1 << vcipher),
vshasig_m = (1 << vshasig),
vpmsumb_m = (1 << vpmsumb),
tcheck_m = (1 << tcheck ),
mfdscr_m = (1 << mfdscr ),
vsx_m = (1 << vsx ),
ldbrx_m = (1 << ldbrx ),
@ -103,7 +101,6 @@ public:
static bool has_lqarx() { return (_features & lqarx_m) != 0; }
static bool has_vcipher() { return (_features & vcipher_m) != 0; }
static bool has_vpmsumb() { return (_features & vpmsumb_m) != 0; }
static bool has_tcheck() { return (_features & tcheck_m) != 0; }
static bool has_mfdscr() { return (_features & mfdscr_m) != 0; }
static bool has_vsx() { return (_features & vsx_m) != 0; }
static bool has_ldbrx() { return (_features & ldbrx_m) != 0; }

11
src/hotspot/cpu/s390/c1_LIRGenerator_s390.cpp
View File

@ -963,17 +963,20 @@ void LIRGenerator::do_If (If* x) {
yin->dont_load_item();
}
LIR_Opr left = xin->result();
LIR_Opr right = yin->result();
set_no_result(x);
// Add safepoint before generating condition code so it can be recomputed.
if (x->is_safepoint()) {
// Increment backedge counter if needed.
increment_backedge_counter(state_for (x, x->state_before()), x->profiled_bci());
increment_backedge_counter_conditionally(lir_cond(cond), left, right, state_for(x, x->state_before()),
x->tsux()->bci(), x->fsux()->bci(), x->profiled_bci());
// Use safepoint_poll_register() instead of LIR_OprFact::illegalOpr.
__ safepoint(safepoint_poll_register(), state_for (x, x->state_before()));
}
set_no_result(x);
LIR_Opr left = xin->result();
LIR_Opr right = yin->result();
__ cmp(lir_cond(cond), left, right);
// Generate branch profiling. Profiling code doesn't kill flags.
profile_branch(x, cond);

6
src/hotspot/cpu/s390/gc/g1/g1BarrierSetAssembler_s390.cpp
View File

@ -96,14 +96,14 @@ void G1BarrierSetAssembler::gen_write_ref_array_post_barrier(MacroAssembler* mas
}
void G1BarrierSetAssembler::load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
const Address& src, Register dst, Register tmp1, Register tmp2, Label *is_null) {
const Address& src, Register dst, Register tmp1, Register tmp2, Label *L_handle_null) {
bool on_oop = type == T_OBJECT || type == T_ARRAY;
bool on_weak = (decorators & ON_WEAK_OOP_REF) != 0;
bool on_phantom = (decorators & ON_PHANTOM_OOP_REF) != 0;
bool on_reference = on_weak || on_phantom;
Label done;
if (on_oop && on_reference && is_null == NULL) { is_null = &done; }
ModRefBarrierSetAssembler::load_at(masm, decorators, type, src, dst, tmp1, tmp2, is_null);
if (on_oop && on_reference && L_handle_null == NULL) { L_handle_null = &done; }
ModRefBarrierSetAssembler::load_at(masm, decorators, type, src, dst, tmp1, tmp2, L_handle_null);
if (on_oop && on_reference) {
// Generate the G1 pre-barrier code to log the value of
// the referent field in an SATB buffer.

2
src/hotspot/cpu/s390/gc/g1/g1BarrierSetAssembler_s390.hpp
View File

@ -65,7 +65,7 @@ class G1BarrierSetAssembler: public ModRefBarrierSetAssembler {
#endif
virtual void load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
const Address& src, Register dst, Register tmp1, Register tmp2, Label *is_null = NULL);
const Address& src, Register dst, Register tmp1, Register tmp2, Label *L_handle_null = NULL);
virtual void resolve_jobject(MacroAssembler* masm, Register value, Register tmp1, Register tmp2);
};

17
src/hotspot/cpu/s390/gc/shared/barrierSetAssembler_s390.cpp
View File

@ -36,9 +36,10 @@ void BarrierSetAssembler::arraycopy_epilogue(MacroAssembler* masm, DecoratorSet
}
void BarrierSetAssembler::load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
const Address& addr, Register dst, Register tmp1, Register tmp2, Label *is_null) {
bool on_heap = (decorators & IN_HEAP) != 0;
bool on_root = (decorators & IN_ROOT) != 0;
const Address& addr, Register dst, Register tmp1, Register tmp2, Label *L_handle_null) {
bool on_heap = (decorators & IN_HEAP) != 0;
bool on_root = (decorators & IN_ROOT) != 0;
bool not_null = (decorators & OOP_NOT_NULL) != 0;
assert(on_heap || on_root, "where?");
switch (type) {
@ -46,16 +47,16 @@ void BarrierSetAssembler::load_at(MacroAssembler* masm, DecoratorSet decorators,
case T_OBJECT: {
if (UseCompressedOops && on_heap) {
__ z_llgf(dst, addr);
if (is_null) {
__ compareU32_and_branch(dst, (intptr_t)0, Assembler::bcondEqual, *is_null);
if (L_handle_null != NULL) { // Label provided.
__ compareU32_and_branch(dst, (intptr_t)0, Assembler::bcondEqual, *L_handle_null);
__ oop_decoder(dst, dst, false);
} else {
__ oop_decoder(dst, dst, true);
__ oop_decoder(dst, dst, !not_null);
}
} else {
__ z_lg(dst, addr);
if (is_null) {
__ compareU64_and_branch(dst, (intptr_t)0, Assembler::bcondEqual, *is_null);
if (L_handle_null != NULL) {
__ compareU64_and_branch(dst, (intptr_t)0, Assembler::bcondEqual, *L_handle_null);
}
}
break;

2
src/hotspot/cpu/s390/gc/shared/barrierSetAssembler_s390.hpp
View File

@ -40,7 +40,7 @@ public:
Register dst, Register count, bool do_return = false);
virtual void load_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
const Address& addr, Register dst, Register tmp1, Register tmp2, Label *is_null = NULL);
const Address& addr, Register dst, Register tmp1, Register tmp2, Label *L_handle_null = NULL);
virtual void store_at(MacroAssembler* masm, DecoratorSet decorators, BasicType type,
const Address& addr, Register val, Register tmp1, Register tmp2, Register tmp3);

3
src/hotspot/cpu/s390/interp_masm_s390.cpp
View File

@ -391,8 +391,7 @@ void InterpreterMacroAssembler::load_resolved_reference_at_index(Register result
bind(index_ok);
#endif
z_agr(result, index); // Address of indexed array element.
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->load_at(this, IN_HEAP, T_OBJECT, Address(result, arrayOopDesc::base_offset_in_bytes(T_OBJECT)), result, tmp, noreg);
load_heap_oop(result, Address(result, arrayOopDesc::base_offset_in_bytes(T_OBJECT)), tmp, noreg);
}
// load cpool->resolved_klass_at(index)

88
src/hotspot/cpu/s390/macroAssembler_s390.cpp
View File

@ -4047,68 +4047,52 @@ void MacroAssembler::compare_heap_oop(Register Rop1, Address mem, bool maybeNULL
BLOCK_COMMENT("} compare heap oop");
}
// Load heap oop and decompress, if necessary.
void MacroAssembler::load_heap_oop(Register dest, const Address &a) {
if (UseCompressedOops) {
z_llgf(dest, a.disp(), a.indexOrR0(), a.baseOrR0());
oop_decoder(dest, dest, true);
void MacroAssembler::access_store_at(BasicType type, DecoratorSet decorators,
const Address& addr, Register val,
Register tmp1, Register tmp2, Register tmp3) {
assert((decorators & ~(AS_RAW | IN_HEAP | IN_HEAP_ARRAY | IN_ROOT | OOP_NOT_NULL |
ON_UNKNOWN_OOP_REF)) == 0, "unsupported decorator");
BarrierSetAssembler* bs = BarrierSet::barrier_set()->barrier_set_assembler();
bool as_raw = (decorators & AS_RAW) != 0;
if (as_raw) {
bs->BarrierSetAssembler::store_at(this, decorators, type,
addr, val,
tmp1, tmp2, tmp3);
} else {
z_lg(dest, a.disp(), a.indexOrR0(), a.baseOrR0());
bs->store_at(this, decorators, type,
addr, val,
tmp1, tmp2, tmp3);
}
}
// Load heap oop and decompress, if necessary.
void MacroAssembler::load_heap_oop(Register dest, int64_t disp, Register base) {
if (UseCompressedOops) {
z_llgf(dest, disp, base);
oop_decoder(dest, dest, true);
void MacroAssembler::access_load_at(BasicType type, DecoratorSet decorators,
const Address& addr, Register dst,
Register tmp1, Register tmp2, Label *is_null) {
assert((decorators & ~(AS_RAW | IN_HEAP | IN_HEAP_ARRAY | IN_ROOT | OOP_NOT_NULL |
ON_PHANTOM_OOP_REF | ON_WEAK_OOP_REF)) == 0, "unsupported decorator");
BarrierSetAssembler* bs = BarrierSet::barrier_set()->barrier_set_assembler();
bool as_raw = (decorators & AS_RAW) != 0;
if (as_raw) {
bs->BarrierSetAssembler::load_at(this, decorators, type,
addr, dst,
tmp1, tmp2, is_null);
} else {
z_lg(dest, disp, base);
bs->load_at(this, decorators, type,
addr, dst,
tmp1, tmp2, is_null);
}
}
// Load heap oop and decompress, if necessary.
void MacroAssembler::load_heap_oop_not_null(Register dest, int64_t disp, Register base) {
if (UseCompressedOops) {
z_llgf(dest, disp, base);
oop_decoder(dest, dest, false);
} else {
z_lg(dest, disp, base);
}
void MacroAssembler::load_heap_oop(Register dest, const Address &a,
Register tmp1, Register tmp2,
DecoratorSet decorators, Label *is_null) {
access_load_at(T_OBJECT, IN_HEAP | decorators, a, dest, tmp1, tmp2, is_null);
}
// Compress, if necessary, and store oop to heap.
void MacroAssembler::store_heap_oop(Register Roop, RegisterOrConstant offset, Register base) {
Register Ridx = offset.is_register() ? offset.register_or_noreg() : Z_R0;
if (UseCompressedOops) {
assert_different_registers(Roop, offset.register_or_noreg(), base);
encode_heap_oop(Roop);
z_st(Roop, offset.constant_or_zero(), Ridx, base);
} else {
z_stg(Roop, offset.constant_or_zero(), Ridx, base);
}
}
// Compress, if necessary, and store oop to heap. Oop is guaranteed to be not NULL.
void MacroAssembler::store_heap_oop_not_null(Register Roop, RegisterOrConstant offset, Register base) {
Register Ridx = offset.is_register() ? offset.register_or_noreg() : Z_R0;
if (UseCompressedOops) {
assert_different_registers(Roop, offset.register_or_noreg(), base);
encode_heap_oop_not_null(Roop);
z_st(Roop, offset.constant_or_zero(), Ridx, base);
} else {
z_stg(Roop, offset.constant_or_zero(), Ridx, base);
}
}
// Store NULL oop to heap.
void MacroAssembler::store_heap_oop_null(Register zero, RegisterOrConstant offset, Register base) {
Register Ridx = offset.is_register() ? offset.register_or_noreg() : Z_R0;
if (UseCompressedOops) {
z_st(zero, offset.constant_or_zero(), Ridx, base);
} else {
z_stg(zero, offset.constant_or_zero(), Ridx, base);
}
void MacroAssembler::store_heap_oop(Register Roop, const Address &a,
Register tmp1, Register tmp2, Register tmp3,
DecoratorSet decorators) {
access_store_at(T_OBJECT, IN_HEAP | decorators, a, Roop, tmp1, tmp2, tmp3);
}
//-------------------------------------------------

