[all-commits] [llvm/llvm-project] dcdd5d: [mlir][LLVMIR] Use insertelement if needed when tr...

[Min-Yih Hsu via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: dcdd5d312f29356fc4329e4c2a9856db870ad617
      https://github.com/llvm/llvm-project/commit/dcdd5d312f29356fc4329e4c2a9856db870ad617
  Author: Min-Yih Hsu <minyihh at uci.edu>
  Date:   2022-06-15 (Wed, 15 Jun 2022)

  Changed paths:
    M mlir/lib/Target/LLVMIR/ConvertFromLLVMIR.cpp
    M mlir/test/Target/LLVMIR/Import/constant-aggregate.ll

  Log Message:
  -----------
  [mlir][LLVMIR] Use insertelement if needed when translating ConstantAggregate

When translating from a llvm::ConstantAggregate with vector type, we
should lower to insertelement operations (if needed) rather than using
insertvalue.

Differential Revision: https://reviews.llvm.org/D127534


  Commit: 719e24d39fad43e5d1f692ff72c1a2a40687b297
      https://github.com/llvm/llvm-project/commit/719e24d39fad43e5d1f692ff72c1a2a40687b297
  Author: Min-Yih Hsu <minyihh at uci.edu>
  Date:   2022-06-15 (Wed, 15 Jun 2022)

  Changed paths:
    M mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp
    M mlir/test/Dialect/LLVMIR/roundtrip.mlir

  Log Message:
  -----------
  [mlir][LLVMIR] Use isScalableVectorType in ShuffleVectorOp::parse

Instead of casting the incoming operand into VectorType to check if it's
scalable or not.
This is the place I missed to fix in f088b99eac74.

Differential Revision: https://reviews.llvm.org/D127535


  Commit: cd8978e19ed90ddd695a193525d50319e74ff507
      https://github.com/llvm/llvm-project/commit/cd8978e19ed90ddd695a193525d50319e74ff507
  Author: Min-Yih Hsu <minyihh at uci.edu>
  Date:   2022-06-15 (Wed, 15 Jun 2022)

  Changed paths:
    M mlir/include/mlir/Dialect/LLVMIR/LLVMOps.td
    M mlir/lib/Dialect/LLVMIR/IR/LLVMDialect.cpp
    M mlir/test/Dialect/LLVMIR/roundtrip.mlir

  Log Message:
  -----------
  [mlir][LLVMIR] Ask ICmpOp to return vector<Nxi1> when needed

If any of the operands for ICmpOp is a vector, returns a vector<Nxi1>
, rather than an i1 type result.

Differential Revision: https://reviews.llvm.org/D127536


Compare: https://github.com/llvm/llvm-project/compare/db6c3ecd2c7a...cd8978e19ed9