How is multitasking performed in operating systems?

How is process-based multitasking achieved by u开发者_运维问答sing multi-threading in each process?

For example, consider when an operating system is running with two background process. Each process supports internally multi-threading features. Now, how does time slicing happen between and inside these processes, and how does time slicing happen between threads?

---

Look at publications by this man: http://en.wikipedia.org/wiki/Andrew_S._Tanenbaum

Or just feed your query into Google. There's many ways to skin the multi-tasking/multi-threading cat.

Come back when you have at least tried to find your own answers and ask some more specific questions.

---

One possible implementation is that the OS just schedules threads. When it switches to a thread, it obviously switches in the address space of the process the thread belongs to, but from a scheduling viewpoint the process is pretty much ignored (e.g., Windows works this way).