Processes in unix

Unix Process Management

31/01/2004

1/30/2004 Operating Systems 1

Unix Process Management
♦ Operating system functions executes within
user process.
♦ 2 modes of execution
– User mode and Kernel mode
♦ 2 types of processes are available
– System processes(Execute OS code)
– User processes( Execute user program code).
♦ System call is used to transfer from user
mode to system mode.


Process states in UNIX


♦ Preemption can only occur when the
process is about to move from Kernel mode
to User mode.
♦ While the process is running in Kernel
mode it may not be preempted. This make
UNIX not suitable for real time processing.
♦ Process 0 is a special process that is created
when the system boots. Process 1 (init
process) is the child of Process 0. All the
other processes in UNIX has Process 1 as
ancestor. All new processes are created
under Process 1 (init process).


Process Description
♦ Elements of process image -- divided into 3
– User level context
• Process text, Process data, User stack and Shared memory.
– Register context
• Program counter, Process status registers, stack pointer,
general purpose registers…
– System level context
• Process table entry, User area, Per process region table, Kernel
stack
– When a process is not running the processor status
information is stored in register context area.


Unix Process Table Entry
♦ Process status
♦ Pointers – to user area & process memory area
(text, data, stack)
♦ Process size – enables OS to know how much
space to allocate
♦ User identifier
– Real user ID - ID of user who is responsible for the
process
– Effective user ID used by process to gain temporary
privilege, while the program is being executed as a part
of process.
– Process identifiers ID of the process


♦ Event descriptor valid when a process is in
sleep state. When the event occurs, the process is
transferred to a ready to run state.
♦ Priority used for scheduling
♦ Signal enumerates signals send to process but
not yet handled.
♦ Timers Process execution time, kernel resource
utilization, and user set timer used to send alarm
signal to a process.
♦ P_link pointer to the next link in the ready
queue.
♦ Memory status indicates the process image is in
the main memory or swapped out.


UNIX User Area
♦ Process table pointer Indicates entry
corresponding to the user area.
♦ User identifiers Real & Effective user Ids, used
to determine user privileges.
♦ Timers Record time that the process spent
executing in user mode & kernel mode.
♦ Signal handling array For each type of signal
defined in the system, indicate how the process
will react to receipt of that signal.
♦ Control terminal Indicate login terminal for this
process, if exist.
♦ Error field Record errors encountered during
system call.

♦ Return value Contain result of s/m call
♦ I/O parameters Describes amount of data
transfer, the address of the source data array in
user space, file offset for I/O.
♦ File parameters Current directory & current
root describe the file system environment of the
process.
♦ User file descriptor table Record the file the
process has open
♦ Limit fields Restrict the size of the process &
size of the file it can write.
♦ Permission mode field Mask mode settings on
files the process creates.


System level context
♦ 2 parts
– 1. Static (Process table entry, User area, Per process
region table).
– 2. Dynamic (Kernel stack)
– Process table entry contains process control information
that is accessible to the kernel to the kernel at all time.
So in VM systems all process table entries are
maintained in main memory.
– User area contains additional process control
information that is needed by the kernel when it
executes in context of a process. It is also used when
swapping process to and from the memory.
– Process reg. table is used by memory management s/m
– Kernel stack – used when the process is executing in
kernel mode and contain information that must be
saved and restored as procedure calls and interrupts
1/30/2004 occurs. Operating Systems 11

#include <stdio.h>
Fork program in C
#include <unistd.h>
int main(int argc, char *argv[])
{
int pid;
/* fork another process */
pid = fork();
if (pid < 0) { /* error occurred */
fprintf(stderr, "Fork Failed");
exit(-1);
}
else if (pid == 0) { /* child process */
execlp("/bin/ls","ls",NULL);
}
else { /* parent process */
/* parent will wait for the child to complete */
wait(NULL);
printf("Child Complete");
exit(0);
}
}

Fork() in UNIX
♦ Allocate slot in the process table for the new
process
♦ Assigns a unique process id to the new process
♦ Make a copy of the process image of the parent,
with the exception of shared memory
♦ It increases counters for any files owned by the
parent, to reflect that an additional process now
also owns these files.
♦ It assigns the child process to a ready to run state
♦ It returns the ID number of the child to the parent
process and a 0 value to the child process.
♦ These all works are done in Kernel of parent
process.

♦ After completing those functions OS will do
the following operations as a part of
dispatcher routine
– Stay in the parent process. Control returns to
the user mode at the point of the fork call of the
parent.
– Transfer control to the child process. The child
process begins executing at the same point in
the code as the parent, namely at the return
from the fork call.
– Transfer control to another process. Both child
and parent are left in the ready to run state.


