feat(dnn/naive): add convolution weight preprocess interface

GitOrigin-RevId: d0fd6c75a6c45922f18e734c134c5ddcb7bfa8d9

dnn/include/megdnn/oprs/nn.h

@@ -131,6 +131,13 @@ public:
         }
     };
 
+    struct PreprocessedFilter {
+        //! user data; its lifetime should be bound to MegDNN Convolution
+        //! operator
+        void* algorithm_id;
+        TensorNDArray tensors;
+    };
+
 protected:
     // Check or deduce output DType
     void check_or_deduce_dtype_fwd(DType src, DType filter, DType& dst) const;
@@ -200,13 +207,27 @@ public:
      * \param[out] dst (n, oc, oh, ow)
      */
     virtual void exec(_megdnn_tensor_in src, _megdnn_tensor_in filter,
-                      _megdnn_tensor_out dst, _megdnn_workspace workspace) = 0;
+                      _megdnn_tensor_out dst,
+                      const PreprocessedFilter* preprocessed_filter,
+                      _megdnn_workspace workspace) = 0;
+    virtual void exec_preprocess(const TensorLayout& src_layout,
+                                 _megdnn_tensor_in filter,
+                                 const TensorLayout& dst_layout,
+                                 PreprocessedFilter* preprocessed_filter,
+                                 _megdnn_workspace workspace) = 0;
     void deduce_dtype(DType src, DType filter, DType& dst);
     void deduce_layout(const TensorLayout& src, const TensorLayout& filter,
                        TensorLayout& dst);
-    virtual size_t get_workspace_in_bytes(const TensorLayout& src,
-                                          const TensorLayout& filter,
-                                          const TensorLayout& dst) = 0;
+    virtual size_t get_workspace_in_bytes(
+            const TensorLayout& src, const TensorLayout& filter,
+            const TensorLayout& dst,
+            PreprocessedFilter* preprocessed_filter) = 0;
+    virtual SmallVector<TensorLayout> deduce_preprocessed_filter_layout(
+            const TensorLayout& src, const TensorLayout& filter,
+            const TensorLayout& dst) = 0;
+    virtual size_t get_preprocess_workspace_in_bytes(
+            const TensorLayout& src, const TensorLayout& filter,
+            const TensorLayout& dst) = 0;
 
 protected:
     CanonizedFilterMeta check_exec(const TensorLayout& src,
@@ -297,17 +318,35 @@ public:
      */
     virtual void exec(_megdnn_tensor_in src, _megdnn_tensor_in filter,
                       _megdnn_tensor_in bias, _megdnn_tensor_in z,
-                      _megdnn_tensor_out dst, _megdnn_workspace workspace) = 0;
+                      _megdnn_tensor_out dst,
+                      const PreprocessedFilter* preprocessed_filter,
+                      _megdnn_workspace workspace) = 0;
+    virtual void exec_preprocess(const TensorLayout& src_layout,
+                                 _megdnn_tensor_in filter,
+                                 const TensorLayout& bias_layout,
+                                 const TensorLayout& z_layout,
+                                 const TensorLayout& dst_layout,
+                                 PreprocessedFilter* preprocessed_filter,
+                                 _megdnn_workspace workspace) = 0;
     void deduce_dtype(DType src, DType filter, DType bias, DType z, DType& dst);
     void deduce_layout(const TensorLayout& src, const TensorLayout& filter,
                        const TensorLayout& bias, const TensorLayout& z,
                        TensorLayout& dst);
 
-    virtual size_t get_workspace_in_bytes(const TensorLayout& src,
-                                          const TensorLayout& filter,
-                                          const TensorLayout& bias,
-                                          const TensorLayout& z,
-                                          const TensorLayout& dst) = 0;
+    virtual size_t get_workspace_in_bytes(
+            const TensorLayout& src, const TensorLayout& filter,
+            const TensorLayout& bias, const TensorLayout& z,
+            const TensorLayout& dst,
+            PreprocessedFilter* preprocessed_filter) = 0;
+    virtual size_t get_preprocess_workspace_in_bytes(
+            const TensorLayout& src, const TensorLayout& filter,
+            const TensorLayout& bias, const TensorLayout& z,
+            const TensorLayout& dst) = 0;
+    virtual SmallVector<TensorLayout> deduce_preprocessed_filter_layout(
+            const TensorLayout& src, const TensorLayout& filter,
+            const TensorLayout& bias, const TensorLayout& z,
+            const TensorLayout& dst) = 0;
+
     enum class BiasMode : uint32_t {
         NO_BIAS = 0,             //!< no bias
         BROADCAST_CHANNEL_BIAS,  //!< broadcast channel bias, [1, c, 1, 1]