[all-commits] [llvm/llvm-project] 84695d: Fix conversion of loops to GPU with no block/threa...

[Stephan Herhut via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 84695dd4d788dd65bb2b83ae1512a86c42109d7e
      https://github.com/llvm/llvm-project/commit/84695dd4d788dd65bb2b83ae1512a86c42109d7e
  Author: Stephan Herhut <herhut at google.com>
  Date:   2020-01-31 (Fri, 31 Jan 2020)

  Changed paths:
    M mlir/lib/Conversion/LoopsToGPU/LoopsToGPU.cpp
    A mlir/test/Conversion/LoopsToGPU/no_blocks_no_threads.mlir

  Log Message:
  -----------
  Fix conversion of loops to GPU with no block/thread dimensions.

Summary:
The current code assumes that one always maps at least one loop to block
dimensions and at least one loop to thread dimensions. If either is not
the case, a loop would get mapped twice.

Differential Revision: https://reviews.llvm.org/D73685