[all-commits] [llvm/llvm-project] 2c9b6c: [AArch64][GlobalISel] Improve codegen for G_VECRED...

[Dhruv Chawla (work) via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 2c9b6c1b36b8185299de083c3058e0c1e7760442
      https://github.com/llvm/llvm-project/commit/2c9b6c1b36b8185299de083c3058e0c1e7760442
  Author: Dhruv Chawla (work) <dhruvc at nvidia.com>
  Date:   2024-02-27 (Tue, 27 Feb 2024)

  Changed paths:
    M llvm/include/llvm/CodeGen/GlobalISel/LegalizerHelper.h
    M llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp
    M llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp
    M llvm/test/CodeGen/AArch64/aarch64-minmaxv.ll
    M llvm/test/CodeGen/AArch64/vecreduce-umax-legalization.ll

  Log Message:
  -----------
  [AArch64][GlobalISel] Improve codegen for G_VECREDUCE_{SMIN,SMAX,UMIN,UMAX} for odd-sized vectors (#82740)

i8 vectors do not have their sizes changed as I noticed regressions in
some tests when that was done.

This patch also adds support for most G_VECREDUCE_* operations to
moreElementsVector in LegalizerHelper.cpp.

The code for getting the "neutral" element is taken almost exactly as it
is in SelectionDAG, with the exception that support for
G_VECREDUCE_{FMAXIMUM,FMINIMUM} was not added.

The code for SelectionDAG is located at
SelectionDAG::getNeutralELement().



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