Remove unused or lefover debugging code.

example/01_gemm/run_gemm_example.inc

@@ -30,28 +30,7 @@ bool run_gemm(const ProblemSize& problem_size, const ExecutionConfig& config)
 
     switch(config.init_method)
     {
-    case 0:
-        ck::utils::FillConstant<ADataType>{1.f}(a_m_k);
-        ck::utils::FillConstant<BDataType>{0.f}(b_k_n);
-
-        // for (ck::index_t m = 0; m < M; ++m)
-        // {
-        //     for (ck::index_t k = 0; k < K; ++k)
-        //     {
-        //         a_m_k(m, k) = (m * M + k) % 5;
-        //     }
-        // }
-
-        for(ck::index_t n = 0; n < N; ++n)
-        {
-            for(ck::index_t k = 0; k < K; ++k)
-            {
-                if(n == k)
-                    b_k_n(k, n) = n * 2;
-            }
-        }
-
-        break;
+    case 0: break;
     case 1:
         ck::utils::FillUniformDistributionIntegerValue<ADataType>{-1.f, 3.f}(a_m_k);
         ck::utils::FillUniformDistributionIntegerValue<BDataType>{-1.f, 3.f}(b_k_n);

include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_v2r4r2.hpp

@@ -508,13 +508,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         return true;
     }
 
-    // __host__ __device__ static auto GetKPad(index_t K, index_t KBatch)
-    // {
-    //     const index_t K0   = math::integer_divide_ceil(K, K1 * K0PerBlock * KBatch) * K0PerBlock;
-    //     const index_t KPad = KBatch * K0 * K1;
-    //     return KPad;
-    // }
-
     __host__ __device__ static constexpr bool CalculateHasMainK0BlockLoop(index_t K0)
     {
         const index_t num_loop = K0 / K0PerBlock;