[all-commits] [llvm/llvm-project] 6afcc5: [SCEV] Infer no-self-wrap via constant ranges

[Philip Reames via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 6afcc54ac7d68fa2b28f0e7cbf9dc1d4ac7fb95e
      https://github.com/llvm/llvm-project/commit/6afcc54ac7d68fa2b28f0e7cbf9dc1d4ac7fb95e
  Author: Philip Reames <preames at rivosinc.com>
  Date:   2023-03-22 (Wed, 22 Mar 2023)

  Changed paths:
    M llvm/lib/Analysis/ScalarEvolution.cpp
    M llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp
    M llvm/test/Analysis/ScalarEvolution/different-loops-recs.ll
    M llvm/test/Analysis/ScalarEvolution/max-backedge-taken-count-guard-info.ll
    M llvm/test/Transforms/LoopRotate/pr51981-scev-problem.ll
    M llvm/test/Transforms/LoopVersioning/lcssa.ll

  Log Message:
  -----------
  [SCEV] Infer no-self-wrap via constant ranges

Without this, pointer IVs in loops with small constant trip counts couldn't be proven no-self-wrap. This came up in a new LSR transform, but may also benefit other SCEV consumers as well.

Differential Revision: https://reviews.llvm.org/D146596