[all-commits] [llvm/llvm-project] d5f113: [AArch64] Default to zero-cycle-zeroing FP registers

[sjoerdmeijer via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: d5f1131c812df57560c7563475cb0d674a101636
      https://github.com/llvm/llvm-project/commit/d5f1131c812df57560c7563475cb0d674a101636
  Author: Sjoerd Meijer <sjoerd.meijer at arm.com>
  Date:   2021-04-06 (Tue, 06 Apr 2021)

  Changed paths:
    M llvm/lib/Target/AArch64/AArch64.td
    M llvm/lib/Target/AArch64/AArch64Subtarget.h
    M llvm/test/CodeGen/AArch64/arm64-fast-isel-materialize.ll
    M llvm/test/CodeGen/AArch64/arm64-fp-contract-zero.ll
    M llvm/test/CodeGen/AArch64/arm64-rev.ll
    M llvm/test/CodeGen/AArch64/arm64-zero-cycle-zeroing.ll
    M llvm/test/CodeGen/AArch64/f16-imm.ll
    M llvm/test/CodeGen/AArch64/fptosi-sat-scalar.ll
    M llvm/test/CodeGen/AArch64/fptosi-sat-vector.ll
    M llvm/test/CodeGen/AArch64/fptoui-sat-scalar.ll
    M llvm/test/CodeGen/AArch64/fptoui-sat-vector.ll
    M llvm/test/CodeGen/AArch64/remat-float0.ll

  Log Message:
  -----------
  [AArch64] Default to zero-cycle-zeroing FP registers

It is generally beneficial to prefer "movi d0, #0" over "fmov s0, wzr" as this
is most efficient across all cores; it is recognised as a zeroing idiom. For
newer cores, fmov instructions can also be eliminated early and there is no
difference with movi, but some implementations lack this so is not true for
other/older cores. Thus this standardises on using movi as this should always
gives the same or better performance than the fmov with wzr.

Differential Revision: https://reviews.llvm.org/D99586