replace the ENV macro with CK_ENV (#1296)

include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl.hpp

@@ -514,7 +514,7 @@ struct DeviceGroupedGemm_Xdl : public DeviceGroupedGemm<ALayout,
 
             for(std::size_t i = 0; i < arg.gemm_desc_kernel_arg_.size(); i++)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "group: " << i << " arg.a_grid_desc_ak0_m_ak1_{"
                               << arg.gemm_desc_kernel_arg_[i].a_grid_desc_ak0_m_ak1_.GetLength(I0)

include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp

@@ -529,7 +529,7 @@ struct DeviceGroupedGemmXdlSplitKCShuffle : public DeviceGroupedGemmSplitK<ALayo
         if((ck::type_convert<ck::index_t>(arg.gemm_kernel_args_.size()) +
             arg.skipped_group_count_) != arg.group_count_)
         {
-            if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+            if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
             {
                 std::cout << "The group count is not equal to sum of skipped groups "
                              "and kernel args size!"
@@ -545,7 +545,7 @@ struct DeviceGroupedGemmXdlSplitKCShuffle : public DeviceGroupedGemmSplitK<ALayo
             bool group_arg_valid = GridwiseGemm::CheckValidity(a);
             if(not group_arg_valid)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "[" << __func__ << "] group id: " << i
                               << " has invalid GridwiseGemm settings!" << std::endl;

include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3.hpp

@@ -935,7 +935,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(!(karg.M % MPerBlock == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M value is not a multiple of MPerBlock! M: " << karg.M << " "
                               << __FILE__ << ":" << __LINE__ << ", in function: " << __func__
@@ -952,7 +952,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(!(karg.N % NPerBlock == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N value is not a multiple of NPerBlock! N: " << karg.N << " "
                               << __FILE__ << ":" << __LINE__ << ", in function: " << __func__
@@ -971,7 +971,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
             auto K_t = karg.KBatch * KPerBlock;
             if(!(karg.K % K_t == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K value is not a multiple of K_Batch * K0PerBlock * K1! K: "
                               << karg.K << " " << __FILE__ << ":" << __LINE__
@@ -995,7 +995,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.K % ABlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K (" << karg.K
                               << ") value is not a multiple of ABlockTransferSrcScalarPerVector ("
@@ -1009,7 +1009,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.M % ABlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M (" << karg.M
                               << ") value is not a multiple of ABlockTransferSrcScalarPerVector ("
@@ -1024,7 +1024,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.N % BBlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N (" << karg.N
                               << ") value is not a multiple of BBlockTransferSrcScalarPerVector ("
@@ -1038,7 +1038,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.K % BBlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K (" << karg.K
                               << ") value is not a multiple of BBlockTransferSrcScalarPerVector ("
@@ -1053,7 +1053,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.N % CShuffleBlockTransferScalarPerVector_NPerBlock != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N (" << karg.N
                               << ") value is not a multiple of "
@@ -1069,7 +1069,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.M % CShuffleBlockTransferScalarPerVector_NPerBlock != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M (" << karg.M
                               << ") value is not a multiple of "
@@ -1084,7 +1084,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
 
         if constexpr(is_same<remove_cvref_t<CDataType>, bhalf_t>::value)
         {
-            if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+            if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
             {
                 std::cout << " KBatch: " << karg.KBatch << " > 1 is not support yet" << __FILE__
                           << ":" << __LINE__ << ", in function: " << __func__ << std::endl;

include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3_multi_abd.hpp

@@ -1113,7 +1113,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(!(karg.M % MPerBlock == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M value is not a multiple of MPerBlock! M: " << karg.M << " "
                               << __FILE__ << ":" << __LINE__ << ", in function: " << __func__
@@ -1130,7 +1130,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(!(karg.N % NPerBlock == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N value is not a multiple of NPerBlock! N: " << karg.N << " "
                               << __FILE__ << ":" << __LINE__ << ", in function: " << __func__
@@ -1149,7 +1149,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
             auto K_t = karg.KBatch * KPerBlock;
             if(!(karg.K % K_t == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K value is not a multiple of K_Batch * K0PerBlock * K1! K: "
                               << karg.K << " " << __FILE__ << ":" << __LINE__
@@ -1173,7 +1173,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.K % ABlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K (" << karg.K
                               << ") value is not a multiple of ABlockTransferSrcScalarPerVector ("
@@ -1187,7 +1187,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.M % ABlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M (" << karg.M
                               << ") value is not a multiple of ABlockTransferSrcScalarPerVector ("
@@ -1202,7 +1202,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.N % BBlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N (" << karg.N
                               << ") value is not a multiple of BBlockTransferSrcScalarPerVector ("
@@ -1216,7 +1216,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.K % BBlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K (" << karg.K
                               << ") value is not a multiple of BBlockTransferSrcScalarPerVector ("
@@ -1231,7 +1231,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.N % CShuffleBlockTransferScalarPerVector_NPerBlock != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N (" << karg.N
                               << ") value is not a multiple of "
@@ -1247,7 +1247,7 @@ struct GridwiseGemm_xdl_cshuffle_v3
         {
             if(karg.M % CShuffleBlockTransferScalarPerVector_NPerBlock != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M (" << karg.M
                               << ") value is not a multiple of "

