[all-commits] [llvm/llvm-project] 9f61fb: [LV] Relax assumption that LCSSA implies single entry

[Philip Reames via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 9f61fbd75ae1757d77988b37562de4d6583579aa
      https://github.com/llvm/llvm-project/commit/9f61fbd75ae1757d77988b37562de4d6583579aa
  Author: Philip Reames <listmail at philipreames.com>
  Date:   2021-01-12 (Tue, 12 Jan 2021)

  Changed paths:
    M llvm/lib/Transforms/Vectorize/LoopVectorizationLegality.cpp
    M llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
    M llvm/test/Transforms/LoopVectorize/first-order-recurrence-complex.ll
    M llvm/test/Transforms/LoopVectorize/loop-form.ll

  Log Message:
  -----------
  [LV] Relax assumption that LCSSA implies single entry

This relates to the ongoing effort to support vectorization of multiple exit loops (see D93317).

The previous code assumed that LCSSA phis were always single entry before the vectorizer ran. This was correct, but only because the vectorizer allowed only a single exiting edge. There's nothing in the definition of LCSSA which requires single entry phis.

A common case where this comes up is with a loop with multiple exiting blocks which all reach a common exit block. (e.g. see the test updates)

Differential Revision: https://reviews.llvm.org/D93725