HDFS-17914. Prevent DFSInputStream from issuing 0-byte reads

[lirui-apache]

Description of PR

Avoid empty buffer read in DFSInputStream#actualGetFromOneDataNode by terminating the loop once the buffer is filled.

How was this patch tested?

Added new test

For code changes:

• Does the title or this PR starts with the corresponding JIRA issue id (e.g. 'HADOOP-17799. Your PR title ...')?
• Object storage: have the integration tests been executed and the endpoint declared according to the connector-specific documentation?
• If adding new dependencies to the code, are these dependencies licensed in a way that is compatible for inclusion under ASF 2.0?
• If applicable, have you updated the LICENSE, LICENSE-binary, NOTICE-binary files?

AI Tooling

If an AI tool was used:

• The PR includes the phrase "Contains content generated by "
  where is the name of the AI tool used.
• My use of AI contributions follows the ASF legal policy
  https://www.apache.org/legal/generative-tooling.html

[Hexiaoqiao]

Thanks for your contribution first. Just one nit concern, is it only one difference that return at L1235(after PR) or return at L1236(before PR)? It will perform a zero-byte transfer in-memory and return immediately if tmp.remaining()=0, So it is OK for me before PR. welcome discussion if I miss something. Thanks again.

[lirui-apache]

Thanks for the review. I hit the issue when I was testing short circuit read. If we call BlockReaderLocal.read with tmp.remaining() == 0, it will hit the slow lane via BlockReaderLocal::readWithBounceBuffer::fillDataBuf, which will actually allocate and read into the bounce buffer. But since user buffer is filled, we cannot drain into it and the read data is thrown away and wasted.