8282047: Enhance StringDecode/Encode microbenchmarks

Backport-of: d4d12ad1d92ae6d353ce9e16b1b72e6416874cd6

Author: GoeLin

test/micro/org/openjdk/bench/java/lang/StringDecode.java

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2021, 2022, Oracle and/or its affiliates. All rights reserved.
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
  * This code is free software; you can redistribute it and/or modify it
@@ -22,46 +22,38 @@
  */
 package org.openjdk.bench.java.lang;
 
-import org.openjdk.jmh.annotations.Benchmark;
-import org.openjdk.jmh.annotations.BenchmarkMode;
-import org.openjdk.jmh.annotations.Fork;
-import org.openjdk.jmh.annotations.Measurement;
-import org.openjdk.jmh.annotations.Mode;
-import org.openjdk.jmh.annotations.OutputTimeUnit;
-import org.openjdk.jmh.annotations.Param;
-import org.openjdk.jmh.annotations.Scope;
-import org.openjdk.jmh.annotations.Setup;
-import org.openjdk.jmh.annotations.State;
-import org.openjdk.jmh.annotations.Warmup;
+import org.openjdk.jmh.annotations.*;
 import org.openjdk.jmh.infra.Blackhole;
 
 import java.nio.charset.Charset;
 import java.util.concurrent.TimeUnit;
 
 @BenchmarkMode(Mode.AverageTime)
 @OutputTimeUnit(TimeUnit.NANOSECONDS)
-@Fork(value = 3)
-@Warmup(iterations = 5, time = 2)
+@Fork(value = 2)
+@Warmup(iterations = 5, time = 3)
 @Measurement(iterations = 5, time = 3)
 @State(Scope.Thread)
 public class StringDecode {
 
-    @Param({"US-ASCII", "ISO-8859-1", "UTF-8", "MS932", "ISO-8859-6", "ISO-2022-KR"})
+    // Reduced by default to only UTF-8, previous coverage:
+    // @Param({"US-ASCII", "ISO-8859-1", "UTF-8", "MS932", "ISO-8859-6", "ISO-2022-KR"})
+    @Param({"UTF-8"})
     private String charsetName;
 
     private Charset charset;
     private byte[] asciiString;
+    private byte[] longAsciiString;
     private byte[] utf16String;
-    private byte[] longUtf16String;
+    private byte[] longUtf16EndString;
     private byte[] longUtf16StartString;
-    private byte[] longLatin1String;
+    private byte[] longUtf16OnlyString;
+    private byte[] latin1String;
+    private byte[] longLatin1EndString;
+    private byte[] longLatin1StartString;
+    private byte[] longLatin1OnlyString;
 
