diff --git a/include/ck/host_utility/hip_check_error.hpp b/include/ck/host_utility/hip_check_error.hpp
index af7bebd9d6..3e44faecb6 100644
--- a/include/ck/host_utility/hip_check_error.hpp
+++ b/include/ck/host_utility/hip_check_error.hpp
@@ -3,8 +3,10 @@
 
 #pragma once
 
+#include <sstream>
 #include <hip/hip_runtime.h>
 
+// To be removed, which really does not tell the location of failed HIP functional call
 inline void hip_check_error(hipError_t x)
 {
     if(x != hipSuccess)
@@ -15,3 +17,16 @@ inline void hip_check_error(hipError_t x)
         throw std::runtime_error(ss.str());
     }
 }
+
+#define HIP_CHECK_ERROR(retval_or_funcall)                                         \
+    do                                                                             \
+    {                                                                              \
+        hipError_t _tmpVal = retval_or_funcall;                                    \
+        if(_tmpVal != hipSuccess)                                                  \
+        {                                                                          \
+            std::ostringstream ostr;                                               \
+            ostr << "HIP Function Failed (" << __FILE__ << "," << __LINE__ << ") " \
+                 << hipGetErrorString(_tmpVal);                                    \
+            throw std::runtime_error(ostr.str());                                  \
+        }                                                                          \
+    } while(0)
diff --git a/include/ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp b/include/ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp
index e203468c61..37867f1eaa 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp
@@ -296,6 +296,28 @@ struct DeviceElementwiseImpl
     {
         return std::make_unique<Invoker>();
     };
+
+    std::string GetTypeString() const override
+    {
+        auto str = std::stringstream();
+
+        // clang-format off
+        str << "DeviceElementwiseImpl<" ;
+        str << "NumDim_" << NumDim << ","; 
+	str << "MPerThread_" << MPerThread << ","; 
+
+        str << "InScalarPerVector"; 
+        static_for<0, InScalarPerVectorSeq::Size(), 1>{}([&](auto i) { str << "_" << InScalarPerVectorSeq::At(i).value; });
+        str << ","; 
+        str << "OutScalarPerVector"; 
+        static_for<0, OutScalarPerVectorSeq::Size(), 1>{}([&](auto i) { str << "_" << OutScalarPerVectorSeq::At(i).value; });
+
+        str << ">";
+        // clang-format on
+
+        return str.str();
+    }
+
 }; // namespace device
 
 } // namespace device
diff --git a/library/include/ck/library/utility/host_common_util.hpp b/library/include/ck/library/utility/host_common_util.hpp
index 20a8f234db..16d6e2233a 100644
--- a/library/include/ck/library/utility/host_common_util.hpp
+++ b/library/include/ck/library/utility/host_common_util.hpp
@@ -22,7 +22,7 @@ static inline void dumpBufferToFile(const char* fileName, T* data, size_t dataNu
     std::ofstream outFile(fileName, std::ios::binary);
     if(outFile)
     {
-        outFile.write(reinterpret_cast<char*>(data), dataNumItems * sizeof(T));
+        outFile.write(reinterpret_cast<const char*>(data), dataNumItems * sizeof(T));
         outFile.close();
         std::cout << "Write output to file " << fileName << std::endl;
     }
diff --git a/library/include/ck/library/utility/host_tensor_generator.hpp b/library/include/ck/library/utility/host_tensor_generator.hpp
index 6b531c4c6f..6fd7ed8aa8 100644
--- a/library/include/ck/library/utility/host_tensor_generator.hpp
+++ b/library/include/ck/library/utility/host_tensor_generator.hpp
@@ -200,10 +200,11 @@ struct GeneratorTensor_3<ck::bf8_t>
 template <typename T>
 struct GeneratorTensor_4
 {
-    std::default_random_engine generator;
+    std::mt19937 generator;
     std::normal_distribution<float> distribution;
 
-    GeneratorTensor_4(float mean, float stddev) : generator(1), distribution(mean, stddev){};
+    GeneratorTensor_4(float mean, float stddev, unsigned int seed = 1)
+        : generator(seed), distribution(mean, stddev){};
 
     template <typename... Is>
     T operator()(Is...)