You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@spark.apache.org by "Michael Munday (Jira)" <ji...@apache.org> on 2020/09/15 13:46:00 UTC

[jira] [Created] (SPARK-32892) Murmur3 and xxHash64 implementations do not produce the correct results on big-endian platforms

Michael Munday created SPARK-32892:
--------------------------------------

             Summary: Murmur3 and xxHash64 implementations do not produce the correct results on big-endian platforms
                 Key: SPARK-32892
                 URL: https://issues.apache.org/jira/browse/SPARK-32892
             Project: Spark
          Issue Type: Bug
          Components: Spark Core, SQL
    Affects Versions: 3.0.1
            Reporter: Michael Munday


The Murmur3 and xxHash64 implementations in Spark do not produce the correct results on big-endian systems. This causes test failures on my target platform (s390x).

These hash functions require that multi-byte chunks be interpreted as integers encoded in *little-endian* byte order. This requires byte reversal when using multi-byte unsafe operations on big-endian platforms.

I have a PR ready for discussion and review.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org