[Mlir-commits] [mlir] [mlir][linalg] Vectorization support for convolution of i1 type (PR #109480)

[Andrzej Warzyński]

================
@@ -61,6 +61,35 @@ func.func @conv1d_nwc_4x2x8_memref(%input: memref<4x6x3xf32>, %filter: memref<1x
 
 // -----
 
+// This test is same as above but for i1 type.
----------------
banach-space wrote:

[nit] It's worth pointing out that the only difference should be `kind` and that's the main that's checked.

Also, I would add "check" lines for `kind` to the original tests (i.e. `@conv1d_nwc_4x2x8_memref` and `@conv1d_ncw_4x8x2_memref_i1`).

https://github.com/llvm/llvm-project/pull/109480