Microprocessor Based

Systems
Lecture No 06 and 07 General
Program Execution Concept

By Nasir Mahmood
 Overview of Last Lecture
  Assembly Language & Machine Language
  How the translation is done
  Compiler, Assembler, OS
  Multistage Pipeline Concept
  Concept of Stages and cycles execution
  Make the execution fast by parallel pipelines

  Superscalar Processor
 Today’s Topics
  Reading from Memory
  Operating Systems
  Multitasking
  New Trends
  Real Time Embedded Systems
 Reading From Memory
  Program throughput is often dependent on the speed of
memory access.

  CPU clock speed might be several giga hertz but the

access to memory might be running at a sluggish 33MHz.
The CPU has to wait one or more clock cycles until
operands are fetched from memory before instruction can
execute.

  Since CPUs are much faster than conventional memory,

computers use CACHE memory to hold the most recent
instructions and data.
4
Memory Read Cycle

5
Operating System level
 Operating System
  An OS is a program that controls the execution of
application programs and acts as an interface b/w
the user and the computer hardware.
  Like other programs, it provides instructions for
the processor. The key difference is the intent of
the program.
  OS directs the µp in the use of system resources
and control timing of its execution of other
programs.
7
 Operating System
  But in order for the up to do anything, it must
cease executing the OS and execute other
programs.
  Therefore, the OS relinquishes control for the
up to do some ‘useful’ work and then resumes
control long enough to prepare the processor to
do the next piece of work.

8
 Multitasking
  A Multitasking OS is capable of running multiple tasks at
the same time
  Most modern OS simultaneously execute tasks that
communicate with hardware, display user interfaces,
perform file processing etc.
  A CPU can really execute only one instruction at a time
  A component of the OS – the scheduler- allocates a slice
of the CPU time to each task. During each time slice the
CPU executes a block of instructions, stopping when the
time slice has ended.

9
 Multitasking (contd.)

  By rapidly switching tasks the CPU creates the

illusion that they are running simultaneously.
  One type of scheduling is called Round Robin
Scheduling

10
 Scheduling
  The key to
multiprogram
ming or
multitasking is
scheduling.

11
Round Robin Scheduling

12
 Process?
  But first lets define a ‘Process’
  A program in execution OR
  The entity to which the processor is assigned

  Process states
  New: A new process is admitted but not executed. OS will
initialize the process, making it ready
  Ready: process is ready & awaiting access to µp
  Running: Process is being executed by µp

  Waiting: Process is suspended from execution waiting for

some sys resource
  Halted: Process has terminated & will be destroyed 13
 Multitasking
Process Control Block

  For each process in system, the OS must

maintain information regarding the state of the
process and other info necessary for process
execution
  Each process is represented in the OS by a
process control block

14
 Process Control Block

  contains:
  Identifier
  State

  Priority

  Program counter

  Memory Pointers

  Context Data

  I/O status Information

  Accounting Info
15
 Multitasking (contd.)
  A Multitasking OS runs on a CPU that
supports task switching, the processor saves
the state of each task before switching to a new
one
  A task’s state consists of the contents of the
processor registers, program counter, and status
flags, along with references to the task’s memory
segments.
  Preemptive Multitasking assigns priorities to
tasks and a higher priority task can interrupt a
low priority task.
16
 Memory Management
  Effective memory management is vital in a
multiprogramming system.
  If only a few processes are in memory, then for
much of the time all processes will be waiting
for I/O and processor will be idle.
  Thus, memory needs to be allocated efficiently
to pack as many processes into memory as
possible.

17
 Memory Management
(Swapping)
  What to do?
  Main memory could be expanded to
accommodate more processes but
  Main memory is expensive
  The appetite of programs for memory grows as fast
as the memory. Thus, larger memory results in
larger processes, not more processes.

18
 Swapping
  Swapping
  Processes stored on disk are brought in one by one
as space becomes available.
  Swapping however, is an I/O operation. Therefore,
there is the potential of making the problem worse.
However, swapping will generally enhance
performance.
  A more, sophisticated scheme, involving virtual
memory improves performance over swapping.

19
 Paging
  Each process is broken down into smaller
chunks known as pages
  Each page of the program could be assigned
available chunks of memory known as frames
  Smaller processes require fewer pages, larger
processes require more.
  When a process is brought in, its pages are
loaded into available frames and a page table is
set up by the OS.
20
 Paging
  The frames do
not have to be
contiguous as the
page table shows
the frame location
for each page of
the process.

21
 Virtual Memory & Demand
Paging
  As a process is executed only in main memory,
that memory is referred to real memory. But
the user sees a much larger memory – which is
allocated on disk.
  This is referred to as the Virtual Memory
which is usually 3 times that of main memory
  Virtual Memory allows for very effective
multiprogramming or multitasking and relieves
the user of the constraints of main memory.
22
Virtual Memory (Windows Task Mgr)

23
 Virtual Memory & Demand
Paging
  Demand Paging simply means that each page of a
process is brought in only when it is needed i.e. on
demand.
  Therefore, it is not necessary to load an entire process
into main memory. Instead, with demand paging the
OS breaks the program into pieces that are loaded one
at a time.
  When the process needs code which is not in
memory, a page fault is triggered which causes the
OS to bring in the desired page.

24
 How Programs Run (Putting it
together)
  Load & Execute Process
  OS searches for the file in the current directory and
then in the Pre-Defined directories called Paths
  If found the OS retrieves the basic info about the
file like the size and physical location
  OS determines the next available location in
memory and loads the program file into memory
  OS executes a branching instruction that begins
execution of the Program’s first machine
instruction, As soon as it starts running it is called
a Process
25
 How Programs Run (contd.)

  OS keeps track of the Process and responds to

requests for Sys Resources
  When the process ends, its handle is removed
and the memory it used is released so it can be
used by other programs.

26
 What is Real-time?
  A system is classified as real-time if it is
required to complete the work on a timely
basis
  It is of utmost importance that the correctness
of the timely behavior of real-time system
should be validated
  Often real-time systems are also safety-critical, a
missing of a deadline an lead to disastrous
consequences
  Hard and Soft deadlines
 Summary
  Reading from Memory
  Bottleneck, Cache Memory
  Operating Systems
  It manages all the process running and do the
memory management
  Multitasking
  Tasks switching
  New Trends
  Real Time Embedded Systems
THE END