aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorAndrew Sun <as-com@users.noreply.github.com>2024-01-07 04:00:58 -0500
committerGitHub <noreply@github.com>2024-01-07 10:00:58 +0100
commitaa0395d01bd800d82dc83b39df27c688e1f03363 (patch)
tree99be90d9accb9906c0fd22eb1e6af203c0ba998d
parent3ebc7d2b9c958ed951457f76dcd5651bdc3ef724 (diff)
Add entry by Andrew Sun
-rwxr-xr-xcalculate_average_asun.sh23
-rw-r--r--src/main/java/dev/morling/onebrc/CalculateAverage_asun.java329
2 files changed, 352 insertions, 0 deletions
diff --git a/calculate_average_asun.sh b/calculate_average_asun.sh
new file mode 100755
index 0000000..8ed6f7e
--- /dev/null
+++ b/calculate_average_asun.sh
@@ -0,0 +1,23 @@
+#!/bin/sh
+#
+# Copyright 2023 The original authors
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+# Uncomment below to use sdk
+# source "$HOME/.sdkman/bin/sdkman-init.sh"
+# sdk use java 21.0.1-graal 1>&2
+
+JAVA_OPTS="--enable-preview --add-modules jdk.incubator.vector -XX:+UnlockExperimentalVMOptions -Xms500m -Xmx500m -XX:CompilationMode=high-only"
+time java $JAVA_OPTS --class-path target/average-1.0.0-SNAPSHOT.jar dev.morling.onebrc.CalculateAverage_asun
diff --git a/src/main/java/dev/morling/onebrc/CalculateAverage_asun.java b/src/main/java/dev/morling/onebrc/CalculateAverage_asun.java
new file mode 100644
index 0000000..88a90ea
--- /dev/null
+++ b/src/main/java/dev/morling/onebrc/CalculateAverage_asun.java
@@ -0,0 +1,329 @@
+/*
+ * Copyright 2023 The original authors
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package dev.morling.onebrc;
+
+import jdk.incubator.vector.*;
+
+import java.io.File;
+import java.io.IOException;
+import java.io.RandomAccessFile;
+import java.lang.foreign.MemorySegment;
+import java.lang.foreign.ValueLayout;
+import java.nio.ByteOrder;
+import java.nio.channels.FileChannel;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.StandardOpenOption;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.List;
+import java.util.TreeMap;
+import java.util.concurrent.ExecutionException;
+import java.util.stream.Collectors;
+
+// based on spullara's submission
+
+public class CalculateAverage_asun {
+ private static final String FILE = "./measurements.txt";
+
+ private static final VectorSpecies<Byte> BYTE_SPECIES = ByteVector.SPECIES_256;
+ private static final VectorSpecies<Integer> INT_SPECIES = IntVector.SPECIES_256;
+ private static final int VECTOR_SIZE = 32;
+
+ private static final ByteVector ASC;
+ static {
+ byte[] bytes = new byte[VECTOR_SIZE];
+ for (int i = 0; i < VECTOR_SIZE; i++) {
+ bytes[i] = (byte) i;
+ }
+
+ ASC = ByteVector.fromArray(BYTE_SPECIES, bytes, 0);
+ }
+
+ public static void main(String[] args) throws IOException, ExecutionException, InterruptedException {
+ long start = System.currentTimeMillis();
+ var filename = args.length == 0 ? FILE : args[0];
+ var file = new File(filename);
+
+ List<FileSegment> fileSegments = getFileSegments(file);
+ // System.out.println(System.currentTimeMillis() - start);
+ var resultsMap = fileSegments.stream().map(segment -> {
+ var resultMap = new ByteArrayToResultMap();
+ long segmentEnd = segment.end();
+ try (var fileChannel = (FileChannel) Files.newByteChannel(Path.of(filename), StandardOpenOption.READ)) {
+ var bb = fileChannel.map(FileChannel.MapMode.READ_ONLY, segment.start(), segmentEnd - segment.start());
+ var ms = MemorySegment.ofBuffer(bb);
+
+ // Up to 100 characters for a city name
+ var buffer = new byte[100 + VECTOR_SIZE];
+ long startLine;
+ long pos = 0;
+ long limit = ms.byteSize();
+ long vectorLimit = limit - VECTOR_SIZE;
+
