[all-commits] [llvm/llvm-project] 4277d9: [LV] Use speculatability within entire loop to avo...

[annamthomas via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 4277d932ef180090f314f7eb7f47d63e76881d30
      https://github.com/llvm/llvm-project/commit/4277d932ef180090f314f7eb7f47d63e76881d30
  Author: Anna Thomas <anna at azul.com>
  Date:   2023-03-21 (Tue, 21 Mar 2023)

  Changed paths:
    M llvm/lib/Analysis/Loads.cpp
    M llvm/test/Transforms/LoopVectorize/X86/load-deref-pred.ll

  Log Message:
  -----------
  [LV] Use speculatability within entire loop to avoid strided load predication

Use existing functionality for identifying total access size by strided
loads. If we can speculate the load across all vector iterations, we can
avoid predication for these strided loads (or masked gathers in
architectures which support it).

Differential Revision: https://reviews.llvm.org/D145616