1JAVA(1) JDK Commands JAVA(1)
2
6 java - launch a Java application
7
9 To launch a class file:
10 java [options] mainclass [args ...]
12 To launch the main class in a JAR file:
14 java [options] -jar jarfile [args ...]
16 To launch the main class in a module:
18 java [options] -m module[/mainclass] [args ...]
20 or
22 java [options] --module module[/mainclass] [args ...]
24 To launch a single source-file program:
26 java [options] source-file [args ...]
28 options
30 Optional: Specifies command-line options separated by spaces.
31 See Overview of Java Options for a description of available op‐
32 tions.
33 mainclass
35 Specifies the name of the class to be launched. Command-line
36 entries following classname are the arguments for the main
37 method.
38 -jar jarfile
40 Executes a program encapsulated in a JAR file. The jarfile ar‐
41 gument is the name of a JAR file with a manifest that contains a
42 line in the form Main-Class:classname that defines the class
43 with the public static void main(String[] args) method that
44 serves as your application's starting point. When you use -jar,
45 the specified JAR file is the source of all user classes, and
46 other class path settings are ignored. If you're using JAR
47 files, then see jar.
48 -m or --module module[/mainclass]
50 Executes the main class in a module specified by mainclass if it
51 is given, or, if it is not given, the value in the module. In
52 other words, mainclass can be used when it is not specified by
53 the module, or to override the value when it is specified.
54 See Standard Options for Java.
56 source-file
58 Only used to launch a single source-file program. Specifies the
59 source file that contains the main class when using source-file
60 mode. See Using Source-File Mode to Launch Single-File
61 Source-Code Programs
62 args ...
64 Optional: Arguments following mainclass, source-file, -jar
65 jarfile, and -m or --module module/mainclass are passed as argu‐
66 ments to the main class.
67
69 The java command starts a Java application. It does this by starting
70 the Java Virtual Machine (JVM), loading the specified class, and call‐
71 ing that class's main() method. The method must be declared public and
72 static, it must not return any value, and it must accept a String array
73 as a parameter. The method declaration has the following form:
74 public static void main(String[] args)
76 In source-file mode, the java command can launch a class declared in a
78 source file. See Using Source-File Mode to Launch Single-File
79 Source-Code Programs for a description of using the source-file mode.
80 Note: You can use the JDK_JAVA_OPTIONS launcher environment
82 variable to prepend its content to the actual command line of
83 the java launcher. See Using the JDK_JAVA_OPTIONS Launcher En‐
84 vironment Variable.
85 By default, the first argument that isn't an option of the java command
87 is the fully qualified name of the class to be called. If -jar is
88 specified, then its argument is the name of the JAR file containing
89 class and resource files for the application. The startup class must
90 be indicated by the Main-Class manifest header in its manifest file.
91 Arguments after the class file name or the JAR file name are passed to
93 the main() method.
94 javaw
96 Windows: The javaw command is identical to java, except that with javaw
97 there's no associated console window. Use javaw when you don't want a
98 command prompt window to appear. The javaw launcher will, however,
99 display a dialog box with error information if a launch fails.
100
102 To launch a class declared in a source file, run the java launcher in
103 source-file mode. Entering source-file mode is determined by two items
104 on the java command line:
105 · The first item on the command line that is not an option or part of
107 an option. In other words, the item in the command line that would
108 otherwise be the main class name.
109 · The --source version option, if present.
111 If the class identifies an existing file that has a .java extension, or
113 if the --source option is specified, then source-file mode is selected.
114 The source file is then compiled and run. The --source option can be
115 used to specify the source version or N of the source code. This de‐
116 termines the API that can be used. When you set --source N, you can
117 only use the public API that was defined in JDK N.
118 Note: The valid values of N change for each release, with new
120 values added and old values removed. You'll get an error mes‐
121 sage if you use a value of N that is no longer supported. Sup‐
122 ported values of N for this release are 7, 8, 9, 10, 11, 12, 13,
123 and 14.
124 If the file does not have the .java extension, the --source option must
126 be used to tell the java command to use the source-file mode. The
127 --source option is used for cases when the source file is a "script" to
128 be executed and the name of the source file does not follow the normal
129 naming conventions for Java source files.
130 In source-file mode, the effect is as though the source file is com‐
132 piled into memory, and the first class found in the source file is exe‐
133 cuted. Any arguments placed after the name of the source file in the
134 original command line are passed to the compiled class when it is exe‐
135 cuted.
136 For example, if a file were named HelloWorld.java and contained a class
138 named hello.World, then the source-file mode command to launch the
139 class would be:
140 java HelloWorld.java
142 The example illustrates that the class can be in a named package, and
144 does not need to be in the unnamed package. This use of source-file
145 mode is informally equivalent to using the following two commands where
146 hello.World is the name of the class in the package:
147 javac -d <memory> HelloWorld.java
149 java -cp <memory> hello.World
150 In source-file mode, any additional command-line options are processed
152 as follows:
153 · The launcher scans the options specified before the source file for
155 any that are relevant in order to compile the source file.
156 This includes: --class-path, --module-path, --add-exports, --add-mod‐
158 ules, --limit-modules, --patch-module, --upgrade-module-path, and any
159 variant forms of those options. It also includes the new --en‐
160 able-preview option, described in JEP 12.
161 · No provision is made to pass any additional options to the compiler,
163 such as -processor or -Werror.
164 · Command-line argument files (@-files) may be used in the standard
166 way. Long lists of arguments for either the VM or the program being
167 invoked may be placed in files specified on the command-line by pre‐
168 fixing the filename with an @ character.
169 In source-file mode, compilation proceeds as follows:
171 · Any command-line options that are relevant to the compilation envi‐
173 ronment are taken into account.
174 · No other source files are found and compiled, as if the source path
176 is set to an empty value.
177 · Annotation processing is disabled, as if -proc:none is in effect.
179 · If a version is specified, via the --source option, the value is used
181 as the argument for an implicit --release option for the compilation.
182 This sets both the source version accepted by compiler and the system
183 API that may be used by the code in the source file.
184 · The source file is compiled in the context of an unnamed module.
186 · The source file should contain one or more top-level classes, the
188 first of which is taken as the class to be executed.
189 · The compiler does not enforce the optional restriction defined at the
191 end of JLS ??7.6, that a type in a named package should exist in a
192 file whose name is composed from the type name followed by the .java
193 extension.
194 · If the source file contains errors, appropriate error messages are
196 written to the standard error stream, and the launcher exits with a
197 non-zero exit code.
198 In source-file mode, execution proceeds as follows:
200 · The class to be executed is the first top-level class found in the
202 source file. It must contain a declaration of the standard pub‐
203 lic static void main(String[]) method.
204 · The compiled classes are loaded by a custom class loader, that dele‐
206 gates to the application class loader. This implies that classes ap‐
207 pearing on the application class path cannot refer to any classes de‐
208 clared in the source file.
209 · The compiled classes are executed in the context of an unnamed mod‐
211 ule, as though --add-modules=ALL-DEFAULT is in effect. This is in
212 addition to any other --add-module options that may be have been
213 specified on the command line.
214 · Any arguments appearing after the name of the file on the command
216 line are passed to the standard main method in the obvious way.
217 · It is an error if there is a class on the application class path
219 whose name is the same as that of the class to be executed.
220 See JEP 330: Launch Single-File Source-Code Programs [http://open‐
222 jdk.java.net/jeps/330] for complete details.
223
225 JDK_JAVA_OPTIONS prepends its content to the options parsed from the
226 command line. The content of the JDK_JAVA_OPTIONS environment variable
227 is a list of arguments separated by white-space characters (as deter‐
228 mined by isspace()). These are prepended to the command line arguments
229 passed to java launcher. The encoding requirement for the environment
230 variable is the same as the java command line on the system. JDK_JA‐
231 VA_OPTIONS environment variable content is treated in the same manner
232 as that specified in the command line.
233 Single (') or double (") quotes can be used to enclose arguments that
235 contain whitespace characters. All content between the open quote and
236 the first matching close quote are preserved by simply removing the
237 pair of quotes. In case a matching quote is not found, the launcher
238 will abort with an error message. @-files are supported as they are
239 specified in the command line. However, as in @-files, use of a wild‐
240 card is not supported. In order to mitigate potential misuse of
241 JDK_JAVA_OPTIONS behavior, options that specify the main class (such as
242 -jar) or cause the java launcher to exit without executing the main
243 class (such as -h) are disallowed in the environment variable. If any
244 of these options appear in the environment variable, the launcher will
245 abort with an error message. When JDK_JAVA_OPTIONS is set, the launch‐
246 er prints a message to stderr as a reminder.
247 Example:
249 $ export JDK_JAVA_OPTIONS='-g @file1 -Dprop=value @file2 -Dws.prop="white spaces"'
251 $ java -Xint @file3
252 is equivalent to the command line:
254 java -g @file1 -Dprop=value @file2 -Dws.prop="white spaces" -Xint @file3
256
258 The java command supports a wide range of options in the following cat‐
259 egories:
260 · Standard Options for Java: Options guaranteed to be supported by all
262 implementations of the Java Virtual Machine (JVM). They're used for
263 common actions, such as checking the version of the JRE, setting the
264 class path, enabling verbose output, and so on.
265 · Extra Options for Java: General purpose options that are specific to
267 the Java HotSpot Virtual Machine. They aren't guaranteed to be sup‐
268 ported by all JVM implementations, and are subject to change. These
269 options start with -X.
270 The advanced options aren't recommended for casual use. These are de‐
272 veloper options used for tuning specific areas of the Java HotSpot Vir‐
273 tual Machine operation that often have specific system requirements and
274 may require privileged access to system configuration parameters. Sev‐
275 eral examples of performance tuning are provided in Performance Tuning
276 Examples. These options aren't guaranteed to be supported by all JVM
277 implementations and are subject to change. Advanced options start with
278 -XX.
279 · Advanced Runtime Options for Java: Control the runtime behavior of
281 the Java HotSpot VM.
282 · Advanced JIT Compiler Options for java: Control the dynamic
284 just-in-time (JIT) compilation performed by the Java HotSpot VM.
285 · Advanced Serviceability Options for Java: Enable gathering system in‐
287 formation and performing extensive debugging.
288 · Advanced Garbage Collection Options for Java: Control how garbage
290 collection (GC) is performed by the Java HotSpot
291 Boolean options are used to either enable a feature that's disabled by
293 default or disable a feature that's enabled by default. Such options
294 don't require a parameter. Boolean -XX options are enabled using the
295 plus sign (-XX:+OptionName) and disabled using the minus sign (-XX:-Op‐
296 tionName).
297 For options that require an argument, the argument may be separated
299 from the option name by a space, a colon (:), or an equal sign (=), or
300 the argument may directly follow the option (the exact syntax differs
301 for each option). If you're expected to specify the size in bytes,
302 then you can use no suffix, or use the suffix k or K for kilobytes
303 (KB), m or M for megabytes (MB), or g or G for gigabytes (GB). For ex‐
304 ample, to set the size to 8 GB, you can specify either 8g, 8192m,
305 8388608k, or 8589934592 as the argument. If you are expected to speci‐
306 fy the percentage, then use a number from 0 to 1. For example, specify
307 0.25 for 25%.
308 The following sections describe the options that are obsolete, depre‐
310 cated, and removed:
311 · Deprecated Java Options: Accepted and acted upon --- a warning is is‐
313 sued when they're used.
314 · Obsolete Java Options: Accepted but ignored --- a warning is issued
316 when they're used.
317 · Removed Java Options: Removed --- using them results in an error.
319
321 These are the most commonly used options supported by all implementa‐
322 tions of the JVM.
323 Note: To specify an argument for a long option, you can use ei‐
325 ther --name=value or --name value.
326 -agentlib:libname[=options]
328 Loads the specified native agent library. After the library
329 name, a comma-separated list of options specific to the library
330 can be used.
331 · Oracle Solaris, Linux, and macOS: If the option -agentlib:foo
333 is specified, then the JVM attempts to load the library named
334 libfoo.so in the location specified by the LD_LIBRARY_PATH
335 system variable (on macOS this variable is DYLD_LIBRARY_PATH).
336 · Windows: If the option -agentlib:foo is specified, then the
338 JVM attempts to load the library named foo.dll in the location
339 specified by the PATH system variable.
340 The following example shows how to load the Java Debug Wire
342 Protocol (JDWP) library and listen for the socket connection
343 on port 8000, suspending the JVM before the main class loads:
344 -agentlib:jdwp=transport=dt_socket,server=y,ad‐
346 dress=8000
347 -agentpath:pathname[=options]
349 Loads the native agent library specified by the absolute path
350 name. This option is equivalent to -agentlib but uses the full
351 path and file name of the library.
352 --class-path classpath, -classpath classpath, or -cp classpath
354 A semicolon (;) separated list of directories, JAR archives, and
355 ZIP archives to search for class files.
356 Specifying classpath overrides any setting of the CLASSPATH en‐
358 vironment variable. If the class path option isn't used and
359 classpath isn't set, then the user class path consists of the
360 current directory (.).
361 As a special convenience, a class path element that contains a
363 base name of an asterisk (*) is considered equivalent to speci‐
364 fying a list of all the files in the directory with the exten‐
365 sion .jar or .JAR . A Java program can't tell the difference
366 between the two invocations. For example, if the directory my‐
367 dir contains a.jar and b.JAR, then the class path element my‐
368 dir/* is expanded to A.jar:b.JAR, except that the order of JAR
369 files is unspecified. All .jar files in the specified directo‐
370 ry, even hidden ones, are included in the list. A class path
371 entry consisting of an asterisk (*) expands to a list of all the
372 jar files in the current directory. The CLASSPATH environment
373 variable, where defined, is similarly expanded. Any class path
374 wildcard expansion that occurs before the Java VM is started.
375 Java programs never see wildcards that aren't expanded except by
376 querying the environment, such as by calling Sys‐
377 tem.getenv("CLASSPATH").
378 --disable-@files
380 Can be used anywhere on the command line, including in an argu‐
381 ment file, to prevent further @filename expansion. This option
382 stops expanding @-argfiles after the option.
383 --enable-preview
385 Allows classes to depend on preview features [https://docs.ora‐
386 cle.com/en/java/javase/12/language/index.html#JS‐
387 LAN-GUID-5A82FE0E-0CA4-4F1F-B075-564874FE2823] of the release.
388 --module-path modulepath... or -p modulepath
390 A semicolon (;) separated list of directories in which each di‐
391 rectory is a directory of modules.
392 --upgrade-module-path modulepath...
394 A semicolon (;) separated list of directories in which each di‐
395 rectory is a directory of modules that replace upgradeable mod‐
396 ules in the runtime image.
397 --add-modules module[,module...]
399 Specifies the root modules to resolve in addition to the initial
400 module. module also can be ALL-DEFAULT, ALL-SYSTEM, and
401 ALL-MODULE-PATH.
402 --list-modules
404 Lists the observable modules and then exits.
405 -d module_name or --describe-module module_name
407 Describes a specified module and then exits.
408 --dry-run
410 Creates the VM but doesn't execute the main method. This
411 --dry-run option might be useful for validating the command-line
412 options such as the module system configuration.
413 --validate-modules
415 Validates all modules and exit. This option is helpful for
416 finding conflicts and other errors with modules on the module
417 path.
418 -Dproperty=value
420 Sets a system property value. The property variable is a string
421 with no spaces that represents the name of the property. The
422 value variable is a string that represents the value of the
423 property. If value is a string with spaces, then enclose it in
424 quotation marks (for example -Dfoo="foo bar").
425 -disableassertions[:[packagename]...|:classname] or -da[:[package‐
427 name]...|:classname]
428 Disables assertions. By default, assertions are disabled in all
429 packages and classes. With no arguments, -disableassertions
430 (-da) disables assertions in all packages and classes. With the
431 packagename argument ending in ..., the switch disables asser‐
432 tions in the specified package and any subpackages. If the ar‐
433 gument is simply ..., then the switch disables assertions in the
434 unnamed package in the current working directory. With the
435 classname argument, the switch disables assertions in the speci‐
436 fied class.
437 The -disableassertions (-da) option applies to all class loaders
439 and to system classes (which don't have a class loader).
440 There's one exception to this rule: If the option is provided
441 with no arguments, then it doesn't apply to system classes.
442 This makes it easy to disable assertions in all classes except
443 for system classes. The -disablesystemassertions option enables
444 you to disable assertions in all system classes. To explicitly
445 enable assertions in specific packages or classes, use the -en‐
446 ableassertions (-ea) option. Both options can be used at the
447 same time. For example, to run the MyClass application with as‐
448 sertions enabled in the package com.wombat.fruitbat (and any
449 subpackages) but disabled in the class com.wombat.fruit‐
450 bat.Brickbat, use the following command:
451 java -ea:com.wombat.fruitbat... -da:com.wombat.fruit‐
453 bat.Brickbat MyClass
454 -disablesystemassertions or -dsa
456 Disables assertions in all system classes.
457 -enableassertions[:[packagename]...|:classname] or -ea[:[package‐
459 name]...|:classname]
460 Enables assertions. By default, assertions are disabled in all
461 packages and classes. With no arguments, -enableassertions
462 (-ea) enables assertions in all packages and classes. With the
463 packagename argument ending in ..., the switch enables asser‐
464 tions in the specified package and any subpackages. If the ar‐
465 gument is simply ..., then the switch enables assertions in the
466 unnamed package in the current working directory. With the
467 classname argument, the switch enables assertions in the speci‐
468 fied class.
469 The -enableassertions (-ea) option applies to all class loaders
471 and to system classes (which don't have a class loader).
472 There's one exception to this rule: If the option is provided
473 with no arguments, then it doesn't apply to system classes.
474 This makes it easy to enable assertions in all classes except
475 for system classes. The -enablesystemassertions option provides
476 a separate switch to enable assertions in all system classes.
477 To explicitly disable assertions in specific packages or class‐
478 es, use the -disableassertions (-da) option. If a single com‐
479 mand contains multiple instances of these switches, then they're
480 processed in order, before loading any classes. For example, to
481 run the MyClass application with assertions enabled only in the
482 package com.wombat.fruitbat (and any subpackages) but disabled
483 in the class com.wombat.fruitbat.Brickbat, use the following
484 command:
485 java -ea:com.wombat.fruitbat... -da:com.wombat.fruit‐
487 bat.Brickbat MyClass
488 -enablesystemassertions or -esa
490 Enables assertions in all system classes.
491 -help, -h, or -?
493 Prints the help message to the error stream.
494 --help Prints the help message to the output stream.
496 -javaagent:jarpath[=options]
498 Loads the specified Java programming language agent. See ja‐
499 va.lang.instrument.
500 --show-version
502 Prints the product version to the output stream and continues.
503 -showversion
505 Prints the product version to the error stream and continues.
506 --show-module-resolution
508 Shows module resolution output during startup.
509 -splash:imagepath
511 Shows the splash screen with the image specified by imagepath.
512 HiDPI scaled images are automatically supported and used if
513 available. The unscaled image file name, such as image.ext,
514 should always be passed as the argument to the -splash option.
515 The most appropriate scaled image provided is picked up automat‐
516 ically.
517 For example, to show the splash.gif file from the images direc‐
519 tory when starting your application, use the following option:
520 -splash:images/splash.gif
522 See the SplashScreen API documentation for more information.
524 -verbose:class
526 Displays information about each loaded class.
527 -verbose:gc
529 Displays information about each garbage collection (GC) event.
530 -verbose:jni
532 Displays information about the use of native methods and other
533 Java Native Interface (JNI) activity.
534 -verbose:module
536 Displays information about the modules in use.
537 --version
539 Prints product version to the output stream and exits.
540 -version
542 Prints product version to the error stream and exits.
543 -X Prints the help on extra options to the error stream.
545 --help-extra
547 Prints the help on extra options to the output stream.
548 @argfile
550 Specifies one or more argument files prefixed by @ used by the
551 java command. It isn't uncommon for the java command line to be
552 very long because of the .jar files needed in the classpath.
553 The @argfile option overcomes command-line length limitations by
554 enabling the launcher to expand the contents of argument files
555 after shell expansion, but before argument processing. Contents
556 in the argument files are expanded because otherwise, they would
557 be specified on the command line until the --disable-@files op‐
558 tion was encountered.
559 The argument files can also contain the main class name and all
561 options. If an argument file contains all of the options re‐
562 quired by the java command, then the command line could simply
563 be:
564 java @argfile
566 See java Command-Line Argument Files for a description and exam‐
568 ples of using @-argfiles.
569
571 The following java options are general purpose options that are specif‐
572 ic to the Java HotSpot Virtual Machine.
