Process Scheduling 1

Process Scheduling

The Operating System Kernel

• Basic set of primitive operations and processes
– Primitive
∗ Like a function call or macro expansion
∗ Part of the calling process
∗ Critical section for the process
– Process
∗ Synchronous execution with respect to the calling process
∗ Can block itself or continuously poll for work
∗ More complicated than primitives and more time and space
• Provides a way to provide protected system services, like supervisor call instruction
– Protects the OS and key OS data structures (like process control blocks) from interference by user programs
– The fact that a process is executing in kernel mode is indicated by a bit in program status word (PSW)
• Execution of kernel
– Nonprocess kernel
∗ Kernel executes outside of any user process
∗ Common on older operating systems
∗ Kernel takes over upon interrupt or system call
∗ Runs in its own memory, and has its own system stack
∗ Concept of process applies only to user programs and not to OS
– Execution with user processes
∗ OS executes in the context of user process
∗ OS is considered to be a collection of functions called by user processes to provide a service
∗ Each user process must have memory for program, data, and stack areas for kernel routines
∗ A separate kernel stack is used to manage code execution in kernel mode
∗ OS code and data are in shared address space and are shared by all processes
∗ Interrupt, trap, or system call execute in user address space but in kernel mode
∗ Termination of kernel’s job allows the process to run with just a mode switch back to user mode
– Process-based kernel
∗ Kernel is implemented as a collection of system processes, or microkernels
∗ Modular OS design with a clean interface between different system processes
• Set of kernel operations
– Process Management: Process creation, destruction, and interprocess communication; scheduling and dispatching;
process switching; management of process control blocks
– Resource Management: Memory (allocation of address space; swapping; page and segment management), sec-
ondary storage, I / O devices, and files
– Input/Output: Transfer of data between memory and I / O devices; buffer management; allocation of I / O channels
and devices to processes
– Interrupt handling: Process termination, I / O completion, service requests, software errors, hardware malfunction
• Kernel in Unix
Process Scheduling 2

– Controls the execution of processes by allowing their creation, termination, suspension, and communication
– Schedules processes fairly for execution on CPU
∗ CPU executes a process
∗ Kernel suspends process when its time quantum elapses
∗ Kernel schedules another process to execute
∗ Kernel later reschedules the suspended process
– Allocates main memory for an executing process
∗ Swapping system: Writes entire process to the swap device
∗ Paging system: Writes pages of memory to the swap device
– Allocates secondary memory for efficient storage and retrieval of user data
– Allows controlled peripheral device access to processes
• Highest Level of User Processes: The shell in Unix
– Created for each user (login request)
– Initiates, monitors, and controls the progress for user
– Maintains global accounting and resource data structures for the user
– Keeps static information about user, like identification, time requirements, I / O requirements, priority, type of pro-
cesses, resource needs
– May create child processes (progenies)
• Process image
– Collection of programs, data, stack, and attributes that form the process
– User data
∗ Modifiable part of the user space
∗ Program data, user stack area, and modifiable code
– User program
∗ Executable code
– System stack
∗ Used to store parameters and calling addresses for procedure and system calls
– Process control block
∗ Data needed by the OS to control the process
– Location and attributes of the process
∗ Memory management aspects: contiguous or fragmented allocation

Data Structures for Processes and Resources

• Used by the OS to keep track of each process and resource
• Cross-referenced or linked in main memory for proper coordination
• Memory tables
– Used to keep track of allocated and requested main and secondary memory
– Protection attributes of blocks of main and secondary memory
– Information to map main memory to secondary memory
• I / O tables
Process Scheduling 3

– Used to manage I / O devices and channels

– State of I / O operation and location in main memory as source/destination of operation

• File tables

– Information on file existence, location in secondary memory, current status, and other attributes
– Part of file management system

• Process control block

– Most important data structure in an OS

– Set of all process control blocks describes the state of the OS
– Read and modified by almost every subsystem in the OS, including scheduler, resource allocator, and performance
monitor
– Constructed at process creation time
∗ Physical manifestation of the process
∗ Set of data locations for local and global variables and any defined constants
– Contains specific information associated with a specific process
∗ The information can be broadly classified as process identification, processor state information, and process
control information
∗ Can be described by Figure 1

