tmp

example/27_layernorm/run_layernorm_example.inc

@@ -89,8 +89,8 @@ int run_groupnorm_example()
         ref_invoker.Run(ref_argument);
 
         y_dev.FromDevice(y.mData.data());
-        pass &= ck::utils::check_err(y, host_y, "Error: Incorrect results", 1e-3, 1e-3);
+        validator.check_err(y, host_y, "Error: Incorrect results", 1e-3, 1e-3);
     }
 
-    return (pass ? 0 : 1);
+    return !validator.is_success();
 }

example/28_grouped_gemm_bias_e_permute/grouped_gemm_bias_e_permute_xdl_fp16.cpp

@@ -458,9 +458,9 @@ int main(int argc, char* argv[])
                 }
             }
 
-            pass &= ck::utils::check_err(e_device_tensors[i], e_ms_ns_host_result);
+            validator.check_err(e_device_tensors[i], e_ms_ns_host_result);
         }
     }
 
-    return pass ? 0 : 1;
+    return !validator.is_success();
 }

example/29_batched_gemm_bias_e_permute/batched_gemm_bias_e_permute_wmma_fp16.cpp

@@ -424,7 +424,8 @@ int main(int argc, char* argv[])
             }
         }
 
-        return ck::utils::check_err(e_gs_ms_ns_device_result, e_gs_ms_ns_host_result) ? 0 : 1;
+        validator.check_err(e_gs_ms_ns_device_result, e_gs_ms_ns_host_result);
+        return !validator.is_success();
     }
 
     return 0;

example/29_batched_gemm_bias_e_permute/batched_gemm_bias_e_permute_xdl_fp16.cpp

@@ -390,7 +390,8 @@ int main(int argc, char* argv[])
             }
         }
 
-        return ck::utils::check_err(e_gs_ms_ns_device_result, e_gs_ms_ns_host_result) ? 0 : 1;
+        validator.check_err(e_gs_ms_ns_device_result, e_gs_ms_ns_host_result);
+        return !validator.is_success();
     }
 
     return 0;

example/30_grouped_conv_fwd_multiple_d/run_grouped_conv_fwd_bias_relu_add_example.inc

@@ -256,12 +256,13 @@ bool run_grouped_conv_fwd_bias_relu_add(const ExecutionConfig& config,
 #ifdef BUILD_INT4_EXAMPLE
         const Tensor<OutUserDataType> out_device_converted(out_device);
 
-        return ck::utils::check_err(
+        validator.check_err(
             out_device_converted, out_host, "Error: incorrect results!", 1e-5f, 1e-4f);
 #else
-        return ck::utils::check_err(
+        validator.check_err(
             out_device, out_host, "Error: incorrect results!", 1e-5f, 1e-4f);
 #endif
+return validator.is_success();
     }
 
     return true;

example/30_grouped_conv_fwd_multiple_d/run_grouped_conv_fwd_bias_relu_add_wmma_example.inc

@@ -254,12 +254,13 @@ bool run_grouped_conv_fwd_bias_relu_add(const ExecutionConfig& config,
 #ifdef BUILD_INT4_EXAMPLE
         const Tensor<OutUserDataType> out_device_converted(out_device);
 
-        return ck::utils::check_err(
+        validator.check_err(
             out_device_converted, out_host, "Error: incorrect results!", 1e-5f, 1e-4f);
 #else
-        return ck::utils::check_err(
+        validator.check_err(
             out_device, out_host, "Error: incorrect results!", 1e-5f, 1e-4f);
 #endif
+return validator.is_success();
     }
 
     return true;

example/30_grouped_conv_fwd_multiple_d/run_grouped_conv_fwd_example.inc

@@ -191,12 +191,13 @@ bool run_grouped_conv_fwd(const ExecutionConfig& config,
 #ifdef BUILD_INT4_EXAMPLE
         const Tensor<OutUserDataType> out_device_converted(out_device);
 
-        return ck::utils::check_err(
+        validator.check_err(
             out_device_converted.mData, out_host.mData, "Error: incorrect results!", 1e-5f, 1e-4f);
 #else
-        return ck::utils::check_err(
+         validator.check_err(
             out_device.mData, out_host.mData, "Error: incorrect results!", 1e-5f, 1e-4f);
 #endif
+return validator.is_success();
     }
 
     return true;

example/31_batched_gemm_gemm/run_batched_gemm_gemm_example.inc

@@ -270,7 +270,8 @@ bool run_batched_gemm_gemm_example(int argc, char* argv[])
         c_g_m_o_device_buf.FromDevice(c_g_m_o_device_result.mData.data());
 #endif
 