573 -Xbatch
575 Disables background compilation. By default, the JVM compiles
576 the method as a background task, running the method in inter‐
577 preter mode until the background compilation is finished. The
578 -Xbatch flag disables background compilation so that compilation
579 of all methods proceeds as a foreground task until completed.
580 This option is equivalent to -XX:-BackgroundCompilation.
581 -Xbootclasspath/a:directories|zip|JAR-files
583 Specifies a list of directories, JAR files, and ZIP archives to
584 append to the end of the default bootstrap class path.
585 Oracle Solaris, Linux, and macOS: Colons (:) separate entities
587 in this list.
588 Windows: Semicolons (;) separate entities in this list.
590 -Xcheck:jni
592 Performs additional checks for Java Native Interface (JNI) func‐
593 tions. Specifically, it validates the parameters passed to the
594 JNI function and the runtime environment data before processing
595 the JNI request. It also checks for pending exceptions between
596 JNI calls. Any invalid data encountered indicates a problem in
597 the native code, and the JVM terminates with an irrecoverable
598 error in such cases. Expect a performance degradation when this
599 option is used.
600 -Xdebug
602 Does nothing. Provided for backward compatibility.
603 -Xdiag Shows additional diagnostic messages.
605 -Xint Runs the application in interpreted-only mode. Compilation to
607 native code is disabled, and all bytecode is executed by the in‐
608 terpreter. The performance benefits offered by the just-in-time
609 (JIT) compiler aren't present in this mode.
610 -Xinternalversion
612 Displays more detailed JVM version information than the -version
613 option, and then exits.
614 -Xlog:option
616 Configure or enable logging with the Java Virtual Machine (JVM)
617 unified logging framework. See Enable Logging with the JVM Uni‐
618 fied Logging Framework.
619 -Xmixed
621 Executes all bytecode by the interpreter except for hot methods,
622 which are compiled to native code. On by default. Use -Xint to
623 switch off.
624 -Xmn size
626 Sets the initial and maximum size (in bytes) of the heap for the
627 young generation (nursery) in the generational collectors. Ap‐
628 pend the letter k or K to indicate kilobytes, m or M to indicate
629 megabytes, or g or G to indicate gigabytes. The young genera‐
630 tion region of the heap is used for new objects. GC is per‐
631 formed in this region more often than in other regions. If the
632 size for the young generation is too small, then a lot of minor
633 garbage collections are performed. If the size is too large,
634 then only full garbage collections are performed, which can take
635 a long time to complete. It is recommended that you do not set
636 the size for the young generation for the G1 collector, and keep
637 the size for the young generation greater than 25% and less than
638 50% of the overall heap size for other collectors. The follow‐
639 ing examples show how to set the initial and maximum size of
640 young generation to 256 MB using various units:
641 -Xmn256m
643 -Xmn262144k
644 -Xmn268435456
645 Instead of the -Xmn option to set both the initial and maximum
647 size of the heap for the young generation, you can use -XX:New‐
648 Size to set the initial size and -XX:MaxNewSize to set the maxi‐
649 mum size.
650 -Xms size
652 Sets the minimum and initial size (in bytes) of the heap. This
653 value must be a multiple of 1024 and greater than 1 MB. Append
654 the letter k or K to indicate kilobytes, m or M to indicate
655 megabytes, g or G to indicate gigabytes. The following examples
656 show how to set the size of allocated memory to 6 MB using vari‐
657 ous units:
658 -Xms6291456
660 -Xms6144k
661 -Xms6m
662 Instead of the -Xms option to set both the minimum and initial
664 size of the heap, you can use -XX:MinHeapSize to set the minimum
665 size and -XX:InitialHeapSize to set the initial size.
666 If you don't set this option, the initial size is set as the sum
668 of the sizes allocated for the old generation and the young gen‐
669 eration. The initial size of the heap for the young generation
670 can be set using the -Xmn option or the -XX:NewSize option.
671 -Xmx size
673 Specifies the maximum size (in bytes) of the heap. This value
674 must be a multiple of 1024 and greater than 2 MB. Append the
675 letter k or K to indicate kilobytes, m or M to indicate
676 megabytes, or g or G to indicate gigabytes. The default value
677 is chosen at runtime based on system configuration. For server
678 deployments, -Xms and -Xmx are often set to the same value. The
679 following examples show how to set the maximum allowed size of
680 allocated memory to 80 MB using various units:
681 -Xmx83886080
683 -Xmx81920k
684 -Xmx80m
685 The -Xmx option is equivalent to -XX:MaxHeapSize.
687 -Xnoclassgc
689 Disables garbage collection (GC) of classes. This can save some
690 GC time, which shortens interruptions during the application
691 run. When you specify -Xnoclassgc at startup, the class objects
692 in the application are left untouched during GC and are always
693 be considered live. This can result in more memory being perma‐
694 nently occupied which, if not used carefully, throws an
695 out-of-memory exception.
696 -Xrs Reduces the use of operating system signals by the JVM. Shut‐
698 down hooks enable the orderly shutdown of a Java application by
699 running user cleanup code (such as closing database connections)
700 at shutdown, even if the JVM terminates abruptly.
701 · Oracle Solaris, Linux, and macOS:
703 · The JVM catches signals to implement shutdown hooks for un‐
705 expected termination. The JVM uses SIGHUP, SIGINT, and
706 SIGTERM to initiate the running of shutdown hooks.
707 · Applications embedding the JVM frequently need to trap sig‐
709 nals such as SIGINT or SIGTERM, which can lead to interfer‐
710 ence with the JVM signal handlers. The -Xrs option is
711 available to address this issue. When -Xrs is used, the
712 signal masks for SIGINT, SIGTERM, SIGHUP, and SIGQUIT aren't
713 changed by the JVM, and signal handlers for these signals
714 aren't installed.
715 · Windows:
717 · The JVM watches for console control events to implement
719 shutdown hooks for unexpected termination. Specifically,
720 the JVM registers a console control handler that begins
721 shutdown-hook processing and returns TRUE for CTRL_C_EVENT,
722 CTRL_CLOSE_EVENT, CTRL_LOGOFF_EVENT, and CTRL_SHUT‐
723 DOWN_EVENT.
724 · The JVM uses a similar mechanism to implement the feature of
726 dumping thread stacks for debugging purposes. The JVM uses
727 CTRL_BREAK_EVENT to perform thread dumps.
728 · If the JVM is run as a service (for example, as a servlet
730 engine for a web server), then it can receive CTRL_LO‐
731 GOFF_EVENT but shouldn't initiate shutdown because the oper‐
732 ating system doesn't actually terminate the process. To
733 avoid possible interference such as this, the -Xrs option
734 can be used. When the -Xrs option is used, the JVM doesn't
735 install a console control handler, implying that it doesn't
736 watch for or process CTRL_C_EVENT, CTRL_CLOSE_EVENT,
737 CTRL_LOGOFF_EVENT, or CTRL_SHUTDOWN_EVENT.
738 There are two consequences of specifying -Xrs:
740 · Oracle Solaris, Linux, and macOS: SIGQUIT thread dumps aren't
742 available.
743 · Windows: Ctrl + Break thread dumps aren't available.
745 User code is responsible for causing shutdown hooks to run, for
747 example, by calling the System.exit() when the JVM is to be ter‐
748 minated.
749 -Xshare:mode
751 Sets the class data sharing (CDS) mode.
752 Possible mode arguments for this option include the following:
754 auto Use shared class data if possible (default).
756 on Require using shared class data, otherwise fail.
758 Note: The -Xshare:on option is used for testing purposes
760 only and may cause intermittent failures due to the use
761 of address space layout randomization by the operation
762 system. This option should not be used in production en‐
763 vironments.
764 off Do not attempt to use shared class data.
766 -XshowSettings
768 Shows all settings and then continues.
769 -XshowSettings:category
771 Shows settings and continues. Possible category arguments for
772 this option include the following:
773 all Shows all categories of settings. This is the default
775 value.
776 locale Shows settings related to locale.
778 properties
780 Shows settings related to system properties.
781 vm Shows the settings of the JVM.
783 system Linux: Shows host system or container configuration and
785 continues.
786 -Xss size
788 Sets the thread stack size (in bytes). Append the letter k or K
789 to indicate KB, m or M to indicate MB, or g or G to indicate GB.
790 The default value depends on the platform:
791 · Linux/x64 (64-bit): 1024 KB
793 · macOS (64-bit): 1024 KB
795 · Oracle Solaris (64-bit): 1024 KB
797 · Windows: The default value depends on virtual memory
799 The following examples set the thread stack size to 1024 KB in
801 different units:
802 -Xss1m
804 -Xss1024k
805 -Xss1048576
806 This option is similar to -XX:ThreadStackSize.
808 --add-reads module=target-module(,target-module)*
810 Updates module to read the target-module, regardless of the mod‐
811 ule declaration. target-module can be all unnamed to read all
812 unnamed modules.
813 --add-exports module/package=target-module(,target-module)*
815 Updates module to export package to target-module, regardless of
816 module declaration. The target-module can be all unnamed to ex‐
817 port to all unnamed modules.
818 --add-opens module/package=target-module(,target-module)*
820 Updates module to open package to target-module, regardless of
821 module declaration.
822 --illegal-access=parameter
824 When present at run time, --illegal-access= takes a keyword pa‐
825 rameter to specify a mode of operation:
826 Note: This option will be removed in a future release.
828 · permit: This mode opens each package in each module in the
830 run-time image to code in all unnamed modules ( such as code
831 on the class path), if that package existed in JDK 8. This
832 enables both static access, (for example, by compiled byte‐
833 code, and deep reflective access) through the platform's vari‐
834 ous reflection APIs. The first reflective-access operation to
835 any such package causes a warning to be issued. However, no
836 warnings are issued after the first occurrence. This single
837 warning describes how to enable further warnings. This mode
838 is the default for the current JDK but will change in a future
839 release.
840 · warn: This mode is identical to permit except that a warning
842 message is issued for each illegal reflective-access opera‐
843 tion.
844 · debug: This mode is identical to warn except that both a warn‐
846 ing message and a stack trace are issued for each illegal re‐
847 flective-access operation.
848 · deny: This mode disables all illegal-access operations except
850 for those enabled by other command-line options, such as
851 --add-opens. This mode will become the default in a future
852 release.
853 The default mode, --illegal-access=permit, is intended to make
855 you aware of code on the class path that reflectively accesses
856 any JDK-internal APIs at least once. To learn about all such
857 accesses, you can use the warn or the debug modes. For each li‐
858 brary or framework on the class path that requires illegal ac‐
859 cess, you have two options:
860 · If the component's maintainers have already released a fixed
862 version that no longer uses JDK-internal APIs then you can
863 consider upgrading to that version.
864 · If the component still needs to be fixed, then you can contact
866 its maintainers and ask them to replace their use of JDK-in‐
867 ternal APIs with the proper exported APIs.
868 If you must continue to use a component that requires illegal
870 access, then you can eliminate the warning messages by using one
871 or more --add-opens options to open only those internal packages
872 to which access is required.
873 To verify that your application is ready for a future version of
875 the JDK, run it with --illegal-access=deny along with any neces‐
876 sary --add-opens options. Any remaining illegal-access errors
877 will most likely be due to static references from compiled code
878 to JDK-internal APIs. You can identify those by running the
879 jdeps tool with the --jdk-internals option. For performance
880 reasons, the current JDK does not issue warnings for illegal
881 static-access operations.
882 --limit-modules module[,module...]
884 Specifies the limit of the universe of observable modules.
885 --patch-module module=file(;file)*
887 Overrides or augments a module with classes and resources in JAR
888 files or directories.
889 --source version
891 Sets the version of the source in source-file mode.
892
894 The following extra options are macOS specific.
895 -XstartOnFirstThread
897 Runs the main() method on the first (AppKit) thread.
898 -Xdock:name=application_name
900 Overrides the default application name displayed in dock.
901 -Xdock:icon=path_to_icon_file
903 Overrides the default icon displayed in dock.
904
906 These java options can be used to enable other advanced options.
907 -XX:+UnlockDiagnosticVMOptions
909 Unlocks the options intended for diagnosing the JVM. By de‐
910 fault, this option is disabled and diagnostic options aren't
911 available.
912 Command line options that are enabled with the use of this op‐
914 tion are not supported. If you encounter issues while using any
915 of these options, it is very likely that you will be required to
916 reproduce the problem without using any of these unsupported op‐
917 tions before Oracle Support can assist with an investigation.
918 It is also possible that any of these options may be removed or
919 their behavior changed without any warning.
920 -XX:+UnlockExperimentalVMOptions
922 Unlocks the options that provide experimental features in the
923 JVM. By default, this option is disabled and experimental fea‐
924 tures aren't available.
925
927 These java options control the runtime behavior of the Java HotSpot VM.
928 -XX:ActiveProcessorCount=x
930 Overrides the number of CPUs that the VM will use to calculate
931 the size of thread pools it will use for various operations such
932 as Garbage Collection and ForkJoinPool.
933 The VM normally determines the number of available processors
935 from the operating system. This flag can be useful for parti‐
936 tioning CPU resources when running multiple Java processes in
937 docker containers. This flag is honored even if UseContainer‐
938 Support is not enabled. See -XX:-UseContainerSupport for a de‐
939 scription of enabling and disabling container support.
940 -XX:AllocateHeapAt=path
942 Takes a path to the file system and uses memory mapping to allo‐
943 cate the object heap on the memory device. Using this option
944 enables the HotSpot VM to allocate the Java object heap on an
945 alternative memory device, such as an NV-DIMM, specified by the
946 user.
947 Alternative memory devices that have the same semantics as DRAM,
949 including the semantics of atomic operations, can be used in‐
950 stead of DRAM for the object heap without changing the existing
951 application code. All other memory structures (such as the code
952 heap, metaspace, and thread stacks) continue to reside in DRAM.
953 Some operating systems expose non-DRAM memory through the file
955 system. Memory-mapped files in these file systems bypass the
956 page cache and provide a direct mapping of virtual memory to the
957 physical memory on the device. The existing heap related flags
958 (such as -Xmx and -Xms) and garbage-collection related flags
959 continue to work as before.
960 -XX:-CompactStrings
962 Disables the Compact Strings feature. By default, this option
963 is enabled. When this option is enabled, Java Strings contain‐
964 ing only single-byte characters are internally represented and
965 stored as single-byte-per-character Strings using ISO-8859-1 /
966 Latin-1 encoding. This reduces, by 50%, the amount of space re‐
967 quired for Strings containing only single-byte characters. For
968 Java Strings containing at least one multibyte character: these
969 are represented and stored as 2 bytes per character using UTF-16
970 encoding. Disabling the Compact Strings feature forces the use
971 of UTF-16 encoding as the internal representation for all Java
972 Strings.
973 Cases where it may be beneficial to disable Compact Strings in‐
975 clude the following:
976 · When it's known that an application overwhelmingly will be al‐
978 locating multibyte character Strings
979 · In the unexpected event where a performance regression is ob‐
981 served in migrating from Java SE 8 to Java SE 9 and an analy‐
982 sis shows that Compact Strings introduces the regression
983 In both of these scenarios, disabling Compact Strings makes
985 sense.
986 -XX:ErrorFile=filename
988 Specifies the path and file name to which error data is written
989 when an irrecoverable error occurs. By default, this file is
990 created in the current working directory and named hs_err_pid‐
991 pid.log where pid is the identifier of the process that encoun‐
992 tered the error.
993 The following example shows how to set the default log file
995 (note that the identifier of the process is specified as %p):
996 -XX:ErrorFile=./hs_err_pid%p.log
998 · Oracle Solaris, Linux, and macOS: The following example shows
1000 how to set the error log to /var/log/java/java_error.log:
1001 -XX:ErrorFile=/var/log/java/java_error.log
1003 · Windows: The following example shows how to set the error log
1005 file to C:/log/java/java_error.log:
1006 -XX:ErrorFile=C:/log/java/java_error.log
1008 If the file exists, and is writeable, then it will be overwrit‐
1010 ten. Otherwise, if the file can't be created in the specified
1011 directory (due to insufficient space, permission problem, or an‐
1012 other issue), then the file is created in the temporary directo‐
1013 ry for the operating system:
1014 · Oracle Solaris, Linux, and macOS: The temporary directory is
1016 /tmp.
1017 · Windows: The temporary directory is specified by the value of
1019 the TMP environment variable; if that environment variable
1020 isn't defined, then the value of the TEMP environment variable
1021 is used.
1022 -XX:+ExtensiveErrorReports
1024 Enables the reporting of more extensive error information in the
1025 ErrorFile. This option can be turned on in environments where
1026 maximal information is desired - even if the resulting logs may
1027 be quite large and/or contain information that might be consid‐
1028 ered sensitive. The information can vary from release to re‐
1029 lease, and across different platforms. By default this option
1030 is disabled.
1031 -XX:FlightRecorderOptions=parameter=value (or)-XX:FlightRecorderOp‐
1033 tions:parameter=value
1034 Sets the parameters that control the behavior of JFR.
1035 The following list contains the available JFR parameter=value
1037 entries:
1038 globalbuffersize=size
1040 Specifies the total amount of primary memory used for da‐
1041 ta retention. The default value is based on the value
1042 specified for memorysize. Change the memorysize parame‐
1043 ter to alter the size of global buffers.
1044 maxchunksize=size
1046 Specifies the maximum size (in bytes) of the data chunks
1047 in a recording. Append m or M to specify the size in
1048 megabytes (MB), or g or G to specify the size in giga‐
1049 bytes (GB). By default, the maximum size of data chunks
1050 is set to 12 MB. The minimum allowed is 1 MB.
1051 memorysize=size
1053 Determines how much buffer memory should be used, and
1054 sets the globalbuffersize and numglobalbuffers parameters
1055 based on the size specified. Append m or M to specify
1056 the size in megabytes (MB), or g or G to specify the size
1057 in gigabytes (GB). By default, the memory size is set to
1058 10 MB.
1059 numglobalbuffers
1061 Specifies the number of global buffers used. The default
1062 value is based on the memory size specified. Change the
1063 memorysize parameter to alter the number of global buf‐
1064 fers.
1065 old-object-queue-size=number-of-objects
1067 Maximum number of old objects to track. By default, the
1068 number of objects is set to 256.
1069 repository=path
1071 Specifies the repository (a directory) for temporary disk
1072 storage. By default, the system's temporary directory is
1073 used.
1074 retransform={true|false}
1076 Specifies whether event classes should be retransformed
1077 using JVMTI. If false, instrumentation is added when
1078 event classes are loaded. By default, this parameter is
1079 enabled.
1080 samplethreads={true|false}
1082 Specifies whether thread sampling is enabled. Thread
1083 sampling occurs only if the sampling event is enabled
1084 along with this parameter. By default, this parameter is
1085 enabled.
1086 stackdepth=depth
1088 Stack depth for stack traces. By default, the depth is
1089 set to 64 method calls. The maximum is 2048. Values
1090 greater than 64 could create significant overhead and re‐
1091 duce performance.
1092 threadbuffersize=size
1094 Specifies the per-thread local buffer size (in bytes).
1095 By default, the local buffer size is set to 8 kilobytes,
1096 with a minimum value of 4 kilobytes. Overriding this pa‐
1097 rameter could reduce performance and is not recommended.
1098 You can specify values for multiple parameters by separating
1100 them with a comma.
1101 -XX:LargePageSizeInBytes=size
1103 Sets the maximum size (in bytes) for large pages used for the
1104 Java heap. The size argument must be a power of 2 (2, 4, 8, 16,
1105 and so on). Append the letter k or K to indicate kilobytes, m
1106 or M to indicate megabytes, or g or G to indicate gigabytes. By
1107 default, the size is set to 0, meaning that the JVM chooses the
1108 size for large pages automatically. See Large Pages.
1109 The following example describes how to set the large page size
1111 to 4 megabytes (MB):
1112 -XX:LargePageSizeInBytes=4m
1114 -XX:MaxDirectMemorySize=size
1116 Sets the maximum total size (in bytes) of the java.nio package,
1117 direct-buffer allocations. Append the letter k or K to indicate
1118 kilobytes, m or M to indicate megabytes, or g or G to indicate
1119 gigabytes. By default, the size is set to 0, meaning that the
1120 JVM chooses the size for NIO direct-buffer allocations automati‐
1121 cally.
1122 The following examples illustrate how to set the NIO size to
1124 1024 KB in different units:
1125 -XX:MaxDirectMemorySize=1m
1127 -XX:MaxDirectMemorySize=1024k
1128 -XX:MaxDirectMemorySize=1048576
1129 -XX:-MaxFDLimit
1131 Disables the attempt to set the soft limit for the number of
1132 open file descriptors to the hard limit. By default, this op‐
1133 tion is enabled on all platforms, but is ignored on Windows.