Process Id
Identification → Parent PID
User Id / Group Id
User registers
Program counter
CPU state → Condition codes
Status info (interrupt enable flags)
Stack pointer
Process state
Priority
Process scheduling → Wait-time
Event being waited on
Privileges
Flags
IPC → Signals
Messages
Unit Owned/
Memory → → ···
Description Shared
Resource Unit Owned/
Resources → → ···
Class Description Shared
Created resources → ↓ → ···
to resource
descriptor
Children

Figure 1: Process Control Block

∗ Identification. Provided by a pointer to the PCB

· Always a unique integer in Unix, providing an index into the primary process table
· Used by all subsystems in the OS to determine information about a process
Process Scheduling 4

· Used for cross-referencing in other tables (memory, I / O)

∗ CPU state
· Provides snapshot of the process
· The program counter register is initialized to the entry point of the program (main() in C programs)
· While the process is running, the user registers contain a certain value that is to be saved if the process is
interrupted
· Typically described by the registers that form the processor status word (PSW) or state vector
· Exemplified by EFLAGS register on Pentium that is used by any OS running on Pentium, including Unix
and Windows NT
∗ Process State. Current activity of the process
· Running. Executing instructions.
· Ready. Waiting to be assigned to a processor; this is the state assigned when the data structure is con-
structed
· Blocked. Waiting for some event to occur.
∗ Allocated address space (Memory map)
∗ Resources associated with the process (open files, I/O devices)
∗ Other Information.
· Progenies/offsprings of the process
· Priority of the process
· Accounting information. Amount of CPU and real time used, time limits, account numbers, etc.
· I/O status information. Outstanding I/O requests, I/O devices allocated, list of open files, etc.
• Resource Descriptors.
– Resource.
∗ Reusable, relatively stable, and often scarce commodity
∗ Successively requested, used, and released by processes
∗ Hardware (physical) or software (logical) components
– Resource class.
∗ Inventory. Number and identification of available units
∗ Waiting list. Blocked processes with unsatisfied requests
∗ Allocator. Selection criterion for honoring the requests
– Contains specific information associated with a certain resource
∗ pˆ.status_data points to the waiting list associated with the resource.
∗ Dynamic and static resource descriptors, with static descriptors being more common
∗ Identification, type, and origin
· Resource class identified by a pointer to its descriptor
· Descriptor pointer maintained by the creator in the process control block
· External resource name ⇒ unique resource id
· Serially reusable or consumable?
∗ Inventory List
· Associated with each resource class
· Shows the availability of resources of that class
· Description of the units in the class
∗ Waiting Process List
· List of processes blocked on the resource class
· Details and size of the resource request
· Details of resources allocated to the process
Process Scheduling 5

∗ Allocator
· Matches available resources to the requests from blocked processes
· Uses the inventory list and the waiting process list
∗ Additional Information
· Measurement of resource demand and allocation
· Total current allocation and availability of elements of the resource class

• Context of a process

– Information to be saved that may be altered when an interrupt is serviced by the interrupt handler
– Includes items such as program counter, registers, and stack pointer

Basic Operations on Processes and Resources

• Implemented by kernel primitives

• Maintain the state of the operating system

• Indivisible primitives protected by “busy-wait” type of locks

• Process Control Primitives

– create. Establish a new process

∗ Assign a new unique process identifier (PID) to the new process
∗ Allocate memory to the process for all elements of process image, including private user address space and
stack; the values can possibly come from the parent process; set up any linkages, and then, allocate space for
process control block
∗ Create a new process control block corresponding to the above PID and add it to the process table; initialize
different values in there such as parent PID, list of children (initialized to null), program counter (set to
program entry point), system stack pointer (set to define the process stack boundaries)
∗ Initial CPU state, typically initialized to Ready or Ready, suspend
∗ Add the process id of new process to the list of children of the creating (parent) process
∗ r0 . Initial allocation of resources
∗ k0 . Initial priority of the process
∗ Accounting information and limits
∗ Add the process to the ready list
∗ Initial allocation of memory and resources must be a subset of parent’s and be assigned as shared
∗ Initial priority of the process can be greater than the parent’s
– suspend. Change process state to suspended
∗ A process may suspend only its descendants
∗ May include cascaded suspension
∗ Stop the process if the process is in running state and save the state of the processor in the process control
block
∗ If process is already in blocked state, then leave it blocked, else change its state to ready state
∗ If need be, call the scheduler to schedule the processor to some other process
– activate. Change process state to active
∗ Change one of the descendant processes to ready state
∗ Add the process to the ready list
– destroy. Remove one or more processes
∗ Cascaded destruction
Process Scheduling 6

