TestParquetReader fails with hadoop 3.4.2

### Describe the bug, including details regarding any error messages, version, and platform.

A couple of test cases of `TestParquetReader` suite started failing with errors like:
```
- testRangeFiltering[0] *** FAILED ***
org.apache.parquet.bytes.TrackingByteBufferAllocator$LeakedByteBufferException: 24 ByteBuffer object(s) is/are remained unreleased after closing this allocator.
at org.apache.parquet.bytes.TrackingByteBufferAllocator.close(TrackingByteBufferAllocator.java:161)
at org.apache.parquet.hadoop.TestParquetReader.closeAllocator(TestParquetReader.java:175)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:77)
at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.base/java.lang.reflect.Method.invoke(Method.java:569)
```
with hadoop 3.4.2 and parquet-mr 1.16.0
The leaks are happening when reading using vectored IO because we never pass the buffers to the releaser.

### Component(s)

_No response_

Commencez par les cas en échec de TestParquetReader, en particulier testRangeFiltering, et examinez le chemin de lecture vectored IO. Vérifiez comment les buffers renvoyés sont traités par le releaser, puis relancez la suite avec Hadoop 3.4.2 et parquet-mr 1.16.0. La tâche est terminée lorsque les tests passent sans erreurs de fuite de TrackingByteBufferAllocator.

Rédigé par le modèle d'indexation à partir du texte de l'issue.