26
src/hotspot/cpu/s390/macroAssembler_s390.hpp
View File

@ -27,6 +27,7 @@
#define CPU_S390_VM_MACROASSEMBLER_S390_HPP
#include "asm/assembler.hpp"
#include "oops/accessDecorators.hpp"
#define MODERN_IFUN(name) ((void (MacroAssembler::*)(Register, int64_t, Register, Register))&MacroAssembler::name)
#define CLASSIC_IFUN(name) ((void (MacroAssembler::*)(Register, int64_t, Register, Register))&MacroAssembler::name)
@ -804,12 +805,25 @@ class MacroAssembler: public Assembler {
int get_oop_base_complement(Register Rbase, uint64_t oop_base);
void compare_heap_oop(Register Rop1, Address mem, bool maybeNULL);
void compare_klass_ptr(Register Rop1, int64_t disp, Register Rbase, bool maybeNULL);
void load_heap_oop(Register dest, const Address &a);
void load_heap_oop(Register d, int64_t si16, Register s1);
void load_heap_oop_not_null(Register d, int64_t si16, Register s1);
void store_heap_oop(Register Roop, RegisterOrConstant offset, Register base);
void store_heap_oop_not_null(Register Roop, RegisterOrConstant offset, Register base);
void store_heap_oop_null(Register zero, RegisterOrConstant offset, Register base);
// Access heap oop, handle encoding and GC barriers.
private:
void access_store_at(BasicType type, DecoratorSet decorators,
const Address& addr, Register val,
Register tmp1, Register tmp2, Register tmp3);
void access_load_at(BasicType type, DecoratorSet decorators,
const Address& addr, Register dst,
Register tmp1, Register tmp2, Label *is_null = NULL);
public:
// tmp1 and tmp2 are used with decorators ON_PHANTOM_OOP_REF or ON_WEAK_OOP_REF.
void load_heap_oop(Register dest, const Address &a,
Register tmp1, Register tmp2,
DecoratorSet decorators = 0, Label *is_null = NULL);
void store_heap_oop(Register Roop, const Address &a,
Register tmp1, Register tmp2, Register tmp3,
DecoratorSet decorators = 0);
void oop_encoder(Register Rdst, Register Rsrc, bool maybeNULL,
Register Rbase = Z_R1, int pow2_offset = -1, bool only32bitValid = false);
void oop_decoder(Register Rdst, Register Rsrc, bool maybeNULL,

27
src/hotspot/cpu/s390/methodHandles_s390.cpp
View File

@ -196,16 +196,19 @@ void MethodHandles::jump_to_lambda_form(MacroAssembler* _masm,
// Load the invoker, as MH -> MH.form -> LF.vmentry
__ verify_oop(recv);
__ load_heap_oop(method_temp,
Address(recv,
NONZERO(java_lang_invoke_MethodHandle::form_offset_in_bytes())));
Address(recv,
NONZERO(java_lang_invoke_MethodHandle::form_offset_in_bytes())),
noreg, noreg, OOP_NOT_NULL);
__ verify_oop(method_temp);
__ load_heap_oop(method_temp,
Address(method_temp,
NONZERO(java_lang_invoke_LambdaForm::vmentry_offset_in_bytes())));
Address(method_temp,
NONZERO(java_lang_invoke_LambdaForm::vmentry_offset_in_bytes())),
noreg, noreg, OOP_NOT_NULL);
__ verify_oop(method_temp);
__ load_heap_oop(method_temp,
Address(method_temp,
NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes())));
Address(method_temp,
NONZERO(java_lang_invoke_MemberName::method_offset_in_bytes())),
noreg, noreg, OOP_NOT_NULL);
__ verify_oop(method_temp);
__ z_lg(method_temp,
Address(method_temp,
@ -405,7 +408,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
NearLabel L_ok;
Register temp2_defc = temp2;
__ load_heap_oop(temp2_defc, member_clazz);
__ load_heap_oop(temp2_defc, member_clazz,
noreg, noreg, OOP_NOT_NULL);
load_klass_from_Class(_masm, temp2_defc, temp3, temp4);
__ verify_klass_ptr(temp2_defc);
__ check_klass_subtype(temp1_recv_klass, temp2_defc, temp3, temp4, L_ok);
@ -431,7 +435,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
if (VerifyMethodHandles) {
verify_ref_kind(_masm, JVM_REF_invokeSpecial, member_reg, temp3);
}
__ load_heap_oop(Z_method, member_vmtarget);
__ load_heap_oop(Z_method, member_vmtarget,
noreg, noreg, OOP_NOT_NULL);
__ z_lg(Z_method, vmtarget_method);
method_is_live = true;
break;
@ -440,7 +445,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
if (VerifyMethodHandles) {
verify_ref_kind(_masm, JVM_REF_invokeStatic, member_reg, temp3);
}
__ load_heap_oop(Z_method, member_vmtarget);
__ load_heap_oop(Z_method, member_vmtarget,
noreg, noreg, OOP_NOT_NULL);
__ z_lg(Z_method, vmtarget_method);
method_is_live = true;
break;
@ -481,7 +487,8 @@ void MethodHandles::generate_method_handle_dispatch(MacroAssembler* _masm,
Register temp3_intf = temp3;
__ load_heap_oop(temp3_intf, member_clazz);
__ load_heap_oop(temp3_intf, member_clazz,
noreg, noreg, OOP_NOT_NULL);
load_klass_from_Class(_masm, temp3_intf, temp2, temp4);
Register Z_index = Z_method;

4
src/hotspot/cpu/s390/s390.ad
View File

@ -2239,8 +2239,8 @@ encode %{
// Go through the vtable. Get receiver klass. Receiver already
// checked for non-null. If we'll go thru a C2I adapter, the
// interpreter expects method in Z_method.
// Use Z_method to temporarily hold the klass oop. Z_R1_scratch is destroyed
// by load_heap_oop_not_null.
// Use Z_method to temporarily hold the klass oop.
// Z_R1_scratch is destroyed.
__ load_klass(Z_method, Z_R2);
int entry_offset = in_bytes(Klass::vtable_start_offset()) + vtable_index * vtableEntry::size_in_bytes();

4
src/hotspot/cpu/s390/stubGenerator_s390.cpp
View File

@ -1300,7 +1300,7 @@ class StubGenerator: public StubCodeGenerator {
unsigned int start_off = __ offset(); // Remember stub start address (is rtn value).
unsigned int size = UseCompressedOops ? 4 : 8;
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -1392,7 +1392,7 @@ class StubGenerator: public StubCodeGenerator {
// Branch to disjoint_copy (if applicable) before pre_barrier to avoid double pre_barrier.
array_overlap_test(nooverlap_target, shift); // Branch away to nooverlap_target if disjoint.
DecoratorSet decorators = 0;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}

4
src/hotspot/cpu/s390/templateInterpreterGenerator_s390.cpp
View File

@ -512,9 +512,7 @@ address TemplateInterpreterGenerator::generate_Reference_get_entry(void) {
__ z_bre(slow_path);
// Load the value of the referent field.
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->load_at(_masm, IN_HEAP | ON_WEAK_OOP_REF, T_OBJECT,
Address(pre_val, referent_offset), pre_val, scratch1, scratch2);
__ load_heap_oop(pre_val, Address(pre_val, referent_offset), scratch1, scratch2, ON_WEAK_OOP_REF);
// Restore caller sp for c2i case.
__ resize_frame_absolute(Z_R10, Z_R0, true); // Cut the stack back to where the caller started.