∗ Only descendant processes may be destroyed

∗ If the process to be “killed” is running, stop its execution
∗ Free all the resources currently allocated to the process
∗ Remove the process control block associated with the killed process
– change priority. Set a new priority for the process
∗ Change the priority in the process control block
∗ Move the process to a different queue to reflect the new priority
• Resource Primitives
– create resource class. Create the descriptor for a new resource class
∗ Dynamically establish the descriptor for a new resource class
∗ Initialize and define inventory and waiting lists
∗ Criterion for allocation of the resources
∗ Specification for insertion and removal of resources
– destroy resource class. Destroy the descriptor for a resource class
∗ Dynamically remove the descriptor for an existing resource class
∗ Resource class can only be destroyed by its creator or an ancestor of the creator
∗ If any processes are waiting for the resource, their state is changed to ready
– request. Request some units of a resource class
∗ Includes the details of request – number of resources, absolute minimum required, urgency of request
∗ Request details and calling process-id are added to the waiting queue
∗ Allocation details are returned to the calling process
∗ If the request cannot be immediately satisfied, the process is blocked
∗ Allocator gives the resources to waiting processes and modifies the allocation details for the process and its
inventory
∗ Allocator also modifies the resource ownership in the process control block of the process
– release. Release some units of a resource class
∗ Return unwanted and serially reusable resources to the resource inventory
∗ Inform the allocator about the return

Organization of Process Schedulers

• Objective of Multitasking: Maximize CPU utilization and increase throughput
• Two processes P0 and P1

P0 t0 i0 t1 i1 ··· in−1 tn

P1 t00 i00 t01 i01 ··· i0m−1 t0m

• Two processes P0 and P1 without multitasking

P0 t0 i0 t1 i1 · · · in−1 tn P0 terminated
P1 P1 waiting t00 i00 t01 i01 · · · i0m−1 t0m

• Processes P0 and P1 with multitasking

P0 t0 t1 ··· tn
P1 t00 t01 ··· t0m
Process Scheduling 7

• Each entering process goes into job queue. Processes in job queue
– reside on mass storage
– await allocation of main memory
• Processes residing in main memory and awaiting CPU time are kept in ready queue
• Processes waiting for allocation of a certain I / O device reside in device queue
• Scheduler
– Concerned with deciding a policy about which process to be dispatched
– After selection, loads the process state or dispatches
– Process selection based on a scheduling algorithm
• Autonomous vs shared scheduling
– Shared scheduling
∗ Scheduler is invoked by a function call as a side effect of a kernel operation
∗ Kernel and scheduler are potentially contained in the address space of all processes and execute as a part of the
process
– Autonomous scheduling
∗ Scheduler (and possibly kernel) are centralized
∗ Scheduler is considered a separate process running autonomously
∗ Continuously polls the system for work, or can be driven by wakeup signals
∗ Preferable in multiprocessor systems as master/slave configuration
· One CPU can be permanently dedicated to scheduling, kernel and other supervisory activities such as I / O
and program loading
· OS is clearly separated from user processes
∗ Who dispatches the scheduler?
· Solved by transferring control to the scheduler whenever a process is blocked or is awakened
· Scheduler treated to be at a higher level than any other process
– Unix scheduler
∗ Autonomous scheduler
∗ Runs between any two other processes
∗ Serves as a dummy process that runs when no other process is ready
• Short-term v/s Long-term schedulers
– Long-term scheduler
∗ Selects processes from job queue
∗ Loads the selected processes into memory for execution
∗ Updates the ready queue
∗ Controls the degree of multiprogramming (the number of processes in the main memory)
∗ Not executed as frequently as the short-term scheduler
∗ Should generate a good mix of CPU-bound and I / O-bound processes
∗ May not be present in some systems (like time sharing systems)
– Short-term scheduler
∗ Selects processes from ready queue
∗ Allocates CPU to the selected process
∗ Dispatches the process
Process Scheduling 8

∗ Executed frequently (every few milliseconds, like 10 msec)