-        return ck::utils::check_err(c_g_m_o_device_result, c_g_m_o_host_result);
+        validator.check_err(c_g_m_o_device_result, c_g_m_o_host_result);
+        return validator.is_success();
     }
 
     return true;

example/32_batched_gemm_scale_softmax_gemm/run_batched_gemm_scale_softmax_gemm.inc

@@ -254,7 +254,8 @@ int run(int argc, char* argv[])
 
         ref_gemm1_invoker.Run(ref_gemm1_argument);
 
-        return ck::utils::check_err(c_g_m_o_device_result.mData, c_g_m_o_host_result.mData) ? 0 : 1;
+         validator.check_err(c_g_m_o_device_result.mData, c_g_m_o_host_result.mData);
+         return !validator.is_success();
     }
 
     return 0;

example/32_batched_gemm_scale_softmax_gemm/run_batched_gemm_scale_softmax_gemm_permute.inc

@@ -265,13 +265,12 @@ int run(int argc, char* argv[])
             atol = 1e-2;
         }
 
-        return ck::utils::check_err(c_gs_ms_os_device_result.mData,
+        validator.check_err(c_gs_ms_os_device_result.mData,
                                     c_gs_ms_os_host_result.mData,
                                     "Error: Incorrect results!",
                                     rtol,
-                                    atol)
-                   ? 0
-                   : 1;
+                                    atol);
+        return !validator.is_success();
     }
 
     return 0;

example/32_batched_gemm_scale_softmax_gemm/run_grouped_gemm_scale_softmax_gemm_permute.inc

@@ -223,7 +223,7 @@ int run(int argc, char* argv[])
     std::cout << "Perf: " << ave_time << " ms, " << tflops << " TFlops, " << gb_per_sec << " GB/s, "
               << gemm.GetTypeString() << std::endl;
 
-    bool pass = true;
+    ck::utils::CorrectnessValidator validator;
     if(do_verification)
     {
         for(std::size_t i = 0; i < group_count; i++)
@@ -309,11 +309,9 @@ int run(int argc, char* argv[])
                 self(idx) = c_g_m_o_host_result(g, idx[2], idx[3]);
             });
 
-            bool pass_ =
-                ck::utils::check_err(c_gs_ms_os_device_result.mData, c_gs_ms_os_host_result.mData);
-            pass &= pass_;
+                validator.check_err(c_gs_ms_os_device_result.mData, c_gs_ms_os_host_result.mData);
         }
     }
 
-    return pass ? 0 : 1;
+    return !validator.is_success();
 }

example/33_multiple_reduce/dual_reduce_common.hpp

@@ -300,15 +300,15 @@ int mean_meansquare_dual_reduce_test(size_t n,
     std::cout << "Perf: " << avg_time << " ms, " << gb_per_sec << " GB/s, " << reduce_name
               << std::endl;
 
-    bool pass = true;
+    ck::utils::CorrectnessValidator validator;
 
     if(do_verification)
     {
         mean_dev.FromDevice(mean.mData.data());
         meansquare_dev.FromDevice(meansquare.mData.data());
-        pass = pass && ck::utils::check_err(mean, mean_ref);
-        pass = pass && ck::utils::check_err(meansquare, meansquare_ref);
+        validator.check_err(mean, mean_ref);
+        validator.check_err(meansquare, meansquare_ref);
     };
 
-    return (pass ? 0 : 1);
+    return !validator.is_success();
 }

example/34_batchnorm/batchnorm_backward_nhwc.cpp

@@ -338,7 +338,7 @@ bool bnorm_bwd_nhwc_test(bool do_verification,
     else
         (void)invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, time_kernel});
 
-    bool pass = true;
+    ck::utils::CorrectnessValidator validator;
 
     if(do_verification)
     {
@@ -394,20 +394,20 @@ bool bnorm_bwd_nhwc_test(bool do_verification,
         dbias_dev.FromDevice(dbias.data());
 
         // clang-format off
-        pass = pass && ck::utils::check_err(dbias.mData, dbias_ref.mData, "dBias result:", 2e-4, 2e-4);
-        pass = pass && ck::utils::check_err(dscale.mData, dscale_ref.mData, "dScale result:", 2e-4, 2e-4);
-        pass = pass && ck::utils::check_err(dx.mData, dx_ref.mData, "dx result:");
+        validator.check_err(dbias.mData, dbias_ref.mData, "dBias result:", 2e-4, 2e-4);
+        validator.check_err(dscale.mData, dscale_ref.mData, "dScale result:", 2e-4, 2e-4);
+        validator.check_err(dx.mData, dx_ref.mData, "dx result:");
         // clang-format on
     };
 
