Process Concept – I – Introduction to Processes

In this post, we first get to know what a process is. This will be followed by a discussion on process states. We will get to know about Process Control block, Process Scheduling using Scheduling queues. Discussion on types of schedulers (Short term, medium term and long term). Finally, we understand what is a context switch.

INTRODUCTION

  • Process : Program in execution
  • unit of work in a modern time-sharing system
  • The terms job [batch process executes jobs] and process used interchangeably

PROCESS

A process is a program in execution. However, it is more than the program code. A process includes the following :
  • Text Section : The Program code
  • Program Counter : represents current activity
  • Process Stack : contains temporary data (e.g. function parameters, return addresses, local variables)
  • Data section : Contains global variables
  • Heap section : Memory dynamically allocated during the process runtime
Process in memory
Process in memory
Program Process
1 Passive entity : consists the text section [program code] only Active entity : Consists of a program counter specifying the next instruction to be executed.
2 Example : File containing  a list of instructions stored on a disk [executable file] Example : Executable file loaded into the memory [Through command line or GUI – double clicking the icon]

NOTE

  1. A Program by itself is not a process. A program becomes a process when an executable file is loaded into the memory.
  2. Although two processes may be associated with the same program, they are considered two separate execution sequences.
    1. Example : User may invoke multiple copies of a web browser program
    2. Each copy is then considered a separate process.
    3. Although they share the text sections (i.e. the program code), the data section, the heap and stack sections may vary.
  3. A process can spawn multiple processes as it runs. Spawning means loading and executing a new child process.
  4. A process can be an execution environment for other code.
    1. Example : JVM (Java virtual machine) is used to execute Java code
    2. The JVM executes as a process that interprets the loaded Java code and takes action via native machine instructions on behalf of that code.

PROCESS STATE

  • The state of a process is used to describe the current activity of that process.
  • As a process executes, it changes state.
  • A process may be in one of the following states.
    • New : Process is being created.
    • Running : Instructions are being executed.
    • Waiting : Process is waiting for some event (e.g. an I/O completion, reception of a signal) to occur.
    • Ready : The process is waiting to be assigned to a processor
    • Terminated : The process has finished execution.
  • Only one process can be running on a processor at any instant of time. However, many processes can be in ready or waiting states. In case of multiple processes, the rest have to wait until the CPU is free and can be rescheduled.
Process States
Process States

PROCESS CONTROL BLOCK

Process Control Block (PCB) or Task Control Block contains many pieces of information associated with a specific process which includes :
  • Process State : can be new/ready/waiting/…
  • Program Counter : Contains address of next instruction to be executed.
  • CPU Registers : include accumulators, index registers, stack pointers, general-purpose registers, condition-code information
  • CPU Scheduling information : include Process priority, pointers to scheduling queues, other scheduling parameters.
  • Memory-management information : includes values of base and limit registers, page tables, and/or segment tables
  • Accounting information : includes process numbers, amount of CPU time and real time used, time limits, account numbers, etc
  • I/O status information : includes list of I/O devices allocated to the process, list of open files, etc
    PCB → Repository of Process Information
Process Control Block
Process Control Block
Linux Representation of PCB
  • In Linux, the PCB is represented as C structure task_struct [found in <linux/sched.h> include file in the kernel source code directory. Refer this for more details.
  • This structure stores all the necessary information representing  a process which includes
    • pointer to process’s parent
    • list of children and siblings
    • state of process
    • scheduling info
    • memory management info
  • Within linux kernel, all active processes represented using a doubly-linked list of task_struct.
  • The kernel maintains a pointer called current to point to the process currently executing in the system.

PROCESS SCHEDULING

Why need Process Scheduling?
  • To meet the objective of Multi-programming (maximize CPU utilization by having some process running at all times)
  • To meet the objective of Time Sharing (switch CPU among processes so frequently that users can interact with each program while it is running)
  • Meeting above objectives via Process Scheduler
    • Process Scheduler selects an available process for program execution (possibly from a list of several processes)

I . PROCESS SCHEDULING USING SCHEDULING QUEUES

The system consists of many queues. Depending on the process state, they are put into appropriate queues.
  • Job Queue : As processes enter the system, the are put into job queue, which consists of all the processes in the system.
  • Ready Queue : Processes residing in Main memory, which are ready and waiting to execute are kept on this list.
    • Implemented as a linked list
    • header contains pointer to first and last PCB in the list.
    • Each PCB includes pointer that points to next PCB in the ready queue.
  • Device Queue : Processes waiting for a particular I/O device are put here. Each device has its own device queue.
The representation of Process Scheduling is done via Queueing Diagrams
  • Contains 2 types of queues : Ready queue and Device queues
  • Rectangles : Each rectangular box represents a queue.
  • Circles : Circles represent resources that serve the queues.
  • Arrows : Arrows represent indicate the flow of processes in the system.
Scheduling Process
  1. New process → Put in Ready queue
  2. Process waits until selected for execution or dispatched.
  3. If selected for execution, CPU is allocated to the process.
    1. Process may issue I/O request → Put in I/O queue.
    2. Process could create new child → Wait for child to finish execution
    3. Process could be interrupted → Put back in ready queue.
  4. Continue this cycle until termination (point at which Process is removed from all queues and has its PCB and resources de-allocated.)
 