∗ Must make a decision quickly ⇒ must be extremely fast

Process or CPU Scheduling

• Major task of any operating system – allocate ready processes to available processors
– Scheduler decides the process to run first by using a scheduling algorithm
– Two components of a scheduler
1. Process scheduling
∗ Decision making policies to determine the order in which active processes compete for the use of CPU
2. Process dispatch
∗ Actual binding of selected process to the CPU
∗ Involves removing the process from ready queue, change its status, and load the processor state
• Desirable features of a scheduling algorithm
– Fairness: Make sure each process gets its fair share of the CPU; also, no process should be starved for CPU
– Efficiency: Keep the CPU busy 100% of the time
– Response time: Minimize response time for interactive users
– Turnaround: Minimize the time batch users must wait for output
– Throughput: Maximize the number of jobs processed per hour
• Types of scheduling
– Preemptive
∗ Temporarily suspend the logically runnable processes
∗ More expensive in terms of CPU time (to save the processor state)
∗ Can be caused by
Interrupt. Not dependent on the execution of current instruction but a reaction to an external asynchronous
event
Trap. Happens as a result of execution of the current instruction; used for handling error or exceptional con-
dition
Supervisor call. Explicit request to perform some function by the kernel
– Nonpreemptive
∗ Run a process to completion
• The Universal Scheduler: specified in terms of the following concepts
1. Decision Mode
– Select the process to be assigned to the CPU
2. Priority function
– Applied to all processes in the ready queue to determine the current priority
3. Arbitration rule
– Applied to select a process in case two processes are found with the same current priority

The Decision Mode

– Time (decision epoch) to select a process for execution
– Preemptive and nonpreemptive decision
– Selection of a process occurs
Process Scheduling 9

1. when a new process arrives

2. when an existing process terminates
3. when a waiting process changes state to ready
4. when a running process changes state to waiting (I/O request)
5. when a running process changes state to ready (interrupt)
6. every q seconds (quantum-oriented)
7. when priority of a ready process exceeds the priority of a running process
– Selective preemption: Uses a bit pair (up , vp )
up set if p may preempt another process
vp set if p may be preempted by another process
The Priority Function
– Defines the priority of a ready process using some parameters associated with the process
– Memory requirements – Important due to swapping overhead
Smaller memory size ⇒ Less swapping overhead
Smaller memory size ⇒ More processes can be serviced
– Attained service time
Total time when the process is in the running state
– Real time in system
Total actual time the process spends in the system since its arrival
– Total service time
Total CPU time consumed by the process during its lifetime
Equals attained service time when the process terminates
Higher priority for shorter processes
Preferential treatment of shorter processes reduces the average time a process spends in the system
– External priorities
Differentiate between classes of user and system processes
Interactive processes ⇒ Higher priority
Batch processes ⇒ Lower priority
Accounting for the resource utilization
– Timeliness – Dependent upon the urgency of a task and deadlines
– System load
Maintain good response time during heavy load
Reduce swapping overhead by larger quanta of time
The Arbitration Rule
– Random choice
– Round robin (cyclic ordering)
– Chronological ordering (FIFO)

• Time-Based Scheduling Algorithms

– May be independent of required service time
• First-in/First-out (FIFO) Scheduling
– Also called First-Come-First-Served (FCFS), or strict queuing
– Simplest CPU-scheduling algorithm
– Nonpreemptive decision mode
– Upon process creation, link its PCB to rear of the FIFO queue
– Scheduler allocates the CPU to the process at the front of the FIFO queue
– Average waiting time can be long
Process Scheduling 10

Process Burst time

P1 24
P2 3
P3 3
– Let the processes arrive in the following order:
P1 , P 2 , P 3
Then, the average waiting time is calculated from:
P1 P2 P3
1 24 25 27 28 30
0+24+27
– Average waiting time = 3 = 17 units
– Priority function P (r) = r
– Tends to favor CPU-bound processes compared to I / O-bound processes, because of its non-preemptive nature
∗ Against cycle stealing
∗ If a CPU-bound process is in wait state when the I / O-bound process initiates I / O, CPU goes idle
· Inefficient use of both CPU and I / O devices
• Last-in/First-out (LIFO) Scheduling
– Similar to FIFO scheduling
– Average waiting time is calculated from:
P3 P2 P1
1 3 4 6 7 30
0+3+6
– Average waiting time = 3 = 3 units
∗ Substantial saving but what if the order of arrival is reversed.
– Priority function P (r) = 1/r
• Shortest Job Next (SJN) Scheduling
– Also called Shortest Job First (SJF) scheduling
– Associate the length of the next CPU burst with each process
– Assign the process with shortest CPU burst requirement to the CPU
– Nonpreemptive scheduling
– Specially suitable to batch processing (long term scheduling)
– Ties broken by FIFO scheduling
– Consider the following set of processes
Process Burst time
P1 6
P2 8
P3 7
P4 3
Scheduling is done as:
P4 P1 P3 P2
1 3 4 9 10 16 17 24
3+16+9+0
Average waiting time = 4 = 7 units
0+6+14+21
– Using FIFO scheduling, the average waiting time is given by 4 = 10.25 units
– Priority function P (t) = 1/t
Process Scheduling 11