-    return (pass);
+    return validator.is_success();
 };
 
 static const double epsilon = std::numeric_limits<float>::epsilon();
 
 int main(int argc, char* argv[])
 {
-    bool pass = true;
+    ck::utils::CorrectnessValidator validator;
 
     if(argc > 1)
     {

example/34_batchnorm/batchnorm_forward_inferring_nhwc.cpp

@@ -308,17 +308,17 @@ bool bnorm_infer_nhwc_test(bool do_verification,
         (void)invoker_ptr_ref->Run(argument_ptr_ref.get());
 
         y_dev.FromDevice(y.mData.data());
-        pass = pass && ck::utils::check_err(y, y_ref);
+        validator.check_err(y, y_ref);
     };
 
-    return (pass);
+    return validator.is_success();
 };
 
 static const double epsilon = std::numeric_limits<float>::epsilon();
 
 int main(int argc, char* argv[])
 {
-    bool pass = true;
+ck::utils::CorrectnessValidator validator;
 
     if(argc > 1)
     {

example/34_batchnorm/batchnorm_forward_training_nhwc.cpp

@@ -362,7 +362,7 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
     else
         (void)invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, time_kernel});
 
-    bool pass = true;
+ck::utils::CorrectnessValidator validator;
 
     if(do_verification)
     {
@@ -414,7 +414,7 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
         (void)invoker_ptr_ref->Run(argument_ptr_ref.get());
 
         y_dev.FromDevice(y.mData.data());
-        pass = pass && ck::utils::check_err(y, y_ref, "Incorrect normalized output values");
+        validator.check_err(y, y_ref, "Incorrect normalized output values");
 
         if(updateMovingAverage)
         {
@@ -424,10 +424,10 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
             resultRunningMean_dev.FromDevice(resultRunningMean.mData.data());
             resultRunningVariance_dev.FromDevice(resultRunningVariance.mData.data());
 
-            pass = pass && ck::utils::check_err(resultRunningMean,
+            validator.check_err(resultRunningMean,
                                                 resultRunningMean_ref,
                                                 "Incorrect running mean values");
-            pass = pass && ck::utils::check_err(resultRunningVariance,
+            validator.check_err(resultRunningVariance,
                                                 resultRunningVariance_ref,
                                                 "Incorrect running variance values");
         };
@@ -442,15 +442,15 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
             resultSaveMean_dev.FromDevice(resultSaveMean.mData.data());
             resultSaveInvVariance_dev.FromDevice(resultSaveInvVariance.mData.data());
 
-            pass = pass && ck::utils::check_err(
+            validator.check_err(
                                resultSaveMean, resultSaveMean_ref, "Incorrect saved mean values");
-            pass = pass && ck::utils::check_err(resultSaveInvVariance,
+            validator.check_err(resultSaveInvVariance,
                                                 resultSaveInvVariance_ref,
                                                 "Incorrect saved invvariance values");
         };
     };
 
-    return (pass);
+    return validator.is_success();
 };
 
 const double epsilon              = std::numeric_limits<float>::epsilon();
@@ -584,7 +584,7 @@ int main(int argc, char* argv[])
                                                             averageFactor,
                                                             epsilon);
 
-        pass = pass && bnorm_fwd_nhwc_test<ck::half_t, float, false>(true,
+        bnorm_fwd_nhwc_test<ck::half_t, float, false>(true,
                                                                      2,
                                                                      false, // don't time kernel
                                                                      {128, 16, 3, 1024},

example/34_batchnorm/batchnorm_forward_training_nhwc_obsolete.cpp

@@ -362,7 +362,7 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
     else
         (void)invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, time_kernel});
 
-    bool pass = true;
+ck::utils::CorrectnessValidator validator;
 
     if(do_verification)
     {
@@ -414,7 +414,7 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
         (void)invoker_ptr_ref->Run(argument_ptr_ref.get());
 
         y_dev.FromDevice(y.mData.data());
-        pass = pass && ck::utils::check_err(y, y_ref, "Incorrect normalized output values");
+        validator.check_err(y, y_ref, "Incorrect normalized output values");
 
         if(updateMovingAverage)
         {
@@ -424,10 +424,10 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
             resultRunningMean_dev.FromDevice(resultRunningMean.mData.data());
             resultRunningVariance_dev.FromDevice(resultRunningVariance.mData.data());
 
-            pass = pass && ck::utils::check_err(resultRunningMean,
+            validator.check_err(resultRunningMean,
                                                 resultRunningMean_ref,
                                                 "Incorrect running mean values");
-            pass = pass && ck::utils::check_err(resultRunningVariance,
+            validator.check_err(resultRunningVariance,
                                                 resultRunningVariance_ref,
                                                 "Incorrect running variance values");
         };
@@ -442,15 +442,15 @@ bool bnorm_fwd_nhwc_test(bool do_verification,
             resultSaveMean_dev.FromDevice(resultSaveMean.mData.data());
             resultSaveInvVariance_dev.FromDevice(resultSaveInvVariance.mData.data());
 
