Porting OS-specific code to a microcontroller

[vreg]

Platforms vs architectures

What is the difference between something which is described as a platform, and a microcontroller/architecture?

For example, some codes are platform-independent and architecture specific. What does that mean? Can a code be architecture-independent but platform-specific?

I also read that sometimes platform and chip level code is mixed. What does that mean exactly? How is portability issue solved if platform and chip level code is mixed?

If a library is created for a specific operating system, what part of it would you need to change so that it can work with a microcontroller such as the Arduino 8-bit AVR platform? How can I make a library that can be compiled for any operating system or microcontroller as desired?

Any help is appreciated.

 

[kam1787]

platform is a more general term

as to changing a library to work with a different controller/architecture it all depends on the system calls and functions used

I doubt you could make a library that covers any OS or controller - certainly anyone that would be a value would be nigh impossible.

 

[vreg]

as to changing a library to work with a different controller/architecture it all depends on the system calls and functions used

What part of the code would you have to change to port a library that works with an embedded OS to work with a microcontroller platform like arduino?

I doubt you could make a library that covers any OS or controller - certainly anyone that would be a value would be nigh impossible.

I believe there are some libraries that work with different operating systems/platforms and you have to just pass some parameter values to them before compilation or something like that. How do they do that?