Fixes for merge and fork node

src/generate_root_modules.cpp

@@ -154,7 +154,8 @@ struct auto_gen_root_modules
     {
         const auto inputs = ins->inputs();
         if(std::any_of(inputs.begin(), inputs.end(), [&](auto input_ins) {
-               if(is_different_subgraph(input_ins, tid))
+               if(tass.find(input_ins) != tass.end() and
+                  tass.at(ins) != tid.value_or(std::numeric_limits<std::size_t>::max()))
                {
                    return true;
                }
@@ -168,7 +169,9 @@ struct auto_gen_root_modules
     {
         const auto outputs = ins->outputs();
         if(std::any_of(outputs.begin(), outputs.end(), [&](auto output_ins) {
-               if(is_different_subgraph(output_ins, tid) and output_ins->name() != "@return")
+               if(tass.find(output_ins) != tass.end() and
+                  tass.at(output_ins) != tid.value_or(std::numeric_limits<std::size_t>::max()) and
+                  output_ins->name() != "@return")
                {
                    return true;
                }
@@ -206,11 +209,7 @@ struct auto_gen_root_modules
             }
             if(not current_tid.has_value())
             {
-                if(tass.find(ins) == tass.end())
-                {
-                    continue;
-                }
-                else
+                if(tass.find(ins) != tass.end())
                 {
                     current_tid = std::make_optional<std::size_t>(tass.at(ins));
                     update_tid_counter(current_tid.value());
@@ -244,11 +243,11 @@ struct auto_gen_root_modules
                 }
                 else
                 {
-                    MIGRAPHX_THROW("Partition: this case shouldn't occur");
+                    MIGRAPHX_THROW("GenerateRootModules: this case shouldn't occur");
                 }
             }
 
-            if(skip_ins.find(ins) == skip_ins.end() and not ins->module_inputs().empty())
+            if(not ins->module_inputs().empty())
             {
                 std::vector<instruction_ref> same_tid_ins_vec_copy        = {};
                 std::unordered_set<instruction_ref> same_tid_ins_set_copy = {};
@@ -308,12 +307,11 @@ struct auto_gen_root_modules
             {
                 (*tmp)->debug_print();
             }
-            std::cout << "\n return ins: \n";
+            std::cout << "return ins: \n";
             for(auto tmp : iterator_for(return_ins))
             {
                 (*tmp)->debug_print();
             }
-            std::cout << "\n";
         }
 
         auto* tmod = p.create_module("target_mod_" + std::to_string(current_tid) + "_" +

test/multi_target/multitarget_test.cpp

@@ -134,6 +134,7 @@ bool check_compiled_program(const migraphx::program& p,
                             const std::vector<migraphx::target>& targets)
 {
     auto mods           = p.get_modules();
+    bool rot_ins        = false;
     bool check_compiled = true;
     for(const auto* mod : mods)
     {
@@ -141,6 +142,7 @@ bool check_compiled_program(const migraphx::program& p,
         {
             if(ins.name() == "run_on_target")
             {
+                rot_ins |= true;
                 auto* mod_input = ins.module_inputs().front();
                 std::size_t target_id =
                     ins.get_operator().to_value()["target_id"].to<std::size_t>();
@@ -156,7 +158,7 @@ bool check_compiled_program(const migraphx::program& p,
             }
         }
     }
-    return check_compiled;
+    return check_compiled and rot_ins;
 }
 
 TEST_CASE(multitarget_compile_cpu_gpu)