[all-commits] [llvm/llvm-project] 1f852b: [PowerPC] Avoid unnecessary fadd for unsigned to p...

[Qiu Chaofan via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 1f852ba8534cc773c114165e77acba411ac25a93
      https://github.com/llvm/llvm-project/commit/1f852ba8534cc773c114165e77acba411ac25a93
  Author: Qiu Chaofan <qiucofan at cn.ibm.com>
  Date:   2020-11-01 (Sun, 01 Nov 2020)

  Changed paths:
    M llvm/lib/CodeGen/SelectionDAG/LegalizeFloatTypes.cpp
    M llvm/test/CodeGen/PowerPC/ppcf128-constrained-fp-intrinsics.ll
    M llvm/test/CodeGen/PowerPC/uint-to-ppcfp128-crash.ll

  Log Message:
  -----------
  [PowerPC] Avoid unnecessary fadd for unsigned to ppcf128

Unsigned 32-bit or shorter integer to ppcf128 conversion are currently
expanded as signed-to-double with an extra fadd to 'complement'. But on
PowerPC we have native instruction to directly convert unsigned to
double since ISA v2.06. This patch exploits it.

Reviewed By: efriedma

Differential Revision: https://reviews.llvm.org/D89786