jmh

问题 I am using the JMH benchmarking framework (http://openjdk.java.net/projects/code-tools/jmh/) to run benchmarks on my code. My understanding is that JMH forks the JVM multiple times during benchmarking in order to discard any profiles built up by the just-in-time (JIT) profiling performed by the JVM during execution. I understand why this is useful in some cases such as the below (copied verbatim from http://java-performance.info/jmh/): By default JHM forks a new java process for each trial

问题 I've seen a lot of threads here that compare and try to answer which is faster: newInstance or new operator . Looking at the source code, it would seem that newInstance should be much slower , I mean it does so many security checks and uses reflection. And I've decided to measure, first running jdk-8. Here is the code using jmh . @BenchmarkMode(value = { Mode.AverageTime, Mode.SingleShotTime }) @Warmup(iterations = 5, time = 2, timeUnit = TimeUnit.SECONDS) @Measurement(iterations = 5, time =

问题 I am using Java 8 streams to iterate over a list with sublists. The outer list size varies between 100 to 1000 (different test runs) and the inner list size is always 5. There are 2 benchmark runs which show unexpected performance deviations. package benchmark; import org.openjdk.jmh.annotations.*; import org.openjdk.jmh.infra.Blackhole; import java.io.IOException; import java.util.concurrent.ThreadLocalRandom; import java.util.*; import java.util.function.*; import java.util.stream.*;

问题 Having a Kotlin project with Gradle setup: apply plugin: 'kotlin' apply plugin: 'kotlin-kapt' dependencies { kapt 'org.openjdk.jmh:jmh-generator-annprocess:1.18' ... } Putting benchmarks under src/main/kotlin works without problems. But when i add a custom source-set for JMH: sourceSets { jmh { compileClasspath += sourceSets.test.runtimeClasspath runtimeClasspath += sourceSets.test.runtimeClasspath } } And move the benchmarks from src/main/kotlin to src/jmh/kotlin , executing the benchmarks

问题 I am learning how to microbenchmark things with JMH. I started with something seemingly simple: string concatenation for StringBuilder vs String += . From my understanding, I should make a State object that contains an instance of StringBuilder because I don't want to benchmark its constructor (nor do I want to an empty one every iteration anyway). Same goes for the String += test - I want a String object in my State to be concatenated with new strings. This is my code: @State(Scope.Thread)

问题 I wanted to do some performance measurements and comparisons of simple for loops and equivalent streams implementations. I believe it's the case that streams will be somewhat slower than equivalent non-streams code, but I wanted to be sure I'm measuring the right things. I'm including my entire jmh class here. import java.util.ArrayList; import java.util.List; import org.openjdk.jmh.annotations.Benchmark; import org.openjdk.jmh.annotations.Scope; import org.openjdk.jmh.annotations.Setup;

问题 I have a folder with structure src/org/openjdk/jmh and in this I have various files such as annotations, generators, infra, profile, results, runner, util. All these files have various java programs in them . Now in src/com/user directory I have a benchmark code "JMHSortBenchmark.java" which uses the import org.openjdk.jmh.annotations.*; import org.openjdk.jmh.runner.*; But when I compile the JMHSortBenchmark.java, I get an error as package org.openjdk.jmh.annotations does not exist package

问题 It is well known that using a loop inside your JMH benchmark is not a good idea because it will be optimized by the JIT compiler and should therefore be avoided. Is there a way to feed my JMH benchmark methods with different values of int inputs (list of inputs) without using a loop. 回答1: Have a look at this example in the JMH documentation. You can use the @Param annotation on a field in order to tell JMH to inject the values of this annotation: @Param({"1", "2"}) public int arg; @Benchmark

问题 I'm playing with jmh and in the section about looping they said that You might notice the larger the repetitions count, the lower the "perceived" cost of the operation being measured. Up to the point we do each addition with 1/20 ns, well beyond what hardware can actually do. This happens because the loop is heavily unrolled/pipelined , and the operation to be measured is hoisted from the loop . Morale: don't overuse loops, rely on JMH to get the measurement right. I tried it myself

问题 I use JMH to specify the complexity of the operation. If you've never worked with JMH, don't worry. JMH will just launch the estimateOperation method multiple times and then get the average time. Question: [narrow] will this program calculate Math.cbrt(Integer.MAX_VALUE) each time? Or it just calculate it once and return cached result afterwards? @GenerateMicroBenchmark public void estimateOperation() { calculate(); } public int calculate() { return Math.cbrt(Integer.MAX_VALUE); } Question: