maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[MLIR][NFC] Improve doc comment and delete stale comment 

Remove duplicate and stale doc comment on affineParallelize. NFC.
This commit is contained in:
Uday Bondhugula 2021-07-16 14:59:16 +05:30
parent 69a5684531
commit edfcfa6f7b
2 changed files with 6 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
mlir/include/mlir/Dialect/Affine/Utils.h
View File

@ -30,11 +30,12 @@ class Operation;
using ReductionLoopMap = DenseMap<Operation *, SmallVector<LoopReduction, 2>>; using ReductionLoopMap = DenseMap<Operation *, SmallVector<LoopReduction, 2>>;
/// Replaces parallel affine.for op with 1-d affine.parallel op. /// Replaces a parallel affine.for op with a 1-d affine.parallel op. `forOp`'s
/// mlir::isLoopParallel detects the parallel affine.for ops. /// body is taken by the affine.parallel op and the former is erased.
/// Parallelizes the specified reductions. Parallelization will fail in presence /// (mlir::isLoopParallel can be used to detect a parallel affine.for op.) The
/// of loop iteration arguments that are not listed in `parallelReductions`. /// reductions specified in `parallelReductions` are also parallelized.
/// There is no cost model currently used to drive this parallelization. /// Parallelization will fail in the presence of loop iteration arguments that
/// are not listed in `parallelReductions`.
LogicalResult LogicalResult
affineParallelize(AffineForOp forOp, affineParallelize(AffineForOp forOp,
ArrayRef<LoopReduction> parallelReductions = {}); ArrayRef<LoopReduction> parallelReductions = {});

4
mlir/lib/Dialect/Affine/Utils/Utils.cpp
View File

@ -129,10 +129,6 @@ static AffineIfOp hoistAffineIfOp(AffineIfOp ifOp, Operation *hoistOverOp) {
return hoistedIfOp; return hoistedIfOp;
} }
/// Replace affine.for with a 1-d affine.parallel and clone the former's body
/// into the latter while remapping values. Parallelizes the specified
/// reductions. Parallelization will fail in presence of loop iteration
/// arguments that are not listed in `parallelReductions`.
LogicalResult LogicalResult
mlir::affineParallelize(AffineForOp forOp, mlir::affineParallelize(AffineForOp forOp,
ArrayRef<LoopReduction> parallelReductions) { ArrayRef<LoopReduction> parallelReductions) {