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

@@ -446,7 +446,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(!(karg.M % MPerBlock == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M value is not a multiple of MPerBlock! M: " << karg.M << " "
                               << __FILE__ << ":" << __LINE__ << ", in function: " << __func__
@@ -463,7 +463,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(!(karg.N % NPerBlock == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N value is not a multiple of NPerBlock! N: " << karg.N << " "
                               << __FILE__ << ":" << __LINE__ << ", in function: " << __func__
@@ -482,7 +482,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
             auto K_t = karg.k_batch * K0PerBlock * K1;
             if(!(karg.K % K_t == 0))
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K value is not a multiple of K_Batch * K0PerBlock * K1! K: "
                               << karg.K << " " << __FILE__ << ":" << __LINE__
@@ -496,7 +496,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(karg.K % ABlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K (" << karg.K
                               << ") value is not a multiple of ABlockTransferSrcScalarPerVector ("
@@ -510,7 +510,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(karg.M % ABlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M (" << karg.M
                               << ") value is not a multiple of ABlockTransferSrcScalarPerVector ("
@@ -525,7 +525,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(karg.N % BBlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N (" << karg.N
                               << ") value is not a multiple of BBlockTransferSrcScalarPerVector ("
@@ -539,7 +539,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(karg.K % BBlockTransferSrcScalarPerVector != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg K (" << karg.K
                               << ") value is not a multiple of BBlockTransferSrcScalarPerVector ("
@@ -554,7 +554,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(karg.N % CBlockTransferScalarPerVector_NWaveNPerXDL != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg N (" << karg.N
                               << ") value is not a multiple of "
@@ -569,7 +569,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         {
             if(karg.M % CBlockTransferScalarPerVector_NWaveNPerXDL != 0)
             {
-                if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+                if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
                 {
                     std::cout << "Arg M (" << karg.M
                               << ") value is not a multiple of "
@@ -584,7 +584,7 @@ struct GridwiseGemm_bk0mk1_bk0nk1_mn_xdlops_v2r4r2
         const auto num_k_loop = karg.K0Padded / K0PerBlock;
         if(!GridwiseGemmPipe::IsSupported(num_k_loop))
         {
-            if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+            if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
             {
                 std::cout << "The number of k loops (" << num_k_loop
                           << ") value is not supported by GridwiseGemm Pipeline."

include/ck/utility/env.hpp

@@ -124,7 +124,7 @@ struct EnvVar
 
 #define CK_DECLARE_ENV_VAR_STR(name) CK_DECLARE_ENV_VAR(name, std::string, "")
 
-#define ENV(name) \
+#define CK_ENV(name) \
     ck::env::name {}
 
 template <class EnvVar>

profiler/include/profiler/profile_grouped_gemm_fixed_nk_impl.hpp

@@ -88,7 +88,7 @@ bool profile_grouped_gemm_fixed_nk_impl(int do_verification,
 
         c_m_n_host_results.push_back(
             Tensor<CDataType>(f_host_tensor_descriptor(Ms[i], Ns[i], StrideCs[i], CLayout{})));
-        if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+        if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
         {
             std::cout << "group: " << i << " a_m_k[" << i << "]:" << a_m_k[i].mDesc << ", b_k_n["
                       << i << "]:" << b_k_n[i].mDesc << ", c_m_n_device_results[" << i

profiler/include/profiler/profile_grouped_gemm_impl.hpp

@@ -87,7 +87,7 @@ bool profile_grouped_gemm_impl(int do_verification,
 
         c_m_n_host_results.push_back(
             Tensor<CDataType>(f_host_tensor_descriptor(Ms[i], Ns[i], StrideCs[i], CLayout{})));
-        if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+        if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
         {
             std::cout << "group: " << i << " a_m_k[" << i << "]:" << a_m_k[i].mDesc << ", b_k_n["
                       << i << "]:" << b_k_n[i].mDesc << ", c_m_n_device_results[" << i

profiler/include/profiler/profile_grouped_gemm_tile_loop_impl.hpp

@@ -82,7 +82,7 @@ bool profile_grouped_gemm_tile_loop_impl(int do_verification,
             Tensor<CDataType>(f_host_tensor_descriptor(Ms[i], Ns[i], StrideCs[i], CLayout{})));
         c_m_n_host_results.push_back(
             Tensor<CDataType>(f_host_tensor_descriptor(Ms[i], Ns[i], StrideCs[i], CLayout{})));
-        if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+        if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
         {
             std::cout << "group: " << i << " a_m_k[" << i << "]:" << a_m_k[i].mDesc << ", b_k_n["
                       << i << "]:" << b_k_n[i].mDesc << ", c_m_n_device_results[" << i

profiler/include/profiler/profile_grouped_gemm_two_stage_impl.hpp

@@ -88,7 +88,7 @@ bool profile_grouped_gemm_two_stage_impl(int do_verification,
 
         c_m_n_host_results.push_back(
             Tensor<CDataType>(f_host_tensor_descriptor(Ms[i], Ns[i], StrideCs[i], CLayout{})));
-        if(ck::EnvIsEnabled(ENV(CK_LOGGING)))
+        if(ck::EnvIsEnabled(CK_ENV(CK_LOGGING)))
         {
             std::cout << "group: " << i << " a_m_k[" << i << "]:" << a_m_k[i].mDesc << ", b_k_n["
                       << i << "]:" << b_k_n[i].mDesc << ", c_m_n_device_results[" << i