merge develop

include/ck/tensor_operation/gpu/device/device_pool2d_fwd_nhwc_nhwc.hpp

@@ -204,7 +204,7 @@ struct DevicePool2dFwd_Input_N_Hi_Wi_C_Output_N_Ho_Wo_C : public DevicePool2dFwd
 
     struct Invoker : public BaseInvoker
     {
-        float Run(const Argument& arg, int nrepeat = 1)
+        float Run(const Argument& arg, const StreamConfig& stream_config = StreamConfig{})
         {
             using gridwise_reduce = GridwiseReduction_mk_to_m_threadwise<InDataType,
                                                                          OutDataType,
@@ -241,8 +241,8 @@ struct DevicePool2dFwd_Input_N_Hi_Wi_C_Output_N_Ho_Wo_C : public DevicePool2dFwd
 
             const index_t grid_size = (ReduceM / ReduceM_BlockTileSize);
 
-            return launch_and_time_kernel(kernel,
-                                          nrepeat,
+            return launch_and_time_kernel(stream_config,
+                                          kernel,
                                           dim3(grid_size),
                                           dim3(BlockSize),
                                           0,
@@ -257,9 +257,10 @@ struct DevicePool2dFwd_Input_N_Hi_Wi_C_Output_N_Ho_Wo_C : public DevicePool2dFwd
                                           arg.p_out_indices_dev_);
         }
 
-        float Run(const BaseArgument* p_arg, int nrepeat = 1) override
+        float Run(const BaseArgument* p_arg,
+                  const StreamConfig& stream_config = StreamConfig{}) override
         {
-            return Run(*dynamic_cast<const Argument*>(p_arg), nrepeat);
+            return Run(*dynamic_cast<const Argument*>(p_arg), stream_config);
         }
     };