Process: an instance of a program in execution

(User) Thread: an execution flow of the process

Pthread (POSIX thread) library
Lightweight process (LWP): used to offer better
support for multithreaded applications
LWP may share resources: address space, open files,
To associate a lightweight process with each thread
Examples of pthread libraries that use LWP: LinuxThreads,
IBMs Next Generation Posix Threading Package (NGPT)
www.QuontraSolutions.Com
task_struct data structure
state: process state
thread_info: low-level information for the process
mm: pointers to memory area descriptors
tty: tty associated with the process
fs: current directory
files: pointers to file descriptors
signal: signals received


www.QuontraSolutions.Com
www.QuontraSolutions.Com
TASK_RUNNING: executing
TASK_INTERRUPTABLE: suspended
(sleeping)
TASK_UNINTERRUPTABLE: (seldom
used)
TASK_STOPPED
TASK_TRACED
EXIT_ZOMBIE
EXIT_DEAD
www.QuontraSolutions.Com
Process descriptor pointers: 32-bit
Process ID (PID): 16-bit (~32767 for compatibility)
Linux associates different PID with each process or
LWP
Programmers expect threads in the same group to
have a common PID
Thread group: a collection of LWPs (kernel 2.4)
The PID of the first LWP in the group
tgid field in process descriptor: using getpid() system call
www.QuontraSolutions.Com
union thread_union {
struct thread_info thread_info;
unsigned long stack[2048];
};
Two data structures in 8KB (2 pages)
thread_info structure (new to 3rd ed.)
Kernel mode process stack

www.QuontraSolutions.Com
www.QuontraSolutions.Com
Obtain the address of thread_info structure from
the esp register
current_thread_info()
movl $0xffffe000, %ecx
andl %esp, %ecx
movl %ecx, p
The process descriptor pointer of the process
currently running on a CPU
The current macro: equivalent to
current_thread_info()->task
movl $0xffffe000, %ecx
andl %esp, %ecx
movl (%ecx), p


www.QuontraSolutions.Com
www.QuontraSolutions.Com
LIST_HEAD(list_name)
list_add(n,p)
list_add_tail(n,p)
list_del(p)
list_empty(p)
list_entry(p,t,m)
list_for_each(p,h)
list_for_each_entry(p,h,m)

www.QuontraSolutions.Com
tasks field in task_struct structure
type list_head
prev, next fields point to the previous and the next
task_struct
Process 0 (swapper): init_task
Useful macros:
SET_LINKS, REMOVE_LINKS: insert and remove a
process descriptor
#define for_each_process(p) \
for (p=&init_task; (p=list_entry((p)->tasks.next), \
struct task_struct, tasks)
\
) != &init_task; )

www.QuontraSolutions.Com
runqueue
run_list field in task_struct structure: type list_head
Linux 2.6 implements the runqueue differently
To achieve scheduler speedup, Linux 2.6 splits the
runqueue into 140 lists of each priority!
array filed of process descriptor: pointer to the
prio_array_t data structure
nr_active: # of process descriptors in the list
bitmap: priority bitmap
queue: the 140 list_heads
enqueue_task(p, array), dequeue_task(p, array)
www.QuontraSolutions.Com
Process 0 and 1: created by the kernel
Process 1 (init): the ancestor of all processes
Fields in process descriptor for parenthood
relationships
real_parent
parent
children
sibling

www.QuontraSolutions.Com
www.QuontraSolutions.Com
To search up the search for the process
descriptor of a PID
Sequential search in the process list is inefficient
The pid_hash array contains four hash tables
and corresponding filed in the process descriptor
pid: PIDTYPE_PID
tgid: PIDTYPE_TGID (thread group leader)
pgrp: PIDTYPE_PGID (group leader)
session: PIDTYPE_SID (session leader)
Chaining is used to handle PID collisions