+ // int[] lastHashMult = new int[]{ 7, 31, 63, 15, 255, 127, 3, 511 };
+ // IntVector lastMul = IntVector.fromArray(INT_SPECIES, lastHashMult, 0);
+
+ vector: while ((startLine = pos) < vectorLimit) {
+ long currentPosition = startLine;
+ ByteVector r;
+ VectorMask<Byte> m;
+ int offset = 0;
+
+ IntVector h = IntVector.zero(INT_SPECIES);
+ while (true) {
+ if (currentPosition >= vectorLimit) {
+ break vector;
+ }
+
+ r = ByteVector.fromMemorySegment(BYTE_SPECIES, ms, currentPosition, ByteOrder.LITTLE_ENDIAN);
+ r.intoArray(buffer, offset);
+ offset += VECTOR_SIZE;
+ m = r.eq((byte) ';');
+ if (m.anyTrue()) {
+ int firstTrue = m.firstTrue();
+ currentPosition += firstTrue;
+ // note: target platform likely does not have AVX-512, so manipulating and using m directly is likely to be slow
+ ByteVector lastMask = (ByteVector) ASC.lt((byte) firstTrue).toVector();
+ h = h.mul(31);
+ h = h.add(r.and(lastMask).reinterpretAsInts());
+ break;
+ }
+ else {
+ currentPosition += VECTOR_SIZE;
+ h = h.mul(31);
+ h = h.add(r.reinterpretAsInts());
+ }
+ }
+
+ // h = h.mul(lastMul);
+
+ int hash = h.reduceLanes(VectorOperators.ADD);
+
+ // currentPosition now has index of semicolon
+ int nameLen = (int) (currentPosition - startLine);
+ currentPosition++;
+
+ if (currentPosition >= limit - 8) {
+ break;
+ }
+
+ long g = ms.get(ValueLayout.JAVA_LONG_UNALIGNED, currentPosition);
+ int negative = (g & 0xff) == '-' ? -1 : 1;
+
+ // 00101101 MINUS
+ // 00101110 PERIOD
+ // 00001101 CR
+ // 00001010 LF
+ // 00110000 0
+ // 00111001 9
+
+ // scan for LF
+ long lf = ~g & 0x20202020202020L;
+ int tzc = Long.numberOfTrailingZeros(lf);
+ long bytesToLF = tzc / 8;
+
+ int shift = 64 - tzc & 0b111000;
+
+ long reversedDigits = Long.reverseBytes(g) >> shift;
+ long digitBits = reversedDigits & (0x1010101010101010L >> shift);
+ long digitsExt = (digitBits >> 1 | digitBits >> 2 | digitBits >> 3 | digitBits >> 4);
+
+ long digitsOnly = Long.compress(reversedDigits, digitsExt);
+
+ long temp = (digitsOnly & 0xf)
+ + 10 * ((digitsOnly >> 4) & 0xf)
+ + 100 * ((digitsOnly >> 8) & 0xf);
+
+ temp *= negative;
+
+ currentPosition += bytesToLF + 1;
+
+ resultMap.putOrMerge(buffer, 0, nameLen, temp, hash);
+ pos = currentPosition;
+
+ }
+
+ while ((startLine = pos) < limit) {
+ long currentPosition = startLine;
+ byte b;
+ int offset = 0;
+
+ while (currentPosition != segmentEnd && (b = ms.get(ValueLayout.JAVA_BYTE, currentPosition++)) != ';') {
+ buffer[offset++] = b;
+ }
+ // Invariant: the remaining length is less than VECTOR_SIZE, so we can just run the last round of hashing
+ int hash = ByteVector.fromArray(BYTE_SPECIES, buffer, 0, ASC.lt((byte) offset))
+ .reinterpretAsInts()
+ // .mul(lastMul)
+ .reduceLanes(VectorOperators.ADD);
+
+ int temp;
+ int negative = 1;
+ // Inspired by @yemreinci to unroll this even further
+ if (ms.get(ValueLayout.JAVA_BYTE, currentPosition) == '-') {
+ negative = -1;
+ currentPosition++;
+ }
+ if (ms.get(ValueLayout.JAVA_BYTE, currentPosition + 1) == '.') {
+ temp = negative * ((ms.get(ValueLayout.JAVA_BYTE, currentPosition) - '0') * 10 + (ms.get(ValueLayout.JAVA_BYTE, currentPosition + 2) - '0'));
+ currentPosition += 3;
+ }
+ else {
+ temp = negative * ((ms.get(ValueLayout.JAVA_BYTE, currentPosition) - '0') * 100
+ + ((ms.get(ValueLayout.JAVA_BYTE, currentPosition + 1) - '0') * 10 + (ms.get(ValueLayout.JAVA_BYTE, currentPosition + 3) - '0')));
+ currentPosition += 4;
+ }
+ if (ms.get(ValueLayout.JAVA_BYTE, currentPosition) == '\r') {
+ currentPosition++;
+ }
+ currentPosition++;
+ resultMap.putOrMerge(buffer, 0, offset, temp, hash);
+ pos = currentPosition;
+ }