-    @Setup
-    public void setup() {
-        charset = Charset.forName(charsetName);
-        asciiString = "ascii string".getBytes(charset);
-        utf16String = "UTF-\uFF11\uFF16 string".getBytes(charset);
-        longUtf16String = """
+    private static final String LOREM = """
              Lorem ipsum dolor sit amet, consectetur adipiscing elit. Aliquam ac sem eu
              urna egestas placerat. Etiam finibus ipsum nulla, non mattis dolor cursus a.
              Nulla nec nisl consectetur, lacinia neque id, accumsan ante. Curabitur et
@@ -70,90 +62,145 @@ public void setup() {
              per conubia nostra, per inceptos himenaeos. Suspendisse erat diam, fringilla
              sed massa sed, posuere viverra orci. Suspendisse tempor libero non gravida
              efficitur. Vivamus lacinia risus non orci viverra, at consectetur odio laoreet.
-             Suspendisse potenti.
-
-             Phasellus vel nisi iaculis, accumsan quam sed, bibendum eros. Sed venenatis
-             nulla tortor, et eleifend urna sodales id. Nullam tempus ac metus sit amet
-             sollicitudin. Nam sed ex diam. Praesent vitae eros et neque condimentum
-             consectetur eget non tortor. Praesent bibendum vel felis nec dignissim.
-             Maecenas a enim diam. Suspendisse quis ligula at nisi accumsan lacinia id
-             hendrerit sapien. \uFF11Donec aliquam mattis lectus eu ultrices. Duis eu nisl\uFF11
-             euismod, blandit mauris vel, \uFF11placerat urna. Etiam malesuada enim purus,
-             tristique mollis odio blandit quis.\uFF11 Vivamus posuere. \uFF11
-             \uFF11
-            """.getBytes(charset);
-        longUtf16StartString = """
-             \uFF11
-             Lorem ipsum dolor sit amet, \uFF11consectetur adipiscing elit. Aliquam ac sem eu
-             urna egestas \uFF11placerat. Etiam finibus ipsum nulla, non mattis dolor cursus a.
-             Nulla \uFF11nec nisl consectetur, lacinia neque id, accumsan ante. Curabitur et
-             sapien in \uFF11magna porta ultricies. \uFF11Sed vel pellentesque nibh. Pellentesque dictum
-             dignissim diam eu ultricies. Class aptent taciti sociosqu ad litora torquent
-             per conubia nostra, per inceptos himenaeos. Suspendisse erat diam, fringilla
-             sed massa sed, posuere viverra orci. Suspendisse tempor libero non gravida
-             efficitur. Vivamus lacinia risus non orci viverra, at consectetur odio laoreet.
-             Suspendisse potenti.
-
-             Phasellus vel nisi iaculis, accumsan quam sed, bibendum eros. Sed venenatis
-             nulla tortor, et eleifend urna sodales id. Nullam tempus ac metus sit amet
-             sollicitudin. Nam sed ex diam. Praesent vitae eros et neque condimentum
-             consectetur eget non tortor. Praesent bibendum vel felis nec dignissim.
-             Maecenas a enim diam. Suspendisse quis ligula at nisi accumsan lacinia id
-             hendrerit sapien. Donec aliquam mattis lectus eu ultrices. Duis eu nisl
-             euismod, blandit mauris vel, placerat urna. Etiam malesuada enim purus,
-             tristique mollis odio blandit quis. Vivamus posuere.
-            """.getBytes(charset);
-
-        longLatin1String = """
-             a\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             b\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             c\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             d\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             e\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             f\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             g\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             h\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             i\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6
-             j\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6\u00F6
-             k\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6\u00F6
-             l\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6\u00F6
-             m\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00F6\u00B6\u00F6\u00F6
-            """.getBytes(charset);
+             Suspendisse potenti.""";
+    private static final String UTF16_STRING = "\uFF11".repeat(31);
+    private static final String LATIN1_STRING = "\u00B6".repeat(31);
+
+    @Setup
+    public void setup() {
+        charset = Charset.forName(charsetName);
+        asciiString = LOREM.substring(0, 32).getBytes(charset);
+        longAsciiString = LOREM.repeat(200).getBytes(charset);
+        utf16String = "UTF-\uFF11\uFF16 string".getBytes(charset);
+        longUtf16EndString = LOREM.repeat(4).concat(UTF16_STRING).getBytes(charset);
+        longUtf16StartString = UTF16_STRING.concat(LOREM.repeat(4)).getBytes(charset);
+        longUtf16OnlyString = UTF16_STRING.repeat(10).getBytes(charset);
+        latin1String = LATIN1_STRING.getBytes(charset);
+        longLatin1EndString = LOREM.repeat(4).concat(LATIN1_STRING).getBytes(charset);
+        longLatin1StartString = LATIN1_STRING.concat(LOREM.repeat(4)).getBytes(charset);
+        longLatin1OnlyString = LATIN1_STRING.repeat(10).getBytes(charset);
     }
 
     @Benchmark
-    public String decodeAsciiCharsetName() throws Exception {
-        return new String(asciiString, charsetName);
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeAsciiShort(Blackhole bh) throws Exception {
+        bh.consume(new String(asciiString, charset));
+        bh.consume(new String(longAsciiString, 0, 15, charset));
+        bh.consume(new String(asciiString, 0, 3, charset));
+        bh.consume(new String(longAsciiString, 512, 512 + 7, charset));
     }
 
     @Benchmark
-    public String decodeAscii() throws Exception {
-        return new String(asciiString, charset);
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeAsciiLong(Blackhole bh) throws Exception {
+        bh.consume(new String(longAsciiString, charset));
+        bh.consume(new String(longAsciiString, 0, 1024 + 31, charset));
     }
 