6
src/hotspot/cpu/s390/templateTable_s390.cpp
View File

@ -200,8 +200,7 @@ static void do_oop_store(InterpreterMacroAssembler* _masm,
Register tmp3,
DecoratorSet decorators) {
assert_different_registers(tmp1, tmp2, tmp3, val, addr.base());
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->store_at(_masm, decorators, T_OBJECT, addr, val, tmp1, tmp2, tmp3);
__ store_heap_oop(val, addr, tmp1, tmp2, tmp3, decorators);
}
static void do_oop_load(InterpreterMacroAssembler* _masm,
@ -212,8 +211,7 @@ static void do_oop_load(InterpreterMacroAssembler* _masm,
DecoratorSet decorators) {
assert_different_registers(addr.base(), tmp1, tmp2);
assert_different_registers(dst, tmp1, tmp2);
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->load_at(_masm, decorators, T_OBJECT, addr, dst, tmp1, tmp2);
__ load_heap_oop(dst, addr, tmp1, tmp2, decorators);
}
Address TemplateTable::at_bcp(int offset) {

80
src/hotspot/cpu/s390/vm_version_ext_s390.cpp Normal file
View File

@ -0,0 +1,80 @@
/*
* Copyright (c) 2013, 2018, 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.
*
*/
#include "jvm.h"
#include "memory/allocation.hpp"
#include "memory/allocation.inline.hpp"
#include "vm_version_ext_s390.hpp"
// VM_Version_Ext statics
int VM_Version_Ext::_no_of_threads = 0;
int VM_Version_Ext::_no_of_cores = 0;
int VM_Version_Ext::_no_of_sockets = 0;
char VM_Version_Ext::_cpu_name[CPU_TYPE_DESC_BUF_SIZE] = {0};
char VM_Version_Ext::_cpu_desc[CPU_DETAILED_DESC_BUF_SIZE] = {0};
// get cpu information.
bool VM_Version_Ext::initialize_cpu_information(void) {
// Not yet implemented.
return false;
}
int VM_Version_Ext::number_of_threads(void) {
initialize_cpu_information();
return _no_of_threads;
}
int VM_Version_Ext::number_of_cores(void) {
initialize_cpu_information();
return _no_of_cores;
}
int VM_Version_Ext::number_of_sockets(void) {
initialize_cpu_information();
return _no_of_sockets;
}
const char* VM_Version_Ext::cpu_name(void) {
if (!initialize_cpu_information()) {
return NULL;
}
char* tmp = NEW_C_HEAP_ARRAY_RETURN_NULL(char, CPU_TYPE_DESC_BUF_SIZE, mtTracing);
if (NULL == tmp) {
return NULL;
}
strncpy(tmp, _cpu_name, CPU_TYPE_DESC_BUF_SIZE);
return tmp;
}
const char* VM_Version_Ext::cpu_description(void) {
if (!initialize_cpu_information()) {
return NULL;
}
char* tmp = NEW_C_HEAP_ARRAY_RETURN_NULL(char, CPU_DETAILED_DESC_BUF_SIZE, mtTracing);
if (NULL == tmp) {
return NULL;
}
strncpy(tmp, _cpu_desc, CPU_DETAILED_DESC_BUF_SIZE);
return tmp;
}

61
src/hotspot/cpu/s390/vm_version_ext_s390.hpp Normal file
View File

@ -0,0 +1,61 @@
/*
* Copyright (c) 2013, 2018, 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.
*
*/
#ifndef CPU_S390_VM_VM_VERSION_EXT_S390_HPP
#define CPU_S390_VM_VM_VERSION_EXT_S390_HPP
#include "utilities/macros.hpp"
#include "vm_version_s390.hpp"
#define CPU_INFO "cpu_info"
#define CPU_TYPE "fpu_type"
#define CPU_DESCRIPTION "implementation"
#define CHIP_ID "chip_id"
#define CORE_ID "core_id"
class VM_Version_Ext : public VM_Version {
private:
static const size_t CPU_TYPE_DESC_BUF_SIZE = 256;
static const size_t CPU_DETAILED_DESC_BUF_SIZE = 4096;
static int _no_of_threads;
static int _no_of_cores;
static int _no_of_sockets;
static char _cpu_name[CPU_TYPE_DESC_BUF_SIZE];
static char _cpu_desc[CPU_DETAILED_DESC_BUF_SIZE];
static bool initialize_cpu_information(void);
public:
static int number_of_threads(void);
static int number_of_cores(void);
static int number_of_sockets(void);
static const char* cpu_name(void);
static const char* cpu_description(void);
};
#endif // CPU_S390_VM_VM_VERSION_EXT_S390_HPP

3
src/hotspot/cpu/sparc/c1_LIRGenerator_sparc.cpp
View File

@ -1219,7 +1219,8 @@ void LIRGenerator::do_If(If* x) {
// add safepoint before generating condition code so it can be recomputed
if (x->is_safepoint()) {
// increment backedge counter if needed
increment_backedge_counter(state_for(x, x->state_before()), x->profiled_bci());
increment_backedge_counter_conditionally(lir_cond(cond), left, right, state_for(x, x->state_before()),
x->tsux()->bci(), x->fsux()->bci(), x->profiled_bci());
__ safepoint(safepoint_poll_register(), state_for(x, x->state_before()));
}

17
src/hotspot/cpu/sparc/jniFastGetField_sparc.cpp
View File

@ -153,8 +153,13 @@ address JNI_FastGetField::generate_fast_get_long_field() {
__ andcc (G4, 1, G0);
__ br (Assembler::notZero, false, Assembler::pn, label1);
__ delayed()->srl (O2, 2, O4);
__ andn (O1, JNIHandles::weak_tag_mask, O1);
__ ld_ptr (O1, 0, O5);
__ mov(O1, O5);
// Both O5 and G1 are clobbered by try_resolve_jobject_in_native.
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->try_resolve_jobject_in_native(masm, /* jni_env */ O0, /* obj */ O5, /* tmp */ G1, label1);
DEBUG_ONLY(__ set(0xDEADC0DE, G1);)
__ add (O5, O4, O5);
assert(count < LIST_CAPACITY, "LIST_CAPACITY too small");
@ -206,8 +211,12 @@ address JNI_FastGetField::generate_fast_get_float_field0(BasicType type) {
__ andcc (G4, 1, G0);
__ br (Assembler::notZero, false, Assembler::pn, label1);
__ delayed()->srl (O2, 2, O4);
__ andn (O1, JNIHandles::weak_tag_mask, O1);
__ ld_ptr (O1, 0, O5);
__ mov(O1, O5);
// Both O5 and G3 are clobbered by try_resolve_jobject_in_native.
BarrierSetAssembler *bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->try_resolve_jobject_in_native(masm, /* jni_env */ O0, /* obj */ O5, /* tmp */ G3, label1);
DEBUG_ONLY(__ set(0xDEADC0DE, G3);)
assert(count < LIST_CAPACITY, "LIST_CAPACITY too small");
speculative_load_pclist[count] = __ pc();

6
src/hotspot/cpu/sparc/stubGenerator_sparc.cpp
View File

@ -2269,7 +2269,7 @@ class StubGenerator: public StubCodeGenerator {
BLOCK_COMMENT("Entry:");
}
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2326,7 +2326,7 @@ class StubGenerator: public StubCodeGenerator {
array_overlap_test(nooverlap_target, LogBytesPerHeapOop);
DecoratorSet decorators = 0;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2446,7 +2446,7 @@ class StubGenerator: public StubCodeGenerator {
BLOCK_COMMENT("Entry:");
}
DecoratorSet decorators = ARRAYCOPY_CHECKCAST;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_CHECKCAST;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}

11
src/hotspot/cpu/x86/c1_LIRGenerator_x86.cpp
View File

@ -1402,16 +1402,19 @@ void LIRGenerator::do_If(If* x) {
yin->dont_load_item();
}
LIR_Opr left = xin->result();
LIR_Opr right = yin->result();
set_no_result(x);
// add safepoint before generating condition code so it can be recomputed
if (x->is_safepoint()) {
// increment backedge counter if needed
increment_backedge_counter(state_for(x, x->state_before()), x->profiled_bci());
increment_backedge_counter_conditionally(lir_cond(cond), left, right, state_for(x, x->state_before()),
x->tsux()->bci(), x->fsux()->bci(), x->profiled_bci());
__ safepoint(safepoint_poll_register(), state_for(x, x->state_before()));
}
set_no_result(x);
LIR_Opr left = xin->result();
LIR_Opr right = yin->result();
__ cmp(lir_cond(cond), left, right);
// Generate branch profiling. Profiling code doesn't kill flags.
profile_branch(x, cond);

6
src/hotspot/cpu/x86/jniFastGetField_x86_64.cpp
View File

@ -188,9 +188,11 @@ address JNI_FastGetField::generate_fast_get_float_field0(BasicType type) {
// robj is data dependent on rcounter.
}
__ clear_jweak_tag(robj);
// Both robj and rtmp are clobbered by try_resolve_jobject_in_native.
BarrierSetAssembler* bs = BarrierSet::barrier_set()->barrier_set_assembler();
bs->try_resolve_jobject_in_native(masm, /* jni_env */ c_rarg0, robj, rtmp, slow);
DEBUG_ONLY(__ movl(rtmp, 0xDEADC0DE);)
__ movptr(robj, Address(robj, 0)); // *obj
__ mov (roffset, c_rarg2);
__ shrptr(roffset, 2); // offset

2
src/hotspot/cpu/x86/macroAssembler_x86.cpp
View File

@ -6276,6 +6276,8 @@ void MacroAssembler::restore_cpu_control_state_after_jni() {
// ((OopHandle)result).resolve();
void MacroAssembler::resolve_oop_handle(Register result, Register tmp) {
assert_different_registers(result, tmp);
// Only 64 bit platforms support GCs that require a tmp register
// Only IN_HEAP loads require a thread_tmp register
// OopHandle::resolve is an indirection like jobject.