1134 The only time that you may need to disable this is on Mac OS,
1135 where its use imposes a maximum of 10240, which is lower than
1136 the actual system maximum.
1137 -XX:NativeMemoryTracking=mode
1139 Specifies the mode for tracking JVM native memory usage. Possi‐
1140 ble mode arguments for this option include the following:
1141 off Instructs not to track JVM native memory usage. This is
1143 the default behavior if you don't specify the -XX:Native‐
1144 MemoryTracking option.
1145 summary
1147 Tracks memory usage only by JVM subsystems, such as Java
1148 heap, class, code, and thread.
1149 detail In addition to tracking memory usage by JVM subsystems,
1151 track memory usage by individual CallSite, individual
1152 virtual memory region and its committed regions.
1153 -XX:ObjectAlignmentInBytes=alignment
1155 Sets the memory alignment of Java objects (in bytes). By de‐
1156 fault, the value is set to 8 bytes. The specified value should
1157 be a power of 2, and must be within the range of 8 and 256 (in‐
1158 clusive). This option makes it possible to use compressed
1159 pointers with large Java heap sizes.
1160 The heap size limit in bytes is calculated as:
1162 4GB * ObjectAlignmentInBytes
1164 Note: As the alignment value increases, the unused space
1166 between objects also increases. As a result, you may not
1167 realize any benefits from using compressed pointers with
1168 large Java heap sizes.
1169 -XX:OnError=string
1171 Sets a custom command or a series of semicolon-separated com‐
1172 mands to run when an irrecoverable error occurs. If the string
1173 contains spaces, then it must be enclosed in quotation marks.
1174 · Oracle Solaris, Linux, and macOS: The following example shows
1176 how the -XX:OnError option can be used to run the gcore com‐
1177 mand to create a core image, and start the gdb debugger to at‐
1178 tach to the process in case of an irrecoverable error (the %p
1179 designates the current process identifier):
1180 -XX:OnError="gcore %p;gdb -p %p"
1182 · Windows: The following example shows how the -XX:OnError op‐
1184 tion can be used to run the userdump.exe utility to obtain a
1185 crash dump in case of an irrecoverable error (the %p desig‐
1186 nates the current process identifier). This example assumes
1187 that the path to the userdump.exe utility is specified in the
1188 PATH environment variable:
1189 -XX:OnError="userdump.exe %p"
1191 -XX:OnOutOfMemoryError=string
1193 Sets a custom command or a series of semicolon-separated com‐
1194 mands to run when an OutOfMemoryError exception is first thrown.
1195 If the string contains spaces, then it must be enclosed in quo‐
1196 tation marks. For an example of a command string, see the de‐
1197 scription of the -XX:OnError option.
1198 -XX:+PrintCommandLineFlags
1200 Enables printing of ergonomically selected JVM flags that ap‐
1201 peared on the command line. It can be useful to know the er‐
1202 gonomic values set by the JVM, such as the heap space size and
1203 the selected garbage collector. By default, this option is dis‐
1204 abled and flags aren't printed.
1205 -XX:+PreserveFramePointer
1207 Selects between using the RBP register as a general purpose reg‐
1208 ister (-XX:-PreserveFramePointer) and using the RBP register to
1209 hold the frame pointer of the currently executing method
1210 (-XX:+PreserveFramePointer . If the frame pointer is available,
1211 then external profiling tools (for example, Linux perf) can con‐
1212 struct more accurate stack traces.
1213 -XX:+PrintNMTStatistics
1215 Enables printing of collected native memory tracking data at JVM
1216 exit when native memory tracking is enabled (see -XX:NativeMemo‐
1217 ryTracking). By default, this option is disabled and native
1218 memory tracking data isn't printed.
1219 -XX:SharedArchiveFile=path
1221 Specifies the path and name of the class data sharing (CDS) ar‐
1222 chive file
1223 See Application Class Data Sharing.
1225 -XX:SharedArchiveConfigFile=shared_config_file
1227 Specifies additional shared data added to the archive file.
1228 -XX:SharedClassListFile=file_name
1230 Specifies the text file that contains the names of the classes
1231 to store in the class data sharing (CDS) archive. This file
1232 contains the full name of one class per line, except slashes (/)
1233 replace dots (.). For example, to specify the classes ja‐
1234 va.lang.Object and hello.Main, create a text file that contains
1235 the following two lines:
1236 java/lang/Object
1238 hello/Main
1239 The classes that you specify in this text file should include
1241 the classes that are commonly used by the application. They may
1242 include any classes from the application, extension, or boot‐
1243 strap class paths.
1244 See Application Class Data Sharing.
1246 -XX:+ShowCodeDetailsInExceptionMessages
1248 Enables printing of improved NullPointerException messages.
1249 When an application throws a NullPointerException, the option
1250 enables the JVM to analyze the program's bytecode instructions
1251 to determine precisely which reference is null, and describes
1252 the source with a null-detail message. The null-detail message
1253 is calculated and returned by NullPointerException.getMessage(),
1254 and will be printed as the exception message along with the
1255 method, filename, and line number. By default, this option is
1256 disabled.
1257 -XX:+ShowMessageBoxOnError
1259 Enables the display of a dialog box when the JVM experiences an
1260 irrecoverable error. This prevents the JVM from exiting and
1261 keeps the process active so that you can attach a debugger to it
1262 to investigate the cause of the error. By default, this option
1263 is disabled.
1264 -XX:StartFlightRecording=parameter=value
1266 Starts a JFR recording for the Java application. This option is
1267 equivalent to the JFR.start diagnostic command that starts a
1268 recording during runtime. You can set the following parame‐
1269 ter=value entries when starting a JFR recording:
1270 delay=time
1272 Specifies the delay between the Java application launch
1273 time and the start of the recording. Append s to specify
1274 the time in seconds, m for minutes, h for hours, or d for
1275 days (for example, specifying 10m means 10 minutes). By
1276 default, there's no delay, and this parameter is set to
1277 0.
1278 disk={true|false}
1280 Specifies whether to write data to disk while recording.
1281 By default, this parameter is enabled.
1282 dumponexit={true|false}
1284 Specifies if the running recording is dumped when the JVM
1285 shuts down. If enabled and a filename is not entered,
1286 the recording is written to a file in the directory where
1287 the process was started. The file name is a system-gen‐
1288 erated name that contains the process ID, recording ID,
1289 and current timestamp, similar to
1290 hotspot-pid-47496-id-1-2018_01_25_19_10_41.jfr. By de‐
1291 fault, this parameter is disabled.
1292 duration=time
1294 Specifies the duration of the recording. Append s to
1295 specify the time in seconds, m for minutes, h for hours,
1296 or d for days (for example, specifying 5h means 5 hours).
1297 By default, the duration isn't limited, and this parame‐
1298 ter is set to 0.
1299 filename=path
1301 Specifies the path and name of the file to which the
1302 recording is written when the recording is stopped, for
1303 example:
1304 · recording.jfr
1306 · /home/user/recordings/recording.jfr
1308 · c:\recordings\recording.jfr
1310 name=identifier
1312 Takes both the name and the identifier of a recording.
1313 maxage=time
1315 Specifies the maximum age of disk data to keep for the
1316 recording. This parameter is valid only when the disk
1317 parameter is set to true. Append s to specify the time
1318 in seconds, m for minutes, h for hours, or d for days
1319 (for example, specifying 30s means 30 seconds). By de‐
1320 fault, the maximum age isn't limited, and this parameter
1321 is set to 0s.
1322 maxsize=size
1324 Specifies the maximum size (in bytes) of disk data to
1325 keep for the recording. This parameter is valid only
1326 when the disk parameter is set to true. The value must
1327 not be less than the value for the maxchunksize parameter
1328 set with -XX:FlightRecorderOptions. Append m or M to
1329 specify the size in megabytes, or g or G to specify the
1330 size in gigabytes. By default, the maximum size of disk
1331 data isn't limited, and this parameter is set to 0.
1332 path-to-gc-roots={true|false}
1334 Specifies whether to collect the path to garbage collec‐
1335 tion (GC) roots at the end of a recording. By default,
1336 this parameter is disabled.
1337 The path to GC roots is useful for finding memory leaks,
1339 but collecting it is time-consuming. Enable this option
1340 only when you start a recording for an application that
1341 you suspect has a memory leak. If the settings parameter
1342 is set to profile, the stack trace from where the poten‐
1343 tial leaking object was allocated is included in the in‐
1344 formation collected.
1345 settings=path
1347 Specifies the path and name of the event settings file
1348 (of type JFC). By default, the default.jfc file is used,
1349 which is located in JRE_HOME/lib/jfr. This default set‐
1350 tings file collects a predefined set of information with
1351 low overhead, so it has minimal impact on performance and
1352 can be used with recordings that run continuously.
1353 A second settings file is also provided, profile.jfc,
1355 which provides more data than the default configuration,
1356 but can have more overhead and impact performance. Use
1357 this configuration for short periods of time when more
1358 information is needed.
1359 You can specify values for multiple parameters by separating
1361 them with a comma.
1362 -XX:ThreadStackSize=size
1364 Sets the Java thread stack size (in kilobytes). Use of a scal‐
1365 ing suffix, such as k, results in the scaling of the kilobytes
1366 value so that -XX:ThreadStackSize=1k sets the Java thread stack
1367 size to 1024*1024 bytes or 1 megabyte. The default value de‐
1368 pends on the platform:
1369 · Linux/x64 (64-bit): 1024 KB
1371 · macOS (64-bit): 1024 KB
1373 · Oracle Solaris (64-bit): 1024 KB
1375 · Windows: The default value depends on virtual memory
1377 The following examples show how to set the thread stack size to
1379 1 megabyte in different units:
1380 -XX:ThreadStackSize=1k
1382 -XX:ThreadStackSize=1024
1383 This option is similar to -Xss.
1385 -XX:-UseBiasedLocking
1387 Disables the use of biased locking. Some applications with sig‐
1388 nificant amounts of uncontended synchronization may attain sig‐
1389 nificant speedups with this flag enabled, but applications with
1390 certain patterns of locking may see slowdowns. .
1391 By default, this option is enabled.
1393 -XX:-UseCompressedOops
1395 Disables the use of compressed pointers. By default, this op‐
1396 tion is enabled, and compressed pointers are used. This will
1397 automatically limit the maximum ergonomically determined Java
1398 heap size to the maximum amount of memory that can be covered by
1399 compressed pointers. By default this range is 32 GB.
1400 With compressed oops enabled, object references are represented
1402 as 32-bit offsets instead of 64-bit pointers, which typically
1403 increases performance when running the application with Java
1404 heap sizes smaller than the compressed oops pointer range. This
1405 option works only for 64-bit JVMs.
1406 It's possible to use compressed pointers with Java heap sizes
1408 greater than 32 GB. See the -XX:ObjectAlignmentInBytes option.
1409 -XX:-UseContainerSupport
1411 The VM now provides automatic container detection support, which
1412 allows the VM to determine the amount of memory and number of
1413 processors that are available to a Java process running in dock‐
1414 er containers. It uses this information to allocate system re‐
1415 sources. This support is only available on Linux x64 platforms.
1416 If supported, the default for this flag is true, and container
1417 support is enabled by default. It can be disabled with
1418 -XX:-UseContainerSupport.
1419 Unified Logging is available to help to diagnose issues related
1421 to this support.
1422 Use -Xlog:os+container=trace for maximum logging of container
1424 information. See Enable Logging with the JVM Unified Logging
1425 Framework for a description of using Unified Logging.
1426 -XX:+UseHugeTLBFS
1428 Linux only: This option is the equivalent of specifying
1429 -XX:+UseLargePages. This option is disabled by default. This
1430 option pre-allocates all large pages up-front, when memory is
1431 reserved; consequently the JVM can't dynamically grow or shrink
1432 large pages memory areas; see -XX:UseTransparentHugePages if you
1433 want this behavior.
1434 See Large Pages.
1436 -XX:+UseLargePages
1438 Enables the use of large page memory. By default, this option
1439 is disabled and large page memory isn't used.
1440 See Large Pages.
1442 -XX:+UseTransparentHugePages
1444 Linux only: Enables the use of large pages that can dynamically
1445 grow or shrink. This option is disabled by default. You may
1446 encounter performance problems with transparent huge pages as
1447 the OS moves other pages around to create huge pages; this op‐
1448 tion is made available for experimentation.
1449 -XX:+AllowUserSignalHandlers
1451 Enables installation of signal handlers by the application. By
1452 default, this option is disabled and the application isn't al‐
1453 lowed to install signal handlers.
1454 -XX:VMOptionsFile=filename
1456 Allows user to specify VM options in a file, for example, ja‐
1457 va -XX:VMOptionsFile=/var/my_vm_options HelloWorld.
1458
1460 These java options control the dynamic just-in-time (JIT) compilation
1461 performed by the Java HotSpot VM.
1462 -XX:AllocateInstancePrefetchLines=lines
1464 Sets the number of lines to prefetch ahead of the instance allo‐
1465 cation pointer. By default, the number of lines to prefetch is
1466 set to 1:
1467 -XX:AllocateInstancePrefetchLines=1
1469 -XX:AllocatePrefetchDistance=size
1471 Sets the size (in bytes) of the prefetch distance for object al‐
1472 location. Memory about to be written with the value of new ob‐
1473 jects is prefetched up to this distance starting from the ad‐
1474 dress of the last allocated object. Each Java thread has its
1475 own allocation point.
1476 Negative values denote that prefetch distance is chosen based on
1478 the platform. Positive values are bytes to prefetch. Append
1479 the letter k or K to indicate kilobytes, m or M to indicate
1480 megabytes, or g or G to indicate gigabytes. The default value
1481 is set to -1.
1482 The following example shows how to set the prefetch distance to
1484 1024 bytes:
1485 -XX:AllocatePrefetchDistance=1024
1487 -XX:AllocatePrefetchInstr=instruction
1489 Sets the prefetch instruction to prefetch ahead of the alloca‐
1490 tion pointer. Possible values are from 0 to 3. The actual in‐
1491 structions behind the values depend on the platform. By de‐
1492 fault, the prefetch instruction is set to 0:
1493 -XX:AllocatePrefetchInstr=0
1495 -XX:AllocatePrefetchLines=lines
1497 Sets the number of cache lines to load after the last object al‐
1498 location by using the prefetch instructions generated in com‐
1499 piled code. The default value is 1 if the last allocated object
1500 was an instance, and 3 if it was an array.
1501 The following example shows how to set the number of loaded
1503 cache lines to 5:
1504 -XX:AllocatePrefetchLines=5
1506 -XX:AllocatePrefetchStepSize=size
1508 Sets the step size (in bytes) for sequential prefetch instruc‐
1509 tions. Append the letter k or K to indicate kilobytes, m or M
1510 to indicate megabytes, g or G to indicate gigabytes. By de‐
1511 fault, the step size is set to 16 bytes:
1512 -XX:AllocatePrefetchStepSize=16
1514 -XX:AllocatePrefetchStyle=style
1516 Sets the generated code style for prefetch instructions. The
1517 style argument is an integer from 0 to 3:
1518 0 Don't generate prefetch instructions.
1520 1 Execute prefetch instructions after each allocation.
1522 This is the default setting.
1523 2 Use the thread-local allocation block (TLAB) watermark
1525 pointer to determine when prefetch instructions are exe‐
1526 cuted.
1527 3 Generate one prefetch instruction per cache line.
1529 -XX:+BackgroundCompilation
1531 Enables background compilation. This option is enabled by de‐
1532 fault. To disable background compilation, specify -XX:-Back‐
1533 groundCompilation (this is equivalent to specifying -Xbatch).
1534 -XX:CICompilerCount=threads
1536 Sets the number of compiler threads to use for compilation. By
1537 default, the number of compiler threads is selected automatical‐
1538 ly depending on the number of CPUs and memory available for com‐
1539 piled code. The following example shows how to set the number
1540 of threads to 2:
1541 -XX:CICompilerCount=2
1543 -XX:+UseDynamicNumberOfCompilerThreads
1545 Dynamically create compiler thread up to the limit specified by
1546 -XX:CICompilerCount. This option is enabled by default.
1547 -XX:CompileCommand=command,method[,option]
1549 Specifies a command to perform on a method. For example, to ex‐
1550 clude the indexOf() method of the String class from being com‐
1551 piled, use the following:
1552 -XX:CompileCommand=exclude,java/lang/String.indexOf
1554 Note that the full class name is specified, including all pack‐
1556 ages and subpackages separated by a slash (/). For easier
1557 cut-and-paste operations, it's also possible to use the method
1558 name format produced by the -XX:+PrintCompilation and -XX:+Log‐
1559 Compilation options:
1560 -XX:CompileCommand=exclude,java.lang.String::indexOf
1562 If the method is specified without the signature, then the com‐
1564 mand is applied to all methods with the specified name. Howev‐
1565 er, you can also specify the signature of the method in the
1566 class file format. In this case, you should enclose the argu‐
1567 ments in quotation marks, because otherwise the shell treats the
1568 semicolon as a command end. For example, if you want to exclude
1569 only the indexOf(String) method of the String class from being
1570 compiled, use the following:
1571 -XX:CompileCommand="exclude,java/lang/String.index‐
1573 Of,(Ljava/lang/String;)I"
1574 You can also use the asterisk (*) as a wildcard for class and
1576 method names. For example, to exclude all indexOf() methods in
1577 all classes from being compiled, use the following:
1578 -XX:CompileCommand=exclude,*.indexOf
1580 The commas and periods are aliases for spaces, making it easier
1582 to pass compiler commands through a shell. You can pass argu‐
1583 ments to -XX:CompileCommand using spaces as separators by en‐
1584 closing the argument in quotation marks:
1585 -XX:CompileCommand="exclude java/lang/String indexOf"
1587 Note that after parsing the commands passed on the command line
1589 using the -XX:CompileCommand options, the JIT compiler then
1590 reads commands from the .hotspot_compiler file. You can add
1591 commands to this file or specify a different file using the
1592 -XX:CompileCommandFile option.
1593 To add several commands, either specify the -XX:CompileCommand
1595 option multiple times, or separate each argument with the new
1596 line separator (\n). The following commands are available:
1597 break Sets a breakpoint when debugging the JVM to stop at the
1599 beginning of compilation of the specified method.
1600 compileonly
1602 Excludes all methods from compilation except for the
1603 specified method. As an alternative, you can use the
1604 -XX:CompileOnly option, which lets you specify several
1605 methods.
1606 dontinline
1608 Prevents inlining of the specified method.
1609 exclude
1611 Excludes the specified method from compilation.
1612 help Prints a help message for the -XX:CompileCommand option.
1614 inline Attempts to inline the specified method.
1616 log Excludes compilation logging (with the -XX:+LogCompila‐
1618 tion option) for all methods except for the specified
1619 method. By default, logging is performed for all com‐
1620 piled methods.
1621 option Passes a JIT compilation option to the specified method
1623 in place of the last argument (option). The compilation
1624 option is set at the end, after the method name. For ex‐
1625 ample, to enable the BlockLayoutByFrequency option for
1626 the append() method of the StringBuffer class, use the
1627 following:
1628 -XX:CompileCommand=option,java/lang/String‐
1630 Buffer.append,BlockLayoutByFrequency
1631 You can specify multiple compilation options, separated
1633 by commas or spaces.
1634 print Prints generated assembler code after compilation of the
1636 specified method.
1637 quiet Instructs not to print the compile commands. By default,
1639 the commands that you specify with the -XX:CompileCommand
1640 option are printed; for example, if you exclude from com‐
1641 pilation the indexOf() method of the String class, then
1642 the following is printed to standard output:
1643 CompilerOracle: exclude java/lang/String.indexOf
1645 You can suppress this by specifying the -XX:CompileCom‐
1647 mand=quiet option before other -XX:CompileCommand op‐
1648 tions.
1649 -XX:CompileCommandFile=filename
1651 Sets the file from which JIT compiler commands are read. By de‐
1652 fault, the .hotspot_compiler file is used to store commands per‐
1653 formed by the JIT compiler.
1654 Each line in the command file represents a command, a class
1656 name, and a method name for which the command is used. For ex‐
1657 ample, this line prints assembly code for the toString() method
1658 of the String class:
1659 print java/lang/String toString
1661 If you're using commands for the JIT compiler to perform on
1663 methods, then see the -XX:CompileCommand option.
1664 -XX:CompilerDirectivesFile=file
1666 Adds directives from a file to the directives stack when a pro‐
1667 gram starts. See Compiler Control [https://docs.ora‐
1668 cle.com/en/java/javase/12/vm/compiler-con‐
1669 trol1.html#GUID-94AD8194-786A-4F19-BFFF-278F8E237F3A].
1670 The -XX:CompilerDirectivesFile option has to be used together
1672 with the -XX:UnlockDiagnosticVMOptions option that unlocks diag‐
1673 nostic JVM options.
1674 -XX:+CompilerDirectivesPrint
1676 Prints the directives stack when the program starts or when a
1677 new directive is added.
1678 The -XX:+CompilerDirectivesPrint option has to be used together
1680 with the -XX:UnlockDiagnosticVMOptions option that unlocks diag‐
1681 nostic JVM options.
1682 -XX:CompileOnly=methods
1684 Sets the list of methods (separated by commas) to which compila‐
1685 tion should be restricted. Only the specified methods are com‐
1686 piled. Specify each method with the full class name (including
1687 the packages and subpackages). For example, to compile only the
1688 length() method of the String class and the size() method of the
1689 List class, use the following:
1690 -XX:CompileOnly=java/lang/String.length,ja‐
1692 va/util/List.size
1693 Note that the full class name is specified, including all pack‐
1695 ages and subpackages separated by a slash (/). For easier cut
1696 and paste operations, it's also possible to use the method name
1697 format produced by the -XX:+PrintCompilation and -XX:+LogCompi‐
1698 lation options:
1699 -XX:CompileOnly=java.lang.String::length,ja‐
1701 va.util.List::size
1702 Although wildcards aren't supported, you can specify only the
1704 class or package name to compile all methods in that class or
1705 package, as well as specify just the method to compile methods
1706 with this name in any class:
1707 -XX:CompileOnly=java/lang/String
1709 -XX:CompileOnly=java/lang
1710 -XX:CompileOnly=.length
1711 -XX:CompileThresholdScaling=scale
1713 Provides unified control of first compilation. This option con‐
1714 trols when methods are first compiled for both the tiered and
1715 the nontiered modes of operation. The CompileThresholdScaling
1716 option has a floating point value between 0 and +Inf and scales
1717 the thresholds corresponding to the current mode of operation
1718 (both tiered and nontiered). Setting CompileThresholdScaling to
1719 a value less than 1.0 results in earlier compilation while val‐
1720 ues greater than 1.0 delay compilation. Setting CompileThresh‐
1721 oldScaling to 0 is equivalent to disabling compilation.
1722 -XX:+DoEscapeAnalysis
1724 Enables the use of escape analysis. This option is enabled by
1725 default. To disable the use of escape analysis, specify
1726 -XX:-DoEscapeAnalysis.
1727 -XX:InitialCodeCacheSize=size
1729 Sets the initial code cache size (in bytes). Append the letter
1730 k or K to indicate kilobytes, m or M to indicate megabytes, or g
1731 or G to indicate gigabytes. The default value depends on the
1732 platform. The initial code cache size shouldn't be less than
1733 the system's minimal memory page size. The following example
1734 shows how to set the initial code cache size to 32 KB:
1735 -XX:InitialCodeCacheSize=32k
1737 -XX:+Inline
1739 Enables method inlining. This option is enabled by default to
1740 increase performance. To disable method inlining, specify
1741 -XX:-Inline.
1742 -XX:InlineSmallCode=size
1744 Sets the maximum code size (in bytes) for already compiled meth‐
1745 ods that may be inlined. Append the letter k or K to indicate
1746 kilobytes, m or M to indicate megabytes, or g or G to indicate
1747 gigabytes. The default value depends on the platform and on
1748 whether tiered compilation is enabled. In the following example
1749 it is set to 1000 bytes:
1750 -XX:InlineSmallCode=1000
1752 -XX:+LogCompilation
1754 Enables logging of compilation activity to a file named
1755 hotspot.log in the current working directory. You can specify a
1756 different log file path and name using the -XX:LogFile option.
1757 By default, this option is disabled and compilation activity
1759 isn't logged. The -XX:+LogCompilation option has to be used to‐
1760 gether with the -XX:UnlockDiagnosticVMOptions option that un‐
1761 locks diagnostic JVM options.
1762 You can enable verbose diagnostic output with a message printed
1764 to the console every time a method is compiled by using the
1765 -XX:+PrintCompilation option.
1766 -XX:FreqInlineSize=size
1768 Sets the maximum bytecode size (in bytes) of a hot method to be
1769 inlined. Append the letter k or K to indicate kilobytes, m or M
1770 to indicate megabytes, or g or G to indicate gigabytes. The de‐
1771 fault value depends on the platform. In the following example
1772 it is set to 325 bytes:
1773 -XX:FreqInlineSize=325
1775 -XX:MaxInlineSize=size
1777 Sets the maximum bytecode size (in bytes) of a cold method to be
1778 inlined. Append the letter k or K to indicate kilobytes, m or M
1779 to indicate megabytes, or g or G to indicate gigabytes. By de‐
1780 fault, the maximum bytecode size is set to 35 bytes:
1781 -XX:MaxInlineSize=35
1783 -XX:MaxTrivialSize=size
1785 Sets the maximum bytecode size (in bytes) of a trivial method to
1786 be inlined. Append the letter k or K to indicate kilobytes, m
1787 or M to indicate megabytes, or g or G to indicate gigabytes. By
1788 default, the maximum bytecode size of a trivial method is set to
1789 6 bytes:
1790 -XX:MaxTrivialSize=6
1792 -XX:MaxNodeLimit=nodes
1794 Sets the maximum number of nodes to be used during single method
1795 compilation. By default the value depends on the features en‐
1796 abled. In the following example the maximum number of nodes is
1797 set to 100,000:
1798 -XX:MaxNodeLimit=100000
1800 -XX:NonNMethodCodeHeapSize=size
1802 Sets the size in bytes of the code segment containing nonmethod
1803 code.
1804 A nonmethod code segment containing nonmethod code, such as com‐
1806 piler buffers and the bytecode interpreter. This code type
1807 stays in the code cache forever. This flag is used only if
1808 -XX:SegmentedCodeCache is enabled.
1809 -XX:NonProfiledCodeHeapSize=size
1811 Sets the size in bytes of the code segment containing nonpro‐
1812 filed methods. This flag is used only if -XX:SegmentedCodeCache
1813 is enabled.
1814 -XX:+OptimizeStringConcat
1816 Enables the optimization of String concatenation operations.
1817 This option is enabled by default. To disable the optimization
1818 of String concatenation operations, specify -XX:-OptimizeString‐
1819 Concat.
1820 -XX:+PrintAssembly
1822 Enables printing of assembly code for bytecoded and native meth‐
1823 ods by using the external hsdis-<arch>.so or .dll library. For
1824 64-bit VM on Windows, it's hsdis-amd64.dll. This lets you to
1825 see the generated code, which may help you to diagnose perfor‐
1826 mance issues.
1827 By default, this option is disabled and assembly code isn't
1829 printed. The -XX:+PrintAssembly option has to be used together
1830 with the -XX:UnlockDiagnosticVMOptions option that unlocks diag‐
1831 nostic JVM options.
1832 -XX:ProfiledCodeHeapSize=size
1834 Sets the size in bytes of the code segment containing profiled
1835 methods. This flag is used only if -XX:SegmentedCodeCache is
1836 enabled.
1837 -XX:+PrintCompilation
1839 Enables verbose diagnostic output from the JVM by printing a
1840 message to the console every time a method is compiled. This
1841 lets you to see which methods actually get compiled. By de‐
1842 fault, this option is disabled and diagnostic output isn't
1843 printed.
1844 You can also log compilation activity to a file by using the
1846 -XX:+LogCompilation option.
1847 -XX:+PrintInlining
1849 Enables printing of inlining decisions. This let's you see
1850 which methods are getting inlined.
1851 By default, this option is disabled and inlining information
1853 isn't printed. The -XX:+PrintInlining option has to be used to‐
1854 gether with the -XX:+UnlockDiagnosticVMOptions option that un‐
1855 locks diagnostic JVM options.
1856 -XX:ReservedCodeCacheSize=size
1858 Sets the maximum code cache size (in bytes) for JIT-compiled
1859 code. Append the letter k or K to indicate kilobytes, m or M to
1860 indicate megabytes, or g or G to indicate gigabytes. The de‐
1861 fault maximum code cache size is 240 MB; if you disable tiered
1862 compilation with the option -XX:-TieredCompilation, then the de‐
1863 fault size is 48 MB. This option has a limit of 2 GB; other‐
1864 wise, an error is generated. The maximum code cache size
1865 shouldn't be less than the initial code cache size; see the op‐
1866 tion -XX:InitialCodeCacheSize.
1867 -XX:RTMAbortRatio=abort_ratio
1869 Specifies the RTM abort ratio is specified as a percentage (%)
1870 of all executed RTM transactions. If a number of aborted trans‐
1871 actions becomes greater than this ratio, then the compiled code
1872 is deoptimized. This ratio is used when the -XX:+UseRTMDeopt
1873 option is enabled. The default value of this option is 50.
1874 This means that the compiled code is deoptimized if 50% of all
1875 transactions are aborted.
1876 -XX:RTMRetryCount=number_of_retries
1878 Specifies the number of times that the RTM locking code is re‐
1879 tried, when it is aborted or busy, before falling back to the
1880 normal locking mechanism. The default value for this option is
1881 5. The -XX:UseRTMLocking option must be enabled.
1882 -XX:+SegmentedCodeCache
1884 Enables segmentation of the code cache. Without the -XX:+Seg‐
1885 mentedCodeCache, the code cache consists of one large segment.
1886 With -XX:+SegmentedCodeCache, we have separate segments for non‐
1887 method, profiled method, and nonprofiled method code. These
1888 segments aren't resized at runtime. The feature is enabled by
1889 default if tiered compilation is enabled (-XX:+TieredCompilation
1890 ) and -XX:ReservedCodeCacheSize >= 240 MB. The advantages are
1891 better control of the memory footprint, reduced code fragmenta‐
1892 tion, and better iTLB/iCache behavior due to improved locality.
1893 iTLB/iCache is a CPU-specific term meaning Instruction Transla‐
1894 tion Lookaside Buffer (ITLB). ICache is an instruction cache in
1895 theCPU. The implementation of the code cache can be found in
1896 the file: /share/vm/code/codeCache.cpp.
1897 -XX:StartAggressiveSweepingAt=percent
1899 Forces stack scanning of active methods to aggressively remove
1900 unused code when only the given percentage of the code cache is
1901 free. The default value is 10%.
1902 -XX:-TieredCompilation
1904 Disables the use of tiered compilation. By default, this option
1905 is enabled.
1906 -XX:UseSSE=version
1908 Enables the use of SSE instruction set of a specified version.
1909 Is set by default to the highest supported version available
1910 (x86 only).
1911 -XX:UseAVX=version
1913 Enables the use of AVX instruction set of a specified version.
1914 Is set by default to the highest supported version available
1915 (x86 only).
1916 -XX:+UseAES
1918 Enables hardware-based AES intrinsics for hardware that supports
1919 it. This option is on by default on hardware that has the nec‐
1920 essary instructions. The -XX:+UseAES is used in conjunction
1921 with UseAESIntrinsics. Flags that control intrinsics now re‐
1922 quire the option -XX:+UnlockDiagnosticVMOptions.
1923 -XX:+UseAESIntrinsics
1925 Enables AES intrinsics. Specifying-XX:+UseAESIntrinsics is
1926 equivalent to also enabling -XX:+UseAES. To disable hard‐
1927 ware-based AES intrinsics, specify -XX:-UseAES -XX:-UseAESIn‐
1928 trinsics. For example, to enable hardware AES, use the follow‐
1929 ing flags:
1930 -XX:+UseAES -XX:+UseAESIntrinsics
1932 Flags that control intrinsics now require the option -XX:+Un‐
1934 lockDiagnosticVMOptions.
1935 -XX:+UseAESCTRIntrinsics
1937 Analogous to -XX:+UseAESIntrinsics enables AES/CTR intrinsics.
1938 -XX:+UseGHASHIntrinsics
1940 Controls the use of GHASH intrinsics. Enabled by default on
1941 platforms that support the corresponding instructions. Flags
1942 that control intrinsics now require the option -XX:+UnlockDiag‐
1943 nosticVMOptions.
1944 -XX:+UseBASE64Intrinsics
1946 Controls the use of accelerated BASE64 encoding routines for ja‐
1947 va.util.Base64. Enabled by default on platforms that support
1948 it. Flags that control intrinsics now require the option
1949 -XX:+UnlockDiagnosticVMOptions.
1950 -XX:+UseAdler32Intrinsics
1952 Controls the use of Adler32 checksum algorithm intrinsic for ja‐
1953 va.util.zip.Adler32. Enabled by default on platforms that sup‐
1954 port it. Flags that control intrinsics now require the option
1955 -XX:+UnlockDiagnosticVMOptions.
1956 -XX:+UseCRC32Intrinsics
1958 Controls the use of CRC32 intrinsics for java.util.zip.CRC32.
1959 Enabled by default on platforms that support it. Flags that
1960 control intrinsics now require the option -XX:+UnlockDiagnos‐
1961 ticVMOptions.
1962 -XX:+UseCRC32CIntrinsics
1964 Controls the use of CRC32C intrinsics for java.util.zip.CRC32C.
1965 Enabled by default on platforms that support it. Flags that
1966 control intrinsics now require the option -XX:+UnlockDiagnos‐
1967 ticVMOptions.
1968 -XX:+UseSHA
1970 Enables hardware-based intrinsics for SHA crypto hash functions
1971 for SPARC hardware. The UseSHA option is used in conjunction
1972 with the UseSHA1Intrinsics, UseSHA256Intrinsics, and Use‐
1973 SHA512Intrinsics options.
1974 The UseSHA and UseSHA*Intrinsics flags are enabled by default on
1976 machines that support the corresponding instructions.
1977 This feature is applicable only when using the sun.securi‐
1979 ty.provider.Sun provider for SHA operations. Flags that control
1980 intrinsics now require the option -XX:+UnlockDiagnosticVMOp‐
1981 tions.
1982 To disable all hardware-based SHA intrinsics, specify the
1984 -XX:-UseSHA. To disable only a particular SHA intrinsic, use
1985 the appropriate corresponding option. For example: -XX:-Use‐
1986 SHA256Intrinsics.
1987 -XX:+UseSHA1Intrinsics
1989 Enables intrinsics for SHA-1 crypto hash function. Flags that
1990 control intrinsics now require the option -XX:+UnlockDiagnos‐
1991 ticVMOptions.
1992 -XX:+UseSHA256Intrinsics
1994 Enables intrinsics for SHA-224 and SHA-256 crypto hash func‐
1995 tions. Flags that control intrinsics now require the option
1996 -XX:+UnlockDiagnosticVMOptions.
1997 -XX:+UseSHA512Intrinsics
1999 Enables intrinsics for SHA-384 and SHA-512 crypto hash func‐
2000 tions. Flags that control intrinsics now require the option
2001 -XX:+UnlockDiagnosticVMOptions.
2002 -XX:+UseMathExactIntrinsics
2004 Enables intrinsification of various java.lang.Math.*Exact()
2005 functions. Enabled by default. Flags that control intrinsics
2006 now require the option -XX:+UnlockDiagnosticVMOptions.
2007 -XX:+UseMultiplyToLenIntrinsic
2009 Enables intrinsification of BigInteger.multiplyToLen(). Enabled
2010 by default on platforms that support it. Flags that control in‐
2011 trinsics now require the option -XX:+UnlockDiagnosticVMOptions.
2012 -XX:+UseSquareToLenIntrinsic
2014 Enables intrinsification of BigInteger.squareToLen(). Enabled
2015 by default on platforms that support it. Flags that control in‐
2016 trinsics now require the option -XX:+UnlockDiagnosticVMOptions.
2017 -XX:+UseMulAddIntrinsic
2019 Enables intrinsification of BigInteger.mulAdd(). Enabled by de‐
2020 fault on platforms that support it. Flags that control intrin‐
2021 sics now require the option -XX:+UnlockDiagnosticVMOptions.
2022 -XX:+UseMontgomeryMultiplyIntrinsic
2024 Enables intrinsification of BigInteger.montgomeryMultiply().
2025 Enabled by default on platforms that support it. Flags that
2026 control intrinsics now require the option -XX:+UnlockDiagnos‐
2027 ticVMOptions.
2028 -XX:+UseMontgomerySquareIntrinsic
2030 Enables intrinsification of BigInteger.montgomerySquare(). En‐
2031 abled by default on platforms that support it. Flags that con‐
2032 trol intrinsics now require the option -XX:+UnlockDiagnosticV‐
2033 MOptions.
2034 -XX:+UseCMoveUnconditionally
2036 Generates CMove (scalar and vector) instructions regardless of
2037 profitability analysis.
2038 -XX:+UseCodeCacheFlushing
2040 Enables flushing of the code cache before shutting down the com‐
2041 piler. This option is enabled by default. To disable flushing
2042 of the code cache before shutting down the compiler, specify
2043 -XX:-UseCodeCacheFlushing.
2044 -XX:+UseCondCardMark
2046 Enables checking if the card is already marked before updating
2047 the card table. This option is disabled by default. It should
2048 be used only on machines with multiple sockets, where it in‐
2049 creases the performance of Java applications that rely on con‐
2050 current operations.
2051 -XX:+UseCountedLoopSafepoints
2053 Keeps safepoints in counted loops. Its default value depends on
2054 whether the selected garbage collector requires low latency
2055 safepoints.
2056 -XX:LoopStripMiningIter=number_of_iterations
2058 Controls the number of iterations in the inner strip mined loop.
2059 Strip mining transforms counted loops into two level nested
2060 loops. Safepoints are kept in the outer loop while the inner
2061 loop can execute at full speed. This option controls the maxi‐
2062 mum number of iterations in the inner loop. The default value
2063 is 1,000.
2064 -XX:LoopStripMiningIterShortLoop=number_of_iterations
2066 Controls loop strip mining optimization. Loops with the number
2067 of iterations less than specified will not have safepoints in
2068 them. Default value is 1/10th of -XX:LoopStripMiningIter.
2069 -XX:+UseFMA
2071 Enables hardware-based FMA intrinsics for hardware where FMA in‐
2072 structions are available (such as, Intel, SPARC, and ARM64).
2073 FMA intrinsics are generated for the java.lang.Math.fma(a, b, c)
2074 methods that calculate the value of ( a * b + c ) expressions.
2075 -XX:+UseRTMDeopt
2077 Autotunes RTM locking depending on the abort ratio. This ratio
2078 is specified by the -XX:RTMAbortRatio option. If the number of
2079 aborted transactions exceeds the abort ratio, then the method
2080 containing the lock is deoptimized and recompiled with all locks
2081 as normal locks. This option is disabled by default. The
2082 -XX:+UseRTMLocking option must be enabled.
2083 -XX:+UseRTMLocking
2085 Generates Restricted Transactional Memory (RTM) locking code for
2086 all inflated locks, with the normal locking mechanism as the
2087 fallback handler. This option is disabled by default. Options
2088 related to RTM are available only on x86 CPUs that support
2089 Transactional Synchronization Extensions (TSX).
2090 RTM is part of Intel's TSX, which is an x86 instruction set ex‐
2092 tension and facilitates the creation of multithreaded applica‐
2093 tions. RTM introduces the new instructions XBEGIN, XABORT,
2094 XEND, and XTEST. The XBEGIN and XEND instructions enclose a set
2095 of instructions to run as a transaction. If no conflict is
2096 found when running the transaction, then the memory and register
2097 modifications are committed together at the XEND instruction.
2098 The XABORT instruction can be used to explicitly abort a trans‐
2099 action and the XEND instruction checks if a set of instructions
2100 is being run in a transaction.
2101 A lock on a transaction is inflated when another thread tries to
2103 access the same transaction, thereby blocking the thread that
2104 didn't originally request access to the transaction. RTM re‐
2105 quires that a fallback set of operations be specified in case a
2106 transaction aborts or fails. An RTM lock is a lock that has
2107 been delegated to the TSX's system.
2108 RTM improves performance for highly contended locks with low
2110 conflict in a critical region (which is code that must not be
2111 accessed by more than one thread concurrently). RTM also im‐
2112 proves the performance of coarse-grain locking, which typically
2113 doesn't perform well in multithreaded applications.
2114 (Coarse-grain locking is the strategy of holding locks for long
2115 periods to minimize the overhead of taking and releasing locks,
2116 while fine-grained locking is the strategy of trying to achieve
2117 maximum parallelism by locking only when necessary and unlocking
2118 as soon as possible.) Also, for lightly contended locks that are
2119 used by different threads, RTM can reduce false cache line shar‐
2120 ing, also known as cache line ping-pong. This occurs when mul‐
2121 tiple threads from different processors are accessing different
2122 resources, but the resources share the same cache line. As a
2123 result, the processors repeatedly invalidate the cache lines of
2124 other processors, which forces them to read from main memory in‐
2125 stead of their cache.