-            pass = pass && ck::utils::check_err(
+            validator.check_err(
                                resultSaveMean, resultSaveMean_ref, "Incorrect saved mean values");
-            pass = pass && ck::utils::check_err(resultSaveInvVariance,
+            validator.check_err(resultSaveInvVariance,
                                                 resultSaveInvVariance_ref,
                                                 "Incorrect saved invvariance values");
         };
     };
 
-    return (pass);
+    return validator.is_success();
 };
 
 const double epsilon              = std::numeric_limits<float>::epsilon();

example/35_splitK_gemm/run_splitK_gemm_example.inc

@@ -121,7 +121,7 @@ bool run_splitK_gemm(const ProblemSize& problem_size, const ExecutionConfig& con
     }
 
     invoker.Run(argument, StreamConfig{nullptr, false});
-    bool pass = true;
+ck::utils::CorrectnessValidator validator;
 
     if(config.do_verification)
     {
@@ -146,12 +146,12 @@ bool run_splitK_gemm(const ProblemSize& problem_size, const ExecutionConfig& con
 
         if(std::is_same<CDataType, ck::half_t>::value)
         {
-            pass &= ck::utils::check_err(
+            validator.check_err(
                 c_m_n_device_result, c_m_n_host_result, "fp16 incorrect result", 3e-3, 1e-3);
         }
         else
         {
-            pass &= ck::utils::check_err(c_m_n_device_result, c_m_n_host_result);
+            validator.check_err(c_m_n_device_result, c_m_n_host_result);
         }
     }
 
@@ -169,7 +169,7 @@ bool run_splitK_gemm(const ProblemSize& problem_size, const ExecutionConfig& con
                   << " GB/s, " << gemm.GetTypeString() << std::endl;
     }
 
-    return pass;
+    return validator.is_success();
 }
 
 bool run_splitK_gemm_example(int argc, char* argv[])

example/36_sparse_embedding/sparse_embedding3_forward_layernorm.cpp

@@ -155,7 +155,7 @@ int main()
         auto invoker_ptr = device_instance.MakeInvokerPointer();
         float time_ms    = invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, time_kernel});
 
-        bool pass = true;
+ck::utils::CorrectnessValidator validator;
         {
             Tensor<OutType> out_from_dev(f_host_tensor_desc_2d(index_length, current_dim));
             ReferenceInstance ref;
@@ -176,7 +176,7 @@ int main()
             ref_invoker.Run(ref_argument);
 
             out_dev.FromDevice(out_from_dev.mData.data());
-            pass &= ck::utils::check_err(out_from_dev, out, "Error: Incorrect results", 1e-3, 1e-3);
+            validator.check_err(out_from_dev, out, "Error: Incorrect results", 1e-3, 1e-3);
         }
 
         double total_read = current_dim * index_length * 3 * sizeof(EmbType) +
@@ -186,7 +186,7 @@ int main()
         double gbps        = (total_read + total_write) / time_ms / 1e6;
 
         std::cout << ", total bytes:" << (total_read + total_write) << ", time:" << time_ms
-                  << ", gbps:" << gbps << ", valid:" << (pass ? "y" : "n") << std::endl
+                  << ", gbps:" << gbps << ", valid:" << (validator.is_success() ? "y" : "n") << std::endl
                   << std::flush;
     });
 

example/37_batched_gemm_add_add_relu_gemm_add/batched_gemm_add_add_relu_gemm_add_xdl_fp16.cpp

@@ -513,8 +513,9 @@ int main(int argc, char* argv[])
         e1_g_m_o_host_result.ForEach([&](auto&, auto idx) {
             cde1_element_op(e1_g_m_o_host_result(idx), c1_g_m_o(idx), d1_g_m_o(idx));
         });
-
-        return ck::utils::check_err(e1_g_m_o_device_result, e1_g_m_o_host_result) ? 0 : 1;
+ck::utils::CorrectnessValidator validator;
+        validator.check_err(e1_g_m_o_device_result, e1_g_m_o_host_result);
+        return !validator.is_success();
     }
 
     return 0;

example/38_grouped_conv_bwd_data_multiple_d/run_grouped_conv_bwd_data_bias_relu_example.inc

@@ -156,8 +156,9 @@ bool run_conv_bwd_data_bias_relu(const ExecutionConfig& config,
             [&](auto&, auto idx) { in_element_op(in_host(idx), c_host(idx), bias(idx)); });
 
         in_device_buf.FromDevice(in_device.mData.data());
-
-        return ck::utils::check_err(in_device, in_host);
+ck::utils::CorrectnessValidator validator;
+        validator.check_err(in_device, in_host);
+        return validator.is_success();
     }
 
     return true;