     @Benchmark
-    public String decodeLatin1Long() throws Exception {
-        return new String(longLatin1String, charset);
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeLatin1Short(Blackhole bh) throws Exception {
+        bh.consume(new String(latin1String, charset));
+        bh.consume(new String(latin1String, 0, 15, charset));
+        bh.consume(new String(latin1String, 0, 3, charset));
+        bh.consume(new String(longLatin1OnlyString, 512, 512 + 7, charset));
     }
 
     @Benchmark
-    public String decodeUTF16Short() throws Exception {
-        return new String(utf16String, charset);
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public String decodeLatin1LongStart() throws Exception {
+        return new String(longLatin1StartString, charset);
     }
 
     @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public String decodeLatin1LongEnd() throws Exception {
+        return new String(longLatin1EndString, charset);
+    }
+
+    @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public String decodeLatin1LongOnly() throws Exception {
+        return new String(longLatin1OnlyString, charset);
+    }
+
+    @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeLatin1Mixed(Blackhole bh) throws Exception {
+        bh.consume(new String(longLatin1EndString, charset));
+        bh.consume(new String(longLatin1StartString, charset));
+        bh.consume(new String(latin1String, charset));
+        bh.consume(new String(longLatin1OnlyString, charset));
+    }
+
+    @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeUTF16Short(Blackhole bh) throws Exception {
+        bh.consume(new String(utf16String, charset));
+        bh.consume(new String(utf16String, 0, 15, charset));
+        bh.consume(new String(utf16String, 0, 3, charset));
+        bh.consume(new String(utf16String, 0, 7, charset));
+    }
+
+    @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
     public String decodeUTF16LongEnd() throws Exception {
-        return new String(longUtf16String, charset);
+        return new String(longUtf16EndString, charset);
     }
 
     @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
     public String decodeUTF16LongStart() throws Exception {
         return new String(longUtf16StartString, charset);
     }
 
     @Benchmark
-    public void decodeUTF16LongMixed(Blackhole bh) throws Exception {
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public String decodeUTF16LongOnly() throws Exception {
+        return new String(longUtf16OnlyString, charset);
+    }
+
+    @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeUTF16Mixed(Blackhole bh) throws Exception {
+        bh.consume(new String(longUtf16StartString, charset));
+        bh.consume(new String(longUtf16EndString, charset));
+        bh.consume(new String(utf16String, charset));
+        bh.consume(new String(longUtf16OnlyString, charset));
+    }
+
+    @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeAllMixed(Blackhole bh) throws Exception {
+        bh.consume(new String(utf16String, charset));
+        bh.consume(new String(longUtf16EndString, charset));
+        bh.consume(new String(utf16String, 0, 15, charset));
         bh.consume(new String(longUtf16StartString, charset));
-        bh.consume(new String(longUtf16String, charset));
+        bh.consume(new String(asciiString, 0, 3, charset));
+        bh.consume(new String(longUtf16OnlyString, charset));
+        bh.consume(new String(latin1String, charset));
+        bh.consume(new String(longLatin1EndString, charset));
+        bh.consume(new String(longLatin1StartString, charset));
+        bh.consume(new String(latin1String, 0, 7, charset));
+        bh.consume(new String(longLatin1OnlyString, charset));
+        bh.consume(new String(asciiString, charset));
+        bh.consume(new String(longAsciiString, charset));
+    }
+
+    @Benchmark
+    @CompilerControl(CompilerControl.Mode.DONT_INLINE)
+    public void decodeShortMixed(Blackhole bh) throws Exception {
+        bh.consume(new String(utf16String, 0, 15, charset));
+        bh.consume(new String(latin1String, 0, 15, charset));
+        bh.consume(new String(asciiString, charset));
+        bh.consume(new String(utf16String, charset));
+        bh.consume(new String(latin1String, 0, 3, charset));
+        bh.consume(new String(asciiString, 0, 3, charset));
+        bh.consume(new String(utf16String, 0, 7, charset));
+        bh.consume(new String(latin1String, charset));
+        bh.consume(new String(asciiString, 0, 7, charset));
+        bh.consume(new String(utf16String, 0, 3, charset));
+        bh.consume(new String(latin1String, 0, 7, charset));
+        bh.consume(new String(asciiString, 0, 15, charset));
     }
 }