[all-commits] [llvm/llvm-project] ced9f4: [MLIR] Modify lowering of gpu.alloc op to llvm (#6...

[Nishant Patel via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: ced9f4f0e808618bb75a3b133150ab3f9213e650
      https://github.com/llvm/llvm-project/commit/ced9f4f0e808618bb75a3b133150ab3f9213e650
  Author: Nishant Patel <nishant.b.patel at intel.com>
  Date:   2023-10-25 (Wed, 25 Oct 2023)

  Changed paths:
    M mlir/lib/Conversion/GPUCommon/GPUToLLVMConversion.cpp
    M mlir/test/Conversion/GPUCommon/lower-alloc-to-gpu-runtime-calls.mlir

  Log Message:
  -----------
  [MLIR] Modify lowering of gpu.alloc op to llvm (#69969)

If gpu.alloc has no asyn deependency ( in case if gpu.alloc has
hostShared allocation), create a new stream & synchronize. This PR is
follow up to #66401