Format

src/targets/gpu/compile_ops.cpp

@@ -163,7 +163,7 @@ struct compile_plan
                   << std::endl;
         std::vector<double> times;
         times.reserve(results.size());
-for(const auto& cr : results)
+        for(const auto& cr : results)
         {
             times.push_back(
                 time_op(*ctx, cr.replace.code_object, to_shapes(cr.ins->inputs()), 20).first);

src/targets/gpu/jit/ck_gemm.cpp

@@ -268,10 +268,10 @@ struct ck_gemm_compiler : compiler<ck_gemm_compiler>
 
     bool can_fold_batch(const std::vector<shape>& inputs) const
     {
-        const auto& a_shape   = inputs[0];
-        const auto& b_shape   = inputs[1];
-        auto rank      = a_shape.lens().size();
-        auto b_strides = b_shape.strides();
+        const auto& a_shape = inputs[0];
+        const auto& b_shape = inputs[1];
+        auto rank           = a_shape.lens().size();
+        auto b_strides      = b_shape.strides();
         return rank >= 3 and b_strides[rank - 3] == 0;
     }
 
@@ -282,8 +282,8 @@ struct ck_gemm_compiler : compiler<ck_gemm_compiler>
         const auto& b_shape = inputs[1];
         const auto& c_shape = inputs.back();
 
-        auto rank      = a_shape.lens().size();
-        
+        auto rank = a_shape.lens().size();
+
         auto batch_count = get_batch_count(c_shape);
         auto m           = c_shape.lens()[rank - 2];
         m                = can_fold_batch(inputs) ? m * batch_count : m;
@@ -293,9 +293,9 @@ struct ck_gemm_compiler : compiler<ck_gemm_compiler>
         const bool trans_a = transposed_matrix(a_shape);
         const bool trans_b = transposed_matrix(b_shape);
         const bool trans_e = transposed_matrix(c_shape);
-        const auto a_type = get_type(a_shape);
-        const auto b_type = get_type(b_shape);
-        const auto e_type = get_type(c_shape);
+        const auto a_type  = get_type(a_shape);
+        const auto b_type  = get_type(b_shape);
+        const auto e_type  = get_type(c_shape);
         std::vector<bool> ds_layout;
         std::transform(inputs.begin() + 2,
                        inputs.end() - 1,
@@ -333,10 +333,10 @@ struct ck_gemm_compiler : compiler<ck_gemm_compiler>
 
     operation compile_op(context& ctx, const std::vector<shape>& inputs, const value& v) const
     {
-        const auto& a_shape      = inputs[0];
-        const auto& b_shape      = inputs[1];
-        const auto& c_shape      = inputs.back();
-        auto tuning_value = v.get("tuning_value", 4);
+        const auto& a_shape = inputs[0];
+        const auto& b_shape = inputs[1];
+        const auto& c_shape = inputs.back();
+        auto tuning_value   = v.get("tuning_value", 4);
         if(not v.contains("tuning_value"))
             tuning_value = get_tuning_for({a_shape, b_shape, c_shape});
         auto batch_count = get_batch_count(c_shape);
@@ -344,7 +344,7 @@ struct ck_gemm_compiler : compiler<ck_gemm_compiler>
 
         const auto include_header   = problem.GetIncludeHeader();
         const auto solutions        = problem.GetSolutions(ctx.get_current_device().get_gfx_name());
-        const auto& solution         = solutions.at(tuning_value);
+        const auto& solution        = solutions.at(tuning_value);
         const auto template_str     = solution.template_str;
         const auto blocks_per_batch = solution.grid_size;
         const auto block_size       = solution.block_size;