A process in operating system informally is a program in execution. A program (usually represented by the executable file) becomes a process only when it is load in main memory.
The process memory layout is represented by the following figure:
- text section: stores the code
- data section: stores the global variables
- stack section: temporary storage when invoking functions
- heap section: dynamically allocated memory
In Linux specifically, each process is represented by a task_struct in a double-linked list which serves as the process table.
References
- A. Silberschatz, P. B. Galvin, e G. Gagne, «Operating System Concepts».
See also
- Tail Call optimization replaces call with jmp — tail call optimization works directly on the stack section: TCO prevents new stack frames from being pushed, reducing stack memory from to
- Linear Recursion is a chain of deferred operations — recursion builds a chain of deferred operations on the stack section; understanding the process memory layout explains why deep recursion causes stack overflow
- SIMD - one operation applied to multiple data points simulaneously — SIMD operates at a lower level than the process: multiple processing elements working on data, bypassing the single-process execution model
- System Programming - Compile utilities — practical tooling for working with processes: ThreadSanitizer and UBSan operate on running processes to detect race conditions and undefined behaviour