+ return resultMap;
+ }
+ catch (IOException e) {
+ throw new RuntimeException(e);
+ }
+ }).parallel().flatMap(partition -> partition.getAll().stream())
+ .collect(Collectors.toMap(e -> new String(e.key()), Entry::value, CalculateAverage_asun::merge, TreeMap::new));
+
+ System.out.println(resultsMap);
+
+ // System.out.println(System.currentTimeMillis() - start);
+
+ Runtime.getRuntime().halt(0);
+ }
+
+ private static List<FileSegment> getFileSegments(File file) throws IOException {
+ int numberOfSegments = Runtime.getRuntime().availableProcessors() * 8;
+ long fileSize = file.length();
+ long segmentSize = fileSize / numberOfSegments;
+ List<FileSegment> segments = new ArrayList<>(numberOfSegments);
+ // Pointless to split small files
+ if (segmentSize < 1_000_000) {
+ segments.add(new FileSegment(0, fileSize));
+ return segments;
+ }
+ try (RandomAccessFile randomAccessFile = new RandomAccessFile(file, "r")) {
+ for (int i = 0; i < numberOfSegments; i++) {
+ long segStart = i * segmentSize;
+ long segEnd = (i == numberOfSegments - 1) ? fileSize : segStart + segmentSize;
+ segStart = findSegment(i, 0, randomAccessFile, segStart, segEnd);
+ segEnd = findSegment(i, numberOfSegments - 1, randomAccessFile, segEnd, fileSize);
+
+ segments.add(new FileSegment(segStart, segEnd));
+ }
+ }
+ return segments;
+ }
+
+ private static Result merge(Result v, Result value) {
+ return merge(v, value.min, value.max, value.sum, value.count);
+ }
+
+ private static Result merge(Result v, long value, long value1, long value2, long value3) {
+ v.min = Math.min(v.min, value);
+ v.max = Math.max(v.max, value1);
+ v.sum += value2;
+ v.count += value3;
+ return v;
+ }
+
+ private static long findSegment(int i, int skipSegment, RandomAccessFile raf, long location, long fileSize) throws IOException {
+ if (i != skipSegment) {
+ raf.seek(location);
+ while (location < fileSize) {
+ location++;
+ if (raf.read() == '\n')
+ break;
+ }
+ }
+ return location;
+ }
+
+ static class Result {
+ long min, max, sum;
+ long count;
+
+ Result(long value) {
+ min = max = sum = value;
+ this.count = 1;
+ }
+
+ @Override
+ public String toString() {
+ return round(min) + "/" + round((sum / 10.0) / count) + "/" + round(max);
+ }
+
+ double round(double v) {
+ return Math.round(v * 10.0) / 10.0;
+ }
+
+ double round(long v) {
+ return v / 10.0;
+ }
+
+ }
+
+ record Entry(byte[] key, Result value) {
+ }
+
+ record FileSegment(long start, long end) {
+ }
+
+ static class ByteArrayToResultMap {
+ public static final int MAPSIZE = 1024 * 128;
+ Result[] slots = new Result[MAPSIZE];
+ byte[][] keys = new byte[MAPSIZE][];
+
+ public void putOrMerge(byte[] key, int offset, int size, long temp, int hash) {
+ int slot = hash & (slots.length - 1);
+ var slotValue = slots[slot];
+ // Linear probe for open slot
+ while (slotValue != null && (keys[slot].length != size || !Arrays.equals(keys[slot], 0, size, key, offset, size))) {
+ slot = (slot + 1) & (slots.length - 1);
+ slotValue = slots[slot];
+ }
+ Result value = slotValue;
+ if (value == null) {
+ slots[slot] = new Result(temp);
+ byte[] bytes = new byte[size];
+ System.arraycopy(key, offset, bytes, 0, size);
+ keys[slot] = bytes;
+ }
+ else {
+ value.min = Math.min(value.min, temp);
+ value.max = Math.max(value.max, temp);
+ value.sum += temp;
+ value.count += 1;
+ }
+ }
+
+ // Get all pairs
+ public List<Entry> getAll() {
+ List<Entry> result = new ArrayList<>(slots.length);
+ for (int i = 0; i < slots.length; i++) {
+ Result slotValue = slots[i];
+ if (slotValue != null) {
+ result.add(new Entry(keys[i], slotValue));
+ }
+ }
+ return result;
+ }
+ }
+}