2126 -XX:+UseSuperWord
2128 Enables the transformation of scalar operations into superword
2129 operations. Superword is a vectorization optimization. This
2130 option is enabled by default. To disable the transformation of
2131 scalar operations into superword operations, specify -XX:-UseSu‐
2132 perWord.
2133
2135 These java options provide the ability to gather system information and
2136 perform extensive debugging.
2137 -XX:+DisableAttachMechanism
2139 Disables the mechanism that lets tools attach to the JVM. By
2140 default, this option is disabled, meaning that the attach mecha‐
2141 nism is enabled and you can use diagnostics and troubleshooting
2142 tools such as jcmd, jstack, jmap, and jinfo.
2143 Note: The tools such as jcmd, jinfo, jmap, and jstack
2145 shipped with the JDK aren't supported when using the
2146 tools from one JDK version to troubleshoot a different
2147 JDK version.
2148 -XX:+ExtendedDTraceProbes
2150 Oracle Solaris, Linux, and macOS: Enables additional dtrace tool
2151 probes that affect the performance. By default, this option is
2152 disabled and dtrace performs only standard probes.
2153 -XX:+HeapDumpOnOutOfMemoryError
2155 Enables the dumping of the Java heap to a file in the current
2156 directory by using the heap profiler (HPROF) when a ja‐
2157 va.lang.OutOfMemoryError exception is thrown. You can explicit‐
2158 ly set the heap dump file path and name using the -XX:HeapDump‐
2159 Path option. By default, this option is disabled and the heap
2160 isn't dumped when an OutOfMemoryError exception is thrown.
2161 -XX:HeapDumpPath=path
2163 Sets the path and file name for writing the heap dump provided
2164 by the heap profiler (HPROF) when the -XX:+HeapDumpOnOutOfMemo‐
2165 ryError option is set. By default, the file is created in the
2166 current working directory, and it's named java_pid<pid>.hprof
2167 where <pid> is the identifier of the process that caused the er‐
2168 ror. The following example shows how to set the default file
2169 explicitly (%p represents the current process identifier):
2170 -XX:HeapDumpPath=./java_pid%p.hprof
2172 · Oracle Solaris, Linux, and macOS: The following example shows
2174 how to set the heap dump file to /var/log/java/java_heap‐
2175 dump.hprof:
2176 -XX:HeapDumpPath=/var/log/java/java_heapdump.hprof
2178 · Windows: The following example shows how to set the heap dump
2180 file to C:/log/java/java_heapdump.log:
2181 -XX:HeapDumpPath=C:/log/java/java_heapdump.log
2183 -XX:LogFile=path
2185 Sets the path and file name to where log data is written. By
2186 default, the file is created in the current working directory,
2187 and it's named hotspot.log.
2188 · Oracle Solaris, Linux, and macOS: The following example shows
2190 how to set the log file to /var/log/java/hotspot.log:
2191 -XX:LogFile=/var/log/java/hotspot.log
2193 · Windows: The following example shows how to set the log file
2195 to C:/log/java/hotspot.log:
2196 -XX:LogFile=C:/log/java/hotspot.log
2198 -XX:+PrintClassHistogram
2200 Enables printing of a class instance histogram after one of the
2201 following events:
2202 · Oracle Solaris, Linux, and macOS: Control+Break
2204 · Windows: Control+C (SIGTERM)
2206 By default, this option is disabled.
2208 Setting this option is equivalent to running the jmap -histo
2210 command, or the jcmd pid GC.class_histogram command, where pid
2211 is the current Java process identifier.
2212 -XX:+PrintConcurrentLocks
2214 Enables printing of java.util.concurrent locks after one of the
2215 following events:
2216 · Oracle Solaris, Linux, and macOS: Control+Break
2218 · Windows: Control+C (SIGTERM)
2220 By default, this option is disabled.
2222 Setting this option is equivalent to running the jstack -l com‐
2224 mand or the jcmd pid Thread.print -l command, where pid is the
2225 current Java process identifier.
2226 -XX:+PrintFlagsRanges
2228 Prints the range specified and allows automatic testing of the
2229 values. See Validate Java Virtual Machine Flag Arguments.
2230 -XX:+PerfDataSaveToFile
2232 If enabled, saves jstat binary data when the Java application
2233 exits. This binary data is saved in a file named hsperfda‐
2234 ta_pid, where pid is the process identifier of the Java applica‐
2235 tion that you ran. Use the jstat command to display the perfor‐
2236 mance data contained in this file as follows:
2237 jstat -class file:///path/hsperfdata_pid
2239 jstat -gc file:///path/hsperfdata_pid
2241 -XX:+UsePerfData
2243 Enables the perfdata feature. This option is enabled by default
2244 to allow JVM monitoring and performance testing. Disabling it
2245 suppresses the creation of the hsperfdata_userid directories.
2246 To disable the perfdata feature, specify -XX:-UsePerfData.
2247
2249 These java options control how garbage collection (GC) is performed by
2250 the Java HotSpot VM.
2251 -XX:+AggressiveHeap
2253 Enables Java heap optimization. This sets various parameters to
2254 be optimal for long-running jobs with intensive memory alloca‐
2255 tion, based on the configuration of the computer (RAM and CPU).
2256 By default, the option is disabled and the heap sizes are con‐
2257 figured less aggressively.
2258 -XX:+AlwaysPreTouch
2260 Requests the VM to touch every page on the Java heap after re‐
2261 questing it from the operating system and before handing memory
2262 out to the application. By default, this option is disabled and
2263 all pages are committed as the application uses the heap space.
2264 -XX:ConcGCThreads=threads
2266 Sets the number of threads used for concurrent GC. Sets threads
2267 to approximately 1/4 of the number of parallel garbage collec‐
2268 tion threads. The default value depends on the number of CPUs
2269 available to the JVM.
2270 For example, to set the number of threads for concurrent GC to
2272 2, specify the following option:
2273 -XX:ConcGCThreads=2
2275 -XX:+DisableExplicitGC
2277 Enables the option that disables processing of calls to the Sys‐
2278 tem.gc() method. This option is disabled by default, meaning
2279 that calls to System.gc() are processed. If processing of calls
2280 to System.gc() is disabled, then the JVM still performs GC when
2281 necessary.
2282 -XX:+ExplicitGCInvokesConcurrent
2284 Enables invoking of concurrent GC by using the System.gc() re‐
2285 quest. This option is disabled by default and can be enabled
2286 only with the -XX:+UseG1GC option.
2287 -XX:G1AdaptiveIHOPNumInitialSamples=number
2289 When -XX:UseAdaptiveIHOP is enabled, this option sets the number
2290 of completed marking cycles used to gather samples until G1
2291 adaptively determines the optimum value of -XX:InitiatingHeapOc‐
2292 cupancyPercent. Before, G1 uses the value of -XX:Initiat‐
2293 ingHeapOccupancyPercent directly for this purpose. The default
2294 value is 3.
2295 -XX:G1HeapRegionSize=size
2297 Sets the size of the regions into which the Java heap is subdi‐
2298 vided when using the garbage-first (G1) collector. The value is
2299 a power of 2 and can range from 1 MB to 32 MB. The default re‐
2300 gion size is determined ergonomically based on the heap size
2301 with a goal of approximately 2048 regions.
2302 The following example sets the size of the subdivisions to 16
2304 MB:
2305 -XX:G1HeapRegionSize=16m
2307 -XX:G1HeapWastePercent=percent
2309 Sets the percentage of heap that you're willing to waste. The
2310 Java HotSpot VM doesn't initiate the mixed garbage collection
2311 cycle when the reclaimable percentage is less than the heap
2312 waste percentage. The default is 5 percent.
2313 -XX:G1MaxNewSizePercent=percent
2315 Sets the percentage of the heap size to use as the maximum for
2316 the young generation size. The default value is 60 percent of
2317 your Java heap.
2318 This is an experimental flag. This setting replaces the -XX:De‐
2320 faultMaxNewGenPercent setting.
2321 -XX:G1MixedGCCountTarget=number
2323 Sets the target number of mixed garbage collections after a
2324 marking cycle to collect old regions with at most G1MixedG‐
2325 CLIveThresholdPercent live data. The default is 8 mixed garbage
2326 collections. The goal for mixed collections is to be within
2327 this target number.
2328 -XX:G1MixedGCLiveThresholdPercent=percent
2330 Sets the occupancy threshold for an old region to be included in
2331 a mixed garbage collection cycle. The default occupancy is 85
2332 percent.
2333 This is an experimental flag. This setting replaces the
2335 -XX:G1OldCSetRegionLiveThresholdPercent setting.
2336 -XX:G1NewSizePercent=percent
2338 Sets the percentage of the heap to use as the minimum for the
2339 young generation size. The default value is 5 percent of your
2340 Java heap.
2341 This is an experimental flag. This setting replaces the -XX:De‐
2343 faultMinNewGenPercent setting.
2344 -XX:G1OldCSetRegionThresholdPercent=percent
2346 Sets an upper limit on the number of old regions to be collected
2347 during a mixed garbage collection cycle. The default is 10 per‐
2348 cent of the Java heap.
2349 -XX:G1ReservePercent=percent
2351 Sets the percentage of the heap (0 to 50) that's reserved as a
2352 false ceiling to reduce the possibility of promotion failure for
2353 the G1 collector. When you increase or decrease the percentage,
2354 ensure that you adjust the total Java heap by the same amount.
2355 By default, this option is set to 10%.
2356 The following example sets the reserved heap to 20%:
2358 -XX:G1ReservePercent=20
2360 -XX:+G1UseAdaptiveIHOP
2362 Controls adaptive calculation of the old generation occupancy to
2363 start background work preparing for an old generation collec‐
2364 tion. If enabled, G1 uses -XX:InitiatingHeapOccupancyPercent
2365 for the first few times as specified by the value of -XX:G1Adap‐
2366 tiveIHOPNumInitialSamples, and after that adaptively calculates
2367 a new optimum value for the initiating occupancy automatically.
2368 Otherwise, the old generation collection process always starts
2369 at the old generation occupancy determined by -XX:Initiat‐
2370 ingHeapOccupancyPercent.
2371 The default is enabled.
2373 -XX:InitialHeapSize=size
2375 Sets the initial size (in bytes) of the memory allocation pool.
2376 This value must be either 0, or a multiple of 1024 and greater
2377 than 1 MB. Append the letter k or K to indicate kilobytes, m or
2378 M to indicate megabytes, or g or G to indicate gigabytes. The
2379 default value is selected at run time based on the system con‐
2380 figuration.
2381 The following examples show how to set the size of allocated
2383 memory to 6 MB using various units:
2384 -XX:InitialHeapSize=6291456
2386 -XX:InitialHeapSize=6144k
2387 -XX:InitialHeapSize=6m
2388 If you set this option to 0, then the initial size is set as the
2390 sum of the sizes allocated for the old generation and the young
2391 generation. The size of the heap for the young generation can
2392 be set using the -XX:NewSize option.
2393 -XX:InitialRAMPercentage=percent
2395 Sets the initial amount of memory that the JVM will use for the
2396 Java heap before applying ergonomics heuristics as a percentage
2397 of the maximum amount determined as described in the -XX:MaxRAM
2398 option. The default value is 1.5625 percent.
2399 The following example shows how to set the percentage of the
2401 initial amount of memory used for the Java heap:
2402 -XX:InitialRAMPercentage=5
2404 -XX:InitialSurvivorRatio=ratio
2406 Sets the initial survivor space ratio used by the throughput
2407 garbage collector (which is enabled by the -XX:+UseParallelGC
2408 and/or -XX:+UseParallelOldGC options). Adaptive sizing is en‐
2409 abled by default with the throughput garbage collector by using
2410 the -XX:+UseParallelGC and -XX:+UseParallelOldGC options, and
2411 the survivor space is resized according to the application be‐
2412 havior, starting with the initial value. If adaptive sizing is
2413 disabled (using the -XX:-UseAdaptiveSizePolicy option), then the
2414 -XX:SurvivorRatio option should be used to set the size of the
2415 survivor space for the entire execution of the application.
2416 The following formula can be used to calculate the initial size
2418 of survivor space (S) based on the size of the young generation
2419 (Y), and the initial survivor space ratio (R):
2420 S=Y/(R+2)
2422 The 2 in the equation denotes two survivor spaces. The larger
2424 the value specified as the initial survivor space ratio, the
2425 smaller the initial survivor space size.
2426 By default, the initial survivor space ratio is set to 8. If
2428 the default value for the young generation space size is used (2
2429 MB), then the initial size of the survivor space is 0.2 MB.
2430 The following example shows how to set the initial survivor
2432 space ratio to 4:
2433 -XX:InitialSurvivorRatio=4
2435 -XX:InitiatingHeapOccupancyPercent=percent
2437 Sets the percentage of the old generation occupancy (0 to 100)
2438 at which to start the first few concurrent marking cycles for
2439 the G1 garbage collector.
2440 By default, the initiating value is set to 45%. A value of 0
2442 implies nonstop concurrent GC cycles from the beginning until G1
2443 adaptively sets this value.
2444 See also the -XX:G1UseAdaptiveIHOP and -XX:G1AdaptiveIHOPNumIni‐
2446 tialSamples options.
2447 The following example shows how to set the initiating heap occu‐
2449 pancy to 75%:
2450 -XX:InitiatingHeapOccupancyPercent=75
2452 -XX:MaxGCPauseMillis=time
2454 Sets a target for the maximum GC pause time (in milliseconds).
2455 This is a soft goal, and the JVM will make its best effort to
2456 achieve it. The specified value doesn't adapt to your heap
2457 size. By default, for G1 the maximum pause time target is 200
2458 milliseconds. The other generational collectors do not use a
2459 pause time goal by default.
2460 The following example shows how to set the maximum target pause
2462 time to 500 ms:
2463 -XX:MaxGCPauseMillis=500
2465 -XX:MaxHeapSize=size
2467 Sets the maximum size (in byes) of the memory allocation pool.
2468 This value must be a multiple of 1024 and greater than 2 MB.
2469 Append the letter k or K to indicate kilobytes, m or M to indi‐
2470 cate megabytes, or g or G to indicate gigabytes. The default
2471 value is selected at run time based on the system configuration.
2472 For server deployments, the options -XX:InitialHeapSize and
2473 -XX:MaxHeapSize are often set to the same value.
2474 The following examples show how to set the maximum allowed size
2476 of allocated memory to 80 MB using various units:
2477 -XX:MaxHeapSize=83886080
2479 -XX:MaxHeapSize=81920k
2480 -XX:MaxHeapSize=80m
2481 The -XX:MaxHeapSize option is equivalent to -Xmx.
2483 -XX:MaxHeapFreeRatio=percent
2485 Sets the maximum allowed percentage of free heap space (0 to
2486 100) after a GC event. If free heap space expands above this
2487 value, then the heap is shrunk. By default, this value is set
2488 to 70%.
2489 Minimize the Java heap size by lowering the values of the param‐
2491 eters MaxHeapFreeRatio (default value is 70%) and MinHeapFreeRa‐
2492 tio (default value is 40%) with the command-line options
2493 -XX:MaxHeapFreeRatio and -XX:MinHeapFreeRatio. Lowering Max‐
2494 HeapFreeRatio to as low as 10% and MinHeapFreeRatio to 5% has
2495 successfully reduced the heap size without too much performance
2496 regression; however, results may vary greatly depending on your
2497 application. Try different values for these parameters until
2498 they're as low as possible yet still retain acceptable perfor‐
2499 mance.
2500 -XX:MaxHeapFreeRatio=10 -XX:MinHeapFreeRatio=5
2502 Customers trying to keep the heap small should also add the op‐
2504 tion -XX:-ShrinkHeapInSteps. See Performance Tuning Examples
2505 for a description of using this option to keep the Java heap
2506 small by reducing the dynamic footprint for embedded applica‐
2507 tions.
2508 -XX:MaxMetaspaceSize=size
2510 Sets the maximum amount of native memory that can be allocated
2511 for class metadata. By default, the size isn't limited. The
2512 amount of metadata for an application depends on the application
2513 itself, other running applications, and the amount of memory
2514 available on the system.
2515 The following example shows how to set the maximum class metada‐
2517 ta size to 256 MB:
2518 -XX:MaxMetaspaceSize=256m
2520 -XX:MaxNewSize=size
2522 Sets the maximum size (in bytes) of the heap for the young gen‐
2523 eration (nursery). The default value is set ergonomically.
2524 -XX:MaxRAM=size
2526 Sets the maximum amount of memory that the JVM may use for the
2527 Java heap before applying ergonomics heuristics. The default
2528 value is the maximum amount of available memory to the JVM
2529 process or 128 GB, whichever is lower.
2530 The maximum amount of available memory to the JVM process is the
2532 minimum of the machine's physical memory and any constraints set
2533 by the environment (e.g. container).
2534 Specifying this option disables automatic use of compressed oops
2536 if the combined result of this and other options influencing the
2537 maximum amount of memory is larger than the range of memory ad‐
2538 dressable by compressed oops. See -XX:UseCompressedOops for
2539 further information about compressed oops.
2540 The following example shows how to set the maximum amount of
2542 available memory for sizing the Java heap to 2 GB:
2543 -XX:MaxRAM=2G
2545 -XX:MaxRAMPercentage=percent
2547 Sets the maximum amount of memory that the JVM may use for the
2548 Java heap before applying ergonomics heuristics as a percentage
2549 of the maximum amount determined as described in the -XX:MaxRAM
2550 option. The default value is 25 percent.
2551 Specifying this option disables automatic use of compressed oops
2553 if the combined result of this and other options influencing the
2554 maximum amount of memory is larger than the range of memory ad‐
2555 dressable by compressed oops. See -XX:UseCompressedOops for
2556 further information about compressed oops.
2557 The following example shows how to set the percentage of the
2559 maximum amount of memory used for the Java heap:
2560 -XX:MaxRAMPercentage=75
2562 -XX:MinRAMPercentage=percent
2564 Sets the maximum amount of memory that the JVM may use for the
2565 Java heap before applying ergonomics heuristics as a percentage
2566 of the maximum amount determined as described in the -XX:MaxRAM
2567 option for small heaps. A small heap is a heap of approximately
2568 125 MB. The default value is 50 percent.
2569 The following example shows how to set the percentage of the
2571 maximum amount of memory used for the Java heap for small heaps:
2572 -XX:MinRAMPercentage=75
2574 -XX:MaxTenuringThreshold=threshold
2576 Sets the maximum tenuring threshold for use in adaptive GC siz‐
2577 ing. The largest value is 15. The default value is 15 for the
2578 parallel (throughput) collector.
2579 The following example shows how to set the maximum tenuring
2581 threshold to 10:
2582 -XX:MaxTenuringThreshold=10
2584 -XX:MetaspaceSize=size
2586 Sets the size of the allocated class metadata space that trig‐
2587 gers a garbage collection the first time it's exceeded. This
2588 threshold for a garbage collection is increased or decreased de‐
2589 pending on the amount of metadata used. The default size de‐
2590 pends on the platform.
2591 -XX:MinHeapFreeRatio=percent
2593 Sets the minimum allowed percentage of free heap space (0 to
2594 100) after a GC event. If free heap space falls below this val‐
2595 ue, then the heap is expanded. By default, this value is set to
2596 40%.
2597 Minimize Java heap size by lowering the values of the parameters
2599 MaxHeapFreeRatio (default value is 70%) and MinHeapFreeRatio
2600 (default value is 40%) with the command-line options -XX:Max‐
2601 HeapFreeRatio and -XX:MinHeapFreeRatio. Lowering MaxHeapFreeRa‐
2602 tio to as low as 10% and MinHeapFreeRatio to 5% has successfully
2603 reduced the heap size without too much performance regression;
2604 however, results may vary greatly depending on your application.
2605 Try different values for these parameters until they're as low
2606 as possible, yet still retain acceptable performance.
2607 -XX:MaxHeapFreeRatio=10 -XX:MinHeapFreeRatio=5
2609 Customers trying to keep the heap small should also add the op‐
2611 tion -XX:-ShrinkHeapInSteps. See Performance Tuning Examples
2612 for a description of using this option to keep the Java heap
2613 small by reducing the dynamic footprint for embedded applica‐
2614 tions.
