[Mlir-commits] [mlir] [mlir][linalg] Support quantized NCHW/FCHW convolution in img2col (PR #225308)

[Federico Bruzzone]

================
@@ -410,8 +410,9 @@ rewriteInIm2Col(RewriterBase &rewriter,
                         transposedResult.getDefiningOp());
 }
 
-FailureOr<std::pair<Operation *, Operation *>>
-rewriteInIm2Col(RewriterBase &rewriter, linalg::Conv2DNchwFchwOp convOp) {
+template <typename ConvOp>
+static FailureOr<std::pair<Operation *, Operation *>>
+rewriteNchwFchwInIm2col(RewriterBase &rewriter, ConvOp convOp) {
----------------
FedericoBruzzone wrote:

Maybe a `static_assert` for `Conv2DNchwFchwOp/Conv2DNchwFchwQOp` here?

https://github.com/llvm/llvm-project/pull/225308