Add new Float16Vector type and enable intrinsification of vector operations supported by auto-vectorizer

XMLWordPrintable

    • Type: Enhancement
    • Resolution: Unresolved
    • Priority: P4
    • tbd
    • Affects Version/s: 26
    • Component/s: hotspot
    • generic
    • generic

      - Add a new HalffloatVector type in addition to existing primitive vector types, maintaining operation parity with the FloatVector type.
      - Add necessary inline expander support.
         - Enable intrinsification for a few vector operations, namely ADD/SUB/MUL/DIV/MAX/MIN/FMA.
      - Use existing Float16 vector IR and backend support.
         
      The idea here is to first be on par with Float16 auto-vectorization support.

            Assignee:
            Jatin Bhateja
            Reporter:
            Jatin Bhateja
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

              Created:
              Updated: