clean

cmake/EnableCompilerWarnings.cmake

@@ -66,7 +66,7 @@ else()
             -Wunreachable-code
             -Wunused
             -Wno-reserved-identifier
-	    #-Werror
+            -Werror
             -Wsign-compare
             -Wno-extra-semi-stmt
         )

example/15_grouped_gemm/grouped_gemm_xdl_splitk_fp16.cpp

@@ -63,7 +63,6 @@ int main(int argc, char* argv[])
 {
     ProblemSize problem_size;
     ExecutionConfig config;
-    ck::index_t kbatch = 1;
 
     problem_size.group_count = 16;
 
@@ -80,21 +79,19 @@ int main(int argc, char* argv[])
         problem_size.stride_Cs.push_back(problem_size.Ns[i]);
     }
 
-    if(argc == 5)
+    if(argc == 4)
     {
         config.do_verification = std::stoi(argv[1]);
         config.init_method     = std::stoi(argv[2]);
         config.time_kernel     = std::stoi(argv[3]);
-        kbatch                 = std::stoi(argv[4]);
     }
     else
     {
         printf("arg1: verification (0=no, 1=yes)\n");
         printf("arg2: initialization (0=no init, 1=integer value, 2=decimal value)\n");
         printf("arg3: time kernel (0=n0, 1=yes)\n");
-        printf("arg4: kbatch\n");
         exit(0);
     }
 
-    return !run_grouped_gemm(problem_size, config, kbatch);
+    return !run_grouped_gemm(problem_size, config);
 }

example/15_grouped_gemm/run_grouped_gemm_example.inc

@@ -20,7 +20,7 @@ struct ExecutionConfig final
     bool time_kernel     = false;
 };
 
-bool run_grouped_gemm(const ProblemSize& problem_size, const ExecutionConfig& config, ck::index_t kbatch = 1)
+bool run_grouped_gemm(const ProblemSize& problem_size, const ExecutionConfig& config)
 {
 #if defined(BUILD_INT4_EXAMPLE) && defined(CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4)
     static_assert(sizeof(ck::int4_t) == sizeof(int8_t));
@@ -172,8 +172,6 @@ bool run_grouped_gemm(const ProblemSize& problem_size, const ExecutionConfig& co
 
     gemm.SetWorkSpacePointer(&argument, gemm_desc_workspace.GetDeviceBuffer());
 
-    gemm.SetKBatchSize(argument, kbatch);
-
     if(!gemm.IsSupportedArgument(argument))
     {
         throw std::runtime_error(

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

@@ -521,8 +521,6 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         auto c_grid_buf = make_dynamic_buffer<AddressSpaceEnum::Global>(
             p_c_grid, c_grid_desc_mblock_mperblock_nblock_nperblock.GetElementSpaceSize());
 
-        // const auto K0 = a_b_k0_m_k1_grid_desc.GetLength(I1);
-
         // divide block work by [KBatch, M, N]
         const auto block_work_idx =
             block_2_ctile_map.CalculateBottomIndex(make_multi_index(get_block_1d_id()));

script/cmake-ck-dev.sh

@@ -11,8 +11,8 @@ cmake
 -D CMAKE_CXX_FLAGS="-std=c++17 -O3 -ftemplate-backtrace-limit=0  -fPIE  -Wno-gnu-line-marker      \
 -save-temps=$PWD"                                                                                 \
 -D CMAKE_BUILD_TYPE=Release                                                                       \
--D BUILD_DEV=OFF                                                                                   \
--D GPU_TARGETS="gfx90a"                                                                    \
+-D BUILD_DEV=ON                                                                                   \
+-D GPU_TARGETS="gfx908;gfx90a"                                                                    \
 -D CMAKE_VERBOSE_MAKEFILE:BOOL=ON                                                                 \
 -D USE_BITINT_EXTENSION_INT4=OFF                                                                  \
 ${MY_PROJECT_SOURCE}