6
src/hotspot/cpu/x86/stubGenerator_x86_32.cpp
View File

@ -837,7 +837,7 @@ class StubGenerator: public StubCodeGenerator {
__ jcc(Assembler::zero, L_0_count);
}
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -1026,7 +1026,7 @@ class StubGenerator: public StubCodeGenerator {
__ jcc(Assembler::zero, L_0_count);
}
DecoratorSet decorators = 0;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -1383,7 +1383,7 @@ class StubGenerator: public StubCodeGenerator {
Address to_element_addr(end_to, count, Address::times_ptr, 0);
Address elem_klass_addr(elem, oopDesc::klass_offset_in_bytes());
DecoratorSet decorators = ARRAYCOPY_CHECKCAST;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_CHECKCAST;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}

10
src/hotspot/cpu/x86/stubGenerator_x86_64.cpp
View File

@ -1820,7 +1820,7 @@ class StubGenerator: public StubCodeGenerator {
setup_arg_regs(); // from => rdi, to => rsi, count => rdx
// r9 and r10 may be used to save non-volatile registers
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -1914,7 +1914,7 @@ class StubGenerator: public StubCodeGenerator {
setup_arg_regs(); // from => rdi, to => rsi, count => rdx
// r9 and r10 may be used to save non-volatile registers
DecoratorSet decorators = 0;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2018,7 +2018,7 @@ class StubGenerator: public StubCodeGenerator {
// r9 and r10 may be used to save non-volatile registers
// 'from', 'to' and 'qword_count' are now valid
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2111,7 +2111,7 @@ class StubGenerator: public StubCodeGenerator {
// r9 and r10 may be used to save non-volatile registers
// 'from', 'to' and 'qword_count' are now valid
DecoratorSet decorators = ARRAYCOPY_DISJOINT;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_DISJOINT;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}
@ -2294,7 +2294,7 @@ class StubGenerator: public StubCodeGenerator {
Address from_element_addr(end_from, count, TIMES_OOP, 0);
Address to_element_addr(end_to, count, TIMES_OOP, 0);
DecoratorSet decorators = ARRAYCOPY_CHECKCAST;
DecoratorSet decorators = IN_HEAP | IN_HEAP_ARRAY | ARRAYCOPY_CHECKCAST;
if (dest_uninitialized) {
decorators |= AS_DEST_NOT_INITIALIZED;
}

2
src/hotspot/cpu/x86/templateInterpreterGenerator_x86.cpp
View File

@ -973,7 +973,7 @@ address TemplateInterpreterGenerator::generate_native_entry(bool synchronized) {
__ testl(t, JVM_ACC_STATIC);
__ jcc(Assembler::zero, L);
// get mirror
__ load_mirror(t, method);
__ load_mirror(t, method, rax);
// copy mirror into activation frame
__ movptr(Address(rbp, frame::interpreter_frame_oop_temp_offset * wordSize),
t);

2
src/hotspot/cpu/x86/vm_version_x86.cpp
View File

@ -923,7 +923,7 @@ void VM_Version::get_processor_features() {
// Only C2 does RTM locking optimization.
// Can't continue because UseRTMLocking affects UseBiasedLocking flag
// setting during arguments processing. See use_biased_locking().
vm_exit_during_initialization("RTM locking optimization is not supported in emulated client VM");
vm_exit_during_initialization("RTM locking optimization is not supported in this VM");
}
if (is_intel_family_core()) {
if ((_model == CPU_MODEL_HASWELL_E3) ||

90
src/hotspot/cpu/zero/vm_version_ext_zero.cpp Normal file
View File

@ -0,0 +1,90 @@
/*
* Copyright (c) 2013, 2018, 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.
*
*/
#include "memory/allocation.hpp"
#include "memory/allocation.inline.hpp"
#include "runtime/os.inline.hpp"
#include "vm_version_ext_zero.hpp"
// VM_Version_Ext statics
int VM_Version_Ext::_no_of_threads = 0;
int VM_Version_Ext::_no_of_cores = 0;
int VM_Version_Ext::_no_of_sockets = 0;
bool VM_Version_Ext::_initialized = false;
char VM_Version_Ext::_cpu_name[CPU_TYPE_DESC_BUF_SIZE] = {0};
char VM_Version_Ext::_cpu_desc[CPU_DETAILED_DESC_BUF_SIZE] = {0};
void VM_Version_Ext::initialize_cpu_information(void) {
// do nothing if cpu info has been initialized
if (_initialized) {
return;
}
int core_id = -1;
int chip_id = -1;
int len = 0;
char* src_string = NULL;
_no_of_cores = os::processor_count();
_no_of_threads = _no_of_cores;
_no_of_sockets = _no_of_cores;
snprintf(_cpu_name, CPU_TYPE_DESC_BUF_SIZE - 1, "Zero VM");
snprintf(_cpu_desc, CPU_DETAILED_DESC_BUF_SIZE, "%s", _features_string);
_initialized = true;
}
int VM_Version_Ext::number_of_threads(void) {
initialize_cpu_information();
return _no_of_threads;
}
int VM_Version_Ext::number_of_cores(void) {
initialize_cpu_information();
return _no_of_cores;
}
int VM_Version_Ext::number_of_sockets(void) {
initialize_cpu_information();
return _no_of_sockets;
}
const char* VM_Version_Ext::cpu_name(void) {
initialize_cpu_information();
char* tmp = NEW_C_HEAP_ARRAY_RETURN_NULL(char, CPU_TYPE_DESC_BUF_SIZE, mtTracing);
if (NULL == tmp) {
return NULL;
}
strncpy(tmp, _cpu_name, CPU_TYPE_DESC_BUF_SIZE);
return tmp;
}
const char* VM_Version_Ext::cpu_description(void) {
initialize_cpu_information();
char* tmp = NEW_C_HEAP_ARRAY_RETURN_NULL(char, CPU_DETAILED_DESC_BUF_SIZE, mtTracing);
if (NULL == tmp) {
return NULL;
}
strncpy(tmp, _cpu_desc, CPU_DETAILED_DESC_BUF_SIZE);
return tmp;
}

54
src/hotspot/cpu/zero/vm_version_ext_zero.hpp Normal file
View File

@ -0,0 +1,54 @@
/*
* Copyright (c) 2013, 2018, 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.
*
*/
#ifndef CPU_ZERO_VM_VM_VERSION_EXT_ZERO_HPP
#define CPU_ZERO_VM_VM_VERSION_EXT_ZERO_HPP
#include "utilities/macros.hpp"
#include "vm_version_zero.hpp"
class VM_Version_Ext : public VM_Version {
private:
static const size_t CPU_TYPE_DESC_BUF_SIZE = 256;
static const size_t CPU_DETAILED_DESC_BUF_SIZE = 4096;
static int _no_of_threads;
static int _no_of_cores;
static int _no_of_sockets;
static bool _initialized;
static char _cpu_name[CPU_TYPE_DESC_BUF_SIZE];
static char _cpu_desc[CPU_DETAILED_DESC_BUF_SIZE];
public:
static int number_of_threads(void);
static int number_of_cores(void);
static int number_of_sockets(void);
static const char* cpu_name(void);
static const char* cpu_description(void);
static void initialize_cpu_information(void);
};
#endif // CPU_ZERO_VM_VM_VERSION_EXT_ZERO_HPP

111
src/hotspot/os/aix/jvm_aix.cpp
View File

@ -1,111 +0,0 @@
/*
* Copyright (c) 1999, 2015, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2013 SAP SE. 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.
*
*/
#include "precompiled.hpp"
#include "jvm.h"
#include "runtime/interfaceSupport.inline.hpp"
#include "runtime/osThread.hpp"
#include <signal.h>
// sun.misc.Signal ///////////////////////////////////////////////////////////
// Signal code is mostly copied from classic vm, signals_md.c 1.4 98/08/23
/*
* This function is included primarily as a debugging aid. If Java is
* running in a console window, then pressing <CTRL-\\> will cause
* the current state of all active threads and monitors to be written
* to the console window.
*/
JVM_ENTRY_NO_ENV(void*, JVM_RegisterSignal(jint sig, void* handler))
// Copied from classic vm
// signals_md.c 1.4 98/08/23
void* newHandler = handler == (void *)2
? os::user_handler()
: handler;
switch (sig) {
/* The following are already used by the VM. */
case SIGFPE:
case SIGILL:
case SIGSEGV:
/* The following signal is used by the VM to dump thread stacks unless
ReduceSignalUsage is set, in which case the user is allowed to set
his own _native_ handler for this signal; thus, in either case,
we do not allow JVM_RegisterSignal to change the handler. */
case BREAK_SIGNAL:
return (void *)-1;
/* The following signals are used for Shutdown Hooks support. However, if
ReduceSignalUsage (-Xrs) is set, Shutdown Hooks must be invoked via
System.exit(), Java is not allowed to use these signals, and the the
user is allowed to set his own _native_ handler for these signals and
invoke System.exit() as needed. Terminator.setup() is avoiding
registration of these signals when -Xrs is present.
- If the HUP signal is ignored (from the nohup) command, then Java
is not allowed to use this signal.
*/
case SHUTDOWN1_SIGNAL:
case SHUTDOWN2_SIGNAL:
case SHUTDOWN3_SIGNAL:
if (ReduceSignalUsage) return (void*)-1;
if (os::Aix::is_sig_ignored(sig)) return (void*)1;
}
void* oldHandler = os::signal(sig, newHandler);
if (oldHandler == os::user_handler()) {
return (void *)2;
} else {
return oldHandler;
}
JVM_END
JVM_ENTRY_NO_ENV(jboolean, JVM_RaiseSignal(jint sig))
if (ReduceSignalUsage) {
// do not allow SHUTDOWN1_SIGNAL,SHUTDOWN2_SIGNAL,SHUTDOWN3_SIGNAL,
// BREAK_SIGNAL to be raised when ReduceSignalUsage is set, since
// no handler for them is actually registered in JVM or via
// JVM_RegisterSignal.
if (sig == SHUTDOWN1_SIGNAL || sig == SHUTDOWN2_SIGNAL ||
sig == SHUTDOWN3_SIGNAL || sig == BREAK_SIGNAL) {
return JNI_FALSE;
}
}
else if ((sig == SHUTDOWN1_SIGNAL || sig == SHUTDOWN2_SIGNAL ||
sig == SHUTDOWN3_SIGNAL) && os::Aix::is_sig_ignored(sig)) {
// do not allow SHUTDOWN1_SIGNAL to be raised when SHUTDOWN1_SIGNAL
// is ignored, since no handler for them is actually registered in JVM
// or via JVM_RegisterSignal.
// This also applies for SHUTDOWN2_SIGNAL and SHUTDOWN3_SIGNAL
return JNI_FALSE;
}
os::signal_raise(sig);
return JNI_TRUE;
JVM_END