– Provably optimal scheduling – Least average waiting time

∗ Moving a short job before a long one decreases the waiting time for short job more than it increase the waiting
time for the longer process
– Problem: To determine the length of the CPU burst for the jobs

• Shortest Remaining Time First (SRTF) Scheduling

– Preemptive version of shortest job next scheduling

– Preemptive in nature (only at arrival time)
– Highest priority to process that need least time to complete
– Priority function P (τ ) = a − τ , where a is the arrival time
– Consider the following processes
Process Arrival time Burst time
P1 0 8
P2 1 4
P3 2 9
P4 3 5
– Schedule for execution
P1 P2 P4 P1 P3
1 2 5 6 10 11 17 18 26
– Average waiting time calculations

• Round-Robin Scheduling

– Preemptive in nature, based on fixed time slices or time quanta q

∗ Reduces the penalty that short jobs suffer with FIFO
– Generate clock interrupts at periodic intervals, or time slices or time quanta
∗ Time quantum between 10 and 100 milliseconds
∗ Each process gets a slice of time before being preempted
– All user processes treated to be at the same priority
– Ready queue treated as a circular queue, effectively giving a variation on FIFO
∗ New processes added to the rear of the ready queue
∗ Preempted processes added to the rear of the ready queue
∗ Scheduler picks up a process from the head of the queue and dispatches it with a timer interrupt set after the
time quantum
– CPU burst < q ⇒ process releases CPU voluntarily
– Timer interrupt results in context switch and the process is put at the rear of the ready queue
– No process is allocated CPU for more than 1 quantum in a row
– Consider the following processes
Process Burst time
P1 24
P2 3
P3 3
q = 4 ms
P1 P2 P3 P1 P1 P1 P1 P1
1 4 5 7 8 10 11 14 15 18 19 22 23 26 27 30
Process Scheduling 12

6+4+7
Average waiting time = 3 = 5.66 milliseconds
– Let there be
n processes in the ready queue, and q be the time quantum, then
each process gets n1 of CPU time in chunks of at most q time units
Hence, each process must wait no longer than (n − 1) × q time units for its next quantum
– Performance depends heavily on the size of time quantum
∗ Large time quantum ⇒ FIFO scheduling
∗ Small time quantum ⇒ Large context switching overhead
∗ Rule of thumb: 80% of the CPU bursts should be shorter than the time quantum

• Multilevel Feedback Queue Scheduling

– Most general CPU scheduling algorithm

– Background
∗ Make a distinction between foreground (interactive) and background (batch) processes
∗ Different response time requirements for the two types of processes and hence, different scheduling needs
∗ Separate queue for different types of processes, with the process priority being defined by the queue
– Separate processes with different CPU burst requirements
– Too much CPU time ⇒ lower priority
– I / O-bound and interactive process ⇒ higher priority
– n different priority levels – Π1 · · · Πn
– Each process may not receive more than TΠ time units at priority level Π
– Let TΠ = 2n−Π Tn , where Tn is the maximum time on the highest priority queue at level n
∗ Each process will spend time Tn at priority n, and time 2i Tn at priority n − i, 1 ≤ i ≤ n
∗ When a process receives time TΠ , decrease its priority to Π − 1
∗ Process may remain at the lowest priority level for infinite time
– CPU always serves the highest priority queue that has processes in it ready for execution
– Aging to prevent starvation

• Policy Driven CPU Scheduling

– Based on a policy function

– Policy function gives the correlation between actual and desired utilization for services
– Services can be measured in terms of time units during which a particular resource is used by the process
– Attempt to strike a balance between actual and desired resource utilization

• Real-time scheduling

– Requires a small scheduling latency for the kernel

– BeOS has a scheduling latency of 250 ms, making it suitable for many real-time applications, and improves respon-
siveness for non-real-time applications
– Unix processes cannot be preempted when they are running in kernel mode; this makes Unix unsuitable for real-time
processing