Types of processes in UNIX
♦ Mainly 3 types
– User process
– Daemon process
– Kernel process
– Daemon process
• Performs the function in a system wide basis. The function can
be of any auxiliary kind, but they are vital in controlling the
computational environment of the system.
• Example Print spooling, Network Management.
• Once created Daemon process can exist throughout the life
time of the Operating System.


Process Tree in a UNIX system


Termination of a process in UNIX
♦ Exit(status_code)
– Status_code indicate the termination status of the
process.
– Kernel does the following after receiving the exit call
• Close all open files of the process
• Releases the memory allotted to it
• Destroy the User area of the process
– It does not destroy the proc structure, this is retained till
the parent of Pi destroy it.
– The terminated process is dead but still exists. Hence it
is called ZOMBIE process.
– The exit call also sends signal to the parent of Pi which
may be ignored by the parent.


Wait statement in UNIX
– Pi can wait for the termination of a child
– Wait(address(xyz)); // xyz variable within the
address space of Pi.
– If process Pi has child processes & at least one of them
has already terminated, the wait call stores the
termination status of the terminated child process into
xyz and immediately return with the id of the
terminated child process. If more child processes exists
their termination status would be made available to Pi
only when it repeats the wait call.
– Process Pi is blocked if it has children but none of them
has terminated, it will be unblocked when one of the
child process terminates. The wait call returns a –1 if Pi
has no children.


Example for wait
Int main()
{
int I,saved_status;
for(I=0;I<3;I++)
{
if (fork()==0)
{
// Child Process
exit();
}
}
while(wait(&saved_status)!= -1); // Loop till all children
terminates
}


Interrupt handling in UNIX
♦ Will execute only one interrupt at a time ( to
avoid race condition).
♦ Each interrupt is assigned an interrupt
priority level. An interrupt priority level is
also assigned to CPU.
♦ When an interrupt at priority level ‘l’ arises,
it is handled only if l> CPU’s interrupt
priority else kept pending till CPU’s
interrupt priority level assumes a lower
value.


System calls in UNIX
– System calls accepts parameters relevant to its
functioning
– When a call occurs these parameters exists on the user
stack of the process which issues the system call.
– The call number is expected to be in register 0
– System call handler obtain this number to determine
which system functionality being invoked.
– From its internal table it knows the address of the
handler for that function.
– Parameters of the system call exists on the user stack of
the process making the call.
– Before passing control to the handler for a specific call
these parameters are copied from the user stack into
some standard places in the user area.


System calls in UNIX Cont…
– A signal can be send to a process/Group of
processes using kill(<pid>,<signum>) system
call. Where pid is the id of the process signal to
be sent.(sender must know receiver address &
receiver must be in the same process tree).
– Pid value 0 signal to be sent to all processes
in the same group as the sender process.
– Pid value –1 to reach processes outside the
process tree of the sender.


Signal handling in UNIX
♦ Oldfunction = signal(<signum>,<function>)
– Where signal is a function in C library which
makes a signal system call. Signum is an
integer & function is the function name.
– The function should be executed on occurrence
of signal <signum>.
– User can specify 0 or 1 instead of function.
Where 0 indicates the default action defined in
the Kernel is to be executed & 1 occurrence
of signal is to be ignored.


♦ Whenever a signal is send to a process, the bit
corresponding to the signal is set to 1 in the proc
structure of the destination process. The kernel
now determines if the signal is being ignored by
the destination process. If not it will makes
provision to deliver the signal to the process. If the
signal is ignored it remains pending and is
delivered when the process is ready to accept it. In
UNIX signal remains pending if the process to
which it is intended is in block state. The signal
would be delivered when the process comes out of
the blocked state. Signals are delivered when
♦ A process returns from a system call, after a
process get unblocked, before a process gets
blocked.

Interesting signals in UNIX
♦ SIGCHLD Child process died or suspended
♦ SIGFPE Arithmetic fault
♦ SIGILL Illegal instruction
♦ SIGINT tty interrupt (Control C)
♦ SIGKILL Kill process
♦ SIGSEGV Segmentation fault
♦ SIGSYS Invalid System call
♦ SIGXCPU Exceeds CPU limit
♦ SIGXFSZ Exceeds file size limit


Processes in unix

Recommended

More Related Content

What's hot (20)

Viewers also liked (8)

Similar to Processes in unix (20)

Processes in unix