Step‑by‑Step: How a Program Is Loaded and Executed by the OS

A program is simply a file that stores a sequence of machine instructions; for example, chrome.exe is a regular file containing millions of instructions that the CPU can execute.

Running the program—by double‑clicking its icon or invoking it from a command line—first requires the operating system to locate the file on disk using the file system.

Once found, the OS parses the executable format (Linux typically uses the ELF format) to identify the locations of code, data, and other sections.

The OS then loads the program by copying its instructions and global variables from the file into RAM, allocating a contiguous free memory region for the new process.

Within that region the OS creates the familiar memory areas: the code (text) segment and data segment come directly from the executable, while the heap and stack are runtime‑allocated abstractions that grow in opposite directions.

If the program depends on shared libraries, the OS loads those libraries into the free space between the heap and stack.

The compiler records the address of the first instruction (the entry point) in the ELF header. Tools like readelf show this as the "Entry point address" field.

When the OS schedules the new process, it initializes the CPU’s instruction pointer with this entry‑point address, allowing the CPU to begin execution at the correct location.

Thus, the seemingly simple action of double‑clicking an icon triggers a chain of OS operations that locate, load, and start the program in memory.