[
https://issues.apache.org/jira/browse/PARQUET-2159?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17681684#comment-17681684
]
ASF GitHub Bot commented on PARQUET-2159:
-----------------------------------------
wgtmac commented on code in PR #1011:
URL: https://github.com/apache/parquet-mr/pull/1011#discussion_r1089859464
##########
parquet-benchmarks/src/main/java/org/apache/parquet/benchmarks/ByteBitPackingVectorBenchmarks.java:
##########
@@ -0,0 +1,83 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.parquet.benchmarks;
+
+import org.apache.parquet.column.values.bitpacking.BytePacker;
+import org.apache.parquet.column.values.bitpacking.Packer;
+import org.openjdk.jmh.annotations.*;
Review Comment:
Please fix the import.
##########
parquet-column/src/main/java/org/apache/parquet/column/values/bitpacking/ParquetReadRouter.java:
##########
@@ -0,0 +1,141 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.parquet.column.values.bitpacking;
+
+import org.apache.parquet.bytes.ByteBufferInputStream;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.nio.ByteBuffer;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.nio.file.Paths;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+/**
+ * Utility class for big data applications (such as Apache Spark and Apache
Flink).
+ * For Intel CPU, Flags containing avx512vbmi and avx512_vbmi2 can have better
performance gains.
+ */
+public class ParquetReadRouter {
+ private static final Logger LOG =
LoggerFactory.getLogger(ParquetReadRouter.class);
+
+ private static final int BITS_PER_BYTE = 8;
+
+ // register of avx512 are 512 bits, and can load up to 64 bytes
+ private static final int BYTES_PER_512VECTOR = 64;
+
+ // values are bit packed 8 at a time, so reading bitWidth will always work
+ private static final int VALUES_PER_PACKED = 8;
Review Comment:
```suggestion
private static final int NUM_VALUES_TO_PACK = 8;
```
##########
parquet-column/src/test/java/org/apache/parquet/column/values/bitpacking/TestParquetReadRouter.java:
##########
@@ -0,0 +1,54 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.parquet.column.values.bitpacking;
+
+import org.apache.parquet.bytes.ByteBufferInputStream;
+import org.junit.Test;
+
+import java.io.IOException;
+import java.nio.ByteBuffer;
+
+import static org.junit.Assert.assertArrayEquals;
+
+public class TestParquetReadRouter {
+
+ /**
+ * The range of bitWidth is 1 ~ 32, change it directly if test other
bitWidth.
+ */
+ private static final int bitWidth = 7;
Review Comment:
It would be good to add cases of different bit widths.
##########
parquet-generator/src/main/resources/ByteBitPacking512VectorLE:
##########
@@ -0,0 +1,3095 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.parquet.column.values.bitpacking;
+
+import jdk.incubator.vector.ByteVector;
+import jdk.incubator.vector.IntVector;
+import jdk.incubator.vector.LongVector;
+import jdk.incubator.vector.ShortVector;
+import jdk.incubator.vector.Vector;
+import jdk.incubator.vector.VectorMask;
+import jdk.incubator.vector.VectorOperators;
+import jdk.incubator.vector.VectorShuffle;
+import jdk.incubator.vector.VectorSpecies;
+
+import java.nio.ByteBuffer;
+
+/**
+ * This is an auto-generated source file and should not edit it directly.
+ */
+public abstract class ByteBitPacking512VectorLE {
Review Comment:
Should we remove this generated file from git? I assume it will be
automatically created at maven build time, like other generators do.
##########
parquet-column/src/main/java/org/apache/parquet/column/values/bitpacking/ParquetReadRouter.java:
##########
@@ -0,0 +1,141 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.parquet.column.values.bitpacking;
+
+import org.apache.parquet.bytes.ByteBufferInputStream;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.nio.ByteBuffer;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.nio.file.Paths;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+/**
+ * Utility class for big data applications (such as Apache Spark and Apache
Flink).
+ * For Intel CPU, Flags containing avx512vbmi and avx512_vbmi2 can have better
performance gains.
+ */
+public class ParquetReadRouter {
+ private static final Logger LOG =
LoggerFactory.getLogger(ParquetReadRouter.class);
+
+ private static final int BITS_PER_BYTE = 8;
+
+ // register of avx512 are 512 bits, and can load up to 64 bytes
+ private static final int BYTES_PER_512VECTOR = 64;
+
+ // values are bit packed 8 at a time, so reading bitWidth will always work
+ private static final int VALUES_PER_PACKED = 8;
+
+ private static volatile Boolean vector;
Review Comment:
Does it make sense to use an enum instead of boolean to represent longest
available vector support? Something like this:
```
enum VectorSupport {
NONE,
VECTOR_128,
VECTOR_256,
VECTOR_512
}
```
This makes it more extensible.
##########
parquet-column/src/main/java/org/apache/parquet/column/values/bitpacking/ParquetReadRouter.java:
##########
@@ -0,0 +1,141 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.parquet.column.values.bitpacking;
+
+import org.apache.parquet.bytes.ByteBufferInputStream;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.nio.ByteBuffer;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.nio.file.Paths;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+/**
+ * Utility class for big data applications (such as Apache Spark and Apache
Flink).
+ * For Intel CPU, Flags containing avx512vbmi and avx512_vbmi2 can have better
performance gains.
