Types are not propagated due to incorrect parameters

[river-li]

Sometimes the parameters of the entry point will directly be passed to a subfunction, without any parameter assignment instruction on the stack. Binary Ninja doesn't recognize these parameters and as a result, the type propagation gets lost here.

Vector35/binaryninja-api#3910

As the above issue mentioned, we probably can temporarily solve this by using set_call_type_adjustments API