2615 -XX:MinHeapSize=size
2617 Sets the minimum size (in bytes) of the memory allocation pool.
2618 This value must be either 0, or a multiple of 1024 and greater
2619 than 1 MB. Append the letter k or K to indicate kilobytes, m or
2620 M to indicate megabytes, or g or G to indicate gigabytes. The
2621 default value is selected at run time based on the system con‐
2622 figuration.
2623 The following examples show how to set the mimimum size of allo‐
2625 cated memory to 6 MB using various units:
2626 -XX:MinHeapSize=6291456
2628 -XX:MinHeapSize=6144k
2629 -XX:MinHeapSize=6m
2630 If you set this option to 0, then the minimum size is set to the
2632 same value as the initial size.
2633 -XX:NewRatio=ratio
2635 Sets the ratio between young and old generation sizes. By de‐
2636 fault, this option is set to 2. The following example shows how
2637 to set the young-to-old ratio to 1:
2638 -XX:NewRatio=1
2640 -XX:NewSize=size
2642 Sets the initial size (in bytes) of the heap for the young gen‐
2643 eration (nursery). Append the letter k or K to indicate kilo‐
2644 bytes, m or M to indicate megabytes, or g or G to indicate giga‐
2645 bytes.
2646 The young generation region of the heap is used for new objects.
2648 GC is performed in this region more often than in other regions.
2649 If the size for the young generation is too low, then a large
2650 number of minor GCs are performed. If the size is too high,
2651 then only full GCs are performed, which can take a long time to
2652 complete. It is recommended that you keep the size for the
2653 young generation greater than 25% and less than 50% of the over‐
2654 all heap size.
2655 The following examples show how to set the initial size of the
2657 young generation to 256 MB using various units:
2658 -XX:NewSize=256m
2660 -XX:NewSize=262144k
2661 -XX:NewSize=268435456
2662 The -XX:NewSize option is equivalent to -Xmn.
2664 -XX:ParallelGCThreads=threads
2666 Sets the number of the stop-the-world (STW) worker threads. The
2667 default value depends on the number of CPUs available to the JVM
2668 and the garbage collector selected.
2669 For example, to set the number of threads for G1 GC to 2, speci‐
2671 fy the following option:
2672 -XX:ParallelGCThreads=2
2674 -XX:+ParallelRefProcEnabled
2676 Enables parallel reference processing. By default, this option
2677 is disabled.
2678 -XX:+PrintAdaptiveSizePolicy
2680 Enables printing of information about adaptive-generation siz‐
2681 ing. By default, this option is disabled.
2682 -XX:+ScavengeBeforeFullGC
2684 Enables GC of the young generation before each full GC. This
2685 option is enabled by default. It is recommended that you don't
2686 disable it, because scavenging the young generation before a
2687 full GC can reduce the number of objects reachable from the old
2688 generation space into the young generation space. To disable GC
2689 of the young generation before each full GC, specify the option
2690 -XX:-ScavengeBeforeFullGC.
2691 -XX:SoftRefLRUPolicyMSPerMB=time
2693 Sets the amount of time (in milliseconds) a softly reachable ob‐
2694 ject is kept active on the heap after the last time it was ref‐
2695 erenced. The default value is one second of lifetime per free
2696 megabyte in the heap. The -XX:SoftRefLRUPolicyMSPerMB option
2697 accepts integer values representing milliseconds per one
2698 megabyte of the current heap size (for Java HotSpot Client VM)
2699 or the maximum possible heap size (for Java HotSpot Server VM).
2700 This difference means that the Client VM tends to flush soft
2701 references rather than grow the heap, whereas the Server VM
2702 tends to grow the heap rather than flush soft references. In
2703 the latter case, the value of the -Xmx option has a significant
2704 effect on how quickly soft references are garbage collected.
2705 The following example shows how to set the value to 2.5 seconds:
2707 -XX:SoftRefLRUPolicyMSPerMB=2500
2709 -XX:-ShrinkHeapInSteps
2711 Incrementally reduces the Java heap to the target size, speci‐
2712 fied by the option -XX:MaxHeapFreeRatio. This option is enabled
2713 by default. If disabled, then it immediately reduces the Java
2714 heap to the target size instead of requiring multiple garbage
2715 collection cycles. Disable this option if you want to minimize
2716 the Java heap size. You will likely encounter performance
2717 degradation when this option is disabled.
2718 See Performance Tuning Examples for a description of using the
2720 MaxHeapFreeRatio option to keep the Java heap small by reducing
2721 the dynamic footprint for embedded applications.
2722 -XX:StringDeduplicationAgeThreshold=threshold
2724 Identifies String objects reaching the specified age that are
2725 considered candidates for deduplication. An object's age is a
2726 measure of how many times it has survived garbage collection.
2727 This is sometimes referred to as tenuring.
2728 Note: String objects that are promoted to an old heap re‐
2730 gion before this age has been reached are always consid‐
2731 ered candidates for deduplication. The default value for
2732 this option is 3. See the -XX:+UseStringDeduplication
2733 option.
2734 -XX:SurvivorRatio=ratio
2736 Sets the ratio between eden space size and survivor space size.
2737 By default, this option is set to 8. The following example
2738 shows how to set the eden/survivor space ratio to 4:
2739 -XX:SurvivorRatio=4
2741 -XX:TargetSurvivorRatio=percent
2743 Sets the desired percentage of survivor space (0 to 100) used
2744 after young garbage collection. By default, this option is set
2745 to 50%.
2746 The following example shows how to set the target survivor space
2748 ratio to 30%:
2749 -XX:TargetSurvivorRatio=30
2751 -XX:TLABSize=size
2753 Sets the initial size (in bytes) of a thread-local allocation
2754 buffer (TLAB). Append the letter k or K to indicate kilobytes,
2755 m or M to indicate megabytes, or g or G to indicate gigabytes.
2756 If this option is set to 0, then the JVM selects the initial
2757 size automatically.
2758 The following example shows how to set the initial TLAB size to
2760 512 KB:
2761 -XX:TLABSize=512k
2763 -XX:+UseAdaptiveSizePolicy
2765 Enables the use of adaptive generation sizing. This option is
2766 enabled by default. To disable adaptive generation sizing,
2767 specify -XX:-UseAdaptiveSizePolicy and set the size of the memo‐
2768 ry allocation pool explicitly. See the -XX:SurvivorRatio op‐
2769 tion.
2770 -XX:+UseG1GC
2772 Enables the use of the garbage-first (G1) garbage collector.
2773 It's a server-style garbage collector, targeted for multiproces‐
2774 sor machines with a large amount of RAM. This option meets GC
2775 pause time goals with high probability, while maintaining good
2776 throughput. The G1 collector is recommended for applications
2777 requiring large heaps (sizes of around 6 GB or larger) with lim‐
2778 ited GC latency requirements (a stable and predictable pause
2779 time below 0.5 seconds). By default, this option is enabled and
2780 G1 is used as the default garbage collector.
2781 -XX:+UseGCOverheadLimit
2783 Enables the use of a policy that limits the proportion of time
2784 spent by the JVM on GC before an OutOfMemoryError exception is
2785 thrown. This option is enabled, by default, and the parallel GC
2786 will throw an OutOfMemoryError if more than 98% of the total
2787 time is spent on garbage collection and less than 2% of the heap
2788 is recovered. When the heap is small, this feature can be used
2789 to prevent applications from running for long periods of time
2790 with little or no progress. To disable this option, specify the
2791 option -XX:-UseGCOverheadLimit.
2792 -XX:+UseNUMA
2794 Enables performance optimization of an application on a machine
2795 with nonuniform memory architecture (NUMA) by increasing the ap‐
2796 plication's use of lower latency memory. By default, this op‐
2797 tion is disabled and no optimization for NUMA is made. The op‐
2798 tion is available only when the parallel garbage collector is
2799 used (-XX:+UseParallelGC).
2800 -XX:+UseParallelGC
2802 Enables the use of the parallel scavenge garbage collector (also
2803 known as the throughput collector) to improve the performance of
2804 your application by leveraging multiple processors.
2805 By default, this option is disabled and the default collector is
2807 used. If it's enabled, then the -XX:+UseParallelOldGC option is
2808 automatically enabled, unless you explicitly disable it.
2809 -XX:+UseSerialGC
2811 Enables the use of the serial garbage collector. This is gener‐
2812 ally the best choice for small and simple applications that
2813 don't require any special functionality from garbage collection.
2814 By default, this option is disabled and the default collector is
2815 used.
2816 -XX:+UseSHM
2818 Linux only: Enables the JVM to use shared memory to set up large
2819 pages.
2820 See Large Pages for setting up large pages.
2822 -XX:+UseStringDeduplication
2824 Enables string deduplication. By default, this option is dis‐
2825 abled. To use this option, you must enable the garbage-first
2826 (G1) garbage collector.
2827 String deduplication reduces the memory footprint of String ob‐
2829 jects on the Java heap by taking advantage of the fact that many
2830 String objects are identical. Instead of each String object
2831 pointing to its own character array, identical String objects
2832 can point to and share the same character array.
2833 -XX:+UseTLAB
2835 Enables the use of thread-local allocation blocks (TLABs) in the
2836 young generation space. This option is enabled by default. To
2837 disable the use of TLABs, specify the option -XX:-UseTLAB.
2838 -XX:+UseZGC
2840 Enables the use of the Z garbage collector. This garbage col‐
2841 lector is best for providing lowest latency with large Java
2842 heaps at some throughput cost. This is an experimental garbage
2843 collector, you need to specify -XX:+UnlockExperimentalVMOptions
2844 before -XX:+UseZGC on the command line.
2845 Example:
2847 -XX:+UnlockExperimentalVMOptions -XX:+UseZGC
2849
2851 These java options are deprecated and might be removed in a future JDK
2852 release. They're still accepted and acted upon, but a warning is is‐
2853 sued when they're used.
2854 -Xfuture
2856 Enables strict class-file format checks that enforce close con‐
2857 formance to the class-file format specification. Developers
2858 should use this flag when developing new code. Stricter checks
2859 may become the default in future releases.
2860 -Xloggc:filename
2862 Sets the file to which verbose GC events information should be
2863 redirected for logging. The -Xloggc option overrides -ver‐
2864 bose:gc if both are given with the same java command. -Xlog‐
2865 gc:filename is replaced by -Xlog:gc:filename. See Enable Log‐
2866 ging with the JVM Unified Logging Framework.
2867 Example:
2869 -Xlog:gc:garbage-collection.log
2871 -XX:+FlightRecorder
2873 Enables the use of Java Flight Recorder (JFR) during the runtime
2874 of the application. Since JDK 8u40 this option has not been re‐
2875 quired to use JFR.
2876 -XX:InitialRAMFraction=ratio
2878 Sets the initial amount of memory that the JVM may use for the
2879 Java heap before applying ergonomics heuristics as a ratio of
2880 the maximum amount determined as described in the -XX:MaxRAM op‐
2881 tion. The default value is 64.
2882 Use the option -XX:InitialRAMPercentage instead.
2884 -XX:MaxRAMFraction=ratio
2886 Sets the maximum amount of memory that the JVM may use for the
2887 Java heap before applying ergonomics heuristics as a fraction of
2888 the maximum amount determined as described in the -XX:MaxRAM op‐
2889 tion. The default value is 4.
2890 Specifying this option disables automatic use of compressed oops
2892 if the combined result of this and other options influencing the
2893 maximum amount of memory is larger than the range of memory ad‐
2894 dressable by compressed oops. See -XX:UseCompressedOops for
2895 further information about compressed oops.
2896 Use the option -XX:MaxRAMPercentage instead.
2898 -XX:MinRAMFraction=ratio
2900 Sets the maximum amount of memory that the JVM may use for the
2901 Java heap before applying ergonomics heuristics as a fraction of
2902 the maximum amount determined as described in the -XX:MaxRAM op‐
2903 tion for small heaps. A small heap is a heap of approximately
2904 125 MB. The default value is 2.
2905 Use the option -XX:MinRAMPercentage instead.
2907 -XX:+TraceClassLoading
2909 Enables tracing of classes as they are loaded. By default, this
2910 option is disabled and classes aren't traced.
2911 The replacement Unified Logging syntax is -Xlog:class+load=lev‐
2913 el. See Enable Logging with the JVM Unified Logging Framework
2914 Use level=info for regular information, or level=debug for addi‐
2916 tional information. In Unified Logging syntax, -verbose:class
2917 equals -Xlog:class+load=info,class+unload=info.
2918 -XX:+TraceClassLoadingPreorder
2920 Enables tracing of all loaded classes in the order in which
2921 they're referenced. By default, this option is disabled and
2922 classes aren't traced.
2923 The replacement Unified Logging syntax is -Xlog:class+pre‐
2925 order=debug. See Enable Logging with the JVM Unified Logging
2926 Framework.
2927 -XX:+TraceClassResolution
2929 Enables tracing of constant pool resolutions. By default, this
2930 option is disabled and constant pool resolutions aren't traced.
2931 The replacement Unified Logging syntax is -Xlog:class+re‐
2933 solve=debug. See Enable Logging with the JVM Unified Logging
2934 Framework.
2935 -XX:+TraceLoaderConstraints
2937 Enables tracing of the loader constraints recording. By de‐
2938 fault, this option is disabled and loader constraints recording
2939 isn't traced.
2940 The replacement Unified Logging syntax is -Xlog:class+load‐
2942 er+constraints=info. See Enable Logging with the JVM Unified
2943 Logging Framework.
2944 -XX:+UseParallelOldGC
2946 Enables the use of the parallel garbage collector for full GCs.
2947 By default, this option is disabled. Enabling it automatically
2948 enables the -XX:+UseParallelGC option.
2949
2951 These java options are still accepted but ignored, and a warning is is‐
2952 sued when they're used.
2953 -XX:+FailOverToOldVerifier
2955 Enables automatic failover to the old verifier when the new type
2956 checker fails. By default, this option is disabled and it's ig‐
2957 nored (that is, treated as disabled) for classes with a recent
2958 bytecode version. You can enable it only for classes with older
2959 versions of the bytecode.
2960 -XX:+UseMembar
2962 Enabled issuing membars on thread-state transitions. This op‐
2963 tion was disabled by default on all platforms except ARM
2964 servers, where it was enabled.
2965 -XX:MaxPermSize=size
2967 Sets the maximum permanent generation space size (in bytes).
2968 This option was deprecated in JDK 8 and superseded by the
2969 -XX:MaxMetaspaceSize option.
2970 -XX:PermSize=size
2972 Sets the space (in bytes) allocated to the permanent generation
2973 that triggers a garbage collection if it's exceeded. This op‐
2974 tion was deprecated in JDK 8 and superseded by the -XX:Metas‐
2975 paceSize option.
2976
2978 No documented java options have been removed in JDK 14.
2979 For the lists and descriptions of options removed in previous releases
2981 see the Removed Java Options section in:
2982 · Java Platform, Standard Edition Tools Reference, Release 13
2984 [https://docs.oracle.com/en/java/javase/13/docs/specs/man/java.html]
2985 · Java Platform, Standard Edition Tools Reference, Release 12
2987 [https://docs.oracle.com/en/java/javase/12/tools/ja‐
2988 va.html#GUID-3B1CE181-CD30-4178-9602-230B800D4FAE]
2989 · Java Platform, Standard Edition Tools Reference, Release 11
2991 [https://docs.oracle.com/en/java/javase/11/tools/ja‐
2992 va.html#GUID-741FC470-AA3E-494A-8D2B-1B1FE4A990D1]
2993 · Java Platform, Standard Edition Tools Reference, Release 10
2995 [https://docs.oracle.com/javase/10/tools/java.htm#JSWOR624]
2996 · Java Platform, Standard Edition Tools Reference, Release 9
2998 [https://docs.oracle.com/javase/9/tools/java.htm#JSWOR624]
2999 · Java Platform, Standard Edition Tools Reference, Release 8 for Oracle
3001 JDK on Windows [https://docs.oracle.com/javase/8/docs/tech‐
3002 notes/tools/windows/java.html#BGBCIEFC]
3003 · Java Platform, Standard Edition Tools Reference, Release 8 for Oracle
3005 JDK on Solaris, Linux, and macOS [https://docs.ora‐
3006 cle.com/javase/8/docs/technotes/tools/unix/java.html#BGBCIEFC]
3007
3009 You can shorten or simplify the java command by using @ argument files
3010 to specify one or more text files that contain arguments, such as op‐
3011 tions and class names, which are passed to the java command. This
3012 let's you to create java commands of any length on any operating sys‐
3013 tem.
3014 In the command line, use the at sign (@) prefix to identify an argument
3016 file that contains java options and class names. When the java command
3017 encounters a file beginning with the at sign (@), it expands the con‐
3018 tents of that file into an argument list just as they would be speci‐
3019 fied on the command line.
3020 The java launcher expands the argument file contents until it encoun‐
3022 ters the --disable-@files option. You can use the --disable-@files op‐
3023 tion anywhere on the command line, including in an argument file, to
3024 stop @ argument files expansion.
3025 The following items describe the syntax of java argument files:
3027 · The argument file must contain only ASCII characters or characters in
3029 system default encoding that's ASCII friendly, such as UTF-8.
3030 · The argument file size must not exceed MAXINT (2,147,483,647) bytes.
3032 · The launcher doesn't expand wildcards that are present within an ar‐
3034 gument file.
3035 · Use white space or new line characters to separate arguments included
3037 in the file.
3038 · White space includes a white space character, \t, \n, \r, and \f.
3040 For example, it is possible to have a path with a space, such as
3042 c:\Program Files that can be specified as either "c:\\Program Files"
3043 or, to avoid an escape, c:\Program" "Files.
3044 · Any option that contains spaces, such as a path component, must be
3046 within quotation marks using quotation ('"') characters in its en‐
3047 tirety.
3048 · A string within quotation marks may contain the characters \n, \r,
3050 \t, and \f. They are converted to their respective ASCII codes.
3051 · If a file name contains embedded spaces, then put the whole file name
3053 in double quotation marks.
3054 · File names in an argument file are relative to the current directory,
3056 not to the location of the argument file.
3057 · Use the number sign # in the argument file to identify comments. All
3059 characters following the # are ignored until the end of line.
3060 · Additional at sign @ prefixes to @ prefixed options act as an escape,
3062 (the first @ is removed and the rest of the arguments are presented
3063 to the launcher literally).
3064 · Lines may be continued using the continuation character (\) at the
3066 end-of-line. The two lines are concatenated with the leading white
3067 spaces trimmed. To prevent trimming the leading white spaces, a con‐
3068 tinuation character (\) may be placed at the first column.
3069 · Because backslash (\) is an escape character, a backslash character
3071 must be escaped with another backslash character.
3072 · Partial quote is allowed and is closed by an end-of-file.
3074 · An open quote stops at end-of-line unless \ is the last character,
3076 which then joins the next line by removing all leading white space
3077 characters.
3078 · Wildcards (*) aren't allowed in these lists (such as specifying *.ja‐
3080 va).
3081 · Use of the at sign (@) to recursively interpret files isn't support‐
3083 ed.
3084 Example of Open or Partial Quotes in an Argument File
3086 In the argument file,
3087 -cp "lib/
3089 cool/
3090 app/
3091 jars
3092 this is interpreted as:
3094 -cp lib/cool/app/jars
3096 Example of a Backslash Character Escaped with Another Backslash
3098 Character in an Argument File
3099 To output the following:
3101 -cp c:\Program Files (x86)\Java\jre\lib\ext;c:\Program Files\Ja‐
3103 va\jre9\lib\ext
3104 The backslash character must be specified in the argument file as:
3106 -cp "c:\\Program Files (x86)\\Java\\jre\\lib\\ext;c:\\Pro‐
3108 gram Files\\Java\\jre9\\lib\\ext"
3109 Example of an EOL Escape Used to Force Concatenation of Lines in an
3111 Argument File
3112 In the argument file,
3114 -cp "/lib/cool app/jars:\
3116 /lib/another app/jars"
3117 This is interpreted as:
3119 -cp /lib/cool app/jars:/lib/another app/jars
3121 Example of Line Continuation with Leading Spaces in an Argument File
3123 In the argument file,
3124 -cp "/lib/cool\
3126 \app/jars???
3127 This is interpreted as:
3129 -cp /lib/cool app/jars
3131 Examples of Using Single Argument File
3133 You can use a single argument file, such as myargumentfile in the fol‐
3134 lowing example, to hold all required java arguments:
3135 java @myargumentfile
3137 Examples of Using Argument Files with Paths
3139 You can include relative paths in argument files; however, they're rel‐
3140 ative to the current working directory and not to the paths of the ar‐
3141 gument files themselves. In the following example, path1/options and
3142 path2/options represent argument files with different paths. Any rela‐
3143 tive paths that they contain are relative to the current working direc‐
3144 tory and not to the argument files:
3145 java @path1/options @path2/classes
3147
3149 Overview
3150 There are occasions when having insight into the current state of the
3151 JVM code heap would be helpful to answer questions such as:
3152 · Why was the JIT turned off and then on again and again?
