[all-commits] [llvm/llvm-project] 4c28d2: [AArch64] Avoid single-element vector fp converts ...

[Benjamin Maxwell via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 4c28d21f6af70ffee33660de35b263283dc32139
      https://github.com/llvm/llvm-project/commit/4c28d21f6af70ffee33660de35b263283dc32139
  Author: Benjamin Maxwell <benjamin.maxwell at arm.com>
  Date:   2024-10-16 (Wed, 16 Oct 2024)

  Changed paths:
    M llvm/lib/Target/AArch64/AArch64InstrInfo.td
    A llvm/test/CodeGen/AArch64/sve-streaming-mode-cvt-fp-int-fp.ll
    M llvm/test/CodeGen/AArch64/sve-streaming-mode-fixed-length-int-to-fp.ll

  Log Message:
  -----------
  [AArch64] Avoid single-element vector fp converts in streaming[-compatible] functions (#112213)

The single-element vector variants of FCVTZS, FCVTZU, UCVTF, and SCVTF
are only supported in streaming[-compatible] functions with `+sme2p2`.

Reference:
-
https://developer.arm.com/documentation/ddi0602/2024-09/SIMD-FP-Instructions/FCVTZS--vector--integer---Floating-point-convert-to-signed-integer--rounding-toward-zero--vector--
-
https://developer.arm.com/documentation/ddi0602/2024-09/SIMD-FP-Instructions/UCVTF--vector--integer---Unsigned-integer-convert-to-floating-point--vector--
-
https://developer.arm.com/documentation/ddi0602/2024-09/SIMD-FP-Instructions/SCVTF--vector--integer---Signed-integer-convert-to-floating-point--vector--

Codegen will be improved in follow up patches.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications