[all-commits] [llvm/llvm-project] 4336fe: [SLP]Do not pre-cache the last instruction for spl...

[Alexey Bataev via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 4336fef7b86f086b62739448f33ceccce4dbdaa3
      https://github.com/llvm/llvm-project/commit/4336fef7b86f086b62739448f33ceccce4dbdaa3
  Author: Alexey Bataev <a.bataev at outlook.com>
  Date:   2026-09-30 (Wed, 30 Sep 2026)

  Changed paths:
    M llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
    A llvm/test/Transforms/SLPVectorizer/X86/split-node-reused-phi-insertion-point.ll

  Log Message:
  -----------
  [SLP]Do not pre-cache the last instruction for split nodes

Their insertion point depends on the vector values of the already emitted
nodes, pre-caching may place them before the values they use.

Fixes https://godbolt.org/z/evG5Tba4j

Reviewers: 

Pull Request: https://github.com/llvm/llvm-project/pull/227878



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