[mlir][Linalg] Account for the case where loop bounds depend on output shape.

[mravishankar]

Summary

Depends On D93076

[mravishankar]

Rebase. This might not be necessary after
https://llvm.discourse.group/t/linalg-and-shapes/2421 is resolved.

[hanchung]

mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp

72

no need for template and use auto

[mravishankar]

Dont need this anymore. Abondoning.

mlir/lib/Dialect/Linalg/IR/LinalgOps.cpp

 SmallVector<Value, 4> LinalgOp::createFlatListOfOperandDims(OpBuilder &b,
                                                             Location loc) {
   SmallVector<Value, 4> res;
   for (Value v : getShapedOperands()) {
     ShapedType t = v.getType().template cast<ShapedType>();
     for (unsigned i = 0, e = t.getRank(); i < e; ++i)
       res.push_back(b.create<DimOp>(loc, v, i));
   }
+  if (getNumInitTensors() == 0 && getOperation()->getNumResults() != 0) {
+    OpBuilder::InsertionGuard guard(b);
+    b.setInsertionPointAfter(getOperation());
+    for (Value v : getOperation()->getResults()) {
+      ShapedType t = v.getType().template cast<ShapedType>();
+      for (unsigned i = 0, e = t.getRank(); i < e; ++i)
+        res.push_back(b.create<DimOp>(loc, v, i));
+    }
+  }
   return res;
 }
 
 SmallVector<Range, 4> LinalgOp::createLoopRanges(OpBuilder &b, Location loc) {
   AffineMap map = getLoopsToShapesMap();
   unsigned numDims = map.getNumDims(), numRes = map.getNumResults();
   auto viewSizes = createFlatListOfOperandDims(b, loc);
   SmallVector<Range, 4> res(numDims);