[all-commits] [llvm/llvm-project] 27d191: [ARM][ParallelDSP] Handle squaring multiplies

[Sam Parker via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 27d19101e9a3bf6cabcc2b481847aa7b86087cc4
      https://github.com/llvm/llvm-project/commit/27d19101e9a3bf6cabcc2b481847aa7b86087cc4
  Author: Sam Parker <sam.parker at arm.com>
  Date:   2020-04-21 (Tue, 21 Apr 2020)

  Changed paths:
    M llvm/lib/Target/ARM/ARMParallelDSP.cpp
    A llvm/test/CodeGen/ARM/ParallelDSP/squaring.ll

  Log Message:
  -----------
  [ARM][ParallelDSP] Handle squaring multiplies

The logic in ARMParallelDSP is setup to merge two 16-bits loads into
a 32-bit load and feed them into the smlads. This requires that four
loads are combined for the four inputs, but there wasn't actually a
check for this.

Differential Revision: https://reviews.llvm.org/D78492