What is the difference between interruptible signals and those that cannot be interrupted in Linux system?

In Linux, signals can be either interruptible or non-interruptible in the context of how they affect processes and system calls.

Interruptible signals are signals that can interrupt a process or system call to handle the signal. For example, if a process is waiting (e.g., in a read system call), and a signal arrives, the system call can be interrupted and return an error EINTR. This allows the process to handle the signal (e.g., terminate or perform other logic).

Non-interruptible signals are signals that do not interrupt system calls. In such cases, the system call will continue to execute despite the arrival of a signal, and the process will handle the signal after the system call completes.

Thus, the main difference is whether the signal affects the interruption of the current process work (e.g., a system call) or not. This is important for correct signal handling and process behavior management in Linux.