3154 · Where has all the code heap space gone?
3156 · Why is the method sweeper not working effectively?
3158 To provide this insight, a code heap state analytics feature has been
3160 implemented that enables on-the-fly analysis of the code heap. The an‐
3161 alytics process is divided into two parts. The first part examines the
3162 entire code heap and aggregates all information that is believed to be
3163 useful or important. The second part consists of several independent
3164 steps that print the collected information with an emphasis on differ‐
3165 ent aspects of the data. Data collection and printing are done on an
3166 "on request" basis.
3167 Syntax
3169 Requests for real-time, on-the-fly analysis can be issued with the fol‐
3170 lowing command:
3171 jcmd pid Compiler.CodeHeap_Analytics [function] [granularity]
3173 If you are only interested in how the code heap looks like after run‐
3175 ning a sample workload, you can use the command line option:
3176 -Xlog:codecache=Trace
3178 To see the code heap state when a "CodeCache full" condition exists,
3180 start the VM with the command line option:
3181 -Xlog:codecache=Debug
3183 See CodeHeap State Analytics (OpenJDK) [https://bugs.openjdk.ja‐
3185 va.net/secure/attachment/75649/JVM_CodeHeap_StateAnalytics_V2.pdf] for
3186 a detailed description of the code heap state analytics feature, the
3187 supported functions, and the granularity options.
3188
3190 You use the -Xlog option to configure or enable logging with the Java
3191 Virtual Machine (JVM) unified logging framework.
3192 Synopsis
3194 -Xlog[:[what][:[output][:[decorators][:output-options[,...]]]]]
3195 what Specifies a combination of tags and levels of the form
3197 tag1[+tag2...][*][=level][,...]. Unless the wildcard (*) is
3198 specified, only log messages tagged with exactly the tags speci‐
3199 fied are matched. See -Xlog Tags and Levels.
3200 output Sets the type of output. Omitting the output type defaults to
3202 stdout. See -Xlog Output.
3203 decorators
3205 Configures the output to use a custom set of decorators. Omit‐
3206 ting decorators defaults to uptime, level, and tags. See Deco‐
3207 rations.
3208 output-options
3210 Sets the -Xlog logging output options.
3211 Description
3213 The Java Virtual Machine (JVM) unified logging framework provides a
3214 common logging system for all components of the JVM. GC logging for
3215 the JVM has been changed to use the new logging framework. The mapping
3216 of old GC flags to the corresponding new Xlog configuration is de‐
3217 scribed in Convert GC Logging Flags to Xlog. In addition, runtime log‐
3218 ging has also been changed to use the JVM unified logging framework.
3219 The mapping of legacy runtime logging flags to the corresponding new
3220 Xlog configuration is described in Convert Runtime Logging Flags to
3221 Xlog.
3222 The following provides quick reference to the -Xlog command and syntax
3224 for options:
3225 -Xlog Enables JVM logging on an info level.
3227 -Xlog:help
3229 Prints -Xlog usage syntax and available tags, levels, and deco‐
3230 rators along with example command lines with explanations.
3231 -Xlog:disable
3233 Turns off all logging and clears all configuration of the log‐
3234 ging framework including the default configuration for warnings
3235 and errors.
3236 -Xlog[:option]
3238 Applies multiple arguments in the order that they appear on the
3239 command line. Multiple -Xlog arguments for the same output
3240 override each other in their given order.
3241 The option is set as:
3243 [tag-selection][:[output][:[decorators][:output-op‐
3245 tions]]]
3246 Omitting the tag-selection defaults to a tag-set of all and a
3248 level of info.
3249 tag[+...] all
3251 The all tag is a meta tag consisting of all tag-sets available.
3253 The asterisk * in a tag set definition denotes a wildcard tag
3254 match. Matching with a wildcard selects all tag sets that con‐
3255 tain at least the specified tags. Without the wildcard, only
3256 exact matches of the specified tag sets are selected.
3257 output-options is
3259 filecount=file-count filesize=file size with optional K,
3261 M or G suffix
3262 Default Configuration
3264 When the -Xlog option and nothing else is specified on the command
3265 line, the default configuration is used. The default configuration
3266 logs all messages with a level that matches either warning or error re‐
3267 gardless of what tags the message is associated with. The default con‐
3268 figuration is equivalent to entering the following on the command line:
3269 -Xlog:all=warning:stdout:uptime,level,tags
3271 Controlling Logging at Runtime
3273 Logging can also be controlled at run time through Diagnostic Commands
3274 (with the jcmd utility). Everything that can be specified on the com‐
3275 mand line can also be specified dynamically with the VM.log command.
3276 As the diagnostic commands are automatically exposed as MBeans, you can
3277 use JMX to change logging configuration at run time.
3278 -Xlog Tags and Levels
3280 Each log message has a level and a tag set associated with it. The
3281 level of the message corresponds to its details, and the tag set corre‐
3282 sponds to what the message contains or which JVM component it involves
3283 (such as, gc, jit, or os). Mapping GC flags to the Xlog configuration
3284 is described in Convert GC Logging Flags to Xlog. Mapping legacy run‐
3285 time logging flags to the corresponding Xlog configuration is described
3286 in Convert Runtime Logging Flags to Xlog.
3287 Available log levels:
3289 · off
3291 · trace
3293 · debug
3295 · info
3297 · warning
3299 · error
3301 Available log tags:
3303 There are literally dozens of log tags, which in the right combina‐
3305 tions, will enable a range of logging output. The full set of avail‐
3306 able log tags can be seen using -Xlog:help. Specifying all instead of
3307 a tag combination matches all tag combinations.
3308 -Xlog Output
3310 The -Xlog option supports the following types of outputs:
3311 · stdout --- Sends output to stdout
3313 · stderr --- Sends output to stderr
3315 · file=filename --- Sends output to text file(s).
3317 When using file=filename, specifying %p and/or %t in the file name ex‐
3319 pands to the JVM's PID and startup timestamp, respectively. You can
3320 also configure text files to handle file rotation based on file size
3321 and a number of files to rotate. For example, to rotate the log file
3322 every 10 MB and keep 5 files in rotation, specify the options file‐
3323 size=10M, filecount=5. The target size of the files isn't guaranteed
3324 to be exact, it's just an approximate value. Files are rotated by de‐
3325 fault with up to 5 rotated files of target size 20 MB, unless config‐
3326 ured otherwise. Specifying filecount=0 means that the log file
3327 shouldn't be rotated. There's a possibility of the pre-existing log
3328 file getting overwritten.
3329 Decorations
3331 Logging messages are decorated with information about the message. You
3332 can configure each output to use a custom set of decorators. The order
3333 of the output is always the same as listed in the table. You can con‐
3334 figure the decorations to be used at run time. Decorations are
3335 prepended to the log message. For example:
3336 [6.567s][info][gc,old] Old collection complete
3338 Omitting decorators defaults to uptime, level, and tags. The none dec‐
3340 orator is special and is used to turn off all decorations.
3341 time (t), utctime (utc), uptime (u), timemillis (tm), uptimemillis
3343 (um), timenanos (tn), uptimenanos (un), hostname (hn), pid (p), tid
3344 (ti), level (l), tags (tg) decorators can also be specified as none for
3345 no decoration.
3346 Decorations Description
3348 ──────────────────────────────────────────────────────────────────────────
3349 time or t Current time and date in ISO-8601 format.
3350 utctime or utc Universal Time Coordinated or Coordinated Universal
3351 Time.
3352 uptime or u Time since the start of the JVM in seconds and mil‐
3353 liseconds. For example, 6.567s.
3354 timemillis or The same value as generated by System.currentTimeMil‐
3355 tm lis()
3356 uptimemillis or Milliseconds since the JVM started.
3357 um
3358 timenanos or tn The same value generated by System.nanoTime().
3359 uptimenanos or Nanoseconds since the JVM started.
3360 un
3361 hostname or hn The host name.
3362 pid or p The process identifier.
3363 tid or ti The thread identifier.
3364 level or l The level associated with the log message.
3365 tags or tg The tag-set associated with the log message.
3366 Convert GC Logging Flags to Xlog
3368 Legacy Garbage Collec‐ Xlog Configura‐ Comment
3369 tion (GC) Flag tion
3370 ───────────────────────────────────────────────────────────────────────────────────────
3371 G1PrintHeapRegions -Xlog:gc+re‐ Not Applicable
3372 gion=trace
3373 GCLogFileSize No configuration Log rotation is handled by the
3374 available framework.
3375 NumberOfGCLogFiles Not Applicable Log rotation is handled by the
3376 framework.
3377 PrintAdaptiveSizePoli‐ -Xlog:gc+er‐ Use a level of debug for most
3378 cy go*=level of the information, or a level
3379 of trace for all of what was
3380 logged for PrintAdaptive‐
3381 SizePolicy.
3382 PrintGC -Xlog:gc Not Applicable
3383 PrintGCApplicationCon‐ -Xlog:safepoint Note that PrintGCApplication‐
3384 currentTime ConcurrentTime and PrintGCAp‐
3385 plicationStoppedTime are logged
3386 on the same tag and aren't sep‐
3387 arated in the new logging.
3388 PrintGCApplication‐ -Xlog:safepoint Note that PrintGCApplication‐
3392 StoppedTime ConcurrentTime and PrintGCAp‐
3393 plicationStoppedTime are logged
3394 on the same tag and not sepa‐
3395 rated in the new logging.
3396 PrintGCCause Not Applicable GC cause is now always logged.
3397 PrintGCDateStamps Not Applicable Date stamps are logged by the
3398 framework.
3399 PrintGCDetails -Xlog:gc* Not Applicable
3400 PrintGCID Not Applicable GC ID is now always logged.
3401 PrintGCTaskTimeStamps -Xlog:gc+task*=de‐ Not Applicable
3402 bug
3403 PrintGCTimeStamps Not Applicable Time stamps are logged by the
3404 framework.
3405 PrintHeapAtGC -Xlog:gc+heap=trace Not Applicable
3406 PrintReferenceGC -Xlog:gc+ref*=debug Note that in the old logging,
3407 PrintReferenceGC had an effect
3408 only if PrintGCDetails was also
3409 enabled.
3410 PrintStringDeduplica‐ `-Xlog:gc+stringdedup*=de‐ ` Not Applicable
3411 tionStatistics bug
3412 PrintTenuringDistribu‐ -Xlog:gc+age*=level Use a level of debug for the
3413 tion most relevant information, or a
3414 level of trace for all of what
3415 was logged for PrintTenur‐
3416 ingDistribution.
3417 UseGCLogFileRotation Not Applicable What was logged for PrintTenur‐
3418 ingDistribution.
3419 Convert Runtime Logging Flags to Xlog
3421 Legacy Runtime Xlog Configuration Comment
3422 Flag
3423 ──────────────────────────────────────────────────────────────────────────────
3424 TraceExceptions -Xlog:exceptions=in‐ Not Applicable
3425 fo
3426 TraceClassLoad‐ -Xlog:class+load=lev‐ Use level=info for regular informa‐
3427 ing el tion, or level=debug for additional
3428 information. In Unified Logging
3429 syntax, -verbose:class equals
3430 -Xlog:class+load=info,class+un‐
3431 load=info.
3432 TraceClassLoad‐ -Xlog:class+pre‐ Not Applicable
3433 ingPreorder order=debug
3434 TraceClassUn‐ -Xlog:class+un‐ Use level=info for regular informa‐
3435 loading load=level tion, or level=trace for additional
3436 information. In Unified Logging
3437 syntax, -verbose:class equals
3438 -Xlog:class+load=info,class+un‐
3439 load=info.
3440 VerboseVerifi‐ -Xlog:verifica‐ Not Applicable
3441 cation tion=info
3442 TraceClassPaths -Xlog:class+path=info Not Applicable
3443 TraceClassReso‐ -Xlog:class+re‐ Not Applicable
3444 lution solve=debug
3445 TraceClassIni‐ -Xlog:class+init=info Not Applicable
3446 tialization
3447 TraceLoaderCon‐ -Xlog:class+load‐ Not Applicable
3448 straints er+constraints=info
3449 TraceClassLoad‐ -Xlog:class+load‐ Use level=debug for regular infor‐
3450 erData er+data=level mation or level=trace for addition‐
3451 al information.
3452 TraceSafepoint‐ -Xlog:safe‐ Not Applicable
3453 CleanupTime point+cleanup=info
3454 TraceSafepoint -Xlog:safepoint=debug Not Applicable
3455 TraceMonitorIn‐ -Xlog:monitorinfla‐ Not Applicable
3456 flation tion=debug
3457 TraceBiased‐ -Xlog:biasedlock‐ Use level=info for regular informa‐
3458 Locking ing=level tion, or level=trace for additional
3459 information.
3460 TraceRede‐ -Xlog:rede‐ level=info, debug, and trace pro‐
3461 fineClasses fine+class*=level vide increasing amounts of informa‐
3462 tion.
3463 -Xlog Usage Examples
3465 The following are -Xlog examples.
3466 -Xlog Logs all messages by using the info level to stdout with uptime,
3468 levels, and tags decorations. This is equivalent to using:
3469 -Xlog:all=info:stdout:uptime,levels,tags
3471 -Xlog:gc
3473 Logs messages tagged with the gc tag using info level to stdout.
3474 The default configuration for all other messages at level warn‐
3475 ing is in effect.
3476 -Xlog:gc,safepoint
3478 Logs messages tagged either with the gc or safepoint tags, both
3479 using the info level, to stdout, with default decorations. Mes‐
3480 sages tagged with both gc and safepoint won't be logged.
3481 -Xlog:gc+ref=debug
3483 Logs messages tagged with both gc and ref tags, using the debug
3484 level to stdout, with default decorations. Messages tagged only
3485 with one of the two tags won't be logged.
3486 -Xlog:gc=debug:file=gc.txt:none
3488 Logs messages tagged with the gc tag using the debug level to a
3489 file called gc.txt with no decorations. The default configura‐
3490 tion for all other messages at level warning is still in effect.
3491 -Xlog:gc=trace:file=gctrace.txt:uptimemillis,pids:filecount=5,file‐
3493 size=1024
3494 Logs messages tagged with the gc tag using the trace level to a
3495 rotating file set with 5 files with size 1 MB with the base name
3496 gctrace.txt and uses decorations uptimemillis and pid.
3497 The default configuration for all other messages at level warn‐
3499 ing is still in effect.
3500 -Xlog:gc::uptime,tid
3502 Logs messages tagged with the gc tag using the default 'info'
3503 level to default the output stdout and uses decorations uptime
3504 and tid. The default configuration for all other messages at
3505 level warning is still in effect.
3506 -Xlog:gc*=info,safepoint*=off
3508 Logs messages tagged with at least gc using the info level, but
3509 turns off logging of messages tagged with safepoint. Messages
3510 tagged with both gc and safepoint won't be logged.
3511 -Xlog:disable -Xlog:safepoint=trace:safepointtrace.txt
3513 Turns off all logging, including warnings and errors, and then
3514 enables messages tagged with safepointusing tracelevel to the
3515 file safepointtrace.txt. The default configuration doesn't ap‐
3516 ply, because the command line started with -Xlog:disable.
3517 Complex -Xlog Usage Examples
3519 The following describes a few complex examples of using the -Xlog op‐
3520 tion.
3521 -Xlog:gc+class*=debug
3523 Logs messages tagged with at least gc and class tags using the
3524 debug level to stdout. The default configuration for all other
3525 messages at the level warning is still in effect
3526 -Xlog:gc+meta*=trace,class*=off:file=gcmetatrace.txt
3528 Logs messages tagged with at least the gc and meta tags using
3529 the trace level to the file metatrace.txt but turns off all mes‐
3530 sages tagged with class. Messages tagged with gc, meta, and
3531 class aren't be logged as class* is set to off. The default
3532 configuration for all other messages at level warning is in ef‐
3533 fect except for those that include class.
3534 -Xlog:gc+meta=trace
3536 Logs messages tagged with exactly the gc and meta tags using the
3537 trace level to stdout. The default configuration for all other
3538 messages at level warning is still be in effect.
3539 -Xlog:gc+class+heap*=debug,meta*=warning,threads*=off
3541 Logs messages tagged with at least gc, class, and heap tags us‐
3542 ing the trace level to stdout but only log messages tagged with
3543 meta with level. The default configuration for all other mes‐
3544 sages at the level warning is in effect except for those that
3545 include threads.
3546
3548 You use values provided to all Java Virtual Machine (JVM) command-line
3549 flags for validation and, if the input value is invalid or
3550 out-of-range, then an appropriate error message is displayed.
3551 Whether they're set ergonomically, in a command line, by an input tool,
3553 or through the APIs (for example, classes contained in the package ja‐
3554 va.lang.management) the values provided to all Java Virtual Machine
3555 (JVM) command-line flags are validated. Ergonomics are described in
3556 Java Platform, Standard Edition HotSpot Virtual Machine Garbage Collec‐
3557 tion Tuning Guide.
3558 Range and constraints are validated either when all flags have their
3560 values set during JVM initialization or a flag's value is changed dur‐
3561 ing runtime (for example using the jcmd tool). The JVM is terminated
3562 if a value violates either the range or constraint check and an appro‐
3563 priate error message is printed on the error stream.
3564 For example, if a flag violates a range or a constraint check, then the
3566 JVM exits with an error:
3567 java -XX:AllocatePrefetchStyle=5 -version
3569 intx AllocatePrefetchStyle=5 is outside the allowed range [ 0 ... 3 ]
3570 Improperly specified VM option 'AllocatePrefetchStyle=5'
3571 Error: Could not create the Java Virtual Machine.
3572 Error: A fatal exception has occurred. Program will exit.
3573 The flag -XX:+PrintFlagsRanges prints the range of all the flags. This
3575 flag allows automatic testing of the flags by the values provided by
3576 the ranges. For the flags that have the ranges specified, the type,
3577 name, and the actual range is printed in the output.
3578 For example,
3580 intx ThreadStackSize [ 0 ... 9007199254740987 ] {pd product}
3582 For the flags that don't have the range specified, the values aren't
3584 displayed in the print out. For example:
3585 size_t NewSize [ ... ] {product}
3587 This helps to identify the flags that need to be implemented. The au‐
3589 tomatic testing framework can skip those flags that don't have values
3590 and aren't implemented.
3591
3593 You use large pages, also known as huge pages, as memory pages that are
3594 significantly larger than the standard memory page size (which varies
3595 depending on the processor and operating system). Large pages optimize
3596 processor Translation-Lookaside Buffers.
3597 A Translation-Lookaside Buffer (TLB) is a page translation cache that
3599 holds the most-recently used virtual-to-physical address translations.
3600 A TLB is a scarce system resource. A TLB miss can be costly because
3601 the processor must then read from the hierarchical page table, which
3602 may require multiple memory accesses. By using a larger memory page
3603 size, a single TLB entry can represent a larger memory range. This re‐
3604 sults in less pressure on a TLB, and memory-intensive applications may
3605 have better performance.
3606 However, large pages page memory can negatively affect system perfor‐
3608 mance. For example, when a large mount of memory is pinned by an ap‐
3609 plication, it may create a shortage of regular memory and cause exces‐
3610 sive paging in other applications and slow down the entire system. Al‐
3611 so, a system that has been up for a long time could produce excessive
3612 fragmentation, which could make it impossible to reserve enough large
3613 page memory. When this happens, either the OS or JVM reverts to using
3614 regular pages.
3615 Oracle Solaris, Linux, and Windows support large pages.
3617 Large Pages Support for Oracle Solaris
3619 Oracle Solaris includes Multiple Page Size Support (MPSS). No addi‐
3620 tional configuration is necessary.
