[Mlir-commits] [mlir] [NFC] Simplify the tiling implementation using cloning. (PR #72178)

[Han-Chung Wang]

================
@@ -139,15 +142,17 @@ static Operation *cloneOpAndUpdateDestinationArgs(RewriterBase &rewriter,
 /// - `loopRanges` specifies the lb, ub and step of the untiled iteration space.
 /// - `tileSizes` is the tile sizes to use. Zero represent untiled loops.
 /// - In `offsets` and `sizes` return the multi-dimensional offset and size of
-/// the
-///   tile processed within the inner most loop.
+///   the tile processed within the inner most loop.
+/// Note that this methods adds `scf.yield` operation for all but the innermost
+/// loop. These yield the value returned by the immediately inner loop. The
+/// caller is expected to add the scf.yield operation for the innermost loop.
 static SmallVector<scf::ForOp> generateTileLoopNest(
     OpBuilder &builder, Location loc, ArrayRef<Range> loopRanges,
     ArrayRef<OpFoldResult> tileSizes, SmallVector<OpFoldResult> &offsets,
-    SmallVector<OpFoldResult> &sizes) {
-  assert(!loopRanges.empty() && "expected at least one loop range");
-  assert(loopRanges.size() == tileSizes.size() &&
----------------
hanhanW wrote:

IIUC, it is for scalar ops. See the check_scalar_operation lit test. It is good to see that we are able to handle 0D tensor/memref in tiling, so we don't have to handle the corner case when we're using it.

https://github.com/llvm/llvm-project/pull/72178