41
src/hotspot/os/aix/os_aix.cpp
View File

@ -622,18 +622,6 @@ extern "C" void breakpoint() {
debug_only(static bool signal_sets_initialized = false);
static sigset_t unblocked_sigs, vm_sigs;
bool os::Aix::is_sig_ignored(int sig) {
struct sigaction oact;
sigaction(sig, (struct sigaction*)NULL, &oact);
void* ohlr = oact.sa_sigaction ? CAST_FROM_FN_PTR(void*, oact.sa_sigaction)
: CAST_FROM_FN_PTR(void*, oact.sa_handler);
if (ohlr == CAST_FROM_FN_PTR(void*, SIG_IGN)) {
return true;
} else {
return false;
}
}
void os::Aix::signal_sets_init() {
// Should also have an assertion stating we are still single-threaded.
assert(!signal_sets_initialized, "Already initialized");
@ -659,13 +647,13 @@ void os::Aix::signal_sets_init() {
sigaddset(&unblocked_sigs, SR_signum);
if (!ReduceSignalUsage) {
if (!os::Aix::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN1_SIGNAL);
}
if (!os::Aix::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN2_SIGNAL);
}
if (!os::Aix::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN3_SIGNAL);
}
}
@ -1377,6 +1365,21 @@ void os::get_summary_os_info(char* buf, size_t buflen) {
snprintf(buf, buflen, "%s %s", name.release, name.version);
}
int os::get_loaded_modules_info(os::LoadedModulesCallbackFunc callback, void *param) {
// Not yet implemented.
return 0;
}
void os::print_os_info_brief(outputStream* st) {
uint32_t ver = os::Aix::os_version();
st->print_cr("AIX kernel version %u.%u.%u.%u",
(ver >> 24) & 0xFF, (ver >> 16) & 0xFF, (ver >> 8) & 0xFF, ver & 0xFF);
os::Posix::print_uname_info(st);
// Linux uses print_libversion_info(st); here.
}
void os::print_os_info(outputStream* st) {
st->print("OS:");
@ -1795,7 +1798,7 @@ static void local_sem_wait() {
}
}
void os::signal_init_pd() {
static void jdk_misc_signal_init() {
// Initialize signal structures
::memset((void*)pending_signals, 0, sizeof(pending_signals));
@ -3020,7 +3023,7 @@ bool unblock_program_error_signals() {
}
// Renamed from 'signalHandler' to avoid collision with other shared libs.
void javaSignalHandler(int sig, siginfo_t* info, void* uc) {
static void javaSignalHandler(int sig, siginfo_t* info, void* uc) {
assert(info != NULL && uc != NULL, "it must be old kernel");
// Never leave program error signals blocked;
@ -3579,6 +3582,10 @@ jint os::init_2(void) {
Aix::signal_sets_init();
Aix::install_signal_handlers();
// Initialize data for jdk.internal.misc.Signal
if (!ReduceSignalUsage) {
jdk_misc_signal_init();
}
// Check and sets minimum stack sizes against command line options
if (Posix::set_minimum_stack_sizes() == JNI_ERR) {

3
src/hotspot/os/aix/os_aix.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2017, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2013, 2016 SAP SE. All rights reserved.
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
*
@ -121,7 +121,6 @@ class Aix {
static void signal_sets_init();
static void install_signal_handlers();
static void set_signal_handler(int, bool);
static bool is_sig_ignored(int sig);
static sigset_t* unblocked_signals();
static sigset_t* vm_signals();

10
src/hotspot/os/aix/os_aix.inline.hpp
View File

@ -1,6 +1,6 @@
/*
* Copyright (c) 1999, 2016, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2016 SAP SE. All rights reserved.
* Copyright (c) 1999, 2018, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 2012, 2018 SAP SE. 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
@ -36,9 +36,9 @@
#include <sys/ioctl.h>
#include <netdb.h>
// File names are case-sensitive on windows only.
inline int os::file_name_strcmp(const char* s1, const char* s2) {
return strcmp(s1, s2);
// File names are case-insensitive on windows only.
inline int os::file_name_strncmp(const char* s1, const char* s2, size_t num) {
return strncmp(s1, s2, num);
}
inline bool os::obsolete_option(const JavaVMOption *option) {

1043
src/hotspot/os/aix/os_perf_aix.cpp Normal file
View File

File diff suppressed because it is too large Load Diff

3
src/hotspot/os/aix/os_share_aix.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2013, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, 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
@ -26,7 +26,6 @@
#define OS_AIX_VM_OS_SHARE_AIX_HPP
// misc
void signalHandler(int, siginfo_t*, ucontext_t*);
void handle_unexpected_exception(Thread* thread, int sig, siginfo_t* info, address pc, address adjusted_pc);
#ifndef PRODUCT
void continue_with_dump(void);

28
src/hotspot/os/bsd/os_bsd.cpp
View File

@ -533,18 +533,6 @@ extern "C" void breakpoint() {
debug_only(static bool signal_sets_initialized = false);
static sigset_t unblocked_sigs, vm_sigs;
bool os::Bsd::is_sig_ignored(int sig) {
struct sigaction oact;
sigaction(sig, (struct sigaction*)NULL, &oact);
void* ohlr = oact.sa_sigaction ? CAST_FROM_FN_PTR(void*, oact.sa_sigaction)
: CAST_FROM_FN_PTR(void*, oact.sa_handler);
if (ohlr == CAST_FROM_FN_PTR(void*, SIG_IGN)) {
return true;
} else {
return false;
}
}
void os::Bsd::signal_sets_init() {
// Should also have an assertion stating we are still single-threaded.
assert(!signal_sets_initialized, "Already initialized");
@ -569,14 +557,14 @@ void os::Bsd::signal_sets_init() {
sigaddset(&unblocked_sigs, SR_signum);
if (!ReduceSignalUsage) {
if (!os::Bsd::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN1_SIGNAL);
}
if (!os::Bsd::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN2_SIGNAL);
}
if (!os::Bsd::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN3_SIGNAL);
}
}
@ -1830,7 +1818,7 @@ int os::sigexitnum_pd() {
static volatile jint pending_signals[NSIG+1] = { 0 };
static Semaphore* sig_sem = NULL;
void os::signal_init_pd() {
static void jdk_misc_signal_init() {
// Initialize signal structures
::memset((void*)pending_signals, 0, sizeof(pending_signals));
@ -1843,7 +1831,7 @@ void os::signal_notify(int sig) {
Atomic::inc(&pending_signals[sig]);
sig_sem->signal();
} else {
// Signal thread is not created with ReduceSignalUsage and signal_init_pd
// Signal thread is not created with ReduceSignalUsage and jdk_misc_signal_init
// initialization isn't called.
assert(ReduceSignalUsage, "signal semaphore should be created");
}
@ -2773,7 +2761,7 @@ extern "C" JNIEXPORT int JVM_handle_bsd_signal(int signo, siginfo_t* siginfo,
void* ucontext,
int abort_if_unrecognized);
void signalHandler(int sig, siginfo_t* info, void* uc) {
static void signalHandler(int sig, siginfo_t* info, void* uc) {
assert(info != NULL && uc != NULL, "it must be old kernel");
int orig_errno = errno; // Preserve errno value over signal handler.
JVM_handle_bsd_signal(sig, info, uc, true);
@ -3301,6 +3289,10 @@ jint os::init_2(void) {
Bsd::signal_sets_init();
Bsd::install_signal_handlers();
// Initialize data for jdk.internal.misc.Signal
if (!ReduceSignalUsage) {
jdk_misc_signal_init();
}
// Check and sets minimum stack sizes against command line options
if (Posix::set_minimum_stack_sizes() == JNI_ERR) {

3
src/hotspot/os/bsd/os_bsd.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2017, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, 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
@ -103,7 +103,6 @@ class Bsd {
static void signal_sets_init();
static void install_signal_handlers();
static void set_signal_handler(int, bool);
static bool is_sig_ignored(int sig);
static sigset_t* unblocked_signals();
static sigset_t* vm_signals();

8
src/hotspot/os/bsd/os_bsd.inline.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2016, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, 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
@ -34,9 +34,9 @@
#include <poll.h>
#include <netdb.h>
// File names are case-sensitive on windows only
inline int os::file_name_strcmp(const char* s1, const char* s2) {
return strcmp(s1, s2);
// File names are case-insensitive on windows only
inline int os::file_name_strncmp(const char* s1, const char* s2, size_t num) {
return strncmp(s1, s2, num);
}
inline bool os::obsolete_option(const JavaVMOption *option) {

