Make fail/pass logic consistent within 01_gemm folder

Removed multiple negations in fail/pass logic to propagate `true` as the success indicator.

example/01_gemm/gemm_xdl_bf16_v3.cpp

@@ -45,4 +45,4 @@ using ReferenceGemmInstance = ck::tensor_operation::host::
 
 #include "run_gemm_example_v2.inc"
 
-int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv)) ? -1 : 0; }
+int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv) ? 0 : -1); }

example/01_gemm/gemm_xdl_fp16_fp8_v3.cpp

@@ -50,4 +50,4 @@ using ReferenceGemmInstance = ck::tensor_operation::host::ReferenceGemm<ADataTyp
 
 #include "run_gemm_example_v2.inc"
 
-int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv)) ? -1 : 0; }
+int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv) ? 0 : -1); }

example/01_gemm/gemm_xdl_fp16_streamk_v3.cpp

@@ -47,5 +47,5 @@ using ReferenceGemmInstance = ck::tensor_operation::host::
 
 int main(int argc, char* argv[])
 {
-    return (run_gemm_universal_streamk_example(argc, argv)) ? -1 : 0;
+    return (run_gemm_universal_streamk_example(argc, argv) ? 0 : -1);
 }

example/01_gemm/gemm_xdl_fp16_v3.cpp

@@ -45,4 +45,4 @@ using ReferenceGemmInstance = ck::tensor_operation::host::
 
 #include "run_gemm_example_v2.inc"
 
-int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv)) ? -1 : 0; }
+int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv) ? 0 : -1); }

example/01_gemm/gemm_xdl_fp8.cpp

@@ -53,4 +53,4 @@ using ReferenceGemmInstanceGPU = ck::tensor_operation::device::ReferenceGemm<ALa
 
 #include "run_gemm_example.inc"
 
-int main(int argc, char* argv[]) { return run_gemm_example(argc, argv) ? 0 : -1; }
+int main(int argc, char* argv[]) { return (run_gemm_example(argc, argv) ? 0 : -1); }

example/01_gemm/gemm_xdl_fp8_v3.cpp

@@ -45,4 +45,4 @@ using ReferenceGemmInstance = ck::tensor_operation::host::
 
 #include "run_gemm_example_v2.inc"
 
-int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv)) ? -1 : 0; }
+int main(int argc, char* argv[]) { return (run_gemm_splitk_example(argc, argv) ? 0 : -1); }

example/01_gemm/run_gemm_example_streamk_v2.inc

@@ -237,7 +237,7 @@ bool run_gemm(const ProblemType& problem_size, const ExecutionConfig& config)
     {
         std::cerr << gemm.GetTypeString() << " does not support this problem" << std::endl;
 
-        return true;
+        return false;
     }
 
     bool pass = true;
@@ -293,7 +293,7 @@ bool run_gemm(const ProblemType& problem_size, const ExecutionConfig& config)
         std::cout << "Perf: " << ave_time << " ms, " << tflops << " TFlops, " << gb_per_sec
                   << " GB/s, " << gemm.GetTypeString() << std::endl;
     }
-    return !pass;
+    return pass;
 }
 
 bool run_gemm_universal_streamk_example(int argc, char* argv[])
@@ -301,5 +301,5 @@ bool run_gemm_universal_streamk_example(int argc, char* argv[])
     ProblemSizeStreamK_universal problem_size;
     ExecutionConfig config;
 
-    return !parse_cmd_args(argc, argv, problem_size, config) || run_gemm(problem_size, config);
+    return parse_cmd_args(argc, argv, problem_size, config) && run_gemm(problem_size, config);
 }

example/01_gemm/run_gemm_example_v2.inc

@@ -216,7 +216,7 @@ bool run_gemm(const ProblemType& problem_size, const ExecutionConfig& config)
     {
         std::cerr << gemm.GetTypeString() << " does not support this problem" << std::endl;
 
-        return true;
+        return false;
     }
 
     bool pass = true;
@@ -273,7 +273,7 @@ bool run_gemm(const ProblemType& problem_size, const ExecutionConfig& config)
         std::cout << "Perf: " << ave_time << " ms, " << tflops << " TFlops, " << gb_per_sec
                   << " GB/s, " << gemm.GetTypeString() << std::endl;
     }
-    return !pass;
+    return pass;
 }
 
 bool run_gemm_splitk_example(int argc, char* argv[])
@@ -281,5 +281,5 @@ bool run_gemm_splitk_example(int argc, char* argv[])
     ProblemSizeSplitK problem_size;
     ExecutionConfig config;
 
-    return !parse_cmd_args(argc, argv, problem_size, config) || run_gemm(problem_size, config);
+    return parse_cmd_args(argc, argv, problem_size, config) && run_gemm(problem_size, config);
 }

include/ck/utility/amd_ck_fp8.hpp

@@ -382,6 +382,8 @@ template <index_t N>
 struct non_native_vector_base<f8_ocp_t, N>
 {
     using data_t = f8_ocp_t::data_type;
+    static_assert(sizeof(f8_ocp_t) == sizeof(data_t),
+                  "non_native_vector_base storage size mismatch");
     using data_v = data_t __attribute__((ext_vector_type(sizeof(data_t) * N)));
     using type   = non_native_vector_base<f8_ocp_t, N>;