Queueing-diagram representation of process scheduling
Queueing-diagram representation of process scheduling

II. SCHEDULERS

A process waits in a scheduling queue until it is selected by the OS in some fashion. This task of selecting the processes is carried out by appropriate Scheduler. Schedulers are of 3 types:
  • Long term Scheduler (or job scheduler) [LTS]
    • Used typically in batch systems, where more jobs are submitted than can be executed immediately.
    • Processes are spooled to a mass storage device, typically a disk, where they are kept for later execution.
    • LTS selects processes from this pool → Loads them into memory for execution.
  • Short term Scheduler (or CPU Scheduler)
    • Used very frequently
    • selects one process from the processes that are ready to execute, and allocates the CPU to that process.
  • Medium Term Scheduler
    • Introduced in some OSs as an intermediate level of Scheduling.
    • Reason : to reduce the degree of multiprogramming
      • MTS carries out Swapping of processes.
    • Swapping :
      • First, a process is removed from memory (and from active contention for the CPU) to reduce the degree of multiprogramming.
      • Later, the process can be re-introduced into the memory, and its execution can be continued where it left-off.
      • Advantages : 
        • Improve Process mix
        • Memory constraints [change in memory requirements has over-committed memory, requiring memory to be freed up
 
Addition of medium-term scheduling to the queuing diagram
Addition of medium-term scheduling to the queuing diagram
A B C
1 Attribute Short-term Scheduler Long-term Scheduler
2 Frequency of execution Selects new process for CPU frequently. executes much-less frequently
3 Time gap between Processes Often, STS executes at least once every 100ms minutes may separate the creation of new processes and the next
4 Speed of execution Because of short time between executions, STS has to be fast. Because of longer interval between executions, LTS can afford to take more time to decide which process should be selected for execution.
  Note :
  • Long-term Scheduler controls degree of multi-programming (number of processes in memory). 
  • If degree of multiprogramming is stable → average rate of process creation == average departure rate of processes leaving the system.
  • I/O bound process : One that spends more of its time doing I/O than spending time on computations
  • CPU bound Process : Process that generates I/O requests infrequently, using more of its time doing computations.
  • It is important that the LTS selects a good process mix of I/O-bound processes and CPU-bound processes.
    • If all processes I/O bound → ready queue empty → short-term scheduler sits idle
    • If all processes CPU bound → I/O waiting queue empty → Devices go unused
    • Good combination of CPU-bound and I/O bound processes → Best Performance of system

CONTEXT SWITCH

  • Context switch : Task of switching CPU to another process [which requires performing a state save of current process and state restore of a different process.]
  • Why Context Switch?
    • When an interrupt occurs, the system needs to save the current context of the process running on the CPU and switch to some other process [kernel routine]
    • This is done to restore the context when processing is done, essentially suspending the process, and then resuming it.
    • Context :
      • saved in PCB of a process
      • includes value of CPU registers,process state,memory-management information
  • Context switch done by the Kernel
  • Context switch time is pure overhead → no useful work done during this time
    • time highly dependent on hardware support
    • e.g. , in some processors, which provide multiple sets of registers, Context Switch → changing pointer to current register set
    • If more complexity involved → more work needs to be done
  • Switching speed : depends upon
    • memory speed
    • number of registers to be copied
    • typically takes few milliseconds
CPU switch from process to process (switch between processes)
CPU switch from process to process (switch between processes)
In the next post, we will continue our discussions on processes further with discussion on operations on processes.

In response to the comments, we differentiate between STS, MTS and LTS in the following manner:

  • Short term scheduling (or dispatching):
    • Determines which of the ready processes can have CPU resources, and for how long. [Running]
  • Long term scheduling:
    • Determines which programs are admitted to the system for execution [New]
    • when, and which processes should be exited. [Terminated]
  • Medium term scheduling:
    • Determines when processes are to be suspended [Waiting] and resumed [Running]

Taking into account the states of a process, we can conclude that

  • Short term Scheduling (dispatching) affects processes in the following states
    • running
    • ready
  • the medium term scheduling affects processes in the following states
    • ready-suspended
    • blocked-suspended
  • the long term scheduling affects processes in the following states
    • new
    • terminated
Sources: 12