8300493: Use ArraysSupport.vectorizedHashCode in j.u.zip.ZipCoder

Reviewed-by: alanb, lancea

Author: cl4es

src/java.base/share/classes/java/lang/System.java

@@ -2461,6 +2461,9 @@ public Stream<ModuleLayer> layers(ClassLoader loader) {
                 return ModuleLayer.layers(loader);
             }
 
+            public int countPositives(byte[] bytes, int offset, int length) {
+                return StringCoding.countPositives(bytes, offset, length);
+            }
             public String newStringNoRepl(byte[] bytes, Charset cs) throws CharacterCodingException  {
                 return String.newStringNoRepl(bytes, cs);
             }

src/java.base/share/classes/java/util/zip/ZipCoder.java

@@ -34,6 +34,7 @@
 import java.nio.charset.CodingErrorAction;
 import java.util.Arrays;
 
+import jdk.internal.util.ArraysSupport;
 import sun.nio.cs.UTF_8;
 
 /**
@@ -209,25 +210,18 @@ int checkedHash(byte[] a, int off, int len) throws Exception {
             if (len == 0) {
                 return 0;
             }
-
             int end = off + len;
-            int h = 0;
-            while (off < end) {
-                byte b = a[off];
-                if (b >= 0) {
-                    // ASCII, keep going
-                    h = 31 * h + b;
-                    off++;
-                } else {
-                    // Non-ASCII, fall back to decoding a String
-                    // We avoid using decoder() here since the UTF8ZipCoder is
-                    // shared and that decoder is not thread safe.
-                    // We use the JLA.newStringUTF8NoRepl variant to throw
-                    // exceptions eagerly when opening ZipFiles
-                    return hash(JLA.newStringUTF8NoRepl(a, end - len, len));
-                }
+            int asciiLen = JLA.countPositives(a, off, len);
+            if (asciiLen != len) {
+                // Non-ASCII, fall back to decoding a String
+                // We avoid using decoder() here since the UTF8ZipCoder is
+                // shared and that decoder is not thread safe.
+                // We use the JLA.newStringUTF8NoRepl variant to throw
+                // exceptions eagerly when opening ZipFiles
+                return hash(JLA.newStringUTF8NoRepl(a, off, len));
             }
-
+            // T_BOOLEAN to treat the array as unsigned bytes, in line with StringLatin1.hashCode
+            int h = ArraysSupport.vectorizedHashCode(a, off, len, 0, ArraysSupport.T_BOOLEAN);
             if (a[end - 1] != '/') {
                 h = 31 * h + '/';
             }

src/java.base/share/classes/jdk/internal/access/JavaLangAccess.java

@@ -302,6 +302,11 @@ public interface JavaLangAccess {
      */
     Stream<ModuleLayer> layers(ClassLoader loader);
 
+    /**
+     * Count the number of leading positive bytes in the range.
+     */
+    int countPositives(byte[] ba, int off, int len);
+
     /**
      * Constructs a new {@code String} by decoding the specified subarray of
      * bytes using the specified {@linkplain java.nio.charset.Charset charset}.

test/micro/org/openjdk/bench/java/util/zip/ZipFileOpen.java

@@ -0,0 +1,93 @@
+/*
+ * Copyright (c) 2023, Oracle and/or its affiliates. All rights reserved.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ * This code is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 only, as
+ * published by the Free Software Foundation.
+ *
+ * This code is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * version 2 for more details (a copy is included in the LICENSE file that
+ * accompanied this code).
+ *
+ * You should have received a copy of the GNU General Public License version
+ * 2 along with this work; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
+ * or visit www.oracle.com if you need additional information or have any
+ * questions.
+ */
+
+package org.openjdk.bench.java.util.zip;
+
+import org.openjdk.jmh.annotations.*;
+
+import java.io.File;
+import java.io.FileOutputStream;
+import java.io.IOException;
+import java.nio.file.Files;
+import java.util.concurrent.TimeUnit;
+import java.util.zip.ZipEntry;
+import java.util.zip.ZipFile;
+import java.util.zip.ZipOutputStream;
+
+/**
+ * Simple benchmark measuring cost of opening zip files, parsing CEN
+ * entries.
+ */
+@BenchmarkMode(Mode.AverageTime)
+@OutputTimeUnit(TimeUnit.NANOSECONDS)
+@State(Scope.Thread)
+@Warmup(iterations = 10, time = 500, timeUnit = TimeUnit.MILLISECONDS)
+@Measurement(iterations = 5, time = 1000, timeUnit = TimeUnit.MILLISECONDS)
+@Fork(3)
+public class ZipFileOpen {
+
+    @Param({"512", "1024"})
+    private int size;
+
+    public File zipFile;
+
+    @Setup(Level.Trial)
+    public void beforeRun() throws IOException {
+        // Create a test Zip file with the number of entries.
+        File tempFile = Files.createTempFile("zip-micro", ".zip").toFile();
+        tempFile.deleteOnExit();
+        try (FileOutputStream fos = new FileOutputStream(tempFile);
+             ZipOutputStream zos = new ZipOutputStream(fos)) {
+
+            // Vary dir and entry sizes, with a bias towards shorter entries.
+            String[] dirPrefixes = new String[] { "dir1", "dir2", "dir3",
+                    "longer-directory-name-", "ridiculously-long-pathname-to-help-exercize-vectorized-subroutines-"};
+            String[] entryPrefixes = new String[] { "e", "long-entry-name-",
+                    "ridiculously-long-entry-name-to-help-exercize-vectorized-subroutines-"};
+
+            for (int i = 0; i < size; i++) {
+                String ename = dirPrefixes[i % dirPrefixes.length] + i + "/";
+                zos.putNextEntry(new ZipEntry(ename));
+
+                ename += entryPrefixes[i % entryPrefixes.length] + "-" + i;
+                zos.putNextEntry(new ZipEntry(ename));
+            }
+        }
+        zipFile = tempFile;
+    }
+
+    @Benchmark
+    public ZipFile openCloseZipFile() throws Exception {
+        // Some shared resources in ZipFile are cached in a shared structure
+        // and needs to be cleaned up to properly capture overhead of creating
+        // a ZipFile - otherwise opening the same zip file again will reuse the
+        // cached data and look artificially fast. By including the ZipFile.close()
+        // we aggressively clear those resources pre-emptively. The operations
+        // appears to be complex enough to not be subject to DCE but care needs
+        // to be taken to check that things like initCEN is properly accounted
+        // for if/when the ZipFile setup improves.
+        ZipFile zf = new ZipFile(zipFile);
+        zf.close();
+        return zf;
+    }
+}