PGIDs

PGIDs stands for process group identifiers. In Unix-like operating systems, every process has a process ID (PID) and belongs to a process group identified by a PGID. A process group is a collection of one or more related processes that can receive signals as a single unit. By convention, the PGID is usually the PID of the group leader, which is the first process in the group to form the group.

Process groups are organized within sessions. A session contains one or more process groups and has a session ID (SID), typically equal to the PID of the session leader. The combination of sessions, process groups, and individual processes enables fine-grained job control, particularly in interactive shells, where the foreground and background status of jobs can be managed collectively.

Key operations and concepts:

- A process can create or join a group using setpgid; if a process calls setpgid(0, pid), it can form a new group or join an existing one.

- The current process’s group can be queried with getpgrp(), getpgid(0), or similar calls.

- A terminal can designate a foreground process group with tcsetpgrp, enabling keyboard input and job control signals to target that group.

- Signals can be sent to an entire group by using a negative PGID in system calls like kill, affecting all members of that group.

- Walls and shells commonly use PGIDs to manage pipelines, foregrounding, stopping (Ctrl-Z), and resuming jobs.

PGIDs are a core part of Unix process management and are primarily relevant to system programming, shells, and other software that coordinates multiple processes. Cross-platform differences exist, as non-POSIX environments implement their own equivalents.