Add grouped conv bwd weight multi d kernel (#1237)

* Add grouped conv bwd weight multi d kernel

* Reference fix

* Fix cmake files

* bwd weight scale only xdl

* Fixes

* Fix client conv fwd example

library/src/tensor_operation_instance/gpu/grouped_conv3d_bwd_weight_bilinear/CMakeLists.txt

+# ONLY XDL_KERNELS
+set(GROUPED_CONV3D_BWD_WEIGHT_BILINEAR
+    xdl/device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp
+    xdl/device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp
+    xdl/device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp)
+
+if((DTYPES MATCHES "fp8" AND DTYPES MATCHES "bf8" AND DTYPES MATCHES "fp16") OR NOT DEFINED DTYPES)
+    list(APPEND GROUPED_CONV3D_BWD_WEIGHT_BILINEAR
+      xdl/device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_f16_comp_bf8_fp8_instance.cpp)
+endif()
+
+add_instance_library(device_grouped_conv3d_bwd_weight_bilinear_instance ${GROUPED_CONV3D_BWD_WEIGHT_BILINEAR})

library/src/tensor_operation_instance/gpu/grouped_conv3d_bwd_weight_bilinear/xdl/device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+#include "ck/library/tensor_operation_instance/gpu/grouped_conv_bwd_weight/device_grouped_conv_bwd_weight_xdl_bilinear_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+// Compilation parameters for in[n, hi, wi, g, c] * wei[g, k, y, x, c] = out[n, ho, wo, g, k]
+void add_device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_bf16_f32_bf16_instances(
+    std::vector<std::unique_ptr<DeviceGroupedConvBwdWeightMultipleD<3,
+                                                                    NDHWGC,
+                                                                    GKZYXC,
+                                                                    NDHWGK,
+                                                                    Tuple<GKZYXC>,
+                                                                    BF16,
+                                                                    F32,
+                                                                    BF16,
+                                                                    Tuple<F32>,
+                                                                    PassThrough,
+                                                                    Bilinear,
+                                                                    PassThrough>>>& instances)
+{
+    // 1. Default
+    add_device_operation_instances(
+        instances,
+        device_grouped_conv_bwd_weight_xdl_c_shuffle_bf16_bilinear_instances<
+            3,
+            NDHWGC,
+            GKZYXC,
+            NDHWGK,
+            ConvBwdWeightDefault>{});
+    // 2. Filter1x1Stride1Pad0
+    add_device_operation_instances(
+        instances,
+        device_grouped_conv_bwd_weight_xdl_c_shuffle_bf16_bilinear_instances<
+            3,
+            NDHWGC,
+            GKZYXC,
+            NDHWGK,
+            ConvBwdWeightFilter1x1Stride1Pad0>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

library/src/tensor_operation_instance/gpu/grouped_conv3d_bwd_weight_bilinear/xdl/device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_f16_comp_bf8_fp8_instance.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/library/tensor_operation_instance/gpu/grouped_conv_bwd_weight/device_grouped_conv_bwd_weight_xdl_bilinear_instance.hpp"
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+void add_device_grouped_conv3d_bwd_weight_xdl_bilinear_ndhwgc_gkzyxc_ndhwgk_f16_comp_bf8_f8_instances(
+    std::vector<std::unique_ptr<DeviceGroupedConvBwdWeightMultipleD<3,
+                                                                    NDHWGC,
+                                                                    GKZYXC,
+                                                                    NDHWGK,
+                                                                    Tuple<GKZYXC>,
+                                                                    F16,
+                                                                    F16,
+                                                                    F16,
+                                                                    Tuple<F16>,
+                                                                    PassThrough,
+                                                                    Bilinear,
+                                                                    PassThrough,
+                                                                    BF8,
+                                                                    F8>>>& instances)
+{
+    // 1. Default
+    add_device_operation_instances(
+        instances,
+        device_grouped_conv_bwd_weight_xdl_c_shuffle_f16_comp_bf8_f8_bilinear_instances<
+            3,
+            NDHWGC,
+            GKZYXC,
+            NDHWGK,
+            ConvBwdWeightDefault>{});
+    // 2. Filter1x1Stride1Pad0
+    add_device_operation_instances(
+        instances,
+        device_grouped_conv_bwd_weight_xdl_c_shuffle_f16_comp_bf8_f8_bilinear_instances<
+            3,
+            NDHWGC,
+            GKZYXC,
+            NDHWGK,
+            ConvBwdWeightFilter1x1Stride1Pad0>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

library/src/tensor_operation_instance/gpu/grouped_conv3d_bwd_weight_scale/CMakeLists.txt

+# ONLY XDL_KERNELS
+set(GROUPED_CONV3D_BWD_WEIGHT_SCALE
+    xdl/device_grouped_conv3d_bwd_weight_xdl_scale_ndhwgc_gkzyxc_ndhwgk_f16_instance.cpp
+    xdl/device_grouped_conv3d_bwd_weight_xdl_scale_ndhwgc_gkzyxc_ndhwgk_f32_instance.cpp
+    xdl/device_grouped_conv3d_bwd_weight_xdl_scale_ndhwgc_gkzyxc_ndhwgk_bf16_instance.cpp)
+
+if((DTYPES MATCHES "fp8" AND DTYPES MATCHES "bf8" AND DTYPES MATCHES "fp16") OR NOT DEFINED DTYPES)
+    list(APPEND GROUPED_CONV3D_BWD_WEIGHT_SCALE
+      xdl/device_grouped_conv3d_bwd_weight_xdl_scale_ndhwgc_gkzyxc_ndhwgk_f16_comp_bf8_fp8_instance.cpp)
+endif()
+
+add_instance_library(device_grouped_conv3d_bwd_weight_scale_instance ${GROUPED_CONV3D_BWD_WEIGHT_SCALE})