llvm/execution: Use 'evaluate' function params when constructing simu…

…lation input (#2467)

The original code assumed 'run' variant would be called.
This is not the case for parallel evaluate that only needs
'run, simulation' variant, resulting in redundant code
generation and compiler calls.

Instead, use the 'evaluate' compiled function that provides the
same binary type of the argument at a different offset.

Tested by asserting that Python-{LLVM,PTX} tests only generate
'run, simulation' variant of the composition function.

Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>

psyneulink/core/llvm/execution.py

@@ -563,8 +563,11 @@ def cuda_execute(self, inputs):
 
     # Methods used to accelerate "Run"
 
-    def _get_run_input_struct(self, inputs, num_input_sets):
-        input_type = self._bin_run_func.byref_arg_types[3]
+    def _get_run_input_struct(self, inputs, num_input_sets, arg=3):
+        # Callers that override input arg, should ensure that _bin_func is not None
+        bin_f = self._bin_run_func if arg == 3 else self._bin_func
+
+        input_type = bin_f.byref_arg_types[arg]
         c_input = (input_type * num_input_sets) * len(self._execution_contexts)
         if len(self._execution_contexts) == 1:
             inputs = [inputs]
@@ -694,8 +697,8 @@ def _prepare_evaluate(self, inputs, num_input_sets, num_evaluations):
         ct_comp_state = self._get_compilation_param('_eval_state', '_get_state_initializer', 1)
         ct_comp_data = self._get_compilation_param('_eval_data', '_get_data_initializer', 6)
 
-        # Construct input variable
-        ct_inputs = self._get_run_input_struct(inputs, num_input_sets)
+        # Construct input variable, the 5th parameter of the evaluate function
+        ct_inputs = self._get_run_input_struct(inputs, num_input_sets, 5)
 
         # Output ctype
         out_ty = bin_func.byref_arg_types[4] * num_evaluations