💼 Process

A process is an instance of a program that is being executed by the ⚙️ Operating System. Each process has its own memory (code, heap, stack), registers, and other resources.

Stack and Heap §

The stack contains local variables and information for local functions, all organized in a stack frame for each function call. The heap contains dynamically-allocated data for anything that’s required out of scope or whose size is only known during runtime.

The stack grows with each stack frame, but it has a static max size. On the other hand, the heap is allocated from a pool of memory with 🍰 Memory Allocation.

States §

A process can be in one of multiple states:

1. Running: process is currently executing instructions.
2. Ready: process is waiting to be scheduled by the kernel.
3. Blocked: process is waiting on some resource.
4. Stopped: process is stopped by the SIGSTOP signal and will only resume with SIGCONT.
5. Zombie: process is finished but still exists in the process table. (This is done so its parent can check its status even if it’s done.)
6. Terminated: after the parent checks a zombie process, the zombie is reaped and completely terminated.

Process Table §

To keep track of all processes that their states, the OS maintains a table of processes, and each process has a process control block (PCB) that contains process id, parent process id, child process ids, process group id, status, and other values.

Fork §

Processes are created from forking another process (via the fork() function). Whenever this happens, both parent and (newly created) child continue running the same code, but they run different branches depending on the fork() output (child_pid for the parent, 0 for the child). The child receives a cloned version of the parent’s memory, making it entirely separate.

Process Groups §

Processes are associated together in groups, which allows us to manage multiple processes at once. For example, 🗼 Signals can be sent to all processes in a group.