+ */
+public class ParquetReadRouter {
+ private static final Logger LOG =
LoggerFactory.getLogger(ParquetReadRouter.class);
+
+ private static final int BITS_PER_BYTE = 8;
+
+ // register of avx512 are 512 bits, and can load up to 64 bytes
+ private static final int BYTES_PER_512VECTOR = 64;
+
+ // values are bit packed 8 at a time, so reading bitWidth will always work
+ private static final int VALUES_PER_PACKED = 8;
+
+ private static volatile Boolean vector;
+
+ // The is just a logical method, most call readBatchUsing512Vector or
readBatch directly given that you should know your computer system
Review Comment:
```suggestion
// Dispatches to use vector when available. Directly call
readBatchUsing512Vector() if you are sure about it.
```
##########
parquet-encoding/src/main/java/org/apache/parquet/column/values/bitpacking/BytePacker.java:
##########
@@ -105,4 +116,26 @@ public void unpack8Values(final byte[] input, final int
inPos, final int[] outpu
public void unpack32Values(byte[] input, int inPos, int[] output, int
outPos) {
unpack32Values(ByteBuffer.wrap(input), inPos, output, outPos);
}
+
+ /**
+ * unpack bitWidth bytes from input at inPos into {unpackCount} values in
output at outPos using Java Vector API.
+ * @param input the input bytes
+ * @param inPos where to read from in input
+ * @param output the output values
+ * @param outPos where to write to in output
+ */
+ public void unpackValuesVector(final byte[] input, final int inPos, final
int[] output, final int outPos) {
+
+ }
+
+ /**
+ * unpack bitWidth bytes from input at inPos into {unpackCount} values in
output at outPos using Java Vector API.
+ * @param input the input bytes
+ * @param inPos where to read from in input
+ * @param output the output values
+ * @param outPos where to write to in output
+ */
+ public void unpackValuesVector(final ByteBuffer input, final int inPos,
final int[] output, final int outPos) {
+
Review Comment:
I mean why is it an empty function? Shouldn't it throw an unimplemented
exception?
##########
parquet-column/src/main/java/org/apache/parquet/column/values/bitpacking/ParquetReadRouter.java:
##########
@@ -0,0 +1,141 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you 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 org.apache.parquet.column.values.bitpacking;
+
+import org.apache.parquet.bytes.ByteBufferInputStream;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.nio.ByteBuffer;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.nio.file.Paths;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+/**
+ * Utility class for big data applications (such as Apache Spark and Apache
Flink).
+ * For Intel CPU, Flags containing avx512vbmi and avx512_vbmi2 can have better
performance gains.
+ */
+public class ParquetReadRouter {
+ private static final Logger LOG =
LoggerFactory.getLogger(ParquetReadRouter.class);
+
+ private static final int BITS_PER_BYTE = 8;
+
+ // register of avx512 are 512 bits, and can load up to 64 bytes
+ private static final int BYTES_PER_512VECTOR = 64;
Review Comment:
```suggestion
private static final int BYTES_PER_VECTOR_512 = 64;
```
> Parquet bit-packing de/encode optimization
> ------------------------------------------
>
> Key: PARQUET-2159
> URL: https://issues.apache.org/jira/browse/PARQUET-2159
> Project: Parquet
> Issue Type: Improvement
> Components: parquet-mr
> Affects Versions: 1.13.0
> Reporter: Fang-Xie
> Assignee: Fang-Xie
> Priority: Major
> Fix For: 1.13.0
>
> Attachments: image-2022-06-15-22-56-08-396.png,
> image-2022-06-15-22-57-15-964.png, image-2022-06-15-22-58-01-442.png,
> image-2022-06-15-22-58-40-704.png
>
>
> Current Spark use Parquet-mr as parquet reader/writer library, but the
> built-in bit-packing en/decode is not efficient enough.
> Our optimization for Parquet bit-packing en/decode with jdk.incubator.vector
> in Open JDK18 brings prominent performance improvement.
> Due to Vector API is added to OpenJDK since 16, So this optimization request
> JDK16 or higher.
> *Below are our test results*
> Functional test is based on open-source parquet-mr Bit-pack decoding
> function: *_public final void unpack8Values(final byte[] in, final int inPos,
> final int[] out, final int outPos)_* __
> compared with our implementation with vector API *_public final void
> unpack8Values_vec(final byte[] in, final int inPos, final int[] out, final
> int outPos)_*
> We tested 10 pairs (open source parquet bit unpacking vs ours optimized
> vectorized SIMD implementation) decode function with bit
> width=\{1,2,3,4,5,6,7,8,9,10}, below are test results:
> !image-2022-06-15-22-56-08-396.png|width=437,height=223!
> We integrated our bit-packing decode implementation into parquet-mr, tested
> the parquet batch reader ability from Spark VectorizedParquetRecordReader
> which get parquet column data by the batch way. We construct parquet file
> with different row count and column count, the column data type is Int32, the
> maximum int value is 127 which satisfies bit pack encode with bit width=7,
> the count of the row is from 10k to 100 million and the count of the column
> is from 1 to 4.
> !image-2022-06-15-22-57-15-964.png|width=453,height=229!
> !image-2022-06-15-22-58-01-442.png|width=439,height=217!
> !image-2022-06-15-22-58-40-704.png|width=415,height=208!
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
