[llvm] [mlir] [MLIR] Generalize expand_shape to take shape as explicit input (PR #69267)

Gaurav Shukla via llvm-commits llvm-commits at lists.llvm.org
Thu Apr 18 13:22:24 PDT 2024


================
@@ -1102,50 +1097,103 @@ def Tensor_ExpandShapeOp : Tensor_ReassociativeReshapeOp<"expand_shape"> {
     rank than the operand `src` whose dimension sizes are a reassociation of
     `src`.
 
-    A reassociation is defined as a continuous grouping of dimensions. It is
-    represented with an array of DenseI64ArrayAttr attribute. Entries in the
-    array are referred to as reassociation maps.
+    A reassociation is defined as a continuous grouping of dimensions and is
+    represented with an array of DenseI64ArrayAttr attribute.  The reassociation
+    maps applied to the result tensor with the higher rank must result in the
+    operand tensor with the smaller rank.
 
-    The reassociation maps are applied to the result shape to obtain the operand
-    shape.
+    The representation for the output shape supports a partially-static
+    specification via attributes specified through the `static_output_shape`
+    argument.  A special sentinel value `ShapedType::kDynamic` encodes that the
+    corresponding entry has a dynamic value.  There must be exactly as many SSA
+    inputs in `output_shape` as there are `ShapedType::kDynamic` entries in
+    `static_output_shape`.
 
     Example:
 
     ```mlir
     // Dimension expansion i -> (i', j') and (k) -> (k')
-    %b = tensor.expand_shape %a [[0, 1], [2]]
-        : tensor<?x?xf32> into tensor<?x?x?xf32>
+    %b = tensor.expand_shape %a [[0, 1], [2]] [%sz0, %sz1, 32]
+        : tensor<?x32xf32> into tensor<?x?x32xf32>
     ```
   }];
+
+  let arguments = (ins AnyTensor:$src, IndexListArrayAttr:$reassociation,
+                       Variadic<Index>:$output_shape,
+                       DenseI64ArrayAttr:$static_output_shape);
+
+  let assemblyFormat = [{
+    $src $reassociation `output_shape`
+    custom<DynamicIndexList>($output_shape, $static_output_shape) attr-dict `:`
+    type($src) `into` type($result)
+  }];
+
   let builders = [
     // Builders using ReassociationIndices.
+    OpBuilder<(ins "Type":$resultType, "Value":$src,
+      "ArrayRef<ReassociationIndices>":$reassociation),
+    [{
+      SmallVector<OpFoldResult> inputShape = 
----------------
Shukla-Gaurav wrote:

Moved the implementation to TensorOps.cpp for the first two builders as they contain non-trivial code. Thanks :)

https://github.com/llvm/llvm-project/pull/69267


More information about the llvm-commits mailing list