In current Library model implementation, each secure function includes the same inline entry point tfm_core_partition_request().
However, it is only required to make NS client check as inline.
The remaining steps in tfm_core_partition_request() increase the memory footprint when they are also inlined in secure functions.

It can save several KB in code size in Library model if the common inline part can be optimized