3
src/hotspot/os/bsd/os_share_bsd.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, 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
@ -26,7 +26,6 @@
#define OS_BSD_VM_OS_SHARE_BSD_HPP
// misc
void signalHandler(int, siginfo_t*, ucontext_t*);
void handle_unexpected_exception(Thread* thread, int sig, siginfo_t* info, address pc, address adjusted_pc);
#ifndef PRODUCT
void continue_with_dump(void);

110
src/hotspot/os/linux/jvm_linux.cpp
View File

@ -1,110 +0,0 @@
/*
* Copyright (c) 1999, 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.
*
*/
#include "precompiled.hpp"
#include "jvm.h"
#include "runtime/interfaceSupport.inline.hpp"
#include "runtime/osThread.hpp"
#include <signal.h>
// sun.misc.Signal ///////////////////////////////////////////////////////////
// Signal code is mostly copied from classic vm, signals_md.c 1.4 98/08/23
/*
* This function is included primarily as a debugging aid. If Java is
* running in a console window, then pressing <CTRL-\\> will cause
* the current state of all active threads and monitors to be written
* to the console window.
*/
JVM_ENTRY_NO_ENV(void*, JVM_RegisterSignal(jint sig, void* handler))
// Copied from classic vm
// signals_md.c 1.4 98/08/23
void* newHandler = handler == (void *)2
? os::user_handler()
: handler;
switch (sig) {
/* The following are already used by the VM. */
case SIGFPE:
case SIGILL:
case SIGSEGV:
/* The following signal is used by the VM to dump thread stacks unless
ReduceSignalUsage is set, in which case the user is allowed to set
his own _native_ handler for this signal; thus, in either case,
we do not allow JVM_RegisterSignal to change the handler. */
case BREAK_SIGNAL:
return (void *)-1;
/* The following signals are used for Shutdown Hooks support. However, if
ReduceSignalUsage (-Xrs) is set, Shutdown Hooks must be invoked via
System.exit(), Java is not allowed to use these signals, and the the
user is allowed to set his own _native_ handler for these signals and
invoke System.exit() as needed. Terminator.setup() is avoiding
registration of these signals when -Xrs is present.
- If the HUP signal is ignored (from the nohup) command, then Java
is not allowed to use this signal.
*/
case SHUTDOWN1_SIGNAL:
case SHUTDOWN2_SIGNAL:
case SHUTDOWN3_SIGNAL:
if (ReduceSignalUsage) return (void*)-1;
if (os::Linux::is_sig_ignored(sig)) return (void*)1;
}
void* oldHandler = os::signal(sig, newHandler);
if (oldHandler == os::user_handler()) {
return (void *)2;
} else {
return oldHandler;
}
JVM_END
JVM_ENTRY_NO_ENV(jboolean, JVM_RaiseSignal(jint sig))
if (ReduceSignalUsage) {
// do not allow SHUTDOWN1_SIGNAL,SHUTDOWN2_SIGNAL,SHUTDOWN3_SIGNAL,
// BREAK_SIGNAL to be raised when ReduceSignalUsage is set, since
// no handler for them is actually registered in JVM or via
// JVM_RegisterSignal.
if (sig == SHUTDOWN1_SIGNAL || sig == SHUTDOWN2_SIGNAL ||
sig == SHUTDOWN3_SIGNAL || sig == BREAK_SIGNAL) {
return JNI_FALSE;
}
}
else if ((sig == SHUTDOWN1_SIGNAL || sig == SHUTDOWN2_SIGNAL ||
sig == SHUTDOWN3_SIGNAL) && os::Linux::is_sig_ignored(sig)) {
// do not allow SHUTDOWN1_SIGNAL to be raised when SHUTDOWN1_SIGNAL
// is ignored, since no handler for them is actually registered in JVM
// or via JVM_RegisterSignal.
// This also applies for SHUTDOWN2_SIGNAL and SHUTDOWN3_SIGNAL
return JNI_FALSE;
}
os::signal_raise(sig);
return JNI_TRUE;
JVM_END

38
src/hotspot/os/linux/os_linux.cpp
View File

@ -425,18 +425,6 @@ extern "C" void breakpoint() {
debug_only(static bool signal_sets_initialized = false);
static sigset_t unblocked_sigs, vm_sigs;
bool os::Linux::is_sig_ignored(int sig) {
struct sigaction oact;
sigaction(sig, (struct sigaction*)NULL, &oact);
void* ohlr = oact.sa_sigaction ? CAST_FROM_FN_PTR(void*, oact.sa_sigaction)
: CAST_FROM_FN_PTR(void*, oact.sa_handler);
if (ohlr == CAST_FROM_FN_PTR(void*, SIG_IGN)) {
return true;
} else {
return false;
}
}
void os::Linux::signal_sets_init() {
// Should also have an assertion stating we are still single-threaded.
assert(!signal_sets_initialized, "Already initialized");
@ -464,13 +452,13 @@ void os::Linux::signal_sets_init() {
sigaddset(&unblocked_sigs, SR_signum);
if (!ReduceSignalUsage) {
if (!os::Linux::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN1_SIGNAL);
}
if (!os::Linux::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN2_SIGNAL);
}
if (!os::Linux::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN3_SIGNAL);
}
}
@ -2532,7 +2520,7 @@ static volatile jint pending_signals[NSIG+1] = { 0 };
static Semaphore* sig_sem = NULL;
static PosixSemaphore sr_semaphore;
void os::signal_init_pd() {
static void jdk_misc_signal_init() {
// Initialize signal structures
::memset((void*)pending_signals, 0, sizeof(pending_signals));
@ -2545,7 +2533,7 @@ void os::signal_notify(int sig) {
Atomic::inc(&pending_signals[sig]);
sig_sem->signal();
} else {
// Signal thread is not created with ReduceSignalUsage and signal_init_pd
// Signal thread is not created with ReduceSignalUsage and jdk_misc_signal_init
// initialization isn't called.
assert(ReduceSignalUsage, "signal semaphore should be created");
}
@ -2884,6 +2872,10 @@ void os::Linux::sched_getcpu_init() {
set_sched_getcpu(CAST_TO_FN_PTR(sched_getcpu_func_t,
(void*)&sched_getcpu_syscall));
}
if (sched_getcpu() == -1) {
vm_exit_during_initialization("getcpu(2) system call not supported by kernel");
}
}
// Something to do with the numa-aware allocator needs these symbols
@ -4452,7 +4444,7 @@ extern "C" JNIEXPORT int JVM_handle_linux_signal(int signo,
void* ucontext,
int abort_if_unrecognized);
void signalHandler(int sig, siginfo_t* info, void* uc) {
static void signalHandler(int sig, siginfo_t* info, void* uc) {
assert(info != NULL && uc != NULL, "it must be old kernel");
int orig_errno = errno; // Preserve errno value over signal handler.
JVM_handle_linux_signal(sig, info, uc, true);
@ -5000,6 +4992,10 @@ jint os::init_2(void) {
Linux::signal_sets_init();
Linux::install_signal_handlers();
// Initialize data for jdk.internal.misc.Signal
if (!ReduceSignalUsage) {
jdk_misc_signal_init();
}
// Check and sets minimum stack sizes against command line options
if (Posix::set_minimum_stack_sizes() == JNI_ERR) {
@ -5241,6 +5237,12 @@ int os::active_processor_count() {
return active_cpus;
}
uint os::processor_id() {
const int id = Linux::sched_getcpu();
assert(id >= 0 && id < _processor_count, "Invalid processor id");
return (uint)id;
}
void os::set_native_thread_name(const char *name) {
if (Linux::_pthread_setname_np) {
char buf [16]; // according to glibc manpage, 16 chars incl. '/0'

1
src/hotspot/os/linux/os_linux.hpp
View File

@ -160,7 +160,6 @@ class Linux {
static void signal_sets_init();
static void install_signal_handlers();
static void set_signal_handler(int, bool);
static bool is_sig_ignored(int sig);
static sigset_t* unblocked_signals();
static sigset_t* vm_signals();

8
src/hotspot/os/linux/os_linux.inline.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2016, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, 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
@ -34,9 +34,9 @@
#include <poll.h>
#include <netdb.h>
// File names are case-sensitive on windows only
inline int os::file_name_strcmp(const char* s1, const char* s2) {
return strcmp(s1, s2);
// File names are case-insensitive on windows only
inline int os::file_name_strncmp(const char* s1, const char* s2, size_t num) {
return strncmp(s1, s2, num);
}
inline bool os::obsolete_option(const JavaVMOption *option) {

12
src/hotspot/os/linux/os_perf_linux.cpp
View File

@ -29,17 +29,7 @@
#include "runtime/os.hpp"
#include "runtime/os_perf.hpp"
#ifdef X86
#include "vm_version_ext_x86.hpp"
#endif
#ifdef ARM
#include "vm_version_ext_arm.hpp"
#endif
#ifndef ARM
#ifdef AARCH64
#include "vm_version_ext_aarch64.hpp"
#endif
#endif
#include CPU_HEADER(vm_version_ext)
#include <stdio.h>
#include <stdarg.h>

3
src/hotspot/os/linux/os_share_linux.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2010, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, 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
@ -26,7 +26,6 @@
#define OS_LINUX_VM_OS_SHARE_LINUX_HPP
// misc
void signalHandler(int, siginfo_t*, ucontext_t*);
void handle_unexpected_exception(Thread* thread, int sig, siginfo_t* info, address pc, address adjusted_pc);
#ifndef PRODUCT
void continue_with_dump(void);

6
src/hotspot/os/bsd/jvm_bsd.cpp → src/hotspot/os/posix/jvm_posix.cpp
View File

@ -30,7 +30,7 @@
#include <signal.h>
// sun.misc.Signal ///////////////////////////////////////////////////////////
// jdk.internal.misc.Signal ///////////////////////////////////////////////////////////
// Signal code is mostly copied from classic vm, signals_md.c 1.4 98/08/23
/*
* This function is included primarily as a debugging aid. If Java is
@ -78,7 +78,7 @@ JVM_ENTRY_NO_ENV(void*, JVM_RegisterSignal(jint sig, void* handler))
case SHUTDOWN2_SIGNAL:
case SHUTDOWN3_SIGNAL:
if (ReduceSignalUsage) return (void*)-1;
if (os::Bsd::is_sig_ignored(sig)) return (void*)1;
if (os::Posix::is_sig_ignored(sig)) return (void*)1;
}
void* oldHandler = os::signal(sig, newHandler);
@ -102,7 +102,7 @@ JVM_ENTRY_NO_ENV(jboolean, JVM_RaiseSignal(jint sig))
}
}
else if ((sig == SHUTDOWN1_SIGNAL || sig == SHUTDOWN2_SIGNAL ||
sig == SHUTDOWN3_SIGNAL) && os::Bsd::is_sig_ignored(sig)) {
sig == SHUTDOWN3_SIGNAL) && os::Posix::is_sig_ignored(sig)) {
// do not allow SHUTDOWN1_SIGNAL to be raised when SHUTDOWN1_SIGNAL
// is ignored, since no handler for them is actually registered in JVM
// or via JVM_RegisterSignal.

