Skip to content

GH-3404: fix CI TestByteBitPacking512VectorLE.unpackValuesUsingVector randomly oom #3405

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
wgtmac merged 1 commit into from
Feb 27, 2026
+32 −30
Merged

GH-3404: fix CI TestByteBitPacking512VectorLE.unpackValuesUsingVector randomly oom #3405

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
62 changes: 32 additions & 30 deletions .../test/java/org/apache/parquet/column/values/bitpacking/TestByteBitPacking512VectorLE.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,8 +22,9 @@

import java.math.BigDecimal;
import java.nio.ByteBuffer;
import java.util.ArrayList;
import java.util.List;
import java.util.Arrays;
import java.util.stream.IntStream;
import java.util.stream.Stream;
import org.junit.Assume;
import org.junit.Test;
import org.slf4j.Logger;
Expand All @@ -41,30 +42,31 @@ public void unpackValuesUsingVector() {
}

private void unpackValuesUsingVectorBitWidth(int bitWidth) {
List<int[]> intInputs = getRangeData(bitWidth);

for (int[] intInput : intInputs) {
int pack8Count = intInput.length / 8;
int byteOutputSize = pack8Count * bitWidth;
byte[] byteOutput = new byte[byteOutputSize];
int[] output1 = new int[intInput.length];
int[] output2 = new int[intInput.length];
int[] output3 = new int[intInput.length];

BytePacker bytePacker = Packer.LITTLE_ENDIAN.newBytePacker(bitWidth);
for (int i = 0; i < pack8Count; i++) {
bytePacker.pack8Values(intInput, 8 * i, byteOutput, bitWidth * i);
}
try (Stream<int[]> intInputs = getRangeData(bitWidth)) {
intInputs.forEach(intInput -> {
int pack8Count = intInput.length / 8;
int byteOutputSize = pack8Count * bitWidth;
byte[] byteOutput = new byte[byteOutputSize];
int[] output = new int[intInput.length];

BytePacker bytePacker = Packer.LITTLE_ENDIAN.newBytePacker(bitWidth);
for (int i = 0; i < pack8Count; i++) {
bytePacker.pack8Values(intInput, 8 * i, byteOutput, bitWidth * i);
}

unpack8Values(bitWidth, byteOutput, output1);
unpackValuesUsingVectorArray(bitWidth, byteOutput, output2);
unpack8Values(bitWidth, byteOutput, output);
assertArrayEquals(intInput, output);
Arrays.fill(output, 0);

ByteBuffer byteBuffer = ByteBuffer.wrap(byteOutput);
unpackValuesUsingVectorByteBuffer(bitWidth, byteBuffer, output3);
unpackValuesUsingVectorArray(bitWidth, byteOutput, output);
assertArrayEquals(intInput, output);
Arrays.fill(output, 0);

Comment on lines +57 to 64
Copy link

Copilot AI Feb 27, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Arrays.fill(output, 0) after each assert is an O(n) operation over the entire output array and adds significant runtime (especially for large chunks). Since each unpack method writes all positions it is responsible for, clearing should be unnecessary if output is sized to intInput.length; if you still need clearing, limit it to the written range (e.g., 0..intInput.length).

Copilot uses AI. Check for mistakes.
assertArrayEquals(intInput, output1);
assertArrayEquals(intInput, output2);
assertArrayEquals(intInput, output3);
ByteBuffer byteBuffer = ByteBuffer.wrap(byteOutput);
unpackValuesUsingVectorByteBuffer(bitWidth, byteBuffer, output);
assertArrayEquals(intInput, output);
Arrays.fill(output, 0);
});
}
}

Expand Down Expand Up @@ -119,8 +121,7 @@ public void unpackValuesUsingVectorByteBuffer(int bitWidth, ByteBuffer input, in
}
}

private List<int[]> getRangeData(int bitWidth) {
List<int[]> result = new ArrayList<>();
private Stream<int[]> getRangeData(int bitWidth) {
int itemMax = 268435456;

long maxValue = getMaxValue(bitWidth);
Expand All @@ -131,9 +132,11 @@ private List<int[]> getRangeData(int bitWidth) {
++itemCount;
}

for (int i = 0; i < itemCount; i++) {
final int finalItemCount = itemCount;

return IntStream.range(0, finalItemCount).mapToObj(i -> {
int len;
if ((i == itemCount - 1) && mode != 0) {
if ((i == finalItemCount - 1) && mode != 0) {
len = mode;
} else {
len = itemMax;
Expand Down Expand Up @@ -162,9 +165,8 @@ private List<int[]> getRangeData(int bitWidth) {
array[j] = value;
j++;
}
result.add(array);
}
return result;
return array;
});
}

private long getMaxValue(int bitWidth) {
Expand Down