Use device name

src/targets/gpu/device_name.cpp

@@ -43,6 +43,11 @@ auto get_arch_name(rank<1>, const HipDeviceProp& props) -> decltype(std::string(
     return std::string(props.gcnArchName);
 }
 
+std::string get_arch_name(const hipDeviceProp_t& props)
+{
+    return get_arch_name(rank<1>{}, props);
+}
+
 int get_device_id()
 {
     int device;
@@ -58,7 +63,7 @@ std::string get_device_name()
     auto status = hipGetDeviceProperties(&props, get_device_id());
     if(status != hipSuccess)
         MIGRAPHX_THROW("Failed to get device properties");
-    return get_arch_name(rank<1>{}, props);
+    return get_arch_name(props);
 }
 
 } // namespace gpu

src/targets/gpu/include/migraphx/gpu/context.hpp

@@ -170,7 +170,9 @@ struct hip_device
 
     std::size_t stream_id() const { return current_stream; }
 
-    std::string get_device_name() const { return device_props.gcnArchName; }
+    std::string get_device_name() const { return get_arch_name(device_props); }
+    
+    std::string get_gfx_name() const { return trim(split_string(get_device_name(), ':').front()); }
 
     std::size_t get_device_major() const { return device_props.major; }
 

src/targets/gpu/include/migraphx/gpu/device_name.hpp

@@ -27,10 +27,14 @@
 #include <migraphx/config.hpp>
 #include <string>
 
+struct hipDeviceProp_t;
+
 namespace migraphx {
 inline namespace MIGRAPHX_INLINE_NS {
 namespace gpu {
 
+std::string get_arch_name(const hipDeviceProp_t& props);
+
 std::string get_device_name();
 
 int get_device_id();

src/targets/gpu/jit/ck_gemm.cpp

@@ -333,7 +333,7 @@ struct ck_gemm_compiler : compiler<ck_gemm_compiler>
                                                          cde_op};
     }
 
-    operation compile_op(context& /* ctx */, const std::vector<shape>& inputs, const value& v) const
+    operation compile_op(context& ctx, const std::vector<shape>& inputs, const value& v) const
     {
         auto a_shape      = inputs[0];
         auto b_shape      = inputs[1];
@@ -343,7 +343,7 @@ struct ck_gemm_compiler : compiler<ck_gemm_compiler>
         auto problem      = create_problem(inputs, v);
 
         const auto include_header   = problem.GetIncludeHeader();
-        const auto solutions        = problem.GetSolutions("gfx90a");
+        const auto solutions        = problem.GetSolutions(ctx.get_current_device().get_gfx_name());
         const auto solution         = solutions.at(tuning_value);
         const auto template_str     = solution.template_str;
         const auto blocks_per_batch = solution.grid_size;