www.QuontraSolutions.Com
Size of each pidhash table: dependent on the
available memory
PID is transformed into table index using
pid_hashfn macro
#define pid_hashfn(x) hash_long((unsigned long)x,
pidhash_shift)
unsigned long hash_long(unsigned long val, unsigned
int bits)
{
unsigned long hash = val * 0x9e370001UL;
return hash >> (32-bits);
}
www.QuontraSolutions.Com
www.QuontraSolutions.Com
pids field of the process descriptor: the pid
data structures
nr: PID number
pid_chain: links to the previous and the next
elements in the hash chain list
pid_list: head of the per-PID list (in thread group)
www.QuontraSolutions.Com
www.QuontraSolutions.Com
do_each_trask_pid(nr, type, task)
while_each_trask_pid(nr, type, task)
find_trask_by_pid_type(type, nr)
find_trask_by_pid(nr)
attach_pid(task, type, nr)
detach_pid(task, type)
next_thread(task)
www.QuontraSolutions.Com
Processes in TASK_STOPPED,
EXIT_ZOMBIE, EXIT_DEAD: not linked in
lists
Processes in TASK_INTERRUPTABLE,
TASK_UNINTERRUPTABLE: wait queues
Two kinds of sleeping processes
Exclusive process
Nonexclusive process: always woken up by the
kernel when the event occurs
www.QuontraSolutions.Com
struct _ _wait_queue_head {
spinlock_t lock;
struct list_head task_list;
};
typedef struct _ _wait_queue_head wait_queue_head_t;
struct _ _wait_queue {
unsigned int flags;
struct task_struct * task;
wait_queue_func_t func;
struct list_head task_list;
};
typedef struct _ _wait_queue wait_queue_t;
www.QuontraSolutions.Com
Wait queue handling functions:
add_wait_queue()
add_wait_queue_exclusive()
remove_wait_queue()
wait_queue_active()
DECLARE_WAIT_QUEUE_HEAD(name)
init_waitqueue_head()
To wait:
sleep_on()
interruptible_sleep_on()
sleep_on_timeout(), interruptible_sleep_on_timeout()
Prepare_to_wait(), prepare_to_wait_exclusive(), finish_wait()
Macros: wait_event, wait_event_interruptible
www.QuontraSolutions.Com
To be woken up:
Wake_up, wake_up_nr, wake_up_all,
wake_up_sync, wake_up_sync_nr,
wake_up_interruptible, wake_up_interruptible_nr,
wake_up_interruptible_all,
wake_up_interruptible_sync,
wake_up_interruptible_sync_nr
www.QuontraSolutions.Com
RLIMIT_AS
RLIMIT_CORE
RLIMIT_CPU
RLIMIT_DATA
RLIMIT_FSIZE
RLIMIT_LOCKS
RLIMIT_MEMLOCK
RLIMIT_MSGQUEUE
RLIMIT_NOFILE
RLIMIT_NPROC
RLIMIT_RSS
RLIMIT_SIGPENDING
RLIMIT_STACK
www.QuontraSolutions.Com
Process switch, task switch, context switch
Hardware context switch: a far jmp (in older Linux)
Software context switch: a sequence of mov
instructions
It allows better control over the validity of data being loaded
The amount of time required is about the same
Performing the Process Switch
Switching the Page Global Directory
Switching the Kernel Mode stack and the hardware
context

www.QuontraSolutions.Com
TSS: a specific segment type in x86
architecture to store hardware contexts

www.QuontraSolutions.Com
In traditional UNIX, resources owned by parent
process are duplicated
Very slow and inefficient
Mechanisms to solve this problem
Copy on Write: parent and child read the same
physical pages
Lightweight process: parent and child share per-
process kernel data structures
vfork() system call: parent and child share the memory
address space

www.QuontraSolutions.Com
clone(fn, arg, flags, child_stack, tls, ptid,
ctid): creating lightweight process
A wrapper function in C library
Uses clone() system call
fork() and vfork() system calls:
implemented by clone() with different
parameters
Each invokes do_fork() function

www.QuontraSolutions.Com
Kernel threads run only in kernel mode
They use only linear addresses greater
than PAGE_OFFSET
www.QuontraSolutions.Com
kernel_thread(): to create a kernel thread
Example kernel threads
Process 0 (swapper process), the ancestor of all
processes
Process 1 (init process)
Others: keventd, kapm, kswapd, kflushd (also
bdflush), kupdated, ksoftirqd,


www.QuontraSolutions.Com
exit() library function
Two system calls in Linux 2.6
_exit() system call
Handled by do_exit() function
exit_group() system call
By do_group_exit() function
Process removal
Releasing the process descriptor of a zombie
process by release_task()
www.QuontraSolutions.Com
Thanks for Your
Attention!
www.QuontraSolutions.Com