[all-commits] [llvm/llvm-project] 632317: [VPlan] Add non-poison propagating LogicalAnd VPIn...

[Florian Hahn via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 632317e9ab5548e991d8974954353033bea62a5b
      https://github.com/llvm/llvm-project/commit/632317e9ab5548e991d8974954353033bea62a5b
  Author: Florian Hahn <flo at fhahn.com>
  Date:   2024-05-14 (Tue, 14 May 2024)

  Changed paths:
    M llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
    M llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
    M llvm/lib/Transforms/Vectorize/VPlan.h
    M llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
    M llvm/test/Transforms/LoopVectorize/vplan-printing.ll
    M llvm/test/Transforms/LoopVectorize/vplan-sink-scalars-and-merge.ll

  Log Message:
  -----------
  [VPlan] Add non-poison propagating LogicalAnd VPInstruction opcode. (#91897)

Add a new opcode to mode non-poison propagating logical AND operations
used when generating edge masks. This follows the similar decision to
model Not as dedicated opcode as well, to improve clarity.

This also helps to simplify the matchers for
https://github.com/llvm/llvm-project/pull/89386.


PR: https://github.com/llvm/llvm-project/pull/91897



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