[PATCH] D72555: [mlir][Linalg] Update the semantics, verifier and test for Linalg with tensors.

[Nicolas Vasilache via Phabricator via llvm-commits]

nicolasvasilache marked 8 inline comments as done.
nicolasvasilache added inline comments.


================
Comment at: mlir/include/mlir/Dialect/Linalg/IR/LinalgStructuredOps.td:694-697
       : tensor<?x?xf32>,
         memref<?x?xf32, stride_specification>,
         tensor<?x?xf32>
         -> (tensor<?x?xf32>)
----------------
herhut wrote:
> So this example has a single output assuming that there is an order constraint on the arguments (inputs first, then outputs). Could the describe this here?
Updated the example, had forgotten this, thanks!


================
Comment at: mlir/lib/Dialect/Linalg/Transforms/LinalgToLoops.cpp:90
 public:
   static void emitScalarImplementation(ArrayRef<Value> allIvs, CopyOp copyOp) {
     auto nPar = copyOp.getNumParallelLoops();
----------------
herhut wrote:
> This code assumes an all buffer representation? Maybe add an assert?
Sprinkled a bunch in the relevant places, thanks!


================
Comment at: mlir/lib/Dialect/Linalg/Transforms/LinalgToLoops.cpp:216
 template <typename IndexedValueType>
 class LinalgScopedEmitter<IndexedValueType, GenericOp> {
 public:
----------------
herhut wrote:
> This code assumes an all buffers representation?
sprinkled a bunch of asserts in the relevant places, thanks!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D72555/new/

https://reviews.llvm.org/D72555