[all-commits] [llvm/llvm-project] 3bc2da: [LoopVectorize] Enable vectorisation of early exit...

[David Sherwood via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 3bc2dade36cb14156ea2442e576d65ff5e5e1a50
      https://github.com/llvm/llvm-project/commit/3bc2dade36cb14156ea2442e576d65ff5e5e1a50
  Author: David Sherwood <david.sherwood at arm.com>
  Date:   2025-01-30 (Thu, 30 Jan 2025)

  Changed paths:
    M llvm/docs/Vectorizers.rst
    M llvm/docs/vplan-early-exit.dot
    M llvm/docs/vplan-early-exit.png
    M llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
    M llvm/lib/Transforms/Vectorize/VPlan.cpp
    M llvm/lib/Transforms/Vectorize/VPlan.h
    M llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp
    M llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
    M llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
    M llvm/lib/Transforms/Vectorize/VPlanTransforms.h
    M llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp
    M llvm/test/Transforms/LoopVectorize/AArch64/simple_early_exit.ll
    M llvm/test/Transforms/LoopVectorize/early_exit_legality.ll
    M llvm/test/Transforms/LoopVectorize/single_early_exit.ll
    M llvm/test/Transforms/LoopVectorize/single_early_exit_live_outs.ll
    M llvm/test/Transforms/LoopVectorize/uncountable-early-exit-vplan.ll

  Log Message:
  -----------
  [LoopVectorize] Enable vectorisation of early exit loops with live-outs (#120567)

This work feeds part of PR
https://github.com/llvm/llvm-project/pull/88385, and adds support for
vectorising
loops with uncountable early exits and outside users of loop-defined
variables. When calculating the final value from an uncountable early
exit we need to calculate the vector lane that triggered the exit,
and hence determine the value at the point we exited.

All code for calculating the last value when exiting the loop early
now lives in a new vector.early.exit block, which sits between the
middle.split block and the original exit block. Doing this required
two fixes:

1. The vplan verifier incorrectly assumed that the block containing
a definition always dominates the block of the user. That's not true
if you can arrive at the use block from multiple incoming blocks.
This is possible for early exit loops where both the early exit and
the latch jump to the same block.
2. We were adding the new vector.early.exit to the wrong parent loop.
It needs to have the same parent as the actual early exit block from
the original loop.

I've added a new ExtractFirstActive VPInstruction that extracts the
first active lane of a vector, i.e. the lane of the vector predicate
that triggered the exit.

NOTE: The IR generated for dealing with live-outs from early exit
loops is unoptimised, as opposed to normal loops. This inevitably
leads to poor quality code, but this can be fixed up later.



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