3621 Large Pages Support for Linux
3623 The 2.6 kernel supports large pages. Some vendors have backported the
3624 code to their 2.4-based releases. To check if your system can support
3625 large page memory, try the following:
3626 # cat /proc/meminfo | grep Huge
3628 HugePages_Total: 0
3629 HugePages_Free: 0
3630 Hugepagesize: 2048 kB
3631 If the output shows the three "Huge" variables, then your system can
3633 support large page memory but it needs to be configured. If the com‐
3634 mand prints nothing, then your system doesn't support large pages. To
3635 configure the system to use large page memory, login as root, and then
3636 follow these steps:
3637 1. If you're using the option -XX:+UseSHM (instead of -XX:+Use‐
3639 HugeTLBFS), then increase the SHMMAX value. It must be larger than
3640 the Java heap size. On a system with 4 GB of physical RAM (or
3641 less), the following makes all the memory sharable:
3642 # echo 4294967295 > /proc/sys/kernel/shmmax
3644 2. If you're using the option -XX:+UseSHM or -XX:+UseHugeTLBFS, then
3646 specify the number of large pages. In the following example, 3 GB
3647 of a 4 GB system are reserved for large pages (assuming a large page
3648 size of 2048kB, then 3 GB = 3 * 1024 MB = 3072 MB = 3072 * 1024 kB =
3649 3145728 kB and 3145728 kB / 2048 kB = 1536):
3650 # echo 1536 > /proc/sys/vm/nr_hugepages
3652 Note: The values contained in /proc resets after you reboot
3654 your system, so may want to set them in an initialization
3655 script (for example, rc.local or sysctl.conf).
3656 · If you configure (or resize) the OS kernel parameters /proc/sys/ker‐
3658 nel/shmmax or /proc/sys/vm/nr_hugepages, Java processes may allocate
3659 large pages for areas in addition to the Java heap. These steps can
3660 allocate large pages for the following areas:
3661 · Java heap
3663 · Code cache
3665 · The marking bitmap data structure for the parallel GC
3667 Consequently, if you configure the nr_hugepages parameter to the size
3669 of the Java heap, then the JVM can fail in allocating the code cache
3670 areas on large pages because these areas are quite large in size.
3671 Large Pages Support for Windows
3673 To use large pages support on Windows, the administrator must first as‐
3674 sign additional privileges to the user who is running the application:
3675 1. Select Control Panel, Administrative Tools, and then Local Security
3677 Policy.
3678 2. Select Local Policies and then User Rights Assignment.
3680 3. Double-click Lock pages in memory, then add users and/or groups.
3682 4. Reboot your system.
3684 Note that these steps are required even if it's the administrator who's
3686 running the application, because administrators by default don't have
3687 the privilege to lock pages in memory.
3688
3690 Application Class Data Sharing (AppCDS) extends class data sharing
3691 (CDS) to enable application classes to be placed in a shared archive.
3692 In addition to the core library classes, AppCDS supports Class Data
3694 Sharing [https://docs.oracle.com/en/java/javase/12/vm/class-data-shar‐
3695 ing.html#GUID-7EAA3411-8CF0-4D19-BD05-DF5E1780AA91] from the following
3696 locations:
3697 · Platform classes from the runtime image
3699 · Application classes from the runtime image
3701 · Application classes from the class path
3703 · Application classes from the module path
3705 Archiving application classes provides better start up time at runtime.
3707 When running multiple JVM processes, AppCDS also reduces the runtime
3708 footprint with memory sharing for read-only metadata.
3709 CDS/AppCDS supports archiving classes from JAR files only.
3711 Prior to JDK 11, a non-empty directory was reported as a fatal error in
3713 the following conditions:
3714 · For base CDS, a non-empty directory cannot exist in the -Xbootclass‐
3716 path/a path
3717 · With -XX:+UseAppCDS, a non-empty directory could not exist in the
3719 -Xbootclasspath/a path, class path, and module path.
3720 In JDK 11 and later, -XX:+UseAppCDS is obsolete and the behavior for a
3722 non-empty directory is based on the class types in the classlist. A
3723 non-empty directory is reported as a fatal error in the following con‐
3724 ditions:
3725 · If application classes or platform classes are not loaded, dump time
3727 only reports an error if a non-empty directory exists in -Xbootclass‐
3728 path/a path
3729 · If application classes or platform classes are loaded, dump time re‐
3731 ports an error for a non-empty directory that exists in -Xbootclass‐
3732 path/a path, class path, or module path
3733 In JDK 11 and later, using -XX:DumpLoadedClassList=class_list_file re‐
3735 sults a generated classlist with all classes (both system library
3736 classes and application classes) included. You no longer have to spec‐
3737 ify -XX:+UseAppCDS with -XX:DumpLoadedClassList to produce a complete
3738 class list.
3739 In JDK 11 and later, because UseAppCDS is obsolete, SharedArchiveFile
3741 becomes a product flag by default. Specifying +UnlockDiagnosticVMOp‐
3742 tions for SharedArchiveFile is no longer needed in any configuration.
3743 Class Data Sharing (CDS)/AppCDS does not support archiving array class‐
3745 es in a class list. When an array in the class list is encountered,
3746 CDS dump time gives the explicit error message:
3747 Preload Warning: Cannot find array_name
3749 Although an array in the class list is not allowed, some array classes
3751 can still be created at CDS/AppCDS dump time. Those arrays are created
3752 during the execution of the Java code used by the Java class loaders
3753 (PlatformClassLoader and the system class loader) to load classes at
3754 dump time. The created arrays are archived with the rest of the loaded
3755 classes.
3756 Extending Class Data Sharing to Support the Module Path
3758 In JDK 11, Class Data Sharing (CDS) has been improved to support ar‐
3759 chiving classes from the module path.
3760 · To create a CDS archive using the --module-path VM option, use the
3762 following command line syntax:
3763 java -Xshare:dump -XX:SharedClassListFile=class_list_file
3765 -XX:SharedArchiveFile=shared_archive_file --mod‐
3766 ule-path=path_to_modular_jar -m module_name
3767 · To run with a CDS archive using the --module-path VM option, use the
3769 following the command line syntax:
3770 java -XX:SharedArchiveFile=shared_archive_file --mod‐
3772 ule-path=path_to_modular_jar -m module_name
3773 The following table describes how the VM options related to module
3775 paths can be used along with the -Xshare option.
3776 Option -Xshare:dump -Xshare:{on,auto}
3778 ────────────────────────────────────────────────────────────────
3779 --module-path[1] mp Allowed Allowed[2]
3780 --module Allowed Allowed
3781 --add-module Allowed Allowed
3782 --upgrade-mod‐ Disallowed (exits Allowed (disables
3783 ule-path[3] if specified) CDS)
3784 --patch-module[4] Disallowed (exits Allowed (disables
3785 if specified) CDS)
3786 --limit-modules[5] Disallowed (exits Allowed (disables
3787 if specified) CDS)
3788 [1] Although there are two ways of specifying a module in a --mod‐
3790 ule-path, that is, modular JAR or exploded module, only modular JARs
3791 are supported.
3792 [2] Different mp can be specified during dump time versus run time. If
3794 an archived class K was loaded from mp1.jar at dump time, but changes
3795 in mp cause it to be available from a different mp2.jar at run time,
3796 then the archived version of K will be disregarded at run time; K will
3797 be loaded dynamically.
3798 [3] Currently, only two system modules are upgradeable (java.compiler
3800 and jdk.internal.vm.compiler). However, these modules are seldom up‐
3801 graded in production software.
3802 [4] As documented in JEP 261, using --patch-module is strongly discour‐
3804 aged for production use.
3805 [5] --limit-modules is intended for testing purposes. It is seldom
3807 used in production software.
3808 If --upgrade-module-path, --patch-module, or --limit-modules is speci‐
3810 fied at dump time, an error will be printed and the JVM will exit. For
3811 example, if the --limit-modules option is specified at dump time, the
3812 user will see the following error:
3813 Error occurred during initialization of VM
3815 Cannot use the following option when dumping the shared archive: --limit-modules
3816 If --upgrade-module-path, --patch-module, or --limit-modules is speci‐
3818 fied at run time, a warning message will be printed indicating that CDS
3819 is disabled. For example, if the --limit-modules options is specified
3820 at run time, the user will see the following warning:
3821 Java HotSpot(TM) 64-Bit Server VM warning: CDS is disabled when the --limit-modules option is specified.
3823 Several other noteworthy things include:
3825 · Any valid combinations of -cp and --module-path are supported.
3827 · A non-empty directory in the module path causes a fatal error. The
3829 user will see the following error messages:
3830 Error: non-empty directory <directory> Hint: enable -Xlog:class+path=info to diagnose the failure Error occurred during initialization of VM Cannot have non-empty directory in paths
3832 · Unlike the class path, there's no restriction that the module path at
3834 dump time must be equal to or be a prefix of the module path at run
3835 time.
3836 · The archive is invalidated if an existing JAR in the module path is
3838 updated after archive generation.
3839 · Removing a JAR from the module path does not invalidate the shared
3841 archive. Archived classes from the removed JAR are not used at run‐
3842 time.
3843 Dynamic CDS archive
3845 Dynamic CDS archive extends AppCDS to allow archiving of classes when a
3846 Java application exits. It improves the usability of AppCDS by elimi‐
3847 nating the trial run step for creating a class list for each applica‐
3848 tion. The archived classes include all loaded application classes and
3849 library classes that are not present in the default CDS archive which
3850 is included in the JDK.
3851 A base archive is required when creating a dynamic archive. If the
3853 base archive is not specified, the default CDS archive is used as the
3854 base archive.
3855 To create a dynamic CDS archive with the default CDS archive as the
3857 base archive, just add the -XX:ArchiveClassesAtExit=<dynamic archive>
3858 option to the command line for running the Java application.
3859 If the default CDS archive does not exist, the VM will exit with the
3861 following error:
3862 ArchiveClassesAtExit not supported when base CDS archive is not loaded
3864 To run the Java application using a dynamic CDS archive, just add the
3866 -XX:SharedArchiveFile=<dynamic archive> option to the command line for
3867 running the Java application.
3868 The base archive is not required to be specified in the command line.
3870 The base archive information, including its name and full path, will be
3871 retrieved from the dynamic archive header. Note that the user could
3872 also use the -XX:SharedArchiveFile option for specifying a regular Ap‐
3873 pCDS archive. Therefore, the specified archive in the
3874 -XX:SharedArchiveFile option could be either a regular or dynamic ar‐
3875 chive. During VM start up the specified archive header will be read.
3876 If -XX:SharedArchiveFile refers to a regular archive, then the behavior
3877 will be unchanged. If -XX:SharedArchiveFile refers to a dynamic ar‐
3878 chive, the VM will retrieve the base archive location from the dynamic
3879 archive. If the dynamic archive was created with the default CDS ar‐
3880 chive, then the current default CDS archive will be used, and will be
3881 found relative to the current run time environment.
3882 Please refer to JDK-8221706 [https://bugs.openjdk.ja‐
3884 va.net/browse/JDK-8221706] for details on error checking during dynamic
3885 CDS archive dump time and run time.
3886 Creating a Shared Archive File and Using It to Run an Application
3888 AppCDS archive
3889 The following steps create a shared archive file that contains all the
3890 classes used by the test.Hello application. The last step runs the ap‐
3891 plication with the shared archive file.
3892 1. Create a list of all classes used by the test.Hello application.
3894 The following command creates a file named hello.classlist that con‐
3895 tains a list of all classes used by this application:
3896 java -Xshare:off -XX:DumpLoadedClassList=hel‐
3898 lo.classlist -cp hello.jar test.Hello
3899 Note that the classpath specified by the -cp parameter must contain
3901 only JAR files.
3902 2. Create a shared archive, named hello.jsa, that contains all the
3904 classes in hello.classlist:
3905 java -Xshare:dump -XX:SharedArchiveFile=hel‐
3907 lo.jsa -XX:SharedClassListFile=hello.classlist -cp hello.jar
3908 Note that the classpath used at archive creation time must be the
3910 same as (or a prefix of) the classpath used at run time.
3911 3. Run the application test.Hello with the shared archive hello.jsa:
3913 java -XX:SharedArchiveFile=hello.jsa -cp hello.jar test.Hel‐
3915 lo
3916 4. Optional Verify that the test.Hello application is using the class
3918 contained in the hello.jsa shared archive:
3919 java -XX:SharedArchiveFile=hello.jsa -cp hello.jar -ver‐
3921 bose:class test.Hello
3922 The output of this command should contain the following text:
3924 Loaded test.Hello from shared objects file by sun/misc/Launcher$AppClassLoader
3926 Dynamic CDS archive
3928 The following steps create a dynamic CDS archive file that contains the
3929 classes used by the test.Hello application and are not included in the
3930 default CDS archive. The second step runs the application with the dy‐
3931 namic CDS archive.
3932 1. Create a dynamic CDS archive, named hello.jsa, that contains all the
3934 classes in hello.jar loaded by the application test.Hello:
3935 java -XX:ArchiveClassesAtExit=hello.jsa -cp hello.jar Hello
3937 Note that the classpath used at archive creation time must be the
3939 same as (or a prefix of) the classpath used at run time.
3940 2. Run the application test.Hello with the shared archive hello.jsa:
3942 java -XX:SharedArchiveFile=hello.jsa -cp hello.jar test.Hel‐
3944 lo
3945 3. Optional Repeat step 4 of the previous section to verify that the
3947 test.Hello application is using the class contained in the hello.jsa
3948 shared archive.
3949 To automate the above steps 1 and 2, one can write a script such as the
3951 following:
3952 ARCHIVE=hello.jsa
3954 if test -f $ARCHIVE; then
3955 FLAG="-XX:SharedArchiveFile=$ARCHIVE"
3956 else
3957 FLAG="-XX:ArchiveClassesAtExit=$ARCHIVE"
3958 fi
3959 $JAVA_HOME/bin/java -cp hello.jar $FLAG test.Hello
3960 Like an AppCDS archive, the archive needs to be re-generated if the Ja‐
3962 va version has changed. The above script could be adjusted to account
3963 for the Java version as follows:
3964 ARCHIVE=hello.jsa
3966 VERSION=foo.version
3967 if test -f $ARCHIVE -a -f $VERSION && cmp -s $VERSION $JAVA_HOME/release; then
3968 FLAG="-XX:SharedArchiveFile=$ARCHIVE"
3969 else
3970 FLAG="-XX:ArchiveClassesAtExit=$ARCHIVE"
3971 cp -f $JAVA_HOME/release $VERSION
3972 fi
3973 $JAVA_HOME/bin/java -cp hello.jar $FLAG test.Hello
3974 Currently, we don't support concurrent dumping operations to the same
3976 CDS archive. Care should be taken to avoid multiple writers to the
3977 same CDS archive.
3978 The user could also create a dynamic CDS archive with a specific base
3980 archive, e.g. named as base.jsa as follows:
3981 java -XX:SharedArchiveFile=base.jsa -XX:ArchiveClassesAtEx‐
3983 it=hello.jsa -cp hello.jar Hello
3984 To run the application using the dynamic CDS archive hello.jsa and a
3986 specific base CDS archive base.jsa:
3987 java -XX:SharedArchiveFile=base.jsa:hello.jsa -cp hello.jar Hel‐
3989 lo
3990 Note that on Windows, the above path delimiter : should be replaced
3992 with ;.
3993 The above command for specifying a base archive is useful if the base
3995 archive used for creating the dynamic archive has been moved. Normal‐
3996 ly, just specifying the dynamic archive should be sufficient since the
3997 base archive info can be retrieved from the dynamic archive header.
3998 Sharing a Shared Archive Across Multiple Application Processes
4000 You can share the same archive file across multiple applications pro‐
4001 cesses. This reduces memory usage because the archive is memory-mapped
4002 into the address space of the processes. The operating system automat‐
4003 ically shares the read-only pages across these processes.
4004 The following steps demonstrate how to create a common archive that can
4006 be shared by different applications. Classes from common.jar, hel‐
4007 lo.jar and hi.jar are archived in the common.jsa because they are all
4008 in the classpath during the archiving step (step 3).
4009 To include classes from hello.jar and hi.jar, the .jar files must be
4011 added to the classpath specified by the -cp parameter.
4012 1. Create a list of all classes used by the Hello application and an‐
4014 other list for the Hi application:
4015 java -XX:DumpLoadedClassList=hello.classlist -cp com‐
4017 mon.jar:hello.jar Hello
4018 java -XX:DumpLoadedClassList=hi.classlist -cp com‐
4020 mon.jar:hi.jar Hi
4021 2. Create a single list of classes used by all the applications that
4023 will share the shared archive file.
4024 Oracle Solaris, Linux, and macOS The following commands combine the
4026 files hello.classlist and hi.classlist into one file, com‐
4027 mon.classlist:
4028 cat hello.classlist hi.classlist > common.classlist
4030 Windows The following commands combine the files hello.classlist
4032 and hi.classlist into one file, common.classlist:
4033 type hello.classlist hi.classlist > common.classlist
4035 3. Create a shared archive named common.jsa that contains all the
4037 classes in common.classlist:
4038 java -Xshare:dump -XX:SharedArchiveFile=com‐
4040 mon.jsa -XX:SharedClassListFile=common.classlist -cp com‐
4041 mon.jar:hello.jar:hi.jar
4042 The classpath parameter used is the common class path prefix shared
4044 by the Hello and Hi applications.
4045 4. Run the Hello and Hi applications with the same shared archive:
4047 java -XX:SharedArchiveFile=common.jsa -cp common.jar:hel‐
4049 lo.jar:hi.jar Hello
4050 java -XX:SharedArchiveFile=common.jsa -cp common.jar:hel‐
4052 lo.jar:hi.jar Hi
4053 Specifying Additional Shared Data Added to an Archive File
4055 The SharedArchiveConfigFile option is used to specify additional shared
4056 data to add to the archive file.
4057 -XX:SharedArchiveConfigFile=shared_config_file
4059 JDK 9 and later supports adding both symbols and string objects to an
4061 archive for memory sharing when you have multiple JVM processes running
4062 on the same host. An example of this is having multiple JVM processes
4063 that use the same set of Java EE classes. When these common classes
4064 are loaded and used, new symbols and strings may be created and added
4065 to the JVM's internal "symbol" and "string" tables. At runtime, the
4066 symbols or string objects mapped from the archive file can be shared
4067 across multiple JVM processes, resulting in a reduction of overall mem‐
4068 ory usage. In addition, archiving strings also provides added perfor‐
4069 mance benefits in both startup time and runtime execution.
4070 In JDK 10 and later, CONSTANT_String entries in archived classes are
4072 resolved to interned String objects at dump time, and all interned
4073 String objects are archived. However, even though all CONSTANT_String
4074 literals in all archived classes are resolved, it might still benefi‐
4075 cial to add additional strings that are not string literals in class
4076 files, but are likely to be used by your application at run time.
4077 Symbol data should be generated by the jcmd tool attaching to a running
4079 JVM process. See jcmd.
4080 The following is an example of the symbol dumping command in jcmd:
4082 jcmd pid VM.symboltable -verbose
4084 Note: The first line (process ID) and the second line (@VER‐
4086 SION ...) of this jcmd output should be excluded from the con‐
4087 figuration file.
4088 Example of a Configuration File
4090 The following is an example of a configuration file:
4091 VERSION: 1.0
4093 @SECTION: Symbol
4094 10 -1: linkMethod
4095 In the configuration file example, the @SECTION: Symbol entry uses the
4097 following format:
4098 length refcount: symbol
4100 The refcount for a shared symbol is always -1.
4102 @SECTION specifies the type of the section that follows it. All data
4104 within the section must be the same type that's specified by @SECTION.
4105 Different types of data can't be mixed. Multiple separated data sec‐
4106 tions for the same type specified by different @SECTION are allowed
4107 within one shared_config_file .
4108
4110 You can use the Java advanced runtime options to optimize the perfor‐
4111 mance of your applications.
4112 Tuning for Higher Throughput
4114 Use the following commands and advanced options to achieve higher
4115 throughput performance for your application:
4116 java -server -XX:+UseParallelGC -XX:+Use‐
4118 LargePages -Xmn10g -Xms26g -Xmx26g
4119 Tuning for Lower Response Time
4121 Use the following commands and advanced options to achieve lower re‐
4122 sponse times for your application:
4123 java -XX:+UseG1GC -XX:MaxGCPauseMillis=100
4125 Keeping the Java Heap Small and Reducing the Dynamic Footprint of
4127 Embedded Applications
4128 Use the following advanced runtime options to keep the Java heap small
4130 and reduce the dynamic footprint of embedded applications:
4131 -XX:MaxHeapFreeRatio=10 -XX:MinHeapFreeRatio=5
4133 Note: The defaults for these two options are 70% and 40% respec‐
4135 tively. Because performance sacrifices can occur when using
4136 these small settings, you should optimize for a small footprint
4137 by reducing these settings as much as possible without introduc‐
4138 ing unacceptable performance degradation.
4139
4141 The following exit values are typically returned by the launcher when
4142 the launcher is called with the wrong arguments, serious errors, or ex‐
4143 ceptions thrown by the JVM. However, a Java application may choose to
4144 return any value by using the API call System.exit(exitValue). The
4145 values are:
4146 · 0: Successful completion
4148 · >0: An error occurred
4150JDK 14 2020 JAVA(1)