12
src/hotspot/os/posix/os_posix.cpp
View File

@ -926,6 +926,18 @@ bool os::Posix::is_valid_signal(int sig) {
#endif
}
bool os::Posix::is_sig_ignored(int sig) {
struct sigaction oact;
sigaction(sig, (struct sigaction*)NULL, &oact);
void* ohlr = oact.sa_sigaction ? CAST_FROM_FN_PTR(void*, oact.sa_sigaction)
: CAST_FROM_FN_PTR(void*, oact.sa_handler);
if (ohlr == CAST_FROM_FN_PTR(void*, SIG_IGN)) {
return true;
} else {
return false;
}
}
// Returns:
// NULL for an invalid signal number
// "SIG<num>" for a valid but unknown signal number

3
src/hotspot/os/posix/os_posix.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1999, 2017, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1999, 2018, 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
@ -64,6 +64,7 @@ public:
// Returns true if signal is valid.
static bool is_valid_signal(int sig);
static bool is_sig_ignored(int sig);
// Helper function, returns a string (e.g. "SIGILL") for a signal.
// Returned string is a constant. For unknown signals "UNKNOWN" is returned.

108
src/hotspot/os/solaris/jvm_solaris.cpp
View File

@ -1,108 +0,0 @@
/*
* Copyright (c) 1998, 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.
*
*/
#include "precompiled.hpp"
#include "jvm.h"
#include "runtime/interfaceSupport.inline.hpp"
#include "runtime/osThread.hpp"
#include <signal.h>
// sun.misc.Signal ///////////////////////////////////////////////////////////
// Signal code is mostly copied from classic vm, signals_md.c 1.4 98/08/23
/*
* This function is included primarily as a debugging aid. If Java is
* running in a console window, then pressing <CTRL-\\> will cause
* the current state of all active threads and monitors to be written
* to the console window.
*/
JVM_ENTRY_NO_ENV(void*, JVM_RegisterSignal(jint sig, void* handler))
// Copied from classic vm
// signals_md.c 1.4 98/08/23
void* newHandler = handler == (void *)2
? os::user_handler()
: handler;
switch (sig) {
/* The following are already used by the VM. */
case SIGFPE:
case SIGILL:
case SIGSEGV:
/* The following signal is used by the VM to dump thread stacks unless
ReduceSignalUsage is set, in which case the user is allowed to set
his own _native_ handler for this signal; thus, in either case,
we do not allow JVM_RegisterSignal to change the handler. */
case BREAK_SIGNAL:
return (void *)-1;
/* The following signals are used for Shutdown Hooks support. However, if
ReduceSignalUsage (-Xrs) is set, Shutdown Hooks must be invoked via
System.exit(), Java is not allowed to use these signals, and the the
user is allowed to set his own _native_ handler for these signals and
invoke System.exit() as needed. Terminator.setup() is avoiding
registration of these signals when -Xrs is present.
- If the HUP signal is ignored (from the nohup) command, then Java
is not allowed to use this signal.
*/
case SHUTDOWN1_SIGNAL:
case SHUTDOWN2_SIGNAL:
case SHUTDOWN3_SIGNAL:
if (ReduceSignalUsage) return (void*)-1;
if (os::Solaris::is_sig_ignored(sig)) return (void*)1;
}
void* oldHandler = os::signal(sig, newHandler);
if (oldHandler == os::user_handler()) {
return (void *)2;
} else {
return oldHandler;
}
JVM_END
JVM_ENTRY_NO_ENV(jboolean, JVM_RaiseSignal(jint sig))
if (ReduceSignalUsage) {
// do not allow SHUTDOWN1_SIGNAL,SHUTDOWN2_SIGNAL,SHUTDOWN3_SIGNAL,
// BREAK_SIGNAL to be raised when ReduceSignalUsage is set, since
// no handler for them is actually registered in JVM or via
// JVM_RegisterSignal.
if (sig == SHUTDOWN1_SIGNAL || sig == SHUTDOWN2_SIGNAL ||
sig == SHUTDOWN3_SIGNAL || sig == BREAK_SIGNAL) {
return JNI_FALSE;
}
}
else if ((sig == SHUTDOWN1_SIGNAL || sig == SHUTDOWN2_SIGNAL ||
sig == SHUTDOWN3_SIGNAL) && os::Solaris::is_sig_ignored(sig)) {
// do not allow SHUTDOWN1_SIGNAL to be raised when SHUTDOWN1_SIGNAL
// is ignored, since no handler for them is actually registered in JVM
// or via JVM_RegisterSignal.
// This also applies for SHUTDOWN2_SIGNAL and SHUTDOWN3_SIGNAL
return JNI_FALSE;
}
os::signal_raise(sig);
return JNI_TRUE;
JVM_END

32
src/hotspot/os/solaris/os_solaris.cpp
View File

@ -291,6 +291,12 @@ void os::Solaris::initialize_system_info() {
(julong)sysconf(_SC_PAGESIZE);
}
uint os::processor_id() {
const processorid_t id = ::getcpuid();
assert(id >= 0 && id < _processor_count, "Invalid processor id");
return (uint)id;
}
int os::active_processor_count() {
// User has overridden the number of active processors
if (ActiveProcessorCount > 0) {
@ -1026,18 +1032,6 @@ bool os::create_thread(Thread* thread, ThreadType thr_type,
debug_only(static bool signal_sets_initialized = false);
static sigset_t unblocked_sigs, vm_sigs;
bool os::Solaris::is_sig_ignored(int sig) {
struct sigaction oact;
sigaction(sig, (struct sigaction*)NULL, &oact);
void* ohlr = oact.sa_sigaction ? CAST_FROM_FN_PTR(void*, oact.sa_sigaction)
: CAST_FROM_FN_PTR(void*, oact.sa_handler);
if (ohlr == CAST_FROM_FN_PTR(void*, SIG_IGN)) {
return true;
} else {
return false;
}
}
void os::Solaris::signal_sets_init() {
// Should also have an assertion stating we are still single-threaded.
assert(!signal_sets_initialized, "Already initialized");
@ -1062,13 +1056,13 @@ void os::Solaris::signal_sets_init() {
sigaddset(&unblocked_sigs, ASYNC_SIGNAL);
if (!ReduceSignalUsage) {
if (!os::Solaris::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN1_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN1_SIGNAL);
}
if (!os::Solaris::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN2_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN2_SIGNAL);
}
if (!os::Solaris::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
if (!os::Posix::is_sig_ignored(SHUTDOWN3_SIGNAL)) {
sigaddset(&unblocked_sigs, SHUTDOWN3_SIGNAL);
}
}
@ -2130,7 +2124,7 @@ void os::Solaris::init_signal_mem() {
memset(ourSigFlags, 0, sizeof(int) * (Maxsignum + 1));
}
void os::signal_init_pd() {
static void jdk_misc_signal_init() {
// Initialize signal semaphore
sig_sem = new Semaphore();
}
@ -2140,7 +2134,7 @@ void os::signal_notify(int sig) {
Atomic::inc(&pending_signals[sig]);
sig_sem->signal();
} else {
// Signal thread is not created with ReduceSignalUsage and signal_init_pd
// Signal thread is not created with ReduceSignalUsage and jdk_misc_signal_init
// initialization isn't called.
assert(ReduceSignalUsage, "signal semaphore should be created");
}
@ -4249,6 +4243,10 @@ jint os::init_2(void) {
Solaris::signal_sets_init();
Solaris::init_signal_mem();
Solaris::install_signal_handlers();
// Initialize data for jdk.internal.misc.Signal
if (!ReduceSignalUsage) {
jdk_misc_signal_init();
}
// initialize synchronization primitives to use either thread or
// lwp synchronization (controlled by UseLWPSynchronization)

3
src/hotspot/os/solaris/os_solaris.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1997, 2017, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1997, 2018, 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
@ -188,7 +188,6 @@ class Solaris {
static void install_signal_handlers();
static void set_signal_handler(int sig, bool set_installed, bool oktochain);
static void init_signal_mem();
static bool is_sig_ignored(int sig);
static void set_our_sigflags(int, int);
static int get_our_sigflags(int);

8
src/hotspot/os/solaris/os_solaris.inline.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1997, 2017, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1997, 2018, 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
@ -37,9 +37,9 @@
#include <netdb.h>
#include <setjmp.h>
// File names are case-sensitive on windows only
inline int os::file_name_strcmp(const char* s1, const char* s2) {
return strcmp(s1, s2);
// File names are case-insensitive on windows only
inline int os::file_name_strncmp(const char* s1, const char* s2, size_t num) {
return strncmp(s1, s2, num);
}
inline bool os::uses_stack_guard_pages() {

58
src/hotspot/os/windows/os_windows.cpp
View File

@ -1745,13 +1745,46 @@ void os::print_memory_info(outputStream* st) {
// value if total memory is larger than 4GB
MEMORYSTATUSEX ms;
ms.dwLength = sizeof(ms);
GlobalMemoryStatusEx(&ms);
int r1 = GlobalMemoryStatusEx(&ms);
st->print(", physical %uk", os::physical_memory() >> 10);
st->print("(%uk free)", os::available_memory() >> 10);
if (r1 != 0) {
st->print(", system-wide physical " INT64_FORMAT "M ",
(int64_t) ms.ullTotalPhys >> 20);
st->print("(" INT64_FORMAT "M free)\n", (int64_t) ms.ullAvailPhys >> 20);
st->print("TotalPageFile size " INT64_FORMAT "M ",
(int64_t) ms.ullTotalPageFile >> 20);
st->print("(AvailPageFile size " INT64_FORMAT "M)",
(int64_t) ms.ullAvailPageFile >> 20);
// on 32bit Total/AvailVirtual are interesting (show us how close we get to 2-4 GB per process borders)
#if defined(_M_IX86)
st->print(", user-mode portion of virtual address-space " INT64_FORMAT "M ",
(int64_t) ms.ullTotalVirtual >> 20);
st->print("(" INT64_FORMAT "M free)", (int64_t) ms.ullAvailVirtual >> 20);
#endif
} else {
st->print(", GlobalMemoryStatusEx did not succeed so we miss some memory values.");
}
// extended memory statistics for a process
PROCESS_MEMORY_COUNTERS_EX pmex;
ZeroMemory(&pmex, sizeof(PROCESS_MEMORY_COUNTERS_EX));
pmex.cb = sizeof(pmex);
int r2 = GetProcessMemoryInfo(GetCurrentProcess(), (PROCESS_MEMORY_COUNTERS*) &pmex, sizeof(pmex));
if (r2 != 0) {
st->print("\ncurrent process WorkingSet (physical memory assigned to process): " INT64_FORMAT "M, ",
(int64_t) pmex.WorkingSetSize >> 20);
st->print("peak: " INT64_FORMAT "M\n", (int64_t) pmex.PeakWorkingSetSize >> 20);
st->print("current process commit charge (\"private bytes\"): " INT64_FORMAT "M, ",
(int64_t) pmex.PrivateUsage >> 20);
st->print("peak: " INT64_FORMAT "M", (int64_t) pmex.PeakPagefileUsage >> 20);
} else {
st->print("\nGetProcessMemoryInfo did not succeed so we miss some memory values.");
}
st->print(", swap %uk", ms.ullTotalPageFile >> 10);
st->print("(%uk free)", ms.ullAvailPageFile >> 10);
st->cr();
}
@ -1993,7 +2026,7 @@ int os::sigexitnum_pd() {
static volatile jint pending_signals[NSIG+1] = { 0 };
static Semaphore* sig_sem = NULL;
void os::signal_init_pd() {
static void jdk_misc_signal_init() {
// Initialize signal structures
memset((void*)pending_signals, 0, sizeof(pending_signals));
@ -2014,10 +2047,8 @@ void os::signal_init_pd() {
// the CTRL-BREAK thread dump mechanism is also disabled in this
// case. See bugs 4323062, 4345157, and related bugs.
if (!ReduceSignalUsage) {
// Add a CTRL-C handler
SetConsoleCtrlHandler(consoleHandler, TRUE);
}
// Add a CTRL-C handler
SetConsoleCtrlHandler(consoleHandler, TRUE);
}
void os::signal_notify(int sig) {
@ -2025,7 +2056,7 @@ void os::signal_notify(int sig) {
Atomic::inc(&pending_signals[sig]);
sig_sem->signal();
} else {
// Signal thread is not created with ReduceSignalUsage and signal_init_pd
// Signal thread is not created with ReduceSignalUsage and jdk_misc_signal_init
// initialization isn't called.
assert(ReduceSignalUsage, "signal semaphore should be created");
}
@ -4132,6 +4163,11 @@ jint os::init_2(void) {
SymbolEngine::recalc_search_path();
// Initialize data for jdk.internal.misc.Signal
if (!ReduceSignalUsage) {
jdk_misc_signal_init();
}
return JNI_OK;
}

6
src/hotspot/os/windows/os_windows.inline.hpp
View File

@ -1,5 +1,5 @@
/*
* Copyright (c) 1997, 2016, Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1997, 2018, 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
@ -33,8 +33,8 @@ inline const char* os::dll_file_extension() { return ".dll"; }
inline const int os::default_file_open_flags() { return O_BINARY | O_NOINHERIT;}
// File names are case-insensitive on windows only
inline int os::file_name_strcmp(const char* s, const char* t) {
return _stricmp(s, t);
inline int os::file_name_strncmp(const char* s, const char* t, size_t num) {
return _strnicmp(s, t, num);
}
inline void os::dll_unload(void *lib) {

7
src/hotspot/os_cpu/aix_ppc/thread_aix_ppc.cpp
View File

@ -40,6 +40,13 @@ frame JavaThread::pd_last_frame() {
return frame(sp, pc);
}
bool JavaThread::pd_get_top_frame_for_profiling(frame* fr_addr, void* ucontext, bool isInJava) {
ucontext_t* uc = (ucontext_t*) ucontext;
*fr_addr = frame((intptr_t*)uc->uc_mcontext.jmp_context.gpr[1/*REG_SP*/],
(address)uc->uc_mcontext.jmp_context.iar);
return true;
}
// Forte Analyzer AsyncGetCallTrace profiling support is not implemented on Aix/PPC.
bool JavaThread::pd_get_top_frame_for_signal_handler(frame* fr_addr, void* ucontext, bool isInJava) {
Unimplemented();

3
src/hotspot/os_cpu/aix_ppc/thread_aix_ppc.hpp
View File

@ -48,4 +48,7 @@
bool pd_get_top_frame_for_signal_handler(frame* fr_addr, void* ucontext,
bool isInJava);
bool pd_get_top_frame_for_profiling(frame* fr_addr, void* ucontext,
bool isInJava);
#endif // OS_CPU_AIX_PPC_VM_THREAD_AIX_PPC_HPP

7
src/hotspot/os_cpu/linux_ppc/thread_linux_ppc.cpp
View File

@ -41,6 +41,13 @@ frame JavaThread::pd_last_frame() {
return frame(sp, pc);
}
bool JavaThread::pd_get_top_frame_for_profiling(frame* fr_addr, void* ucontext, bool isInJava) {
ucontext_t* uc = (ucontext_t*) ucontext;
*fr_addr = frame((intptr_t*)uc->uc_mcontext.regs->gpr[1/*REG_SP*/],
(address)uc->uc_mcontext.regs->nip);
return true;
}
// Forte Analyzer AsyncGetCallTrace profiling support is not implemented on Linux/PPC.
bool JavaThread::pd_get_top_frame_for_signal_handler(frame* fr_addr, void* ucontext, bool isInJava) {
Unimplemented();

2
src/hotspot/os_cpu/linux_ppc/thread_linux_ppc.hpp
View File

@ -49,4 +49,6 @@
bool pd_get_top_frame_for_signal_handler(frame* fr_addr, void* ucontext, bool isInJava);
bool pd_get_top_frame_for_profiling(frame* fr_addr, void* ucontext, bool isInJava);
#endif // OS_CPU_LINUX_PPC_VM_THREAD_LINUX_PPC_HPP

7
src/hotspot/os_cpu/linux_s390/thread_linux_s390.cpp
View File

@ -41,6 +41,13 @@ frame JavaThread::pd_last_frame() {
return frame(sp, pc);
}
bool JavaThread::pd_get_top_frame_for_profiling(frame* fr_addr, void* ucontext, bool isInJava) {
ucontext_t* uc = (ucontext_t*) ucontext;
*fr_addr = frame((intptr_t*)uc->uc_mcontext.gregs[15/*REG_SP*/],
(address)uc->uc_mcontext.psw.addr);
return true;
}
// Forte Analyzer AsyncGetCallTrace profiling support is not implemented on Linux/S390x.
bool JavaThread::pd_get_top_frame_for_signal_handler(frame* fr_addr, void* ucontext, bool isInJava) {
Unimplemented();

2
src/hotspot/os_cpu/linux_s390/thread_linux_s390.hpp
View File

@ -48,4 +48,6 @@
bool pd_get_top_frame_for_signal_handler(frame* fr_addr, void* ucontext, bool isInJava);
bool pd_get_top_frame_for_profiling(frame* fr_addr, void* ucontext, bool isInJava);
#endif // OS_CPU_LINUX_S390_VM_THREAD_LINUX_S390_HPP

8
src/hotspot/os_cpu/linux_zero/thread_linux_zero.hpp
View File

@ -110,6 +110,14 @@
return false; // silence compile warning
}
bool pd_get_top_frame_for_profiling(frame* fr_addr,
void* ucontext,
bool isInJava) {
ShouldNotCallThis();
return false; // silence compile warning
}
// These routines are only used on cpu architectures that
// have separate register stacks (Itanium).
static bool register_stack_overflow() { return false; }

2
src/hotspot/share/c1/c1_Canonicalizer.cpp
View File

@ -779,7 +779,7 @@ void Canonicalizer::do_If(If* x) {
if (cmp->x() == cmp->y()) {
do_If(canon);
} else {
if (compilation()->profile_branches()) {
if (compilation()->profile_branches() || compilation()->count_backedges()) {
// TODO: If profiling, leave floating point comparisons unoptimized.
// We currently do not support profiling of the unordered case.
switch(cmp->op()) {

Some files were not shown because too many files have changed in this diff Show More