Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-32950

No need for some big-endian specific code paths in {On,Off}HeapColumnVector

    XMLWordPrintableJSON

Details

    • Improvement
    • Status: Resolved
    • Trivial
    • Resolution: Fixed
    • 3.0.1
    • 3.1.0
    • SQL

    Description

      There is no need for a separate code path for big-endian platforms in putFloats and putDoubles in OnHeapColumnVector and OffHeapColumnVector. Since SPARK-26985 was fixed the values have been copied in native byte order so the code required to perform this operation can be the same on both little- and big-endian platforms.

      Attachments

        Activity

          People

            mundaym Michael Munday
            mundaym Michael Munday
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: