Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-8204353

AARCH64: optimize FPU load and stores in macroAssembler

XMLWordPrintable

    • Icon: Enhancement Enhancement
    • Resolution: Fixed
    • Icon: P3 P3
    • 11
    • 10, 11
    • hotspot
    • b19
    • aarch64
    • linux

        push_call_clobbered_registers, pop_call_clobbered_registers, push_CPU_state, pop_CPU_state methods are using 2-fpu-registers-per-instruction approach, while it can use 4-fpu-registers-per-instruction, which will make respective code up to 2 times shorter

              dpochepk Dmitrij Pochepko
              dpochepk Dmitrij Pochepko
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated:
                Resolved: