[all-commits] [llvm/llvm-project] 65d3dd: [VPlan] Add first VPlan version of sinkScalarOpera...

[Florian Hahn via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 65d3dd7c883fd46c855afe71a2f04cf84750e730
      https://github.com/llvm/llvm-project/commit/65d3dd7c883fd46c855afe71a2f04cf84750e730
  Author: Florian Hahn <flo at fhahn.com>
  Date:   2021-05-24 (Mon, 24 May 2021)

  Changed paths:
    M llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
    M llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
    M llvm/lib/Transforms/Vectorize/VPlanTransforms.h
    M llvm/test/Transforms/LoopVectorize/X86/small-size.ll
    M llvm/test/Transforms/LoopVectorize/first-order-recurrence-sink-replicate-region.ll
    M llvm/test/Transforms/LoopVectorize/if-pred-stores.ll
    M llvm/test/Transforms/LoopVectorize/vplan-sink-scalars-and-merge-vf1.ll
    M llvm/test/Transforms/LoopVectorize/vplan-sink-scalars-and-merge.ll

  Log Message:
  -----------
  [VPlan] Add first VPlan version of sinkScalarOperands.

This patch adds a first VPlan-based implementation of sinking of scalar
operands.

The current version traverse a VPlan once and processes all operands of
a predicated REPLICATE recipe. If one of those operands can be sunk,
it is moved to the block containing the predicated REPLICATE recipe.
Continue with processing the operands of the sunk recipe.

The initial version does not re-process candidates after other recipes
have been sunk. It also cannot partially sink induction increments at
the moment. The VPlan only contains WIDEN-INDUCTION recipes and if the
induction is used for example in a GEP, only the first lane is used and
in the lowered IR the adds for the other lanes can be sunk into the
predicated blocks.

Reviewed By: Ayal

Differential Revision: https://reviews.llvm.org/D100258